scieee Science in your language
[en] (orig)
Metal-organic nanowires:
Microfluidic-based synthesis, optical and
electrical characterization and label-free
sensing applications
vorgelegt von
Master of Science (M. Sc.)
Yanlong Xing
aus Zibo, China
Von der Fakultät II – Mathematik und Naturwissenschaften
der Technischen Universität Berlin
zur Erlangung des akademischen Grades
doctor rerum naturalium (Dr. rer. nat.)
genehmigte Dissertation
Promotionsausschuss:
Vorsitzende: Prof. Dr. Birgit Kanngießer (Technische Universität Berlin)
Gutachter: Prof. Dr. Norbert Esser (Technische Universität Berlin)
Gutachter: Prof. Dr. Peter Hildebrandt (Technische Universität Berlin)
Gutachterin: Prof. Dr. Petra. S. Dittrich (Eidgenössische Technische Hochschule
Zürich)
Tag der wissenschaftlichen Aussprache: 21. Oktober 2016
Berlin 2016
For my family…
V
List of publications
Journal Articles
1. Yanlong Xing, Andreas Wyss, Norbert Esser and Petra S. Dittrich*, Label-
free biosensors based on in situ formed and functionalized microwires in
microfluidic devices, Analyst, 2015, 140, 7896-7901.
2. Yanlong Xing, Norbert Esser, Petra S. Dittrich*, Conductive single
nanowires formed and analysed on microfluidic devices. Journal of
Material Chemistry C, 2016, 4, 9235-9244.
3. Yanlong Xing, Guoguang Sun, Eugen Speiser, Norbert Esser and Petra S.
Dittrich*, Localized synthesis of conductive Cu-TCNQ nanostructures in
ultra-small microchambers for nanoelectronics. Ready for submission. Oct.
2016.
4. Yanlong Xing, Eugen Speiser, Petra S. Dittrich and Norbert Esser*,
Polarized confocal Raman microscopy of TTF crystal and single Au-TTF
nanowires. Oct. 2016, Manuscript prepared.
Conference Proceedings
1. Yanlong Xing, Eugen Speiser, Dheeraj Kumar Singh, Petra S. Dittrich,
Norbert Esser, the 2016 Raman Fest Conference, May 2016, Berlin,
Germany (Poster, 3rd prize of Best Poster Award).
2. Yanlong Xing, Eugen Speiser, Princia Salvatore, Petra S. Dittrich, Norbert
Esser, the 8th congress on X-ray analytics for industrial processes (PRORA
2015), November 2015, Berlin, Germany (Poster).
3. Mario Lenz, Yanlong Xing, Petra. S. Dittrich, 4th International Symposium
on Sensor Science, July 2015, Basel, Switzerland (Poster).
4. Yanlong Xing, Andreas Wyss, Norbert Esser, Petra. S. Dittrich, 9.
Deutsches BioSensor Symposium, March 2015, Munich, Germany (Poster).
5. Yanlong Xing, Norbert Esser, Petra S. Dittrich, EMBL Conference,
Microfluidics 2014 , July 2014, Heidelberg, Germany (Poster).
VI
VII
Abstract
Metal-organic nanowires exhibit not only the properties of one-dimensional
structures including ultra-small scale, large surface-to-volume ratio etc., but also
obvious advantages in their tunable properties and label-free sensing ability by
optical or electrical readout. Thus, the evaluation of metal-organic materials by
the use of transition metal ions and organic ligands including tetrathiafulvalene
(TTF) and tetracyanoquinodimethane (TCNQ) were focused in this thesis.
Concerning the synthesis of metal-organic nanowires, microfluidics offers
various benefits, e.g. laminar flow, reduced sample/reagent consumption and
control of self-assembly of nanostructures. Therefore, microfluidic techniques
have been mainly applied to the synthesis and application of nano-
/microstructures.
In the first part of this thesis, label-free biosensors based on in situ formed and
functionalized gold-tetrathiafulvalene (Au-TTF) wires were developed using an
integrated microfluidic system. Au-TTF microwires were formed and
immobilized inside the microchip. Then, different surface modification
protocols were applied to modify Au-TTF wires which were used for sensitive
label-free detection of catecholamines and human IgG by Raman spectroscopy.
Following, a study of molecular self-organization in individual Au-TTF nano-
/microwire by polarized confocal Raman spectroscopy was performed to
understand the growth mechanism of Au-TTF. Single nanowires were analysed
using non-destructive polarized Raman spectroscopy. Angular polarization
Raman measurement of a single TTF crystal and single nanowire showed the
periodic variations in typical Raman bands, indicating preferential ordering of
molecules in both crystal and Au-TTF wire. Based on the density functional
theory (DFT) calculation and simulation of depolarization ratio, the molecular
assembly in a single TTF crystal was confirmed. The tilted stacking of TTF units
in single Au-TTF nanowire along the long axis was also proved.
Afterwards, the formation of fibres and particles made of metal salts and TTF
derivatives on a microfluidic device and in a conventional reaction flask was
investigated. Their morphologies, optical properties and electrical
conductivities were characterized. This study provides a comprehensive
overview of the morphologies of the products obtained from reactions between
metals and different commercially available TTF derivatives.
VIII
Finally, a microfluidic-assisted synthesis of copper-tetracyanoquinodimethane
(Cu-TCNQ) nanostructures based on TCNQ was performed. A two-layer
microfluidic device comprising parallel actuated microchambers was used for
the synthesis, and enabled the excellent fluid handling for the continuous and
multiple chemical reactions in confined ultra-small chambers. The as-prepared
Cu-TCNQ wire bundles showed good conductivity and hysteresis reversing
memory effect, which proved the possibility in using them to build advanced
nanoelectronics.
IX
Zusammenfassung
Metall-organische Nanodrähte besitzen nicht nur die Eigenschaften
eindimensionaler Materialien wie z.B. großes Oberflächen-Volumen-Verhältnis,
Teilchengrößen im Nanometerbereich usw., sondern haben auch bedeutende
Vorteile durch ihre durchstimmbaren Eigenschaften und den Einsatz als
markierungsfreie Sensoren für optische oder elektronische Messungen. Deshalb
legt die vorliegende Arbeit den Fokus auf die Charakterisierung von metall-
organischen Materialien, die aus Übergangsmetallionen und organischen
Liganden wie Tetrathiafulvalen (TTF) und Tetracyanochinodimethan (TCNQ)
synthetisiert werden. Für die Synthese von metall-organischen Nanodrähten
eignen sich Mikrofluidik-Techniken besonders gut, da sie laminare Strömungen
erzeugen, in denen die Zufuhr von Substanzen sehr genau kontrolliert warden
kann.
Der erste Teil dieser Dissertation enthält die Entwicklung von
markierungsfreien Biosensoren basierend auf in situ produzierten und
funktionalisierten Gold-Tetrathiafulvalen (Au-TTF)-Drähten mittels eines
integrierten mikrofluidischen Systems. Zunächst wurden Au-TTF-Mikrodrähte
in einem Mikrochip gebildet und immobilisiert. Daraufhin wurden verschiedene
Modifikationsprotokolle angewandt um die Au-TTF-Drähte für den sensitiven
und markierungsfreien Nachweis von Katecholaminen und menschlichem IgG
mit Raman-Spektroskopie vorzubereiten.
Im Anschluss an diesen Teil folgt eine Studie der molekularen
Selbstorganisation in individuellen Au-TTF-Nano-/Mikrodrähten mit
zerstörungsfreier, konfokaler Polarisations-Raman-Spektroskopie um den
Wachstumsmechanismus der Drähte genauer zu verstehen. Polarisations-
Raman-Spektren eines einzelnen TTF-Kristalls und eines einzelnen TTF-
Nanodrahts zeigten die üblichen periodischen Abhängigkeiten der
polarisationsabhängigen Raman-Moden und ließen so auf eine bevorzugte
räumliche Anordnung der Moleküle sowohl im Kristall als auch in den Drähten
schließen. Basierend auf DFT-Berechnungen und der Simulation des
Depolarisierungsverhältnisses konnte die molekulare Anordnung in einem
einzelnen TTF-Kristall bestätigt werden. Die Stapelung von gekippten TTF-
Einheiten in einzelnen Au-TTF-Nanodrähten entlang der langen Achse konnte
ebenfalls bewiesen werden.
X
Im dritten Teil der vorliegenden Arbeit wurde die Bildung von Fasern und
Teilchen aus Metallsalzen und TTF-Derivaten in einem mikrofluidischen
System und einem konventionellen Reagenzglas untersucht. Morphologien,
optische Eigenschaften und elektrische Leitfähigkeiten wurden charakterisiert.
Die Ergebnisse geben eine umfassende Übersicht über die Beschaffenheit der
Reaktionsprodukte zwischen Metallsalzen und verschiedenen kommerziell
erwerblichen TTF-Derivaten.
Schlussendlich wurde die Synthese von Kupfer-Tetracyanochinodimethan (Cu-
TCNQ)-Nanostrukturen durchgeführt. Als Basis diente eine mikrofluidische
Doppelzelle bestehend aus zwei parallel betriebenen Mikrokammern, wodurch
eine exzellente Steuerung der Flüssigkeiten für kontinuierliche und multiple
chemische Reaktionen innerhalb der Kammern möglich war. Die hergestellten
Cu-TCNQ-Bündel zeigten gute Leitfähigkeitswerte und ein Hystereseverhalten
und bestätigten somit die Verwendungsmöglichkeit als innovative
nanoelektronische Komponenten.
XI
List of Figures
1.1. Chemical structures of TTF, its cation radical TTF+ and dication TTF2+.
1.2. Chemical structures of TCNQ and its anion radical.
1.3. Stacking of charge-transfer salts.
1.4. Crystal structure of TTF-TCNQ 2:1 charge-transfer compound (ac face).
1.5. The formation of Au-TTF nanowires by charge-transfer.
1.6. The synthesis of Ag-TCNQ complex.
2.1. Microfluidic synthesis based on continuous laminar flow.
2.2. Microfluidic device with valves.
2.3. Microfluidic-assisted application of conductive nanostructures.
2.4. Images of chrome masks and wafers used in this thesis.
2.5. A photograph of a glass slide with microelectrodes.
2.6. The formation of PDMS by catalytic cross-linking reaction.
2.7. Photographs of two microchips used in this project.
3.1. Simplified Jablonski diagrams of light scattering.
3.2. Raman spectrum of TCNQ under laser excitation of 532 nm.
3.3. Components of a polarized confocal Raman microscope with
backscattering geometry.
4.1. Images of TTF crystal and Au-TTF nano-/microwires.
4.2. Photograph of microchips and optical image of microchannels.
4.3. Schematic illustration of Cu-TCNQ wire formation in microchamber.
4.4. Optical images of a microchamber showing the operation of control layer.
4.5. Photographs of conductivity measurement devices.
5.1. Graphic abstract showing the sensing mechanism of the in situ label-free
sensor in this work.
5.2. The open-donut microchip and its features.
5.3. Images of Au-TTF wire formation.
5.4. Characterization of Au-TTF wires.
5.5. A second microchip design for wire formation.
5.6. Fluorescent images of wires with surface modification for biosensors.
5.7. Control experiment without GA solution.
5.8. Raman spectra of various bioamines on functionalized Au-TTF wires.
5.9. Quantitative Raman measurement of DA
based on the functionalized
wire for biosensor
.
XII
5.10.
Data analysis method of Raman spectrum.
5.11.
Fluorescent images of wires after surface modification as immunosensor.
5.12. Raman measurement results of functionalized wire for immunosensor.
5.13. Graphic abstract of the polarization Raman measurement of TTF crystal
and Au-TTF wire.
5.14. DFT modelling of TTF and TTF cation.
5.15. The view from three faces of crystalline TTF and the calculated angle
between two TTF molecules in a unit cell of TTF crystal.
5.16. Sketch of TTF crystal and a Raman spectrum.
5.17. Angular variations of Raman spectra from different faces of TTF crystal
under parallel and crossed polarization configurations.
5.18. Simulated depolarization ratio of TTF crystal.
5.19. Raman spectra of single Au-TTF wire under parallel (black curves) and
crossed (grey curve) polarization configurations.
5.20. Angular variation of Raman spectra of single Au-TTF nanowire under
parallel and crossed polarization configurations.
5.21. Simulated depolarization ratio of six Au-TTF nanowires.
5.22. The 3D model of Au-TTF wire showing rotation of molecules.
5.23. Graphic abstract of the synthesis and analysis of metal-TTF (derivatives)
in this project.
5.24. Chemical structures of TTF derivatives.
5.25. Synthesis of wires, exemplary for (TTF)7(CuCl2)3 (2a).
5.26. SEM images of (TTF)2CuCl2 (2b), (TTF)4Cu(NO3)2 (3a) and
(TTF)4Cu(NO3)2 (3b).
5.27. SEM images of (TTF)5FeCl3 (4a), (TTF)5FeCl3 (4b), (TTF)5Fe(NO3)3
(5a) and (TTF)5Fe(NO3)3 (5b).
5.28. Optical images of products derived from compounds 6, 7, 8, 9, 10, 11, 12,
13, 14.
5.29. EDX-SEM spectra of (TTF)2CuCl2 (2b), (TTF)4Cu(NO3)2 (3a) and
(TTF)4Cu(NO3)2 (3b).
5.30. EDX-SEM spectra of (TTF)5FeCl3 (4a), (TTF)5FeCl3 (4b),
(TTF)5Fe(NO3)3 (5a) and (TTF)5Fe(NO3)3 (5b).
5.31. UV-Vis absorption spectra of solutions of 2a, 2b, 3a, 3b, 4a, 4b, 5a, 5b,
6, 7, 8, 9, 10, 11, 12, 13, 14.
5.32. IR spectra of compounds (TTF)7(CuCl2)3 (2a) and (TTF)4Cu(NO3)2 (3a).
5.33. Raman spectra of compounds 2a, 2b, 3b, 4a, 4b, 5a and 5b at 532 nm
excitation.
XIII
5.34. Representative I-V curves of single nanowires of M-TTF complexes.
5.35. Linear range of the I-V curves in Figure 5.34 with the electrical
conductivities of single nanowires of 2a, 2b, 3a, 3b, 4a, 4b, 5a and 5b.
5.36. UV-Vis absorption spectra of the solutions of M-FTTF.
5.37. Optical and SEM images of M-FTTF generated by bulk synthesis and on
microchip.
5.38. Optical images of structures from 17, 18, 19, 20, 21 and 22 in bulk and
on microchips.
5.39. IR spectra of M-FTTF complexes.
5.40. EDX-SEM spectra of M-TTF complexes.
5.41. I-V curves of M-FTTF nanowires.
5.42. Optical images of compounds 28, 29 and 30.
5.43. UV-Vis absorption spectra of solutions of M-BEDT-TTF.
5.44. Optical images of compounds 23, 24, 25, 26 and 27 formed in bulk
synthesis.
5.45. SEM images and I-V curves of structures of 23, 25 and 27.
5.46. SEM images and I-V curves of structures of 24 and 26.
5.47. Linear I-V curves and calculated electrical conductivity of single wires of
23, 25 and 27.
5.48. EDX-SEM spectra of M-BEDT-TTF.
5.49. Sensing of organic gases by single Au-TTF wire.
5.50. Graphic abstract of the synthesis, characterization and analysis of Cu-
TCNQ.
5.51. Microchip device for the synthesis of Cu-TCNQ structures.
5.52. Photograph of Cu0 and Cu-TCNQ.
5.53. The Cu0 layer and Cu-TCNQ formed by using different concentrations of
Cu2+ and reductant.
5.54. SEM images and EDX spectra of Cu-TCNQ nanostructures.
5.55. UV-Vis absorption spectra of TCNQ (black) and Cu-TCNQ (red)
solutions in CH3CN.
5.56. FT-IR and Raman analysis of TCNQ and Cu-TCNQ.
5.57. EDX-SEM spectra of a) glass slide and b) Cu-TCNQ microsctructures.
5.58. FT-IR and Raman Characterization of Cu-TCNQ microsctructures.
5.59. I-V characteristics of Cu0 and Cu-TCNQ at room-temperature.
7.1. Structures of catecholamines and aromatic amino acids.
XIV
List of Schemes
2.1. Schematic of the microfabrication steps of a SU-8 master mold.
2.2. Schematic of the microfabrication steps of an AZ master mold.
5.1. Functionalization on Au-TTF wire for the sensing of amines (RNH2).
5.2. Functionalization of Au-TTF wire for immunoassay.
5.3. Reaction mechanisms for binding of fluorophores to the SAM of 4-ATPh.
5.4. Reaction mechanism for the formation of Cu0 and Cu-TCNQ.
XV
List of Tables
2.1. An overview of the microdevices used in different Chapters of this thesis.
5.1. Raman tensors for the symmetry classes of the C2v and D2h point groups.
5.2. Calculated Raman tensor values.
5.3.
Raman assignments of TTF, TTFn+ (0<n<1) and TTF+.
5.4. An overview of the experimental parameters, morphology, conductiv-
-ity and characterization of metal-TTF (and derivatives) complexes.
5.5. Elemental analysis of compounds 2a, 2b, 3a, 3b, 4a, 4b, 5a and 5b.
5.6. IR assignments of TTF0 and TTF+.
5.7. IR assignments of other M-TTF compounds 2b, 3b, 4a, 4b, 5a and 5b.
5.8. Elemental analysis of 15 and 16 complexes.
5.9. Morphology of structures formed under different reaction conditions on
microchip.
XVI
List of Equations
1.1. Electrical resistive switching properties of M-TCNQ.
2.1. Reynolds number in microfluidic system.
3.1. Energy equation in IR absorption.
3.2. Definition of induced molecular dipole moment ind.
3.3. Definition of nuclear displacement Qk.
3.4. Definition of molecular polarizability α.
3.5. General expression of induced dipole moment.
3.6. The Raman scattering intensity I.
3.7. Definition of depolarization ratio P.
5.1. Definition of depolarization ratio with a constant factor D.
5.2. The reversible switching states of Cu-TCNQ upon voltage sweeping.
XVII
Abbreviations and Acronyms
A
Acceptor
Ag-TCNQ
Silver-tetracyanoquinodimethane
4-ATPh
4-aminothiophenol
Au-TTF
Gold-tetrathiafulvalene
BEDT-TTF
Bis(ethylenedithio)tetrathiafulvalene
CAD
Computer-aided design
CEA
Cysteamine
CH3CN
Acetonitrile
C6H5O7Na3
Sodium citrate dehydrate
CoCl2·6H2O
Cobalt(II) chloride hexahydrate
Co(NO3)2
Cobalt(II) nitrate
CuCl
Copper(I) chloride
CuCl2
Copper(II) chloride
Cu(NO3)2
Copper(II) nitrate
CuSO4
Copper(II) sulfate
Cu-TCNQ
Copper-tetracyanoquinodimethane
D
Donor
1D
One dimensional
DA
Dopamine
DFT
Density functional theory
DI
Deionized
DMSO
Dimethyl sulfoxide
EDC
1-Ethyl-3-(3-dimethylaminopropyl)carbodiimide
hydrochloride
EDX
Energy dispersive X-ray
XVIII
EPI
Epinephrine
FET
Field-effect transistor
FeCl3
Iron(III) chloride
FeCl2·4H2O
Iron(II) chloride tetrahydrate
Fe(NO3)3
Iron(III) nitrate
FTTF
Formyl-tetrathiafulvalene
GA
Glutaldehyde
HWP
Half-wave plate
HAuCl4
Hydrotetrachloroaurate
IgG
Immunoglobulin G
IR
Infrared
ISP
Isoprenaline hydrochloride
I-V
Current-voltage
K-TCNQ
Potassium-tetracyanoquinodimethane
LP
Linear polarizer
MnCl2
Manganese(II) chloride
μTAS
Micro total analysis systems
NA
Numerical aperture
NaBH3CN
Cyanoborohydride
NaBH4
Sodium borohydride
NaOH
Sodium hydroxide
Na-TCNQ
Sodium-tetracyanoquinodimethane
Nd:YAG
Neodymium yttrium aluminium garnet
NE
Norepinephrine hydrochloride
NHS
N-Hydroxysuccinimide
Ni(NO3)2
Nickel(II) nitrate hexahydrate
nm
nanometre
PBS
Phosphate buffered saline
XIX
PDMS
Polydimethylsiloxane
Phe
Phenylalanine
PA
Polarizer analyser
Pt
Platinum
SAM
Self-assembled monolayer
SEM
Scanning electron microscopy
SERS
Surface-enhanced Raman scattering
SWNT
Single-walled carbon nanotube
TCE-TTF
2,3,6,7-Tetrakis(2-cyanoethylthio)tetrathiafulvalene
TCNQ
Tetracyanoquinodimethane
TET-TTF
Tetrakis(ethylthio)tetrathiafulvalene
THF
Tetrahydrofuran
TTF
Tetrathiafulvalene
TTF-TCNQ
Tetrathiafulvalene-tetracyanoquinodimethane
Tyr
Tyrosine
μm
Micrometre
μM
Micromolar
VOC
Volatile organic compounds
ZnCl2
Zinc chloride
Zn(NO3)2
Zinc nitrate hexahydrate
XX
XXI
Contents
1 Introduction .................................................................................................. 1
1.1 Electron donor and acceptor molecules .................................................. 1
1.2 Metal-organic charge-transfer complexes ............................................... 4
1.3 One-dimensional (1D) nanostructures .................................................... 7
1.4 Scope of the thesis .................................................................................. 9
2 Methods-1: Microfluidics and microfabrication ..................................... 13
2.1 Introduction to microfluidics ................................................................ 13
2.2 Microfluidic-guided synthesis of nano-/microwires ............................. 14
2.3 Microfluidic-assisted application of nano-/microwires ......................... 18
2.4 Microfabrication ................................................................................... 20
2.4.1 Wafers and masters ........................................................................ 20
2.4.2 Soft lithography ............................................................................. 23
2.4.3 Microchip fabrication .................................................................... 24
3 Methods-2: Raman microspectroscopy .................................................... 27
3.1 Raman scattering ................................................................................... 27
3.1.1 Stokes, anti-Stokes and resonance Raman scattering .................... 27
3.1.2 Raman spectroscopy and fingerprint ............................................. 29
3.1.3 Raman and Infrared spectroscopy ................................................. 31
3.2 Theoretical basis of Raman scattering .................................................. 31
3.2.1 Normal modes ............................................................................... 31
3.2.2 Polarizability and Raman selection rules ....................................... 32
3.2.3 Raman tensor ................................................................................. 33
3.3 Light polarization and depolarization ratio ........................................... 33
3.4 Confocal and polarized Raman microscopy: principles and setups ...... 34
3.4.1 Confocal Raman microspectroscopy ............................................. 34
XXII
3.4.2 Polarized Raman microspectroscopy ............................................. 36
3.4.5 In situ Raman measurements ......................................................... 37
3.5 Instrumentation for Raman spectroscopy.............................................. 37
4 Methods-3: Sample preparation and characterization ........................... 39
4.1 Microchip operation .............................................................................. 39
4.2 In situ synthesis, immobilization and functionalization of Au-TTF
microwires .................................................................................................. 39
4.2.1 In situ synthesis and immobilization of Au-TTF microwires ........ 39
4.2.2 Functionalization protocol for biosensing applications ................. 40
4.2.3 Functionalization protocol for immunosensing applications ......... 41
4.3 Synthesis of TTF crystal and Au-TTF nano-/microwires ..................... 42
4.4 Microfluidic synthesis of metal-TTF structures .................................... 43
4.5 In situ formation of Cu-TCNQ nano-/microstructures .......................... 45
4.6 Sample characterization ........................................................................ 47
4.6.1 Electrical characterization ............................................................. 47
4.6.2 Other characterization techniques .................................................. 49
4.7 Methods for data analysis ..................................................................... 50
4.7.1 Data analysis of fluorescent images, Raman, IR spectra and I-V
curves ..................................................................................................... 50
4.7.2 DFT calculations ........................................................................... 50
4.7.3 Single TTF crystal analysis ........................................................... 51
5 Results and discussion ................................................................................ 53
5.1 Label-free biosensors based on in situ formed and functionalized
microwires in microfluidic devices ............................................................. 53
5.1.1 Microchip and synthesis of Au-TTF microwires ........................... 55
5.1.2 Label-free biosensing of catecholamines ....................................... 59
5.1.3 Label-free immunosensors for detection of human IgG ................ 65
5.2 Study of molecular self-organization in TTF crystals and individual Au-
TTF nano-/microwires by polarized confocal Raman spectroscopy ........... 71
XXIII
5.2.1 DFT calculations of Raman tensor ................................................ 72
5.2.2 Polarized confocal Raman spectra of single TTF crystals ............. 74
5.2.3 Polarized confocal Raman spectra of single Au-TTF nanowires .. 81
5.3 Conductive single nanowires formed and analysed on microfluidic
devices.... .................................................................................................... 87
5.3.1 Microfluidic guided synthesis of charge-transfer complexes ........ 88
5.3.2 Metal-TTF complexes ................................................................... 93
5.3.3 Metal-FTTF complexes ............................................................... 108
5.3.4 TCE-TTF, TET-TTF and BEDT-TTF ......................................... 113
5.3.5 Metal-BEDT-TTF nano-/microstructures .................................... 114
5.3.6 Sensing of organic gases by single TTF-based wires .................. 120
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-
small microchambers for nanoelectronics ................................................. 123
5.4.1 Microchips with a microchamber array ....................................... 125
5.4.2 Reaction mechanism .................................................................... 126
5.4.3 Controlled synthesis of Cu and Cu-TCNQ structures .................. 126
5.4.4 Characterization of Cu-TCNQ nanostructures ............................ 130
5.4.5 Characterization of Cu-TCNQ microstructures ........................... 134
5.4.6 Electrical properties of Cu-TCNQ nano-/microstructures ........... 135
6 Conclusions and outlook .......................................................................... 139
Bibliography ................................................................................................ 143
Appendix ...................................................................................................... 159
Acknowledgement ....................................................................................... 161
Statement of authorship.............................................................................. 163
XXIV
1.1 Electron donor and acceptor molecules
1
1 Introduction
1.1 Electron donor and acceptor molecules
Electron donors refer to those molecules rich in electrons, which can act as a
reducing agent and transfer electrons to electron acceptors in reactions. It is
reported that the most promising candidates for good organic electron donors
are multisulphur heterocycles, such as TTF. Since its discovery in 1970,1 this
well-known organic electron donor has received intense focus on synthesizing
conductive compounds via charge-transfer. The reaction between TTF and the
strong electron acceptor, TCNQ,2 results in the formation of TTF-TCNQ
(tetrathiafulvalene-tetracyanoquinodimethane) nanowires which show metallic
conductive behaviour.3,4 This explains ongoing research interest in developing
organic conductive materials, as well as studying the mechanism of organic
electronics based on TTF.
TTF consists of two five-membered rings connected by a carbon-carbon double
bond. It is an excellent electron donor because each 1,3-dithiole ring has seven
π-electrons (one from each carbon and two from each sulphur), so that the
monocationic state has six π-electrons.5 In early studies,6 it was considered that
TTF had a planar structure. However, in modern studies it is widely accepted
that the neutral TTF shows boat-like conformation while charged TTF has a
planar molecular structure, on the basis of theoretical calculation results.7,8 Upon
charge-transfer, TTF can form the cations TTF+ and TTF2+ (Figure 1.1).
Figure 1.1. Chemical structures of TTF, its cation radical TTF+ and dication TTF2+.9
In a charge-transfer reaction, when the electron acceptor (e.g. metal ion) is in
excess, both TTF+ and TTF2+ can be formed.10 Thus, a mixed-valence of TTF
units (TTF+ and TTF2+) exists in the final charge-transfer compounds. The
compounds with TTF2+ showed low conductivity (<10-4 S cm-1).10,11 In contrast,
1 Introduction
2
when TTF is in excess, only TTF+ is formed. As a result, the mixed-valence
states, i.e., TTF0 and TTF+ are present in the obtained complexes.10,12,13 It has
been reported that the common feature of conductive donor-acceptor complexes
is the mixed-valence (or partial-oxidation) state of constituent molecules in the
compound. Thus, appropriate oxidants are needed for the preparation of
conductive TTF-based complexes which can oxidize TTF0 partially to TTF n+
(n<1). In this work, in order to obtain conductive nano-/microstructures, the
concentration of TTF solution is higher than metal salt, so that partially oxidized
TTF (TTF0 and TTF+) were formed in the charge-transfer complexes (detailed
results can be found in Chapter 5.3).
Apart from TTF, its selena and tellura analogues show good conductivities for
organic conductors. These two analogues were not used in this project, so no
further discussion in this thesis. Detailed information can be found in literature.5
Various TTF derivatives are reported for building organic conductors, and this
makes TTF to be the widely used electron donor molecule. For example, a TTF
derivative bis(ethylenedithio)tetrathiafulvalene (BEDT-TTF) which exhibits
predominant electrical properties, has been used as electron donors to form
charge-transfer compounds.14
Electron acceptors refer to the molecules that accept electrons from electron
donors and act as the oxidizing agent in reaction. TCNQ is a well-known strong
electron acceptor with the strong electron withdrawing cyano-group in the
structure (Figure 1.2). It was first synthesized in 1973, and was found to yield a
class of anion-radical derivatives with high stability and low resistivity.2 Upon
the reaction with electron donors, the strong π-acid TCNQ can undergo
complete electron transfer to form anion radical TCNQ-·. The resulting
complexes have two series of crystalline structures, with the first series of
M+TCNQ-· which shows high electrical resistivity, and the second one of
M+(TCNQ-·)TCNQ, involving neutral TCNQ in the structure and exhibiting low
electrical resistivity. Figure 1.2 shows the molecular structure of TCNQ and its
anion radical TCNQ-·.
1.1 Electron donor and acceptor molecules
3
Figure 1.2. Chemical structures of TCNQ and its anion radical.2
TTF can react with different electron acceptors to form a variety of electrical
conductive materials from insulators to superconductors.10 The electrical
properties of TTF-based complexes depend on the charge-transfer and also the
stacking mode of TTF in the obtained compounds.15 As for the charge-transfer
salts, the mixed-valence TTF systems showed very high electrical conductivities
for organic solids.16,17 On the other hand, for the orientation and alignment of
TTF units, the uniform spacing between the components in such complexes can
contribute to the good conductive behaviour. Charge-transfer compounds will
exhibit high conductivity performance when donor (D) and acceptor (A) units
alternately stacked to form a mix-stack structure (Figure 1.3a).5 In a special case,
when neutral molecules (D0A0) and charged molecules (D+A-) stacked, partial
charge-transfer complexes can be obtained, with a segregated stack structure
consisting of two independent columns of donors and acceptors (Figure 1.3b).18
A typical example is TTF-TCNQ, which shows segregated columnar stacks of
cations (TTF+) and anions (TCNQ-), as can been see from the crystal structure
(Figure 1.4).19
Figure 1.3. Stacking of charge-transfer salts. a) Mixed-stack, and b) Segregated
arrangements of D (hatched) and A (open) molecules in charge-transfer salts.5
1 Introduction
4
Figure 1.4. Crystal structure of TTF-TCNQ 2:1 charge-transfer compound (ac face).19,20
Reproduced with permission of the International Union of Crystallography.
1.2 Metal-organic charge-transfer complexes
Metal-organic frameworks refers to coordination polymers with metal ions
coordinated to organic ligands to form one, two, or three-dimensional
structures.21 Another class of metal-organic materials are charge-transfer
complexes. For metal-organic charge-transfer compounds, organic conductors
which are characterized as strong electron donor and acceptors are used as
organic ligands. When TTF is used as electron donor, an oxidized state of metal
ions will be the electron acceptor.12 When TCNQ is applied as oxidizing agent,
a neutral metal is required in spontaneous electrolysis process,22 while metal
ions can be used in electro-crystallization synthesis.23
Because of the properties mentioned in Chapter 1.1, TTF has been widely used
not only for synthesis of numerous organic conductors, but also as a precursor
for low scale self-assembled structures formed upon reaction with transition
metal salts. Various metal-TTF charge-transfer compounds have been
synthesized, either in bulk reactions24-26 or by microfluidics.27,28 These metal-
organic complexes showed different morphologies, such as dendritic,26 rod
1.2 Metal-organic charge-transfer complexes
5
like,24 wire like,27 etc., among which wire-like structures exhibited the most
intriguing characteristics. TTF-based metal-organic charge-transfer
nanostructures were synthesized by self-organization of the 1D stacking of TTF
component. A typical structure was Au-TTF, which was generated by the redox
reaction between TTF and hydrotetrachloroaurate (HAuCl4) solutions (Figure
1.5).24 The electron transfer between TTF and Au(III) induced the formation of
TTF radical cation (TTF+·) and neutral Au0. The 1D crystalized nanowires were
formed with a backbone of the mixed stacking of neutral TTF (or derivatives)
and TTF (or derivatives) radical cation.24,25
Figure 1.5. The formation of Au-TTF nanowires by charge-transfer. Reprinted with
permission from (Naka, K.; Ando, D.; Wang, X.; Chujo, Y. Synthesis of Organic-Metal
Hybrid Nanowires by Cooperative Self-Organization of Tetrathiafulvalene and Metallic
Gold via Charge-Transfer, Langmuir 2007, 23, 3450-3454). 24 Copyright (2007)
American Chemical Society.
In the present work, the synthesis of Au-TTF nano-/microwires was achieved
both by bulk synthesis and on microchips. The diameter of these wires varied
according to the synthesis method, either by stirring in bulk, or by laminar flow
or diffusion-based synthesis on microchips (see details in Chapters 5.1 and 5.3).
Due to the existence of the mixed-valence of TTF, Au-TTF wires showed good
1 Introduction
6
conductivity behaviour. The source-drain current of single Au-TTF wires were
measured, with no gate effect observed in the measurement, indicating the
metallic behaviour of Au-TTF charge-transfer wires.27 Furthermore, due to the
existence of Au0, surface modification on the active surface of Au-TTF was
possible by adding Au-active molecules into the system. For example,
fluorescent carboxylate-modified nanoparticles were used to bind on Au-TTF
nano-/microwire surface.27,29
Metal-organic charge-transfer salts based on TCNQ exhibit good conducting
and magnetic properties.5,30,31 The synthesis of TCNQ-based materials has
received considerable research interest due to potential applications in optical,
electrical and sensing devices.32,33 TCNQ acts as a bidentate, as well as a
tetradentate organic bridging ligand, which allows the formation of products
with 1:1 and 1:2 metal-TCNQ ratios.34-36 TCNQ can react with metal ions to
form stable 1:1 charge-transfer complexes like K-TCNQ and Na-TCNQ, which
are highly 1D compounds, consisting of TCNQ stacks and showing
semiconductor behaviour above room temperature.37 The 1:1 charge-transfer
complexes of Ag and Cu are of intense interest due to the fact that these
complexes undergo electric field induced bistable switching effect.38,39 In
particular, Ag-TCNQ and Cu-TCNQ can form quasi-1D wire-like structures,
which make them intriguing in the field of nanoelectronics.40 Apart from their
good resistive-switching properties, Ag-TCNQ and Cu-TCNQ have showed
their advantages over other organic memory materials because they can be
formed via self-assembly on the corresponding metal by spontaneous chemical
reaction with TCNQ (either in gas41 or solution39 phase). This aspect is very
promising to fabricate novel nanoelectronic memory devices. Figure 1.6 shows
a typical synthesis of Ag-TCNQ by spontaneous electrolysis method.
Figure 1.6. The synthesis of Ag-TCNQ complex. a) Structure of TCNQ molecule. b)
Equations showing the reactions towards Ag-TCNQ wire formation. (1) The silver ion
salt (AgNO3) is reduced by reductant (Red) to metallic silver (Ag0). (2) Neutral Ag
1.3 One-dimensional (1D) nanostructures
7
spontaneously reacts with TCNQ to form the charge-transfer complexes Ag+TCNQ-. For
the formation of Ag-TCNQ, two steps can be illustrated: with (2a) TCNQ reduced to
TCNQ- and simultaneously (2b) Ag0 is oxidized to Ag+.28 Reprinted with permission
from (Cvetkovic, B. Z.; Puigmarti-Luis, J.; Schaffhauser, D.; Ryll, T.; Schmid, S.;
Dittrich, P. S. Confined Synthesis and Integration of Functional Materials in Sub-
nanoliter Volumes, ACS Nano 2013, 7, 183-190). Copyright (2013) American Chemical
Society.
As organometallic semiconductors, Ag-TCNQ and Cu-TCNQ exhibit bistable
reproducible and electrical resistive switching properties because of the changes
in composition. M+(TCNQ-·) (M represents metal, also as following) exhibits
high resistance of 104-1012 Ω·cm, while M+(TCNQ-·)(TCNQ0) with neutral
TCNQ shows low resistance of 0.01-100 Ω·cm.42 When a reversible voltage is
applied to M-TCNQ, it can show the reversible electronic behaviour between
high resistance (OFF) and low resistance (ON) states (Equation 1.1. Reprinted
with permission from Applied Physics Letters. Copyright (2006) AIP
Publishing LLC). 38,43
(1.1)
Although the exact switching mechanism has not yet been determined, reported
studies have already proved the existence of TCNQ0 after the switching process
to “ON” state.44 Thus, in the low resistance state, neutral TCNQ contributes to
the stack with TCNQ anions to form the mixed-valence structure. These results
proved the importance of the organic memory complex M+TCNQ- in the
switching process.45
1.3 One-dimensional (1D) nanostructures
1D or quasi-1D structures such as nanowires, nanofibers, nanorods, and
nanoparticles have been widely studied due to their physical properties of ultra-
small scale, large surface-to-volume ratio and possible applications in the field
of electronics, optics, magnetic devices, sensors, etc.46-48 For sensors, the size of
substrate compared to that of target molecules is one of the most important
factors determining the detection limit. Thus, nanowires and nanofibers are
considered as optimal building blocks for sensing substrate. Their large surface-
1 Introduction
8
to-volume ratio is suitable to study chemical compounds and biomolecules with
high sensitivity.49 In the past decades, different classes of materials have been
extensively reported as sensor substrates, such as silicon nanowires, carbon
nanotubes, metal nanowires, etc.47,49,50 However, metal-organic structures have
not been explored to a large extent. Compared to other materials, nanowires
made from metal-organic compounds have showed their obvious advantages in
tunable properties based on the use of different metal units and organic ligands
and label-free sensing by optical or(/and) conductive readout.28,51 Thus, using
such metal-organic materials as sensing platform is one of the focuses of this
research.
1D or quasi-1D structures exhibit interesting and unique properties because of
the anisotropic structures, quantum confinement and high surface-to-volume
ratios. The mechanical, electronic and optical properties of such structures
greatly depend on their particle size. Furthermore, nanostructures are the
smallest structures employed for electron transport, and thus of special interest
in nanoelectronics. Therefore, the development of nano-/microstructures
represents an important issue for the miniaturization of electronic devices.50,52,53
In the past decades, various nanostructures have been studied and reported,54-57
among which metal-organic nanomaterials58-60 showed their unique optical and
electrical properties, compared to other materials such as metals, organic
materials, inorganic materials and conducting polymers.
To obtain nano-/microstructures, different chemical methods have been
established, such as template-assisted top-down synthesis, which is most widely
used for metallic or inorganic nanowires,61-63 while organic polymers, bio-
organic and metal-organic nanomaterials can form anisotropic structures via
bottom-up process.57,58 The variabilities in chemical reactants allow the ease of
fabrication of anisotropic nanostructures, as well as fine-tuning of their physical
and chemical properties. For metal-(bio)organic structures, the most common
approach is based on the self-assembly of metal ions and (bio-)organic ligands
in solution under certain reaction conditions,64,65 or with the assistance of other
advanced synthesis techniques, such as electrospinning and ultrasonic.66-68
However, bulk synthesis based on above-mentioned bottom-up approaches
faces the challenges including the size-selective synthesis, controlling and
guiding the assembly of precursors, synthesis under less harsh experimental
conditions, etc.69,70
1.4 Scope of the thesis
9
1.4 Scope of the thesis
The overall goal of this project is the microfluidic-assisted formation of
nanowire-based sensors that are capable of detecting chemical molecules and
biomolecules. With this aim, novel materials (transition metal and organic
ligands, TTF and TCNQ) for metal-organic nanowire synthesis will be used and
evaluated. The binding of analytes on the wire structures can be quickly
determined by changes of the electrical conductivity, while the identification of
the bound molecules can be done by Raman spectroscopy. Thus, highly
sensitive nanosensors for the label-free detection of gaseous and liquid
bioanalytes will be developed. However, several challenges such as the
immobilization of the nanowires in withstand liquid flow, the optimization of
the microfluidic platform to Raman spectroscopy have to be addressed.
Specifically, the main objectives of this doctoral project are in four aspects:
a) Formation of nanowires with suitable materials: Various precursors need
to be tested. In particular, transition metal salts (such as Au, Ag, Cu and Fe salts)
in combination with the important organic molecules TTF and TCNQ are used
for synthesis in liquid. The formation of nanowires will be performed on
microfluidic platforms.
b) Optical and electrical characterization of such nanowires: to determine
their optical (mainly Raman) properties and conductive behaviour, depending
on the size (length and diameter) as well as the material composition (Ag, Au,
etc.). For Raman analysis, the resonance enhancement and resonance energy are
necessary. This was done via polarization dependent optical analysis in the
present work.
c) Localized synthesis and immobilization of wires: nano-/microwires will be
synthesized in confined area of microchannel and immobilized for further
modification. Both the synthesis and immobilization of the nanowires will be
finished in withstand liquid flow, in this case, different valves in microchips will
be applied for synthesis and trapping purposes.
d) Functionalization of the nano-/microwires for specific detection of
selected molecules: to test the sensing behaviour of as-prepared nano-
/microwires, these structures will be surface modified for binding target
molecules. In reported work, Au-TTF nano-/microwires have shown to sense
different gases by changes in electrical readout.28 In addition, the Raman signal
from the TTF ligand has been obviously enhanced. Based on these findings, the
1 Introduction
10
use Au-TTF wires as sensing elements combined to on-chip Raman
measurements can be proposed.
Based on the above mentioned objectives, the following parts of work were
accomplished, with each part addressed one or more aspects of the objectives:
1) Label-free biosensors based on in situ formed and functionalized Au-TTF
wires were developed on an integrated microfluidic system. After the synthesis
of Au-TTF microwires in microchip, the immobilization of wires was achieved
by pressing the control layer with nitrogen gas. With an active Au surface on
Au-TTF wires, different surface modification protocols were applied to modify
these wires which were used for sensitive label-free detection of catecholamines
and human IgG via Raman spectroscopy. The proposed method enabled an easy
functionalization of in situ formed structures and could favour their use as
sensing elements in microfluidic devices. In this part, the objectives of localized
synthesis and immobilization of wires and the functionalization for sensing
application were achieved.
2) Study of molecular self-organization in single TTF crystal and individual Au-
TTF nano-/microwire by polarized confocal Raman spectroscopy was
performed. A detailed structure analysis of Au-TTF wires was important to
understand the orientation of TTF molecules in a nanowire, the growth
mechanism of Au-TTF and the influence of the absorbent on the surface. The
proposed method based on polarized Raman spectroscopy and simulation was
first tested by studying the orientation of TTF molecules in TTF crystal. Then,
the same analysis was applied to the single Au-TTF nanowire. The results of
angular polarization Raman measurements and the simulation of depolarization
ratio demonstrated the tilted stacking of TTF units in single Au-TTF nanowire
along the long axis. This is the first study on molecular orientation of metal-
organic charge-transfer nanowires via polarized Raman spectroscopy. In this
part, the objective of characterizing and understanding wire structures was
achieved.
3) The formation of fibres and particles made of metal salts and derivatives of
TTF in a microfluidic device and in a conventional reaction flask was
comparatively investigated. Their morphologies, optical properties and
electrical conductivities were characterized. A series of uniform 1D structures
were successfully formed via charge-transfer interactions at the interface of two
laminar streams in the microdevice. The reaction occurred between metal and
different TTF derivatives indicated that the high planarity and strong molecular
interaction of TTF derivatives are beneficial to the formation of nanowires. This
1.4 Scope of the thesis
11
study provided a comprehensive overview of the morphologies of the products
obtained from reactions between metals and different commercially available
TTF derivatives such as TTF, formyl-TTF (FTTF), and BEDT-TTF. In this
subproject, the aims of finding suitable materials for nanowire sensors,
characterizing and understanding the structure were achieved.
4) The synthesis based on another organic ligand TCNQ was addressed. The
microfluidic-assisted synthesis of Cu-TCNQ nanostructures at ambient
environment is reported for the first time. A two-layer microfluidic device
comprising parallel actuated microchambers was used for the synthesis, and
enabled an excellent fluid handling for continuous and multiple chemical
reactions in confined ultra-small chambers. The localized synthesis of copper
and in situ transformation to Cu-TCNQ wires in solution was achieved by
applying different gas pressures in the control layer. The as-prepared Cu-TCNQ
wire bundles showed good electrical conductivity and hysteresis reversing
memory effect, which proved the possibility of using them to fabricate advanced
nanoelectronic components. In this subproject, the aims of locally synthesizing
of wires, and formation of suitable materials for potential sensing applications
were achieved.
The outline of this thesis is as follows: the thesis is divided into 6 Chapters, in
Chapter 1 (Introduction), a general introduction to the research is presented:
concepts of electron donor and acceptor molecules (Chapter 1.1), metal-organic
charge-transfer complexes (Chapter 1.2), one-dimensional nanostructures
(Chapter 1.3) are illustrated. The scope of thesis is introduced in Chapter 1.4.
Microfluidic technology is described in Chapter 2, which includes an
introduction to microfluidics (Chapter 2.1), microfluidic-guided synthesis
(Chapter 2.2) and microfluidic-assisted applications (Chapter 2.3) of nano-
/microwires, and microfabrication (Chapter 2.4). The experimental techniques
used in this work are described in Chapters 3, including optical techniques
(Raman spectroscopy in Chapter 3.1) and electrical techniques (Chapter 3.2).
The detailed sample preparation and characterization including materials,
microchip operation, preparation of solutions and characterization techniques
are described in Chapter 4. Results are discussed in Chapter 5 in four parts. In
the first part, Chapter 5.1, label-free biosensors based on in situ formed and
functionalized microwires in microfluidic devices are discussed. In Chapter 5.2,
the study of molecular self-organization in single TTF crystal and individual
Au-TTF nano-/microwire by polarized confocal Raman spectroscopy is
discussed. In Chapter 5.3, a comprehensive overview of the morphologies of the
products obtained from reactions of metal salts and TTF and TTF derivative are
1 Introduction
12
discussed. In Chapter 5.4, the synthesis and electric properties of Cu-TCNQ
which are based on a different organic ligand are described. Chapter 6 contains
conclusion and the outlook based on this project. Finally, molecular structures
of the main chemicals used for this work, as well as the materials and details on
suppliers are reported in Appendix.
2.1 Introduction to microfluidics
13
2 Methods-1: Microfluidics and microfabrication
2.1 Introduction to microfluidics
Microfluidic systems refer to miniaturized systems in micrometre (μm)
dimensions. Miniaturization of analytical systems for chemical and biological
applications has attracted great attention in the past decades. The first
micrometre-sized device was fabricated in the 1970s,71 and used as a
miniaturized gas chromatograph. After the introduction and great development
of micro total analysis systems (μTAS) in the 1990s,72 an increasing research
interest has emerged in this field worldwide, with a growing number of
publications on device fabrication and applications. Microfluidics involves two
aspects, on one hand the miniaturized system with dimensions of tens to
hundreds of micrometres, and on the other hand the technology in manipulating
small volumes (from nanolitre (10-9 L) to attolitre (10-18 L) range) of fluids with
this system.73 Since the miniaturized “lab-on-a-chip” device enables the
manipulation of ultra-small volumes, it offers obvious advantages over
conventional devices. A most important effect is that in microfluidics, viscous
forces are stronger than inertial forces, leading to the absence of turbulences, i.e.
laminar flow. Whether the flow of liquid is laminar or turbulent, can be
described by the Reynolds number (Equation 2.1), which is an important
dimensionless parameter in fluid mechanics,
𝑅𝑒 =𝑣𝜌𝑙
𝜇=𝑖𝑛𝑒𝑟𝑡𝑖𝑎𝑙 𝑓𝑜𝑟𝑐𝑒𝑠
𝑣𝑖𝑠𝑐𝑜𝑢𝑠 𝑓𝑜𝑟𝑐𝑒𝑠 (2.1)
with flow velocity v density
length scale l and dynamic viscosity of the fluid
For the flow at low Reynolds numbers (less than 2000), the viscous forces
dominate the fluid properties in microchannel, and the liquid flow is laminar.
The transition between laminar and turbulent is in the range of 2000-3000,74
when Reynolds number is larger than 3000, inertial forces are dominant,
resulting in a typical turbulent flow behaviour. The Reynolds number of
microfluidic systems is typically in the range of less than 10, indicating the
laminar flow behaviour in microfluidic devices. If two reactant solutions flow
into one channel and the flow is laminar, a stable interface between two reactive
2 Methods-1: Microfluidics and microfabrication
14
streams can be established, and the mixture of the solutions will be determined
by diffusion. Therefore, the reaction will take place preferentially at the
interface of the reactant solutions. Since reaction is determined by both the
length of the channel and the flow rates, tuning of reaction times inside a
microfluidic system is possible by controlling the flow rates of solutions, and
this can be an advantage for fast chemical reactions.
For the substrate of microfluidics, it was firstly performed on etched and bonded
silicon and glass, which required a clean room environment and thus was limited
in use because of the high fabrication cost. In the year 1998, Duffy et al.
introduced the soft lithography technology for microfluidics fabrication and use
polymers such as poly(dimethlysiloxane) (PDMS) to replicate microstructures
from silicon wafers in normal laboratories.75 Since only silicon wafers need to
be prepared in clean room, soft lithography became an important technique in
making microchips due to advantages such as low cost, ease of fabrication and
rapid prototyping. After further development of monolithic micro-fabricated
valves and pumps by multilayer soft lithography in 2000 by Unger et al.,76
microchips based on soft lithography fabrication turned to be an more advanced
technology.
In general, microfluidics offers various benefits for different research fields.
Miniaturized systems are very promising and efficient in cell analysis, due to
the ultra-small size. Apart from the easyhandling of fluids, the use of
microfluidic reactors for chemical processes offers several advantages. The
decreased sample consumption can be advantageous for the generation and
screening of huge compound libraries, but also allows to reduce the costs of
expensive biological samples and the volume of generated chemical wastes,
with a great benefit for environment. Combined with increasing number of
advanced microchip devices, microfluidics technique has become more and
more important with growing applications in many different subjects and
modern analytical fields, such as chemistry,77,78 biology,79,80 material
sciences,81,82 drug discovery,83 and cell analysis.84,85
2.2 Microfluidic-guided synthesis of nano-/microwires
In the past couple of years, much effort has been focused on the development of
new techniques to produce uniform nanowires using top-down (lithography)
and bottom-up (self-assembly) approaches.86-88 For bottom-up processes, a
variety of solution-based deposition techniques have been developed including
2.2 Microfluidic-guided synthesis of nano-/microwires
15
chemical synthetic approaches, self-assembled monolayers, Langmuir-Blodgett
films, electrochemical methods, etc.89,90 Although these techniques made much
progress in the fabrication of nanowires, it is still challenging to obtain uniform
wires, to control and guide the assembly of structures at the nanoscale.
In recent years, microfluidics techniques have been demonstrated to be an
attractive approach for fabricating micro- and nanostructures due to the
excellent properties when scaling-down the dimensions of a system. It can offer
numerous benefits, e.g. laminar flow, reduced sample/reagent consumption, fast
analysis and separation.91,92 Compared to conventional large-scale methods,
microfluidic systems provide unique advantages in precise fluid handling and
metering, which enable spatial localization of small reagent volumes in
microchannels.92-95 In particular, under the non-turbulent laminar flow
conditions, a stable interface between two co-flowing streams can be established.
In this case, reaction and structure assembly will directly take place at the
interface and the diffusion zone of the reactants (Figure 2.1).91,96 In addition,
microfluidics facilitates the reproducible and well-controlled supply of reagents,
leading to the formation of high-quality homogeneous micro- and
nanostructures. Hence, microfluidics is considered as an advanced tool for
forming low scale structures and many new microfluidic devices have been
presented in recent years for these purposes. Thus, different microfluidics
devices were also applied for the synthesis of nano-/microstructures in this
project.
Figure 2.1. Microfluidic synthesis based on continuous laminar flow. a) Schematic
illustration of the fabrication of 1D coordination polymer nanostructures using laminar
2 Methods-1: Microfluidics and microfabrication
16
flow in a microfluidic platform. b) Optical microscope image showing the guided
assembly of 1D nanostructure bundles created at the interface between aqueous Ag(I)
metal ions and cysteine solutions. Scale bar: 100 m. c) Optical micrograph of a silver
wire deposited in a zigzag channel at the laminar flow interface between solutions
containing the components of an electroless silver plating solution. Scale bar is 500 m
and 80 m for the low and high magnification of channel. Figures a) and b) were
reprinted with permission from (Puigmartí-Luis, J.; Rubio-Martínez, M.; Hartfelder, U.;
Imaz, I.; Maspoch, D.; Dittrich, P. S. Coordination Polymer Nanofibers Generated by
Microfluidic Synthesis, Journal of the American Chemical Society 2011, 133, 4216-
4219).58 Copyright (2011) American Chemical Society. Figure c) was reprinted with
permission from (Science 1999, 285, 83-85). 91 Copyright (1999) The American
Association for the Advancement of Science.
Furthermore, microfluidics systems can be integrated with pneumatically
actuated valves29,97-100 that can create micrometre-sized reaction chambers
inside the channels. This is particularly useful to handle ultra-small amount of
liquids and to enable the chemical reactions in confined chambers. The valves
can have a round shape like a “donut” or a line shape, depending on various
applications. A typical “donut”-shaped valve shown in Figure 2.2a was firstly
developed to realize the on-chip immobilization of small particles and used to
obtain different lengths of metal-organic wires. Another two-valve microchip
design was developed based on the “donut” valve for single cells trapping and
analysis.101 Linear gas control was used to control solution supply, usually in a
parallel way (Figure 2.2 b). A microchamber array composed of ten parallel
microchambers was controlled by three gas control channels. By applying
different pressure to different channels, the solution supply is controlled to allow
the reaction in the confined microchambers.28 These static micrometre-sized
reactors can be designed in a highly parallel way and used for biochemical and
chemical reactions.76
2.2 Microfluidic-guided synthesis of nano-/microwires
17
Figure 2.2. Microfluidic devices with valves. a) Microfluidic device with donut shape
valve. i) Schematic illustration depicting the principle of wire trapping. The bottom
fluidic layer is used for nanowire formation when co-flowing the reactant solutions and
the top layer is the control layer that can be pressurized by N2 gas. Donut-like features
and N2 gas supply channels are impressed into this layer. A flexible PDMS membrane
separates both layers. Wires formed at the interface can be selectively immobilized by
pressurizing the donut, thereby defining the length of the trapped nanowires. ii)
Micrograph of the device. The channels in the fluid layer and the control layer are filled
with food dye. Scale bar: 1mm. iii) Magnification of the donut features. Scale bar: 300
Pm. Figures are adapted from Reference 29 with permission from The Royal Society of
Chemistry. b) Microfluidic device with linear valve and multi-chamber arrays. i)
Micrograph of the multilayer PDMS chip with an array of 10 parallel microchambers in
the fluid layer (red channels, 10 μm high). Fluid metering and supply is controlled by
three pneumatically actuated valves (control layer, blue channels, 100 μm high). Scale
bar: 750 μm. ii) Magnification of two microchambers. The final microchambers are
confined by the two side valves and encapsulate a volume of 675 pL. Scale bar: 150 μm.
iii)-vi) Series of micrographs showing the operation of the valves and the diffusive
mixing of two food dyes in a typical reaction procedure, corresponding schematic side
views are shown on the bottom of each figure. iii) Introduction of the reagents, iv)
Compartmentalization of two reagent volumes on either side, v) opening of the central
valve, and vi) diffusive mixing. Scale bars: 100 μm. Reprinted with permission from
(Cvetkovic, B. Z.; Puigmarti-Luis, J.; Schaffhauser, D.; Ryll, T.; Schmid, S.; Dittrich, P.
S. Confined Synthesis and Integration of Functional Materials in Sub-nanoliter Volumes,
ACS Nano 2013, 7, 183-190).28 Copyright (2013) American Chemical Society.
2 Methods-1: Microfluidics and microfabrication
18
2.3 Microfluidic-assisted application of nano-/microwires
For applications in nanoelectronics, nanostructures must be not only synthesized,
but also positioned and integrated with electrodes. On one hand, these multiple-
steps can be done separately, that is, by aligning pre-synthesized nanostructures
onto microchip devices for analysis. For example, a single-walled carbon
nanotube (SWNT) prepared by chemical-vapour-deposition growth was aligned
to a PDMS microchip and connected to Pd/Au source and drain contacts.
Because a microchip can be used as a microdevice with a delivering system for
further reactions on the nanowires formed/trapped inside the microchip. This
SWNT transistors was successfully used to sense the local potential generated
by a fluidic flow of ionic solutions on charged surfaces.102 On the other hand,
the synthesis and application can be in principle totally performed in one
microfluidic device. To achieve this goal, a microchip with prefabricated
microelectrodes is needed.98 Together with the localized formation of structures,
direct analysis of electronical characteristics of conductive nanostructures is
possible, without further complicated manipulation (Figure 2.3).
Thus, features as the ease of fabrication, laminar flow, elimination of local
variation, control of reaction times by diffusion and total integrated system
make microfluidics systems very attractive and promising for synthesis and
analysis of nanoscale structures. Therefore, a huge effort has been made
worldwide in order to explore the appealing features that microfluidics offers
for the formation and analysis of various micro and nanoscale structures and
devices.
2.3 Microfluidic-assisted application of nano-/microwires
19
Figure 2.3. Microfluidic-assisted application of conductive nanostructures. a) Schematic
diagram of a nanotube-based flow sensor showing a nanotube integrated with source,
drain and counter electrodes. The device is covered by a PDMS membrane with a micro-
fabricated channel. The channel is placed over the device and liquid is caused to flow
through it. Figure was adapted from Reference 102 with permission from Nature
Publishing Group. b) Optical image of the electrochemical fabrication of conducting
polymer nanowires in the microfluidic channel (left) and the SEM image showing the
nanowire grown in the channel on top of the pre-existing Pt electrodes. Figure adapted
from Reference 98. c) i) Scheme showing a microchamber integrated with prefabricated
electrodes. ii) Polarized micrograph of Au-TTF wires formed inside a microreactor
chamber, showing the alignment of the wire on top of the microelectrodes. Scale bars:
50 μm. Reprinted with permission from (Cvetkovic, B. Z.; Puigmarti-Luis, J.;
Schaffhauser, D.; Ryll, T.; Schmid, S.; Dittrich, P. S. Confined Synthesis and Integration
of Functional Materials in Sub-nanoliter Volumes, ACS Nano 2013, 7, 183-190). 28
Copyright (2013) American Chemical Society.
Reported studies have shown the possibilities of using microfluidics to form
metal nanowires,91 metal-(bio)organic nanostructures,58 polymer fibres,98 etc.,
as well as to develop microfluidic-synthesized nano-/microstructures for sensors
and nanoelectronics.
2 Methods-1: Microfluidics and microfabrication
20
2.4 Microfabrication
Before the fabrication of a microchip, the first step is the development of a
microchip design for the specific application. The microchip design is prepared
as a schematic drawing showing the specific microchip properties, with the help
of a computer-aided design (CAD) software. Secondly, this CAD drawing has
to be transferred onto a film mask. Afterwards, the micropatterns on the film
mask are transferred into a photosensitive polymer (photoresist) to form a mold
by photolithography. Finally, the microchip is fabricated by soft lithography
using PDMS and bonded to a glass slide to form the final device. The
introductory knowledge on the fabrication of silicon wafers, masters, glass slide
with microelectrodes is shown in Chapter 2.4.1. The mechanism of soft
lithography is explained in Chapter 2.4.2, followed by the detailed description
of the microchip preparation protocols applied in this work (Chapter 2.4.3).
2.4.1 Wafers and masters
a) Silicon wafers
Nowadays, silicon wafers used for building microfluidic devices are
predominantly fabricated by photolithography, due to its possibility to prepare
structures down to sub-micron size. Photolithography is a technique to transfer
a structure from a photomask to photoresist using light, e.g. UV light at 365 nm
or 405 nm. In general, depending on the working mechanism of photoresists,
two different behaviours can be observed upon UV-irradiation. A negative
photoresist e.g. SU-8, the illuminated areas polymerize when exposed to UV
light, whereas for a positive photoresist, e.g. AZ 9260, the non-exposed areas
polymerize. These two different photoresist are used for the fabrication of
different channels, for rectangular channels SU-8 is used, while for round
channels the AZ resistant is applied. Depending on the photoresist used for
wafer fabrication, different photomasks are applied. For negative resist SU-8, a
dark field mask is prepared because features in white will result in a structure
(Figures 2.4a and 2.4c). In contrast, for positive resist AZ, bright field mask will
be used since features in black generate the structures (Figures2.4b and 2.4 d).
2.4 Microfabrication
21
Figure 2.4. Images of chrome masks and wafers used in this thesis. Test plot images of
a) SU-8 dark field chrome mask and b) AZ bright field chrome mask. Photographs of 4-
inch wafers of c) SU-8 produced from a) and d) AZ fabricated from b). Note: mirror
images of wafers are shown in test plots.
b) SU-8 silicon masters
To fabricate the master forms for preparing microchip layers, a 4-inch silicon
wafer is first spin-coated with SU-8 photoresist and soft baked. The microchip
features is transferred from the silicon wafer onto SU-8 by the use of photomask
and UV-light exposure. After the post-exposure baking step, the non-exposed
SU-8 is developed and hard-baked to get the final device. All master forms were
silanized overnight under vacuum using 1H,1H,2H,2H-perfluorodecyl-
dimethylchlorosilane in order to avoid PDMS adhesion during chip fabrication.
For the fabrication of different master molds with various heights, different
types of SU-8 photoresists with their own process parameters can be applied.
An introductive schematic illustration is shown in Scheme 2.1. Details on the
fabrication method can be found in literature.103
2 Methods-1: Microfluidics and microfabrication
22
Scheme 2.1. Schematic of the microfabrication steps of a SU-8 master mold. Scheme
reproduced from Reference 103.
c) AZ silicon masters
For the fabrication of pneumatic actuated valves, round-shaped fluidic channels
are needed. The master mold for the liquid channel was prepared using a 4-inch
silicon wafer. A silicon wafer was first spin-coated with AZ photoresist and soft
baked. The desired microchip features was transferred onto AZ using a
photomask and UV-light exposure. After the development of the non-exposed
AZ, a reflow procedure was performed to get the final device. An introductive
schematic illustration is shown in Scheme 2.2. More details on the fabrication
method can be found in literature.103
Scheme 2.2. Schematic of the microfabrication steps of an AZ master mold. Scheme
reproduced from Reference 103.
2.4 Microfabrication
23
d) Glass slide with microelectrodes
For electrical characterization of conductive wires, platinum (Pt) patterned
microelectrodes were fabricated on top of glass coverslips (24×40 mm, No. 5,
Menzel Gläser) by conventional photolithography and physical vapour
deposition. A detailed description of the fabrication method is reported
elsewhere.28,103
Figure 2.5. A photograph of a glass slide with microelectrodes (left, used in the
experiments reported in Chapter 5.3 and Chapter 5.4. Details of the four-point
microelectrodes under optical microscope are shown in the right image (Scale bar: 50
Pm).
2.4.2 Soft lithography
Soft lithography refers to the technique that involves an elastomer as the mask,
stamp, or mold.104 It is a widely used to fabricate microfluidic devices.
Nowadays, the most popular elastomeric material for microchips is the
commercially available PDMS, because of its useful properties including low
cost, low toxicity, high biocompatibility, chemical inertness, mechanical
flexibility and durability, as well as the ease in fabrication. PDMS consists of
two components, an oligomer (elastomer) base and a curing agent (Pt catalyst).
In the oligomer base, there exist vinyl group-terminated siloxane oligomers and
cross-linking oligomers (Figure 2.6). The curing agent is composed of Pt-based
catalyst that can catalyse the cross-linking reaction. To produce a PDMS-based
microchip, the two gels are mixed (Figure 2.6). In addition, the hardness of the
PDMS microchip can be controlled by using different weight ratio of the
components (PDMS oligomer: Curing agent, 5:1-20:1). If the ratio of curing
agent is increased, higher cross-linking reaction will result in a harder PDMS
microchip, while a lower ratio of curing agent will form softer microchip.103,105
2 Methods-1: Microfluidics and microfabrication
24
Figure 2.6. The formation of PDMS by catalytic cross-linking reaction. Catalytic
formation of PDMS with the addition of the Si-H bond (cross-linking copolymer) to the
vinyl groups (siloxane oligomer), forming Si-CH2-CH2-Si linkages. The multiple
reaction sites on both the base and cross-linking oligomers allow for three dimensional
cross-linking. Figure reproduced from Ref. 103.
2.4.3 Microchip fabrication
All microchips used in this work were fabricated by soft lithography, introduced
in Chapter 2.4.2. The microchannel features imprinted in the PDMS layer need
to be sealed with a glass slide to get the final microfluidic device. In most cases,
a strong permanent bonding between the PDMS layer and glass slide is required,
in order to use high pressure to pump liquid into the microchannel. Thus, an
oxygen plasma is applied to activate the surfaces for bonding. However, when
the structures formed inside the microchannel are positioned and needed for
further characterization, a non-bonded microchip is used, where the PDMS layer
and glass slide are hold together in a custom-made clamping device. The
fabrication protocols of one layer and double layer PDMS microchips (Figure
2.7) are reported in the following.
2.4 Microfabrication
25
Figure 2.7. Photographs of two microchips used in this project. a) A one layer microchip
filled with dark red food dye and b) A double-layer microchip filled with food dyes to
show the different layers, top layer in dark red and bottom layer in orange.
a) Single layer microchip fabrication
For preparation of a microchip layer in this project, PDMS oligomer and curing
agent were mixed at a ratio of 10:1, degassed, poured onto the wafer bearing the
respective features and degassed again, then heated in oven at 80 °C for 3 h.
After this, the layer was removed from the wafer and holes were punched with
a biopsy puncher (1.5 mm diameter, Miltex, York PA). Afterwards, the PDMS
layer and a cleaned glass slide (Thermo Scientific, Menzel-Gläser, Nr. 3) were
treated in oxygen plasma, and then carefully aligned and bonded together.
Finally, the bonded chip was heated at 90 °C for 30 min to confirm the
permanent bonding.
b) Double layer microchip fabrication
For the multilayer microfluidic chip, layers were prepared separately by casting
PDMS from the respective master forms. Alignment, assembly and bonding of
both layers resulted in the final device. For preparation of the control layer,
PDMS oligomer and curing agent were mixed at a ratio of 10:1, degassed,
poured onto the control layer wafer and degassed again, then heated in oven at
80°C for 3 h. After this, the layer was removed from the wafer and holes were
punched with a biopsy puncher (1 mm diameter, Miltex, York PA). For the fluid
layer, softer PDMS (mixing ratio elastomer: curing agent of 15:1) was spin-
coated at 2000 rpm on the structured four-inch fluidic master mold wafer to
create a 23 μm-high membrane, and then the wafer was heated at 80 °C for 1 h.
Both layers were activated in oxygen plasma using a plasma cleaner (PDC-32G
plasma cleaner, Harrick Plasma, power 18 W, time 45 s) and afterwards aligned
under a Multizoom AZ100M microscope (Nikon Corporation, Switzerland).
The assembly was cured for 2 h at 80°C to form a permanent bond between both
2 Methods-1: Microfluidics and microfabrication
26
layers. Afterwards, the device was removed from the master form and access
holes for the tubing (1.5 mm diameter, Miltex) were punched. The assembled
chip and a clean glass slide (Thermo Scientific, Menzel-Gläser, Nr. 1) were
treated in oxygen plasma, and then carefully aligned and bonded together.
Finally, permanent bonding was achieved by heating the well-assembled chip
was heated at 90 °C for 30 min.
Both single layer and double layer microchips were used in this project. Table
2.1 shows an overview of the microchips used in this work, as well as their
applications. Detailed microchip features can be found in separate Chapters.
Table 2.1. An overview of the microdevices used in different Chapters of this work.
Chapter
Topic
Microdevices
Function
5.1
Label
-
free biosensors based
on in situ
formed and
functionalized microwires
in microfluidic devices
Synthesis of
Au-
TTF
microwire;
Development
of sensors
5.2
Study of molecular self-
organization in individual
Au-TTF nano-
/microwire
by polarized confocal
Raman spectroscopy
Synthesis of
Au
-
TTF nano-
/microwires
5.3
Conductive single
nanowires formed and
analyse
d on microfluidic
devices
Synthesis of
metal-
TTF
(and
derivatives)
5.4
Localized synthesis of
conductive Cu-
TCNQ
nanostructures in ultra
-
small microchambers for
nanoelectronics
Synthesis of
Cu-
TCNQ;
Conductivity
measurement
3.1 Raman scattering
27
3 Methods-2: Raman microspectroscopy
Raman spectroscopy is a non-destructive optical technique which can be applied
to study the molecular vibrations of diverse materials on solid (such as metal
and semiconductors) at interfaces of solid-liquid, in liquid, and in complicated
bio-samples etc.106-108 Thus, in this project, Raman spectroscopy was the main
optical technique used for monitoring vibrational changes of molecules
absorbed on nanostructures (Chapter 5.1) in liquids and also the molecules
inside nanostructures (Chapter 5.2). Following, an introduction of Raman
scattering (including Stokes, anti-Stokes and resonance Raman scattering,
Raman spectroscopy and fingerprint, Raman and Infrared (IR) spectroscopy),
theoretical basis of Raman scattering (including normal modes, selection rules,
Raman tensor) and confocal and polarized Raman microspectroscopy (principle
and setup) is presented.
3.1 Raman scattering
3.1.1 Stokes, anti-Stokes and resonance Raman scattering
When an incident electromagnetic radiation encounters a molecule, it can be
absorbed by the molecule if the energy of the incoming photon matches the
energy difference between the ground state and the excited state of the molecule.
However, in most cases the incoming radiation does not match any molecular
transition. Alternatively, the photons are scattered from the molecule. Most
photons are elastically scattered. They have the same energy as the incoming
photons. This elastic scattering process is called Rayleigh scattering. In a few
cases, photons are inelastically scattered and energy is exchanged between
photons and molecules. This means that the molecule changes its vibrational
state upon interaction with the incoming photon, resulting in changes in the
energy or frequency of the scattered photons. This inelastic scattering effect is
termed Raman scattering. Apart from scattering at vibrations also other
elementary excitations may be involved, which however is not relevant here.
Raman scattering requires monochromatic irradiation for exciting molecular
vibrations. After inelastic scattering, a portion of the incident photons is
scattered inelastically. Thus, the energy of the scattered photons (һʋR) differs
3 Methods-2: Raman microspectroscopy
28
from that of the incident photons (hʋ0). According to the law of conservation of
energy, the energy difference corresponds to the energy change of the molecule,
and refers to the transition between two vibrational states.109 If the molecule
changes from a ground state to an excited vibrational state after scattering, this
means the molecule absorbs energy from incoming photons and the scattered
photons possess less energy than that of incident photon (hʋR < hʋ0). This
scattering is called Stokes Raman scattering. Conversely, when a molecule
relaxed from an initial excited state to to ground state, the energy of the scattered
photons is higher than incident photons (hʋR > hʋ0). This scattering is known as
anti-Stokes Raman scattering. For an anti-Stokes process to take place, the
molecule must be in an excited vibrational state, which requires thermal
excitation. According to the equilibrium on temperature (T) of Boltzmann factor:
exp(-E/KBT) (KB is constant of distribution), the higher the energy of the
vibration E is, the intensity of the scattering peaks becomes exponentially
weaker. Thus, anti-Stokes Raman scattering is weaker than the Stokes scattering
process.106 The Jablonski diagrams in Figure 3.1 illustrates the energy changes
in Rayleigh, non-resonance (Stokes and anti-Stokes) and resonance Raman
scattering processes.
The initial excited state of a molecule in Raman scattering is also known as a
virtual state. In quantum physics, a virtual state is short-lived, unobservable,
intermediate quantum state. In a special case, if the virtual state coincides with
a real electronic state of the molecule (for example excited state S1, Figure 3.1),
the scattering process is resonant. Resonant Raman scattering has important
implications for both the magnitude and the selection rules of the effect in
Raman scattering.106
3.1 Raman scattering
29
Figure 3.1. Simplified Jablonski diagrams of light scattering. Schematic illustration of
Rayleigh scattering, non-resonance Raman scattering (Stokes and anti-Stokes) and
resonance Raman scattering processes.
3.1.2 Raman spectroscopy and fingerprint
The energy lost by the photons in the scattering event is called the Raman shift
and is defined in energy as (hʋ0-hʋR). It is therefore positive for a Stokes and
negative for an anti-Stokes process. Raman shifts are commonly expressed in
wavenumbers (cm-1), with the frequencies corresponding to particular
vibrational modes of molecules. A Raman spectrum corresponds to the
wavelength- (or energy) dependence of the Raman scattered intensity at a given
incident wavelength.106 It provides information by which molecules can be
identified. As mentioned above, the Stokes Raman intensity is higher than anti-
Stokes Raman intensity. So the normal Raman scattering refers to Stokes Raman
scattering. The Raman shift in Stokes Raman scattering is called Stokes Raman
shift, which is very useful to identify molecular vibrational states. Typically, the
frequency range from 400 cm-1 to 1500 cm-1 is known as the fingerprint region
for identifying molecules. Raman spectroscopy is the spectroscopic technique
based on measurement and analysis of the signals arising from the Raman effect.
3 Methods-2: Raman microspectroscopy
30
It is recognized as a non-invasive optical technique which can provide detailed
molecular information on the measured samples.110,111 It enables the direct
detection of molecules by analysing the fingerprints in their separate Raman
spectrum. Thus, Raman spectroscopy has general and wide applications in
various (bio-)chemistry fields including analytical chemistry, material science,
biosciences, pharmaceutical studies, etc.106
In thermodynamic equilibrium, the excited state is less populated than the
ground state. Therefore, the rate of transitions from the ground state to the
excited state (Stokes) will be higher than that in the opposite direction (anti-
Stokes). Thus, the ratio of Raman intensities (IStokes/Ianti-Stokes) highly depends on
the temperature, which is a useful tool for temperature-dependent measurements
in solid-state physics. Figure 3.2 shows an example of the Raman spectrum of
TCNQ. The peak positions of Stokes and anti-Stokes Raman scattering are
symmetric around Rayleigh scattering (Δʋ=0) because they correspond to the
energy changes between the same excited and ground resonant states. However,
the Raman intensities are temperature dependent, with Stokes scattering peaks
much stronger than anti-Stokes scattering peaks.
Figure 3.2. Raman spectrum of TCNQ under laser excitation of 532 nm. Rayleigh
scattering (Δʋ = 0) in the middle, with symmetric spectrum of Stokes (red) and anti-
Stokes (blue) shift. The dash line indicates the position where Δʋ is 0.
3.2 Theoretical basis of Raman scattering
31
3.1.3 Raman and Infrared spectroscopy
Molecular vibrations can be excited not only by the inelastic scattering of
photons (Raman scattering), but also by the absorption of light quanta. When a
polychromatic light irradiates molecules, if the light energy (hʋ0’) matches the
energy difference (hʋk) between the ground state (hʋi) and a vibrational state
(hʋj) of the molecule, direct absorption of photons is achieved (Equation 3.1).
ℎ𝑣0′~ℎ𝑣𝑘=ℎ𝑣𝑗−ℎ𝑣𝑖

 (3.1)
As these energy differences are in the order of 0.5 and 0.005 eV, light with
wavelengths longer than 2.5 m (IR light) is sufficient to induce the vibrational
transitions. Thus, vibrational spectroscopy that is based on the direct absorption
of light quanta is defined as IR absorption.109 The absorption of incoming light
in the infrared region results in a spectrum that corresponds to specific
vibrational modes in molecules measured. Raman and IR spectroscopy are both
optical techniques based on the changes in molecular vibration, showing unique
fingerprint of the molecule separately. The modes appearing in the Raman and
IR spectra are different, however, the two techniques provide complementary
information of the vibrational structure of molecules.109 In this work, IR
spectroscopy was also applied for the characterization of metal-organic
compounds (see Chapters 5.3 and 5.4).
3.2 Theoretical basis of Raman scattering
3.2.1 Normal modes
In the Cartesian coordinate system, each atom in a molecule can be displaced in
the x-, y- and z-directions, corresponding to three degrees of freedom.112 A
molecule of N atoms has in total 3N degrees of freedom. However, not all of
them correspond to vibrational degrees of freedom. If all atoms are displaced in
the x-, y-, and z-directions by the same increments, the entire molecule moves
in a certain direction, representing one of the three translational degrees of
freedom. The same with displacements of the atoms that correspond to rotation
of the molecule: a nonlinear molecule has three rotational degrees of freedom,
while a linear molecule has two. Thus, the remaining 3N-6 and 3N-5 degrees of
freedom correspond to the vibrations of a nonlinear and a linear molecule,
respectively. In molecules, movements of atoms create a vibrational motion,
which is known as the normal mode of a molecule. Normal mode analysis shows
the characteristics of a certain molecule in the frequencies and molecular
properties.109
3 Methods-2: Raman microspectroscopy
32
3.2.2 Polarizability and Raman selection rules
In a Raman scattering process, the electron cloud of a molecule is distorted from
its normal shape by the electric field vector of an incoming electromagnetic
wave, which induces an electric dipole moment. The extent of the electron cloud
distortion depends on its polarizability, α, which is defined to relate the induced
dipole moment ind to the electric field E causing it (Equation 3.2)
µ𝑖𝑛𝑑 =𝛼∙𝐸 =𝛼∙𝐸0𝑐𝑜𝑠(2𝜋𝑣0𝑡) . (3.2)
where E0 is the amplitude and ʋ0 is the frequency of the electric field component
of the incoming magnetic wave that produces the dipole moment.113-115 The
polarizability α is actually a tensor changing over time, and describes the
response of the electron distribution to the movements of the nuclei that oscillate
with the normal vibration. In the simplest case of a diatomic molecule (harmonic
oscillator) vibrating with a frequency ʋk, the nuclear displacement Qk is as
follows (Equation 3.3):
𝑄𝑘=𝑄0𝑐𝑜𝑠(2𝜋𝑣𝑘𝑡) (3.3)
where Q0 is the amplitude (maximum displacement) of the molecular vibration.
For small amplitudes of the molecular vibration, α can be approximated by a
function of Qk (Equation 3.4):
𝛼 =𝛼0+(𝜕𝛼
𝜕𝑄𝑘)0𝑄𝑘 . (3.4)
Here α0 refers to the polarizability at equilibrium position, and (∂α/∂Qk)0 refers
to the derivative of polarizability α with respect to the change in Qk, evaluated
at the equilibrium position. Combining Equations. 3.2 and 3.4, the general
expression for the induced dipole moment indcan be written as follows
(Equation 3.5):
µ𝑖𝑛𝑑 =𝐸0{𝛼0𝑐𝑜𝑠 (2𝜋𝑣0𝑡)+(𝜕𝛼
𝜕𝑄𝑘)0𝑄0𝑐𝑜𝑠[2𝜋(𝑣0+𝑣𝑘)𝑡]+
( 𝜕𝛼
𝜕𝑄𝑘)0𝑄0𝑐𝑜𝑠 [2𝜋(𝑣0−𝑣𝑘)𝑡]} (3.5)
Equation 3.5 includes three contributions to the scattered radiation appearing as
separate terms with ʋ0 for the Rayleigh scattering, while (ʋ0+ʋk) for anti-Stokes
3.3 Light polarization and depolarization ratio
33
Raman scattering and (ʋ0-ʋk) for Stokes Raman scattering. This equation implies
that if the derivative (∂α/∂Qk) is zero, that means the polarizability of molecule
does not change during molecular vibration, then it is not Raman active. Thus,
Raman scattering can be activated only in the presence of non-zero
polarizability derivative with respect to the normal coordinate (∂α/∂Qk 0). This
is also known as the gross selection rules for Raman scattering.109,115
3.2.3 Raman tensor
Derived from Equation 3.5, (∂α/∂Qk )0Qk is the Raman tensor. In general, Raman
tensor is a second rank, three dimensional tensor.116 Considering Ei and Es as the
unit vectors of the polarization of the incoming and scattered light, respectively.
The Raman scattering intensity I is described as follows (Equation 3.6):117
𝐼 ∝ |𝐸𝑖·𝑅·𝐸𝑠|2 (3.6)
where R is the Raman tensor. R is a symmetric tensor and has the same
symmetry as the corresponding molecular vibration.116 In real experiment, under
certain experimental conditions regarding the incoming and scattered
polarization, whether a Raman-active molecular vibration can be observed or
not is determined by the Raman scattering intensity (if the resonance Raman
scattering is already taken into account). This is specific Raman selection rules.
Thus, under different scattering geometries, the symmetry of the Raman tensor
and the underlying molecular vibrations can be identified by Raman
experiments. Results on Raman tensor calculation of molecules and the
polarized Raman measurement are shown in Chapter 5.2.
3.3 Light polarization and depolarization ratio
Light travelling in free space or isotropic medium can be considered as
transverse waves, with the electric field vector and magnetic field in directions
perpendicular to the direction of wave propagation.118 By using a linear polarizer,
the oscillation of these fields can be in a single direction which is termed linear
polarization. Conventionally, the orientation of linearly polarized light is
defined by the direction of the electric field vector, parallel or vertically
polarized.
In Raman measurements, the scattered light is the sum of the parallel component
and the perpendicular component. To obtain more information about the Raman
3 Methods-2: Raman microspectroscopy
34
modes of molecules, polarized analysis can be performed by using the vibration
and the 90 degree scattering configuration. In this way, the scattered light can
be separated into two configurations, with one parallel to the incident
polarization and the other one perpendicular to both the incident beam and the
incident polarization. In Raman spectroscopy, a concept widely used to
quantitative analysis of the sample polarization response is named
depolarization ratio P.119,120 It is defined as follows (Equation 3.7).112
𝑃 =𝐼𝑥𝑦
𝐼𝑥𝑥 =𝐼perpendicular
𝐼parallel =𝐼⊥
𝐼∥ (3.7)
Here, Ixy represents the Raman intensity at the perpendicular (crossed) laser
configuration, Ixx represents the Raman intensity at the parallel laser
configuration.
3.4 Confocal and polarized Raman microscopy: principles and setups
In this work, confocal Raman microscopy was exploited to focus on nano-
/microstructures and get their Raman spectra (Chapters 5). In particular, this is
beneficial to single wire analysis inside microchip (Chapter 5.1) as all the
signals recorded only from the structure, due to the sub-micrometre scale
resolution (focus diameter ~400 nm). Polarized confocal Raman measurement
was also performed to study the molecular structures of single nano-/microwire
(Chapter 5.2). In the following, the principles and experimental setups of
confocal and polarized Raman microscopy are introduced.
3.4.1 Confocal Raman microspectroscopy
In the past decades, Raman spectroscopy has been greatly developed in
scientific research and real applications, due to the great progress in
instrumentation. A macro-Raman spectrometer includes some basic
components: a monochromatic light source, i.e. a laser source to excite the
samples, a sample area with collection optics, a light dispersing unit
(spectrometer) and a detector recording the signal and passing the signal to a
computer. In Raman measurements, strong Raleigh scattering needs to be
avoided. Thus, modern instrumentation almost universally employ a notch filter
to filter out the strong Raleigh scattering light. Also a sensitive CCD detector is
used. A macro-Raman spectrometer has only a spatial resolution from 100 μm
to 1 mm. To measure even small samples at micro and nanoscale, Raman
3.4 Confocal and polarized Raman microscopy: principles and setups
35
spectrometer equipped with a microscope to magnify its spatial resolution is
required.
Micro-Raman spectroscopy is very useful to study the vibrational spectra
characteristics of very small samples, especially, when combined with confocal
microscopy. This can be more powerful and useful technique to study nanoscale
samples. With a pinhole added to the conjugate plane of the detection beam path,
out-of-focus light is filtered out. The focused laser beam can be calculated with
the spot diameter d (0.61 O/NA) (NA, numerical aperture) and focus depth l
(0.89 O/NA2). For example, with a 100× microscope (NA 0.90), a laser
wavelength (O of 532 nm, the spatial resolution are calculate to be sub-micron
range (d ≈ 0.36 μm, l ≈ 0.58 μm). Apart from the improved resolution in the
lateral plane (O/2, down to 200 nm), the advantages of confocal microscopy over
conventional microscopy also lie in reduced background signal, and the
possibility to perform two dimensional (2D) (surface) and three dimensional
(3D) (with depth profiling) Raman mapping on ultra-small samples in order to
understand the location and amount of different components. Figure 3.3 is a
schematic view of a micro-Raman setup. A microscope objective is added to the
system, with the illumination, scattering light collection passing through the
microscope objective in a back-scattering mode.
Figure 3.3. Components of a polarized confocal Raman microscope with back-scattering
geometry (blue solid lines indicate the incoming light and red dash lines represent back
3 Methods-2: Raman microspectroscopy
36
scattered light from the sample). Note the existence of linear polarizer (LP) after the laser
excitation, the polarizer analyser (PA) on the scattered light beam path and the half-wave
plate (HWP), which is used for the angular polarization measurement. Figure was
adapted from Reference 121 with permission from OSA publishing.
In the present work, mainly confocal Raman microscopy is employed to study
nano-/microstructures to obtain normal Raman spectra. In the attempt to achieve
increased signal and higher sensitivity of a given sample, a surface-enhanced
Raman scattering (SERS) substrate (e.g. Au, Ag) is required. There are a lot of
reported work on the application of SERS labels to reach very high sensitivity
and low limit of detection of chemical and biomolecules.122,123 However, this
technique is not used in this work.
3.4.2 Polarized Raman microspectroscopy
During the last years, polarized Raman spectroscopy has been proved to be a
very useful approach to determine the orientation of a molecular layer composed
of small organic molecules in organic thin layers or single crystals.112 Analysis
of the dependence of the Raman scattering intensity upon the angular rotation
of the sample around the substrate normal allows the determination of the
molecular orientation. The molecular vibrational modes change during the
rotation of the sample with a periodicity, suggesting a preferential molecular
orientation. Combined with confocal microscopy, the polarized confocal Raman
spectroscopy has been employed to study the composition of nano-/micrometre
scaled substrate due to its high spatial resolution and non-destructive
characteristics.124-126
Polarized Raman measurements are performed using a polarizer that allow
passing light of a desired polarization while blocking waves of other
polarizations.127 To study the nano-/microstructures, a linear polarizer needs to
be used to get linear polarized light. Then a polarization analyser inserted in the
scattered light beam path between the sample and the spectrometer, allowing
the Raman polarization to be selected. In a typical angular polarization
experiment of wire structure, the angle between the long axis of the structure
and the polarization direction of the incident laser is adjusted using a half-wave
plate. A detailed schematic illustration of a polarized confocal Raman
microscope can be seen in Figure 3.3.121
3.5 Instrumentation for Raman spectroscopy
37
In Raman scattering experiment, Porto’s notation is used to indicate the
configuration of the scattering 128, and express the orientation of the crystal with
respect to the polarization of the laser in both the excitation and analysing
directions. For example, in a light configuration with a back scattering geometry,
the direction of the propagation of the scattered light is in the opposite sense of
the direction of the incident light, for example X(YZ)X
. For polarization
measurement on samples, normally two configurations are used, parallel and
crossed configurations, with the direction of polarized light in a parallel and
orthogonal geometry, respectively.124 Polarized Raman measurements on a
single crystal and a single nanowire are discussed in Chapter 5.2 of this thesis.
3.4.5 In situ Raman measurements
For microfluidic-assisted reactions, a confocal Raman microscope allows to
detect small samples in a micrometre-scaled microfluidic channel and to gain
insight into the reaction occurring in microchips by combing the advantages of
both Raman spectroscopy and microfluidics.129 This method is known as in situ
Raman analysis and exploited to obtain a fast signal readout and in situ
monitoring of ultra-small samples, such as the Au-TTF nanostructures
investigated in the present work. The binding of analytes can be identified by
changes of the electrical conductivity/resistance (e.g. Chapter 5.3), the
identification of the bound molecules on wire surface can be done by changes
in Raman response (e.g. Chapter 5.1).
3.5 Instrumentation for Raman spectroscopy
a) Raman spectra of Au-TTF structures in open-donut microchip were recorded
on a confocal Raman microscope (CRM 200, WITec GmbH, Germany)
(neodymium-doped yttrium aluminium garnet, Nd:YAG laser, 532 nm) using
an upright 100× objective (Nikon, Japan) with a NA of 0.90. The laser power
was 1.4 mW, and the accumulation time was 2 s. In all the measurement, the
signals were collected from large Au-TTF wires (diameter ~ 2 μm) through a
relatively small laser point (diameter ~ 400 nm). Raman spectra were recorded
from six wires and averaged to represent the Raman results.
b) Raman spectra of metal-TTF nanowires were also obtained on the same
Raman microscope as that in a) (λ = 532 nm, 50× objective, NA = 0.50). Since
the wires are deposited on glass slide and directly applied for Raman
measurement in air atmosphere, a lower laser power of 0.37 mW than that of
3 Methods-2: Raman microspectroscopy
38
Au-TTF wires in liquid was used to avoid the heating effect. The accumulation
time was 15 s.
c) Raman spectra of Cu-TCNQ were obtained on in-house assembled confocal
Raman microscope equipped with a diode-pumped solid-state (DPSS) laser
(Cobolt Samba, λ = 532 nm), a 50× Olympus objective (NA = 0.75) and a liquid
N2 cooled CCD as detector. The laser power was 0.4 mW, and the accumulation
time was 60 s. A nano-piezo stage (Ratis, Nanoscan Technology) with a
minimum scan step of 0.1 nm in the lateral directions was used to move samples
during the measurement.
The polarized confocal Raman microscopy analysis of Au-TTF was also carried
out with the same in-house assembled Raman microscope. The excitation
wavelength generated by an Ar+ ion laser was fixed at O= 532 nm and the
ingoing polarization controlled using a half-wave plate. Focusing on the sample
was done with a 50× objective (Olympus MPL, NA = 0.75, for TTF crystals)
and a 100× objective (Zeiss EC Epiplan-NEOFLUAR, NA = 0.90, for Au-TTF).
Different laser intensities were applied for TTF crystal and Au-TTF wire, in
order to obtain the optimal Raman response. The laser intensity was about 15
PW at the focal point for the analysis of TTF crystals with an acquisition time
of 120 s per spectrum. For Au-TTF wires, a lower laser intensity (about 7 PW)
than the crystal and an acquisition time of 60 s per spectrum were applied to
lower the heating effect on wire. The Raman signal was collected under a
backscattering geometry. A polarizer served as an analyser just in front of the
entrance slit of the spectrometer. Two polarizations were adopted in the
measurement, by setting the polarization of the analyser on “0o” and “270o” for
parallel and crossed polarization, respectively. A half-wave plate served to carry
out the angular polarization experiments of wire structures. The angle of the
polarized incident light was varied in the range of 0-360o by rotating the half-
wave plate with the angle within 0-180o.
4.1 Microchip operation
39
4 Methods-3: Sample preparation and characterization
In this chapter, the experimental procedure including microchip operation;
sample preparation of Au-TTF, metal-TTF, TTF crystal and Cu-TCNQ; and the
detailed characterization method and setups are introduced.
4.1 Microchip operation
In this work, both single layer and double layer microchips were used. For single
layer microchip (used in Chapters 5.2 and 5.3), the reagents were loaded into
plastic syringes and introduced into the chip through Teflon tubings using a
syringe pump system (neMESYS module, Cetoni GmbH Korbußen, Germany)
that controlled the individual flow rates for the microchannels. The double layer
microchip has a top (control) layer and a bottom (fluid) layer. Custom-made
metal connectors and silicone tubing were used to supply N2 gas into the control
layer. Control channels were filled with water by pressurization before each
experiment to get rid of bubbles. Regarding the fluid layer, regents were
supplied into microchannel via a syringe-pump system (NanoJet, ChemyxInc,
Germany). For the project discussed in Chapter 5.1, the control layer was
pressurized up to 3 bar for closing the donuts and immobilizing the wires, and
then released to 2 bar in a functionalization step. For the project described in
Chapter 5.4, the gas control channels was pressurized up to 3 bar for confining
the microchamber, followed by a low pressure of 2 bar to allow the slow
diffusion of solutions into microchamber. Detailed microchip operation
procedures can be found in following chapters.
4.2 In situ synthesis, immobilization and functionalization of Au-
TTF microwires
This part shows the experimental procedure related to the results in Chapter 5.1.
In situ synthesis and immobilization of Au-TTF wires (Chapter 4.2.1), two
functionalization methods applied to Au-TTF wires for biosensors (Chapter
4.2.2) and immunosensors (Chapter 4.2.3) are discussed here.
4.2.1 In situ synthesis and immobilization of Au-TTF microwires
4 Methods-3: Sample preparation and characterization
40
Au-TTF wires were synthesized either by injecting the two solutions (TTF, 24
mM; HAuCl4, 6 mM, both in acetonitrile (CH3CN)) with a pipette into the
microchip from either channel, or by diffusing HAuCl4 solution (6 mM) into the
microchannel filled with TTF solution (24 mM). In this project, the latter
diffusion technique was mainly applied. After wire formation, the control layer
was pressured to 3 bar by N2 gas in order to trap the wires formed in the fluid
layer. Non-trapped wires were washed away by the injected solvent. Afterwards,
different solutions were supplied at low flow rates (2 to 10 μL/min) into the
microchannel, in order to avoid the flow forces to remove the structures. Also,
air bubbles were carefully avoided to keep the biomolecules, particularly the
antibodies, functional during the experiments. By using the microchips
described in Chapter 5.1.1, the wires could be easily flushed away, just by
releasing the pressure on the control layer and flushing with the solvent.
Afterwards, new wires could be easily formed for further experiment by
supplying new TTF and gold solutions into the channel. Thus, there was no need
to recover the wires after an experiment, instead, they could be easily rebuilt.
4.2.2 Functionalization protocol for biosensing applications
Dopamine (DA) hydrochloride was dissolved in an ethanol/phosphate buffered
saline (PBS) (1:1 solvent mixture (pH 7.2, PBS final concentration 10 mM) and
diluted to various concentrations (5 μM, 15 μM, 25 μM, 35 μM, 50 μM, 75 μM
and 100 μM). The solvent mixture ethanol/PBS (1:1, pH 7.2) was used unless
otherwise stated. Cysteamine (CEA) solution (10 mM) was firstly supplied into
the channel (5 μL/min for 15 min), then the flow was stopped for 15 min. After
washing away excess CEA, glutaldehyde (GA) solution (5% v/v) was supplied
into the channel (2 μL min for 20 min) and subsequently washed. Then, different
concentrations of DA solutions were pumped into the channel (2 μL/min for 20
min). After removal of excess DA and subsequent incubation with sodium
cyanoborohydride (NaBH3CN, 1 mM, 2 μL/min for 15 min), the microchip was
washed by PBS buffer (5 μL/min for 5 min) and then directly used for Raman
measurements.
In addition, the surface modification was proved by fluorescence microscopy.
Generally, a lissamine rhodamine B sulfonylchloride solution (100 μM, 5
μL/min for 10 min) was used after CEA incubation. After removal of excess
lissamine rhodamine B sulfonyl chloride by the solvent mixture (5 μL/min, 15
min), fluorescence images were taken upon excitation with green light. On
another chip, 6-aminofluorescein (100 μM) and NaBH3CN (1 mM) solutions
4.2 In situ synthesis, immobilization and functionalization of Au-TTF microwires
41
were used successively after supplying GA. Samples without the GA treatment
were exposed to 6-aminofluorescein and NaBH3CN solutions as a negative
control. Blue light and appropriate optical filters were used for taking
fluorescent images.
DL-Norepinephrine hydrochloride (NE), L-Epinephrine (EPI), Isoprenaline
hydrochloride (ISP), L-Phenylalanine (Phe) and L-Tyrosine (Tyr) were
dissolved in the solvent mixture to a concentration of 100 μM separately. The
functionalization procedures are the same as those applied for DA.
4.2.3 Functionalization protocol for immunosensing applications
A dimethylsulfoxide (DMSO)/PBS 2:3 solvent mixture (PBS final
concentration 10 mM) was prepared and the pH value was adjusted to 7.2. This
solution was used as a solvent unless otherwise stated. Antigens (human IgG
protein) was dissolved in the solvent mixture and diluted into 5 nM, 10 nM, 30
nM, 50 nM and 70 nM. 4-aminothiophenol (4-ATPh) (15 mM in CH3CN) was
supplied into the channel at a flow rate of 5 μL/min for 15 min, then stopped the
flow and kept reacting for another 15 min. At the same time, anti-human IgG
antibody (20 μg/mL in 10 mM PBS buffer, pH 7.4) was incubated inside a 1 mL
tube with 1-Ethyl-3-(3-dimethyl- aminopropyl) carbodiimide (EDC, 100 mM
in PBS) and N-Hydroxysuccinimide (NHS, 25 mM in DMSO) for 1 h. After
washing off non-reacted 4-ATPh with CH3CN (10 μL/min for 10 min) and the
solvent mixture (5 μL/min for 10 min), structures inside donuts were incubated
with the antibody solution for 30 min by pipette diffusion. Then, solvent mixture
was used to wash non-bonded antibodies (5 μL/min for 10 min). Followed by
incubation with CEA (10 mM) as a blocking reagent, different concentrations
of antigen (human IgG protein) solutions were supplied into the channel by
pipette diffusion and incubated for 20 min. After washing away the excess
antigen by PBS buffer (5 μL/min for 10 min), the chip was applied for in situ
Raman measurements.
As in the case of the biosensor, fluorescent microscopy served to prove the
surface modifications. Lissamine rhodamine B (100 μM) was used after the
incubation of 4-ATPh. Besides, FITC-labelled human IgG antibody (20 μg/mL
in PBS buffer, pH 7.4) was first incubated inside a 1 mL tube with EDC (100
mM, in PBS) and NHS (25 mM in DMSO) for 1 h, then supplying into the
micro-channel by pipette diffusion for 30 min. After washing away surplus
4 Methods-3: Sample preparation and characterization
42
FITC-labelled human IgG antibody with PBS buffer (5 μL/min for 10 min),
fluorescent images were taken under blue light excitation.
4.3 Synthesis of TTF crystal and Au-TTF nano-/microwires
This part shows the experimental procedure related to the results discussed in
Chapter 5.2.
Formation of TTF crystal. Single TTF crystals were grown by recrystallization
of commercially obtained TTF samples which were dissolved in hexane
(saturated solution).130,131 TTF crystals formed bright orange needles, up to 2 cm
long and about 0.25 mm2 in cross section (Figure 4.1a). A single TTF crystal
was selected under microscope for polarized Raman spectroscopy measurement
at room temperature.
Formation of Au-TTF nano-/microwires. Different diameters of Au-TTF wires
was synthesized on four-inlet microchips by controlling the flow rates of
reactants27 (see Figures 4.1b and 4.1c).
Figure 4.1. Images of TTF crystal and Au-TTF nano-/microwires. a) A picture of an
orange TTF crystal (on a 24 × 40 mm glass slide). Images of Au-TTF wires under Raman
microscope b) nanometre scaled, average diameter: 780 nm; c) micrometre scaled,
average diameter: 4 Pm. Various Au-TTF wires were synthesized on a four-inlet
microchip. Scale bars: 10 Pm.
4.4 Microfluidic synthesis of metal-TTF structures
43
4.4 Microfluidic synthesis of metal-TTF structures
The experimental procedures described in this chapter relate to the results in
Chapter 5.3.
General method: The reactions between transition metal salts with TTF and its
derivatives were performed on a single layer four-inlet microchip, where the
inner channels are used for the precursor supply, and the outer channels for
supply of pure solvent. The additional solvent streams focus the inner streams,
and are referred to as sheath flow. The width of the microchannels for the inlets
and main channel was 100 and 300 Pm, respectively. In some cases, a two-inlet
microchips was used as a comparison. Here, the inlet channels were 150 μm
wide and the main channel had a width of 300 μm. The height of all the channels
was 20 Pm. The total length of the main reactor channel was 1 cm. For the
microchannels of the four-inlet chip, a flow rate ratio of 10 was applied as for
the synthesis of Au-TTF.27 The two reactants were supplied through the middle
channels at 50 PL/min. Two side streams of solvents were supplied through the
side channels at 500 PL/min (Figure 4.2).
Figure 4.2. Photograph of microchips and optical image of microchannels. a) Photograph
of the single layer four-inlet microchip (top) and two-inlet microchip (bottom) filled with
green food dye (a one Swiss Franc is used for scale). b) Optical image of the
microchannels of the four-inlet microchip filled with different food dyes. Metal ion
solution in red, TTF (and its derivatives) solution in orange, side streams of solvents in
green. Scale bar: 100 Pm.
Solutions eluted from the microchips were collected and diluted 5 times with
solvent (final concentration, approx. 0.06 mM) for UV-Vis measurements. The
eluted structures were collected, filtered, washed by solvents and dried under
4 Methods-3: Sample preparation and characterization
44
vacuum for FT-IR spectroscopy. In addition, the eluted structures from the chips
were dropped on clean glass slides (Thermo Scientific, Menzel-Gläser, No. 3)
and dried under N2 gas. Then the glass slides with structures were gently rinsed
with deionized (DI) water for 30 s and dried under a slow stream of N2 gas
before Raman analysis, SEM imaging and EDX measurements. In SEM and
EDX experiments, the glass slide with samples were sputtered with Au/Pd (9/1)
(or carbon for samples with gold) first and mounted on a sample holder for
measurement.
In some cases, structures generated on a two-inlet chip was compared with those
formed on the four-inlet chip. Here, only the metal salts and TTF derivative
solutions were supplied at the flow rates of 100; 100 L/min.
a. Synthesis of (TTF)nCuX2 complexes: the two reactant solutions of CuX2 (X
= Cl-, NO3-) (6 mM in CH3CN) and TTF (24 mM in CH3CN) were supplied
from middle channels of the four-inlet chip, with the side channels used for
CH3CN. Structures were formed right after the injection of the solutions, elutes
from the chips showed dark purple precipitates. An analogue method was used
for the reactions in methanol (CH3OH) by changing the solvents into CH3OH.
b. Synthesis of (TTF)nFeX3 complexes: the solutions of FeX3 (X =Cl-, NO3-) (6
mM in CH3OH) and TTF (21 mM in CH3OH) were injected from middle
channels of the four-inlet chip, with side channels for CH3OH. Structures were
formed immediately, elutes from the outlet showed dark purple precipitates. An
analogue method was used for the reactions in CH3CN. As Fe(NO3)3 cannot
fully dissolve in CH3CN, a filtered solution was used for the reaction. So, here
Fe(NO3)3 had a concentration of approx. 6 mM.
c. Synthesis of Cu-FTTF complexes: similar synthesis method to the synthesis
of (TTF)nCuX2 complexes were used by changing the organic ligand to FTTF
(24 mM in CH3CN). Here, only the reaction in CH3CN was performed.
d. Synthesis of Fe-FTTF complexes: similar synthesis method to the synthesis
of (TTF)nFeX3 complexes were used by changing the organic ligand to FTTF
(24 mM in CH3OH). Here, only the reaction in CH3OH was performed.
e. Synthesis of M-BEDT-TTF: the solutions of HAuCl4, CuX2 (X =Cl-, NO3-),
FeX3 (X =Cl-, NO3-) (0.06 mM in CH3OH) and BEDT-TTF (0.24 mM in
tetrahydrofuran (THF)) were supplied into middle channel of the four-inlet chip,
the solvent side streams were both CH3OH.
4.5 In situ formation of Cu-TCNQ nano-/microstructures
45
4.5 In situ formation of Cu-TCNQ nano-/microstructures
The experimental procedures described here relate to the results discussed in
Chapter 5.4.
Solution preparation. Copper sulphate (CuSO4) and sodium citrate (C6H5O7Na3)
powders were dissolved in DI water separately to form solutions of
concentrations up to 0.5 M. (C6H5O7Na3 is used as an organic chelating agent
that interacts with CuSO4 and solubilizes Cu2+ ions present in solution, to form
a copper-chelator complex.) A 1 M NaBH4 solution was prepared using NaOH
solution (pH=12) as solvent. A high pH value was used to stabilize NaBH4.
TCNQ was dissolved in CH3CN to get the solutions of 2 mM and 5 mM. Other
concentrations of solutions were obtained by diluting the above-mentioned
solutions using their separate solvents. In addition, to avoid air in the experiment,
all the solutions were bubbled using Ar gas for 20 min before use.
Synthesis of Cu0 layer and Cu-TCNQ. As illustrated in Figure 4.3 the formation
of Cu0 and Cu-TCNQ was achieved by the following steps:
i) The middle control channel was pressed to 3 bar by N2 gas to close the
middle of the chamber, Cu(C6H5O7)24+ and NaBH4 solutions were
supplied from two inlets.
ii) The two side gas channels was pressured to 3 bar to stop supplying
solutions, then the middle gas channel was released to allow two
solutions to mix. This led to the formation of Cu0.
iii) DI water was supplied from the inlets. The left gas channel was
released to 2 bar to allow water slowly diffusing into the chamber and
wash Cu0.
iv) CH3CN was supplied (instead of water) from the left inlet to diffuse
into the chamber and wash Cu0.
v) TCNQ solution was supplied instead of water to diffuse into the
microchamber and react with Cu0. Different morphology of structures
were observed after the reaction (see Chapter 5.4.3 for details).
vi) The solvent CH3CN was introduced to wash off the TCNQ solution in
the channel and microchambers.
Finally, the microchip was put in vacuum and allowed the solvent to evaporate.
The resulted nano-/microstructures were obtained in the microchambers.
4 Methods-3: Sample preparation and characterization
46
Figure 4.3. Schematic illustration of Cu-TCNQ wire formation in microchamber. Solid
lines showed one of the sub-nanolitre microchambers, dashed lines showed the gas
control channels.
The operation of gas control channels to enable the formation of Cu0 and Cu-
TCNQ is shown in Figure 4.4.
Figure 4.4. Optical images of a microchamber showing the operation of control layer. a)
4.6 Sample characterization
47
The N2 gas channel in the middle was pressured to 3 bar and totally close the chambers.
b) Magnification image of that in a) which showing clearly the closing of microchamber.
c) The image after releasing the middle gas channel and pressured the side gas channels
to 3 bar to stop flows and allow the reaction in microchamber. Notice the totally closed
chambers at the two side gas channels. d) The image after releasing a side gas channel
from 3 bar to 2 bar to allow slow diffusing of solutions into microchambers. Notice the
partially closed chamber under the gas channel with 2 bar N2 gas pressure. Scale bars:
50 m.
4.6 Sample characterization
4.6.1 Electrical characterization
TTF and TCNQ are organic semiconductors which can be used to form charge-
transfer complexes with transition metal ions. These hybrid products show
different electrical conductive behavior compared with that of pure organic
molecules. In this work, electrical properties of TTF and TCNQ-based metal-
organic charge-transfer complexes were measured by applying current/voltage
sweeping between source and drain electrodes, with the main focus of
understanding the non-linear electrical conductive response of these
compounds.5
a) Label-free sensing application in nanoelectronics
Nano- and micrometre-sized conductive wires, integrated into devices and
involving electronic transduction, could potentially be used as highly sensitive
gas sensors e.g. for volatile organic compounds (VOC) or could be enriched
with biological recognition capabilities via a proper functionalization and
integration of small biomolecules, antibodies, enzymes, etc. The advantages
over other sensing technologies such as electrochemical or fluorescent ones lie
firstly on the capability of delivering a response that is label-free using a simple
electronic readout setup, and secondly the scaled device at the nanometre level.
The sensing ability of Au-TTF wires was first proven by Cvetković and
coworkers after direct exposure of the wires to polar and non-polar gases.28 Au-
TTF was proved to efficiently work as electrical sensing elements with the
merits of an easy fabrication, a fast read-out, and nanometre-sized dimensions
of the sensing devices. Other TTF-based metal-organic compounds are expected
to have similar sensing properties, when formed using similar charge-transfer
4 Methods-3: Sample preparation and characterization
48
techniques. A challenge in label-free sensing of nanodevices lies in the
development of single wire devices, which enables the miniaturization of
sensing elements. Much effort has been done in this project to synthesize and
build single nanowire devices using TTF-based complexes. Detailed results and
discussion on these aspects can be found in Chapter 5.3.
b) Conductivity measurement
The I-V curve measurements were performed to understand the electrical
behaviour of metal-organic wire structures. For Au-TTF nano-/microstructures,
the single wire conductivity was addressed by Dittrich group.27 The source-drain
current was measured while sweeping the applied source-drain voltage in a two-
point configuration. The nonlinear I-V characteristics are an indication of a non-
perfect contact interfaces, including small barriers, as the wires simply lay on
the electrodes with no topside metallization and subsequent annealing step
performed. No gate effect was observed for the different devices, indicating the
metallic behaviour of the Au-TTF charge-transfer structure.
To assess the conductivity of the other metal-TTF (derivatives) nanostructures,
single wires were aligned on microelectrodes that were prefabricated on a glass
slide by a home-built micro-manipulator. The conductivities of the aligned
structures were obtained by carrying out four-point measurements. A current
was applied through the source electrode pair (source high and source low), and
the resulting potential on the sensing electrode pair (sense high and sense low)
was monitored using a Keithley 2612A system source meter (Figures 4.5a and
4.5b). The electrical resistance of a single Au-TTF wire was determined by
using the four-probe configuration program of Keithley 3706 system
switch/multimeter (Figures 4.5c and 4.5d).
For Cu-TCNQ nano-/microstructures, the microchip was dried in vacuum after
the formation of the nanostructures. Then, a four-point measurement method
was applied. To investigate the electrical memory effect, a voltage sweep was
applied on the source electrodes.
4.6 Sample characterization
49
Figure 4.5. Photographs of conductivity measurement devices. a) Source meter for I-V
curve measurements; b) Four-point measurement configuration (see Figure 5.24 right
bottom for microelectrode configuration); c) Multimeter used for measurements of
electrical resistance; d) 8 pairs of micro-probes aligned and connected to the
microelectrodes on the glass slide.
4.6.2 Other characterization techniques
The UV-Vis spectra were recorded on a Genesys 10 UV scanning
spectrophotometer (scanning capability from 190-1100 nm, Thermo Scientific)
with disposable polystyrene (PS) semi-micro cuvettes (BRAND). Absorption
spectra were obtained on a UV-Vis spectrophotometer (V-650, JASCO, Japan).
IR spectra for metal-TTF and derivatives were obtained via a FT-IR Tensor 27
Infrared spectrophotometer (Bruker) equipped with a Bruker Golden Gate
diamond ATR (Attenuated Total Reflection) cell. The IR transmission
measurements of Cu-TCNQ were performed using TENSOR 37 (Bruker,
Germany) FT-IR spectrometer with a DTGS detector. The resolution of spectra
was 4 cm-1. Elemental analysis was performed by the micro-laboratory group at
the Laboratory of Organic Chemistry, ETH Zürich. SEM images and the energy
dispersive X-ray spectroscopy-scanning electron microscopy (EDX-SEM)
spectra were obtained using a FEI Quanta 200 FEG at the Scientific Centre for
Optical and Electron Microscopy (ScopeM), ETH Zürich. Coloured and
polarized images were recorded using a stereo microscope (AZ-100M, Nikon)
4 Methods-3: Sample preparation and characterization
50
equipped with a digital camera (Digital sight DS-Fi1, Nikon). Fluorescent and
other optical images were recorded on an inverted microscope (IX71, Olympus)
equipped with a digital camera (UK1117, ABS) and standard filters.
4.7 Methods for data analysis
4.7.1 Data analysis of fluorescent images, Raman, IR spectra and I-V
curves
Fluorescent images were recorded, false-coloured and background-corrected
using the Image J software. Raman spectra reported in Chapters 5.1 and 5.3 were
recorded using WiTec Project software and analysed with Origin Pro 9.1
software (Academic, OriginLab Corporation). The implemented peak analyser
tool of Origin Pro 9.1 served to get background-corrected spectra, peak position
and intensity values.
Raman spectra reported in Chapters 5.2 and 5.4 were recorded using Nspec
software (NanoScan Technology, GmbH). IR spectra were acquired using the
OPUS software (21 CFR Part 11 compliant). The Raman and IR spectra of
TCNQ and Cu-TCNQ were analysed with Origin Pro 9.1 (Academic, OriginLab
Corporation). Conductivity data were recorded by TSP Express Software Tool
and analysed with Origin Pro 9.1 software.
4.7.2 DFT calculations
In this work, the DFT calculations were used to obtain the exact values of Raman
tensor of TTF molecules (neutral and TTF cation). The calculations carried out
using the B3LYP level of theory supplemented with the standard 6-31+G (d,p)
basis set and the Gaussian 09 program package.132 The geometry optimization
and harmonic Raman frequencies of the different normal modes of neutral TTF
(C2v point group) and TTF cation (D2h point group) in gas phase were calculated
using DFT calculations. In order to get the Raman tensor element, the
quadrupole moment was taken into account using the polar keyword in DFT
calculations. The global minima on the potential energy surface were confirmed
by the real harmonic vibrational wavenumbers calculated for both neutral TTF
(C2v point group) and TTF cation (D2h point group). In the studies reported
recently, DFT calculations with B3LYP method133-137 provided useful results for
understanding the observed phenomenon (e.g. self-association and hydrogen
4.7 Methods for data analysis
51
bonding in molecules, molecular vibration). The results on DFT calculation in
this work are discussed in Chapter 5.2.
4.7.3 Single TTF crystal analysis
Crystal data were obtained from the Cambridge Structural Database (reference
code BDTOLE 10) and analysed by Mercury 3.6 software, which was
downloaded from the Cambridge Crystallographic Data Centre. Calculations of
molecular planes and angle measurements were carried out using the “Calculate
planes” and “Picking mode: Measure Angles” tools implemented in the software.
52
5.1 Label-free biosensors based on in situ formed and functionalized microwires in
microfluidic devices
53
5 Results and discussion
The results of four subprojects are discussed in this chapter. Chapter 5.1
describes the label-free biosensing application of in situ synthesized and
functionalized Au-TTF microwires in a microchip. Chapter 5.2 illustrates the
study on the molecular structure of TTF crystal and Au-TTF nano-/microwires
by polarized confocal Raman spectroscopy. Chapter 5.3 shows the various
products formed from different transition metal salts with TTF and its
derivatives. In the last part (Chapter 5.4), the investigation of TCNQ-based
metal-organic complexes is described.
5.1 Label-free biosensors based on in situ formed and functionalized
microwires in microfluidic devices
Figure 5.1 Graphic abstract of the sensing mechanism for the in situ label-free sensor in
this work. An Au-TTF wire was formed and trapped by the open-donut feature, then two
different surface modification procedures were done for two kinds of biosensors. The
upper one is a surface modified Au-TTF with the ability to bind analytes with amino
group e.g. dopamine. The lower one is a surface modified Au-TTF with antibody to
detect respective antigens, e.g. human immunoglobulin G (IgG). Both sensing elements
5 Results and discussion
54
achieved quantitatively measurement of biomolecules. Results on this work has been
published in the peer-reviewed journal Analyst (2015, 140, 7896-7901. RSC publishing).
Label-free sensing of small molecules and proteins is of high interest for
chemical gas and pH sensing, medical diagnostics, and pharmaceutical and
biological applications. Novel tools for biosensing have been developed over
the last years. Among them 1D nanostructures, i.e. fibres, wires and tubes have
shown to be very promising and to work as sensing elements in an operating
device. Although experimentally difficult to realize, it is indeed an intriguing
outlook that only a single nanostructure is needed for the building of
extraordinary small and portable instruments. In particular, when combined with
microfluidic techniques, 1D nanostructure sensors can reveal their full potential
for the detection of biological processes and multiplexing for parallel analyses
of different target compounds. Microfluidic devices allow the formation,
alignment, positioning and immobilization of nano- and micrometresized fibres,
while at the same time the fluidic channels can be used as delivery systems to
supply analytes to the sensing structures.138 One of the intriguing prospects is
that the integrated synthesis of the 1D structures and the final use in sensing
applications on a single microfluidic device, rendering any nano-/microstructure
manual handling or micromanipulation unnecessary.28,139,140
In the past years, various materials have been utilized to form 1D structures for
sensing purposes. For example, crystalline metal oxide nanowires like ZnO or
SnO2 are formed141 and used as capable building elements for conduct metric
gas sensors. More difficult is the creation of a sensor for biomolecules which
are dissolved in aqueous solution because special care is required for
immobilization of the wires. Si-nanowire, which can be prepared as p- or n-type
material and configured as field effect transistor based sensor device has been
utilized to detect biomolecules such as DNA and proteins.142 In this context,
semiconducting SWNTs should be also mentioned as they also have been
successfully implemented into a device for sensing (bio)molecules in
solution.143,144 Alternatively, hybrid systems created from metal salts and
organic compounds have attracted a lot of attention, because of their tunable
properties and applications in building nanodevices.145 Au-TTF is a conductive
metal-organic hybrid structure and can be formed in a controlled way by using
microfluidic-based technology.27 Dittrich group reported recently on the use of
the wires for sensing of water based on conductivity,28 and the sensing of
vapours of polar organic solvents was also observed. In addition, the same group
also showed the enhancement of the TTF Raman signal on Au-TTF wires.51
5.1 Label-free biosensors based on in situ formed and functionalized microwires in
microfluidic devices
55
Based on these results, the possibility to use Au-TTF wires as sensing elements
was proposed, with conductivity measurements for fast and online readout, and
Raman spectroscopy for further identification of the detected molecules.
However, to broaden the use of the wires for detection of a wide range of
analytes, several challenges have to be solved including a method to position
and functionalize the wires.
In this project, the realization of on-chip functionalization of single or few Au-
TTF wires with different molecules146 that have the ability to interact with gold
are shown. These wires have diameters of a few hundred nanometres up to a
few micrometres. To demonstrate that the integrated system is versatile for
biological analysis, selective label-free sensing of catecholamines or human IgG
by applying different modification approaches (Figure 5.1) are shown. The first
one enables the direct sensing of catecholamines by confocal Raman
spectroscopy and is based on wire functionalization by CEA and GA. The
second one enables indirect sensing147 by using the Raman tag 4-ATPh.148 Here,
an immunoassay is employed, where the antibody is bound to 4-ATPh and
captures the target molecules (human IgG). To the best of knowledge, this is the
first attempt to demonstrate the covalent binding of molecules on Au-TTF wire.
In addition, the experimental method presented here can be proposed as a
general means for the functionalization of 1D nano- and microstructures by
single- or multi-step reactions.
5.1.1 Microchip and synthesis of Au-TTF microwires
The microdevice used is based on the previously reported microchip design of
Dittrich group.29 It consists of two layers made of PDMS, a fluid layer and a
control layer that are separated by a 20 μm thin and flexible PDMS membrane.
The control layer has several round shaped, donut-like features that have
different diameters (50-250 μm) with an opening of 25 μm (Figures 5.2a and
5.2b).The two layers were first aligned and then bonded to a clean glass slide to
form the final device (Figure 5.2c). The multi-layered microfluidic chip
combines the ability to synthesize structures inside the fluid layer channel and
to immobilize them by pressurizing the control layer to 3 bar with N2 via the gas
inlets. There are two advantages in using this design, firstly, the structures
trapped inside the donuts are not exposed to the shear forces of the fluid flow,
and secondly, fluids can diffuse into the donut traps through the openings, which
enables the reaction between structures and solutions. In addition, extra fluids
5 Results and discussion
56
can also be easily washed off solvents. As shown in Figure 5.2d, the fluid
channel was filled with orange food dye to visualize the above-mentioned
features of this microchip design. After pressurized the control layer, the food
dye was flushed off by PBS buffer solution at a flow rate of 5 μL/min in two
minutes, which proved the efficiency in using this chip design for fluid change.
Figure 5.2. The open-donut microchip and its features. a) A photograph of the
microdevice with two layers, the control layer filled with dark red dyes and the fluid
layer filled with orange dyes. A one Swiss Franc was used for scale. b) A micrograph of
the fluid channel and open-donut features. Scale bar: 300 μm. c) Schematic of the
fabrication of the multilayer chip. d) The optical pictures of a bonded chip filled with
food dye after washing with PBS buffer at 5 μL/min in 110 s. Scale bars: 300 μm.
In the synthesis of Au-TTF wires, dissolved TTF in CH
3
CN (24 mM) was
added into the device and afterwards, HAuCl
4
(6 mM in CH
3
CN) was
introduced into the channel through the inlet reservoirs. The slow diffusion
enabled the electron transfer between the two reactants and the direct
formation of Au-TTF wires in the microchannel, which were easily
monitored on an optical microscope (Figures 5.3 and 5.4a). By applying N
2
through gas inlets, the wires formed inside the pneumatic cages could be
5.1 Label-free biosensors based on in situ formed and functionalized microwires in
microfluidic devices
57
trapped easily for further functionalization (Figure 5.4b). The morphology
and composition of Au-TTF wires formed with this diffusion technique were
demonstrated by SEM (Figure 5.4c) and EDX-SEM (Figures 5.4d and 5.4e).
The EDX spectrum in Figure 5.4e clearly indicates the presence of Au and
S in the Au-TTF wire, compared to that of the glass slide (Figure 5.4d). The
formed wires were about 2 μm in diameter and varied in length from tens to
several hundred micrometres. Thus, different traps in the control layer were
used to immobilize different lengths of wires. A similar chip with eight small
diameter traps (50-100 μm) was also used in our experiment. In the case of
a diffusion formation of Au-TTF microwires, as observed in this work, an
increased number of traps with smaller diameter can effectively trap more
wires (Figure 5.5).
Figure 5.3. Images of Au-TTF wire formation. Time sequence of micrographs under
optical microscope during the diffusion-based formation of Au-TTF microwires. Scale
bars: 50 μm.
5 Results and discussion
58
Figure 5.4. Characterization of Au-TTF wires. a) Optical microscope images of Au-TTF
wires formed inside a chip by diffusion. (Scale bar: 100 μm). b) A wire formed inside
the open-donut (Scale bar: 100 μm). c) SEM image of Au-TTF wires synthesized by
solution diffusion on a non-bonded chip. In this case the long wire is approx. 2 μm in
width and 110 μm in length. Scale bar: 10 μm. EDX spectra of d) the glass slide and e)
an Au-TTF wire.
Figure 5.5. A second microchip design for wire formation. a) A micrograph of the micro-
channel and 8 open-donuts with diameters of 50 μm and 100 μm in an alternated way.
5.1 Label-free biosensors based on in situ formed and functionalized microwires in
microfluidic devices
59
Scale bar: 300 μm. b) An optical image showing the bundles of wires trapped inside the
microchips. Scale bar: 100 μm.
5.1.2 Label-free biosensing of catecholamines
Due to the presence of Au
0
on Au-TTF wires, further modification can be
achieved by using molecules that have the ability to interact with gold. The
surface functionalization protocols applied for sensing aims are reported
here. Firstly, bioamines e.g. catecholamines, which play vital role in human
body and bioprocess.
149
To investigate the possibility in using Au-TTF wires
as sensing elements for label-free detection of catecholamines, a surface
modification of the wires was applied and demonstrated.
a) Functionalization of Au-TTF wires for biosensing
The functionalization of Au-TTF as biosensors for catecholamines is
indicated in Scheme 5.1 (Step 1-Step 3). First of all, CEA was used to form
a self-assembled monolayer (SAM) on wire surface by Au-S bond (Step 1),
then GA was bond to CEA by the reaction between carboxylic and amine
group to form a Schiff base (Step 2).
150
Afterwards, the aldehyde terminal
group of GA was exploited for the detection of biomolecules with amine
group and Raman activity (Step 3, 1)). Finally, NaCNBH
3
(Step 3, 2)) was
used to reduce Schiff bases to more stable second amine.
151
Using this
method, catecholamines could be covalently bonded to the wire surface and
directly monitored by their Raman signals by a label-free fast readout. The
selectivity of this
method depends on the analytes themselves which can be
differentiated by their own typical Raman peaks.
5 Results and discussion
60
Scheme 5.1. Functionalization on Au-TTF wire for the sensing of amines (RNH2).
Step 1: Formation of SAM of CEA on Au-TTF wire surface; Step 2: Binding of GA
to amino group; Step 3: 1) Binding of bioanalytes with amino group RNH2 and 2)
following treatment with NaCNBH3. Reaction schemes for the binding of lissamine-
Rhodamine B to amino group (step 2’) and aminofluorescein to aldehyde group (step
3’). This protocol can be used for detecting other chemical or biomolecules with
amino group, the detection of bioamines (e.g. catecholomines) is the focus of this
work.
To confirm the formation of the SAM of CEA and the following GA
bonding, fluorescent probes that reacted with the terminal groups of the
substrates were added after each step. As shown in Scheme 5.1 the reaction
mechanisms (Step 2’), the amino-active fluorescent dye Lissamine
Rhodamine B sulfonylchloride was utilized after Step 1. The binding of
fluorescent dye molecules to amino group exhibits its fluorescence, and
proves that CEA formed a SAM on the wire surface. Similarly, 6-
aminofluorescein which is aldehyde-active was used after Step 2. The
successful wire modification with GA can be demonstrated by the
fluorescence from aminofluorescein (Step 3’).
5.1 Label-free biosensors based on in situ formed and functionalized microwires in
microfluidic devices
61
Based on this method, the wires were firstly trapped on different chips
(Figures 5.6a and 5.6b). Dark images were taken before supplying
fluorescent dyes into microchips, even under laser excitation (Figures 5.6c
and 5.6d). However, after using fluorophores separately, the fluorescent
wires (Figure 5.6e) indicated the reaction between the amino groups and
Lissamine Rhodamine B sulfonylchloride, which proved the formation of
SAM of CEA on the wire surface. Similarly, the fluorescent wires in Figure
5.6f confirmed the reaction between the aldehyde group and 6-
aminofluorescein, and demonstrated the successful binding of GA. In
addition, to explain the important role of GA, a control experiment without
GA was performed and showed no fluorescence even after the incubation
with 6-aminofluorescein (Figure 5.7).
Figure 5.6.
Fluorescent images of wires after surface modification for biosensors.
a)
and b) Bright field images of nanowires immobilized by different donuts. c) and d)
Fluorescent images after incubation with CEA (10 mM) separately, before supply of the
fluorophores. Scale bars: 100 μm. False-coloured fluorescent images of trapped wires
after reaction with e) 10 mM CEA, 100 μM Lissamine Rhodamine B sulfonylchloride
(green laser excitation) and f) 10 mM CEA, 5% GA, 100 μM 6-aminofluorescein and 1
mM NaCNBH3 (blue laser excitation) separately. Scale bars: 100 μm.
5 Results and discussion
62
Figure 5.7. Control experiment without GA solution. a) Bright field image of nanowire
immobilized by donut. b) False-coloured fluorescent image after incubation with CEA
(10 mM) and 6-aminofluorescein (100 μM). c) False-coloured fluorescent image after
supplying of NaCNBH3 (1 mM) and washing with PBS. Blue light excitation was used
for fluorescent images. Scale bars: 100 μm.
b) Direct sensing of catecholamines
In this part, direct sensing means that Raman signals come from the analytes.
After demonstrating the feasibility of the first protocol, catecholamines
including the well-known neurotransmitter DA,152 as well as NE, EPI and ISP
(molecular structures in Appendix) were detected separately by the above-
mentioned direct sensing protocol. Micro-Raman spectroscopy was performed
to get signals from every single wire. It has been reported that Au-TTF wire has
a typical Raman peak at around 1426 cm-1 (C=C stretching).51 Compared to the
spectra of the single Au-TTF wire, two additional Raman peaks appeared at
1268 cm-1 (C-O stretching) and 1476 cm-1 (phenyl C=C stretching) after using
DA (Figure 5.8a), which clearly demonstrated the binding of DA to the modified
wire.153 Another important neurotransmitter is NE, which has a structure similar
to DA. In the spectrum of NE, typical peaks at 1268 cm-1 (C-O stretching) and
1479 cm-1 (phenyl C=C stretching) were observed. The spectrum is 0-3 cm-1
shifted with respect to the corresponding bands in DA and has a slightly lower
intensity (Figure 5.8a). The other two catecholamines, EPI and ISP, only
showed very weak Raman peaks (Figure 5.8a), which may due to their own
weak Raman intensities.154 To prove the selectivity of the system, aromatic
amino acids that are related to the formation of neurotransmitters in bioprocess,
Phe and Tyr (Figure 5.8b), were also measured using the same method applied
to catecholomines. However, Phe and Tyr exhibited totally different Raman
bands with respect to those of DA (Figure 5.8a). Thus, using the proposed
technique, on the basis of the observed Raman bands, catecholamines can be
distinguished from other bioamines in the wavenumber range of 1200 cm-1-1700
cm-1.
5.1 Label-free biosensors based on in situ formed and functionalized microwires in
microfluidic devices
63
Figure 5.8. Raman spectra of various bioamines on functionalized Au-TTF wires.
a) Comparison of the Raman spectra of single Au-TTF wire and after separate
reaction with 100 μM different catecholamines: NE (1268 cm
-1
, C-O stretching
and 1479 cm
-1
, phenyl C=C stretching), EPI (1273 cm
-1
, C-O stretching and 1483
cm
-1
, phenyl C=C stretching) and ISP (1270 cm
-1
, C-O stretching and 1484 cm
-1
,
phenyl C=C stretching). b) Raman spectra of Au-TTF wires after reaction with
100 μM aromatic amino acids including Phe (1004 cm
-1
, ring breathing; 1032 cm
-
1
, in plane ring bending and 1207cm
-1
, C
6
H
5
-C stretching) and Tyr (1179 cm
-1
, in
plane C-H bending and 1217 cm
-1
, benzene ring stretching) separately. In this
project, the functionalization of Au-TTF is on focus, thus, only typical Raman
peaks are shown here. For detailed Raman spectra analysis of Au-TTF, see the
results discussed in Chapter 5.2.
Moreover, a quantitative analysis was carried out using dopamine. Although DA
has a weak Raman signal,155 changes in the Raman intensities with an increasing
concentration ranging from 5 μM to 100 μM (Figure 5.9a) were observed. The
signal intensities at 1476 cm-1 (phenyl C=C stretching) was plotted against the
DA concentration and used for the quantification of DA. A baseline-corrected
peak intensities at 1476 cm-1 are shown in Figure 5.10. It is evident in Figure
5.9b, increasing Raman intensities were detected with the increment of
micromolar concentrations of DA. A linear response of the Raman intensities
and DA with respect to the concentration of 5 to 50 μM was obtained (Figure
5.9b inset). In addition, a detection limit of ~5 PM was achieved via this
approach, and resulted lower than a recently reported Raman-based method
(~100 PM).155
5 Results and discussion
64
Figure 5.9. Quantitative Raman measurement of DA based on the functionalized
wire for biosensor. a) Representative Raman spectra of the biosensor for
increasing concentrations of DA from 5 μM-100 μM. b) Raman intensity at around
1476 cm
-1
for different concentrations of DA after data correction. The inset figure
shows the linear relationship in the concentration of 5-50 μM (coefficient of
determination: R
2
=0.9968). Error bars indicate standard deviations from six
measurements.
Figure 5.10. Data analysis method of Raman spectrum. The Raman spectrum of
Au-TTF wire after reaction with 100 μM DA (black) and the modelled baseline by
Origin Pro 9.1 software (blue).
5.1 Label-free biosensors based on in situ formed and functionalized microwires in
microfluidic devices
65
5.1.3 Label-free immunosensors for detection of human IgG
a) Functionalization of Au-TTF wire for immunosensing purposes
Evaluation of such microfluidic integrated Au-TTF wire systems for sensitive
label-free immunoassay was done by using another surface modification
protocol and testing how human IgG as molecular targets.
In this approach, the
Raman reporter molecule, 4-ATPh served as a linker between the wire and
the capture antibody. Reported studies have proved that 4-ATPh acts as a
nano-mechanical stress sensor and can be employed as an indicator of the
antigen-antibody binding events.
148
As the Raman signal changes originated
from the Raman reporter 4-ATPh while not the analysed antigens
themselves, this method is also known as an indirect sensing method. The
sensing mechanism is based on the deformation of the stretching mode of
the benzene ring of 4-ATPh upon binding of antibodies and antigens. This
can result in a change of Raman frequencies of the Raman reporter
molecule.
148,156
Thus, Raman shifts of 4-ATPh correspond to different
concentrations of antigens binding to antibodies. Scheme 5.2 shows the
procedure for the functionalization of the wire. Firstly, the SAM of 4-ATPh
was formed on the wire surface (Scheme 5.2, Step 1). Immobilization of the
capturing antibody (anti-human IgG) was realized in two steps: firstly,
activating the carboxyl terminal of the antibody with EDC/NHS, a wide used
method for activating the carboxyl group,
157
and secondly by the reaction of
the antibodies with the amino group of 4-ATPh to form an amide bond
(Scheme 5.2, Step 2). To improve the specific binding of antigens, CEA was
used as the blocking reagent (Scheme 5.2, Step 3). Finally, human IgG was
applied as antigens (Scheme 5.2, Step 4).
5 Results and discussion
66
Scheme 5.2. Functionalization of Au-TTF wire for immunoassay. Step 1: SAM of 4-
ATPh; Step 2: Immobilization of the capturing antibody (anti-human IgG) by using
EDC/NHS. Step 3: Blocking unreacted wire surface by CEA and Step 4: Antigen (human
IgG) binding.
As in the label-free sensing of catecholamines (Figure 5.6), fluorescently
tagged compounds were used to test the surface modification. Lissamine
Rhodamine B was used to prove the binding of 4-ATPh to the wire (Scheme
5.3, Step 2a), while the covalently bond of
FITC-labelled anti-human IgG
antibody to show the attachment of other antibodies
(Scheme 5.3, Step 2b).
After the immobilization and functionalization of the wires
separately
(Figures 5.11a and 5.11b
), the dark images were obtained before fluorophores
were supplied
(Figures 5.11c and 5.11d
). Then fluorescent images were
recorded after introducing fluorophores, which clearly demonstrated the
reaction between Lissamine Rhodamine B and the SAM of 4-ATPh (Figure
5.11e), as well as the successful binding of pre-treated FITC-labelled anti-
human IgG antibody to 4-ATPh (Figure 5.11f).
5.1 Label-free biosensors based on in situ formed and functionalized microwires in
microfluidic devices
67
Scheme 5.3. Reaction mechanisms for binding of fluorophores to the SAM of 4-ATPh:
for lissamine Rhodamine B (Step 2a), and for the binding of FITC-labelled antibody after
pre-treatment with EDC/NHS (Step 2b).
Figure 5.11.
Fluorescent images of wires after the surface modification as
immunosensor.
a) and b) Bright field images of the trapped wires inside the donuts, c)
5 Results and discussion
68
Fluorescent image after reaction with 4-ATPh (15 mM) but without Lissamine
Rhodamine B sulfonylchloride (green light excitation), d) Fluorescent image after
reaction with 4-ATPh (15 mM), but without FITC-labelled human IgG (blue light
excitation). Fluorescent images of trapped wires after reaction with e) 15 mM 4-ATPh
and 100 μM Lissamine rhodamine B sulfonylchloride; and f) 15 mM 4-ATPh and 20
μg/mL FITC-labelled human IgG antibody (pre-treated by EDC/NHS). Scale bars: 100
μm.
b) Indirect Sensing of human IgG
In this part, indirect sensing means that the Raman signals originates not from
the analyte but a Raman reporter molecule. On the basis of the proposed
immune-sensor approach, Raman spectroscopy was applied to the label-free
indirect immunoassay of human IgG. Figure 5.12a shows the Raman spectra of
pure solid 4-ATPh, Au-TTF and the modified Au-TTF/4-ATPh. Pure 4-ATPh
crystal showed typical bands at 390 cm-1 (CS bending), 470 cm-1 (CCC
bending), 1092 cm-1 (CS stretching) and 1597 cm-1 (benzene ring stretching)
(curve i), all of the bands belonging to a1 modes.158,159 Au-TTF wire showed its
characteristic Raman peaks at 506 cm-1 (in plane ring bending) and 1426 cm-1
(C=C stretching) (curve ii). Compared to the Raman spectrum of pure 4-ATPh
(curve i), the one from the Au-TTF modified with 4-ATPh (curve iii) showed
differences such as the absence of the peak at 470 cm-1 and the peak shift from
1092 cm-1 to 1080 cm-1, as well as the shift from 1597 cm-1 to 1581cm-1. This
indicated the formation of the SAM of 4-ATPh on Au-TTF wire.158 Spectra of
the antibody-conjugated wire (Au-TTF/4-ATPh/anti-human IgG) incubated
with different concentrations of antigen (0 nM to 70 nM) are shown in Figure
5.12b. The typical Raman peak around 1580 cm-1 (which corresponds to the
benzene ring stretching of 4-ATPh molecule) was responsive to the supplying
of antigens after their selective binding to antibodies.156 As shown in the
magnified spectra in Figure 5.12c, the Raman frequencies of the sensor shifted
to higher wavenumbers for about 5.7 cm-1 with the increasing concentration of
human IgG ranging from 0 to 70 nM, in agreement with the previously reported
works.148,156 Figure 5.12d showed the calibration curve of the immunoassay with
a plateau at larger concentrations. This indicated the saturation of the antigen-
antibody complex. A good linear correlation of Raman shift with antigen
concentrations was shown in the range below 50 nM (Figure 5.12d, inset). This
approach demonstrated its high sensitivity and a low detection range (5 nM).
In these experiment, due to the larger diameter of the wires (approx. 2 μm,
Figure 5.4c) than that of the confocal Raman laser point (approx. 400 nm), the
5.1 Label-free biosensors based on in situ formed and functionalized microwires in
microfluidic devices
69
Raman signals were collected from the single wires. In addition, the reasonable
small error bars of Raman signals obtained from different structures indicated
the reproducibility of the two protocols proposed in this project (Figures 5.8b
and 5.12d).
Figure 5.12.
Raman measurement data of functionalized wire for immunosensor.
a)
Raman spectra of pure 4-ATPh (curve i), Au-TTF wire (curve ii) and Au-TTF/4-ATPh
(curve iii). b) Representative Raman spectra of the immunosensor in the presence of
increasing concentrations of human IgG from 0 nM to 70 nM. c) Magnification of Raman
peaks at around 1580 cm-1 (benzene ring stretching) of those in b). d) Corresponding
Raman shifts to different antigen concentrations. The inset figure shows the linear
relationship within the concentration range of 0 - 50 nM (coefficient of determination,
R2 = 0.9855). Error bars indicate standard deviations from six measurements.
5 Results and discussion
70
After proving the functionalization and sensing application of Au-TTF wires,
the growth mechanism of Au-TTF wire is of great interest since Au-TTF can
form wire structures with a different diameter. Thus, the molecular assembly in
a single wire structure needs to be further addressed. In the following chapter,
the structural analysis of Au-TTF wires via non-destructive polarized Raman
spectroscopy is discussed in detail.
5.2 Study of molecular self-organization in TTF crystals and individual Au-TTF nano-
/microwires by polarized confocal Raman spectroscopy
71
5.2 Study of molecular self-organization in TTF crystals and
individual Au-TTF nano-/microwires by polarized confocal Raman
spectroscopy
Figure 5.13. Graphic abstract of the polarization Raman measurement of TTF crystal and
Au-TTF wire. The molecular ordering of TTF in a single crystal is studied by rotational
polarization Raman spectroscopy. The same analysis was applied to a single Au-TTF
nanowire. The tilted stacking of TTF molecules along the wire growth direction was
demonstrated as shown in the 3D model.
In this work, the molecular orientation in a single TTF crystal and an Au-TTF
nanowire was firstly studied via polarized confocal Raman spectroscopy, by
taking into account the rotational polarization dependence of the molecular
vibrational modes (Figure 5.13). To determine molecular ordering in a
crystalline assembly, the Euler’s angles of the molecules need to be determined
using polarized Raman spectroscopy and known Raman tensors of TTF
molecule. Upon rotation of incoming laser on different faces of the crystal, the
5 Results and discussion
72
Raman intensity of molecular vibrational modes changed with a certain
periodicity, suggesting a preferential molecular orientation in the crystal. Based
on the DFT calculations of the Raman tensor and the simulation of
depolarization ratio, the molecular assembly in a single TTF crystal was
confirmed, in accordance with the single crystal X-ray diffraction data. Thus,
the efficiency of the proposed method to study TTF-based materials was proved.
Afterwards, the same technique was applied to investigate the self-organization
of TTF molecules and TTF cations in single Au-TTF charge-transfer nanowires.
Polarization-dependent spectra of single Au-TTF nano-/microwire clearly
proved that TTF molecules are assembled in a certain orientation in the wire
structure. The stacking way of TTF units inside a single wire was modelled
based on the simulated depolarization ratio data. Finally, the tilt stacking of TTF
units along the wire growth direction was obtained.
5.2.1 DFT calculations of Raman tensor
Optimized molecular structures of TTF and TTF+ are shown in Figure 5.14. The
optimization of TTF molecule forms the non-planar structures in such a way
that the symmetry of TTF drops to C2v point group. The optimization of TTF+
radical shows that TTF+ forms a planar structure having D2h point group of
symmetry. It is more likely that the torsional changes are produced by the four
sulphur atoms presents in TTF molecule. The DFT calculation results showed
that neutral TTF has a boat-like structure with C2v symmetry, while the radical
cation has a planar D2h symmetry (Figure 5.14), which is both in accordance
with reported work.160
Figure 5.14. DFT modelling of TTF and TTF cation.
5.2 Study of molecular self-organization in TTF crystals and individual Au-TTF nano-
/microwires by polarized confocal Raman spectroscopy
73
Molecular vibrations can be described by the sum of normal internal vibrations
of atoms. It is possible to classify the normal vibrations by applying the group
theory.161 Each normal vibration will have a symmetry corresponding to one of
the irreducible representations of the molecular point group. If a molecule has a
centre of inversion, the vibrations that are centrosymmetric show Raman active
properties, which are usually labelled with the subscript g. The symmetry of a
given vibrational mode is reflected in its corresponding Raman tensor.112 The
Raman tensors for C2v (TTF neutral) and D2h (TTF cation) are show in Table
5.1, the two point groups are isomeric. The Raman tensor values calculated for
both TTF and TTF+ are depicted in Table 5.2 (detailed calculation method see
Chapter 4.7.2).
Table 5.1. Raman tensors for the symmetry classes of the C2v and D2h point groups*
Symmetry class
A1g A2g B1g B2g
Symmetry class
A1g B1g B2g B3g
* Raman tensors of point group are from Bilbao Crystallographic Server.
Table 5.2. Calculated Raman tensor values
Molecule
Point
group
Traceless Quadrupole moment (field-independent basis, Debye-Ang)*
XX (a)
YY (b)
ZZ (c)
XY (d)
XZ (e)
YZ (f)
TTF
C2v
-5.0235
16.8375
-11.8140
0.0000
0.0000
0.0000
TTF+
D2h
-24.3185
-7.8936
32.2121
0.0000
0.0000
0.0000
*XYZ coordinates of the molecules refers to Figure 5.14.
00
00
000
00
000
00
000
00
00
00
00
00
f
f
e
e
d
d
c
b
a
00
00
000
00
000
00
00
00
00
00
00
00
f
f
e
e
d
d
d
c
b
a
5 Results and discussion
74
5.2.2 Polarized confocal Raman spectra of single TTF crystals
a) X-ray crystal structure of single TTF crystals
TTF appears in two polymorphs, one is the orange crystal with monoclinic form
which is named α-TTF, the other one is the yellow crystal with triclinic form,
known as
E
-TTF. The single crystal structure of α-TTF has been extensively
studied at room temperature by Cooper et al. in early 1970s 162,163 It has been
reported that at low temperature (about 190 K), the monoclinic α-TTF crystal
undergoes a reversible second-order phase transition to
E
-TTF.130 This project
is focused on the study of the orange monoclinic TTF crystal at room
temperature. TTF (refers to α-TTF, so as in the following text) shows a based-
centred monoclinic lattice, with two molecules per unit cell, as can be seen from
X-ray diffraction data (Figure 5.15).6,164 The view from ab face and bc face
indicated that there is an angle between the two molecule planes (49.45o) (Figure
5.15).
Figure 5.15. The view from three faces of crystalline TTF and the calculated angle
between two TTF molecules in a unit cell of TTF crystal. a) b) c) Views of TTF crystal
(Cambridge Structural Database, reference code: BDTOLE10) from three faces of the
unit cell (ab, bc, ac separately) in Mercury 3.6 software. Angle between molecular planes
were calculated to be 49.45o. The structure was reported in Reference 163.
5.2 Study of molecular self-organization in TTF crystals and individual Au-TTF nano-
/microwires by polarized confocal Raman spectroscopy
75
Following, in order to prove the efficiency in using our method to study TTF-
based structures, the polarized Raman measurement on TTF crystal and
simulation of depolarization ratio were carried out. Afterwards, the calculated
angle between two molecules in TTF crystal will be compared to the data
obtained from the reported work.
b) Raman spectrum of single TTF crystals
The TTF crystal structure with the definition of abc faces and xyz configuration
for incoming laser is shown in Figure 5.16a. The Raman analysis was carried
out at different faces with the laser beam in the backscattering configuration.
The polarization of the laser light was selected by using a polarizer under
parallel and perpendicular configurations. A full spectrum of TTF crystal from
bc face at room temperature showed the assignments of typical Raman peaks
(Figure 5.16b), with the peaks at 315 cm-1, 471 cm-1, 1089 cm-1 and 1516 cm-1
belonging to Ag vibration, while the peak at 58 cm-1 belonging to Bg
vibration.6,165 In this work, only the Ag modes were investigated at room
temperature. A detailed work in studying Bg mode at low temperature (80 K)
was reported in literature.6
Figure 5.16. Sketch of TTF crystal and a Raman spectrum. a) A sketch of TTF crystal
with the definition of abc faces and xyz configuration for incoming laser. b) A full
spectrum of TTF crystal from bc face.
5 Results and discussion
76
c) Polarized Raman spectra of ac, bc and ab faces of TTF crystal
With the aim to study the molecular orientation of TTF molecules in a single
crystal, polarization-dependent Raman spectra measurements were carried out.
The rotational polarization dependence of the molecular vibrational modes was
studied upon rotating the incoming laser polarization with steps of 15o. Two
polarized configurations, parallel and crossed, were used. As indicated in
Figure 5.17, the Porto notations (see Chapter 3.4.2) of three faces of TTF crystal,
x(zz)x (Figure 5.17a) and x(zy)x (Figure 5.17b) represent the parallel/crossed
polarization at ab face, while y(zz)y (Figure 5.17c) and y(zx)y (Figure 5.17d)
representing those for bc face, and z(xx)z (Figure 5.17e) and z(xy)z (Figure
5.17f) for ac face. Upon the rotation of incoming laser, the Raman intensity of
molecular Ag vibrational modes changed with a periodicity under both parallel
and crossed configurations, suggesting a preferential molecular orientation in
the crystal. Typically, the central C=C stretching Ag mode at ~1516 cm-1 can be
used for both qualitative and quantitative analysis. As seen from Figure 5.17,
the Raman intensities of Ag modes of the crystal varied with a period of 180o in
the parallel polarization (Figures 5.17a, 5.17c and 5.17e), while with a period of
90o in the crossed polarized configuration (Figures 5.17b, 5.17d and 5.17f). The
highest Raman signal in the parallel polarization was obtained when incoming
laser was perpendicular (rotation angle 90o) to the starting axis in the
corresponding faces.
5.2 Study of molecular self-organization in TTF crystals and individual Au-TTF nano-
/microwires by polarized confocal Raman spectroscopy
77
5 Results and discussion
78
Figure 5.17. Angular variations of Raman spectra from different faces of TTF crystal
under parallel and crossed polarization configurations. a) and b) Angular polarization
Raman spectra of ab face under parallel and crossed polarization respectively. c) and d)
5.2 Study of molecular self-organization in TTF crystals and individual Au-TTF nano-
/microwires by polarized confocal Raman spectroscopy
79
Those spectra of bc face. e) and f) Those spectra of ac faces. The abc axes of the crystal
and xyz configuration of incoming laser refer to Figure 5.16a, x, y, z indicate the
direction of backscattered light. For ab and bc faces, angle 0o is always defined along z
axis, while for ac face angle 0 is defined along x axis.
d) Simulation of depolarization ratio
To describe the translational movement of a molecule, three coordinates are
employed. Furthermore, if the molecule is non-linear, three additional
coordinates as the Euler’s angles (typically denoted as α, β, γ, or φ, θ, ψ) are
necessary to define the orientation of the molecule in space. When molecules
building a crystal, the total number of normal coordinates is related to the
number of molecules in the unit cell. For TTF crystal, there are two molecules
in the monoclinic unit cell (see Figure 5.15). Thus, a two molecule system is
adapted for the simulation of depolarization ratio.
In addition, it is noteworthy that the depolarization ratio is nearly zero, when
assuming the scattered light is totally polarized. This is an ideal situation.
However, taking the real experimental values into consideration, the intensities
of scattered light in parallel and perpendicular need to be expressed with a
constant factor (or depolarization error) D. Thus, the depolarization ratio P is
expressed as in Equation 5.1112:
𝑃 =(1−𝐷)𝐼𝑥𝑦+𝐷∗𝐼𝑥𝑥
𝐷∗𝐼𝑥𝑦+(1−𝐷)𝐼𝑥𝑥 (5.1)
In ideal situation, when D = 0, P can be described as P = Ixx /Ixy (Equation 3.7,
Chapter 3.1.3).
The constant factor D is ~0.12 when fitting for the three faces. It has been
reported that the depolarization ratio is useful in analysing the polarization
response of a substrate.161 The experiment and simulated data of depolarization
ratio of the 1516 cm-1 mode (C=C stretching) under angular variation of three
faces (ab, bc, ac) of the crystal are shown in Figure 5.18. Symbols represent the
experimental data. The blue line is the ideal depolarization ratio simulation (D
= 0) and the red line is the simulated depolarization ratio with a constant factor
D. A high correlation of experimental data with simulated results from three
different faces of the crystal was observed.
5 Results and discussion
80
Figure 5.18. Simulated depolarization ratio of TTF crystal. Experimental (symbols) and
simulated (lines) depolarization ratio of assignment at ~1516 cm-1 obtained upon rotation
with certain angles around the normal of three faces of TTF crystal. Blue line is the ideal
simulation, red line is the simulation with depolarization error. abc axes of the crystal as
in Figure 5.16a.
Based on the simulation of depolarization ratio, the angle between the two
molecules in the crystal were obtained. The angles simulated from different
faces were 49.4o (from ac face), 48o (from bc face), 49.2o (from ab face). All the
angles obtained from simulated results (Figure 5.18) were in the range of 49.45o
± 1.5o, indicating a good correlation of the simulated data with the experimental
data (49.45o, Figure 5.15). Thus, the proposed method was proven to be efficient
to model the molecular orientation of TTF-based materials.
5.2 Study of molecular self-organization in TTF crystals and individual Au-TTF nano-
/microwires by polarized confocal Raman spectroscopy
81
5.2.3 Polarized confocal Raman spectra of single Au-TTF nanowires
a) Polarized Raman spectra of Au-TTF nanowires
After proving the efficiency of the proposed method on TTF crystal, the same
method was applied to the analysis of single Au-TTF nano-/microwires. Firstly,
the spectra of a single Au-TTF wire under parallel and crossed polarization
configurations were obtained. The typical peaks for the partially charged TTF
shifted from those of neutral TTF molecules (Figure 5.16b), showing typical
Raman peaks at 504 cm-1 (in-plane vibration), 1011 cm-1 (C-H in-plane bend)
and 1415 cm-1 (C=C stretch) (Figure 5.19). A comparison of the Raman
assignments of TTF, TTFn+ (0<n<1) and TTF+ is shown in Table 5.3. In addition,
the different spectra from two parallel polarizations ሺሺሻ
തand ሺሻ
ത,
Figure 5.19) indicate that TTF molecules are assembled in a certain orientation
in the wire.
Figure 5.19. Raman spectra of single Au-TTF wire under parallel (black curves) and
crossed (grey curve) polarization configurations. The full spectrum of under ሺሻ
ത
configuration shows the typical assignments. Figure inset indicates the XYZ direction on
a single wire.
ത indicates the direction of backscattered light.
5 Results and discussion
82
Table 5.3. Raman assignments of TTF, TTFn+ (0<n<1) and TTF+
Vibrational modes
(Ag symmetry)*
TTF
TTFn+
(0<n<1)
TTF+ *
ʋ2 C=C stretch
1560 cm-1
1513 cm-1
1510 cm-1
ʋ2 C=C stretch
1548 cm-1
1484 cm-1
1481 cm-1
ʋ3 C=C stretch (ring-ring)
1516 cm-1
1415 cm-1
1416 cm-1
ʋ4 C-H in plane
1089 cm-1
1011 cm-1
1088 cm-1
ʋ5 C-S stretch
795 cm-1
753 cm-1
758 cm-1
ʋ6 In-plane ring bend
471 cm-1
504 cm-1
506 cm-1
* The symmetry species numbers, frequencies of TTF+ see Reference 10. Raman shifts
for TTF0 and TTFn+ (0<n<1) are the results from this work.
b) Angular polarization Raman spectra of Au-TTF nanowire
Following, the angular based polarization Raman spectra of Au-TTF were
obtained upon rotation of incoming laser every step of 15o. As shown in Figure
5.20, the Raman intensities of Ag modes (504 cm-1 and 1415 cm-1) varied with
a period of 180o in parallel polarization (Z(XX)Z
, Figure 5.20a), and 90o in
crossed polarization configuration (Z(XY)Z
, Figure 5.20b). The highest Raman
signal in parallel polarization was obtained when incoming laser was nearly
perpendicular (rotation angle 90o) to the long axis of the wire.
5.2 Study of molecular self-organization in TTF crystals and individual Au-TTF nano-
/microwires by polarized confocal Raman spectroscopy
83
Figure 5.20. Angular variation of Raman spectra of single Au-TTF nanowire under
parallel and crossed polarization configurations. The XYZ configurations refer to Figure
5.19,
ത indicate the direction of backscattered light of Z. The 00 angle is defined along
the X axis of wire.
c) Simulation of depolarization ratio
To study the depolarization ratio of Au-TTF nanowire, six different wires with
the diameter of ~700 nm were randomly chosen for polarized Raman
measurement. The simulation method for TTF crystals was applied to simulate
also Au-TTF nanowires. The simulated values of depolarization ratio of 1516
cm-1 peak under angular variation are shown in Figure 5.21. Symbols represent
5 Results and discussion
84
the experimental data. The blue line is the ideal depolarization ratio simulation
(D = 0), and the red line is the simulated depolarization ratio of that with a
constant factor D. Here, a D value of ~0.1 was applied for all the fitting curves.
Figure 5.21. Simulated depolarization ratio of six Au-TTF nanowires. Experimental
(symbols) and simulated (lines) depolarization ratio of assignment at ~1415 cm-1
obtained upon rotation with certain angles around X axis of Au-TTF wire (Figure 5.19).
The blue line is the ideal simulation, while the red line is the simulation with
depolarization constant factor.
5.2 Study of molecular self-organization in TTF crystals and individual Au-TTF nano-
/microwires by polarized confocal Raman spectroscopy
85
For the simulation of depolarization ratio of Au-TTF nanowires, a two-molecule
system was needed. This resulted in a good fitting with the experiment results.
Thus, this suggests two different molecular orientation exist in a single nanowire.
The angle between the two molecules was calculated as 126 ± 0.7o. With this
angle, all the simulated curves (Figure 5.21, red line) showed high correlation
with experimental data (Figure 5.21, black solid square). However, the angular
based depolarization ratio of the six nanowires showed different variations
(Figure 5.21, experimental data). Since all these data can be simulated using the
angle of 126 ± 0.7o, the difference in experimental data can be explained by the
rotation of the wires when they were randomly deposited on the glass slide for
measurement.
d) Molecular orientation of TTF0 and TTF+ in single nanowire
The 3D model was obtained from the simulation, with molecules rotated
according to the Euler angles. As indicated in Figure 5.22, the yellow cylinder
indicates Au-TTF nanowire with the two molecules inside the wire structure.
Blue XYZ coordinates refers to the wire coordinates (the same as that shown in
Figure 5.19 inset), where the blue X is the wire growth direction and the Red
XYZ is the molecule coordinates, which in accordance with the ones used in the
DFT calculations (Figure 5.14). A partially charged TTF molecule A is
positioned according to the Euler angle (α, β, γ) derived from the simulation,
then a second molecule B is defined with the angle of 126 ± 0.7o to the molecular
plane of A.
It can be seen from the 3D model that the two molecules are oriented along the
wire growth direction, with an obvious tilt angle with respect to the long axis of
the wire. Similar molecular orientation results were obtained from the six
different wires, which proved the reproducibility of the method and similar
properties of the two substrates. The above-mentioned results give the direct
evidence to explain the growth of Au-TTF wire: firstly, the molecules are
assembled along the long axis of the wire and this enables the stacking of
charged TTF molecules and the growth of wire in length; and secondly, with the
tilt angle of the molecules to the long X axis (blue axis), TTF molecules can be
stacked along Y (blue axis) and Z (blue axis) directions. This results in the
different diameter of Au-TTF wires from nanometre to several micrometres (see
Chapter 4.4 and References 27,28,100).
5 Results and discussion
86
Figure 5.22. The 3D models of Au-TTF wire showing rotation of molecules. Blue XYZ
coordinates refers to the wire coordinates, while red XYZ coordinates showing the
molecule coordinates.
To conclude, the molecular orientation in an individual Au-TTF nanowire was
determined and used to understand the growth mechanism of Au-TTF wires. It
was deducted that the orientation of TTF molecules in a charge-transfer
nanowire could influence the properties of the wire, e.g. absorbing molecules
on the surface. In the following chapter, the microfluidic-assisted synthesis
based on different metal-organic complexes using transition metal ions and TTF
derivatives is of interest.
5.3 Conductive single nanowires formed and analysed on microfluidic devices
87
5.3 Conductive single nanowires formed and analysed on
microfluidic devices
Figure 5.23. Graphic abstract of the synthesis and analysis of metal-TTF (derivatives).
Nanostructures were firstly synthesized by microfluidic-based method with side streams,
then the obtained structures were well characterized. For those which formed nano-
/microwires, a single wire was aligned on microelectrodes for the conductivity
measurement. Results on this work has been published in the peer-reviewed journal
Journal of Material Chemistry C (2016, 4, 9235-9244, RSC publishing).
Single nanowire devices are the key materials in nanoelectronics and
nanosensors.
55,166,167
An increasing research interest has been focused on
controlling the morphology and size distribution of nanostructures. In this
project, the synthesis of various TTF-based nanowires is discussed.
As discussed in Chapter 1.2, TTF can react with transition metal ions to
form charge-transfer complexes.
168-170
In previous work, the charge-
transfer between transition metal ions (Cu(II), Fe(III), Ru(III), Rh(III)
etc.) and TTF derivatives has been studied, as well as the conductivity of
their powder products.
12,13,171
In addition, the formation of fibre structures
formed from Au
172
and Pb
173
with TTF and its derivative has been
5 Results and discussion
88
reported. In TTF-based charge-transfer complexes, the mixed-valence
state of TTF composed of neutral molecules and cation radicals has been
proved.
13,24
The common feature in all these structures is the stacking of
the TTF units, with metal units coordinated to the sulphur atoms in the
stacks.
12,24,174
The morphology of these complexes is affected by metals,
different TTF derivatives, as well as the experimental variables e.g.
solvents, etc.
For the above mentioned TTF-based compounds (Cu-TTF, Fe-TTF, etc.),
a systematic study is missing, in which the reaction products of the various
combinations of metal precursors and TTF derivatives are thoroughly
characterized. This detailed study on these parameters will be useful for
the development of advanced single nanowire devices. In addition, nano-
/microwires derived from metal-TTF charge-transfer compounds like Au-
TTF, have been proved to act as sensing elements by fast electronic
28
or
optical
100
readout. Thus, for further application purposes, it is of great
importance to sort out the metal-TTF (and derivatives) which can easily
form quasi-1D single wire structures under certain experimental
conditions.
In this work, the microfluidic-guided synthesis of a series of TTF-based
metal-organic structures by using variable transition metal ions and TTF
derivatives in different solvents was carried out (Figure 5.23). The aim is
to explore single nanowires with good conducting behaviour for novel
advanced nanodevices. Transition metal ions and organic ligands which
have TTF redox-active moieties were used. These reactions were
performed on a microfluidic device,
27,58,175
where the precursors reacted at
the interface of two laminar streams. The so formed structures were
characterized by SEM and optical microscopy, and the obtained fibres
were further analysed by UV-Vis absorption, IR and Raman spectroscopy,
elemental analysis,
EDX and conductivity measurements. In addition, the
effects of substitute groups of TTF derivatives on wire formation and conductive
properties were compared and discussed.
5.3.1 Microfluidic guided synthesis of charge-transfer complexes
In this work, the microfluidic-guided synthesis of a series of TTF-based
metal-organic structures on a four-inlet microchip by using variable
transition metal ions and TTF derivatives is described. The aim is to
explore the single nanowires with good conductivity behaviour for further
5.3 Conductive single nanowires formed and analysed on microfluidic devices
89
developing advanced nanodevices. Generally, metal ions (including
Au(III), Cu(I), Cu(II), Fe(II), Fe(III), Co(II), Ni(II), Mn(II), Zn(II)) and
TTF-based organic ligands (including TTF, FTTF, TET-TTF, TCE-TTF
and BEDT-TTF), structures shown in Figure 5.24) were applied to the
reaction. The two reactants were supplied from two middle channels,
while solvents were introduced from side channels (at flow rate of
500;50;50;500
P
L/min for the four channels) (see Chapter 4.4 for
experimental details). An overview of the metal salts, solvents, the
morphologies of metal-TTF (and derivatives) obtained from bulk
synthesis and microchips, as well as the single wire conductivity is given
in Table 5.4. In the following, the major findings on the reaction of
TTF
with various metals, particularly Cu(II) and Fe(III) salts, are discussed. Finally,
the results for the TTF derivatives are presented.
Figure 5.24. Chemical structures of TTF derivatives. The labelling of elements of
TTF molecules refers to Reference 164.
Table 5.4. An overview of the main experimental parameters, morphology,
conductivity and characterization of metal-TTF (and derivatives) complexes
# σ refers to the conductivities of the specific measured single wires (see Figure 5.34
inset) at room temperature.
† Bulk results from this work; diameters of wires formed by microfluidics depend on the
applied method; wire formed by diffusion (~2Pm) conductivity was calculated from a
previous work (Reference 28).
* For BEDT-TTF a two-solvent protocol was used, with metal salts in CH3OH and
BEDT-TTF in THF.
** n.a. not available
“-” means no result shown here.
5 Results and discussion
90
Metal
salts
No.
Solvent
Compound
Formula
Properties
σ
#
S cm
-1
Characterization of
structures
Ref.
diameter
bulk [
P
P
m]
diameter
P
fluidic
[
P
m]
I. Tetrathiafulvalene (TTF)
HAuCl
4
†
1
CH
3
CN
(TTFCl
0.78)Au0.12
0.41
-4.43
0.2
-
2.0
0.021
SEM/ EDX
-
SEM/
UV
-
Vis/ IR/ EA (see
Ref.)
24
,
27,
28
,
51,
100
CuCl
2
2a
CH
3
CN
(TTF
)7(CuCl2)3
0.75 ± 0.39
Fig.
5.25
e_2a
0.31 ± 0.15
Fig.
5.25
e_2a
1.13
Fig.
5.35
a
SEM (Figs. 5.2
5b-
d)/
EDX
-SEM (Fig.
5.2
5f)/ UV-
Vis (Fig.
5.31
a)/ IR (Fig.
5.3
2
)/ Raman (Fig.
5.3
3
)/ EA (Table 5.5)
10
,
13,
171
2b
CH
3
OH
(TTF)
2CuCl2
0.91 ± 0.43
Fig.
5.2
5e_2b
0.69 ± 0.25
Fig.
5.2
5e_2b
0.34
Fig.
5.3
5b
SEM (Figs. 5.2
6a-
b)/
EDX
-SEM (Fig.
5.29
a)/ UV-
Vis (Fig.
5.31
b)/ IR (Table
5.7)/ Raman (Fig.
5.3
3
)/ EA (Table 5.5)
10
,
13,
171
Cu(NO
3
)2
3a
CH
3
CN
(TTF)
4Cu(NO3)
1.35 ± 0.67
Fig.
5.2
5e_3a
0.68 ± 0.27
Fig.
5.2
5e_3a
0.31
Fig.
5.3
5c
SEM (Figs. 5.2
6c-
d)/
EDX
-SEM (Figs.
5.
29b)/ UV-
Vis (Fig.
5.31
b)/ IR (Fig.
5.3
2
)/ Raman (Fig.
5.3
3)/ EA
(Table 5.5)
10
,
171
3b
CH
3
OH
(TTF)
4Cu(NO3)2
1.53 ± 0.41
Fig.
5.2
5e_3b
0.55 ± 0.22
Fig.
5.2
5e_
3b
0.15
Fig.
5.3
5d
SEM (Figs. 5.2
6e-
f)/
EDX
- SEM (Fig.
5.29
c)/ UV-
Vis (Fig.
5.31
b)/ IR (Table
5.7)/ Raman (Fig.
5.3
3
)/ EA (Table 5.5)
10
,
171
FeCl
3
4a
CH
3
CN
(TTF)
5FeCl3
3.82 ± 0.71
Fig.
5.2
5e_4a
0.93 ± 0.09
Fig.
5.2
5e_4a
0.11
Fig.
5.3
5e
SEM (Figs. 5.2
7a-
b)/
EDX
- SEM (Fig.
5.30
a)/ UV-
Vis (Fig.
5.31
c)/ IR (Table
5.7)/ Raman (Fig.
5.3
3
)/ EA (Table 5.5)
This
work
4b
CH
3
OH
(TTF)
5FeCl3
1.35 ± 0.28
Fig.
0.32 ± 0.05
Fig.
0.40
Fig.
SEM (Figs. 5.2
7c-
d)/
EDX-SEM (Fig.
12
5.3 Conductive single nanowires formed and analysed on microfluidic devices
91
5.2
5e_4b
5.2
5e_4b
5.3
5f
5.30
b)/ UV-
Vis (Fig.
5.3
1c)/ IR (Table
5.7)/ Raman (Fig.
5.3
3
)/ EA (Table 5.5)
Fe(NO
3
)
3
5a
CH
3
CN
(TTF)
5Fe(NO3)3
6.05 ± 1.25
Fig.
5.2
5e_5a
0.83 ± 0.36
Fig.
5.2
5e_5a
0.35
Fig.
5.3
5g
SEM (Figs. 5.2
7e-
f)/
EDX
-SEM (Fig.
5.30
c)/ UV-Vis
(Fig.
5.31
c)/ IR (Table
5.7)/ Raman (Fig.
5.3
3
)/ EA (Table 5.5)
This
work
5b
CH
3
OH
(TTF)
5Fe(NO3)3
1.51 ± 0.61
Fig.
5.2
5e_5b
0.82 ± 0.21
Fig.
5.2
5e_5b
0.53
Fig.
5.3
5h
SEM (Figs. 5.2
7g-
h)/
EDX
-SEM (Fig.
5.30
d)/ UV-Vis
(Fig.
5.3
1c)/ IR (Table
5.7)/ Raman (Fig.
5.3
3
)/ EA (Table 5.5)
This
work
Cu(I)Cl
6
CH
3
CN
─
─
No
structure
F
ig. 5.28
a
─
UV
-Vis
Figs. 5.31
e-f
This
work
FeCl
2
7
CH
3
OH
─
Needles
Fig. 5.2
8
b
Needles
Fig.
5.28
c
─
UV
-Vis
Figs. 5.3
1e-f
This
work
CoCl
2
8
CH
3
CN
─
─
No
structure
Fig. 5.2
8
d
─
UV
-Vis
Figs. 5.31
e-f
This
work
Co(NO
3
)2
9
CH
3
CN
─
Flake
Fig. 5.2
8
e
Particles
(Fig. 5.2
8
f)
─
UV
-Vis
Figs. 5.3
1e-f
This
work
MnCl
2
10
CH
3
OH
─
─
Particles
Fig. 5.2
8
g
─
UV
-Vis
Figs. 5.3
1e-f
This
work
NiCl
2
11
CH
3
OH
─
Particles
Fig. 5.2
8
h
Particles
Fig. 5.28i
─
UV
-Vis
Figs. 5.3
1e-f
This
work
Ni(NO
3
)
2
12
CH
3
CN
─
Particles
Fig. 5.2
8
j
Particles/
needles
Fig. 5.2
8
k
─
UV
-Vis
Figs. 5.3
1e-f
This
work
ZnCl
2
13
CH
3
CN
─
Particles
F
ig. 5.28
l
Particles
Fig.
5.28m
─
UV
-Vis
Figs. 5.31
e-f
This
work
Zn(NO
3
)2
14
CH
3
CN
─
Rods
Fig. 5.2
8
n
Needles
Fig. 5.2
8
o
─
UV
-Vis
Figs. 5.3
1e-f
This
work
II. 2
-formyl-tetrathiafulvalene (FTTF)
5 Results and discussion
92
HAuCl
4
15
CH
3
CN
(FTTF)
3AuCl3
Dendrites
Fig. 5.3
7
a
0.26 ± 0.14
Fig. 5.3
7
c
0.66
Fig.
5.41
c
EDX
-SEM (Fig.
5.40
a)/ UV-
Vis (Fig.
5.3
6a)/ IR (Fig.
5.
39
a)/ EA (Table
5.8)
This
work
CuCl
2
16
CH
3
CN
(FTTF)
2CuCl2
Dendrites
Fig. 5.3
7
d
0.20 ± 0.08
Fig.5.3
7f
0.84
Fig.
5.4
1d
EDX
-SEM (Fig.
5.40
b)/ UV-
Vis (Fig.
5.3
6a)/ IR (Fig.
5.
39
b)/ EA (Table
5.8)
This
work
Cu(NO
3
)2
17
CH
3
CN
─
Clusters
Fig. 5.3
8
a
Particles
Fig.
5.38
c
─
UV
-Vis
Fig. 5.3
6a
This
work
Fe(NO
3
)
3
18
CH
3
OH
─
Clusters
Fig.
5.38
d
Particles
Fig.
5.38
f
─
UV
-Vis
Fig.5.3
6b
This
work
FeCl
3
19
CH
3
OH
─
Dendrites
Fig. 5.3
8
g
Dendrites
Fig.
5.38
i
─
UV
-Vis
Fig.5.3
6b
This
work
FeCl
2
20
CH
3
OH
─
FTTF
Dendrites
Fig. 5.3
8
j
Dendrites
Fig.
5.38
k
─
UV
-Vis
Fig. 5.3
6c
This
work
ZnCl
2
21
CH
3
CN
─
FTTF
Crystals
Fig. 5.3
8
l
Particles
Fig.
5.38m
─
UV
-Vis
Fig. 5.3
6c
This
work
Zn(NO
3
)2
22
CH
3
CN
─
FTTF
Crystals
Fig. 5.3
8
n
Particles
Fig.
5.38
o
─
UV
-Vis
Fig. 5.3
6c
This
work
III. B
is(ethylenedithio)tetrathiafulvalene (BEDT-TTF)*
HAuCl
4
23
CH
3
OH
-
THF
─
1.73 ± 0.61
Fig.
5.44
a
0.56 ± 0.12
Fig. 5.4
5
a
6.3×10-3
Fig.
5.4
7a
SEM (5.46a)/ UV
-
Vis (Fig. 5.4
3a)/
EDX
-SEM (Fig.
5.4
8b)
172
CuCl
2
24
CH
3
OH
-
THF
(BEDTTTF)
1.5
Cu
Cl
2
1.79
±
0.31
Fig. 5.4
4
b
0.59 ± 0.14
Fig.5.4
6
a
n.a.**
Fig.
5.4
6b
SEM (5.47a)/ UV
-
Vis (Fig. 5.4
3b)
174,
176,
177
Cu(NO
3
)2
25
CH
3
OH
-
THF
─
3.16 ± 0.82
Fig.
5.44
c
2.11 ± 0.53
Fig.5.4
5b
4.8×10-3
Fig.
5.4
7b
SEM (5.46b)/ UV
-
Vis (Fig. 5.4
3b)/
EDX
-SEM (Fig.
5.4
8c)
This
work
5.3 Conductive single nanowires formed and analysed on microfluidic devices
93
FeCl
3
26
CH
3
OH
-
THF
─
1.68 ± 0.95
Fig. 5.4
4
d
0.50 ± 0.12
Fig.5.4
6c
n.a.**
Fig.
5.4
6d
SEM (5.47c)/ UV
-
Vis (Fig. 5.43
c)
This
work
Fe(NO
3
)
3
27
CH
3
OH
-
THF
─
1.29 ± 0.59
Fig. 5.4
4
e
0.47 ± 0.11
Fig.5.4
5
c
1.2×10-2
Fig.
5.4
7c
SEM (5.46c)/ UV
-
Vis (Fig. 5.4
3c)/
EDX
-SEM (Fig.
5.4
8d)
This
work
IV
. Tetrakis(ethylthio)tetrathiafulvalene (TET-TTF)
HAuCl
4
28
CH
3
CN
─
Crystals
Fig. 5.4
2
a
─
─
─
This
work
CuCl
2
29
CH
3
CN
─
Crystals
Fig. 5.4
2
b
─
─
─
This
work
V. 2,3,6,7
-Tetrakis(2-cyanoethylthio) tetrathiafulvalene (TCE-TTF)
HAuCl
4
30
CH
3
CN
─
Crystals
Fig. 5.4
2
c
─
─
─
This
work
5.3.2 Metal-TTF complexes
a) Synthesis and analysis by optical microscopy and SEM
Figure 5.25a shows exemplary the reaction of CuCl2 and TTF carried out on a
microdevice by supplying the two reactant solutions through channels B and C
both at flow rates of 50 μL/min. The solvent, here CH3CN, was supplied through
channels A and D. The solvent streams with a flow rate of 500 μL/min focused
the inner streams to a narrow reaction zone.27 The reaction to (TTF)7(CuCl2)3
(2a) occurred immediately as can be seen by the black product in Figure 5.25a.
The structures were flushed out of the channel and observed by SEM (Figure
5.25b). For comparison, the reaction in a two-inlet microchip, i.e. without sheath
flows (Figure 5.25c), and a standard glass flask (Figure 5.25d) was also carried
out. As can be seen from the SEM images (Figure 5.25b), the reaction in the
four-inlet microdevice yielded the longest fibres with a smaller diameter than
those obtained by the other approaches (Figures 5.25c and 5.25d). The four-inlet
microfluidic chip is preferred in the following because it provides a defined
interface between the two laminar flowing precursor streams and confines the
reaction zone by use of the sheath flow. The wires assemble in this reaction zone
5 Results and discussion
94
in direction of the flow. Furthermore, this chip design has the benefits that the
wires are flushed out immediately and do not block the channel. Once outside,
they are immediately diluted and do not react further.
Figure 5.25. Synthesis of wires, exemplary for (TTF)7(CuCl2)3 (2a). a) Optical
micrograph of the four inlet microchip. Wire bundles are formed at the interface of the
CuCl2 (channel B, flow rate 50 μL/min) and TTF solutions (channel C, flow rate 50
μL/min) in CH3CN. The solvent was also used as sheath flow, supplied from channel A
and D, at flow rates of 500 μL/min. b) SEM image of wire structures obtained on the
four-inlet microchip. c) and d) are SEM images of wire structures obtained by two-inlet
chip (flow rates of 100;100 μL/min) and bulk synthesis. e) is the comparison of the
diameters (μm) of wires obtained from different M-TTF complexes synthesized in bulk
(black square) and on four-inlet microchips (red dot). The average diameter and standard
deviation were obtained from ten different wires for each complex. For the compounds
in the horizontal axis (2a, 2b, 3a, 3b, 4a, 4b, 5a, 5b), refer to numbers in Table 5.4
(column 2). f) EDX-SEM spectrum of 2a.
5.3 Conductive single nanowires formed and analysed on microfluidic devices
95
Figure 5.26. SEM images of (TTF)2CuCl2 (2b), (TTF)4Cu(NO3)2 (3a) and
(TTF)4Cu(NO3)2 (3b).
2b generated a) by bulk synthesis and b) on four-inlet microchip.
(Figure b inset is a magnified image of the end of a needle structure showing bundles of
wires). c) and d) SEM images of 3a generated by bulk synthesis and on microchip in
CH3CN. e) and f) SEM images of 3b generated by bulk synthesis and on microchip. The
numbers of compounds are in accordance with the numbers in Table 5.4.
The morphology of structures obtained from the reaction of TTF with Cu(II)
and Fe(III) salts was studied (2b, 3a, 3b, 4a, 4b, 5a, 5b), which showed clearly
differences between bulk synthesis and four-inlet microchips (Figure 5.25e,
SEM images in Figures 5.26 and 5.27). All the wires resulted from the four-inlet
microchip are in nanometre range and smaller than from the bulk synthesis,
confirming that microfluidic synthesis is an advanced tool to control the
formation of nanostructures.
5 Results and discussion
96
Figure 5.27. SEM images of (TTF)5FeCl3 (4a), (TTF)5FeCl3 (4b), (TTF)5Fe(NO3)3 (5a)
and (TTF)5Fe(NO3)3 (5b). a) and b): 4a formed in bulk synthesis and on four-inlet
microchip. c) and d): 4b formed in bulk and on microchip. e) and f): 5a generated by
bulk synthesis and on microchip. g) and h): 5b generated by bulk synthesis and on
microchip. The numbers of compounds are in accordance with the numbering in Table
5.4.
In addition, the products of reactions between TTF and other metal salts
including CuCl (
6
), FeCl
2
(
7
), CoCl
2
(
8
), Co(NO
3
)
2
(
9
), MnCl
2
(
10
), NiCl
2
(
11
), Ni(NO
3
)
2
(
12
), ZnCl
2
(
13
) and Zn(NO
3
)
2
(
14
) were investigated by
optical microscopy. Here, the products showed various morphologies (e.g.
particles, crystals, mixtures of particles and needles, etc.), while not
uniform 1D nano-/microstructures (Figure 5.28).
5.3 Conductive single nanowires formed and analysed on microfluidic devices
97
Figure 5.28. Optical images of products derived from compounds
6
,
7
,
8
,
9
,
10
,
11
,
12
,
13
,
14
. a) CuCl-TTF (
6
) formed on four-inlet microchip, no obvious structures;
FeCl
2
-TTF (
7
) formed b) in bulk, mixture of needle and small particles, and c) on
two-inlet microchip, large branches of orange crystals; d) CoCl
2
-TTF (
8
) on four-
inlet microchip, no obvious structures; Co(NO
3
)
2
(
9
) formed e) in bulk, flakes and
f) on four-inlet microchip, particles; g) MnCl
2
-TTF (
10
) on four-inlet microchip,
particles; NiCl
2
-TTF (
11
) formed h) in bulk, mixture of small rods segments and
particles, and i) on microchip, particles; Ni(NO
3
)
2
-TTF (
12
) formed j) in bulk,
5 Results and discussion
98
mixture of small rods and particles, and k) on four-inlet microchip, particles;
ZnCl
2
-TTF (
13
) formed l) in bulk, mixture of crystals, rods and particles, and m)
on two-inlet microchip, large particles; Zn(NO
3
)
2
-TTF (
14
) formed n) in bulk,
cluster of small rods and particles, and o) on two-inlet microchip, needles and
crystals. The numbers of compounds are in accordance with the numbering in
Table 5.4.
b) EDX and elemental (CHN) analysis
To determine the elemental composition of the nanowires obtained from
microchips, EDX spectroscopy was performed. The spectrum of
2a
(Figure 5.25f) clearly showed C, Cu, S, Cl elements, indicating that the
nanowires are composed of CuCl
2
and TTF (C
6
H
4
S
4
). The results for the
other Cu(II) salts are depicted in Figure 5.29, which proved the
composition of CuCl
2
and TTF for
2b
(Figure 5.29a), as well as the
composition of Cu(NO
3
)
2
and TTF for
3a
(Figure 5.29b) and
3b
(Figure
5.29c). The EDX spectra of Fe(III) salts are shown in Figure 5.30, which
also proved the composition of FeCl
3
and TTF for
3a
(Figure 5.30a) and
3b
(Figure 5.30b), and the composition of Fe(NO
3
)
3
and TTF for
3c
(Figure 5.30c) and
3d
(Figure 5.30d).
5.3 Conductive single nanowires formed and analysed on microfluidic devices
99
Figure 5.29. EDX-SEM spectra of (TTF)
2
CuCl
2
(
2b
)
,
(TTF)
4
Cu(NO
3
)
2
(
3a
)
and
(TTF)
4
Cu(NO
3
)
2
(
3b)
. Note the existence of Cu and Cl in a) and the existence of
N and Cu in c) and d). The small peaks of N was due to its low percentage in the
complexes.
Figure 5.30. EDX-SEM spectra of (TTF)5FeCl3 (4a), (TTF)5FeCl3 (4b),
(TTF)5Fe(NO3)3 (5a) and (TTF)5Fe(NO3)3 (5b). Note the existence of Fe and Cl in a)
and b), the existence of N and Fe in c) and d). The small peaks of N was due to its low
percentage in the complexes.
To get the formula of the above-mentioned TTF-based compounds, elemental
analysis (CHN) was performed. The results are summarized in Table 5.5, which
confirmed the chemical composition of the nanostructures of 2a, 2b, 3a, 3b, 4a,
4b, 5a and 5b, quantitatively. The compound formula are in accordance with
reported work for bulk synthesized compounds obtained using the same metal
salts and solvents.
5 Results and discussion
100
Table 5.5. Elemental analysis of compounds 2a, 2b, 3a, 3b, 4a, 4b, 5a and 5b
Metal salts
Solvent
Anal. % found Compound
No. (calculated) Formula
C H N
CuCl
2
CH
3
CN
28.21 1.53 (TTF)
7
(CuCl
2
)3
2a (27.51) (1.54)
CuCl
2
CH
3
OH
27.20 1.57 (TTF)
2
CuCl
2
2b (26.54) (1.48)
Cu(NO
3
)2
CH
3
CN
28.98 1.77 3.50 (TTF)
4
Cu(NO
3
)2
3a (28.68) (1.60) (2.79)
Cu(NO
3
)2
CH
3
OH
26.88 1.58 2.88 (TTF)
4
Cu(NO
3
)2
3b (28.68) (1.60) (2.79)
FeCl
3
CH
3
CN
30.91 2.23 (TTF)
5
FeCl
3
4a (30.43) (1.70)
FeCl
3
CH
3
OH
29.99 2.00 (TTF)
5
FeCl
3
4b (30.43) (1.70)
Fe(NO
3
)
3
CH
3
CN
28.77 1.81 3.52 (TTF)
5
Fe(NO
3
)3
5a (28.52) (1.60) (3.32)
Fe(NO
3
)
3
CH
3
OH
28.54 1.81 3.37 (TTF)
5
Fe(NO
3
)3
5b (28.52) (1.60) (3.32)
c) UV-Vis spectroscopy
The eluted solution from the chips after reaction were diluted and directly
applied for UV-Vis investigations. Three conclusions can be drawn from
the UV-Vis spectra analysis. Firstly, the electron transfer between metal
ions and TTF happened after reaction: as indicated in Figure 5.31a, TTF
has a weak absorbance at 436 nm, while CuCl
2
only shows negligible
absorption signal. In comparison, the spectrum of eluted solution of
2a
exhibits an intense band at 436 nm and a new absorption band at 580 nm,
which indicates the existence of TTF radical cation, due to the electron
transfer from TTF to Cu
2+
.
26,171
Secondly, there is an influence from
solvents on the electron transfer reaction. The absorption spectra of the
complexes (
2b, 3a, 3b, 4a, 4b, 5a, 5b
) at 580 nm are summarized in Figure
5.31b, which
indicating the electron transfer between TTF and Cu(II), as well
as that of TTF and Fe(III) (
Figure 5.31
c). No
interferences on absorption
from metal ions (Figure 5.31d). However, the differences in absorption
intensities clearly shows that Cu(II) is more powerful oxidizing agent in
CH
3
CN than in CH
3
OH (Figures 5.31a and Figure 5.31b,
A
2a
>A
3a
>A
2b
>A
3b
), while Fe(III) showing higher oxidative ability in
5.3 Conductive single nanowires formed and analysed on microfluidic devices
101
CH
3
OH than in CH
3
CN (Figure 5.31c, A
5b
>A
5a
>A
4b
>A
4a
). This is in
accordance with reported work.
10,178
Figure 5.31. UV-Vis absorption spectra of solutions of 2a, 2b, 3a, 3b, 4a, 4b, 5a, 5b, 6,
5 Results and discussion
102
7, 8, 9, 10, 11, 12, 13 and 14. a) CuCl2 in CH3CN (0.06 mM), TTF in CH3CN and CH3OH
(both are 0.24 mM) and the solution after the reaction of CuCl2 and TTF in CH3CN. A
new absorption band at 580 nm (A = 0.237) appeared after formation of the complex 2a.
Dash lines indicated the position of 436 nm and 580 nm. The absorption band at 580 nm
of different charge-transfer complexes in different solvents: b) 3a (A = 0.189), 2b (A =
0.141) and 3b (A = 0.115); c) 5b (A = 0.28), 4b (A = 0.257), 5a (A = 0.262) and 4a (A
= 0.179); d) Fe(III) and Cu(II) salts in CH3CN and CH3OH; e) metal salts of Cu(I), Fe(II),
Co(II), Mn(II), Ni(II), Zn(II); f) solution after the reaction of TTF and metal ions in e).
6, 7, 8, 9, 10, 11, 12, 13, 14. The final concentrations of all metal salts were 0.06 mM.
The elutes of microreactions were diluted to approx. 0.06 mM. Dash lines in Figures b)
and c) indicated the position of 580 nm. The numbering of compounds are in accordance
with the numbers in Table 5.4.
Thirdly, the influence of various metals ions and different oxidize state of
the same metal ions on the electron transfer was understood. For the
precursors of Cu(I), Fe(II), Co(II), Mn(II), Ni(II), Zn(II) ions and the
products after their reactions with TTF, only negligible absorptions were
observed at 580 nm (Figures 5.31e and 5.31f). Following, a comparison
of the spectra from different oxidised states of the same metal was
performed. In contrast to the spectra of solutions resulted from Cu(II) and
TTF (
2a
,
2b
, Figures 5.31a and 5.31b), the spectrum of Cu(I) and TTF
exhibits no obvious absorbance at 580 nm (
6,
Figure 5.31f). Similar results
were observed for the spectra of solutions from Fe(III) and TTF (
4a
,
4b
,
Figure 5.31c), and that of Fe(II) and TTF (
7
, Figure 5.31f). Thus, only
higher oxidised state of Cu(II) and Fe(III) can lead to the electron transfer
from TTF to corresponding metal ions in the experimental conditions
applied in this work.
d) IR spectroscopy
The IR spectra were obtained for the M-TTF complexes which formed wire
structures. Table 5.6 shows the typical IR vibrations of neutral TTF and TTF
cation.
5.3 Conductive single nanowires formed and analysed on microfluidic devices
103
Table 5.6. IR assignments of TTF0 and TTF+*
Vibrational modes*
TTF0
TTF+
ʋ23 CCH bend
1254 cm-1
1237 cm-1
ʋ15 CCH bend
1090 cm-1
1072 cm-1
ʋ16 C-S stretch
781 cm-1
836 cm-1
ʋ25 ring SCC bend
794 cm-1
825 cm-1
ʋ17 C-S stretch
734 cm-1
751 cm-1
* The symmetry species numbers (ʋ23, ʋ15, ʋ16, ʋ25, ʋ17 ) see literature.179
As shown in the IR spectrum of 2a (Figure 5.32, 2a), modes at 1242 cm-1 (ʋ23,
CCH bend), 1082 cm-1 (ʋ15, CCH bend), 814 cm-1 (ʋ16, CS stretch), 804 cm-1
(ʋ25, SCC bend), 740 cm-1 (ʋ17, CS stretch) are all between the IR vibrations for
TTF0 molecules and TTF+ cations (Table 5.6). Such IR spectrum indicates the
mixed-valence state of this TTF-based material in the neutral and cation radical
states.13,179 Similar TTF bands were also observed for 3a at 1245 cm-1 (ʋ23, CCH
bend), 817 cm-1 (ʋ16, CS stretch) and 733 cm-1 (ʋ17, CS stretch) (Figure 5.32).
The observed peak at 1312 cm-1 indicated the existence of nitrate.10,180 which
formed anion columns in the resulted structures. In addition, the mixed-valence
state of neutral and TTF cations in other TTF-based complexes (2b, 3a, 3b, 4a,
4b, 5a, 5b) was also confirmed by their IR spectral modes (Table 5.7). Thus,
similar structures were formed for these M-TTF wires, with a backbone scaffold
of TTF neutral and cation radicals stabilized with counter ions forming the 1D
nanowires. To further prove the TTF components in the nanostructures,
confocal Raman measurements were performed. Raman peaks at around 500
cm-1 and 1420 cm-1 (Figure 5.33) were observed, similarly to the spectra of Au-
TTF (1) in previous works.51,100 The difference between 1 and the complexes
(2b, 3a, 3b, 4a, 4b, 5a, 5b) is that in the former structure Au(III) ions were
reduced to Au0 ,24 while in the latter structures, metal ions existed as low
oxidized Cu(I) and Fe(II) when excess TTF was used. The common features of
these donor-acceptor complexes structures are firstly the mix-valence state of
TTF stacking, and secondly the coordinated metal (Au0)24 and metal ion (Cu(I)
and Fe(II)) to the sulphur atoms in the TTF stacks.12
5 Results and discussion
104
Figure 5.32. IR spectra of compounds (TTF)7(CuCl2)3 (2a) and (TTF)4Cu(NO3)2 (3a). In
the spectrum of 2a, characteristic vibration modes at 1242 cm-1 (ʋ23, CCH bend), 1082
cm-1 (ʋ15, CCH bend), 814 cm-1 (ʋ16, CS stretch), 804 cm-1 (ʋ25, SCC bend), 740 cm-1 (ʋ17,
CS stretch) are shown. In the spectrum of 3a, typical bands were observed at 1245 cm-1
ʋ23, CCH bend), 817 cm-1 (ʋ16, CS stretch) and 733 cm-1 (ʋ17, CS stretch), the peak at
1312 cm-1 indicates the presence of nitrate (NO3-). Products were obtained on four-inlet
microchips. The wavenumbers of the different vibrational bands are shown for TTF0 (T0,
solid line) and its cation TTF+ (T+, dash line). The numbering of all the IR vibrational
modes is given in reported work.13,179
5.3 Conductive single nanowires formed and analysed on microfluidic devices
105
Table 5.7. IR assignments of other M-TTF compounds 2b, 3b, 4a, 4b, 5a and 5b
Comp-
ound
IR spectra* (cm-1)
NO3-

ʋ
23
,
CCH bend

ʋ
15
,
CCH bend

ʋ
16
,
CS stretch

ʋ
25
,
SCC bend

ʋ
17
,
CS stretch
2b
1245
1078
814
803
3b
1319
1247
820
4a
1243
1076
823
800
743
4b
1236
1076
823
5a
1318
1243
827
5b
1314
1247
818
Notes: Compounds were obtained on four-inlet chips. The numbers of compounds are in
accordance with the numbers in Table 5.4.
*The numbering of the vibrational modes is given in literature.13,179 The IR spectra
information of nitrate is given in literature.180
Figure 5.33. Raman spectra of compounds 2a, 2b, 3b, 4a, 4b, 5a and 5b at 532 nm
excitation. (a) 2a (b) 2b, (c) 3a, (d) 3b, (e) 5a, (f) 5b (g) 4a, (h) 4b. The Raman peaks at
around 500 cm-1 and 1420 cm-1.
500 1000 1500
(h)
(g)
(f)
(e)
(c)
(b)
(d)
Raman Intensity (a.u.)
Raman Shift (cm
-1
)
(a)
5 Results and discussion
106
e) Electrical properties
It has been reported that partially charged TTF salts exhibit higher
electrical conductivity than simple salts at room temperature.
179
Thus,
nanowires obtained from M-TTF were used to address conductivity via
four-point probe measurements by aligning a single nanowire on pre-
fabricated microelectrodes. I-V curves of these complexes were obtained
by measuring the source-drain voltage, while sweeping the applied
source-drain current at room temperature. As shown in Figure 5.34, the I-
V curves for
2b
,
3a
,
3b
,
4a
,
4b
,
5a
and
5b
are non-linear. It can be assumed
as an indication of the non-perfect contact, as the nanowires were simply
aligned on the microelectrodes, without topside metallization. A similar
conductive behaviour of
1
was observed in previous work.
28
Different
voltage ranges were observed for different nanowires with various
diameters and lengths when the same current was applied. Thus, a
comparison of the linear conductance of all the I-V curves in Figure 5.34
was made and shown in Figure 5.35. The results indicated that all the
nanowires exhibited high electrical conductivities from 10
-1
to 10 S cm
-1
at room temperature (Figure 5.35). In particular,
2a
showed the highest
conductivity (σ = 1.13 S cm
-1
), followed by
5b
(σ = 0.53 S cm
-1
) and
4b
(σ = 0.40 S cm
-1
).
Figure 5.34. Representative I-V curves of single nanowires of M-TTF complexes. The
curves were obtained by four-point measurements (the two outer microelectrodes
represent source high and source low, while the two inner electrodes represent sense high
5.3 Conductive single nanowires formed and analysed on microfluidic devices
107
and sense low). a) (TTF)7(CuCl2)3 in CH3CN (2a), b) (TTF)2CuCl2 in CH3OH (2b),
(TTF)4Cu(NO3)2 in c) CH3CN (3a) and in d) CH3OH (3b), (TTF)5FeCl3 in e) CH3CN
(4a) and in f) CH3OH (4b), (TTF)5Fe(NO3)3 in g) CH3CN (5a) and in h) CH3OH (5b).
Figure insets show the alignment of single nanowires on the four microelectrodes, all the
electrodes connected to the source meter. The gaps between electrodes (scale bars) are 5
μm.
Figure 5.35. Linear range of the I-V curves in Figure 5.34 with the electrical
conductivities of single nanowires of 2a, 2b, 3a, 3b, 4a, 4b, 5a and 5b. a) 2a, σ = 1.13 S
cm-1, b) 2b, σ = 0.34 S cm-1; c) 3a, σ = 0.31 S cm-1 ; d) 3b, σ = 0.15 S cm-1 (here the I-V
curve showed linearity between -0.02PA and 0.02 P$) ; e) 4a, σ = 0.11 S cm-1 ; f) 4b, σ
= 0.40 S cm-1; g) 5a, σ = 0.35 S cm-1; h) 5b, σ = 0.53 S cm-1. σ refers to the conductivity
at room temperature.
Thus, with the so proposed characterization technique, it was possible to
measure the conductivities of single nanowires, for which the non-linear
curve were caused by contact effects between the wire and
microelectrodes, while the linear parts of I-V curves show typical ohmic
behaviour (Figure 5.35). In addition, the electron transfer of TTF to metal
ions and conductive behaviour of induced fibres were influenced by the
solvent used in the system.
5 Results and discussion
108
5.3.3 Metal-FTTF complexes
Furthers studies on the reaction between transition metal salts and commercially
available TTF derivatives were conducted. FTTF, consisting of an aldehyde
functional group at the C1 substitution of TTF unit, was further explored. The
reactions of FTTF with HAuCl4, Cu(II) and Zn(II) salts were performed in
CH3CN, while the reaction with Fe(III) salts in CH3OH.
a) UV-Vis spectroscopy
In the UV-Vis spectral analysis of M-FTTF (Figure 5.36), compared to pure
FTTF (Figures 5.36a and 5.36b), all the spectra for compounds 15, 16, 17, 18
and 19 exhibited the absorption at 437nm and 580 nm, suggesting a charge-
transfer between these metal ions and FTTF. However, in the reaction of Zn(II),
Fe(II) with FTTF (Figure 5.36c, 20, 21, 22), no peaks at 437 nm or 580 nm were
observed.
Figure 5.36. UV-Vis absorption spectra of the solutions of M-FTTF. a) 15, 17, 16 and
pure FTTF in CH3CN, b) 19, 18 and pure FTTF in CH3OH. c) 22, 21 and 20. Final
concentrations for FTTF were 0.24 mM and for other solutions about 0.06 mM. The dash
lines in Figures a) and c) indicates the position of 580 nm.
5.3 Conductive single nanowires formed and analysed on microfluidic devices
109
b) Morphologies
Various morphologies were observed for the products of M-FTTF obtained
from bulk synthesis and microchips. The reaction of FTTF with HAuCl4 and
CuCl2 yielded large dendrites in bulk synthesis (Figures 5.37a and 5.37d),
however, 1D wire structures were obtained on both two-inlet and four-inlet
microchips for the compounds 15 (Figures 5.37b and 5.37c) and 16 (Figures
5.37e and 5.37f).
Figure 5.37. Optical and SEM images of M-FTTF generated by bulk synthesis and on
microchip. a) optical image of Au-FTTF by bulk synthesis, only dendrites structures.
SEM images of Au-FTTF (15) generated on b) two-inlet microchip at flow rates of 100
PL/min (average diameter, 0.39 ± 0.11 Pm) and c) four-inlet microchip at flow rates of
500;50;50;500 PL/min (average diameter, 0.26 ± 0.14Pm). d) optical image of Cu-FTTF
(16) by bulk synthesis, only dendrites structures. SEM images of Cu-FTTF generated on
e) two-inlet microchip (average diameter, 0.27 ± 0.13 Pm) and f) four-inlet microchip
(0.2 ± 0.08 Pm).
Although Cu(NO3)3, Fe(NO3)3 and FeCl3 induced the charge-transfer with FTTF
(UV-Vis spectra, Figure 5.36b), only dendrites and particles while not 1D wire
structures were observed (17, Figures 5.38a-c) (18, Figures 5.38d-f) (19, Figures
5.38g-i). Similarly, for the other metal salts, including FeCl2-FTTF (20, Figures
5.38j-k), ZnCl2-FTTF (21, Figures 5.38l-m) and Zn(NO3)2-FTTF (22, Figures
5.38n-o), particles, dendrites or crystals were formed after reaction with FTTF.
5 Results and discussion
110
Figure 5.38. Optical images of structures from 17, 18, 19, 20, 21 and 22 in bulk and on
microchips: 17 a) in bulk (clusters), b) on two-inlet chip (clusters) and c) four-inlet chip
(particles). 18 d) in bulk (clusters), e) on two-inlet chip (small dendrites) and f) four-inlet
chip (particles). 19 g) in bulk (small dendrites), h) on two-inlet chip (small dendrites) and
i) four-inlet chip (large dendrites). 20 j) in bulk (large dendrites) and k) on four-inlet chip
5.3 Conductive single nanowires formed and analysed on microfluidic devices
111
(small dendrites). 21 l) in bulk (well ordered crystal trees) and m) on four-inlet chip (large
particles). 22 n) in bulk (well ordered crystal trees) and o) on four-inlet chip (large
particles). Notice the well ordered crystal trees in j), l) and n) are FTTF crystals.
c) Characterization of M-FTTF nanowires
IR spectra of 15 and 16 (obtained from microchips) and pure FTTF are displayed
in Figure 5.40. The so synthesised nanowires 15 and 16 showed evidently
shifted bands with the spectrum of pure FTTF in the range of 700 cm-1-1500 cm-
1 (Figure 5.39), because of the charge-transfer reaction (UV-Vis spectra, Figure
5.36a). As emerged from the comparison of the IR vibrational modes of
compounds 15, 16 with FTTF, the aldehyde group (C=O) centred at about 1640
cm-1 remained after the formation of nanowires (Figure 5.39). This result
suggests the possibility of further modifications of the wire using compounds
which can bond to aldehyde groups e.g. amines.
Figure 5.39. IR spectra of M-FTTF complexes. Au-FTTF (15), Cu-FTTF (16) and pure
FTTF.
5 Results and discussion
112
EDX-SEM spectra showed the presence of C, O, Au, S and Cl elements in Au-
FTTF (15) (Figure 5.40a), and C, O, Cu, S, Cl in Cu-FTTF (16) (Figure 5.40b).
Elemental analysis of 15 and 16 (Table 5.8) determined the composition of these
two charge-transfer compounds of metal chloride and FTTF. These results
demonstrated that the C1 substitution of aldehyde group on TTF unit (Figure
5.24) did not block the stacking of TTF units, while active side groups could be
preserved in the charge-transfer complexes.
Figure 5.40. EDX-SEM spectra of M-FTTF complexes. a) Au-TTF (15) and b) Cu-TTF
(16). Note the presence of Cl in a), and the presence of Cu and Cl in b).
Table 5.8. Elemental analysis of 15 and 16 complexes
Metal salts
Solvent
No. Anal. % found Compound
(calculated) Formula
C H
HAuCl
4
CH
3
CN
26.82 1.20 (FTTF)
3
AuCl3
15
(25.21) (1.20)
CuCl
2
CH
3
CN
28.97 1.72 (FTTF)
2
CuCl2
16
(28.07) (1.34)
Note: The compounds are numbered as in Table 5.4.
5.3 Conductive single nanowires formed and analysed on microfluidic devices
113
d) Electronic properties
I-V curves were measured for single nanowires of 15 and 16 formed on the four-
inlet microchip, after connecting the nanowires to microelectrodes. As shown in
Figures 5.41a and 5.41b, the nanowires showed metallic-like linear I-V curve.
The electrical conductivities were calculated considering only the linear range
of the curves in Figures 5.41a and 5.41b. And were found to be 0.66 S cm-1 and
0.84 S cm-1 for 15 (Figure 5.41c) and 16 (Figure 5.41d), respectively. Thus,
these results indicated a conductive behaviour of M-FTTF nanowire structures
at room temperature.
Figure 5.41. I-V curves of M-FTTF nanowires. a) Single Au-FTTF and b) Cu-FTTF
nanowire formed on four-inlet microchip. Figure insets in a) and b) showed the aligned
single nanowires on microelectrodes. The gaps between the microelectrodes are 5 Pm as
the scale bars. c) and d) Linear range of the I-V curves showing in Figures a) and b).
5.3.4 TCE-TTF, TET-TTF and BEDT-TTF
In an attempt to better understand the role of side groups in TTF
derivatives, other TTF derivatives which own additional sulphur atoms at
the periphery of the TTF skeleton
181
(such as TET-TTF, TCE-TTF and
BEDT-TTF) were chosen to react with transition metal ions. All these
5 Results and discussion
114
three compounds were added to a solutions of a metal salt (HAuCl
4
and
Cu(II) salts). After reacting with the latter, only BEDT-TTF resulted into
wire structures (details will be discussed in the following chapter). This
could be attributed to their different molecular structures, or more
specifically, to the presence of short side groups at C
1
, C
2
-position (-S-
CH
2
-CH
3
and -S-CH
2
-CN) for TET-TTF and TCE-TTF, respectively,
instead of a planar S-containing aromatic ring as in BEDT-TTF (Figure
5.24). The short side groups in TET-TTF and TCE-TTF had low stacking
capability, so that difficult to form wire structures.
172
Figure 5.42 shows the outcome of the reaction between TET-TTF, TCE-
TTF with HAuCl
4
and CuCl
2
, proving the formation of macro-flake
structures. Details on metal-BEDT-TTF nano-/microstructures are given
in Chapter 5.3.5.
Figure 5.42. Optical images of compounds 28, 29 and 30. a) 28 (1.5 mM of HAuCl4 and
6 mM of TET-TTF, both in CH3CN), b) 29 (1.5 mM of CuCl2 and 6 mM of TET-TTF,
both in CH3CN) and c) 30 (0.06 mM of HAuCl4 and 0.24 mM of TCE-TTF, both in
CH3CN) generated on two-inlet microchips, respectively. All the compounds formed
flake structures but no wire structures. Scale bars: 100 Pm.
5.3.5 Metal-BEDT-TTF nano-/microstructures
a) UV-Vis spectroscopy
BEDT-TTF, with planar S-containing aromatic ring frames substituted at
C
1
,C
2
-position of TTF (Figure 5.24),
176
was used to react with HAuCl
4
,
Cu(II) and Fe(III) salts and formed compounds
23
,
24
,
25
,
26
and
27
(Table 5.4). Here, a two-solvent protocol was applied, with metal salts
dissolved in CH
3
OH (0.06 mM) and BEDT-TTF in THF (0.24 mM),
because of its low solubility in CH
3
CN. As seen in Figure 5.43, the
5.3 Conductive single nanowires formed and analysed on microfluidic devices
115
absorbance shoulder at ~580 nm in all the spectra was observed and
proved the charge-transfer between metal ions and BEDT-TTF.
Figure 5.43. UV-Vis absorption spectra of solutions of M-BEDT-TTF. a) 23, BEDT-
TTF in THF, HAuCl4 in CH3OH, b) 24 and 25. c) 27 and 26. Final concentrations for
BEDT-TTF were 0.24 mM while for other solutions about 0.06 mM. Note the absorbance
at 580 nm for 23, 24, 25, 26 and 27.
b)
Morphology and electrical property of M-BEDT-TTF complexes
While M-BEDT-TTF showed rods or large wire-like structures in the case
of bulk synthesis (Figure 5.44), the structures obtained from microchips
all exhibited nano-/microwire morphology (Figures 5.45a-c, 5.46a and
5.46c). Au-BEDT-TTF (
23
) and Fe-BEDT-TTF (
27
) formed nanowires
(Figures 5.45a and 5.45c), while Cu-BEDT-TTF (
25
) resulted in
microwires (Figure 5.45b). Morphology of M-TET-TTF, M-TCE-TTF
and M-BEDT-TTF suggested that among the C
1
, C
2
-position substituted
TTF derivatives with sulphur studied in this work, the aromatic ring
5 Results and discussion
116
structure was more favourable than short linear groups for a strong
molecular interaction to form stacking structures.
172,182
Figure 5.44. Optical images of compounds 23, 24, 25, 26 and 27 formed in bulk synthesis.
a) 23 (1.73 ± 0.61Pm), b) 24 (1.79 ± 0.31 Pm), c) 25 (3.16 ± 0.82 Pm), d) 26 (1.68 ±
0.95 Pm), and e) 27 (1.29 ± 0.59 Pm) formed in bulk synthesis seperately. All of them
formed wire like structures. Scale bars: 50 Pm.
Figure 5.45. SEM images and I-V curves of structures of 23, 25 and 27. a) Au-BEDT-
5.3 Conductive single nanowires formed and analysed on microfluidic devices
117
TTF (23) (diameter: 0.56 ± 0.12 μm), b) Cu-BEDT-TTF (25) (diameter: 2.11 ± 0.53 μm)
and c) Fe-BEDT-TTF (27) (average diameter: 0.47 ± 0.11 μm) formed on the four-inlet
microchip. d), e) and f) I-V curves of single wires from 23, 25 and 27, respectively. Figure
insets in d), e) and f) show the interconnecting single nano-/microwire on
microelectrodes.
Figure 5.46. SEM images and I-V curves of structures of 24 and 26. a) 24 (average
diameter: 0.59 ± 0.14 Pm). b) I-V curve of a single nanowire of 24. c) SEM images of
26 (average diameter: 0.50 ± 0.12 Pm). d) I-V curve of a single nanowire of 26. Figure
insets of b) and d) show the alignment of single nanowires on the microelectrodes. The
gaps between the microelectrodes were 5 Pm as the scale bar. A current overload in
Figures b) and d) was observed when the low current sweep from -0.1 PA to +0.1 PA
applied with a Keithley 2612 A system source meter.
BEDT is known as an excellent electron donor and to form various charge-
transfer compounds with different electrical behaviours from insulators to
semiconductors, or also to organic metal, according to their specific
composition.
14,183
In this work, the conductivity of single M-BEDT-TTF
nano-/microwires was tested and measured by aligning single wires on
interconnecting microelectrodes, as for the other M-TTF-based
nanostructures. As shown in Figures 5.45d to 5.45f, all the structures
5 Results and discussion
118
derived from the BEDT-TTF-based complexes (
23
,
25
and
27
) exhibited
a non-linear behaviour in the current sweeping range from -1
P
A to +1
P
A, proving the conductive nature of these wires at room temperature.
However, the attempt in acquiring I-V curves of structures from
24
and
26
was not successful in this work with the source meter (Figure 5.46b and
5.46d, respectively). For the compounds
23
,
25
and
27
, the electrical
conductivity was calculated taking into account the linear part of I-V
curves (Figure 5.45) and found to be 6.3 × 10
-3
S cm
-1
, 4.8 × 10
-3
S cm
-1
and 1.2 × 10
-2
S cm
-1
,
respectively (Figure 5.47). The conductivity values
so obtained are comparable to the values reported for other Cu-BEDTTTF
powder products.
174,176
Figure 5.47. Linear I-V curves and calculated electrical conductivity of the single wires
of 23, 25 and 27. a) 23, σ = 6.3 × 10-3 S cm-1, b) 25, σ = 4.8 × 10-3 S cm-1 and c) 27, σ =
1.2× 10-2 S cm-1. Such I-V characteristics represent the linear range of the curves in
Figures 5.45d to 5.45f. σ refers to the conductivities at room temperature.
5.3 Conductive single nanowires formed and analysed on microfluidic devices
119
c) Characterization of M-BEDT-TTF wires
Due to the low product yield of reactions between metal salts and BEDT-
TTF, IR and EA analysis was not carried out. EDX-SEM spectra of
23
,
25
and
27
are shown in
Figure 5.48. Taking into account the spectrum of a
glass slide (Figure 5.48a), the presence of C, S, Au and Cl elements was
confirmed for the compound Au-BEDT-TTF (
23
) (Figure 5.48b), while
C, N, Cu and S elements for Cu-BEDT-TTF (
25
) (Figure 5.48c) and C, N,
Fe, S for Fe-BEDT-TTF (
27
) (Figure 5.48d). Thus, the relative
composition of respective metal salts and BEDT-TTF could be determined
in each complexes.
Figure 5.48. EDX-SEM spectra of M-BEDT-TTF. a) the glass slide, b) Au-BEDT-TTF
(23), c) Cu-BEDT-TTF (25) and d) Fe-BEDT-TTF (27). Note the existance of Cu and
Fe in Figures c) and d). The low intensity of N-peaks could be attributed to the low
concentration of N in the wires.
5 Results and discussion
120
5.3.6 Sensing of organic gases by single TTF-based wires
Since TTF-based charge-transfer compounds are conductive and can be applied
as sensing platforms. Following, a first experiment was carried out using single
Au-TTF hybrid wires as sensing elements for organic vapours, such as organic
acids (e.g. acetic acid) and organic thiol (e.g. dodecanethiol), which are
considered to be toxic in high concentrations. Such an experiment was
conducted in a closed chamber filled with a saturated organic vapour gas.
Single Au-TTF wires were deposited and aligned on interconnecting
microelectrode, and then exposed to different gases separately. The resistance
response was obtained by using a four-point probe and measuring for 60 min at
intervals of 10 min each. Different resistance signal (R) was observed for
different analytes. As shown in Figure 5.49, negligible changes in the
normalized resistance ratio (R/R0) (R0 represents the resistance of Au-TTF wires
before exposure to gases) were observed for the wires exposed to dodecanethiol.
However, when exposed to acetic acid, the resistance ratio of single Au-TTF
wire increased gradually after 15 min and increased after 30 min. This indicates
the higher selectivity of TTF towards acetic acid than dodecanethiol. Although
such preliminary data showed that Au-TTF is relatively slow in sensing gases
(with clear signal changes observable after 10 min), the high resistance ratio
indicated the possible use of this structure as a gas sensor. Analogously, other
TTF-based nano-/microwires are expected to exhibit similar sensing abilities, as
they all have TTF-based structures and show common features in their
conductive response. In the future, more work will be done on studying this
sensing mechanism and developing fast responding gas sensors by using and
improving different elements e.g. quantitative gas flow and closed
microchamber.
5.3 Conductive single nanowires formed and analysed on microfluidic devices
121
Figure 5.49. Sensing of organic gases by single Au-TTF wire. Normalized resistance
response ratio (R/R0) of single Au-TTF wires after exposing to two volatile organic
vapours, acetic acid and dodecanethiol separately. (Exposure time: 60 min in total, with
10 min steps). Error bars indicate the standard deviation obtained from six measurements.
In Chapters 5.1-5.3, TTF-based metal-organic charge-transfer compounds were
discussed. The sensing application and growth mechanism of Au-TTF, as well
as characteristics of various metal-TTF based complexes were well understood.
In Chapter 5.4, the synthesis and characterization of other metal-organic
complex using a TCNQ as an organic ligand will be shown and discussed.
122
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-small
microchambers for nanoelectronics
123
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in
ultra-small microchambers for nanoelectronics
Figure 5.50. Graphic abstract of the synthesis, characterization and analysis of Cu-TCNQ.
The synthesis was carried out in a 50 Pm microchamber. The obtained Cu-TCNQ was
characterized by SEM. A direct measurement of the I-V curves was achieved using the
prefabricated microelectrodes after the synthesis. The results on this project has been
disscussed and the manuscript is ready for submission.
In the field of materials science, metal-organic complexes based on TCNQ have
attracted considerable research interest, due to their intriguing electronic and
magnetic properties, and their potential applications in building advanced
conductive materials, sensors, magnetic devices, as well as energy and data
storage substrates.32,184-187 Various metal-TCNQ complexes were synthesized
and studied in the past decades, using different metals, such as Ag-TCNQ, Cu-
TCNQ, Mn-TCNQ, Fe-TCNQ, Co-TCNQ and Ni-TCNQ.23,36,38,188 Among
these charge-transfer salts, Ag-TCNQ and Cu-TCNQ are more advantageous
than the others because they can be formed by self-assembly on the
corresponding metal when exposed to TCNQ (either in gas41 or solution189
phase).
In nanomaterials and nanotechnology field, Cu-TCNQ has received great
attention after the discovery of its quasi-1D crystalline nanostructures.190,191
Because of their large surface-to-volume ratio and size effects, Cu-TCNQ
nanowires showed their advantages for building nanoelectronic devices over
their widely studied bulk and thin film counterparts. For this reason, much effort
has been put in developing new approaches for the control of the Cu-TCNQ
nanostructure growth. Reported methods fall into three categories: firstly,
solution reaction of Cu(I) precursors with TCNQ solution in CH3CN;39 secondly,
“spontaneous electrolysis” technique with the reaction between TCNQ
dissolved in CH3CN and metallic copper (Cu0);39,190 and thirdly, vapour
5 Results and discussion
124
deposition of TCNQ on Cu0 metal surfaces.22,45 However, all these techniques
required inert atmosphere since Cu0 is quickly oxidized when directly exposed
to air.192 In some cases, high temperatures (over 100 °C) were also needed. Thus,
the synthesis of Cu-TCNQ in less harsh environment at room temperature is still
challenging. In addition, precise manipulating and positioning of nanostructures
onto pre-fabricated microelectrodes is highly desirable to finely control their
alignment, since this would make the overall assembly of final nanodevices
easier. Another difficulty in the preparation of nano-sized crystal structures for
electronic devices is the reproducibility with respect to uniform morphology.
To solve the abovementioned problems, microfluidics technique was shown to
be an optimal approach, due to its possibility of a precise fluid handling,
allowing a spatially controllable formation of structures inside locally confined
microchambers. In addition, in combination with integrated electrodes in the
microchambers, a direct contact of the nanostructures with microelectrodes is
achievable without further complicated manipulation.76,93,193 Recently, the
confined synthesis and integration of Ag-TCNQ in microchambers using a two-
layer microchip have been reported.28 Ag-TCNQ was synthesized by depositing
TCNQ solution on dried Ag film, which is stable in air.
Figure 5.50 shows the general idea of the synthesis, characterization and
analysis of Cu-TCNQ. In this work, the microchamber arrays were greatly
improved for the localized formation of Cu-TCNQ nano-/microscale structures
in ultra-small microchambers. This method allows not only a continuous and
multiple chemical reactions in confined microchambers, but also an excellent
fluid handling which ensures the in situ change of trapped Cu0 nanoparticles into
Cu-TCNQ nanostructures. Furthermore, since the synthesis if carried out in a
solution phase inside the microchip, exclusion of exogenous oxygen into the
system is possible when pretreated (Ar gas bubbled) solutions are used. For the
synthesis of Cu-TCNQ structures, the spontaneous electrolysis of Cu0 reacting
with TCNQ solutions in microchambers was adopted. With the proposed
method, the morphologies of Cu-TCNQ nano-/micro structures could be
controlled by adjusting the concentration of precursors and the reaction time.
Outcomes of this method were analysed by SEM. Moreover, the conductive
property and memory effect of such synthesized Cu-TCNQ nanostructures was
tested by using integrated microelectrode arrays without further manipulation.
This work represents the first attempt to synthesize Cu-TCNQ nanostructures
by microfluidic-assisted techniques in “normal” environment at room
temperature, with direct integration into an electronic system.
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-small
microchambers for nanoelectronics
125
5.4.1 Microchips with a microchamber array
As shown in Figure 5.51, the microfluidic device was a double-layer microchip
made of PDMS via soft lithography, consisting of a top control layer with three
gas channels and a fluid layer with eight parallel microchambers, separated by
a thin and flexible PDMS membrane. The multilayer chip was assembled and
integrated to a glass slide with pre-patterned microelectrodes to form the final
device (Figures 5.51a and 5.51b). The microchambers could be operated by the
three parallel control channels upon pressurization with N2 gas (see Chapter 4.5).
Compared to the microchamber array reported for Ag-TCNQ,28 several aspects
were improved in this chip design for Cu-TCNQ. Firstly, 50 Pm wide
microchambers were fabricated to reduce the volume of solutions required to a
sub-nanolitre scale or even lower. Secondly, the microchannel towards inlet and
outlet of the fluid layer were changed in order to better control the fluids into
ultra-small chambers. Thirdly, for the electrical measurements of the structures
formed in microchambers, a four-electrode system (source high, source low,
sense high and sense low) was used in the four-point probe measurements
(Figures 5.51c), instead of the two-point method adopted for the measurement
of Ag-TCNQ. The synthesis of Cu0 in microchamber by controlling the gas
channels is shown in Figure 5.51d to 5.51f (detailed experimental procedure see
Chapter 4.5).
Figure 5.51. Microchip device for the synthesis of Cu-TCNQ structures. a) Schematic of
the alignment of the microchip; b) Photograph of the final device with two reservoirs
added to the inlets after the alignment of the microchip. A one Euro coin is shown for
5 Results and discussion
126
scale; c) Top: optical image of the microchip, with orange and red food dyes in the control
and fluidic layers, respectively; Bottom: magnification of one reaction microchamber
(50Pm in width) with aligned microelectrodes. The four electrodes used in this
experiment are shown in this image, with source high and source low electrodes outside,
sense high and sense low electrodes insides. Scale bars: 200 Pm; d)-f) Operation of gas
channels with d) the middle gas channel B pressured to 3 bar to totally close the chamber;
e) the two side gas channels A and C pressured to 3 bar to stop flows and allow the
reaction inside the microchamber (black particles started to form from the middle
indicating the formation of Cu0) and f) the channel A was released to 2 bar to allow the
slow solvent exchange in the microchamber and wash the Cu0 formed inside the chamber.
Scale bars: 100 Pm.
5.4.2 Reaction mechanism
Cu-TCNQ structures were synthesized by using Cu0 and TCNQ dissolved in
CH3CN. As shown in Scheme 5.4, Cu0 nanoparticles were produced by a two-
step reduction process. In the first step, Cu(C6H5O7)24+ was obtained by mixing
CuSO4 and C6H5O7Na3 solutions (Scheme 5.4, Equation 1). In the second step,
NaBH4 dissolved in NaOH was used to reduce Cu(C6H5O7)24+ to Cu0 (Scheme
5.4, Equation 2). After the formation of Cu0, the TCNQ solution in CH3CN was
added and, by reacting with Cu0, resulted in the formation of Cu-TCNQ, because
of the ion transfer effect between the two precursors (Scheme 5.4, Equation 3).
Scheme 5.4. Reaction mechanism for the formation of Cu0 and Cu-TCNQ.
5.4.3 Controlled synthesis of Cu and Cu-TCNQ structures
Since Cu0 can be easily to be oxidized when in contact with air,194 special
precautions were taken in order to get rid of oxygen in the microchips. First of
all, all the solutions were bubbled with argon gas for 20 min before each
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-small
microchambers for nanoelectronics
127
experiment. Then, two reservoirs with water or CH3CN were used for the two
inlets of the chambers (Figure 5.51b), with the purpose of avoiding air bubbles
into the microchannels during the experiment. With the so designed microchip,
the localized synthesis of Cu0 inside the microchamber was achieved by
adjusting the pressure into the gas channels to 3 bar (Figures 5.51d and 5.51e,
and Figure 4.3 (i) and (ii)). After the formation and washing of Cu0, the direct
synthesis of charge-transfer compound Cu-TCNQ was achieved by slightly
opening the valves to allow the diffusion of TCNQ solutions into confined
microchambers with trapped Cu0 particles (Figure 5.51f and Figure 4.3 (v), see
Chapter 4.5 for further details).
The copper formed via this method appeared brown in colour and shiny as a
metal. As observed by eye on a glass slide (Figure 5.52a), as well as under an
optical microscope on a glass slide (Figure 5.52c) and in a microchannel
(Figures 5.52d and 5.52e). This indicates the presence of metal copper. After
the reaction with TCNQ solutions on a glass slide, a dark blue product was
obtained (Figure 5.52b), indicating the formation of Cu-TCNQs. The wire
morphologies were observed under microscope in a microchannel (Figure 5.52f).
Figure 5.52. Photograph of Cu0 and Cu-TCNQ. a) Cu0 right after its formation on a non-
5 Results and discussion
128
bonded chip by direct injection of Cu2+ and NaBH4 solutions; b) Cu-TCNQ after
dropping of TCNQ solutions on Cu0 film formed on a non-bonded chip. Optical images
of c) Cu0 formed on a non-bonded chip after dried in vacuum; d) in a microchamber
(reduced from 250 mM Cu2+) and e) in a microchamber (reduced from 50 mM Cu2+); f)
dark blue Cu-TCNQ wires formed in microchannel. Scale bars: 50 m. Note that the
brown Cu0 in Figure a) turned into black after exposure to air in tens of seconds,
indicating the fast oxidation of Cu0.
With the microfluidic techniques, the density of Cu0 layer and morphologies of
Cu-TCNQ could be formed in a controlled way, for example, by using different
concentration of precursors and controlling the reaction time (Table 5.9). High
concentration of Cu(C6H5O7)24+/NaBH4 (250 mM/ 500 mM) resulted in a thick
Cu0 layer (Figure 5.53a). Upon reaction with 5 mM TCNQ at a flow rate of 2
L/min for 10 min, dense short nano-/microrods (0.4-1 m in diameter, 5-10
m in length) were observed (Figures 5.53b and 5.53c). However, five times
diluted Cu(C6H5O7)24+/NaBH4 solutions (50 mM/100 mM) resulted in large
particles of Cu0, followed by the formation of long nano-/microrods after
reacting with 5 mM TCNQ for 10 min (Figures 5.53d-5.53f, structures with 0.8-
1.5 m in diameter, 10-50 m in length). In contrast, a very low concentration
of the Cu(C6H5O7)24+/NaBH4 solution (25 mM/50 mM) was used and led to the
formation of a thin Cu0 layer in the channel (Figure 5.53g), after reacting with
low concentrated TCNQ solution (2 mM) for 10 min. In this case, bundles of
nanowire structures were observed (Figure 5.53h), as also seen by SEM results
(Figures 5.54a and 5.54b). In addition, a longer reaction time (20 min) caused
the formation of larger structures (Figure 5.53i), compared to those in Figure
5.53h. This may due to the aggregation of wire structures.190 Both the
composition of Cu0 and resulted Cu-TCNQ structures were determined by EDX
(Figures 5.54c and 5.54d). Using high concentrations of Cu(C6H5O7)24+/NaBH4
(250 mM/500 mM), bubbles were oberserved in the channel, likely due to the
release of hydrogen in the redox reaction between NaBH4 and Cu(C6H5O7)24+.
Instead, this was not observed when low concentrated solutions were used. Thus,
low concentration of precursors (Cu(C6H5O7)24+/NaBH4, 25 mM/50 mM) and
short reaction time (10 min) are ideal for the synthesis of nanometer scaled Cu-
TCNQ structures.
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-small
microchambers for nanoelectronics
129
Table 5.9. Morphology of structures formed under different reaction conditions on
microchip
Cu(C
6H5O7)24+/
NaBH4
Concentration
TCNQ
concentration
Reaction
time
Morphology
of products*
Φ diameter, L length
250 mM/ 500 mM
5 mM
10 min
Short nano-/microrods
Φ 0.4-1
P
m. L 5-10
P
m
50 mM/ 100 mM
5 mM
10 min
Long nano-/microrods
Φ 0.8-1.5 Pm. L 10-50 P
m
25 mM/ 50 mM
2 mM
10 min
Nanowire bundles
Φ 0.1-0.8
P
m. L 5-20
P
m
25 mM/ 50 mM
2 mM
20 min
Microrods
* The optical and SEM images are shown in Figures 5.53 and 5.54.
Figure 5.53. The Cu0 layer and Cu-TCNQ formed by using different concentrations of
Cu2+ and reductant. a)-c) represents the a) Cu0 layer reduced from 250 mM
Cu(C6H5O7)24+/500 mM NaBH4 b) Cu-TCNQ in microchamber after reaction with 5 mM
5 Results and discussion
130
TCNQ at flow rate of 2 PL/min for 10 min, and c) SEM image of Cu-TCNQ (short nano-
/microrods, 0.4-1 Pm in diameter, 5-10 Pm in length). d) Cu0 layer reduced from 50 mM
Cu(C6H5O7)24+/100 mM NaBH4 e) Cu-TCNQ in microchamber after reaction with 5 mM
TCNQ at flow rate of 2 PL/min for 10 min, and f) SEM image of Cu-TCNQ (long nano-
/micro-rods, 0.8-1.5 Pm in diameter, 10-50 Pm in length). g) Cu0 layer reduced from 25
mM Cu(C6H5O7)24+/50 mM NaBH4 h) Cu-TCNQ in microchamber after reaction with 2
mM TCNQ at flow rate of 2 PL/min for 10 min, and i) after 20 min.
5.4.4 Characterization of Cu-TCNQ nanostructures
a)
SEM and EDX spectroscopy
Both the composition of the copper layer and formed Cu-TCNQ
nanowires were determined by EDX. Compared to the spectrum of Cu
0
(Figure 5.54c), the appearance of N and the higher concentration of C
proved that TCNQ (molecular formula C
12
H
4
N
4
) is part of the resulted
nanowires (Figure 5.54d). Combined with the Cu signal, it is evident that
the nanowires are composed of Cu and TCNQ.
Figure 5.54. SEM images and EDX spectra of Cu-TCNQ nanostructures. The SEM
images of a) Cu-TCNQ nanowire bundles (100-800 nm in diameter, 5-20 Pm in length)
and b) a wire bundle; c) The EDX spectrum of c) Cu layer (notice the strong Cu element
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-small
microchambers for nanoelectronics
131
signal), and d) Cu-TCNQ nanowire bundles. The strong O signals in the EDX spectra
mainly obtained from glass slide (Figure 5.57a) and partially from the oxidation of
samples upon exposure to air.
b) UV-Vis absorption
The UV-Vis spectra of neutral TCNQ and Cu-TCNQ were obtained, with
the latter obtained from eluted solution of a non-bonded microchip after
the synthesis. As indicated in Figure 5.55, neutral TCNQ showed only one
peak at 394 nm. In contrast, Cu-TCNQ solution exhibited a broad peak at
416 nm and absorption peaks at 746 nm and 840 nm, which could be
assigned to TCNQ anion radicals in the sample.
190
Figure 5.55. UV-Vis absorption spectra of TCNQ (black) and Cu-TCNQ (red) solutions
in CH3CN.
c) FT-IR and Raman spectroscopy
In reported studies, two polymorphs of Cu-TCNQ have been identified:
one with quasi-1D structures and high electric conductivity (phase I), the
other with pallets morphology and low conductivity (phase II).
39
These
two phases could be identified by their distinct morphology, IR spectra
and electronic characteristics.
190
In this work, Cu-TCNQ crystals showed
5 Results and discussion
132
wire-like nano-/microstructures, suggesting the formation of Cu-TCNQ in
phase I.
FT-IR and Raman measurements were carried out in order to determine
the oxidation state of TCNQ molecules in neutral TCNQ and Cu-TCNQ
nanostructures (Figure 5.56). As can be seen from the IR spectra (Figure
5.56a), compared to the spectrum of pure TCNQ, the C=C-H mode of Cu-
TCNQ shifted from 861 cm
-1
to 825 cm
-1
. In addition, the C=C ring
stretching peak at 1543 cm
-1
split into two peaks at 1506 cm
-1
and 1577
cm
-1
. Typically, the peak at 1506 cm
-1
denoted the formation of TCNQ
anion radicals.
190
The C=C wing-stretching mode was also represented by
the peak at 1354 cm
-1
. Moreover, the strong band at ~2200 cm
-1
was
associated to the C≡N stretching mode. Here, the strong absorption at
2200 cm
-1
with a small shoulder peak at about 2172 cm
-1
proved the
formation of Cu-TCNQ (phase I), because the phase II would exhibit two
strong sharp peaks.
39
The Raman spectral analysis of Cu-TCNQ (Figure 5.56b) showed the
main characteristic vibration modes at 1205 cm
-1
(C=C-H bending), 1377
cm
-1
(C–CN stretching), 1606 cm
-1
(C=C ring stretching), and 2209 cm
-1
(C≡N stretching). Compared to the Raman spectrum of neutral TCNQ
crystal, the C–CN stretching mode in Cu-TCNQ shifts from 1455 cm
-1
to
1377 cm
-1
, because of complete charge-transfer between Cu
0
and TCNQ
molecules.
194,195
Thus, the FT-IR and Raman spectra clearly indicate the
formation of Cu-TCNQ (phase I) in these nanostructures.
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-small
microchambers for nanoelectronics
133
Figure 5.56. FT-IR and Raman analysis of TCNQ and Cu-TCNQ. a) FT-IR and b) Raman
spectra of TCNQ and Cu-TCNQ nanowires obtained from non-bonded microchips at
concentration of Cu(C6H5O7)24+/NaBH4 (25 mM/ 50 mM) after reacting with 2 mM
TCNQ.
5 Results and discussion
134
5.4.5 Characterization of Cu-TCNQ microstructures
The larger microstructures formed at higher concentration of precursors
(Figures 5.54e and5.54f) were also characterized by EDX, FT-IR and Raman
analysis. Compared to the EDX spectrum of glass slide (Figure 5.57a), the
spectrum of the microstructures clearly showed the compositions of Cu-TCNQ
with C, N and Cu (Figure 5.57b). FT-IR and Raman spectra of Cu-TCNQ
microstructures (Figure 5.58) showed charateristics similar to those observed
for nanowires (Figure 5.56), and suggested a similar composition of these
structures. Thus, these results clearly proved the formation of Cu-TCNQ (phase
I) microwire structures using this in situ synthesis technique.
Figure 5.57. EDX-SEM spectra of a) glass slide and b) Cu-TCNQ microstructures.
Compared to a), the spectrum b) clearly indicates the increase in C and presence of N
and Cu in Cu-TCNQ.
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-small
microchambers for nanoelectronics
135
Figure 5.58. FT-IR and Raman characterization of Cu-TCNQ microsctructures. a) FT-IR
and b) Raman spectra of Cu-TCNQ microstructures. Samples used here were synthesized
by the reaction of 5 mM TCNQ with a thick Cu0 layer (250 mM Cu(C6H5O7)24+/ 500 mM
NaBH4).
5.4.6 Electrical properties of Cu-TCNQ nano-/microstructures
To confirm the possibility of a direct integration and to characterize the
electrical properties of Cu-TCNQ nanostructures, the I-V curves of Cu0 layer
5 Results and discussion
136
before and after reacting with TCNQ solutions were obtained on-chip by four-
point probe measurements after solvent evaporation in vacuum. The graph in
Figure 5.59a shows a linear I-V sweep for the copper layer, indicating its typical
metallic nature. In contrast, after the formation of nanostructures inside different
microchambers, non-linear I-V curves were obtained (Figure 5.59b). The quasi-
symmetric characteristics of the I-V curves for wire bundles indicated the
formation of Cu-TCNQ crystals and the good contact with the prefabricated
microelectrodes. Reproducible I-V responses indicated the possible application
of this technique for parallelized integration of conductive nanowires. Moreover,
the electrical properties of the synthesized nanowire devices were proven to be
stable in air. By applying a voltage from 0 V to 10 V and back to 0 V, reversible
hysteresis I-V curves were observed (Figure 5.59c). The two switchable states
were confirmed and attributed to low and high conductivity states, or “OFF”
and “ON” states, respectively (Equation 5.2). A detailed discussion of the “ON”
and “OFF” states for M-TCNQ can be found in Chapter 1.4 and literature45. The
reversible hysteresis I-V curves proved the good memory behaviour of Cu-
TCNQ nanowire bundles. The conductive behaviour of larger microstructures
was also tested (Figure 5.59d) and the nonlinearity of the I-V curve provided an
evidence that the Cu-TCNQ was in the conductive phase I.
5.4 Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-small
microchambers for nanoelectronics
137
Figure 5.59. I-V characteristics of Cu0 and Cu-TCNQ at room-temperature. a) Cu0 film
and b) multiple sweeps of different integrated Cu-TCNQ nanowire bundles in different
microchambers. c) I-V curves recorded for Cu-TCNQ nanowire bundles showing
hysteresis in the electrical response after sweeping the applied voltage from 0 V to 10 V.
Forward voltage curve relates to the “ON” state, while reverse voltage curve relates to
the “OFF” state. d) I-V curve of Cu-TCNQ microstructures. The non-linear
characteristics indicates the formation of Cu-TCNQ crystals of conductive phase I.
(5.2)
The reversible conductivity response of the Cu-TCNQ nano-/microstructures
obtained in this work showed consistency with other reported work.45,195 This
proved also the efficiency of the microfluidic-based synthesis used here. In
addition, the so proposed method demonstrated to achieve a well-control of
conductive Cu-TCNQ (phase I) structures in ambient environment, showing the
advantage of using microfluidic techniques, instead of the conventional bulk
synthesis.
138
6 Conclusions and outlook
139
6 Conclusions and outlook
To summarize, various metal-organic nano-/microstructures were
synthesized on microfluidic systems and their optical and structural
properties and their use in sensing applications were investigated.
Regarding the first part of this project “Label-free biosensors based on
in situ formed and functionalized microwires in microfluidic devices”
(Chapter 5.1), an integrated Au-TTF wire-based microfluidic system
with sensitive label-free biosensing capability was reported. With this
method, the direct sensing of catecholamines involving dopamine and
the indirect sensing of human IgG were tested by Raman measurements,
both with a detection limit comparable to the one recently reported in
other studies,
148,155
but with the advantage in this work to conduct all the
measurements on single wires. However, the so proposed biosensors
could not distinguish dopamine from its carbon-hydroxylated product
norepinephrine. This is likely attributed to the high similarity in
molecular structure of these catecholamines, so that their Raman spectra
overlap. This has been found also in previous studies by Raman
spectroscopy.
149
However, the two approaches reported here are of
general interest for the synthesis and characterization of various metal-
organic structures with an active surface Moreover, this work opens a
new route to sensing devices based on in situ formed micro- and
nanowires with an integrated microfluidic delivering system. In the
future, by optimising the alignment and positioning of the wires formed
inside the microchip, single micro-/nanowire sensing systems could also
be fabricated and used as advanced sensing tools for chemicals and
biomolecules.
In the frame of the second part “Study of molecular self-organization in
TTF crystals individual Au-TTF nano-/microwire by polarized confocal
Raman spectroscopy” (Chapter 5.2), the orientation and self-assembly
of molecules in both a single TTF crystal and Au-TTF nanowire were
studied. The non-destructive polarized confocal Raman spectroscopic
analysis on a sing TTF crystal proved the preferential ordering of
molecules inside the crystal. Supported by the DFT calculations and
6 Conclusions and outlook
140
simulations of depolarization ratio, the angle between two molecules in
a single TTF unit cell was calculated and showed little deviation from
that calculated from single crystal X-ray diffraction data. In addition, a
single Au-TTF nanowire was studied with the same technique. The
result obtained indicated that there were two different molecule
configurations in the wire, both tilted with respected to the long axis of
wire.
In the frame of the third part “
Conductive single nanowires formed and
analysed on microfluidic devices” (Chapter 5.3)
, a comprehensive overview
of different metal and TTF-based compounds (Table 5.4) was presented
and highlighted the influence of the different preparation conditions
(such as bulk (non-flow) vs laminar flow conditions) on the resulting
morphology. The data obtained proved four-inlet microchip with
laminar flow and solvent sheath streams worked efficiently to obtain
long and thin nanowires. The conductive behaviour of single nanowires
at room temperature proved by electrical characterization and suggested
the use of such materials for fabricating nanodevices. A comparison
between M-TTF and M-TTF derivatives indicated that M-TTF, M-FTTF
and M-BEDT-TTF can lead to the formation of wire structures, while
M-TET-TTF and M-TCE-TTF can only form crystals. Thus, for C
1
, C
2
-
position substituted TTF derivatives with sulphur, the planar aromatic
ring structure in BEDT-TTF can form stacking structures, much easier
than those with short side groups (TET-TTF and TCE-TTF) under the
reaction conditions described in this work. Among the M-TTF (and
derivatives) systems studied here, the nanowires derived from M-TTF
and M-FTTF exhibited high electrical conductivities from 10
-1
to 10 S
cm
-1
at room temperature, one or two orders higher than those for M-
BEDT-TTF, in accordance with previously reported work.
10,13,174
The
resulting overview table (Table 5.4) is a means to select TTF-based
structures with desired structural and electrical properties, e.g. for
sensing applications with a high conductivity. In the future, the use of
different TTF-based wire structures for sensing of various analytes e.g.
volatile organic compounds, cancer biomarkers could be studied and
compared to the sensing platforms proposed so far.
Finally, a second organic ligand TCNQ was discussed in the last part
“Localized synthesis of conductive Cu-TCNQ nanostructures in ultra-small
6 Conclusions and outlook
141
microchambers for nanoelectronics” (Chapter 5.4). In particular the
microfluidic-based synthesis of conductive Cu-TCNQ in ambient
environment at room temperature was achieved. Different morphologies of
Cu-TCNQ nano-/microstructures could be synthesized in a controlled way by
tuning the concentration of precursors and by changing the reaction time
inside microchips. With this technique, reactions were confined in parallel
ultra-small microchambers. Thus, the nanostructures were locally formed and
integrated with prefabricated microelectrodes, and moreover, their electrical
properties could be measured directly, without further micromanipulation. Cu-
TCNQ nanostructures formed with the proposed method showed a conductive
behaviour and a memory effect, indicating the possibility of using this
technique for the building of nanoelectronics and nanoswitches in an easy and
efficient way.
In general, each of these four subprojects addresses one or more aspects of the
overall topic, i.e., firstly, to find suitable materials for nanowire sensors
(Chapter 5.3 and 5.4); secondly, to characterize and understand how the
structure is (Chapter 5.2 and 5.3); thirdly, to locally synthesize the wires
(Chapter 5.1 and 5.4); lastly, to functionalize them for application as a sensor
(Chapter 5.1). By this work, the formation mechanism, synthesis technique,
structural properties of TTF and TCNQ-based metal-organic charge-transfer
complexes were better understood. However, the synthesis and application of
these structures have not been fully investigated, especially for the sensing
applications based on conductive wires. In the future, deeper insight into three
aspects can be gained: 1) molecular self-organization in TCNQ-based
nanostructures, such as Ag-TCNQ and Cu-TCNQ; 2) synthesis of novel metal-
organic compounds using different transition metals and bio-ligands; 3)
development of novel electrical or optical sensors based on metal-TTF and
metal-TCNQ structures widely characterized in this work. In particular,
preliminary result in Chapter 5.3.6 had shown the sensing behaviour of
conductive Au-TTF for organic gases. A further study in developing gas
sensors using TTF/TCNQ-based conductive wires will be very important in
building label-free sensing platforms with miniaturized size and high
sensitivity.
142
Bibliography
143
Bibliography
[1] Wudl, F.; Smith, G. M.; Hufnagel, E. J. Bis-1,3-dithiolium chloride: an
unusually stable organic radical cation, Journal of the Chemical Society D: Chemical
Communications 1970, 1453-1454.
[2] Acker, D. S.; Harder, R. J.; Hertler, W. R.; Mahler, W.; Melby, L. R.; Benson,
R. E.; Mochel, W. E. 7,7,8,8-TETRACYANOQUINODIMETHANE AND ITS
ELECTRICALLY CONDUCTING ANION-RADICAL DERIVATIVES, Journal of the
American Chemical Society 1960, 82, 6408-6409.
[3] Coleman, L. B.; Cohen, M. J.; Sandman, D. J.; Yamagishi, F. G.; Garito, A. F.;
Heeger, A. J. Superconducting fluctuations and the peierls instability in an organic solid,
Solid State Communications 1973, 12, 1125-1132.
[4] Ferraris, J.; Cowan, D. O.; Walatka, V.; Perlstein, J. H. Electron transfer in a
new highly conducting donor-acceptor complex, Journal of the American Chemical
Society 1973, 95, 948-949.
[5] Mori, T.; Kawamoto, T. Organic conductors-from fundamentals to nonlinear
conductivity, Annual Reports Section "C" (Physical Chemistry) 2007, 103, 134-172.
[6] Temkin, H.; Fitchen, D. B.; Wudl, F. Raman scattering in TTF crystals, Solid
State Communications 1977, 24, 87-92.
[7] Pou-Amérigo, R.; Ortí, E.; Merchán, M.; Rubio, M.; Viruela, P. M. Electronic
Transitions in Tetrathiafulvalene and Its Radical Cation:  A Theoretical Contribution,
The Journal of Physical Chemistry A 2002, 106, 631-640.
[8] Pou-Amérigo, R.; Viruela, P. M.; Viruela, R.; Rubio, M.; Ortı, E. Electronic
spectra of tetrathiafulvalene and its radical cation: analysis of the performance of the
time-dependent DFT approach, Chemical Physics Letters 2002, 352, 491-498.
[9] Nielsen, M. B.; Lomholt, C.; Becher, J. Tetrathiafulvalenes as building blocks
in supramolecular chemistry II, Chemical Society Reviews 2000, 29, 153-164.
[10] Siedle, A. R.; Candela, G. A.; Finnegan, T. F.; Vanduyne, R. P.; Cape, T.;
Kokoszka, G. F.; Woyciejes, P. M.; Hashmall, J. A. Copper and Gold
Metallotetrathiaethylenes, Inorganic Chemistry 1981, 20, 2635-2640.
[11] Honda, M.; Katayama, C.; Tanaka, J.; Tanaka, M. Structure of the 2/1 complex
dibenzotetrathiafulvalenium hexachlorocuprate(II), 2C14H8S4+.Cu2Cl62−, Acta
Crystallographica Section C 1985, 41, 197-199.
Bibliography
144
[12] Kim, Y. I.; Hatfield, W. E. Electrical, magnetic and spectroscopic properties
of tetrathiafulvalene charge transfer compounds with iron, ruthenium, rhodium and
iridium halides, Inorganica Chimica Acta 1991, 188, 15-24.
[13] Inoue, M. B.; Inoue, M.; Fernando, Q.; Nebesny, K. W. Highly
electroconductive tetrathiafulvalenium salts of copper halides, Inorganic Chemistry 1986,
25, 3976-3980.
[14] Williams, J. M.; Emge, T. J.; Wang, H. H.; Beno, M. A.; Copps, P. T.; Hall, L.
N.; Carlson, K. D.; Crabtree, G. W. Synthetic metals based on
bis(ethylenedithio)tetrathiafulvalene (BEDT-TTF): synthesis, structure, and ambient-
pressure superconductivity in (BEDT-TTF)2I3, Inorganic Chemistry 1984, 23, 2558-
2560.
[15] Ouahab, L. Organic/Inorganic Supramolecular Assemblies and Synergy
between Physical Properties, Chemistry of Materials 1997, 9, 1909-1926.
[16] Wudl, F.; Wobschall, D.; Hufnagel, E. J. Electrical conductivity by the bis(1,3-
dithiole)-bis(1,3-dithiolium) system, Journal of the American Chemical Society 1972, 94,
670-672.
[17] Scott, B. A.; La Placa, S. J.; Torrance, J. B.; Silverman, B. D.; Welber, B. The
crystal chemistry of organic metals. Composition, structure, and stability in the
tetrathiafulvalinium-halide systems, Journal of the American Chemical Society 1977, 99,
6631-6639.
[18] Dahm, D. J.; Horn, P.; Johnson, G. R.; Miles, M. G.; Wilson, J. D.
Classification of electrically conducting ‘electron-transfer’ compounds by crystal
structure, Journal of Crystal and Molecular Structure, 5, 27-34.
[19] Phillips, T. E.; Kistenmacher, T. J.; Ferraris, J. P.; Cowan, D. O. Crystal
structure of the radical-cation radical-anion salt from 2,2[prime or minute]-bi-1,3-
dithiole and 7,7,8,8-tetracyanoquinodimethane, Journal of the Chemical Society,
Chemical Communications 1973, 471-472.
[20] Kistenmacher, T. J.; Phillips, T. E.; Cowan, D. O. The crystal structure of the
1:1 radical cation-radical anion salt of 2,2'-bis-l,3-dithiole (TTF) and 7,7,8,8-
tetracyanoquinodimethane (TCNQ), Acta Crystallographica Section B 1974, 30, 763-
768.
[21] Furukawa, H.; Cordova, K. E.; O’Keeffe, M.; Yaghi, O. M. The Chemistry and
Applications of Metal-Organic Frameworks, Science 2013, 341.
[22] Zheng, K. B.; Shen, H. T.; Ye, C. N.; Li, J. L.; Sun, D. L.; Chen, G. R. The
electrical switching characteristics of single copper tetra-cyanoquinodimethane nanowire,
Nano-Micro Letters 2009, 1, 23-26.
[23] Nafady, A.; Bond, A. M.; Bilyk, A.; Harris, A. R.; Bhatt, A. I.; O'Mullane, A.
P.; De Marco, R. Tuning the Electrocrystallization Parameters of Semiconducting
Bibliography
145
Co[TCNQ]2- Based Materials To Yield either Single Nanowires or Crystalline Thin
Films, Journal of the American Chemical Society 2007, 129, 2369-2382.
[24] Naka, K.; Ando, D.; Wang, X.; Chujo, Y. Synthesis of Organic-Metal Hybrid
Nanowires by Cooperative Self-Organization of Tetrathiafulvalene and Metallic Gold
via Charge-Transfer, Langmuir 2007, 23, 3450-3454.
[25] Tanaka, K.; Matsumoto, T.; Ishiguro, F.; Chujo, Y. Conductivity regulation of
the mixed-valence tetrathiafulvalene nanowire/poly(methyl methacrylate) composites
using heterogeneous tetrathiafulvalene derivatives, Journal of Materials Chemistry 2011,
21, 9603-9607.
[26] Wang, X.; Itoh, H.; Naka, K.; Chujo, Y. Tetrathiafulvalene-Assisted Formation
of Silver Dendritic Nanostructures in Acetonitrile, Langmuir 2003, 19, 6242-6246.
[27] Puigmarti-Luis, J.; Schaffhauser, D.; Burg, B. R.; Dittrich, P. S. A Microfluidic
Approach for the Formation of Conductive Nanowires and Hollow Hybrid Structures,
Advanced Materials 2010, 22, 2255-2259.
[28] Cvetkovic, B. Z.; Puigmarti-Luis, J.; Schaffhauser, D.; Ryll, T.; Schmid, S.;
Dittrich, P. S. Confined Synthesis and Integration of Functional Materials in Sub-
nanoliter Volumes, ACS Nano 2013, 7, 183-190.
[29] Kuhn, P.; Puigmarti-Luis, J.; Imaz, I.; Maspoch, D.; Dittrich, P. S. Controlling
the length and location of in situ formed nanowires by means of microfluidic tools, Lab
on a Chip 2011, 11, 753-757.
[30] Taliaferro, M. L.; Palacio, F.; Miller, J. S. The magnetic behaviors of the
metamagnetic and ferromagnetic phases of [Fe(C5Me5)2][TCNQ] (TCNQ = 7,7,8,8-
tetracyano-p-quinodimethane). Determination of the phase diagram for the
metamagnetic phase, Journal of Materials Chemistry 2006, 16, 2677-2684.
[31] Mo, X.-L.; Chen, G.-R.; Cai, Q.-J.; Fan, Z.-Y.; Xu, H.-H.; Yao, Y.; Yang, J.;
Gu, H.-H.; Hua, Z.-Y. Preparation and electrical/optical bistable property of potassium
tetracyanoquinodimethane thin films, Thin Solid Films 2003, 436, 259-263.
[32] Teresa Azcondo, M.; Ballester, L.; Golhen, S.; Gutierrez, A.; Ouahab, L.;
Yartsev, S.; Delhaes, P. Structural, magnetic, electrical and optical characterization of
systems built from [M([9]aneN3)2]2+ (M=CuII or NiII) and TCNQ or TCNQF4, Journal
of Materials Chemistry 1999, 9, 1237-1244.
[33] Wooster, T. J.; Bond, A. M. Ion selectivity obtained under voltammetric
conditions when a TCNQ chemically modified electrode is presented with aqueous
solutions containing tetraalkylammonium cations, Analyst 2003, 128, 1386-1390.
[34] Bartley, S. L.; Dunbar, K. R. [{Re2Cl4(dppm)2}2 (μ-TCNQ)], a Novel Charge
Transfer Complex Derived from a Donor with a Metal-Metal Multiple Bond,
Angewandte Chemie International Edition in English 1991, 30, 448-450.
Bibliography
146
[35] Oshio, H.; Ino, E.; Mogi, I.; Ito, T. A weak antiferromagnetic interaction
between manganese(2+) centers through a TCNQ column: crystal structures and magnetic
properties of [MnII(tpa)(TCNQ)(CH3OH)](TCNQ)2.CH3CN, [MnII(tpa)(-
O2CCH3)]2(TCNQ)2.2CH3CN, and [MnII(tpa)(NCS)2].CH3CN (tpa = tris(2-
pyridylmethyl)amine), Inorganic Chemistry 1993, 32, 5697-5703.
[36] Zhao, H.; Heintz, R. A.; Dunbar, K. R.; Rogers, R. D. Unprecedented Two-
Dimensional Polymers of Mn(II) with TCNQ-• (TCNQ = 7,7,8,8-
Tetracyanoquinodimethane), Journal of the American Chemical Society 1996, 118,
12844-12845.
[37] Torrance, J. B. An Overview of Organic Charge-Transfer Solids: Insulators,
Metals, and the Neutral-Ionic Transition, Molecular Crystals and Liquid Crystals 1985,
126, 55-67.
[38] Potember, R. S.; Poehler, T. O.; Cowan, D. O. Electrical switching and
memory phenomena in Cu‐TCNQ thin films, Applied Physics Letters 1979, 34, 405-407.
[39] Heintz, R. A.; Zhao, H.; Ouyang, X.; Grandinetti, G.; Cowen, J.; Dunbar, K.
R. New Insight into the Nature of Cu(TCNQ):  Solution Routes to Two Distinct
Polymorphs and Their Relationship to Crystalline Films That Display Bistable Switching
Behavior, Inorganic Chemistry 1999, 38, 144-156.
[40] Huibiao, L.; Junbo, L.; Changshi, L.; Changshui, H.; Yuliang, L.; Zhong Lin,
W.; Daoben, Z. Morphological tuning and conductivity of organic conductor nanowires,
Nanotechnology 2007, 18, 495704.
[41] Liu, H.; Zhao, Q.; Li, Y.; Liu, Y.; Lu, F.; Zhuang, J.; Wang, S.; Jiang, L.; Zhu,
D.; Yu, D.; Chi, L. Field Emission Properties of Large-Area Nanowires of Organic
Charge-Transfer Complexes, Journal of the American Chemical Society 2005, 127,
1120-1121.
[42] Kepler, R. G.; Bierstedt, P. E.; Merrifield, R. E. Electronic Conduction and
Exchange Interaction in a New Class of Conductive Organic Solids, Physical Review
Letters 1960, 5, 503-504.
[43] Sato, C.; Wakamatsu, S.; Tadokoro, K.; Ishii, K. Polarized memory effect in
the device including the organic charge‐transfer complex, copper‐
tetracyanoquinodimethane, Journal of Applied Physics 1990, 68, 6535-6537.
[44] Kamitsos, E. I.; Tzinis, C. H.; Risen Jr, W. M. Raman study of the mechanism
of electrical switching in Cu TCNQ films, Solid State Communications 1982, 42, 561-
565.
[45] Müller, R.; De Jonge, S.; Myny, K.; Wouters, D. J.; Genoe, J.; Heremans, P.
Organic CuTCNQ integrated in complementary metal oxide semiconductor copper back
end-of-line for nonvolatile memories, Applied Physics Letters 2006, 89, 223501.
Bibliography
147
[46] Law, M.; Goldberger, J.; Yang, P. SEMICONDUCTOR NANOWIRES AND
NANOTUBES, Annual Review of Materials Research 2004, 34, 83-122.
[47] Kong, J.; Franklin, N. R.; Zhou, C.; Chapline, M. G.; Peng, S.; Cho, K.; Dai,
H. Nanotube Molecular Wires as Chemical Sensors, Science 2000, 287, 622-625.
[48] Han, S. W.; Lee, S.; Hong, J.; Jang, E.; Lee, T.; Koh, W.-G. Mutiscale
substrates based on hydrogel-incorporated silicon nanowires for protein patterning and
microarray-based immunoassays, Biosensors and Bioelectronics 2013, 45, 129-135.
[49] Lee, M.; Baik, K. Y.; Noah, M.; Kwon, Y. K.; Lee, J. O.; Hong, S. Nanowire
and nanotube transistors for lab-on-a-chip applications, Lab on a Chip 2009, 9, 2267-
2280.
[50] Cui, Y.; Wei, Q.; Park, H.; Lieber, C. M. Nanowire Nanosensors for Highly
Sensitive and Selective Detection of Biological and Chemical Species, Science 2001, 293,
1289-1292.
[51] Puigmarti-Luis, J.; Stadler, J.; Schaffhauser, D.; del Pino, A. P.; Burg, B. R.;
Dittrich, P. S. Guided assembly of metal and hybrid conductive probes using floating
potential dielectrophoresis, Nanoscale 2011, 3, 937-940.
[52] Hoeben, F. J. M.; Jonkheijm, P.; Meijer, E. W.; Schenning, A. P. H. J. About
Supramolecular Assemblies of π-Conjugated Systems, Chemical Reviews 2005, 105,
1491-1546.
[53] Gan, H.; Liu, H.; Li, Y.; Zhao, Q.; Li, Y.; Wang, S.; Jiu, T.; Wang, N.; He, X.;
Yu, D.; Zhu, D. Fabrication of Polydiacetylene Nanowires by Associated Self-
Polymerization and Self-Assembly Processes for Efficient Field Emission Properties,
Journal of the American Chemical Society 2005, 127, 12452-12453.
[54] Clayton, D. A.; McPherson, T. E.; Pan, S.; Chen, M.; Dixon, D. A.; Hu, D.
Spatial and temporal variation of surface-enhanced Raman scattering at Ag nanowires in
aqueous solution, Physical Chemistry Chemical Physics 2013, 15, 850-859.
[55] Park, K. S.; Cho, B.; Baek, J.; Hwang, J. K.; Lee, H.; Sung, M. M. Single-
Crystal Organic Nanowire Electronics by Direct Printing from Molecular Solutions,
Advanced Functional Materials 2013, 23, 4776-4784.
[56] Chen, X.; Kis, A.; Zettl, A.; Bertozzi, C. R. A cell nanoinjector based on carbon
nanotubes, Proceedings of the National Academy of Sciences 2007, 104, 8218-8222.
[57] Kemp, N. T.; McGrouther, D.; Cochrane, J. W.; Newbury, R. Bridging the Gap:
Polymer Nanowire Devices, Advanced Materials 2007, 19, 2634-2638.
[58] Puigmartí-Luis, J.; Rubio-Martínez, M.; Hartfelder, U.; Imaz, I.; Maspoch, D.;
Dittrich, P. S. Coordination Polymer Nanofibers Generated by Microfluidic Synthesis,
Journal of the American Chemical Society 2011, 133, 4216-4219.
Bibliography
148
[59] Imaz, I.; Rubio-Martínez, M.; Saletra, W. J.; Amabilino, D. B.; Maspoch, D.
Amino Acid Based Metal−Organic Nanofibers, Journal of the American Chemical
Society 2009, 131, 18222-18223.
[60] Rieter, W. J.; Taylor, K. M. L.; An, H.; Lin, W.; Lin, W. Nanoscale
Metal−Organic Frameworks as Potential Multimodal Contrast Enhancing Agents,
Journal of the American Chemical Society 2006, 128, 9024-9025.
[61] Hou, S.; Harrell, C. C.; Trofin, L.; Kohli, P.; Martin, C. R. Layer-by-Layer
Nanotube Template Synthesis, Journal of the American Chemical Society 2004, 126,
5674-5675.
[62] Li, Y.; Yang, X.-Y.; Feng, Y.; Yuan, Z.-Y.; Su, B.-L. One-Dimensional Metal
Oxide Nanotubes, Nanowires, Nanoribbons, and Nanorods: Synthesis, Characterizations,
Properties and Applications, Critical Reviews in Solid State and Materials Sciences 2012,
37, 1-74.
[63] Xia, Y.; Yang, P.; Sun, Y.; Wu, Y.; Mayers, B.; Gates, B.; Yin, Y.; Kim, F.;
Yan, H. One-Dimensional Nanostructures: Synthesis, Characterization, and Applications,
Advanced Materials 2003, 15, 353-389.
[64] Spokoyny, A. M.; Kim, D.; Sumrein, A.; Mirkin, C. A. Infinite coordination
polymer nano- and microparticle structures, Chemical Society Reviews 2009, 38, 1218-
1227.
[65] Zhang, X.; Chen, Z.-K.; Loh, K. P. Coordination-Assisted Assembly of 1-D
Nanostructured Light-Harvesting Antenna, Journal of the American Chemical Society
2009, 131, 7210-7211.
[66] McDowell, J. J.; Zacharia, N. S.; Puzzo, D.; Manners, I.; Ozin, G. A.
Electroactuation of Alkoxysilane-Functionalized Polyferrocenylsilane Microfibers,
Journal of the American Chemical Society 2010, 132, 3236-3237.
[67] Wang, J.-J.; Dai, L.-X.; Gao, Q.; Wu, P.-F.; Wang, X.-B. Electrospun
nanofibers of polyferrocenylsilanes with different substituents at silicon, European
Polymer Journal 2008, 44, 602-607.
[68] Zhang, S.; Yang, S.; Lan, J.; Tang, Y.; Xue, Y.; You, J. Ultrasound-Induced
Switching of Sheetlike Coordination Polymer Microparticles to Nanofibers Capable of
Gelating Solvents, Journal of the American Chemical Society 2009, 131, 1689-1691.
[69] Hui, J. K. H.; MacLachlan, M. J. Metal-containing nanofibers via coordination
chemistry, Coordination Chemistry Reviews 2010, 254, 2363-2390.
[70] Carne, A.; Carbonell, C.; Imaz, I.; Maspoch, D. Nanoscale metal-organic
materials, Chemical Society Reviews 2011, 40, 291-305.
[71] Terry, S. C.; Jerman, J. H.; Angell, J. B. A gas chromatographic air analyzer
fabricated on a silicon wafer, IEEE Transactions on Electron Devices 1979, 26, 1880-
1886.
Bibliography
149
[72] Manz, A.; Graber, N.; Widmer, H. M. Miniaturized total chemical analysis
systems: A novel concept for chemical sensing, Sensors and Actuators B: Chemical 1990,
1, 244-248.
[73] Whitesides, G. M. The origins and the future of microfluidics, Nature 2006,
442, 368-373.
[74] Avila, K.; Moxey, D.; de Lozar, A.; Avila, M.; Barkley, D.; Hof, B. The Onset
of Turbulence in Pipe Flow, Science 2011, 333, 192-196.
[75] Duffy, D. C.; McDonald, J. C.; Schueller, O. J. A.; Whitesides, G. M. Rapid
Prototyping of Microfluidic Systems in Poly(dimethylsiloxane), Analytical Chemistry
1998, 70, 4974-4984.
[76] Unger, M. A.; Chou, H.-P.; Thorsen, T.; Scherer, A.; Quake, S. R. Monolithic
Microfabricated Valves and Pumps by Multilayer Soft Lithography, Science 2000, 288,
113-116.
[77] Schaffhauser, D. F.; Andrini, O.; Ghezzi, C.; Forster, I. C.; Franco-Obregon,
A.; Egli, M.; Dittrich, P. S. Microfluidic platform for electrophysiological studies on
Xenopus laevisoocytes under varying gravity levels, Lab on a Chip 2011, 11, 3471-3478.
[78] Küster, S. K.; Fagerer, S. R.; Verboket, P. E.; Eyer, K.; Jefimovs, K.; Zenobi,
R.; Dittrich, P. S. Interfacing Droplet Microfluidics with Matrix-Assisted Laser
Desorption/Ionization Mass Spectrometry: Label-Free Content Analysis of Single
Droplets, Analytical Chemistry 2013, 85, 1285-1289.
[79] Stamou, D.; Duschl, C.; Delamarche, E.; Vogel, H. Self-Assembled
Microarrays of Attoliter Molecular Vessels, Angewandte Chemie 2003, 115, 5738-5741.
[80] Cooksey, G. A.; Elliott, J. T.; Plant, A. L. Reproducibility and Robustness of a
Real-Time Microfluidic Cell Toxicity Assay, Analytical Chemistry 2011, 83, 3890-3896.
[81] Park, J. I.; Saffari, A.; Kumar, S.; Günther, A.; Kumacheva, E. Microfluidic
Synthesis of Polymer and Inorganic Particulate Materials, Annual Review of Materials
Research 2010, 40, 415-443.
[82] Song, Y.; Hormes, J.; Kumar, C. S. S. R. Microfluidic Synthesis of
Nanomaterials, Small 2008, 4, 698-711.
[83] Dittrich, P. S.; Manz, A. Lab-on-a-chip: microfluidics in drug discovery,
Nature Reviews Drug Discovery 2006, 5, 210-218.
[84] El-Ali, J.; Sorger, P. K.; Jensen, K. F. Cells on chips, Nature 2006, 442, 403-
411.
[85] Gerber, D.; Maerkl, S. J.; Quake, S. R. An in vitro microfluidic approach to
generating protein-interaction networks, Nature Methods 2009, 6, 71-74.
Bibliography
150
[86] Hong, B. H.; Bae, S. C.; Lee, C.-W.; Jeong, S.; Kim, K. S. Ultrathin Single-
Crystalline Silver Nanowire Arrays Formed in an Ambient Solution Phase, Science 2001,
294, 348-351.
[87] Lee, J.-K.; Koh, W.-K.; Chae, W.-S.; Kim, Y.-R. Novel synthesis of organic
nanowires and their optical properties, Chemical Communications 2002, 138-139.
[88] Tian, B.; Zheng, X.; Kempa, T. J.; Fang, Y.; Yu, N.; Yu, G.; Huang, J.; Lieber,
C. M. Coaxial silicon nanowires as solar cells and nanoelectronic power sources, Nature
2007, 449, 885-889.
[89] Gomar-Nadal, E.; Puigmarti-Luis, J.; Amabilino, D. B. Assembly of functional
molecular nanostructures on surfaces, Chemical Society Reviews 2008, 37, 490-504.
[90] García, M.; Batalla, P.; Escarpa, A. Metallic and polymeric nanowires for
electrochemical sensing and biosensing, TrAC Trends in Analytical Chemistry 2014, 57,
6-22.
[91] Kenis, P. J. A.; Ismagilov, R. F.; Whitesides, G. M. Microfabrication Inside
Capillaries Using Multiphase Laminar Flow Patterning, Science 1999, 285, 83-85.
[92] Pumera, M. Nanomaterials meet microfluidics, Chemical Communications
2011, 47, 5671-5680.
[93] Dendukuri, D.; Pregibon, D. C.; Collins, J.; Hatton, T. A.; Doyle, P. S.
Continuous-flow lithography for high-throughput microparticle synthesis, Nature
Materials 2006, 5, 365-369.
[94] Edel, J. B.; Fortt, R.; deMello, J. C.; deMello, A. J. Microfluidic routes to the
controlled production of nanoparticles, Chemical Communications 2002, 1136-1137.
[95] Wang, Z.; Bao, R.; Zhang, X.; Ou, X.; Lee, C.-S.; Chang, J. C.; Zhang, X. One-
Step Self-Assembly, Alignment, and Patterning of Organic Semiconductor Nanowires
by Controlled Evaporation of Confined Microfluids, Angewandte Chemie International
Edition 2011, 50, 2811-2815.
[96] Jahn, A.; Vreeland, W. N.; Gaitan, M.; Locascio, L. E. Controlled Vesicle Self-
Assembly in Microfluidic Channels with Hydrodynamic Focusing, Journal of the
American Chemical Society 2004, 126, 2674-2675.
[97] Eyer, K.; Stratz, S.; Kuhn, P.; Küster, S. K.; Dittrich, P. S. Implementing
Enzyme-Linked Immunosorbent Assays on a Microfluidic Chip To Quantify
Intracellular Molecules in Single Cells, Analytical Chemistry 2013, 85, 3280-3287.
[98] Wang, J.; Bunimovich, Y. L.; Sui, G.; Savvas, S.; Wang, J.; Guo, Y.; Heath, J.
R.; Tseng, H.-R. Electrochemical fabrication of conducting polymer nanowires in an
integrated microfluidic system, Chemical Communications 2006, 3075-3077.
[99] Puigmartí-Luis, J.; Rubio-Martínez, M.; Imaz, I.; Cvetković, B. Z.; Abad, L.;
Pérez del Pino, A.; Maspoch, D.; Amabilino, D. B. Localized, Stepwise Template
Bibliography
151
Growth of Functional Nanowires from an Amino Acid-Supported Framework in a
Microfluidic Chip, ACS Nano 2014, 8, 818-826.
[100] Xing, Y.; Wyss, A.; Esser, N.; Dittrich, P. S. Label-free biosensors based on in
situ formed and functionalized microwires in microfluidic devices, Analyst 2015, 7896-
7901.
[101] Eyer, K.; Kuhn, P.; Hanke, C.; Dittrich, P. S. A microchamber array for single
cell isolation and analysis of intracellular biomolecules, Lab on a Chip 2012, 12, 765-
772.
[102] Bourlon, B.; Wong, J.; Miko, C.; Forro, L.; Bockrath, M. A nanoscale probe
for fluidic and ionic transport, Nature Nanotechnology 2007, 2, 104-107.
[103] Cvetković, B. Z. PhD thesis, ETH ZURICH, 2012.
[104] Xia, Y.; Whitesides, G. M. Soft Lithography, Angewandte Chemie
International Edition 1998, 37, 550-575.
[105] Lisensky, G. C.; Campbell, D. J.; Beckman, K. J.; Calderon, C. E.; Doolan, P.
W.; Rebecca, M. O.; Ellis, A. B. Replication and Compression of Surface Structures with
Polydimethylsiloxane Elastomer, Journal of Chemical Education 1999, 76, 537.
[106] Etchegoin, E. C. L. R. a. P. G. In Principles of Surface-Enhanced Raman
Spectroscopy; Elsevier: Oxford UK, 2009.
[107] Kneipp, J.; Schut, T. B.; Kliffen, M.; Menke-Pluijmers, M.; Puppels, G.
Characterization of breast duct epithelia: a Raman spectroscopic study, Vibrational
Spectroscopy 2003, 32, 67-74.
[108] Speiser, E.; Hinrichs, K.; Prete, P.; Lovergine, N.; Esser, N. Vibrational Raman
scattering from surfaces of III–V semiconductors: Microscopic and macroscopic surface
modes, physica status solidi (b) 2015, 252, 11-18.
[109] Friedrich Siebert, P. H. In Vibrational Spectroscopy in Life Science; WILEY-
VCH: Weinheim, 2008; Vol. 2.
[110] Long, D. A. The Raman Effect: A Unified Treatment of the Theory of Raman
Scattering by Molecules; John Wiley & Sons Ltd: Chichester, 2002.
[111] Ewen Smith, G. D. Modern Raman Spectroscopy: A Practical Approach; John
Wiley & Sons, Ltd., 2005.
[112] Zahn, D. R. T.; Gavrila, G. N.; Salvan, G. Electronic and Vibrational
Spectroscopies Applied to Organic/Inorganic Interfaces, Chemical Reviews 2007, 107,
1161-1232.
[113] Griffiths, D. J. Introduction to Electrodynamics; 3rd ed.; Pearson Education,
Dorling Kindersley, 2007.
Bibliography
152
[114] Kasap, S. O. In Principles of Electronic Materials and Devices; 3rd ed.; S., M.,
Ed.; McGraw-Hill: NewYork, 2006.
[115] Opilik, L.; Schmid, T.; Zenobi, R. Modern Raman Imaging: Vibrational
Spectroscopy on the Micrometer and Nanometer Scales, Annual Review of Analytical
Chemistry 2013, 6, 379-398.
[116] Pézolet, M.; Nafie, L. A.; Peticolas, W. L. Complete polarization
measurements for non-symmetric Raman tensors: Symmetry assignments of
ferrocytochrome C vibrations, Journal of Raman Spectroscopy 1973, 1, 455-464.
[117] Cardona, M. In Light Scattering in Solids II; Güntherodt, M. C. a. G., Ed.;
Springer: Berlin, Germany, 1982.
[118] Brosseau, C. Fundamentals of polarized light: a statistical optics approach;
Wiley: Germany, 1998.
[119] John R. Ferraro, K. N. a. C. W. B. Introductory Raman Spectroscopy;
Academic Press: San Diego: San Diego, CA 1994.
[120] Allemand, C. D. Depolarization Ratio Measurements in Raman Spectrometry,
Applied Spectroscopy 1970, 24, 348-353.
[121] Ko, A. C. T.; Choo-Smith, L.-P. i.; Hewko, M.; Sowa, M. G.; Dong, C. C. S.;
Cleghorn, B. Detection of early dental caries using polarized Raman spectroscopy,
Optics Express 2006, 14, 203-215.
[122] Wang, Y.; Yan, B.; Chen, L. SERS Tags: Novel Optical Nanoprobes for
Bioanalysis, Chemical Reviews 2013, 113, 1391-1428.
[123] Fabris, L.; Dante, M.; Nguyen, T.-Q.; Tok, J. B. H.; Bazan, G. C. SERS
Aptatags: New Responsive Metallic Nanostructures for Heterogeneous Protein Detection
by Surface Enhanced Raman Spectroscopy, Advanced Functional Materials 2008, 18,
2518-2525.
[124] Pauzauskie, P. J.; Talaga, D.; Seo, K.; Yang, P.; Lagugné-Labarthet, F.
Polarized Raman Confocal Microscopy of Single Gallium Nitride Nanowires, Journal of
the American Chemical Society 2005, 127, 17146-17147.
[125] Fréchette, J.; Carraro, C. Resolving Radial Composition Gradients in Polarized
Confocal Raman Spectra of Individual 3C-SiC Nanowires, Journal of the American
Chemical Society 2006, 128, 14774-14775.
[126] Yu, T.; Zhao, X.; Shen, Z. X.; Wu, Y. H.; Su, W. H. Investigation of individual
CuO nanorods by polarized micro-Raman scattering, Journal of Crystal Growth 2004,
268, 590-595.
[127] Johnsen, S. The Optics of Life: A Biologist's Guide to Light in Nature;
Princeton Univ. Press. , 2012.
[128] In Porto's notation; Bilbao Crystallographic Server.
Bibliography
153
[129] Lim, C.; Hong, J.; Chung, B. G.; deMello, A. J.; Choo, J. Optofluidic platforms
based on surface-enhanced Raman scattering, Analyst 2010, 135, 837-844.
[130] Batsanov, A. S. Tetrathiafulvalene revisited, Acta Crystallographica Section
C-Crystal Structure Communications 2006, 62, O501-O504.
[131] Wudl, F.; Kaplan, M. L.; Hufnagel, E. J.; Southwick, E. W. Convenient
synthesis of 1,4,5,8-tetrahydro-1,4,5,8-tetrathiafulvalene, The Journal of Organic
Chemistry 1974, 39, 3608-3609.
[132] Frisch, M. J. T., G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.;
Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A.; Nakatsuji,
H.; Caricato, M.; Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.;
Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida,
M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, J. A.; Jr.;
Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.;
Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant,
J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J.
E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.;
Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.;
Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.;
Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.;
Fox, D. J. ; Gaussian, I. W. C., Ed. 2009.
[133] Singh, D.; Vikram, K.; Singh, D. K.; Kiefer, W.; Singh, R. K. Raman and DFT
study of hydrogen-bonded 2- and 3-chloropyridine with methanol, Journal of Raman
Spectroscopy 2008, 39, 1423-1432.
[134] Singh, D. K.; Srivastava, S. K.; Schlücker, S.; Singh, R. K.; Asthana, B. P.
Self-association and hydrogen bonding of propionaldehyde in binary mixtures with water
and methanol investigated by concentration-dependent polarized Raman study and DFT
calculations, Journal of Raman Spectroscopy 2011, 42, 851-858.
[135] Singh, D. K. A., B. P.; Srivastava, S. K. Modeling the weak hydrogen bonding
of pyrrole and dichloromethane through Raman and DFT study, Journal of Molecular
Modeling 2012, 18 (8), 3541-3552.
[136] Lam, A. T. N.; Yoon, J.; Ganbold, E.-O.; Singh, D. K.; Kim, D.; Cho, K.-H.;
Son, S. J.; Choo, J.; Lee, S. Y.; Kim, S.; Joo, S.-W. Adsorption and desorption of tyrosine
kinase inhibitor erlotinib on gold nanoparticles, Journal of Colloid and Interface Science
2014, 425, 96-101.
[137] Lam, A. T. N.; Yoon, J.; Ganbold, E.-O.; Singh, D. K.; Kim, D.; Cho, K.-H.;
Lee, S. Y.; Choo, J.; Lee, K.; Joo, S.-W. Colloidal gold nanoparticle conjugates of
gefitinib, Colloids and Surfaces B: Biointerfaces 2014, 123, 61-67.
Bibliography
154
[138] Lee, M.; Baik, K. Y.; Noah, M.; Kwon, Y.-K.; Lee, J.-O.; Hong, S. Nanowire
and nanotube transistors for lab-on-a-chip applications, Lab on a Chip 2009, 9, 2267-
2280.
[139] Hou, S.; Wang, S.; Yu, Z. T. F.; Zhu, N. Q. M.; Liu, K.; Sun, J.; Lin, W.-Y.;
Shen, C. K. F.; Fang, X.; Tseng, H.-R. A Hydrodynamically Focused Stream as a
Dynamic Template for Site-Specific Electrochemical Micropatterning of Conducting
Polymers, Angewandte Chemie International Edition 2008, 47, 1072-1075.
[140] Kim, J.; Li, Z.; Park, I. Direct synthesis and integration of functional
nanostructures in microfluidic devices, Lab on a Chip 2011, 11, 1946-1951.
[141] Comini, E.; Baratto, C.; Faglia, G.; Ferroni, M.; Vomiero, A.; Sberveglieri, G.
Quasi-one dimensional metal oxide semiconductors: Preparation, characterization and
application as chemical sensors, Progress in Materials Science 2009, 54, 1-67.
[142] Patolsky, F.; Zheng, G.; Lieber, C. M. Nanowire-Based Biosensors, Analitical
Chemistry 2006, 78, 4260-4269.
[143] Chen, R. J.; Bangsaruntip, S.; Drouvalakis, K. A.; Wong Shi Kam, N.; Shim,
M.; Li, Y.; Kim, W.; Utz, P. J.; Dai, H. Noncovalent functionalization of carbon
nanotubes for highly specific electronic biosensors, Proceedings of the National
Academy of Sciences 2003, 100, 4984-4989.
[144] Besteman, K.; Lee, J.-O.; Wiertz, F. G. M.; Heering, H. A.; Dekker, C.
Enzyme-Coated Carbon Nanotubes as Single-Molecule Biosensors, Nano Letters 2003,
3, 727-730.
[145] Huynh, W. U.; Dittmer, J. J.; Alivisatos, A. P. Hybrid Nanorod-Polymer Solar
Cells, Science 2002, 295, 2425-2427.
[146] Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Self-
Assembled Monolayers of Thiolates on Metals as a Form of Nanotechnology, Chemical
Reviews 2005, 105, 1103-1170.
[147] Han, X. X.; Chen, L.; Ji, W.; Xie, Y.; Zhao, B.; Ozaki, Y. Label-Free Indirect
Immunoassay Using an Avidin-Induced Surface-Enhanced Raman Scattering Substrate,
Small 2011, 7, 316-320.
[148] Kho, K. W.; Dinish, U. S.; Kumar, A.; Olivo, M. Frequency Shifts in SERS
for Biosensing, ACS Nano 2012, 6, 4892-4902.
[149] Volkan, M.; Stokes, D. L.; Vo-Dinh, T. Surface-Enhanced Raman of
Dopamine and Neurotransmitters Using Sol-Gel Substrates and Polymer-Coated Fiber-
Optic Probes, Applied Spectroscopy 2000, 54, 1842-1848.
[150] Wang, Y.; Lee, K.; Irudayaraj, J. SERS aptasensor from nanorod-nanoparticle
junction for protein detection, Chemical Communications 2010, 46, 613-615.
Bibliography
155
[151] Miller, A. W.; Robyt, J. F. Sodium cyanoborohydride in the immobilization of
proteins to glutaraldehyde-activated aminoalkyl silica, Biotechnology and Bioengeering
1983, 25, 2795-2800.
[152] Sasso, L.; Heiskanen, A.; Diazzi, F.; Dimaki, M.; Castillo-Leon, J.; Vergani,
M.; Landini, E.; Raiteri, R.; Ferrari, G.; Carminati, M.; Sampietro, M.; Svendsen, W. E.;
Emneus, J. Doped overoxidized polypyrrole microelectrodes as sensors for the detection
of dopamine released from cell populations, Analyst 2013, 138, 3651-3659.
[153] Mcglashen, M. L.; Davis, K. L.; Morris, M. D. Surface-Enhanced Raman-
Scattering of Dopamine at Polymer-Coated Silver Electrodes, Analitical Chemistry 1990,
62, 846-849.
[154] Lee, N. S.; Hsieh, Y. Z.; Paisley, R. F.; Morris, M. D. Surface-enhanced Raman
spectroscopy of the catecholamine neurotransmitters and related compounds, Analytical
Chemistry 1988, 60, 442-446.
[155] Oh, Y. J.; Jeong, K. H. Optofluidic SERS chip with plasmonic nanoprobes self-
aligned along microfluidic channels, Lab on a Chip 2014, 14, 865-868.
[156] Perumal, J.; Kong, K. V.; Dinish, U. S.; Bakker, R. M.; Olivo, M. Design and
fabrication of random silver films as substrate for SERS based nano-stress sensing of
proteins, RSC Advances 2014, 4, 12995-13000.
[157] Li, M.; Cushing, S. K.; Zhang, J.; Suri, S.; Evans, R.; Petros, W. P.; Gibson, L.
F.; Ma, D.; Liu, Y.; Wu, N. Three-Dimensional Hierarchical Plasmonic Nano-
Architecture Enhanced Surface-Enhanced Raman Scattering Immunosensor for Cancer
Biomarker Detection in Blood Plasma, ACS Nano 2013, 7, 4967-4976.
[158] Hu, X. G.; Wang, T.; Wang, L.; Dong, S. J. Surface-enhanced Raman
scattering of 4-aminothiophenol self-assembled monolayers in sandwich structure with
nanoparticle shape dependence: Off-surface plasmon resonance condition, Journal of
Physical Chemistry C 2007, 111, 6962-6969.
[159] Cao, L.; Diao, P.; Tong, L.; Zhu, T.; Liu, Z. Surface-Enhanced Raman
Scattering of p-Aminothiophenol on a Au(core)/Cu(shell) Nanoparticle Assembly,
ChemPhysChem 2005, 6, 913-918.
[160] Katan, C. First-Principles Study of the Structures and Vibrational Frequencies
for Tetrathiafulvalene TTF and TTF-d4 in Different Oxidation States, The Journal of
Physical Chemistry A 1999, 103, 1407-1413.
[161] Ferraro, J. R. N., K. In Introductory Raman Spectroscopy; Academic Press:
San Diego: CA, 1999.
[162] Cooper, W. F.; Kenny, N. C.; Edmonds, J. W.; Nagel, A.; Wudl, F.; Coppens,
P. Crystal and molecular structure of the aromatic sulphur compound 2,2[prime or
minute]-bi-1,3-dithiole. Evidence for d-orbital participation in bonding, Journal of the
Chemical Society D: Chemical Communications 1971, 889-890.
Bibliography
156
[163] Cooper, W. F. E., J. W.; Wudl, F.; Coppens, P. The 2-2'-bi-1,3-dithiole, Crystal
Structure Communications 1974, 3, 23-26.
[164] Sandman, D. J.; Epstein, A. J.; Chickos, J. S.; Ketchum, J.; Fu, J. S.; Scheraga,
H. A. Crystal lattice and polarization energy of tetrathiafulvalene, The Journal of
Chemical Physics 1979, 70, 305-313.
[165] Berlinsky, A. J.; Hoyano, Y.; Weiler, L. Raman spectra of tetrathiofulvalene
(TTF), Chemical Physics Letters 1977, 45, 419-421.
[166] Bangar, M. A.; Shirale, D. J.; Chen, W.; Myung, N. V.; Mulchandani, A. Single
Conducting Polymer Nanowire Chemiresistive Label-Free Immunosensor for Cancer
Biomarker, Analytical Chemistry 2009, 81, 2168-2175.
[167] Pescaglini, A.; Iacopino, D. Metal nanoparticle-semiconductor nanowire
hybrid nanostructures for plasmon-enhanced optoelectronics and sensing, Journal of
Materials Chemistry C 2015, 3, 11785-11800.
[168] Bryce, M. R. Tetrathiafulvalenes as π-Electron Donors for Intramolecular
Charge-Transfer Materials, Advanced Materials 1999, 11, 11-23.
[169] Suzuki, W.; Fujiwara, E.; Kobayashi, A.; Fujishiro, Y.; Nishibori, E.; Takata,
M.; Sakata, M.; Fujiwara, H.; Kobayashi, H. Highly Conducting Crystals Based on
Single-Component Gold Complexes with Extended-TTF Dithiolate Ligands, Journal of
the American Chemical Society 2003, 125, 1486-1487.
[170] Segura, J. L.; Martín, N. New Concepts in Tetrathiafulvalene Chemistry,
Angewandte Chemie International Edition 2001, 40, 1372-1409.
[171] Choi, S. N.; Jung, W. S.; Lee, M. J.; Lee, Y. M.; Kim, Y. I. Synthesis and
physicochemical properties of charge transfer compounds derived from the reaction of
tetrathiafulvalene with copper(II), Polyhedron 2004, 23, 2111-2115.
[172] Naka, K.; Ando, D.; Chujo, Y. Effect of substituent groups for formation of
organic-metal hybrid nanowires by charge-transfer of tetrathiafulvalene derivatives with
metal ion, Synthetic Metals 2009, 159, 931-934.
[173] Geng, Y.; Wang, X.-J.; Chen, B.; Xue, H.; Zhao, Y.-P.; Lee, S.; Tung, C.-H.;
Wu, L.-Z. Semiconducting Neutral Microstructures Fabricated by Coordinative Self-
Assembly of Intramolecular Charge-Transfer Tetrathiafulvalene Derivatives, Chemistry
– A European Journal 2009, 15, 5124-5129.
[174] Kanehama, R.; Umemiya, M.; Iwahori, F.; Miyasaka, H.; Sugiura, K.-i.;
Yamashita, M.; Yokochi, Y.; Ito, H.; Kuroda, S.-i.; Kishida, H.; Okamoto, H. Novel ET-
Coordinated Copper(I) Complexes:  Syntheses, Structures, and Physical Properties (ET
= BEDT-TTF = Bis(ethylenedithio)tetrathiafulvalene), Inorganic Chemistry 2003, 42,
7173-7181.
Bibliography
157
[175] Thangawng, A. L.; Howell Jr, P. B.; Richards, J. J.; Erickson, J. S.; Ligler, F.
S. A simple sheath-flow microfluidic device for micro/nanomanufacturing: fabrication
of hydrodynamically shaped polymer fibers, Lab on a Chip 2009, 9, 3126-3130.
[176] Mas-Torrent, M.; Hadley, P.; Bromley, S. T.; Ribas, X.; Tarres, J.; Mas, M.;
Molins, E.; Veciana, J.; Rovira, C. Correlation between crystal structure and mobility in
organic field-effect transistors based on single crystals of tetrathiafulvalene derivatives,
Journal of the American Chemical Society 2004, 126, 8546-8553.
[177] Geiser, U.; Wang, H. H.; Hammond, C. E.; Firestone, M. A.; Beno, M. A.;
Carlson, K. D.; Nunez, L.; Williams, J. M. Structure of di[3,4;3',4'-bis(ethylenedithio)-
2,2',5,5'-tetrathiafulvalenium] dichlorocuprate(I), (BEDT-TTF)2CuCl2, Acta
Crystallographica Section C 1987, 43, 656-659.
[178] Hathaway, B. J.; Holah, D. G.; Postlethwaite, J. D. 630. The preparation and
properties of some tetrakis(methylcyanide)copper(I) complexes, Journal of the Chemical
Society (Resumed) 1961, 3215-3218.
[179] Bozio, R.; Zanon, I.; Girlando, A.; Pecile, C. Vibrational spectroscopy of
molecular constituents of one‐dimensional organic conductors. Tetrathiofulvalene (TTF),
TTF+, and (TTF+)2 dimer, The Journal of Chemical Physics 1979, 71, 2282-2293.
[180] Curtis, N. F.; Curtis, Y. M. Some Nitrato-Amine Nickel(II) Compounds with
Monodentate and Bidentate Nitrate Ions, Inorganic Chemistry 1965, 4, 804-809.
[181] Jorgensen, T.; Hansen, T. K.; Becher, J. Tetrathiafulvalenes as building-blocks
in supramolecular chemistry, Chemical Society Reviews 1994, 23, 41-51.
[182] Novoa, J. J.; Rovira, M. C.; Rovira, C.; Veciana, J.; Tarrés, J. C.H…S and
S…S: Two major forces in organic conductors, Advanced Materials 1995, 7, 233-237.
[183] J. M. Williams, J. R. F., R. J. Thorn, K. D. Carlson, U. Geiser, H. H. Wang, A.
M. Kini, and M.-H. Whangbo Organic superconductors (including fullerenes): Synthesis,
structure, properties, and theory. ; Prentice Hall, Englewood Cliffs, NJ, 1992.
[184] Inoue, M. B.; Inoue, M.; Fernando, Q.; Nebesny, K. W.
Tetracyanoquinodimethan salts of copper chelate with
tetrabenzo[b,f,j,n][1,5,9,13]tetraazacyclohexadecine: electrical properties and mixed-
valence states, The Journal of Physical Chemistry 1987, 91, 527-530.
[185] Kaim, W.; Moscherosch, M. The coordination chemistry of TCNE, TCNQ and
related polynitrileπ acceptors, Coordination Chemistry Reviews 1994, 129, 157-193.
[186] Miller, J. S.; Calabrese, J. C.; Harlow, R. L.; Dixon, D. A.; Zhang, J. H.; Reiff,
W. M.; Chittipeddi, S.; Selover, M. A.; Epstein, A. J. Ferromagnetic behavior of linear
chain electron-transfer complexes. Decamethylferrocene electron-transfer salts of 2,5-
disubstituted-7,7,8,8-tetracyano-p-quinodimethanes. Magnetic characterization of
[Fe(C5Me5)2].+[TCNQI2].- and structures of [TCNQI2]n- (n = 0, 1, 2), Journal of the
American Chemical Society 1990, 112, 5496-5506.
Bibliography
158
[187] Potember, R. S.; Hoffman, R. C.; Hu, H. S.; Cocchiaro, J. E.; Viands, C. A.;
Poehler, T. O. Electronic Devices from Conducting Organics and Polymers, Polymer
Journal 1987, 19, 147-156.
[188] Zhao, H.; Heintz, R. A.; Ouyang, X.; Dunbar, K. R.; Campana, C. F.; Rogers,
R. D. Spectroscopic, Thermal, and Magnetic Properties of Metal/TCNQ Network
Polymers with Extensive Supramolecular Interactions between Layers, Chemistry of
Materials 1999, 11, 736-746.
[189] Liu, S.-G.; Liu, Y.-Q.; Wu, P.-J.; Zhu, D.-B. Multifaceted Study of CuTCNQ
Thin-Film Materials. Fabrication, Morphology, and Spectral and Electrical Switching
Properties, Chemistry of Materials 1996, 8, 2779-2787.
[190] Liu, Y.; Ji, Z.; Tang, Q.; Jiang, L.; Li, H.; He, M.; Hu, W.; Zhang, D.; Jiang,
L.; Wang, X.; Wang, C.; Liu, Y.; Zhu, D. Particle-Size Control and Patterning of a
Charge-Transfer Complex for Nanoelectronics, Advanced Materials 2005, 17, 2953-
2957.
[191] O'Mullane, A. P.; Fay, N.; Nafady, A.; Bond, A. M. Preparation of
Metal−TCNQ Charge-Transfer Complexes on Conducting and Insulating Surfaces by
Photocrystallization, Journal of the American Chemical Society 2007, 129, 2066-2073.
[192] Li, D.; Sutton, D.; Burgess, A.; Graham, D.; Calvert, P. D. Conductive copper
and nickel lines via reactive inkjet printing, Journal of Materials Chemistry 2009, 19,
3719-3724.
[193] Hansen, C. L.; Skordalakes, E.; Berger, J. M.; Quake, S. R. A robust and
scalable microfluidic metering method that allows protein crystal growth by free
interface diffusion, Proceedings of the National Academy of Sciences 2002, 99, 16531-
16536.
[194] Gucciardi, P. G.; Trusso, S.; Vasi, C.; Patane, S.; Allegrini, M. Nano-Raman
imaging of Cu-TCNQ clusters in TCNQ thin films by scanning near-field optical
microscopy, Physical Chemistry Chemical Physics 2002, 4, 2747-2753.
[195] Xiao, K.; Ivanov, I. N.; Puretzky, A. A.; Liu, Z. Q.; Geohegan, D. B. Directed
integration of tetracyanoquinodimethane-Cu organic nanowires into prefabricated device
architectures, Advanced Materials 2006, 18, 2184-2188.
Appendix
159
Appendix
A) Materials
SU-8 and developers for the resists were purchased from Microchem (Newton,
MA, USA). 1H,1H,2H,2H-perfluorodecyl-dimethylchlorosilane was purchased
from ABCR (Karlsruhe, Germany), and poly(dimethylsiloxane) (PDMS)
(Sylgard 184) was obtained from Dow Corning (Midland, MI, USA).
Acetonitrile (CH3CN, 99.8+%), 4-Aminothiophenol (4-ATPh, 97%), 6-
Aminofluorescein (95%), Cobalt(II) chloride hexahydrate (CoCl2·6H2O, 98%),
Copper(I) chloride powder (CuCl, 97%), Copper(II) chloride powder (CuCl2,
99%), Copper(II) sulfate (CuSO4, anhydrous, powder, ≥99.99% trace metals
basis), Dimethylsulfoxid (DMSO, 99.8+%), DL-Norepinephrine hydrochloride,
(>=97% TLC), Dopamine hydrochloride, Ethanol (99.8+%), Ethanol (99.8+%),
Gold(III) chloride trihydrate (HAuCl4, 99+% trace metal basis), Iron(II)
chloride tetrahydrate (FeCl2·4H2O, >=99%), Iron(III) nitrate tetrahydrate
(Fe(NO3)3·4H2O, >=98%), Isoprenaline hydrochloride, L-Epinephrine, L-
Phenylalanine (>=99%), L-Tyrosine (>99%), Methanol (CH3OH, 99.8+%), N-
Hydroxysuccinimide (NHS, 98%), Nickel(II) chloride hexahydrate
(NiCl2·6H2O, 99.9%), Phosphate buffered saline (PBS) tablet, Sodium
borohydride (NaBH4, powder, ≥98.0%), Sodium citrate dehydrate
(C6H5O7Na3·2H2O), Sodium hydroxide (NaOH, reagent grade, ≥98%, pellets),
and tetrahydrofuran (THF, anhydrous, ≥99.9%) were all purchased from Sigma-
Aldrich (Buchs, Switzerland). 2-Aminoethanethiol hydrochloride (Cysteamine
hydrochloride, 98%), Iron(III) chloride anhydrous (FeCl3, 98%), Manganese(II)
chloride tetrahydrate (MnCl2·4H2O, 99+%) were obtained from Acros Organics
(Basel, Switzerland). Copper(II) nitrate trihydrate (Cu(NO3)2·3H2O, 98+%),
Glutaraldehyde (GA, 50% in water solution) and Zinc chloride (ZnCl2, >98%)
were purchased from Fluka (Buchs, Switzerland). Bis(ethylenedithio)-
tetrathiafulvalene (BEDT-TTF), 1-Ethyl-3-(3-dimethylaminopropyl)carbodii-
mide hydrochloride (EDC, 98%), 2-Formyltetrathiafulvalene (FTTF, >98%),
7,7,8,8-Tetracyanoquinodimethane (TCNQ, >98.0%, HPLC), Tetrakis-
(ethylthio)tetrathiafulvalene (TET-TTF), 2,3,6,7-Tetrakis(2-cyanoethylthio)-
tetrathiafulvalene (TCE-TTF) and Tetrathiaful--valene (TTF, 99+%) were
obtained from TCI (Eschborn, Germany). Lissamine rhodamine B sulfonyl
chloride (mixed isomers) and PBS solution (pH 7.40, 10 mM) were obtained
Appendix
160
from Invitrogen (Lucerne, Switerland). FITC-labeled human IgG antibody (0.5
mg/mL), Human IgG antibody (mAb mouse, 0.5 mg/mL) and Human IgG
protein (4 mg/mL) were purchased from GenScript (Piscataway, NJ, USA).
Cobalt(II) nitrate hexahydrate (Co(NO3)2·6H2O, >= 99.0%), Zinc nitrate
hexahydrate (Zn(NO3)2·6H2O, extra pure) was purchased from MERCK (Merck
Schuchardt OHG, Hohenbrunn, Germany). Nickel (II) nitrate hexahydrate
(Ni(NO3)2·6H2O, GR for analysis) was purchased from VWR International AG
(Dietikon, Switzerland). Ultrapure deionized (DI) water (Mill-Q purifiers, 18.2
MΩ·cm at 25 °C) was used throughout the experiment.
B) Structures of important chemicals used Chapter 5.1
Figure 7.1. Structures of catecholamines and aromatic amino acids.
Catecholamines including Dopamine (DA), Norepinephrine (NE), L-Epinephrine
(EPI) and Isoprenaline (ISP). Aromatic amino acids including L-Phenylalanine
(Phe) and L-Tyrosine (Tyr).
Dopamine Norepinephrine
Isoprenaline L-Phenylalanine L-Tyrosine
L-Epinephrine
Acknowledgement
161
Acknowledgement
First of all, I am very grateful to my supervisors, Prof. Dr. Norbert Esser and
Prof. Dr. Petra S. Dittrich. It is because of their supervision and great support
during the past three and a half years that I can finally manage to finish my
doctoral research and this thesis. Great thanks to Prof. Esser for offering me the
chance to work on this interesting and challenging project. Thanks for his
constant encouragement, positive attitude and tireless effort in discussion on the
project throughout the whole period. Thanks him for sharing inspiring and
motivating insights in the interdisciplinary project.
Great thanks go to my second supervisor Prof. Petra S. Dittrich for giving me
the opportunity to do research in her group in ETH Zürich. I am deeply
impressed by her rich knowledge in microfluidics and bioanalytics, and also
highly encouraged by her enthusiasm in doing research! Thanks for allowing
me to follow my interests and small projects. This helped me not only to gain a
broad understanding of my research field, but also to know how to refine
research projects in the future.
Furthermore, I would like to thank all the examiners. Many thanks to Prof.
Hildebrandt for agreeing to be the examiner with no hesitate. Thank all of them
for their careful reading and revisions on this thesis. I appreciate all the
suggestions, corrections and the time they invested. Great thanks goes to Prof.
Birgit Kanngießer who agreed to be the chairman of my oral examination.
Following, I want to express my gratitude to my colleagues in Bioanalytics
group of ETH Zürich: Dr. Eva Bönzli and Simone Stratz for their help in
microfluidics experiments. My officemates in HCI F301, Dr. Tom Robinson for
the valuable discussion and suggestions; Klaus Eyer for sharing biological
samples and great help in sample preparation; Simon Küster for always keeping
the nice working environment in the office. Also, my thanks goes to Dr. Daniel
Schaffhauser for his great help in electrical measurements and valuable
discussion on the results. Many thanks also for Pascal Verboket for chip design
and microfabrication. Thank the whole group for the free time we spent together,
badminton, Friday Bistro time, etc. These were so important to keep a free mind
for better work. In addition, I want to address my gratitude to Christoph Bärtschi
from the workshop, Dr. Karsten Kunze from the ScopeM Center for their help.
Great thanks to my colleagues in ISAS Berlin: Dr. Eugen Speiser, for his
great help in my project, the valuable suggestions in experiment and useful
Acknowledgement
162
discussion in the results! Many thanks to Dr. Princia Salvatore for the very
valuable discussion, useful suggestions and detailed corrections on this
thesis. I want to thank Maximilian Ries for helping in translating the
German abstract. Thank Dr. Maciej Neumann for his help in explaining
the promotion system. Many thanks go to my officemates, Johannes
Falkenburg (for the nice German-Chinese learning time and good
suggestions on my presentation), Julian Plaickner, Timo Seemke for the
good atmosphere in the office. It is very nice to share the office with them.
I would like to thank the technician, Mr. Karsten Roland, also the other
group members for their help.
Thank my collaborators, Andreas Wyss (D-MATL, ETH Zürich) for the
Raman measurement, Dr. Guoguang Sun (ISAS Berlin) for the IR
measurement and proof reading, Dr. Dheeraj Kumar Singh (Jacobs
University, Bremen) for the DFT calculation.
Also, I would like to acknowledge the funding from DFG, with which I could
do my research in Berlin and Zürich, also joining conferences to share and learn
ideas. I enjoyed the period being a fellow of SALSA (School of Analytical
Sciences Adlershof) which supports women in science during the past over three
years. I want to thank the management director, Ms. Katharina Schultens, for
her great support in feedback, discussion and suggestions. Also, deeply gratitude
to Ms. Esther Santel for her help with everything. I will never forget the moment
when I received her phone call in China in 2013. Thank other SALSA staff for
their help and support. Thanks to my friends from SALSA school, Benita
Schmidt and Natalia Ogorodnikova for their support and the friendship.
With the next few lines I want to direct my gratitude to my friends in Zürich and
Berlin, Chengjun, Xiuxiu, Wenjie, Yuanyuan, Xiaojuan, Zhiyang and Jing, who
always standing at my side. I never felt lonely when I was away from my family.
Thanks for helping me settling down in the two cities and sharing the friendship,
also the nice chats after work, about life, listening and advice.
Finally, I want to give my special and deep thanks to my family. Thanks for the
support and love from my husband and my son during the separation period.
Thanks to my husband Ming for sharing my depression when I encountered
problems, and the joy when I made progress in work. Also thank my lovely son
Zhitao who gives me a happy smile and a big hug every day. I could not manage
to finish without their support. Their love is the power always encouraging me
to pursue my life in scientific research, in the past and in the future!
Statement of authorship
163
Statement of authorship
This doctoral thesis will be submitted for the degree of Doctor rerum naturalium
(Dr. rer. nat.). I hereby certify that this thesis was composed by myself about
my work during the doctoral research period. All the collaborations and support
from agencies and people have been specifically acknowledged. All references
supporting this work have been quoted properly. I promise that this thesis does
not contain any work extracted from a thesis, or research paper previously
presented for another doctoral degree at this or other universities. I have
prepared this thesis specifically for the degree of Dr. rer. nat., under supervision
of Prof. Dr. Norbert Esser at the Leibniz-Institut für Analytische Wissenschaften
– ISAS – e. V. Berlin during the period of May 2013 to October 2016 (with part
of the cooperation work done in Eidgenössische Technische Hochschule Zürich).