Investigation of dimensional and structural properties of dye aggregates / Richárd Szopkó [original]

Investigation of dimensional and structural properties

of dye aggregates

Von der Fakultät für Naturwissenschaften

Department Chemie

der Universität Paderborn

zur Erlangung des Grades eines

Doktors der Naturwissenschaften

Dr. rer. nat.

genehmigte Dissertation

von

Richárd Péter Szopkó

aus Hajdúdorog (Ungarn)

Paderborn 2009

Die vorliegende Arbeit wurde in der Zeit von Februar 2005 bis April 2009 im

Fachgebiet Physikalische Chemie am Department Chemie der Fakultät für

Naturwissenschaften der Universität Paderborn unter Anleitung von Prof. Dr.

Claudia Schmidt angefertigt.

Referent: Prof. Dr. Claudia Schmidt

Korreferent: PD. Dr. Hans Egold

Eingereicht am: 23.04.2009

Mündliche Prüfung am: 29.05.2009

First of all, I would like to address my thanks to my supervisor Prof. Dr. Claudia

Schmidt, who gave me the possibility to work on this interesting topic. Thank you

for your support and patience in completing my thesis!

By the same token, I wish to express my thanks to Prof. Dr. Ferenc Joó, who

encouraged me to start my PhD work.

Further thanks to:

PD. Dr. H. Egold for being a valuable source of ideas and suggestions in the topic of

NMR spectroscopy and moreover for agreeing to write the Korreferat.

Prof. Dr. K. Huber for providing the dyes.

Prof. Dr. U. Olsson for the opportunity to work in the Department of Physical

Chemistry 1 at Lund University.

Prof. Dr. D. Topgaard, Dr. S. Lasic, Dr. M. Nilsson and I. Aslund for the technical

support of the diffusion NMR measurements at Lund University.

B. Medronho from Coimbra and Lund University for the effective collaboration.

Dr. F. Kleinschmidt for the diffusion NMR measurements in Freiburg.

Dr. J. Sehnert for help with the quantum chemical calculations in Bayreuth.

M. Perez for help with NMR sample preparation.

My colleagues, Dr. T. Witte, Dr. A. Hoischen, Dr. S. Benning, Dr. M. Lauhof, Dr. H.

Matthias, A. Lorenz, A. Redler, S. Lages, M. Bayer, Dr. S. Shafaei and G. Ar for a

positive working atmosphere.

G. Jünnemann and S. Keuker-Baumann for the technical support and the every-day

conversation in German.

I. Koralewicz for the fast and perfect all-around help.

Special thanks to Dr. L. Germán, O. Germán, Dr. B. Elsaesser, Dr. L. Majoros and B.

Wolf for their personal and scientific advice and help.

Finally, I extend my honest thanks to my family for all they have shared and their

continuous help!

Presentations and publications

R. Szopko, K. Huber, C. Schmidt: NMR investigation of the aggregation of dye molecules,

Poster, 72nd Annual Meeting of the German Physical Society, Berlin, Germany, February

2008

R. Szopko, C. Schmidt:

1

H chemical shift analysis and diffusion study of dye aggregates,

Poster, 7

th

Annual Surface and Colloid Symposium Nanostructures from self-assembly in

solution, Lund, Sweden, November 2007

R. Szopko, C. Schmidt: Diffusion NMR study of dye aggregates, Poster, 21

th

Conference of

the European Council of International Schools, Geneva, Switzerland, September 2007

R. Szopko, C. Schmidt: Aggregation of dye molecules: diffusion measured by pulsed field

gradient NMR method, Talk, NMR meeting, Berlin, Germany, March 2007

R. Szopko: Structure determination of dye aggregates, Poster, GDCh Magnetic Resonance

Division 28

th

Discussion Meeting, Poster, Tübingen, Germany, September 2006

R. Szopko: Structure determination of dyes, Talk, NMR meeting, Pottenstein, Germany,

March 2006

B. Medronho, S. Shafaei, R. Szopko, M. G. Miguel, U. Olsson, C. Schmidt: Shear-induced

transitions between a planar phase and multi-lamellar vesicles: Continuous versus

discontinuous transformation, Langmuir 2008, 24, 6480-6486

B. Medronho, S. Shafaei, R. Szopko, M. G. Miguel, U. Olsson, C. Schmidt: Shear-induced

structural transformations of the lyotropic lamellar phase: Continuous or discontinuous

transitions?, Poster, 21

th

Conference of the European Council of International Schools,

Geneva, Switzerland, September 2007

CHAPTER 1 INTRODUCTION 1

CHAPTER 2 THEORETICAL ASPECTS 4

2.1

O

PTICAL PROPERTIES OF

J-

AND

H-

AGGREGATES

4

2.2

NMR

SPECTROSCOPY

7

2.2.1

O

NE

-D

IMENSIONAL

NMR

TECHNIQUES

7

2.2.2

T

WO

-D

IMENSIONAL

NMR

TECHNIQUES

10

2.2.3

D

IFFUSION AND

R

ELAXATION

11

2.3

C

OMPUTER

S

IMULATION

20

CHAPTER 3 EXPERIMENTAL 22

3.1

M

ATERIALS

22

3.1.1

S

AMPLE PURIFICATION

23

3.1.2

S

AMPLE PREPARATION

24

3.2

T

ECHNIQUES

28

3.2.1

P

OLARIZING

M

ICROSCOPY

28

3.2.2

UV

SPECTROSCOPY

28

3.2.3

NMR

SPECTROSCOPY

28

3.2.4

C

OMPUTER SIMULATION

29

CHAPTER 4 RESULTS AND DISCUSSION 30

4.1

O

PTICAL PROPERTIES OF DYES

30

4.1.1

P

OLARIZING MICROSCOPY

30

4.1.2

UV/VIS

SPECTROSCOPY

32

4.2

N

UCLEAR

M

AGNETIC RESONANCE

36

4.2.1

P

EAK ASSIGNMENT

36

4.2.2

1

H

C

HEMICAL SHIFT

A

NALYSIS

50

4.2.3

S

IGNAL INTENSITIES

57

4.2.4

C

HEMICAL

E

XCHANGE AND

L

INE WIDTH

59

4.2.5

NMR

DIFFUSOMETRY

69

4.3

C

OMPUTER SIMULATION

88

4.3.1

DFT

STRUCTURE OPTIMIZATION OF THE MONOMER

88

4.3.2

DFT

SIMULATION OF

NMR

SPECTRA OF THE MONOMER

89

4.2.4

DFT

SIMULATION OF

UV

SPECTRA OF THE MONOMER

90

4.2.5

DFT

STRUCTURE OPTIMIZATION OF THE DIMER

91

4.2.5

DFT

NMR

SPECTRUM SIMULATION OF THE DIMER

93

CHAPTER 5 SUMMARY 96

APPENDIX 98

A1

D

IFFUSION OF THE MIXTURE

98

A2

D

IFFUSION DATA OF

YD 102

A3

P

ULSE PROGRAM FOR DIFFUSION EXPERIMENTS

108

LIST OF SYMBOLS 118

REFERENCES 120

Introduction

1

CHAPTER 1 INTRODUCTION

Many small molecules form aggregates in solution. This phenomenon of

aggregation plays an important role in technology and biology. The self assembly of

molecules or particles is often an unwanted effect, for example in the case of

dyestuffs

1

, where it leads to several problems during the dyeing process like

clogging of filters or uneven dye distribution on the fiber. Another example of

aggregation is found for small amyloid polypeptides.

2,3

Their aggregation leads to

the formation of the plaques that have been related to diseases. Therefore, a

control of aggregation is in demand and a better physicochemical understanding of

aggregation processes and of the properties of the aggregates is required.

In the formation of organic dye aggregates two different kinds of forces are

important: dispersion forces due to the interaction between the π system of the

molecules and forces resulting from the hydrophobic effect

4

, which becomes

relevant when hydrophobic substances are added into a hydrophilic medium, like

water. When molecules in aqueous solution form aggregates the single molecules

release their individual ordered solvation shells and the entropy of the system

increases. In this case the attraction between the dyestuff molecules is less relevant,

but the interaction of the water molecules which results in a highly organized

water structure and the elimination of the water-dyestuff contact through

aggregate building drives the aggregation. Between the surfaces of flat aromatic

rings van der Waals interaction and H-bonds can stabilize the aggregates

5

and the

molecules tend to arrange themselves like a stack of coins.

The aromatic interactions are ubiquitous in nature. They are believed to provide

stability to J- and H-aggregates or to the duplex DNA, they have been proposed to

contribute to the unique properties of proteins

6

, they may play a role in the

aggregation of amyloid in Alzheimer’s disease

7

and they are common motifs in

biomolecular recognition. Extensive work has focused on these interactions to

determine their importance

8-12

. π-π interactions are caused by intermolecular

overlapping of p-orbitals in π-conjugated systems, so they become stronger as the

Introduction

2

number of π-electrons increases. The change of the electron cloud around the

molecule, due to the stacking of the ring, affects the behavior of the system, thus

NMR chemical shift analysis and optical spectroscopy can provide information

about the aggregates.

The phenomenon of the aggregation process of dyestuff has been observed in 1936

by Scheibe

13

and Jelley

14

who investigated cyanine dyestuff by optical spectroscopy

and described the aggregation of the molecules. These aggregate structures are an

intermediate stage of the dyestuff from the dissociated stage to the crystals. The

aggregate formation can be followed by studying the absorption properties of

solutions with different concentrations in the UV/VIS range. During the

aggregation process the intensity of the main absorption band increases or

disappears while several new absorption bands appear in the spectra either at

lower or higher wavelengths and an intense change in the color of the solution can

be observed. These aggregates are usually referred to as J- or, occasionally, Scheibe-

aggregates. If the aggregate absorption band is blue shifted (towards higher

absorption energy) with respect to that of the monomer the aggregates are called

H-aggregates (hypsochromic shift). Many spectroscopic investigations on the dye

aggregates have been conducted since their first discovery

15-27

.

In several dye systems a concentration-dependent variation in the NMR chemical

shifts of aromatic protons

28-34

was observed as a result of the stacking of aromatic

rings. If there are molecules in the aggregates that contain protons held in position

directly over an aromatic ring a characteristic upfield chemical shift value is

observed by NMR spectroscopy.

The chemical shift of a proton can be calculated by quantum chemical methods

and together with the measured NMR spectra structural information can be

acquired.

35-39

Furthermore, the shape and the dimension of molecules and

aggregates are reflected in their diffusion properties. The fact that molecular self

diffusion can be measured by NMR methods was realized in the early days of NMR

spectroscopy.

40,41

The most practical pulse sequence for measuring diffusion

coefficients by NMR spectroscopy was introduced by Stejskal and Tanner.

42-44

Introduction

3

Nowadays diffusion NMR spectroscopy is used in many different fields

45-50

, such as

medical

51

and material science.

52-54

The dyes “Rot 2G” (RD) and “Gelb GA” (YD) used in this study, which are anionic

direct dyes are employed as colorants in the production of tinted cellulose. A

mixture of those dyes is used to produce orange-hued paper. However, in the

presence of alkaline-earth ions, such as Mg

2+

, aggregates are formed, which

interfere with this dyeing process. Time-resolved static light scattering observed

within 15 − 20 minutes after mixing the two dye solutions, has shown the

formation of short aggregates with a contour length of 250 − 300 nm in the

presence of Na

2

SO

4

, whereas substituting Na

+

ions with Mg

2+

ions leads to a nearly

tenfold increase of the contour length.

55,56

A RD/YD molar ratio 1/1.16 for the

aggregates was found.

57

In this doctoral thesis, the pure YD was used (no salt added) as a model molecule

in order to demonstrate the capabilities of methods for studying dimensional and

structural properties of the aggregates. NMR relaxometry

58-62

and diffusion NMR

(using the pulsed field gradient technique),

28,45-47,63-65

as well as other methods, such

as polarizing light microscopy and UV/VIS spectroscopy were applied to investigate

the aggregates. The NMR chemical shift analysis together with the quantum-

chemical computatios

66,67

was carried out to study the structure of aggregates.

Theroretical aspects

4

CHAPTER 2 THEORETICAL ASPECTS

2.1

O

PTICAL PROPERTIES OF

J-

AND

H-

AGGREGATES

The optical properties of aggregates can be described by Davydov´s exciton

68

model by using a quasi-classical vector-model which considers the electrostatic

interaction of the transition dipole moments

69

. The excited state resonance

interaction is approximated by considering the electrostatic interaction of

transition dipole moments. If there is an electrostatic repulsion between the

transition dipole moments, the energy of the dimer will be higher compared to the

energy of the monomer.

The absorption of the electromagnetic irradiation is connected to the transition

dipole vector µ.

70

F

IGURE

2.1: Vector model - energy diagram for the alignment of the transition dipoles in

dimers and the monomer

70

(Figure from ref. 70).

In the easiest case the dimers are coupled at their chromophores whereby the

transition dipoles can interact. The excited state is split into two separate levels.

Theroretical aspects

5

Depending on the relative directions of the transition dipoles of the aggregate

there can be different borderline-cases. The total transition state moment arises

from the sum of the vectors in the single molecules. If we are stacking the

chromophores then both parallel and antiparallel dipole directions are possible

(Fig. 2.1B). If this direction is antiparallel then the transition dipole is 0, i.e. the

transition is electrically forbidden

71

.

However, if the transition dipoles are parallel then a light absorption can occur

which is shifted hypsocrome in comparison to the monomer since the dipoles of

the excited state are higher due to the electrostatic repulsion. This is called an H-

aggregation (H like hypsochrome), because the absorption peak of the aggregate

appears at lower wavelengths compared to the monomer absorption.

If the chromophores are linked head to tail behind each other and their transition

dipoles point into the same direction then the sum of the transition moments can

differ from zero only if the transition dipoles are parallel (Fig. 2.1C). In this case, the

energy of the exited state is lower than that of the monomer and we can observe a

bathochromically shifted J-band (called J after the discoverer, Jelly) in the

absorption spectra. Such aggregates are called J-aggregates. If the chromophores

(and the transitions dipoles) of the dimer are tilted then both the attractive and

repulsive combination of the dipoles result in a non-zero transition moment,

where the vectors of the transition dipoles for the two transitions are vertical to

each other (Fig. 2.1D). These aggregates show two different absorptions where one

has a bathochromic and the other one has a hypsochromic shift in comparison to

the absorption band of the monomer.

For aggregate stacking an important property is the shearing angle α ≤ 90°, which

is the acute angle between the axis of the transition dipoles and the line

connecting the centers of neighboring molecules (Fig. 2.2).

Theroretical aspects

6

F

IGURE

2.2: Representation of the terms shearing (A) and screwing (B) for the orientation

of the transition dipoles of the dimer

70

(Figure from ref. 70).

If the aggregate is even helically twisted then this torsion can be described by a

dihedral angle Θ as shown in Fig. 2.2. According to the point dipole approximation

the interaction energy V

i,j

of the neighboring transition dipoles depends on the

value of the transition moment of the monomer, the intermolecular distance r

i,j

and the geometry of the angles in the aggregate α

i

, α

j

and Θ (Fig. 2.2 B2). The

excition-splitting (or Davydov-Splitting) of an aggregate of N molecules can be

described as follows:

)coscos2sinsin(cos

r

M

N

1N

4V2

jiji

3

j,i

2

j,i

αα−ααΘ























−

=

→

[2.1]

In a dimer stacked in a parallel way without screwing (α

i

= α

j

, Θ = 0°) at a shearing

angle of α = arc cos 1/√3 = 54.7° the excition splitting is zero. α > 54.7° corresponds

to an H-aggregate and for α < 54.7° to a J-aggregate. In both cases there is only one

allowed transition, which can be described by the in-phase interaction of the

Theroretical aspects

7

transition dipoles. If the molecules in the aggregate are twisted a dihedral angle Θ

≠ 0° occurs and the intensity of the formerly forbidden transition dipole increases

and an additional absorption is allowed. One exemption is the case when α = Θ =

90°, where the transition dipoles are vertical to each other and therefore the

dipole-dipole interaction becomes zero. The general case (α

i

≠α

j

) considers an

additional tilted orientation of the dipoles.

In an aggregate of N linked chromophores an exciton band with N excited states

can be observed. Significantly, in a classical H- or J-aggregate only the highest and

the deepest excited state of the aggregate can be obtained via electrically allowed

transitions.

71

Several examples about the investigation of the optical properties of J- and H-

aggregates are available in the literature.

16,27,72,73,26,74-77

2.2

NMR

SPECTROSCOPY

Nuclear magnetic resonance spectroscopy exploits the magnetic properties of the

nuclei

78-81

and it can provide physical, chemical, electronic and structural

information about molecules in the solid or liquid phase

82-86

.

2.2.1 O

NE

-D

IMENSIONAL

NMR

TECHNIQUES

Nuclei with non zero spin (I≠0), when immersed in a static magnetic field, adopt

2I+1 spin orientations, each with different energy. The difference between the

energy levels depends on the magnetic moment of the nuclei and the strength of

the magnetic field. Transitions of the nuclei between the energy levels can be

induced by a radio frequency (rf) pulse (if the frequency of the electromagnetic

radiation fulfills the resonance condition ∆E=hυ). This is the principle of

NMR.

79,87,81,80

Theroretical aspects

8

One-dimensional (1D) NMR experiments contain two stages: preparation and

detection. The preparation phase is the excitation by using one rf pulse or several

pulses, while during the detection phase the resulting signal is recorded.

Chemical shift

A general feature of NMR spectroscopy is the dependence of the observed

resonance frequencies on the local environments of the individual nuclei. The

differences in frequency are referred to as chemical shift; they offer the possibility

to distinguish between nuclei in different chemical environments.

78

The effect of the small magnetic field, which is caused by the motion of the

electrons, is called nuclear shielding. The strength of the shield depends on the

molecular environment in that the nucleus is embedded. Therefore the changes in

molecular environment lead to changes in the peak position. For example in case

of aromatic stacking a change of the

1

H NMR chemical shift can occur due to the

shielding of the aromatic protons by neighboring molecules.

66,67

Chemical exchange

60,61

means that a nucleus moves from one environment to

another. For example this can occur in case of isomers or aggregating systems.

Exchange processes are classified by their rate relative to the NMR timescale. There

is slow exchange, which is defined by a situation where characteristic individual

sites can be observed in the spectra by their individual signals, and fast exchange,

which yields a spectrum time-averaged over the different sites.

47

Between these extremes a rich variety of chemical exchange line shapes can be

observed, as illustrated for azapropazone in Fig 2.3.

60

Based on the conjugation

between the nitrogen ion pairs and the aromatic π system a partial double bond is

formed which is a significant barrier to the rotation of the dimethylamino group.

The barrier height is comparable to thermal energies at accessible temperatures, so

Theroretical aspects

9

that the phenomenon of chemical exchange between the two methyl sites can be

seen in the NMR spectra. At low temperature, two separate methyl peaks are seen,

but with increasing temperature the rate of rotation about the bond increases. The

lines broaden and then coalesce.

In case of an aggregation, there can be fast or slow exchange between the

monomer and the aggregates, thus there can appear just one average signal in the

1

H NMR spectrum or individual peaks of both the monomers and the aggregates.

F

IGURE

2.3: Proton NMR spectra at 300 MHz of the N-methyl signals in a derivative of

azapropazone (structure shown in the figure) as a function of temperature. The lowest

spectrum is at 223 K, and then at 243, 253, 263, and 273 K

60

(Figure from ref. 60).

Theroretical aspects

10

2.2.2 T

WO

-D

IMENSIONAL

NMR

TECHNIQUES

Two-dimensional (2D) NMR spectra provide more information about a molecule

than one-dimensional NMR spectra and are especially useful in determining the

structure of a molecule, particularly for molecules that yield spectra too

complicated to be interpreted using one-dimensional NMR or when more detailed

investigations are required.

88

A two-dimensional NMR method involves a series of one-dimensional experiments.

Each experiment consists of a sequence of radio frequency pulses with delay

periods in between them. It is the timing, frequencies, and intensities of these

pulses that distinguish different NMR experiments from one another. During some

of the delays, the nuclear spins are allowed to freely process for a determined

length of time known as the evolution time. The frequencies of the nuclei are

detected after the final pulse by recording the resonance signals. By incrementing

the evolution time in successive experiments, a two-dimensional data set is

generated from a series of one-dimensional experiments.

In this thesis the following 2D NMR techniques were used:

79,81

COSY (Correlation Spectroscopy)

COSY shows homonuclear correlation due to J couplings between protons. Two-

and three-bond correlations yield COSY signals.

HMQC (Heteronuclear Multiple Quantum Coherence)

The HMQC experiment provides correlations between protons and their attached

heteronuclei through the heteronuclear J coupling. This is the most common two-

dimensional experiment because of its high sensitivity.

HMBC (Heteronuclear Multiple Bond Correlation)

The HMBC experiment detects long range coupling between proton and carbon

(two or three bonds away) with great sensitivity.

Theroretical aspects

11

2.2.3 D

IFFUSION AND

R

ELAXATION

Self-diffusion is the random translational motion of species

(molecules/ions/aggregates) driven by internal kinetic energy.

54

Translational

diffusion is the most fundamental form of transport and is responsible for all

chemical reactions, since the reacting species must collide before they can react.

Diffusion is also closely related to molecular size. According to Einstein the

diffusion coefficient D depends on the friction factor f:

f

Tk

D

B

=

[2.2]

where k

B

is the Boltzmann constant and the T is temperature. For the simple case

of a spherical particle with an effective hydrodynamic radius R

h

in a solution of

viscosity η the friction factor is given by the Stokes relationship:

h

R6f

πη

=

[2.3]

In NMR spectroscopy there are two main ways to study self-diffusion coefficients,

which are also known as tracer diffusion or intradiffusion coefficients: analysis of

relaxation data and pulsed-field gradient (PFG) NMR. However, the two methods

report on motions with very different time scales and thus, even though a

translational diffusion coefficient can be derived in both cases. The two estimates

will agree only under certain circumstances (in case of medium size and not very

mobile species) since the relaxation method is in fact sensitive to rotational

diffusion, which is proportional to the relaxation time, thus the molecular size,

whereas the PFG method measures translational diffusion form the signal

attenuation.

80

Theroretical aspects

12

Relaxation and Linewidth

Any excited magnetic moment relaxes back to equilibrium, which is called

relaxation

79,80

. In other words, relaxation times describe how fast spins “forget" the

direction in which they are oriented. There are two types of relaxation for isotropic

systems in the absence of chemical exchange: longitudinal or spin-lattice (T

1

) and

transverse or spin-spin (T

2

) relaxation. Different physical processes are responsible

for the relaxation of the components of the nuclear spin magnetization vector M

parallel (longitudinal) and perpendicular (transverse) to the external magnetic

field, B

0

(which is conventionally oriented along the z axis).

In principle in liquids T

2

can be obtained by measuring the signal width at half-

height:

2

T

1

π

=υ∆

[2.4]

However, the line width for non-viscous liquids is most often dominated by field

inhomogeneity. The experimental relaxation rate, extracted from the line width, is

called 1/T

2

*

hom)in(22

*

2

T

1

T

1

T

1+=

[2.5]

In modern NMR spectroscopy T

2(inhom)

/ π is on the order of 1 Hz.

The transverse relaxation rate (T

2-1

) values are approximately proportional to the

overall rotational correlation time

80

of the big molecules, such as proteins or

aggregates and thus depend on molecular mass and shape of the molecules,

solvent viscosity and temperature. In case of small molecules this is not true base

on the high mobility of the molecules.

Theroretical aspects

13

The correlation time for rotational diffusion can be measured experimentally or

calculated by using a variety of hydrodynamic theories. In the absence of more

accurate information, the simplest theoretical approach for approximately

spherical molecules calculates the isotropic rotational correlation time from Stokes

law:

Tk3

R4

B

3

h

c

πη

=τ

[2.6]

where η is the viscosity of the solvent, R

h

is the effective hydrodynamic radius of

the molecule, k

B

is the Boltzmann constant and T is the temperature.

80

Diffusion NMR − PFG pulse sequences

Pulsed-field gradient NMR spectroscopy has become a convenient method for

measuring diffusion in solution.

86,89,90,45-48,50,91,49

The diffusion coefficient of a

molecule is a function of its effective molecular weight, size, and shape, so the

measured diffusion data can be used to study molecular dimensions. Gradient

NMR spectroscopy is a powerful tool not only for studying diffusion and

dimensions of the species but it also provides structure information about cavities

in cells or zeolites in the range of 0.1–100 mm when the diffusion is restricted on

the NMR timescale.

86

The basis for diffusion measurements is that magnetic field gradients can be used

to label the spatial position of nuclear spins through their Larmor frequency,

00

B

γ

=

ω

[2.7]

given in radians per second. γ is the gyromagnetic ratio, given in rad/Ts and B

0

is

the strength of the static magnetic field. If B

0

is spatially homogeneous throughout

the sample, ω

0

is the same over the whole sample but using an additional space

Theroretical aspects

14

dependent B

0

field (gradient pulse) the effective Larmor frequency will depend on

the position of the nuclei.

Gr)r(

0eff

γ

+

ω

=

ω

[2.8]

r is the coordinate vector of the nuclei. G is the gradient of the magnetic field,

according to equation [2.9], where i, j, and k are the unit vectors in the x, y, and z

directions, respectively, of the laboratory frame of reference. The gradient can be

applied along three directions.

k

z

B

j

y

B

i

x

B

BG

zzz

0

∂

+

∂

+

∂

=∇=

[2.9]

If a gradient of known magnitude is applied over a defined time period, the Larmor

precession yields an additional phase shift that is dependent on the spatial position

of the spin, the direction of the gradient, and the duration and strength of the

gradient.

If the gradient is applied along the z axis (parallel to B

0

) the phase shift for a single

spin is described by equation [2.10]

dt)t(z)t(GB)(

0

∫

δ

γ+δγ=δφ

[2.10]

δ is the duration of the gradient and t is the integration variable and z(t) the time

dependent z position of the nucleus.

In order to measure diffusion the Hahn spin echo

41,40

pulse sequence was modified

by Stejskal and Tanner.

43,42

The pulsed-field gradient spin echo (PFG-SE) sequence

contains a 90° pulse as an excitation pulse and a 180° pulse, which flips the

Theroretical aspects

15

magnetization of the spin. Thus exactly the same gradient pulses (length, strength

and direction of the pulses) can be applied (Fig. 2.4)

F

IGURE

2.4: Schematic representation of the Stejskal and Tanner pulse sequence. In each

delay, τ, a gradient pulse of duration δ and magnitude G is inserted.

After the 90° pulse the magnetization of the system flips from the z plane to the x-y

plane. The first gradient pulse with duration δ and magnitude G is applied and the

spins experience a phase shift, according to [2.10]. Due to the inversion of the

phase shifts acquired from the first gradient by the 180° pulse, the second gradient

pulse will refocus the magnetization of all the spins dephased by the first gradient.

The spins that did not change their z-position are fully refocused, but those spins

that moved during the evolution interval ∆ are not fully refocused. Since the

phases are averaged over all spins in the sample the echo is not phase shifted but

attenuated.

In addition to the echo attenuation due to diffusion there is also a decrease of the

echo amplitude due to relaxation. The echo amplitude is given by equation [2.11].

Theroretical aspects

16

43421

diffusion

relaxation

2

0

)D,,G,(f

T

2

expS)2(S ∆δ













τ

−=τ

[2.11]

If τ is kept constant during all the experiments, the relaxation-induced signal

attenuation is constant and can be separated from diffusion-induced attenuation.

The relationship between the diffusion and the observed signal attenuation is

described by equation [2.11]. A step-wise time-dependent integration of the Bloch

equation for the given pulse sequence results in:













δ

−∆δγ−=

3

222

DGSln

[2.12]

The attenuation of the signal depends on the gyromagnetic ratio (γ) of the

observed nuclei and the parameters of the diffusion measurement, such as gradient

pulse length (δ), gradient strength (G) and diffusion time (Δ).

For small molecules the spin echo sequence can be applied because T

2

relaxation is

slow and sufficient signal is obtained even for long diffusion time (Δ). In case of

medium size molecules the T

1

relaxation time has a minimum and it increases for

very small and very big molecules. Contrarily the T

2

relaxation time is high for small

molecules and decreases for increasing molecular size. Thus the T

2

relaxation time

of big molecules is much shorter than their T

1

relaxation time (T

1

>>T

2

). Because of

the limitation of the T

2

relaxation time the spin echo experiment cannot be

applied for large molecules.

The stimulated-echo (STE) pulse sequence with pulsed-field gradients, which is

presented in Fig. 2.5, is applied for measuring diffusion of macromolecules or

aggregates. In this pulse sequence the magnetization is “stored” along the z axis

during the τ

2

period and therefore subjected only to the lower longitudinal T

1

relaxation.

Theroretical aspects

17

F

IGURE

2.5: The stimulated-echo (STE) pulse sequence with pulsed-field gradients. During

the τ

2

period, magnetization is stored along the z axis and therefore subjected only to

longitudinal T

1

relaxation. G

s

indicates a spoil gradient, which eliminates the remaining x-y

components.

Temperature gradient and convection – DSTE sequence

A typical problem affecting diffusion measurements is the convection within a

sample, especially at elevated temperature.

92

Convection currents are generated by

small temperature gradients in the sample and cause additional signal decay. This

artifact can corrupt the diffusion measurements. Some improvement can be

achieved, for instance by sample rotation,

93

special sample cells or use of transverse

gradients, but the perfect elimination of this problem is difficult. Jerschow and

Müller developed a pulse sequence

94

which suppresses the convection effects in

stimulated echo diffusion experiments to first order, provided that the convection

current has a constant laminar flow profile during the diffusion interval of the

pulse sequence. This condition is fulfilled in many systems of practical relevance.

However, this method is not effective for turbulent convection. Convection

compensation sequences are based on the pulse scheme shown in Fig. 2.6 in which

only the gradient pulses are shown.

Theroretical aspects

18

F

IGURE

2.6: Basis of convection compensation sequence.

As can be seen in Fig. 2.6 part 1 and part 2 of the pulse sequence only differ in the

signs of the gradients. If there is laminar convection present, the effect during the

first part is expressed by:













∆

δγ 2

Givexp

[2.13]

while during the second part this is













∆

δγ− 2

Givexp

[2.14]

which results in a total effect of:

1

2

Givexp

2

Givexp =













∆

δγ−













∆

δγ

[2.15]

where v is the flow velocity, G is gradient strength and Δ is the diffusion time.

By choosing both values of Δ in equation [2.15] equal the effect of laminar

convection is completely removed. The physical reason for this is that in the case of

laminar convection the displacement of a molecule during the second part of the

sequence is exactly the same as during the first part, and thus the value of the

Theroretical aspects

19

phase difference obtained is also the same. However, because the gradient is

oppositely signed, the phase difference is also oppositely signed. Therefore the total

phase difference is zero. In case of turbulent convection the displacement in the

first and second parts are not equivalent so the phase differences do not have the

same value.

Theroretical aspects

20

2.3

C

OMPUTER

S

IMULATION

Recent years have witnessed an increase in the number of people using

computational chemistry to understand a problem more completely.

35,37,95,96,66,97,98

There has been an enormous progress in the field of computational investigations

of chemical and biochemical systems over the last decade. There are some

properties of a molecule that can be obtained computationally more easily than by

experimental means. Additional insights into molecular bonding can be obtained

from the results of computations that cannot be obtained from any experimental

method. Molecular modeling and simulation methods can address fundamental

questions that cannot easily be answered experimentally. Thus, many experimental

chemists are now using computational modeling also to gain additional

understanding of the compounds being examined in the laboratory.

NMR spectroscopy is a valuable technique for harvesting molecular information.

The structure of the molecule and the NMR spectra is in strong relation, thus the

spectrum is very sensitive to a change of the molecular structure.

39

Before an NMR spectrum can be calculated, a good model of the true molecular

structure must be obtained by a computational procedure, known as structure

optimization, in which the energy minimum is searched for by variation of the

atomic coordinates.

Density functional theory

38

(DFT) is a quantum mechanical theory used in physics

and chemistry to investigate the electronic structure (principally the ground state)

of many-body systems, in particular atoms, molecules and the condensed phases,

where the electron density is expressed as a linear combination of basis functions

similar in mathematical form to Hartree-Fock (HF) orbitals. A determinant is then

formed from these functions, called Kohn-Sham orbitals, which is then used to

compute the energy. The major problem with DFT is that the exact functionals for

exchange and correlation are not known except for the free electron gas. However,

approximations exist which permit the calculation of certain physical quantities

quite accurately. One of the most popular functional is the BLYP (from the name

Theroretical aspects

21

Becke for the exchange part and Lee, Yang and Parr for the correlation part). Even

more widely used is B3LYP, which is a hybrid functional in which the exchange

energy, in this case from Becke's exchange functional, is combined with the exact

energy from Hartree-Fock theory. Along with the component exchange and

correlation functionals, three parameters define the hybrid functional, specifying

how much of the exact exchange is mixed in. The adjustable parameters in hybrid

functionals are generally fitted to a 'training set' of molecules. Hence in the current

DFT approach it is not possible to estimate the error of the calculations without

comparing them to other methods or experiments.

After having obtained an accurate structure, the spectra can be calculated, such as

NMR or UV spectra. NMR chemical shifts can be computed from the shielding

tensor of atoms

37,95,98

in simple model systems incorporating functional groups that

exert through-space effects. Once the shielding tensors have been computed, the

chemical shifts can be determined by subtracting the isotropic shielding values for

the molecule of interest from the TMS values. Computing shielding tensors is

difficult because of gauge problems (dependence on the coordinate system's

origin). It is extremely important that the shielding tensors be computed for

equilibrium geometries with the same method and basis that were used to

complete the geometry optimization.

One of the most popular techniques is called GIAO. This originally stood for gauge

invariant atomic orbitals. More recent versions have included ways to relax this

condition without loss of accuracy and subsequently the same acronym was

renamed gauge including atomic orbitals. The GIAO method is based on

perturbation theory. It is a means for computing shielding tensors from HF or DFT

wave functions.

It is also possible to calculate the electronic excited states of a molecule. These

calculations are an important tool for the analysis of UV spectroscopy.

Experimental

22

CHAPTER 3 EXPERIMENTAL

3.1 M

ATERIALS

The azodyes “Rot 2G” (RD) and “Gelb GA” (YD) (Fig.3.1) are anionic direct dyes

that were provided by Ciba Spezialitätenchemie AG. The dyes in form of their salts,

RDNa

4

and YDNH(C

2

H

4

-OH)

3

are employed as colorants in the production of

tinted cellulose.

S

SS

S

N

NN

N

NN

N

H

HH

H

N

NN

N

NN

N

NN

N

O

OO

O

N

NN

N

H

HH

H

C

CC

CN

NN

N

H

HH

H

3

33

3

C

CC

C

S

SS

SO

OO

O

3

33

3

H

HH

HO

OO

O

N

NN

N

NN

N

H

HH

H

S

SS

SO

OO

O

3

33

3

O

OO

O

S

SS

SO

OO

O

3

33

3

N

NN

N

H

HH

H

N

NN

N

H

HH

H

O

OO

O

OO

O

N

NN

N

NN

N

H

HH

H

O

OO

O

3

33

3

S

SS

S

O

OO

O

3

33

3

S

SS

S

Y

YY

Y

D

DD

D

R

RR

RD

DD

D

F

IGURE

3.1: Chemical structures of the azodyes used: anion of the yellow dyestuff (YD) and

anion of the red dyestuff (RD).

For the investigations, the two dyes were used in their sodium salt form (the molar

mass of the anions are M

YD

= 505 g/mol and M

RD

= 1061 g/mol). In case of YD an ion

exchange was necessary. The purification of the dyestuffs was achieved by dialysis

of aqueous solutions of the raw material to produce the pure sodium salt of the

dyes which was used for further investigations.

Experimental

23

3.1.1 S

AMPLE PURIFICATION

The dialysis was carried out with dialysis hoses (type: 50303, Reichelt). The

diameter and length of the hoses were 12 and 25 cm, respectively. For each

dyestuff, two hoses were filled with a 250 ml of concentrated solution (4 g/L) of

dyestuff. Each hose was sealed with clamps at both ends and placed in a 10 L vessel.

The vessel was completely filled with the dialysis medium. Dialysis was performed

at first against an aqueous solution (1.2 g/L) of NaCl for 2 days, and the dialyzing

medium was changed twice a day.

At the beginning of the fourth day, the NaCl solution was replaced by distilled

water as the dialyzing medium, and the dialysis was continued for 3 days against

distilled water which was again changed twice a day. The endpoint of the dialyzing

process was indicated by subjecting the dialyzing medium to a precipitation test

for Cl

–

ions with an Ag

+

solution. In all cases, the test was negative after 3 days

meaning that there were no Cl

–

ions in the dye solution.

After the dialysis, the water content in the hoses had increased by about 50 %

because of water penetrating the membrane. To recover the dyestuff, their

aqueous solutions were first concentrated by rotation evaporation of about 90 %

of the water at 30 °C under reduced pressure. The remaining water was then

removed by freeze-drying of the concentrated dyestuff solutions.

Experimental

24

3.1.2 S

AMPLE PREPARATION

Light microscopy

Three different samples in aqueous solution were prepared. 66 mg dye (YD, RD, or

1:1 w/w% MIXTURE) and 4 mg MgSO

4.

7H

2

O (Magnesium sulfate, Merk, p.a., > 99,5

%) were dissolved in 1 mL H

2

O in each case. Moreover three further samples were

prepared in DMSO (Dimethyl sulfoxide-d6, Deutero GmbH, 99.8%). After mixing

the components, one droplet of the solution was placed on a glass slide and then

covered by another one.

The concentrations of the dye solutions used are listed in Table 3.1 and Table 3.2.

T

ABLE

3.1

Dye samples in DMSO solution for light microscopy measurements.

The dyes were dissolved in 1 mL DMSO

Dye Dye-amount (mg) c

Dye

(mM)

YD 66 130

RD 66 62

1:1 w/w MIX 66 96

T

ABLE

3.2

Dye samples in aqueous solution for light microscopy measurements.

The dyes were dissolved in 1 mL magnesium sulfate solution ([MgSO

4

]=33 mM) in D

2

O

Dye Dye-amount (mg) c

Dye

(mM)

YD 66 130

RD 66 62

1:1 w/w MIX 66 96

Experimental

25

UV measurements

A small amount of the dyes was weighted using an analytical balance and the

desired amount of solvent was added with a micropipette. In order to prepare the

0.07 mM samples the 0.7 mM sample was diluted. Due to the limited amount of

the dyes, the preparation of a dye stock solution was not possible. The

measurements were performed at room temperature immediately after the

preparation. Because of the high absorbance in the visual range of YD a special

thin sample holder was created from two agglutinated glass slides (Menzel Glaeser

76 x 26 mm) which were separated by a 25 μm foil. A foil frame was glued to the

first glass slide and a small drop pf the solution was added to the center of the slide

and then covered by the second slide. Finally the slides were glued together, thus it

was not refillable. The glass sides have no absorption in the UV range so the

measurements are not affected by the glass. The reference sample holder was not

chanced during the measurements contrary to the sample holder with the dye

solutions. Thus the background correction can be imprecise.

T

ABLE

3.3

Dye samples for UV measurements

Solvent YD (mg) c

YD

(mM)

1 ml DMSO 0.074

0.07

1 ml D

2

O 0.074 0.07

1 ml D

2

O 0.74 0.7

1 ml D

2

O 1.33 2.5

1 ml D

2

O 2.65 5

1

H,

13

C, COSY, HMQC and HMBC NMR spectroscopy

5 mg of the sample (YD or RD) was put in a plastic sample holder and 0.5 ml of

DMSO was added. The sample holder was shaken manually for about a minute and

then the solution was injected in a 5 mm NMR tube. A

1

H NMR spectrum was

Experimental

26

recorded immediately after the sample preparation (time zero). The chemical shift

values are given in ppm relative to TMS (based on the known chemical shift of the

solvent peak). The measurements were carried out at room temperature. Before

and after each 2D measurement a 1D

1

H spectrum was recorded to confirm the

stability of the sample.

T

ABLE

3.4

Dye samples for the

1

H,

13

C, COSY, HMQC and HMBC measurements

Dyes Dye amount (mg) c

YD

(mM)

YD + 0.5 ml DMSO 5

20

RD + 0.5 ml DMSO 5 10

1

H Diffusion NMR and temperature dependent

1

H NMR spectroscopy

An amount of the sample determined by the desired analytical concentration was

put in a plastic tube and 0.5 ml of D

2

O, containing 0.18 mM DSS (2,2-dimethyl-2-

silapentane-5-sulfonic acid) as internal standard, was added. The sample holder

was shaken manually for about a minute and then the solution was injected in a 5

mm NMR tube. The

1

H NMR spectrum of time zero was recorded immediately and

the change of the spectra was monitored afterwards as a function of time. The

solutions were stable for weeks, since the corresponding

1

H NMR spectrum was

unchanged. Before and after each 2D measurement a

1

H spectrum was recorded to

confirm the stability of the sample. The viscosity data of D

2

O were taken from the

literature.

99

Experimental

27

T

ABLE

3.5

Dye samples for the

1

H diffusion NMR and temperature dependent

1

H NMR

measurements

Solvent YD (mg) c

YD

(mM)

0.5 ml DMSO 5

20

0.5 ml D

2

O 0.36 0.7

0.5 ml D

2

O 0.66 2.5

0.5 ml D

2

O 1.36 5

0.5 ml D

2

O 2.73 10

0.5 ml D

2

O 4.18 15

0.5 ml D

2

O 5.31 20

0.5 ml D

2

O 6.58 25

0.5 ml D

2

O 8.10 30

Experimental

28

3.2 T

ECHNIQUES

3.2.1 P

OLARIZING

M

ICROSCOPY

The microscopy experiments were performed in transmission mode under crossed

polarizers using a Leitz ORTHOLUX 2 POL microscope. A lens (L5) with fivefold

magnification was used. The pictures were recorded with a JVC TK-C1381 color

video camera. The samples were measured at room temperature.

3.2.2 UV

SPECTROSCOPY

The UV absorption spectra were recorded with a Lambda 19 Spectrometer from

Perkin Elmer. The wavelength interval of the measurement was from 350 nm to

600 nm, the resolution was 0.1 nm and the scanning speed was 240 nm/min.

Distillated water or DMSO was used as reference depending on the solvent of the

sample. Before every measurement background correction was carried out to

reduce artifacts.

3.2.3 NMR

SPECTROSCOPY

One- and two dimensional (1D, 2D) NMR experiments in DMSO, such as

1

H,

13

C,

COSY, HMQC, HMBC, were performed on a BRUKER ARX 500 NMR spectrometer

operating at 500 and 125 MHz for

1

H and

13

C, respectively, using Bruker standard

pulse sequences. The NMR chemical shift values are given in ppm relative to TMS.

The diffusion measurements and the

1

H NMR measurements for temperature

dependent

1

H chemical shift and linewidth analysis were performed on a Bruker

DMX-200 spectrometer (Lund University, Sweden) operating at 200.11 MHz for

1

H.

The spectrometer was equipped with a Bruker diffusion probe, capable of

Experimental

29

generating field gradients up to 9 T/m. The temperatures of the measurements

were varied between 10 and 70 °C. The spectra were taken forty minutes after

changing the set temperature.

Up to 30 °C the standard Bruker pulsed-gradient stimulated echo sequence was

used and at higher temperature the double stimulated echo sequence with square

shaped gradients. The pulse programs are shown in the appendix. The gradient

pulse duration was 0.5 ms. The diffusion times and gradient strengths were

optimized for each experiment. The diffusion times were 20 ms, 40 ms and 100 ms.

The gradient strength was varied between 0 and 9 T/m in steps 0.5 T/m. The

measured data were analyzed by the Bruker software package (Topspin) and a

Matlab script (The original script from Lund University).

DSS as internal standard was used in aqueous solution. The ppm chemical shift

values were measured with respect to the internal standard peak, but reported

relative to TMS. The recorded NMR spectra with or without the internal standard

were identical, except for the presence of the internal standard signal.

In all NMR experiments the optimal number of scans (16-256) was used to reach a

good signal to noise ratio depending on the dye concentration.

3.2.4 C

OMPUTER SIMULATION

The program package NWCHEM was used to find the starting structure of the

monomer and the dimer. The Gaussian03 computational chemistry software was

used to determine the structure of YD and to simulate the NMR and UV spectra.

The simulations were performed on the RZ cluster at RWTH Aachen and on the

Arminius cluster of the University of Paderborn using 4-16 processors. GaussView

and MOLDEN programs were used to visualize the input and output structures of

the molecules.

Results and Discussion

30

CHAPTER 4 RESULTS AND DISCUSSION

4.1 O

PTICAL PROPERTIES OF DYES

Precipitation was observed by visual inspections with the naked eye in case of dye

mixtures and yellow dye at high (> 62 mM) concentrations in aqueous solution in

the presence of positively charged ions. The motivation for microscopic

experiments was to investigate the size and the shape of the aggregates at higher

concentration where they are big enough to be detected by microscopy (the size of

the particles is in the µm range). Thus, yellow (YD) and red (RD) dyes, as well as

their 1:1 mixture (MD) were investigated by light microscopy in aqueous solution

with Mg

2+

added. UV/Vis spectroscopy provided further information of the

aromatic stacking (Chapter 2.1) about the pure YD based on the color change by

increasing the dye concentration.

4.1.1 P

OLARIZING MICROSCOPY

Light microscopy was used to obtain information regarding the size and shape of

the aggregates. This method, however, is applicable only in the μm diameter range,

which means, in our case, that it can be applied when the concentration of the dye

and the Mg

2+

is relatively high (over 62 mM, dye concentration). During these

experiments, only aqueous solutions of the dyes were examined (since no

aggregates are formed in DMSO).

Fig. 4.1 shows the polarizing microscopy picture of RD. No aggregates are visible

under these conditions ([dye]= 62 mM; [MgSO

4

]= 33 mM at room temperature).

Hence the dye is in perfect solution or any aggregates are below the detectable size.

Under similar circumstances ([dye]= 130 mM, [MgSO

4

]= 33 mM at room

temperature) YD forms aggregates, which are in the observable size range as shown

in Fig. 4.2. The aqueous dye solution contains clearly visible aggregates of various

Results and Discussion

31

shapes and sizes. Treating the 1:1 mixture of the two dyes in a similar way, the

resulting system is composed of huge aggregates as shown in Fig. 4.3.

F

IGURE

4.1: Light microscope picture of RD in aqueous solution [dye]= 62 mM,

[MgSO

4

]= 33 mM at room temperature. The picture was taken two hours after sample

preparation. No aggregates can be observed.

F

IGURE

4.2: Light microscope picture of YD in aqueous solution [dye]= 130 mM,

[MgSO

4

]= 33mM at room temperature. The picture was taken two hours after sample

preparation. The size of the aggregates is up to hundred micrometers.

Results and Discussion

32

F

IGURE

4.3: Light microscope picture of 1:1 w/w mixture of YD:RD in aqueous solution

[dye]= 96 mM, [MgSO

4

]= 33 mM at room temperature. The picture was taken two hours

after sample preparation. Huge aggregates can be observed.

It is obvious that, although optical microscopy is a powerful technique, it has its

limitation, being unable to detect aggregates smaller than about one μm. Since this

method can only be used at higher concentrations and with positively charged

metal ions (such as Mg

2+

) added, other techniques are necessary to investigate

small aggregates.

In case on YD and the mixture of YD and RD aggregation was observed in aqueous

solution. Because of the simple NMR spectrum of YD and the limited amounts of

dye available most of the following studies will focus on YD. The results of the

diffusion NMR measurements of the dye mixture are presented in the Appendix.

4.1.2 UV/VIS

SPECTROSCOPY

In Fig. 4.4A the UV/Vis absorption spectra of YD in water/DMSO at different

concentrations at room temperature are shown. Increasing concentration shifts

the maximum absorbance to lower wavelengths (blue shift). In case of the 0.07

mM solution in D

2

O the absorption between 500 and 600 nm is higher compared

to the other samples. Due to the home made sample holders used in the UV

measurement there may be additional absorption of the glass slides.

Results and Discussion

33

Plotting the maximum absorbance as a function of [YD] yields by Fig. 4.4B where

the aforementioned shift can be seen clearly. The absorbance at 0 concentration is

a hypothetical value referring to the solution in DMSO, where no aggregation

occurred. Within this concentration range (0-5 mM) the measured hypsochromic

shift in wavelengths is 16 nm, which is similar to what was observed in the case of

other dyes.

30,100-102,55

The results of these optical measurements provide us a hint of parallel transition

dipoles. In case of aromatic rings the molecular dipole moment, which is in the

molecular plane, is parallel to the transition dipoles.

69,71

Thus based on the exciton

model (Chapter 2.1), the shearing angle α is in the 54.7-90° range. In case of blue

shift there is a vertical stacking of the dye molecules in the aggregate, which is

called an H aggregate.

Fig. 4.5 shows the molar extinction coefficient as a function of wavelength for the

four different aqueous solutions of YD. The molar extinction coefficient is reduced

substantially at concentrations higher than 0.07 mM. In case of big aggregates the

light is also scattered and the intensity of the detected light is lower.

Results and Discussion

34

350 400 450 500 550 600

0.0

0.2

0.4

0.6

0.8

1.0

5 mM in D

2

O

2.5 mM in D

2

O

0.7 mM in D

2

O

0.07 mM in D

2

O

0.07 mM in DMSO

2

Absorbance

λ

(nm)

012345

416

418

420

422

424

426

428

430

432

434

436

λ (nm)

[YD] (mM)

F

IGURE

4.4: UV/Vis absorption spectra of YD in water/DMSO at different concentrations as

a function of wavelength at room temperature (A) and wavelength of maximum

absorbance as a function of YD concentration (B). The 0 concentration refers to the

absorbance in DMSO.

B

A

Results and Discussion

35

350 400 450 500 550

0

50

100

150

200

ε

(M

-1

cm

-1

)

λ

(nm)

0.07 mM

0.7 mM

2.5 mM

5 mM

F

IGURE

4.5: Molar extinction coefficients, of YD in water at different concentrations at

room temperature as a function of wavelength.

The UV/VIS spectra prove the formation of aggregates in dilute solutions even

without any salt added, where no visible precipitation occurred. Using further

optical studies

102

could provide more details about the aggregating system, but in

this work the results of the optical techniques was used primarily to confirm the

NMR experiments discussed in the following.

Results and Discussion

36

4.2 N

UCLEAR

M

AGNETIC RESONANCE

The main technique used in this thesis was NMR spectroscopy, which is a powerful

tool that provides essential information regarding the dimensional and structural

properties of the dyes in the molecular level. The chemical structure of the

molecules was reconfirmed by using high resolution liquid NMR methods, such as

1

H,

13

C, COSY, HMQC and HMBC. The peak assignment of the NMR spectra is also

required for interpreting the changes of chemical shifts with concentration and

temperature in order to get information about the structure of the aggregates. The

dimension of the species was estimated from the

1

H NMR linewidth and from the

pulsed field gradient diffusion NMR data.

4.2.1 P

EAK ASSIGNMENT

In the initial step of the NMR measurements an organic solution (DMSO) of YD

(Fig. 4.6) was examined by one- and two-dimensional NMR techniques to assign

the peaks of the spectra. Peak assignment is of importance for the detailed study of

the aggregation process in aqueous solution.

S

N

H

N

O

N

H

CN

H

3

C

SO

3

H

O

2

22

2

3

33

3

4

44

4

5

55

5

6

66

6

7

77

7

8

88

8

9

99

9

1

11

10

00

0

1

11

1

11

12

22

2

1

11

13

33

3

1

11

14

44

4

1

11

15

55

5

1

11

16

66

6

1

11

17

77

7

1

11

18

88

8

1

11

19

99

9

1

11

1

F

IGURE

4.6: Chemical structure of YD with numbered atoms. The numbers refer to the

carbon atoms and the directly attached hydrogen atoms.

Results and Discussion

37

The samples were prepared with deuterated DMSO as solvent to prevent

aggregation. The following NMR methods were used:

1

H-NMR,

13

C-NMR, COSY,

HMQC and HMBC. The results of these experiments are detailed below.

1

H-NMR

The measurements were preceded by a simple and fast estimation of

1

H and

13

C

chemical shifts using the ChemOffice Ultra 10.0 software to predict the peak

positions in the NMR spectra. The software carries out an estimation procedure

based on empirical methods. The simplest empirical calculation called group

additivity method is applied, so these can be performed very quickly on small

desktop computers. As a drawback, however, the estimation is of limited value for

atoms with unique or undocumented chemical environments.

The

1

H-NMR spectrum in DMSO is shown in Fig. 4.7. In accordance with

literature,

99

the solvent peak appears at 2.50 ppm. Since DMSO always contains

small amounts of water, a second solvent peak (HDO) is visible at 3.4 ppm.

According to the literature data aliphatic hydrogen peaks are expected to appear

at lower ppm (2-3) values. This, together with the fact that YD contains a single

aliphatic group, the methyl peak can be readily identified (2.7 ppm). The molecule

has six aromatic hydrogen atoms. Since 10 and 11 as well as 12 and 13 have similar

chemical environments, only four peaks are visible in the spectrum. The two peaks

showing double intensity are assigned to the hydrogen pairs 10-11, and 12-13. All

aromatic peaks appear as doublets due to scalar couplings. Identification of the

peaks referring to 10-11, 12-13, 3 and 4 is impossible based solely on the

1

H NMR

spectra.

Results and Discussion

38

F

IGURE

4.7: Measured

1

H NMR spectrum of YD ([YD]=130 mM) at room temperature in

DMSO solution. The expansions show the aromatic regime and the methyl peak.

OH and NH protons appear as broad peaks in

1

H NMR spectra as a result of the

chemical exchange between the hydrogen atoms of these groups and the solvent

molecules, making them easily recognizable. This broadening can be so excessive

that it leads to the disappearance of these peaks into the baseline. In the spectrum

of YD two such broad peaks can be observed, both with smaller maximum height

due to the line broadening by chemical exchange. These peaks at 10.9 and 15.7 are

assigned to OH and NH groups.

COSY

Correlated spectroscopy was used for determining signals which arise from protons

in YD that are coupled to each other through bonds. Two, three as well as four

bond correlations yield COSY signals, which is shown in Fig.4.8.

Results and Discussion

39

F

IGURE

4.8: Measured COSY spectrum of YD in DMSO-d6 at room temperature.

This spectrum was used to identify the peaks of the hydrogen atoms 3 and 4. The

1

H peak arising at 2.7 ppm from the methyl hydrogen atoms was already assigned

to hydrogen 1; therefore it was chosen as a starting point. The COSY spectrum

shows two cross peaks for hydrogen 1 which are encircled in Fig. 4.8: one with the

peak at 7.30 ppm (not very strong but visible) and another even weaker one with

the peak at 7.85 ppm. Since there are only two H-atoms (3 and 4) in the aromatic

ring that contains the methyl group, the cross peaks can be assigned to these

atoms. The stronger correlation peak belongs to those atoms, which are closer to

each other, thus the peak at 7.3 ppm is assigned hydrogen 3.

In Fig. 4.9 an enlargement of the aromatic region is shown. The peaks assigned to

hydrogen 3 and 4 show the expected cross peaks with each other. The remaining

two doublets at 7.65 ppm and 8.15 ppm, which are correlated only with each other,

had been assigned to hydrogen 10-11 and 12-13 already. The COSY results confirm

Results and Discussion

40

that the hydrogen atom pairs 10-11 and 12-13 are on the same ring but it is

impossible to identify which peak belongs to 10-11 and which one to 12-13.

F

IGURE

4.9: Aromatic hydrogen region of the measured

COSY spectrum of YD in DMSO.

Thus, the only new information from the COSY experiment is the assignment of

peaks 3 and 4 – everything else was known from the 1D spectrum.

13

C-NMR

The

13

C spectrum of YD in DMSO-d6 is shown in Fig. 4.10. According to the

literature data

99

the broad solvent peak appears at 40 ppm. YD molecule contains

only one aliphatic group, namely the methyl group, which appears at 20.74 ppm in

the

13

C spectrum.

Results and Discussion

41

F

IGURE

4.10: Measured

13

C NMR spectra of YD in DMSO.

The peaks of the aromatic carbon atoms appear between 110-180 ppm, resulting in

a somewhat crowded spectrum. Fig. 4.11 presents this range in higher resolution.

The peaks of the quaternary carbon atoms have long relaxation times (T

1

) because

there is no strong dipolar coupling that could provide a relaxation mechanism.

These peaks appear with low intensity and are broad (170.16, 163.84, 162.03,

122.60, 118.69 ppm). In order to gain further information, 2D methods were

applied.

Results and Discussion

42

F

IGURE

4.11: 110-180 ppm range of the measured

13

C NMR spectrum of YD in DMSO. The

unlabelled peaks with very low intensity result from impurities.

HMQC

The HMQC spectra of YD show the one-bond correlations between

13

C atoms and

1

H atoms (Fig. 4.12). Based on the previously assigned hydrogen atoms, the carbon

atoms can be assigned, which are directly bonded.

In the aliphatic range, the methyl carbon appears at 20.74 ppm. Furthermore the

solvent peak is recognizable at 40.05 ppm.

Results and Discussion

43

F

IGURE

4.12: HMQC spectrum of YD in DMSO at room temperature.

An enlargement of the aromatic region of the spectrum is shown in Fig. 4.13.

By using the estimated data (Table 4.1) and the previous assignments of hydrogen

atoms 3 (7.30 ppm), 4 (7.85 ppm), 10-11 (8.15 ppm) and 12-13 (7.65 ppm) the

directly attached carbon atoms are assigned: 3 (130.53 ppm), 4 (122.97 ppm) and

10-11 or 12-13 (128.83 ppm and 116.79 ppm)

Results and Discussion

44

F

IGURE

4.13: Aromatic region of the HMQC spectrum of YD.

HMBC

Although HMBC is a heteronuclear long range technique yielding single cross peaks

for

2

J

CH

,

3

J

CH

and

4

J

CH

couplings, occasionally

1

J

CH

couplings are also visible in the

spectra as two symmetric peaks positioned on a line parallel to the

1

H frequency

axis and centered at the crossing point of the

13

C and

1

H frequency. In this case we

can see the carbons and the directly attached protons.

Results and Discussion

45

F

IGURE

4.14: HMBC spectrum of YD in DMSO at room temperature.

The HMBC spectrum shown in Fig. 4.14 confirms the findings of the

1

H-NMR and

COSY techniques, since the carbon of the methyl group correlates with hydrogen

atoms 3 and 4. Certainly, the correlation with 3 is stronger than with 4 due to the

shorter distance. The one-bond correlation between the methyl carbon and

hydrogen atoms can also be observed. Examination of the correlation of methyl

hydrogen atoms and aromatic carbons usually provides valuable information, but

in this case the peaks overlap and therefore no additional information can be

obtained. Correlations between the aromatic carbon and hydrogen atoms appear

in the crowded part of the spectrum. A higher resolution plot of that region is

shown in Fig. 4.15.

Results and Discussion

46

F

IGURE

4.15: Aromatic region of the HMBC spectrum of YD at room temperature.

Results and Discussion

47

The first challenge of the spectrum assignment is the selection of a suitable point

of reference. In this instance, carbon 7 was the best possible choice because of its

expected strong correlation with hydrogen atoms 10-11 and a weaker one with

hydrogen atoms 12-13. Moreover, a very weak correlation with hydrogen 4 and

almost no correlation with hydrogen 3 are expected.

The positions of the hydrogen atoms are known from the COSY and

1

H spectra.

The only carbon peak in the HMBC spectrum showing correlations with all of the

hydrogen atoms designated 10-11, 12-13, 3 and 4 is the one at 168.97 ppm.

Therefore it can be assigned to carbon 7 (168.97 ppm). By using the HMQC data

carbons 10-11, 12-13, 3, and 4 were already assigned. (carbons 10-11 at 128.83 ppm

and hydrogen atoms 10-11 at 8.15 ppm; carbons 12-13 at 116.79 ppm and

hydrogen atoms 12-13 at 7.65 ppm; carbon 3 at 130.53 ppm and hydrogen 3 at 7.30

ppm; carbon 4 at 122.97 ppm and hydrogen 4 at 7.85 ppm).

The next step is the identification of carbons 9 and 14. As expected, the strong

correlations with hydrogen pairs 10-11 and 12-13 were the only ones detectable,

while the available estimation data suggest that 14 appears in the higher ppm

region of the spectrum (carbon 9 at ~129 ppm; carbon 14 at ~144 ppm).

Considering the first ring which contains the methyl group, carbons 3 and 4 are

already assigned. Besides these, there are four (2, 5, 6, 8) other carbon atoms that

should display a correlation with hydrogen atoms 3 and 4. Three of these four

peaks appear at chemical shifts values close to each other, namely at 132.47, 133.43

and 140.39 ppm, which refer to carbons 8, 2 and 6 as these atoms all have their

estimated values between 131 and 134 ppm. Moreover the peak with appears at

152.72 ppm, based on the estimation, is assigned to 5. As a consequence of similar

chemical environments of 2, 6 and 8, however, the available information is not

sufficient to assign the peaks to the distinctive carbons.

The last phase of the spectrum assignment takes the last ring of the molecule

containing exclusively quaternary carbon atoms into consideration. The single

hydrogen in the OH group is not informative due to the exchange with the solvent.

Thus, no correlation in the HMBC spectrum can be seen, which obstructs the

selection of a reference point.

Results and Discussion

48

Quaternary carbons, because of their extremely long relaxation times, appear with

low intensity. Using the estimated data the quaternary peaks can be assigned as

follows: carbon 19 at 118.69 ppm; carbon 16 at 170.16 ppm; carbon 18 at 162.03

ppm; carbon 15 at 122.60 ppm; carbon 17 at 163.84.

The summary of the peak assignment based on the different methods is listed in

Tables 4.1 and 4.2.

T

ABLE

4.1

1

H NMR chemical shifts of YD

Chemical shift (ppm)

Designation

of H atoms measured estimated

1 2.70

2.64

3 7.30 7.61

4 7.85 7.89

10, 11 8.15 7.76

12, 13 7.65 6.69

NH 15.70 13.24

OH 10.90 12.57

Results and Discussion

49

T

ABLE

4.2

13

C NMR chemical shifts of YD (unclear assignments are highlighted in italic).

Chemical shift (ppm)

Designation

of C atoms measured estimated

1 20.74 21.6

2, 6 or 8 133.43 133.0

3 130.53 127.6

4 122.97 126.5

5 152.72 152.4

6, 2 or 8 140.39 133.4

7 168.97 166.5

8, 2 or 6 132.47 131.8

9 129.88 121.0

10, 11 128.83 129.6

12, 13 116.79 116.8

14 144.33 143.0

15 122.60 154.0

16 170.16 163.0

17 163.84 193.0

18 162.03 163.0

19 118.69 118.0

The concentration and the temperature dependence of the chemical shifts of the

1

H NMR peaks were investigated for dye concentrations between 0.7 and 130 mM

at various temperatures from 10° C to 70° C. No change of chemical shift is

observed, meaning that there are no aggregates in this organic solvent. The

recorded spectra belong to the monomeric state of the dye.

Results and Discussion

50

4.2.2

1

H

C

HEMICAL SHIFT

A

NALYSIS

In order to investigate the aggregates the organic solvent DMSO was substituted

by deuterated water, D

2

O. This change in conditions leads to the formation of

aggregates as demonstrated by optical techniques before. Using the peak

assignments in DMSO the aggregation process observed in aqueous solution can

be more thoroughly understood.

F

IGURE

4.16:

1

H NMR spectrum of YD in 30 mM concentration D

2

O solution at 25°C

(recorded by 200 MHz spectrometer).

The

1

H NMR spectrum of YD in aqueous solution (Fig. 4.16) displays line

broadening because of the short transverse relaxation time (T

2

) of bigger molecules

produces broader lines

80

. Three distinct groups of peaks appear in the spectrum.

The unresolved four doublets from 6.7 to 7.7 ppm correspond to six aromatic

hydrogen atoms; the large peak at 5.8 ppm is the solvent peak, while the remaining

one at 2.46 ppm arises from the methyl group.

Results and Discussion

51

The previously used concentration, which was used for the microscopy

measurements, is too high to be able to provide additional information about the

aggregation process by using NMR. At high concentration of the dye the spectrum

is unresolved (due to the broad lines from the big aggregates) and the peaks have

low intensity (due to precipitation). Therefore a dilute solution of YD (0.7 mM

concentration) was investigated, which is in the concentration range used for the

UV/VIS measurements.

A comparison of the line widths of the spectra measured in DMSO as solvent and

in the most dilute aqueous solution measured in NMR displays considerable

agreement. (Fig. 4.17)

F

IGURE

4.17:

1

H NMR spectra of YD in DMSO and dilute D

2

O solution at room

temperature.

At the high concentration of 30 mM the resulting spectrum demonstrates

significantly broader and shifted peaks. Therefore a series of measurements (0.7 –

Results and Discussion

52

30 mM) was carried out to verify the suspected dependence on the dye

concentration [YD].

The aromatic regions of several selected spectra measured in D

2

O at different dye

concentrations are presented in Fig. 4.18. For comparison the spectrum in DMSO

solution, where no aggregation occurs, is shown on top. The concentration of the

dye in D

2

O was increased between 0.7 mM and 15 mM.

F

IGURE

4.18: Aromatic regions of

1

H NMR spectra of YD in DMSO and D

2

O solution.

As can be seen in Fig. 4.18 increasing the dye concentration leads to an upfield shift

of all peaks. Fig. 4.19 shows this effect in more detail and the related chemical shift

values are presented in Table 4.3. The change in chemical shifts is larger at low

Results and Discussion

53

concentrations; while it levels off at higher dye concentrations. The chemical shift

value of the methyl group displays a smaller variation.

0 5 10 15 20 25 30

2,5

3,0

6,5

7,0

7,5

8,0

Hydrogen 1

Hydrogen 3

Hydrogen12,13

Hydrogen 4

Hydrogen 10, 11

Chemical shift (ppm)

Concentration of YD (mM)

F

IGURE

4.19: Proton NMR chemical shifts as a function of [YD] in aqueous solution at room

temperature.

T

ABLE

4.3

Concentration dependence of

1

H NMR chemical shift of YD at room temperature

1

H NMR chemical shifts (ppm)

[YD] (mM) 1 3 12, 13 4 10, 11

0.0 2.7 7.35 7.61 7.86 8.13

0.7 2.64 7.31 7.39 7.76 7.81

2.5 2.64 7.27 7.28 7.72 7.74

5.0 2.6 7.17 7.17 7.63 7.63

10.0 2.55 7.12 7.05 7.55 7.51

15.0 2.54 7.08 6.98 7.5 7.44

20.0 2.45 6.98 6.85 7.39 7.31

25.0 2.44 6.95 6.8 7.36 7.27

30.0 2.46 6.96 6.8 7.36 7.26

Results and Discussion

54

The concentration dependence of

1

H chemical shift values indicates a stacking of

the aromatic rings and formation of aggregates. With the increase of the

concentration the peaks are subjected to an upfield shift, which is more

pronounced for atoms located in the central regions of the molecule, namely

hydrogen pairs 10-11 and 12-13. This shift occurs as a result of the shielding

property of π-electrons, since the molecules are stacked upon each other. The

methyl group, which is located in the outer part of the molecule, is less affected by

π-shielding. These results correspond with the UV/VIS data very well, where a shear

angle of 54.7-90° was found.

Another aspect of the aggregation process worth examining is its temperature

dependence. The most dilute solution was investigated at three different

temperatures, and the results were compared with those of the DMSO solution.

Fig. 4.20 shows the temperature dependent spectra in both D

2

O and DMSO.

Substantial changes are observed in the spectra at the lowest measured dye

concentration in aqueous solution as the temperature increases, while almost no

temperature dependence is found with DMSO as solvent. The small high field shift

is within experimental error. No internal standard was used in case of DMSO

solutions and the solvent peak was used as reference, which can be shifted by

changing the temperature. These observations suggest that DMSO solutions

contain individual molecules, but in D

2

O there is already some aggregation even at

lower concentrations. With increasing temperature the aggregates dissolve or

become smaller, which leads to the observed changes of the chemical shifts.

Results and Discussion

55

F

IGURE

4.20: Aromatic regions of

1

H NMR spectra demonstrating the temperature

dependence of chemical shifts in YD solutions at the measured lowest concentration in

organic and inorganic solvents.

Results and Discussion

56

F

IGURE

4.21: Aromatic regions of

1

H NMR spectra demonstrating the concentration

dependence of chemical shifts of YD in aqueous solution.

Fig. 4.21 shows the temperature dependence for the aromatic regions of

1

H NMR

spectra in aqueous solutions with different YD concentrations of 2.5 mM (a) and

30 mM (b). For each concentration spectra recorded at room temperature and at

70° C, which was the upper temperature limit of the instrument, are shown. In case

of the more dilute solution elevating the temperature results in a spectrum

substantially down-field shifted, suggesting that there is less aggregation at higher

temperature. Under identical conditions the more concentrated solution shows a

similar phenomenon, but to a much less degree. Theoretically, given the possibility

of elevating the temperature above the inherent limit of the system, a spectrum

most similar to the one describing the monomer state might be obtained (aside

from solvent effects).

Results and Discussion

57

4.2.3 S

IGNAL INTENSITIES

In case of YD aggregates there are no additional peaks of the aggregates in the

1

H

NMR spectra (Fig. 4.16) in comparison with the monomer state spectra in DMSO,

indicating that there is a fast chemical exchange between the aggregates and the

individual dye molecules. The observed spectra are time-averaged spectra over the

aggregates and the monomers. Fast exchange was confirmed by diffusion NMR

measurements (Chapter 4.2.3.). The signals in the high resolution liquid NMR

spectrum arise from both dissolved molecules and aggregates. Based on the

formation of dye aggregates the maximum signal intensity can be decreased, when

the aggregates are too big and/or rigid to be detected by classical liquid NMR

methods and their signals are not contributing to the measured NMR spectrum.

Even aggregates in solution, which have not precipitated may not be observed.

A quantitative analysis of dissolved species was performed by measuring the signal

intensity of the methyl peak at various concentrations and temperatures. The

results are presented in Fig. 4.22.

At low temperatures the measured peak intensities are always lower than at high

temperatures. For each concentration the intensity increases with increasing

temperature and a constant plateau is reached. The plateau values are reached at

lower temperature for the lower concentrations. This indicates that signal intensity

is lost at low temperature due to big aggregates, which cannot be detected by the

solution state NMR method used here. For example in case of 30 mM sample at

10°C the maximum signal intensity is about 40 and at 70°C it is about 72, thus the

loss of signal can be around (72-40)/72= 44 %.

Results and Discussion

58

10 20 30 40 50 60 70

0

10

20

30

40

50

60

70

80

Maximum Signal Intensity

(a.u.)

T (

o

C)

5 mM

15 mM

20 mM

30 mM

F

IGURE

4.22: Maximum signal intensity of the methyl proton as a function of temperature

in D

2

O solutions of different concentrations.

Results and Discussion

59

4.2.4 C

HEMICAL

E

XCHANGE AND

L

INE WIDTH

The line width of the peaks, which is defined as the full-with at half height of a

resonance line, depends on the T

2

relaxation time of the molecules. There are two

other effects for additional broadening, namely chemical exchange between the

individual dye molecules and the aggregates and the broadening due to the

magnetic field inhomogeneity.

Under the experimental conditions with the 500 MHz spectrometer these two

additional effects are negligibly small due to the fast chemical exchange and the

high resolution of the NMR instrument. For the measurements of the 200 MHz

spectrometer those effects were taken account.

The very small and the very big molecules have extremely long T

1

relaxation time,

but the T

2

relaxation time, which is relevant for the line width

80

and remains low

for big molecules results in broad peaks in the spectra. Even midsize molecules,

such as the NMR detectable YD aggregates, have short T

2

relaxation time which

produces broad peaks.

Similar to the signal intensity analysis the methyl peak line width was analyzed for

each case. The overlapping signals of aromatic protons cannot be measured with

high accuracy so only the methyl peak was analyzed. The linewidth of the methyl

signal as a function of temperature in aqueous solution at various dye

concentrations is shown in Fig. 4.23. The same data are presented in Fig. 4.24

showing linewidth versus dye concentration at various temperatures.

At higher concentrations and at lower temperatures the lines get broader as a

result of bigger aggregates which have shorter relaxation times and produce

broader peaks in the NMR spectra. The discontinuity of the 20 mM sample can be

a measurement error, such as shimming.

Results and Discussion

60

10 20 30 40 50 60 70

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

Linewidth of Methyl peak (ppm)

T (oC)

0.7 mM

5 mM

10 mM

15 mM

20 mM

25 mM

30 mM

20 mM in DMSO

F

IGURE

4.23: Linewidth of the methyl protons as a function of temperature in aqueous

solutions at different dye concentrations in DMSO at a concentration of 20 mM.

Results and Discussion

61

0 5 10 15 20 25 30 35

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.10

Linewidth of Methyl peak (ppm)

[YD] (mM)

10

o

C

15

o

C

20

o

C

25

o

C

30

o

C

35

o

C

40

o

C

45

o

C

50

o

C

55

o

C

60

o

C

65

o

C

70

o

C

F

IGURE

4.24: Linewidth of the methyl peaks as a function of YD concentration at different

temperature in aqueous solution.

Based on equation [2.4] the measured line widths are converted to T

2

relaxation

times, which are shown in Fig. 4.25.

1/T

2

is approximately proportional to the overall rotational correlation time (τ

c

),

which depends on the mass and shape of the molecule. τ

c

is approximately the

average time for the molecule to rotate by one radian.

In order to obtain τ

c

from the measured line widths a calculated relationship from

the literature was used.

80

This relationship is shown in Fig. 4.26.

The principal uncertainties in the calculation are due to the following factors:

anisotropic rotational diffusion of nonspherical molecules, differential contribution

from internal motions, cross correlation effect,

1

H dipolar interaction with nearby

protons, chemical shift anisotropy. In light of these uncertainties, the result,

presented in Fig. 4.26 should be regarded as an approximate guideline.

80

For

Results and Discussion

62

example

1

H measured linewidths in ubiquitin are 6-9 Hz. This value is consistent

with 5 Hz in Fig. 4.26.

0 5 10 15 20 25 30

16

18

20

22

24

26

28

30

32

34

36

38

40

42

44

46

48

[YD] (mM)

T

2

(ms)

10 °C

15 °C

20 °C

25 °C

30 °C

35 °C

40 °C

45 °C

50 °C

60 °C

65 °C

70 °C

F

IGURE

4.25: Relaxation times as a function of YD concentration at different temperature.

According to Fig. 4.26 the correlation times were estimated from the linewidths,

the resulting values shown in Fig. 4.27.

Results and Discussion

63

2 4 6 8 10 12 14 16

2

4

6

8

10

12

14

16

18

20

Y = A + BX

Parameter Value

-------------------------

A 0,6

B 1,2E9

∆ν (Hz)

τ

c

(ns)

F

IGURE

4.26: Resonance linewidths of

1

H spins are shown as a function of rotational

correlation time. Adapted from the literature

80

where the linewidth was calculated as a

function of the rotational correlation time τ

c

for various types of nuclei. The line shows

the data for the case of

1

H spins. The line is represented by ∆υ=0.6+1.2E9 τ

c

.

Results and Discussion

64

0 5 10 15 20 25 30

4

5

6

7

8

9

10

11

12

13

14

15

10 °C

15 °C

20 °C

25 °C

30 °C

35 °C

40 °C

45 °C

50 °C

55 °C

60 °C

65 °C

70 °C

τ

c

(ns)

[YD] (mM)

F

IGURE

4.27: Estimated rotational correlation times are shown as a function of YD

concentration at various temperatures.

The correlation time varies with molecular size, solvent viscosity and temperature.

The simplest theoretical approach for approximately spherical globular particles

calculates the isotropic rotational correlation time from Stokes law. Using equation

[2.6] and the values of the estimated

τ

c

, R

h

was calculated. The results are shown in

Fig. 4.28.

Results and Discussion

65

0 5 10 15 20 25 30

1,65

1,70

1,75

1,80

1,85

1,90

1,95

2,00

2,05

2,10

2,15

2,20

2,25

2,30

2,35

2,40 10 °C

15 °C

20 °C

25 °C

30 °C

35 °C

40 °C

45 °C

50 °C

55 °C

60 °C

65 °C

70 °C

R

h

(nm)

[YD] (mM)

F

IGURE

4.28: Estimated hydrodynamic radii are shown as a function of YD concentration at

various temperatures.

The measured linewidth used in the previous analysis can be much bigger than the

homogeneous linewidth especially in case of sharp peaks (small molecules).

Shimming of the 200 MHz magnet used for the diffusion NMR experiments was

not perfect and the magnet field inhomogeneity cannot be neglected. To decrease

this problem a correction of the linewidth was carried out.

The correction of the linewidth was calculated from the methyl peak of YD in

DMSO. Based on computer simulations one half of the molecular length is 0.7 nm.

It was taken as R

h

of the monomer. The rotational correlation time was calculated

according to equation [2.6] and the homogeneous linewidth corresponding to

τ

c

was obtained form Fig. 4.26. Then all of the measured linewidth values were

corrected by the difference of the measured and the calculated value of the methyl

peak width in DMSO. The measured linewidth of the methyl protons was 6.9 Hz

Results and Discussion

66

and the calculated one was 1.345 Hz. Thus the broadening of the peak due to the

magnetic field inhomogeneity and other effects, which was discussed before, is

5.555 Hz.

All of the corrected linewidths are shown in Fig. 4.29.

0 5 10 15 20 25 30

0

1

2

3

4

5

6

7

8

9

10

11

12

13

10 °C

15 °C

20 °C

25 °C

30 °C

35 °C

40 °C

45 °C

50 °C

55 °C

60 °C

65 °C

70 °C

∆ν

(Hz)

[YD] (mM)

F

IGURE

4.29: Estimated proton resonance homogeneous linewidths are shown as a

function of YD concentration at various temperatures.

From the estimated homogeneous linewidth the rotational correlation time (Fig.

4.30) was calculated, which then was used for the R

h

determination (Fig. 4.31).

Results and Discussion

67

0 5 10 15 20 25 30

0

1

2

3

4

5

6

7

8

9

10

11

10 °C

15 °C

20 °C

25 °C

30 °C

35 °C

40 °C

45 °C

50 °C

55 °C

60 °C

65 °C

70 °C

τ

c

(ns)

[YD] (mM)

F

IGURE

4.30: Corrected rotational correlation times are shown as a function of YD

concentration at various temperature.

Results and Discussion

68

0 5 10 15 20 25 30

0,6

0,7

0,8

0,9

1,0

1,1

1,2

1,3

1,4

1,5

1,6

1,7

1,8

1,9

10 °C

15 °C

20 °C

25 °C

30 °C

35 °C

40 °C

45 °C

50 °C

55 °C

60 °C

65 °C

70 °C

R

h

(nm)

[YD] (mM)

F

IGURE

4.31: Corrected estimated hydrodynamic radii are shown as a function of YD

concentration at various temperatures.

The radius of the particles increases with increasing concentration and decreasing

temperature up to 1.9 nm. There is some scattering of the data due to the

estimation error and the accuracy of the magnet shimming.

Results and Discussion

69

4.2.5 NMR

DIFFUSOMETRY

NMR diffusometry was used to determine the diffusion coefficients of the dye

molecules/aggregates at various concentrations (0.7–30 mM) and temperatures

(10–70 °C). The same samples were used for the high resolution and the diffusion

NMR measurements.

Determination of the diffusion coefficient

The data analysis of the PGSTE diffusion experiments is presented in the following.

Fig. 4.31a shows an example of the diffusion NMR spectra of YD. The strength of

the pulsed-field gradient along the z axis was varied between 0 and 650 G/cm. At

the highest gradient almost no signal was detected.

The protons of the water disappear at lower gradient strength than the aromatic-

and the methyl protons of the dye. In this sample the dye forms aggregates and the

observed diffusion of the dye molecules is much slower than that of the solvent

molecules.

Results and Discussion

70

ppm

0.0

1.0

2.0

3.0

4.0

5.0

Aromatic protons

Water peak

Methly protons

F

IGURE

4.31a:

1

H PGSTE diffusion experiment of YD. ([YD]= 30 mM, T= 10 °C, Δ= 100 ms,

δ= 500 μs, g

max

= 650 G/cm, n

spectrum

=64, ns= 128) 64 spectra were recorded, but only every

fifth one is shown (starting with number 5 at the bottom and ending with number 50 at

the top) and the last 3 spectra were mainly noise and they were neglected. The signal

attenuation of methyl- and aromatic protons is very similar while the water proton

intensity decreases faster.

The integral of each peak versus the gradient strength is shown in Fig. 4.32. The

normalization was necessary to get a better comparison.

The intensity of the overlapping aromatic protons was measured together. The

integration was performed using the same ppm intervals at each case. For

comparison the attenuation of the methyl peak was also analyzed.

50

45

40

35

30

25

20

15

10

5

Results and Discussion

71

0 100 200 300 400 500 600 700

0.0

0.2

0.4

0.6

0.8

1.0

Signal intensity (norm.)

Aromatic protons

Water peak

Methyl peak

g (G/cm)

F

IGURE

4.32: Normalized signal intensity from the diffusion experiment shown in Fig. 4.31

as a function of gradient strength.

Plotting the logarithm of the signal intensity versus kg

2

(with k=∆δ

2

γ

2

) results in

straight lines, which is shown in Fig. 4.33. The linearity of the attenuation of the

signal shows that there is only one diffusing component based on the before

mentioned fast chemical exchange. The slope of each line is related to the diffusion

coefficients of the component. The diffusion coefficient values of the methyl signal

and the aromatic proton signal are in a good agreement. The calculated

measurement error is around 2 %.

The linear form of the resonance signal decays are presented in Fig. 4.33.

Results and Discussion

72

0,0 0,5 1,0 1,5

0,01

0,10

1,00

Aromatic protons D= 6,9e-11 m

2

/s

Water peak D= 1,2e-9 m

2

/s

Methyl peak D= 7,1e-11 m

2

/s

ln(Signal Intensity)

g

2

γ

2

δ

2

∆

slop = -D

x10

8

F

IGURE

4.33: Normalized signal intensity from the diffusion experiment shown in Fig. 4.31

as a function of kg

2

.

Results and Discussion

73

The Stokes-Einstein equation: from D to R

h

The Stokes-Einstein equation shows the relationship between the diffusion

coefficient and the structural properties of the diffusing particles, such as the

hydrodynamic radius (R

h

) (equation [2.2] and [2.3]). Based on this equation the

measured diffusion coefficients were converted to the hydrodynamic radius. The

calculated dimensions of the above mentioned sample are collected in Table 4.4

T

ABLE

4.4

Estimated molecular dimension of YD

Derived from

Aromatic protons

Derived from

Methyl protons

Derived from

Water protons

D

(m

2

/s)

R

h

(nm) D

(m

2

/s)

R

h

(nm) D

(m

2

/s)

R

h

(nm)

6.9e-11 1.95 7.1e-11 1.9 1.2e-9 0.1

The length of YD molecule, based on the simulation data discussed later, is around

1.44 nm. The value of R

h

, shown in Table 4.4, is a complicated average based on the

measured diffusion coefficient, which is weight-averaged over monomers and

aggregates. Using a simplified model of monodisperse compact spheres, for which

the hydrodynamic radius is the same as the radius of the sphere and assuming a

density of 1 g/cm

3

, R

h

= 1.9 nm can be converted to an aggregation number of 37.

By using another model of loose spheres, the estimated aggregation number is 22.

Both models can be considered as limiting cases, thus the real aggregation number

is between 22 and 37.

The measured value of the water diffusion coefficient is in good agreement with

the literature data

103

(at 25°C D

water

= 1.872x10

-9

m

2

/s and at 5°C D

water

= 1.015x10

-9

m

2

/s). In case of dye solution the diffusion of the water is a bit slower based on the

obstruction.

It is important to recall that the Stokes-Einstein equation is valid for spherical

particles of colloidal dimension (much larger size than the solvent molecules),

which move with uniform velocity in a fluid continuum.

Results and Discussion

74

In case of dye aggregation only the average dimension of monomer and aggregates

can be estimated because there is no individual peak of the monomer and the

aggregates based on the fast chemical exchange. Moreover the aggregates might

have ellipsoidal, elongated or wormlike shape.

Using the above discussed steps of the diffusion NMR data analysis the average

hydrodynamic radius of the aggregating dye system was estimated at various

concentrations and temperatures, under conditions similar to those of the

1

H

chemical shift measurements (Chapter 4.2.2). Before discussing the results, some

difficulties of the diffusion measurements and the data analysis are described, such

as convection artifacts and fitting problems of the decay curve.

Temperature gradient and convection

Convection within a sample of low viscosity is a serious problem affecting diffusion

measurements especially at higher temperature. Convection currents are caused by

small temperature gradients in the sample. It causes additional signal attenuation

leading to erroneous diffusion coefficients. This artifact can be compensated by

using the double stimulated echo experiment, described in Chapter 2.2.3.

A comparison of the results obtained with the double stimulated (DSTE) echo and

the simple stimulated echo (STE) pulse sequence at various temperatures (T= 10,

30 and 50 °C) is presented in Fig. 4.34, Fig. 4.35 and Fig. 4.36. At each temperature

the DSTE and STE measurement conditions (diffusion time and gradient) were

exactly the same.

Results and Discussion

75

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5

0,01

0,10

1,00

DSTE D=6,92 m

2

/s

STE D=6,93 m

2

/s

ln(Signal Intensity-norm.)

g

2

(a.u.)

x10

4

F

IGURE

4.34: Comparison of double stimulated echo (DSTE) and stimulated echo (STE)

pulse sequence at T=10°C in case of YD in aqueous solution. ([YD]= 30 mM, T= 10 °C, Δ=

100 ms, δ= 500 μs, g

max

= 568 G/cm, n

spectrum

= 64, ns= 128).

At low temperatures the DSTE and STE signal decays are the same (Fig. 4.34),

indicating that no convection occurs. A higher accuracy of the measurement can

be achieved by increasing the number of scans or by decreasing diffusion time

and/or gradient strength.

Results and Discussion

76

0,0 0,5 1,0 1,5 2,0 2,5 3,0 3,5

0,01

0,10

1,00

ln(Signal Intensity-norm.)

g

2

(a.u.)

DSTE D=1,58e-10 m

2

/s

STE D=1,65e-10 m

2

/s

x10

4

F

IGURE

4.35: Comparison of double stimulated echo (DSTE) and stimulated echo (STE)

pulse sequence at T= 30°C in case of YD in aqueous solution. ([YD]= 30 mM, T= 30 °C,

Δ=100 ms, δ= 500 μs, g

max

= 568 G/cm, n

spectrum

=64 (not all shown), ns= 128).

At somewhat elevated temperature (30 °C) DSTE and STE experiments are still

comparable. The signal decay in case of STE is only slightly faster compared to the

DSTE experiment. Close to room temperature no convection occurs.

Results and Discussion

77

0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4

0,01

0,10

1,00

ln (Signal Intensity-norm.)

g

2

(a.u.)

DSTE D= 3,03e-10 m

2

/s

STE D= 7,28e-9 m

2

/s

x10

4

F

IGURE

4.36: Comparison of double stimulated echo (DSTE) and stimulated echo (STE)

pulse sequence at T= 50 °C in case of YD in aqueous solution. ([YD]= 30 mM T= 50 °C

Δ= 100 ms δ= 500 μs g

max

= 568 G/cm n

spectrum

=64 ns= 128). In case of STE the attenuation

of the signal is very fast because convection occurs.

The convection artifact at low temperature (up to 30 °C) is negligible, but at higher

temperature (Fig. 4.36) convection occurs in the sample, leading to false values for

D. Therefore the double stimulated echo pulse sequence must be used in the

diffusion NMR experiments.

On the basis of these results the diffusion NMR experiments were carried out at

low temperature (up to 30 °C) with STE pulse sequence and at higher temperature

DSTE was applied.

In principle the DSTE pulse sequence can be used at low temperature as well but it

is not recommended because of the bigger signal loss due to the longer sequence.

Results and Discussion

78

Reproducibility and fitting problem of the diffusion experiment

In this part the reproducibility of the diffusion measurement is investigated and

guidelines for the optimum choice of experimental parameters, such as diffusion

time and gradient strength are given.

The first experiment of verification was carried out at T= 60 °C using the DSTE

pulse sequence which was repeated twice using exactly the same settings on the

same day and the same sample. The data analysis was carried out using exactly the

same frequency range for integration. The good reproducibility of the diffusion

NMR experiment can be recognized in Fig. 4.37.

0,0 0,2 0,4 0,6 0,8 1,0 1,2 1,4

0,10

1,00

ln(Signal Intensity-norm.)

g

2

(a.u.)

DSTE 1 D= 4.27e-10 m

2

/s

DSTE 2 D= 4.29e-10 m

2

/s

DSTE 3 D= 4.39e-10 m

2

/s

F

IGURE

4.37: Reproducibility of the diffusion NMR experiment [YD]=15 mM, T= 60 °C, Δ=

20 ms, δ= 500 μs, g

max

= 3,5 T/m).

Results and Discussion

79

The diffusion measurement is a fairly time-consuming experiment. Therefore it is

not possible to use high numbers of scans, diffusion times, and gradient strengths.

So it is very important to find the optimal parameters of the diffusion NMR

experiment. For example when too long diffusion time and strong gradient are

used there are only few accurate points within the entire curve and a lot of points,

with almost zero intensity, leading to big errors. On the other hand, with short

diffusion times and weak gradients, all measured points are in the first part of the

decay curve and deviations from Gaussian diffusion may not be detected. Fig. 4.38

shows the calculated diffusion coefficient values versus the number of points fitted

to determine the required number of points.

The whole curve contains 64 points. The first diffusion coefficient of the diagram

comes from the fitting of the first 3 points and the rest was neglected. The fitting

was done from 3 up to 64 points of the decay curve and the diffusion coefficients

were calculated for each case.

The calculated diffusion coefficient is not accurate enough if only the first few

decay points are considered. At least 30-35 points are required, which corresponds

to the full range of data shown in Fig. 4.35. Of course it also plays a role how the

given numbers of points is distributed over the whole decay curve, which is not

further analyzed here.

Results and Discussion

80

0 10 20 30 40 50 60 70

5.8

6.0

6.2

6.4

6.6

6.8

7.0

D (m2/s) x10e-11

Number of points

F

IGURE

4.38: The calculated diffusion coefficient versus the number of point in the decay

curve of Fig. 4.34 ([YD]= 30 mM T= 10 °C Δ= 100 ms δ= 500 μs g

max

= 650 G/cm).

Dye Diffusion

The diffusion coefficients were calculated at various solutions, dye concentrations

and temperatures and the dimensions of the aggregates were estimated by using

the Stokes-Einstein equation.

A simple stimulated echo pulse sequence was used up to 30 °C to obtain a better

signal to noise ratio while at higher temperatures the double stimulated echo

sequence was used to compensate the convection in the sample.

Because of the fast chemical exchange between the monomer and the aggregates

every signal attenuation curve shows monoexponential decay independently of the

solvent, temperature or dye concentration. The average diffusion coefficient of the

system contains the diffusion of the fast/small monomer molecules and the

Results and Discussion

81

slow/big aggregates which are still in solution and detectable by NMR

spectroscopy.

Diffusion of dye in DMSO

As shown in Chapter 4.2.1 and 4.2.2 the

1

H NMR chemical shifts and line widths of

YD in DMSO solution are independent of the concentration and the temperature.

To get information about the dimensional properties under the same

concentrations and temperatures diffusion NMR measurement were carried out.

In Fig. 4.39 is shown the diffusion coefficient of YD in DMSO versus the

temperature. The dye diffusion gets faster with increasing temperature because

the viscosity of the solvent decreases.

30 40 50 60 70

1.50

2.00

2.50

3.00

3.50

4.00

D (m

2

/s) x10e-10

T (

o

C)

F

IGURE

4.39: Average diffusion coefficient of YD in DMSO as a function of temperature.

([YD]= 20 mM Δ= 20-60 ms δ= 500 μs g

max

= 150- 650 G/cm).

Results and Discussion

82

Fig. 4.40 shows the size of YD particles estimated on the basis of the Stokes-Einstein

equation as a function of temperature in DMSO. The temperature independence

of YD dimension is well observable.

The estimated hydrodynamic radii of YD solution and the one calculated by DFT

simulations, presented in section 4.3, are in very good agreement. Based on DFT

simulation the molecular length of a planar YD molecule is about 1.4 nm and the

estimated hydrodynamic radius is about one half of it, 0.7 nm.

The diffusion measurements clearly show that in DMSO there are only individual

YD molecules and no aggregation occurs. This corresponds in an excellent way to

the result of the chemical shift and linewidth analysis of the

1

H NMR

experiments.

30 40 50 60 70

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

R

h

(nm)

T (

o

C)

F

IGURE

4.40: Estimated hydrodynamic radius of YD as a function of temperature.

Results and Discussion

83

Dye diffusion in aqueous solution

YD forms aggregates in aqueous solution as mentioned in Chapter 4.2.2. In D

2

O

diffusion measurements were carried out under various conditions, such as various

diffusion times and gradient strengths. The measured diffusion coefficient is

independent of the diffusion time, there is free diffusion. The results of the

diffusion coefficient at various diffusion times and for repeated experiments were

averaged and used for later analysis.

10 20 30 40 50 60 70

0

1

2

3

4

5

6

7

D (m

2

/s) x10e-10

T (

o

C)

0,7 mM

5 mM

10 mM

15 mM

20 mM

25 mM

30 mM

Figure 4.41: Diffusion coefficient of YD versus temperature at various dye-concentrations

in aqueous solution ([YD]= 0.7 – 30 mM).

Fig. 4.41 shows the diffusion coefficient versus temperature at various dye

concentrations (0.7 mM to 30 mM). The diffusion of the system becomes faster

upon increasing temperature and decreasing concentration of the dye.

Results and Discussion

84

There is equilibrium between the monomers and the aggregates. At higher dye

concentration the possibility of the aggregate formation is higher due to the large

number of accidental collisions. Based on the equilibrium n[M] ↔ [M]

n

increasing

the dye concentration an exponential growth of the aggregate dimension can

occur. The temperature-dependent diffusion measurements show that at higher

temperature the equilibrium is shifted to the formation of the monomer.

10 20 30 40 50 60 70

0.50

0.75

1.00

1.25

1.50

1.75

2.00

2.25

R

h

(nm)

T (

o

C)

0,7 mM

5 mM

10 mM

15 mM

20 mM

25 mM

30 mM

F

IGURE

4.42: Estimated hydrodynamic radii of YD versus temperature at various dye-

concentrations in aqueous solution in case of sphere shape aggregate. ([YD]= 0.7 – 30

mM).

The estimated hydrodynamic radius as a function of temperature is shown in Fig.

4.42. At low dye concentration and high temperature the estimated hydrodynamic

radii are in the range of 0.75 nm, which is comparable with the monomer

dimension. Under these conditions the individual dye molecules do not show an

inclination to form aggregates. At higher dye concentration and lower temperature

Results and Discussion

85

the estimated average radius of the dye system increases up to 2 nm. The reason of

this small average size is that the really big aggregates do not contribute to the

NMR signal.

In order to understand how many units constitute the aggregate it is important to

know the hydrodynamic volume of the single dye molecule (in case of a sphere:

V

h0

=1.437 nm

3

). This was calculated from the diffusion data in DMSO solution

where no aggregation was observed (R

h

= 0.7 nm)

The aggregation number (N) is a parameter which quantifies the degree of

aggregation and in case of non compact spheres can be estimated from

o

h

V

N= [4.1]

N represents the average number of monomers constituting the aggregate.

In Fig. 4.43a the aggregation number is plotted as a function of temperature at

various dye concentrations. The number of the monomers in the aggregate

depends on the dye concentration and the temperature. At low concentrations

and high temperatures only individual monomer exist in aqueous solution. With

increasing concentration and decreasing temperature the aggregation number

increases up to 22. This estimation is very rough because it assumes spherical

monomers and does not take into account the shape of the aggregates.

Results and Discussion

86

10 20 30 40 50 60 70

0

2

4

6

8

10

12

14

16

18

20

22

24

0,7 mM

5 mM

10 mM

15 mM

20 mM

25 mM

30 mM

Aggregation number (N)

T (oC)

F

IGURE

4.43a: Estimated aggregation number of YD versus temperature at various dye-

concentrations in aqueous solution in case of spherical aggregates. ([YD]= 0.7 – 30 mM).

In case of compact spheres the aggregation number can be calculated according to

equation [4.2]

molecule

A

3

h

molecule

3

h

M

N

R

3

4

V

R

3

4

Nϕ

π=

π

=

[4.2]

Based on equation [4.2] and the estimated R

h

values from the diffusion data the

aggregation number N as a function of temperature is shown in Fig. 4.43b.

Results and Discussion

87

10 20 30 40 50 60 70

0

5

10

15

20

25

30

35

40

0,7 mM

5 mM

10 mM

15 mM

20 mM

25 mM

30 mM

Aggregation number (N)

T (°C)

F

IGURE

4.43b: Estimated aggregation number of YD versus temperature at various dye-

concentrations in aqueous solution in case of compact spheres. ([YD]= 0.7 – 30 mM).

Results and Discussion

88

4.3 C

OMPUTER SIMULATION

In order to get more information about the dye system a high accuracy density

functional theory (DFT) method was used to obtain an optimized structure of the

yellow dye (YD) monomer in the gas phase and subsequently its NMR and UV

spectra were calculated.

4.3.1 DFT

STRUCTURE OPTIMIZATION OF THE MONOMER

The first input structure was drawn by GaussView. A planar structure was chosen

and converted to a NWChem input file. For the optimization of yellow dye (YD)

the 6-311g(p,d) basis set was used along with a DFT-based method with b3lyp

exchange correlation implemented into the computational chemistry software

package of NWChem.

F

IGURE

4.44: DFT optimized structure of YD (b3lyp/6-311g(d,p)).

An energy minimum was found by this very time consuming optimization,

corresponding to the structure shown in Fig. 4.44, which was used for the

simulations discussed later. The optimized yellow dye has a planar structure, where

the carbonyl oxygen atoms are in the enol form and the hydrogen atoms are

Results and Discussion

89

building strong H-bonds with the neighboring nitrogen atoms. To validate this

structure the NMR and UV spectra were computed.

4.3.2 DFT

SIMULATION OF

NMR

SPECTRA OF THE MONOMER

For the calculation of the NMR spectra of YD its DFT optimized structure was

applied as starting geometry using the same 6-311g(p,d) basis set of the Gaussian

03’ program package. After the structure optimization the isotropic shielding

values of the proton were calculated by using GIAO (Gauge-Including Atomic

Orbital) method. The calculated shielding values were extracted from the Gaussian

output, the value of shielding tensors of equivalent hydrogen atoms are averaged.

The isotropic shielding value of the hydrogen atom was then subtracted from the

value of the shielding value of TMS.

Table 4.5 shows the comparison of the calculated and measured

1

H chemical shifts

and Fig. 4.45 represents the calculated NMR spectra.

As shown in Table 4.5 the calculated and measured chemical shifts are in good

agreement, suggesting that the geometry obtained by the quantum-chemical

optimization is a possible structure and can be used for later calculations.

T

ABLE S

4.5

Measured and calculated

1

H NMR chemical shift of YD

H atom number

Measured

1

H

chemical shift

(ppm)

Calculated

1

H

chemical shift

(ppm)

Deviation

(pmm)

1 2.69 2.32 0.37

3 7.35 6.87 0.48

4 7.86 7.57 0.29

10,11 8.13 8.13 0.00

12,13 7.61 7.63 0.02

Results and Discussion

90

9 8 7 6 5 4 3 2

34

10

1

12

NMR chemical shift (ppm)

F

IGURE

4.45: Calculated

1

H NMR spectrum of YD in gas phase (b3lyp/6-311g(d,p)).

4.2.4 DFT

SIMULATION OF

UV

SPECTRA OF THE MONOMER

Another verification of our computational model is the calculation of the UV

spectra. The input structure was the optimized structure of YD obtained by using

the Gaussian program package. In Fig. 4.46 the simulated and measured UV spectra

are shown.

The measured and the calculated spectra are in very good agreement. However,

there is additional absorbance in the measured spectra (410 - 360 nm), which can

be an effect of the aggregate formation. One other reason can be additional

absorption at lower wavelength by the glass of the home made sample holders.

Results and Discussion

91

360 380 400 420 440 460 480 500 520 540

0.0

0.2

0.4

0.6

0.8

1.0

Absorbance

λ (nm)

DFT calculation

Measured spectra

F

IGURE

4.46: Measured and calculated UV spectrum of YD in gas phase (td b3lyp/6-

311g(d,p).

4.2.5 DFT

STRUCTURE OPTIMIZATION OF THE DIMER

The input for the dimer structure simulation was a stack of two monomers. The

central rings were placed exactly upon each other and one molecule was rotated

180° with respect to the other, such that the methyl groups and the SO

3-

groups

are most distant. There are several possibilities to put the molecules upon each

other but due to the big amount of computer time consumed by just this one

calculation, only one variation was optimized. The starting distance of the two

planar molecules was determined by energy calculation; a value of 2.5 Å was

obtained. Starting from this stack of two planar molecules, all constrains on

Results and Discussion

92

dihedral angles were relieved in the further optimization. The optimized structure

is shown in Fig. 4.47.

F

IGURE

4.47: DFT optimized structure of a stack of two YD (b3lyp/6-311g(d,p)).

The initially planar structure has changed. A twist of the molecular structure can

be observed and the dimer tends to form H bonds. This conformational change of

the structure may result from the deficiency of the chosen simulation method,

which is not very accurate for π-π interactions.

Results and Discussion

93

4.2.5 DFT

NMR

SPECTRUM SIMULATION OF THE DIMER

After the structure optimization the NMR spectra of the dimer was calculated by

the same method as before (b3lyp/6-311g(d,p)). The

1

H NMR chemical shifts are

shown in Table 4.6 and Fig. 4.48 shows a schematic representation of the calculated

1

H NMR spectrum.

T

ABLE

4.6

Structure optimation of the dimer

H atom number

Calculated

1

H

chemical shift of

the dimer

(ppm)

Difference of

calculated

1

H

chemical shift of

the monomer and

the dimer (ppm)

Maximum shift

observed in

experiment

1 2.54 0.23 2.46

3 6.78 -0.09 6.96

4 7.45 -0.12 7.36

10,11 7.69 -0.44 7.26

12,13 7.35 -0.28 6.8

The NMR chemical shifts of the aromatic protons are shifted to the low ppm

values while the methyl peak is shifted to the higher ppm values.

Results and Discussion

94

9 8 7 6 5 4 3 2

1

34

10 12

NMR chemical shift (ppm)

F

IGURE

4.48: The calculated

1

H NMR spectrum of YD dimer model in the gas phase

(b3lyp/6-311g(d,p)).

The calculated upfield shift of the aromatic protons in the dimer agrees well with

the experimental observations for the aggregates. This shift can be explained by

additional shielding of π electron system, which is above or below the adjacent

ring. In case of the methyl peak the calculation show a downfield shift whereas the

measurements show a small upfield chemical shift change at various dye

concentrations and temperatures (Fig.4.19). The reason can be the insufficient

accuracy of the calculated dimer structure. Another possible explanation for the

difference of calculated and observed changes in chemical shift upon dimer

formation may be that the relative orientation of the monomer assumed in the

model does not agree with the aggregate structure.

During the structure optimization the planar structure is changed to the twisted

structure, thus the position and the chemical environment of the methyl protons

are also changed resulting in chemical shift changes, which do not agree with

experimental observation. In the central part of the molecules there is almost no

Results and Discussion

95

twist (hydrogen atoms 10-13) and calculated and measured chemical shift changes

have the same direction.

To extract more information about the structures of the aggregates more

calculations for different models are needed.

Summary

96

CHAPTER 5 SUMMARY

The formation of aggregates from single molecules is a well-known phenomenon,

which can be observed, for example, in case of dye molecules or peptides. Previous

studies by light scattering had shown that mixtures of anionic direct dyes, namely

“Rot 2G” (RD) and “Gelb GA” (YD) form very large aggregates in the presence of

Mg

2+

ions. This was confirmed in this study by light microscopy, which showed that

even the pure YD formed aggregates in aqueous solutions containing Mg

2+

ions.

However, no visible aggregates were found for the RD. Therefore YD was chosen as

a typical model molecule for the investigation of the aggregation process in

aqueous solution. Different NMR techniques, such as chemical shift analysis,

relaxometry, pulsed-field gradient diffusion NMR in combination with UV

spectroscopy and quantum chemical calculation were used to obtain information

about the structure and the size of the aggregates.

The peaks in

1

H and

13

C NMR spectra of YD in DMSO solution, where no

aggregates occurs, were assigned using two-dimensional methods, such as COSY,

HMQC and HMBC. Good agreement is found between the measured chemical

shifts and the ones obtained from quantum chemical calculation for the monomer.

In aqueous solutions broad NMR peaks are observed due to aggregation.

Compared to the monomer spectra in DMSO, the peaks are shifted but no

additional peaks are found, indicating that there is fast exchange between

monomers and aggregates.

1

H spectra in aqueous solution were measured for

concentrations from 0.7 mM to 30 mM and in the temperature range 10-70° C. A

growth of the aggregates with increasing dye concentration and decreasing

temperature is found, which can be seen from the corresponding broadening of

the NMR peaks. In particular at low temperature and high concentrations the

measured peak intensities are up to 40 % lower than theoretically (based on the

concentration) expected. This shows that there are aggregates which are too big to

be detected by the NMR method used. From the line widths an average

hydrodynamic radius, corresponding to the NMR-visible fraction of the aggregates,

Summary

97

can be estimated. The values obtained are between 0.7 nm at a dye concentration

of 0.7 mM and a temperature of 70° C and up to 1.9 nm for the 30 mM solution at

T= 10° C. As the aggregates increase with increasing dye concentration and

decreasing temperature an upfield shift (towards lower ppm) of the

1

H NMR peaks

is found, which is more pronounced for the aromatic protons in the central region

of the molecules than for the methyl protons of the outer part of the molecules.

This shift occurs as a result of the shielding property of the π-electrons of

neighboring molecules. Since the molecules are stacked upon each other, the effect

is more significant for the middle section. The model of stacked molecules is in

good agreement with UV spectroscopy data, which show a blue shift of the

absorption maximum from 434 nm at a dye concentration 0.07 mM to 418 nm at 5

mM.

Diffusion NMR measurements of YD in aqueous solutions (0.7 – 30 mM) yield a

monoexponential decay, confirming that there is fast exchange between

monomers and aggregates. The diffusion coefficients were analyzed in terms of the

Stokes-Einstein model, yielding values of the hydrodynamic radius between 0.7 and

1.9 nm in very good agreement with the estimates obtained from the line width,

where from 0.2 nm up to 2 nm were found. The hydrodynamic radius of 1.9 nm

corresponds to an aggregation number between 22 and 37, depending on whether

a model of loosely aggregated molecules or a model of compact spherical

aggregates is used.

Appendix

98

APPENDIX

A1

D

IFFUSION OF THE MIXTURE

Figure A1.1: 1:1 mixture of RD and YD in aqueous solution [dye] = 45 mM [Mg

2+

] =

0.68 mM STE sequence (G = 0 – 9.63 T/m, = 0.5 ms, Δ = 100 ms).

Appendix

99

Analysis:

Figure A1.2: D

(fast)

= 9.58 · 10

–11

m

2

/s D

(slow)

= 9.06 · 10

–12

m

2

/s Fast component = 83

% R

h (small)

= 2 nm R

h (big)

= 20 nm. The aromatic region of the dyes was analyzed. A

non-exponential decay is observed, which can be fitted by two exponentials,

meaning that there are two species, slow/big aggregates and fast/small molecules

in the solution.

X10

10

Appendix

100

0 20 40 60 80 100 120 140

0,0

2,0x10

-11

4,0x10

-11

6,0x10

-11

8,0x10

-11

1,0x10

-10

1,2x10

-10

84 %

88 %

85 %

90 - 92 %

Diffusion coefficient D (m

2

/s)

[Mg

2+

] (mM)

small (fast) component (45 mM dye)

big (slow) component (45 mM dye)

small (fast) component (11 mM dye)

big (slow) component (11 mM dye)

(%) fraction of fast component

in observed signal

Figure A1.3: Diffusion coefficients of the dyes as a function of [Mg

2+

] at two

different dye concentrations. From the biexponential fit two diffusion coefficients

were extracted, showing that there are fast/small and slow/big aggregates.

Table A1.1

Diffusion coefficient as a function of [Mg

2+

] at two different dye concentrations

[dye] = 45 mM [dye] = 11 mM [Mg

2+

]

(mM) R

h

(small)

(nm)

R

h

(big)

(nm)

R

h

(small)

(nm)

R

h

(big)

(nm)

0.00 2.5 40 1.8 20

0.69 2.1 95 1.7 35

6.87 2.0 96

137.18 2.2 181

Appendix

101

0 20 40 60 80 100 120 140

1

2

3

4

5

6

7

Observed Signal Intensity

[Mg

2+

] (mM)

High (45 mmol/L) dye concentration

Low (11 mmol/L) dye concentration

F

IGURE

A1.4:

Observed signal intensity at various [Mg

2+

] at two dye concentrations. The

lower dye concentration show higher signal intensity. The reason of these observations is

that the big aggregates do not give signal (too big, rigid or precipitated).

Appendix

102

A2

D

IFFUSION DATA OF

YD

Table A2.1 Diffusion Data in aqueous solution – Aromatic region

[YD]

(mM)

T

(°C)

D

(m

2

/s)

G

max

(T/m)

∆

(s)

All

points

Used

Points Pulse

0,7 30 2,81E-10

3,8058 0,01982

24 22 STE

10 30 2,19E-10

5,6875 0,01998

24 22 STE

15 10 7,89E-11

5,6875 0,01999

32 30 DSTE

15 10 8,57E-11

5,6875 0,03999

32 30 DSTE

15 10 9,13E-11

5,6875 0,09999

32 13 DSTE

15 10 9,35E-11

7,5075 0,02004

16 14 STE

15 10 9,36E-11

5,5055 0,03992

16 13 STE

15 15 1,15E-10

9,0098 0,02037

16 14 STE

15 15 1,16E-10

7,0078 0,02013

24 22 STE

15 20 1,34E-10

5,6875 0,02999

32 28 DSTE

15 20 1,42E-10

5,6875 0,09999

32 17 DSTE

15 20 1,41E-10

6,507 0,01981

16 14 STE

15 20 1,42E-10

6,0062 0,0399 16 14 STE

15 25 1,75E-10

6,0062 0,01978

16 14 STE

15 25 1,70E-10

5,0057 0,03992

16 14 STE

15 30 1,78E-10

5,6875 0,01999

32 30 DSTE

15 30 2,15E-10

5,6875 0,09999

32 16 DSTE

15 30 2,04E-10

5,5055 0,01984

16 14 STE

15 30 2,04E-10

4,0042 0,03984

16 14 STE

15 35 2,46E-10

5,0057 0,01982

16 14 STE

15 35 2,64E-10

3,5034 0,03986

16 14 STE

15 35 2,10E-10

5,6856 0,02006

16 14 DSTE

15 40 2,51E-10

5,6875 0,01999

32 28 DSTE

15 40 2,75E-10

5,6875 0,09999

32 12 DSTE

15 40 2,55E-10

5,6875 0,01999

32 30 DSTE

15 40 2,86E-10

3,5034 0,0811 32 19 DSTE

15 45 3,39E-10

4,5049 0,04133

16 10 DSTE

15 45 3,11E-10

2,0021 0,03971

16 14 DSTE

15 50 3,36E-10

5,6875 0,01999

32 30 DSTE

15 50 3,71E-10

3,5034 0,03936

32 30 DSTE

15 60 4,34E-10

3,5034 0,01936

32 30 DSTE

15 60 4,49E-10

3,5034 0,01936

32 30 DSTE

15 60 4,47E-10

3,5034 0,01936

32 30 DSTE

15 60 4,73E-10

3,5034 0,03936

32 23 DSTE

Appendix

103

[YD]

(mM)

T

(°C)

D

(m

2

/s)

Gmax

(T/m)

∆

(s)

All

points

Used

Points Pulse

15 60 4,38E-10

5,6875 0,01999

32 23 DSTE

15 60 5,52E-10

3,0036 0,03922

16 14 DSTE

15 65 4,96E-10

3,5034 0,01936

32 30 DSTE

15 65 5,16E-10

2,5028 0,03863

16 14 DSTE

15 70 5,96E-10

3,0036 0,01922

32 30 DSTE

15 70 6,16E-10

2,0021 0,03908

16 14 DSTE

15 30 3,79E-10

5,6875 0,00911

16 9 STE

15 30 4,07E-10

3,5034 0,01884

16 14 STE

15 30 2,83E-10

3,5034 0,03936

16 14 DSTE

15 30 2,06E-10

3,5034 0,03989

24 22 STE

15 30 2,06E-10

5,6875 0,01998

24 22 STE

20 10 7,16E-11

3,5034 0,01936

32 30 DSTE

20 10 7,40E-11

5,6875 0,01999

32 30 DSTE

20 10 7,73E-11

5,6875 0,03999

32 30 DSTE

20 20 1,22E-10

5,6875 0,02999

32 29 DSTE

20 30 1,62E-10

5,6875 0,01999

32 30 DSTE

20 40 2,30E-10

5,6875 0,01999

32 30 DSTE

20 50 3,35E-10

3,5034 0,03936

32 30 DSTE

20 60 4,40E-10

3,5034 0,02936

32 30 DSTE

20 70 5,38E-10

3,5034 0,01936

32 28 DSTE

20 30 1,82E-10

5,6875 0,01998

24 22 STE

25 30 1,73E-10

5,6875 0,01998

24 22 STE

30 10 6,92E-11

5,6875 0,09999

32 26 DSTE

30 10 6,93E-11

5,8059 0,09998

32 26 STE

30 10 6,67E-11

5,6875 0,09999

32 23 DSTE

30 10 6,87E-11

9,63 0,09998

32 18 STE

30 10 6,95E-11

6,507 0,10006

16 14 STE

30 10 6,91E-11

6,507 0,10001

64 62 STE

30 15 8,76E-11

9,0098 0,02011

16 14 STE

30 15 8,75E-11

7,0078 0,03998

24 22 STE

30 20 1,09E-10

8,5091 0,02005

16 14 STE

30 20 1,09E-10

6,507 0,03992

24 22 STE

30 30 1,58E-10

5,6875 0,03999

32 30 DSTE

30 30 1,65E-10

5,8059 0,03998

32 30 STE

30 30 1,53E-10

5,6875 0,03999

32 23 DSTE

30 30 1,46E-10

5,6875 0,01999

32 30 DSTE

30 40 2,05E-10

5,6875 0,01999

32 30 DSTE

30 50 3,05E-10

3,5034 0,03936

32 30 DSTE

30 50 1,67E-08

3,5034 0,0398 32 30 STE

30 50 2,76E-10

5,6875 0,01999

32 30 DSTE

30 60 3,71E-10

5,6875 0,01999

32 24 DSTE

30 70 4,72E-10

3,5034 0,01936

32 30 DSTE

30 30 1,65E-10

9,63 0,04082

19 13 STE

30 30 1,66E-10

5,6875 0,01968

24 22 STE

Appendix

104

[YD]

(mM)

T

(°C)

D

(m

2

/s)

G

max

(T/m)

∆

(s)

All

points

Used

Points Pulse

5 10 9,20E-11

3,5034 0,01936

32 30 DSTE

5 10 1,09E-10

5,6875 0,01999

32 30 DSTE

5 20 1,59E-10

5,6875 0,01999

32 30 DSTE

5 30 2,44E-10

3,5034 0,03936

32 30 DSTE

5 40 3,64E-10

3,5034 0,03936

32 26 DSTE

5 50 4,12E-10

3,5034 0,01936

32 30 DSTE

5 60 5,40E-10

3,5034 0,01936

32 30 DSTE

5 70 6,63E-10

3,5034 0,01936

32 26 DSTE

5 30 2,56E-10

5,6875 0,01998

24 22 STE

Table A2.2 Diffusion Data in DMSO solution – Aromatic region

[YD]

(mM)

T

(°C)

D

(m

2

/s)

G

max

(T/m)

∆

(s)

All

points

Used

Points Pulse

20 30 1,64E-10

3,5034 0,01936

32 30 DSTE

20 30 1,65E-10

5,6875 0,01999

32 30 DSTE

20 70 3,52E-10

1,5013 0,01908

16 14 DSTE

20 70 3,87E-10

2,5028 0,03908

16 14 DSTE

20 60 3,22E-10

5,6875 0,01999

32 30 DSTE

20 60 3,07E-10

5,6875 0,01999

32 30 DSTE

20 30 1,62E-10

6,0062 0,02006

24 21 DSTE

20 30 2,31E-10

3,5034 0,03971

16 14 STE

20 30 1,75E-10

3,0036 0,06013

16 13 DSTE

20 30 1,99E-10

5,6875 0,01998

24 22 STE

Appendix

105

Table A2.3 Diffusion Data in aqueous solution – Methyl peak

[YD]

(mM)

T

(°C)

D

(m

2

/s)

G

max

(T/m)

∆

(s)

All

points

Used

Points Pulse

0,7 30 2,92E-10 3,8058

0,01982

24 22 STE

10 30 2,22E-10 5,6875

0,01998

24 22 STE

15 10 8,58E-11 5,6875

0,01999

32 30 DSTE

15 10 8,96E-11 5,6875

0,03999

32 30 DSTE

15 10 1,02E-10 5,6875

0,09999

32 13 DSTE

15 10 9,43E-11 7,5075

0,02004

16 14 STE

15 10 9,40E-11 5,5055

0,03992

16 13 STE

15 15 1,17E-10 9,0098

0,02037

16 14 STE

15 15 1,16E-10 7,0078

0,02013

24 22 STE

15 20 1,38E-10 5,6875

0,02999

32 28 DSTE

15 20 1,48E-10 5,6875

0,09999

32 17 DSTE

15 20 1,43E-10 6,507 0,01981

16 14 STE

15 20 1,44E-10 6,0062

0,0399 16 14 STE

15 25 1,69E-10 6,0062

0,01978

16 14 STE

15 25 1,71E-10 5,0057

0,03992

16 14 STE

15 30 1,87E-10 5,6875

0,01999

32 30 DSTE

15 30 2,16E-10 5,6875

0,09999

32 16 DSTE

15 30 2,05E-10 5,5055

0,01984

16 14 STE

15 30 2,07E-10 4,0042

0,03984

16 14 STE

15 35 2,48E-10 5,0057

0,01982

16 14 STE

15 35 2,65E-10 3,5034

0,03986

16 14 STE

15 35 2,13E-10 5,6856

0,02006

16 14 DSTE

15 40 2,60E-10 5,6875

0,01999

32 28 DSTE

15 40 2,81E-10 5,6875

0,09999

32 12 DSTE

15 40 2,56E-10 5,6875

0,01999

32 30 DSTE

15 40 2,71E-10 3,5034

0,0811 32 19 DSTE

15 45 3,33E-10 4,5049

0,04133

16 10 DSTE

15 45 3,27E-10 2,0021

0,03971

16 14 DSTE

15 50 3,39E-10 5,6875

0,01999

32 30 DSTE

15 50 3,70E-10 3,5034

0,03936

32 30 DSTE

15 60 4,69E-10 3,5034

0,01936

32 30 DSTE

15 60 4,77E-10 3,5034

0,01936

32 30 DSTE

15 60 4,73E-10 3,5034

0,01936

32 30 DSTE

15 60 4,93E-10 3,5034

0,03936

32 23 DSTE

Appendix

106

[YD]

(mM)

T

(°C)

D

(m

2

/s)

G

max

(T/m)

∆

(s)

All

points

Used

Points Pulse

15 60 4,31E-10 5,6875

0,01999

32 23 DSTE

15 60 5,51E-10 3,0036

0,03922

16 14 DSTE

15 65 5,03E-10 3,5034

0,01936

32 30 DSTE

15 65 5,34E-10 2,5028

0,03863

16 14 DSTE

15 70 5,88E-10 3,0036

0,01922

32 30 DSTE

15 70 6,83E-10 2,0021

0,03908

16 14 DSTE

15 30 3,58E-10 5,6875

0,00911

16 9 STE

15 30 4,46E-10 3,5034

0,01884

16 14 STE

15 30 3,01E-10 3,5034

0,03936

16 14 DSTE

15 30 2,07E-10 3,5034

0,03989

24 22 STE

15 30 2,04E-10 5,6875

0,01998

24 22 STE

20 10 7,50E-11 3,5034

0,01936

32 30 DSTE

20 10 7,54E-11 5,6875

0,01999

32 30 DSTE

20 10 7,85E-11 5,6875

0,03999

32 30 DSTE

20 20 1,21E-10 5,6875

0,02999

32 29 DSTE

20 30 1,67E-10 5,6875

0,01999

32 30 DSTE

20 40 2,32E-10 5,6875

0,01999

32 30 DSTE

20 50 3,42E-10 3,5034

0,03936

32 30 DSTE

20 60 4,43E-10 3,5034

0,02936

32 30 DSTE

20 70 5,41E-10 3,5034

0,01936

32 28 DSTE

20 30 1,83E-10 5,6875

0,01998

24 22 STE

25 30 1,74E-10 5,6875

0,01998

24 22 STE

30 10 7,02E-11 5,6875

0,09999

32 26 DSTE

30 10 6,94E-11 5,8059

0,09998

32 26 STE

30 10 6,92E-11 5,6875

0,09999

32 23 DSTE

30 10 6,96E-11 9,63 0,09998

32 18 STE

30 10 6,97E-11 6,507 0,10006

16 14 STE

30 10 6,97E-11 6,507 0,10001

64 62 STE

30 15 8,81E-11 9,0098

0,02011

16 14 STE

30 15 8,93E-11 7,0078

0,03998

24 22 STE

30 20 1,10E-10 8,5091

0,02005

16 14 STE

30 20 1,10E-10 6,507 0,03992

24 22 STE

30 30 1,60E-10 5,6875

0,03999

32 30 DSTE

30 30 1,66E-10 5,8059

0,03998

32 30 STE

30 30 1,52E-10 5,6875

0,03999

32 23 DSTE

30 30 1,47E-10 5,6875

0,01999

32 30 DSTE

30 40 2,10E-10 5,6875

0,01999

32 30 DSTE

30 50 3,03E-10 3,5034

0,03936

32 30 DSTE

30 50 1,57E-08 3,5034

0,0398 32 30 STE

30 50 2,79E-10 5,6875

0,01999

32 30 DSTE

30 60 3,72E-10 5,6875

0,01999

32 24 DSTE

30 70 4,86E-10 3,5034

0,01936

32 30 DSTE

30 30 1,62E-10 9,63 0,04082

19 13 STE

30 30 1,63E-10 5,6875

0,01968

24 22 STE

Appendix

107

[YD]

(mM)

T

(°C)

D

(m

2

/s)

G

max

(T/m)

∆

(s)

All

points

Used

Points Pulse

5 10 1,23E-10 3,5034

0,01936

32 30 DSTE

5 10 1,12E-10 5,6875

0,01999

32 30 DSTE

5 20 1,63E-10 5,6875

0,01999

32 30 DSTE

5 30 2,43E-10 3,5034

0,03936

32 30 DSTE

5 40 3,70E-10 3,5034

0,03936

32 26 DSTE

5 50 4,06E-10 3,5034

0,01936

32 30 DSTE

5 60 5,60E-10 3,5034

0,01936

32 30 DSTE

5 70 6,60E-10 3,5034

0,01936

32 26 DSTE

5 30 2,57E-10 5,6875

0,01998

24 22 STE

Table A2.4 Diffusion Data in DMSO solution – Methyl peak

[YD]

(mM)

T

(°C)

D

(m

2

/s)

G

max

(T/m)

(s)

All

points

Used

Points Pulse

20 30 1,78E-10

3,5034

0,01936

32

30

DSTE

20 30 1,83E-10

5,6875

0,01999

32

30

DSTE

20 70 4,67E-10

1,5013

0,01908

16

14

DSTE

20 70 4,60E-10

2,5028

0,03908

16

14

DSTE

20 60 3,52E-10

5,6875

0,01999

32

30

DSTE

20 60 3,34E-10

5,6875

0,01999

32

30

DSTE

20 30 1,82E-10

6,0062

0,02006

24

21

DSTE

20 30 2,40E-10

3,5034

0,03971

16

14

STE

20 30 1,97E-10

3,0036

0,06013

16

13

DSTE

20 30 2,06E-10

5,6875

0,01998

24

22

STE

Appendix

108

A3

P

ULSE PROGRAM FOR DIFFUSION EXPERIMENTS

STE

;diffSte

;2D stimulated echo sequence

;new version including decoupling 13.05.04 KLZ

;comments updated 24.06.04 KLZ

;decoupling corrected 02.07.04 KLZ

;CLASS, DIM, TYPE, define list<gradient>, acqt0 added 24.05.06 KLZ

;blanking syntax updated 20.02.2007 KLZ

;$CLASS=diff

;$DIM=2D

;$TYPE=exp

;$OWNER=Bruker

#include <Grad.incl>

#include <Avance.incl>

define list<gradient> diff_ramp=<diff_ramp>

"acqt0=0"

ze

10u

5m pl1:f1 ;set rf power level

start, 1u

if (l14) { ; if decoupling in use

1u do:f2 ; decoupler off during d1

} else {

1u

}

if (l12) { ; if lock in use

d1 LOCKH_OFF ; lock on during d1

d11 UNBLKGRAD ; unblank gradient amplifier, lock hold during

experiment

} else { ; if locnuc off

Appendix

109

d1

d11 UNBLKGRAMP ; unblank gradient amplifier

}

;-------------------------- Start of dummy gradient loop ---------------------------

if (l3) { ; dummy gradient pulses in use

dummy, p17:gp1*diff_ramp ; trapezoidal gradient pulse

p18:gp2*diff_ramp ; trapezoidal gradient pulse

p17:gp3*diff_ramp ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

d9 BLKGRAMP ; tau

if (l11) { ; if spoiler in use

d11 UNBLKGRAMP ; unblank gradient amplifier

p17:gp4 ; trapezoidal gradient pulse

p19:gp5 ; trapezoidal gradient pulse

p17:gp6 ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

}

if (l12) { ; if lock in use

d5 BLKGRAD ; blank gradient amp., lock on during long

tau

d11 UNBLKGRAD ; unblank gradient amplifier

} else { ; if locnuc off

d5 BLKGRAMP ; long tau

d11 UNBLKGRAMP ; unblank gradient

amplifier

}

lo to dummy times l13 ; l13 number of dummy gradient pulses

}

;-------------------------- Start of experiment -----------------------------------

p1:f1 ph1 ; 90 degree pulse

p17:gp1*diff_ramp ; trapezoidal gradient pulse

p18:gp2*diff_ramp ; trapezoidal gradient pulse

p17:gp3*diff_ramp ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

d9 BLKGRAMP ; tau

p1:f1 ph2 ; 90 degree pulse

if (l11) { ; if spoiler in use

d11 UNBLKGRAMP ; unblank gradient amplifier

p17:gp4 ; trapezoidal gradient pulse

p19:gp5 ; trapezoidal gradient pulse

Appendix

110

p17:gp6 ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

}

if (l12) { ; if lock in use

d5 BLKGRAD ; blank gradient amp., lock on during long

tau

d11 UNBLKGRAD ; unblank gradient amplifier

} else { ; if locnuc off

d5 BLKGRAMP ; long tau

d11 UNBLKGRAMP ; unblank gradient amplifier

}

p1:f1 ph3 ; 90 degree pulse

p17:gp1*diff_ramp ; trapezoidal gradient pulse

p18:gp2*diff_ramp ; trapezoidal gradient pulse

p17:gp3*diff_ramp ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

d10 ph0 BLKGRAMP ; tau

if (l14) { ; if f2 on

go=start ph31 cpd2:f2 ; start acquisition with decoupling

} else { ; if f2 off

go=start ph31 ; start acquisition

}

100u wr #0 if #0 zd igrad diff_ramp ; store data, increment gradient ramp

lo to start times td1 ; td1 = number of gradientsteps

if (l14) { ; if decoupling in use

100m do:f2 ; wait for data storage, decoupler off

} else {

100m ; wait for data storage

}

if (l12) { ; if lock in use

100m rf #0 LOCKH_OFF ; reset file pointer, lock on

} else { ; if locnuc off

100m rf #0 ; reset file pointer

}

lo to start times l1 ; l1 = Number of repetitions

exit

ph0=0

ph1= 0 0 0 0 2 2 2 2 1 1 1 1 3 3 3 3

ph2= 1 3 0 2

Appendix

111

ph3= 1 3 0 2

ph31=0 0 2 2 2 2 0 0 3 3 1 1 1 1 3 3

;pl1: f1 channel - power level for pulse (default)

;p1: f1 channel - 90 degree pulse

;p17: gradient ramp time

;p18: gradient duration - p17

;p19: spoil gradient duration - 2*p17

;d1: relaxation delay; 1-5 * T1

;d2: gradient stabilisation time

;d5: DELTA remainder

;d9: tau remainder

;d10: tau remainder, used to shift trigger position

;d11: gradient amplifier unblank delay 200 us

;gpnam1: ramp up of trapezoidal

;gpnam2: plateau of trapezoidal

;gpnam3: ramp down of trapezoidal

;gpnam4: ramp up of trapezoidal

;gpnam5: plateau of trapezoidal

;gpnam6: ramp down of trapezoidal

;gpx1: x-diffusion gradient amplitude

;gpx2: x-diffusion gradient amplitude

;gpx3: x-diffusion gradient amplitude

;gpx4: x-spoiler gradient amplitude

;gpx5: x-spoiler gradient amplitude

;gpx6: x-spoiler gradient amplitude

;gpy1: y-diffusion gradient amplitude

;gpy2: y-diffusion gradient amplitude

;gpy3: y-diffusion gradient amplitude

;gpy4: y-spoiler gradient amplitude

;gpy5: y-spoiler gradient amplitude

;gpy6: y-spoiler gradient amplitude

;gpz1: z-diffusion gradient amplitude

;gpz2: z-diffusion gradient amplitude

;gpz3: z-diffusion gradient amplitude

;gpz4: z-spoiler gradient amplitude

;gpz5: z-spoiler gradient amplitude

;gpz6: z-spoiler gradient amplitude

Appendix

112

;NS: 16 * n

;td1: number of experiments

;l1: Repetitions of the whole experiment

;l3: dummy gradient pulses off/on 0/1

;l11: spoil gradient off/on 0/1

;l12: lock off/on 0/1

;l13: number of dummy gradient pulses

;l14: decoupling off/on 0/1

;l21: diffusion gradient list type

;l27: use taumin off/on 0/1

;l28: use default parameters off/on 0/1

;l29: use userdefined pulse program off/on 0/1

Appendix

113

DSTE

;diffDste

;2D double stimulated echo sequence

;new version including decoupling 09.06.04 KLZ

;comments updated 24.06.04 KLZ

;decoupling corrected 02.07.04 KLZ

;CLASS, DIM, TYPE, define list<gradient>, acqt0 added 24.05.06 KLZ

;blanking syntax updated 20.02.2007 KLZ

;$CLASS=diff

;$DIM=2D

;$TYPE=exp

;$OWNER=Bruker

#include <Grad.incl>

#include <Avance.incl>

3m ; do not remove this delay

define list<gradient> diff_ramp=<diff_ramp>

ze

10u

5m pl1:f1 ;set rf power level

start, 1u

if (l14) { ; if decoupling in use

1u do:f2 ; decoupler off during d1

} else {

1u

}

if (l12) { ; if lock in use

d1 LOCKH_OFF ; lock on during d1

d11 UNBLKGRAD ; unblank gradient amplifier, lock hold during

experiment

} else { ; if locnuc off

d1

d11 UNBLKGRAMP ; unblank gradient amplifier

}

;-------------------------- Start of dummy gradient loop ---------------------------

Appendix

114

if (l3) { ; dummy gradient pulses in use

dummy, p17:gp1*diff_ramp ; trapezoidal gradient pulse

p18:gp2*diff_ramp ; trapezoidal gradient pulse

p17:gp3*diff_ramp ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

d9 BLKGRAMP ; tau

if (l11) { ; if spoiler in use

d11 UNBLKGRAMP ; unblank gradient amplifier

p17:gp4 ; trapezoidal gradient pulse

p19:gp5 ; trapezoidal gradient pulse

p17:gp6 ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

}

if (l12) { ; if lock in use

d5 BLKGRAD ; blank gradient amp., lock on during

long tau

d11 UNBLKGRAD ; unblank gradient amplifier

} else { ; if locnuc off

d5 BLKGRAMP ; long tau

d11 UNBLKGRAMP ; unblank gradient

amplifier

}

lo to dummy times l13 ; l13 number of dummy gradient pulses

}

;-------------------------- Start of experiment -----------------------------------

p1:f1 ph1 ; 90 degree pulse

p17:gp1*diff_ramp ; trapezoidal gradient pulse

p18:gp2*diff_ramp ; trapezoidal gradient pulse

p17:gp3*diff_ramp ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

d9 BLKGRAMP ; tau

p1:f1 ph2 ; 90 degree pulse

if (l11) { ; if spoiler in use

d11 UNBLKGRAMP ; unblank gradient amplifier

p17:gp4 ; trapezoidal gradient pulse

p19:gp5 ; trapezoidal gradient pulse

p17:gp6 ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

}

if (l12) { ; if lock in use

Appendix

115

d5 BLKGRAD ; blank gradient amp., lock on during long

tau

d11 UNBLKGRAD ; unblank gradient amplifier

} else { ; if locnuc off

d5 BLKGRAMP ; long tau

d11 UNBLKGRAMP ; unblank gradient amplifier

}

p1:f1 ph3 ; 90 degree pulse

p17:gp1*diff_ramp ; trapezoidal gradient pulse

p18:gp2*diff_ramp ; trapezoidal gradient pulse

p17:gp3*diff_ramp ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

d9 ph0 ; tau

p17:gp1*diff_ramp ; trapezoidal gradient pulse

p18:gp2*diff_ramp ; trapezoidal gradient pulse

p17:gp3*diff_ramp ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

d9 BLKGRAMP ; tau

p1:f1 ph4 ; 90 degree pulse

if (l11) { ; if spoiler in use

d11 UNBLKGRAMP ; unblank gradient amplifier

p17:gp4*1.13 ; trapezoidal gradient pulse

p19:gp5*1.13 ; trapezoidal gradient pulse

p17:gp6*1.13 ; trapezoidal gradient pulse

d2

}

if (l12) { ; if lock in use

d5 BLKGRAD ; blank gradient amp., lock on during long

tau

d11 UNBLKGRAD ; unblank gradient amplifier

} else { ; if locnuc off

d5 BLKGRAMP ; long tau

d11 UNBLKGRAMP ; unblank gradient amplifier

}

p1:f1 ph5 ; 90 degree pulse

p17:gp1*diff_ramp ; trapezoidal gradient pulse

p18:gp2*diff_ramp ; trapezoidal gradient pulse

p17:gp3*diff_ramp ; trapezoidal gradient pulse

d2 ; gradient stabilisation time

d10 ph0 BLKGRAMP ; tau

Appendix

116

if (l14) { ; if f2 on

go=start ph31 cpd2:f2 ; start acquisition with decoupling

} else { ; if f2 off

go=start ph31 ; start acquisition

}

100u wr #0 if #0 zd igrad diff_ramp ; store data, increment gradient ramp

lo to start times td1 ; td1 = number of gradientsteps

if (l14) { ; if decoupling in use

100m do:f2 ; wait for data storage, decoupler off

} else {

100m ; wait for data storage

}

if (l12) { ; if lock in use

100m rf #0 LOCKH_OFF ; reset file pointer, lock on

} else { ; if locnuc off

100m rf #0 ; reset file pointer

}

lo to start times l1 ; l1 = Number of repetitions

exit

ph0=0

ph1= 0 1 2 3

ph2= 0

ph3= 2 3

ph4= 2 2 2 2 0 0 0 0

ph5= 0

ph6= 0

ph7= 0

ph31=0 0 2 2 2 2 0 0

;pl1: f1 channel - power level for pulse (default)

;p1: f1 channel - 90 degree pulse

;p17: gradient ramp time

;p18: gradient duration - p17

;p19: spoil gradient duration - 2*p17

;d1: relaxation delay; 1-5 * T1

;d2: gradient stabilisation time

;d5: DELTA/2 remainder

;d9: tau remainder

;d10: tau remainder, used to shift trigger position

Appendix

117

;d11: gradient amplifier unblank delay 200 us

;gpnam1: ramp up of trapezoidal

;gpnam2: plateau of trapezoidal

;gpnam3: ramp down of trapezoidal

;gpnam4: ramp up of trapezoidal

;gpnam5: plateau of trapezoidal

;gpnam6: ramp down of trapezoidal

;gpx1: x-diffusion gradient amplitude

;gpx2: x-diffusion gradient amplitude

;gpx3: x-diffusion gradient amplitude

;gpx4: x-spoiler gradient amplitude

;gpx5: x-spoiler gradient amplitude

;gpx6: x-spoiler gradient amplitude

;gpy1: y-diffusion gradient amplitude

;gpy2: y-diffusion gradient amplitude

;gpy3: y-diffusion gradient amplitude

;gpy4: y-spoiler gradient amplitude

;gpy5: y-spoiler gradient amplitude

;gpy6: y-spoiler gradient amplitude

;gpz1: z-diffusion gradient amplitude

;gpz2: z-diffusion gradient amplitude

;gpz3: z-diffusion gradient amplitude

;gpz4: z-spoiler gradient amplitude

;gpz5: z-spoiler gradient amplitude

;gpz6: z-spoiler gradient amplitude

;NS: 8 * n

;td1: number of experiments

;l1: Repetitions of the whole experiment

;l3: dummy gradient pulses off/on 0/1

;l11: spoil gradient off/on 0/1

;l12: lock off/on 0/1

;l13: number of dummy gradient pulses

;l14: decoupling off/on 0/1

;l21: diffusion gradient list type

;l27: use taumin off/on 0/1

;l28: use default parameters off/on 0/1

;l29: use userdefined pulse program off/on 0/1

List of Symbols

118

LIST OF SYMBOLS

13

C NMR Carbon NMR

1D One Dimensional

1

H NMR Proton NMR

2D Two Dimensional

α Shearing angle

B

0

External magnetic field

COSY Correlation Spectroscopy

δ Gradient duration

∆ Diffusion time

D

2

O Deuterated water

DFT Density functional theory

DMSO Dimethyl sulfoxide

[dye] Dye concentration (mM)

∆ν Line width

DSS Sodium 4,4-dimethyl-4-silapentane-1-sulfonate

DST Double stimulated echo

ε Molar extinction coefficient

f Friction coefficient

γ Gyromagnetic ratio

g Gradient strength

GIAO Gauge invariant atomic orbitals

G

s

Spoil gradient

h Planck constant

η Viscosity

HMBC Heteronuclear multiple bond correlation

HMQC Heteronuclear multiple quantum coherence

I Spin quantum number

k

B

Boltzmann constant

List of Symbols

119

λ Wavelength (nm)

µ Transition dipole vector

M Magnetization vector

[MgSO

4

] Magnesium sulfate concentration

M

RD

Molar mass of the RD anions

M

YD

Molar mass of YD anions

ν Frequency

N Aggregation number

NMR Nuclear Magnetic Resonance

PFG Pulsed field gradient

r Coordinate vector of the nuclei

RD Red dye

rf Radio frequency

R

h

Hydrodynamic radius

S Signal attenuation

STE Stimulated echo

T Temperature

τ Delay between pulses

T

1

Spin-lattice relaxation time

T

2

Spin-spin relaxation time

τ

c

Rotational correlation time

TMS Tetra methyl silane

UV Ultra violet (λ= 190-400 nm)

v Flow velocity

V

h

Hydrodynamic volume

V

i,j

Interaction energy

ω

0

Larmor frequency

YD Yellow dye

Θ Dihedral angle

References

120

REFERENCES

1. Tiddy, G. J. T.; Mateer, D. L.; Ormerod, A. P.; Harrison, W. J.; Edwards, D. J. Langmuir

1995, 11, 390-393.

2. Yamazaki, T.; Chang, T. Y.; Haass, C.; Ihara, Y. Journal of Biological Chemistry 2001,

276, 4454-4460.

3. Bayer, T. A.; Breyhan, H.; Duan, K.; Rettig, J.; Wirths, O. European

Neuropsychopharmacology 2007, 17, S205-S206.

4. Tanford, C. The hydrophobic effect: formation of micelles and biological

membranes, 2d ed.; Wiley: New York, 1980.

5. Evans, D. F.; Wennerström, H. The colloidal domain: where physics, chemistry,

biology, and technology meet; VCH Publishers: New York, NY, 1994.

6. Kannan, N.; Vishveshwara, S. Protein Engineering 2000, 13, 753-761.

7. Gazit, E. Faseb Journal 2002, 16, 77-83.

8. McGaughey, G. B.; Gagne, M.; Rappe, A. K. Journal of Biological Chemistry 1998,

273, 15458-15463.

9. Cockroft, S. L.; Perkins, J.; Zonta, C.; Adams, H.; Spey, S. E.; Low, C. M. R.; Vinter, J. G.;

Lawson, K. R.; Urch, C. J.; Hunter, C. A. Organic & Biomolecular Chemistry 2007, 5,

1062-1080.

10. Hunter, C. A.; Lawson, K. R.; Perkins, J.; Urch, C. J. Journal of the Chemical Society-

Perkin Transactions 2 2001, 651-669.

11. Waters, M. L. Biopolymers 2004, 76, 435-445.

12. Waters, M. L. Current Opinion in Chemical Biology 2002, 6, 736-741.

13. Scheibe, G. Angewante Chemie 1936, 49, 563.

14. Jelley, E. E. Nature 1936, 138, 1009.

15. Rohatgi, K. K.; Singhal, G. S. Journal of Physical Chemistry 1966, 70, 1695-&.

16. Hida, M.; Yabe, A.; Murayama, H.; Hayashi, M. Bulletin of the Chemical Society of

Japan 1968, 41, 1776-&.

17. Mukerjee, P.; Ghosh, A. K. Journal of the American Chemical Society 1970, 92, 6419-

&.

18. Mukerjee, P. Journal of Pharmaceutical Sciences 1972, 61, 478-&.

19. Pugh, D.; Giles, C. H.; Duff, D. G. Transactions of the Faraday Society 1971, 67, 563-

&.

20. Monahan, A. R.; Germano, N. J.; Blossey, D. F. Journal of Physical Chemistry 1971,

75, 1227-&.

21. Monahan, A. R.; Deluca, A. F.; Ward, A. T. Abstracts of Papers of the American

Chemical Society 1971, 24-&.

22. Monahan, A. R.; Blossey, D. F. Abstracts of Papers of the American Chemical Society

1970, 141-&.

23. Monahan, A. R.; Blossey, D. F. Journal of Physical Chemistry 1970, 74, 4014-&.

24. Reeves, R. L.; Maggio, M. S.; Harkaway, S. A. Journal of Physical Chemistry 1979, 83,

2359-2368.

25. Roda, A.; Hofmann, A. F.; Mysels, K. J. Journal of Biological Chemistry 1983, 258,

6362-6370.

26. Hamada, K.; Yamada, K.; Mitsuishi, M.; Ohira, M.; Mesuda, K. Journal of the

Chemical Society-Faraday Transactions 1995, 91, 1601-1605.

27. Hamada, K.; Take, S.; Iijima, T.; Amiya, S. Journal of the Chemical Society-Faraday

Transactions I 1986, 82, 3141-3148.

References

121

28. Sanna, C.; La Mesa, C.; Mannina, L.; Stano, P.; Viel, S.; Segre, A. Langmuir 2006, 22,

6031-6041.

29. Katsuyama, I.; Khalil, A. A.; Dunbar, C.; Zjawiony, J. K. Spectroscopy Letters 2003, 36,

477-485.

30. Neumann, B. Langmuir 2001, 17, 2675-2682.

31. Mitra, A.; Seaton, J.; Capitani, J. F.; Assarpour, A. Abstracts of Papers of the

American Chemical Society 1999, 217, U169-U169.

32. Mitra, A.; Seaton, P. J. Abstracts of Papers of the American Chemical Society 2000,

219, U185-U185.

33. Mitra, A.; Seaton, P. J.; Assarpour, R. A.; Williamson, T. Tetrahedron 1998, 54,

15489-15498.

34. Mitra, A.; Seaton, P. J.; Capitani, J. F.; Assarpour, A. Journal of the Indian Chemical

Society 1998, 75, 823-830.

35. Rodriguez-Ropero, F.; Casanovas, J.; Aleman, C. Journal of Computational Chemistry

2008, 29, 69-78.

36. Silverstein, R. M.; Bassler, G. C.; Morrill, T. C. Spectrometric identification of organic

compounds, 5th ed.; Wiley: New York, 1991.

37. Sehnert, J., Universität Bayreuth, 2007.

38. Cramer, C. J. Essentials of computational chemistry: theories and models, 2nd ed.;

Wiley: Chichester, West Sussex, England; Hoboken, NJ, 2004.

39. Young, D. C. Computational chemistry: a practical guide for applying techniques to

real-world problems; Wiley-Interscience: New York, 2001.

40. Hahn, E. L. Physical Review 1950, 77, 746-746.

41. Hahn, E. L. Physical Review 1950, 80, 580-594.

42. Stejskal, E. O.; Tanner, J. E. Journal of Chemical Physics 1965, 42, 288-&.

43. Stejskal, E. O. Journal of Chemical Physics 1965, 43, 3597-&.

44. Stejskal, E. O. Advances in Molecular Relaxation Processes 1972, 3, 27-&.

45. Beck, S.; Geyer, A.; Brintzinger, H. H. Chemical Communications 1999, 2477-2478.

46. Cohen, Y.; Avram, L.; Frish, L. Angewandte Chemie-International Edition 2005, 44,

520-554.

47. Dehner, A.; Kessler, H. Chembiochem 2005, 6, 1550-1565.

48. Adalsteinsson, T.; Dong, W. F.; Schonhoff, M. Journal of Physical Chemistry B 2004,

108, 20056-20063.

49. Adalsteinsson, T.; Schoenhoff, M. Abstracts of Papers of the American Chemical

Society 2005, 229, U725-U725.

50. Johnson, C. S. Progress in Nuclear Magnetic Resonance Spectroscopy 1999, 34, 203-

256.

51. Lauterbu.Pc. Nature 1973, 242, 190-191.

52. Price, W. S.; Nara, M.; Arata, Y. Biophysical Chemistry 1997, 65, 179-187.

53. Price, W. S. Concepts in Magnetic Resonance 1998, 10, 197-237.

54. Price, W. S. Concepts in Magnetic Resonance 1997, 9, 299-336.

55. Katzenstein, A.; Huber, K. Langmuir 2002, 18, 7049-7056.

56. Ingles, S. E.; Katzenstein, A.; Schlenker, W.; Huber, K. Langmuir 2000, 16, 3010-3018.

57. Sinemus, T., Master Thesis, Universität Paderborn, 2006.

58. Aksnes, D. W.; Gjerdaker, L. Journal of Molecular Structure 1999, 475, 27-34.

59. Aksnes, D. W.; Kimtys, L. Journal of Molecular Structure 1998, 440, 25-33.

60. Bain, A. D. Progress in Nuclear Magnetic Resonance Spectroscopy 2003, 43, 63-103.

61. Chen, A. D.; Johnson, C. S.; Lin, M.; Shapiro, M. J. Journal of the American Chemical

Society 1998, 120, 9094-9095.

References

122

62. Moren, A. K.; Nyden, M.; Soderman, O.; Khan, A. Langmuir 1999, 15, 5480-5488.

63. Cotts, R. M.; Hoch, M. J. R.; Sun, T.; Markert, J. T. Journal of Magnetic Resonance

1989, 83, 252-266.

64. Geil, B. Concepts in Magnetic Resonance 1998, 10, 299-321.

65. Macchioni, A.; Ciancaleoni, G.; Zuccaccia, C.; Zuccaccia, D. Chemical Society

Reviews 2008, 37, 479-489.

66. Martin, N. H.; Allen, N. W.; Moore, J. C. Journal of Molecular Graphics & Modelling

2000, 18, 242-+.

67. Martin, N. H.; Allen, N. W.; Moore, K. D.; Vo, L. Journal of Molecular Structure-

Theochem 1998, 454, 161-166.

68. Davydov, A. S. Theory of Molecular Excitons; Plenum Press: New York, London,

1971.

69. Kasha, M. Radiation Research 1963, 20, 55-&.

70. Korger, M., Heinrich-Heine-Universität Düsseldorf, 2005.

71. Kasha, M.; Rawls, H. R.; El-Bayoumi, M. A. Pure Appl. Chem. 1965, 11, 371-392.

72. Abbott, L. C.; Batchelor, S. N.; Oakes, J.; Gilbert, B. C.; Whitwood, A. C.; Smith, J. R. L.;

Moore, J. N. Journal of Physical Chemistry A 2005, 109, 2894-2905.

73. Abbott, L. C.; Batchelor, S. N.; Oakes, J.; Smith, J. R. L.; Moore, J. N. Journal of Physical

Chemistry B 2004, 108, 13726-13735.

74. Harrison, W. J.; Mateer, D. L.; Tiddy, G. J. T. Faraday Discussions 1996, 139-154.

75. Harrison, W. J.; Mateer, D. L.; Tiddy, G. J. T. Journal of Physical Chemistry 1996, 100,

2310-2321.

76. Herzog, B.; Huber, K.; Stegemeyer, H. Langmuir 2003, 19, 5223-5232.

77. Zhang, Y. Z.; Xiang, J. F.; Tang, Y. L.; Xu, G. Z.; Yan, W. P. Dyes and Pigments 2008,

76, 88-93.

78. Hore, P. J. Nuclear magnetic resonance; Oxford University Press: Oxford [England] ;

New York, 1995.

79. Keeler, J. Understanding NMR spectroscopy; Wiley: Chichester, England ; Hoboken,

NJ, 2005.

80. Cavanagh, J. Protein NMR spectroscopy: principles and practice; Academic Press:

San Diego, 1996.

81. Macomber, R. S. A complete introduction to modern NMR spectroscopy; Wiley: New

York, 1998.

82. Aksnes, D. W.; Kimtys, L. L. Magnetic Resonance in Chemistry 1991, 29, 698-705.

83. Breitmaier, E. Structure elucidation by NMR in organic chemistry: a practical guide,

3rd rev. ed.; Wiley: Chichester, West Sussex, England, 2002.

84. Gogoll, A.; Tanner, D. Magnetic Resonance in Chemistry 1989, 27, 863-871.

85. Hassanzadeh, A.; Zeini-Isfahani, A.; Habibi, M. H.; Heravi, M. R. A. P.; Abdollahi-

Alibeik, M. Spectrochimica Acta Part a-Molecular and Biomolecular Spectroscopy

2006, 63, 247-254.

86. Kleinschmidt, F., Albert-Ludwigs-Universität, 2005.

87. Levitt, M. H. Spin dynamics: basics of nuclear magnetic resonance; John Wiley &

Sons: Chichester ; New York, 2001.

88. Imran, M.; Ahmed, E.; Malik, A. Magnetic Resonance in Chemistry 2007, 45, 785-

788.

89. Pettirossi, S.; Bellachioma, G.; Ciancaleoni, G.; Zuccaccia, C.; Zuccaccia, D.;

Macchioni, A. Chemistry-a European Journal 2009, 15, 5337-5347.

90. Inglesby, M. K.; Zeronian, S. H. Dyes and Pigments 2001, 50, 3-11.

91. Viel, S.; Capitani, D.; Mannina, L.; Segre, A. Biomacromolecules 2003, 4, 1843-1847.

References

123

92. Saveyn, P.; Topgaard, D.; Ghent University, Lund University: Lund, 2006.

93. Lounila, J.; Oikarinen, K.; Ingman, P.; Jokisaari, J. Journal of Magnetic Resonance

Series A 1996, 118, 50-54.

94. Jerschow, A.; Muller, N. Journal of Magnetic Resonance 1997, 125, 372-375.

95. Tait, K. M.; Parkinson, J. A.; Jones, A. C.; Ebenezer, W. J.; Bates, S. P. Chemical Physics

Letters 2003, 374, 372-380.

96. Baldridge, K. K.; Siegel, J. S. Journal of Physical Chemistry A 1999, 103, 4038-4042.

97. Truflandier, L.; Paris, M.; Boucher, F. Physical Review B 2007, 76, -.

98. Kim, D. H.; Eun, H. M.; Choi, H. S. Bulletin of the Korean Chemical Society 2000, 21,

148-150.

99. Gottlieb, H. E.; Kotlyar, V.; Nudelman, A. Journal of Organic Chemistry 1997, 62,

7512-7515.

100. Neumann, B. Dyes and Pigments 2002, 52, 47-53.

101. Neumann, B.; Huber, K.; Pollmann, P. Physical Chemistry Chemical Physics 2000, 2,

3687-3695.

102. Neumann, B.; Pollmann, P. Physical Chemistry Chemical Physics 2001, 3, 4508-4514.

103. Mills, R. J. Phys. Chem. 1973, 77, 685.