Synthesis, Characterization and Properties of Copper Clusters

UNIVERSIDAD DE SANTIAGO DE COMPOSTELA Facultad de Química Synthesis, Characterization and Properties of Copper Clusters Noelia Vilar Vidal PhD Dissertation Santiago de Compostela 2012 Universidad de Santiago de Compostela Facultad de Química Departamento de Química Física Synthesis, Characterization and Properties of Copper Clusters Memoria presentada por Noelia Vilar Vidal para optar al grado de Doctor en Quimica por la Universidad de Santiago de Compostela Santiago de Compostela, Septiembre 2012 D. M. Arturo López Quintela, Catedrático del Departamento de Química Física de la Universidad de Santiago de Compostela y D. José Rivas Rey, Catedrático de Física y director del International Iberian Nanotechnology Laboratory (INL), informan que: La memoria adjunta titulada “Synthesis, Characterization and Properties of Copper Clusters”, presentada por la Licenciada en Química Dª Noelia Vilar Vidal para optar al Grado de Doctor en Química, ha sido realizado bajo su dirección en el Departamento de Química Física de la Universidad de Santiago de Compostela. Y para hacer constancia, firmamos la presente autorización: Fdo. D. M .Arturo López Quintela Fdo. D. José Rivas Rey Santiago de Compostela, Septiembre 2012. A mi familia “Doubt is one of the names of intelligence”. Jorge Luis Borges. I List of abbreviatures: AFM: atomic force microscopy AES: Auger spectroscopy AgNFs: silver nanofibers BE: binding energy CLs: clusters Cu: copper CuNPs: copper nanoparticles CV: cyclic voltammetry e-: electrons EFermi: Fermi energy Egap: energy gap ESI-TOF: electrospray ionization-time of flight EtOH: ethanol FWHM: full width at half maximum H2O: water HOMO: homo highest occupied molecular orbital HPLC: high-performance liquid chromatography HRTEM: high-resolution transmission electron microscopy KE: kinetic energy LDI-TOF: laser desporption ionization-time of flight LUMO: lowest unoccupied molecular orbital MALDI-TOF: matrix assisted laser desorption/ionization mass spectrometrytime of flight MeCN: acetonitrile MS: mass spectrometry N2: nitrogen nm: nanometer ORR: oxygen electro-reduction PA: Auger parameter PL: photoluminescence SPB: surface plasmond band Tª: temperature TBAAcO: tetrabutylammonium acetate TBANO3 tetrabutylammonium nitrate TBABr: tetrabutylammonium bromide TEM: transmission electron microscopy UV-Vis: ultraviolet-visible XPS: X-ray photoelectron spectroscopy λexc: excitation wavelength λemi: emission wavelength III SUMMARY Metal clusters are receiving a great interest nowadays due to their distinctive photoluminescence (PL) and catalytic properties owing their quantum confinement effect. The main purpose of this thesis has been the synthesis of copper clusters as an affordable alternative to the most studied clusters up to now: gold and silver clusters. In this dissertation, firstly, we have explored the synthesis of different copper clusters (CuCLs) sizes and their subsequent purification. Various synthesis and post-treatments parameters were examined in order to improve the synthesis method. The as-synthesized CuCLs were thoroughly characterized by using different techniques, such as: UV-Vis spectroscopy, fluorescence, mass spectrometry, X-ray photoelectron spectroscopy, non-contact atomic force microscopy, transmission electron microscopy…in order to study the cluster size, their morphology and photoluminescent properties. Beside this, the great photoluminescent results obtained for our copper clusters lead us to a deep study on their photoluminescence stability and on their use as potential ion nanosensors. Finally, the catalytic activity of the copper clusters has been also investigated, focusing on their size dependent behavior. CuCLs were found to be useful as efficient and reusable catalyst in the redox reaction of the methylene blue and display good results working as photocatalyst on the silver nanofibers photodissolution reaction. In summary, the simple electrochemical synthesis of copper clusters with excellent photoluminescent and catalytic properties will provide a wide range of potential applications in different fields: as nanosensors, catalyst and photocatalyst. V RESUMEN Los clústeres metálicos despiertan un gran interés hoy en día por sus fantásticas propiedades fluorescentes y catalíticas que surgen debido al confinamiento cuántico. El principal objetivo de esta tesis se ha centrado en la síntesis de clústeres de cobre como una alternativa más asequible a aquellos clústeres más estudiados hasta ahora: los clústeres de oro y plata. En primer lugar en esta tesis se ha desarrollado la síntesis de diferentes tamaños de clústeres de cobre (CuCLs) así como su posterior purificación. Se han examinado diferentes parámetros de síntesis y tratamientos posteriores con el objetivo de mejorar el método de síntesis. Los clústeres de cobre sintetizados fueron caracterizados por medio de diferentes técnicas, tales como: espectroscopía de absorción UV-Vis, fluorescencia, espectrometría de masas, espectroscopía fotoelectrónica de rayos X, microscopía de fuerzas atómicas en modo no contacto, microscopía de transmisión electrónica, etc. con el objetivo de determinar el tamaño del clúster, su morfología y estudiar sus propiedades fluorescentes. Además de esto, los excelentes resutados de fluorescencia dieron lugar a la realización de un estudio detallado en la estabilidad de fluorescencia así como en su posible aplicación como nanosensores de iones. Finalmente, se ha estudiado la actividad catalítica de los clústeres de cobre la cual muestra un comportamiento dependiente del tamaño. Los clústeres de cobre son eficientes y reciclables en la reacción redox del azul de metileno y presentan buenos resultados como fotocatalizadores en la reacción de fotodisolución de nanofibras de plata. En resumen, mediante una síntesis electroquímica simple se obtienen clústeres de cobre con excelentes propiedades fluorescentes y catalíticas que serán muy útiles en un amplio rango de aplicaciones tales como nanosensores, catalizadores y fotocatalizadores. VII OBJECTIVES The main objective of this thesis is the electrochemical synthesis of different sized copper clusters with great photoluminescent properties. Besides, a thorough characterization process of the as-synthesized clusters and the study of different purification methods is needed to include in order to identify the copper cluster nature. Finally, the study of their fluorescent properties and their catalytic activity will be developed. Therefore, this work can be divided into four main parts: A brief summary about metal clusters properties, different synthesis techniques of metal clusters and novel applications will be shown in Chapter 1. The different experimental techniques employed in this work will be summarized on Chapter 2. Chapter 3, 4 and 5 will describe the synthesis, purification and characterization of different copper clusters sizes: small copper clusters (CuCLs) with 5 up to 13 copper atoms, medium CuCLs up to 20 copper atoms and larger CuCLs with 25 copper atoms, respectively. Different characterization techniques such as UV-Vis absorption, fluorescence, mass spectrometry, non-contact atomic force microscopy (NC-AFM) will be employed along the three chapters. The fluorescence stability of the as-synthesized copper clusters (CuCLs) and their potential use as ion recyclable nanosensors will be shown on Chapter 6. Catalytic and photocatalytic studies by using different CuCLs sizes will be described on Chapter 7, concluding on a size dependent catalytic behavior. CHAPTER 1. Introduction Introduction 6 compact atomic shell structure for the metallic core, the satisfying of all such conditions complicates the calculations of the structures enormously. Recently, research on the structure of metal clusters has mainly been focused on alkanethiolate-stabilized gold and silver clusters. However, due to the difficulty in synthesizing homogeneous and high-quality sub-nanometer sized crystals, only limited cluster systems have been investigated. For example: Jadzinsky et al. 15 reported the X-ray structure determination of a Au102(p-MBA)44 (p-MBA, pmercaptobenzoic acid) single crystal. They found an Au49 core surrounded by two groups of 20 Au atoms face camping the Au49 and then 13 Au atoms dispersed without apparent symmetry. Zhu 16 and Heaven 17 proposed a similar Au combination for the Au25(SR)18 cluster: an icosahedral Au13 core plus the exterior 12 Au atoms in the form of six –RS–Au–RS–Au–RS– motifs. Besides this, Zhu et al. found that all the proposed Au25(SR)18 structure was independent of the employed ligands. 1.2.3. Clusters Band-Gap. One of the consequences of the quantum-size regime (with the presence of discrete states in metal clusters) is the appearance of a sizable HOMO-LUMO band gap (similar to that of semiconductors). As can be seen on Figure 1.5, such a semiconductor like behavior is particular significant for smaller clusters with band gaps widely exceeding of 1 eV. Zheng et al. 18 established a correlation of the emission energy with the number of atoms (N) for small gold clusters by observing that the energy emission (Egap) decreases with the increasing number of cluster atoms. They quantitatively fitted this correlation by the expression: Eemission=Egap= EFermi/N1/3, where EFermi is the Fermi energy of bulk gold (5.32 eV), as predicted by the Jellium model6, this being a very good approximation for small metal clusters with N<20. However, for larger clusters, it was observed that a small anharmonic distortion term is required for N≥25. Chapter 1 7 Figure 1. 5| Schematic comparison between band-gaps of some silver, gold and copper clusters (MN, M=metal and N= number of atoms) and those of well-known semiconductors. Bandgaps (Eg) were calculated from the spherical Jellium model (Eg = EF/N1/3 ; EF = Fermi level) and the position of the conduction band, ECB, was estimated by the formula ECB =  – EF – ½ Eg (  = electronegativity). Figure 1.5 shows characteristics band gaps of different metal clusters. It is known that, in general, bandgap decreases when the cluster size increases, having a nice agreement between the DFT calculated bandgap values and those predicted by the simple Jellium model, and also that ligands also play a minor role. Table 1.1 shows the change of the band gap with the cluster size and the ligands including the band gap calculated by Jellium for comparison (clusters sizes with N   25 atoms a correction of -0.4eV was used because of the anharmonicity observed in large clusters).18 Introduction 8 Table 1. 1| Comparison (adapted from reference 14) between: (1) experimentally determined band gaps for free gas-phase gold cluster anions from photoelectron spectroscopy, (2) theoretical (Density Functional Theory) values for HOMO-LUMO gaps of passivated gold cluster compounds that correspond to 8, 34, and 58 conduction-electron shell closings, and (3) band gaps calculated through the Jellium model for clusters with the same number of gold atoms; *for these clusters a correction of -0.4eV was used. (1) Experiment (2) Theory (3) Jellium model Shell closing Cluster Gap/eV Cluster compound Gap/eV Eg=5.32/N1/3 /eV 8e (1S21P6) Au11(PH3)7(SMe)3 1.5 2.4 8e Au11(PH3)7Cl3 2.1 2.4 8e Au13(PH3)10Cl23+ 1.8 2.3 8e Au25(SMe)181.2 1.3* 34e(8e + 1D102S21F14) Au341.0 Au39Cl6(PH3)140.8 1.0* 58e (34e + 2P61G18) Au580.6 Au102(p-MBA)44 0.5 0.6* 58e Au102(SMe)44 0.5 0.6* Chapter 1 9 1.3. UNIQUE PROPERTIES OF METAL CLUSTERS. 1.3.1. Optical properties. Significantly different from large nanoparticles, optical absorption spectra of clusters exhibit molecular like optical transitions (between the discrete energy levels corresponding to the last occupied orbital and the first unoccupied orbitals of the delocalized conduction electrons). 19 , 20 Bands at lower energies will appear with the increase of the cluster size. As it can be seen in Figure 1.6.A Cu nanoparticles show SPB around 580nm (interband transition 5d106sp1) 21 according to the literature 22 and copper clusters show well-defined absorption bands indicating a molecular-like behavior (Figure 1.6.B) as a consequence of the energy level space increase and to their band structure separated at discrete levels. 23 The multiple absorptions could be attributed to the intraband (sp) HOMO-LUMO transmission, interband transition (d-sp), or mixed sp-sp intraband and d-sp interband transitions. Figure 1.6| UV-Vis absorption spectra of CuNPs displaying the typical plasmon band at  567nm (A) and large CuCLs with well-defined molecular-like absorption bands (B). 1.3.2. Photoluminescence properties. Luminescence is one of the major properties of metal clusters. Fluorescence of Au, Ag and Cu clusters has been extensively studied during the past decade. Several works have demonstrated that the photoluminescence could be assigned to the electronic transitions between the highest occupied orbital and A) B) 200 300 400 500 600 700 800 0.03 0.04 0.05 0.06 Copper Nanoparticles 567 nm Absorbance/AU Wavelength/nm 200 300 400 500 600 700 800 0.0 0.3 0.6 Copper Clusters Absorbance/AU Wavelength/nm Introduction 10 the lowest unoccupied orbital (HOMO-LUMO). 24 Photoluminescence was first observed in noble metal by Mooradian 25 , who observed visible emission from copper and gold films with a quantum efficiency of ~10-10. Size-dependent fluorescence properties of Au quantum dots were studied by the Dickson group.18 They found that excitation and emission bands shift to smaller wavelengths with decreasing cluster size. The excitation and emission spectra of Au clusters composed of a different number of atoms is showed in Figure 1.7. The synthesized Au5, Au8, Au13, Au23 and Au31 clusters show UV (385 nm), blue (455 nm), green (510 nm), red (760 nm) and near IR (866 nm) fluorescence, respectively. One can see that the excitation and emission bands shift to longer wavelengths (low energy) with increasing cluster size. Fluorescence properties of metal clusters are very sensitive to their chemical environments, including the cluster size, solvent and surface protecting ligands. 26 , 27 By increasing the charge transfer from the surface ligands to the metal core an enhancement of the fluorescence was observed.27 This enhancement can be achieved by (1) increasing the electron donation capability of the ligands, (2) improving the electropositivity of the metal core and (3) using the protecting ligands with electron-rich atoms and groups. Figure 1. 7| Excitation (dashed) and emission (solid) spectra of different gold clusters. Excitation and emission maxima shift to longer wavelengths with increasing the cluster size from Au5 up to Au31 (a). Emission from the three shortest wavelengths emitting gold cluster solutions (from left to right) under longwavelength UV lamp irradiation (366nm). (b) Figure extracted from reference 18. The discrete emission of these clusters open up new opportunities, which can be used as biological labels or light-emitting sources in nanoelectronics as an alternative to semiconductor quantum dots especially due to their biocompatibility and their reduced photobleaching. Chapter 1 11 1.3.3. Catalytic properties. Metals including gold, silver and copper, are usually quite inert and show little activity for reactive molecules adsorption, but, when their dimensions are diminished to the nanoscale, the properties of the materials exhibit a dramatic deviation from those of the bulk. For instance, gold clusters were found to be reactive at room temperature with oxygen and other molecules. Quantum chemical calculations indicate that such high reactivity is due to undercoordination of the metal atoms forming the cluster. 28 In recent years, Au clusters have been extensively examined as active catalysts for CO oxidation and oxygen electro-reduction. 29 Although this is, a field not explored much so far, these electrocatalytic properties make metal clusters promising materials in fuelcell applications. Oxygen electro-reduction (ORR) represents a critical cathodic reaction in fuel cells. Despite extensive research progress, wide-spread commercialization of fuel cells has been hindered partly because of the sluggish reaction dynamics for the ORR at the cathode and the high cost of Ptbased electrocatalyst. Therefore, numerous studies have focused on non-Pt catalyst like for example Au clusters. 29b, 30 Ag 31 and Cu 32 clusters.The results showed that the activity of clusters against ORR increases as the cluster core size decreases. Introduction 12 1.4. SYNTHESIS OF METAL CLUSTERS. Due to the highly necessary control of experimental conditions and a suitable purification and isolation method for the synthesis of metal clusters, more accurate synthesis techniques will be required than for nanoparticles. Metal clusters can generally be prepared through “bottom-up” or “top-down” synthesis. The “bottom-up’’ approach involves metal precursors which are reduced to atoms with a reducing reagent forming the metal clusters by the nucleation of the zerovalent metal atoms. The ‘‘top-down’’ methods consist of a ligand etching process from nanoparticles to clusters. Here we will summarize some different synthesis techniques of metal clusters such as: electrochemical synthesis, photoreduction, microemulsion technique, chemical reduction, templating techniques and etching methods. A) Chemical reduction by modified Brust-Schiffrin method: Here, metal precursors are first dissolved in an aqueous solution, subsequently, organic protecting ligands and reducing reagents are added into the solution to generate metal clusters and then they are transferred to an organic solvent by phase-transferring reagents such as tetraoctylammonium bromide. The core size and the surface properties of the metal clusters can be controlled effectively by adjusting the experimental parameters, such as the metal to ligand ratio, chemical structure of the protecting ligands, the nature of the reducing agent, reaction temperature and time, pH of the solution, etc. By using this method different metal clusters such as Au 33 , Pt 34 , Ag 35 and Cu 36 clusters have been successfully synthesized. B) Template-based synthesis.- Template-based synthesis provides a predetermined environment for the clusters formation, which is favorable to produce clusters with well-controlled size and shape. Over the past few decades, template-based methods have proved to be efficient synthetic techniques for preparing fluorescent metal clusters, using polymers, 37 proteins, 38 , 39 , 40 , 41 dendrimers 42 , 43 or even DNA 44 , 45 , 46 as templates. Copper clusters have been synthesized by a modified poly(amidoamine) (PAMAM) dendrimer as a template. 47 C) Ligand induced etching of metal nanoparticles.- The etching capability of some ligands (like thiols) is used to synthesized clusters by removing the surface atoms from metal nanoparticles leading to stable quantum clusters (shell-closing magic numbers). Chapter 1 13 Scheme 1. 1| Schematic representation of two possible routes for the formation of gold clusters by etching of mercaptosuccinic acid-capped gold nanoparticles. Figure extracted from reference 48. Examples of this technique are the synthesis of AuCLs (Au8 or Au25) from mercaptosuccinic acid (MSA)-protected Au nanoparticles (4-5nm core diameter) by etching with excess glutathione varying the etching pH (~7-8 to Au8 and ~3 for Au25). 48 Although the detailed mechanisms of the ligandand precursorinduced nanocrystal etching have not yet been specified clearly (Scheme 1.1) this new synthetic strategy holds promise for preparing new types of sub-nanometer sized clusters containing only a few metal atoms. D) Electrochemical synthesis.- Electrochemical synthesis, first developed by Reetz 49 in 1994, is a very promising technique to produce metal clusters because of their simplicity. Compared to chemical reduction, the electrochemical method exhibits many advantages in the preparation of metal clusters, such as the low reaction temperature, large-scale yield, low cost of the initial materials, and easy manipulation of cluster size by tuning current, voltage, electrolyte, concentration of stabilizers etc. For instance, Reetz et al. prepared 1–5 nm palladium nanoparticles just by changing the current density. 50 Based on this mechanism, this thesis will be focused on the use of this synthetic method in order to prepare photoluminescent copper clusters. E) Microemulsion method.- Nanodroplets of water dispersed in oil using surfactants and/or amphiphilic blockcopolymers will be used as nanoreactors in order to obtain clusters with well-defined sizes. The cluster size can be tuned by Introduction 14 adjusting the liquid core dimensions of microemulsions. Silver clusters AgN (N ≤ 10) were prepared in this way by Ledo et al. 51 achieving novel photoluminescent and magnetic properties. More recently, small copper clusters as CuN, with N≤13, showing UV photoluminescent properties have also been synthesized by this method. 52 As shown in Scheme 1.2, the core size of the copper clusters can be easily controlled by varying the amount of the reducing agent (NaBH4). Scheme 1. 2| Schematic picture of the evolution of the copper cluster size dependent on α (being α the used moles of NaBH4 needed for the stoichiometric reduction of Cu(II) ions). Estimated wavelengths of the plasmon band for different sized copper clusters are shown at the bottom. Imagen extracted from reference 50. F) Photoreduction synthesis.- Photoreduction is a simple method where metal salts are reduced by irradiating the system by UV-light in presence of some capping agent. Since the first report of size-dependent fluorescent Ag and Au clusters prepared by Zheng and co-workers 53 photoreduction synthesis of metal clusters has received more and more attention and copper clusters are being to be studied. 54 Copper clusters are being more studied nowadays, but their recent discovering has given us few references in the literature. Therefore, we summarize in the Table 1.2 the different types of CuCLs synthesis employed until now, the obtained cluster size and their PL range. Chapter 1 15 Table 1.2| Literature summary of CuCLs synthesis. Synthesis Reference Cluster size/N (CuN) PL (λemi /nm) QY % Dendrimer Template 43 - - - 47 Cu4-Cu65 - - Microemulsions 52 CuN N≤13 330nm - Modified Brust 36 CuN N≤8 425-593nm 4.4 Polyol 55 Cu9 460nm 0.65 BSA protein 41 Cu13 410nm 0.15 DNA templated 46 - 593nm Photoreduction 54 Cu5 600nm 2.2 Introduction 22 45 Petty J.T.; Zheng J.; Hud N.V.; Dickson R.M. J. Am. Chem. Soc. 126, 5207 (2004). 46 Jia X.; Li J.; Han L.; Ren J.; Yang X.; Wang E. ACS Nano 6 , 3311 (2012). 47 Balogh L.; Tomalia D.A. J. Am. Chem. Soc. 120, 7355 (1998). 48 Muhammed M.H.; Ramesh S.; Sinha S.; Pal S.; Pradeep T. Nano Res. 1, 333 (2008). 49 Reetz M. T.; Helbig W. J. Am. Chem. Soc. 116, 7401 (1994). 50 Reetz M.T.; Winter M.; Breinbauer R.; Thurn-Albrecht T.; Vogel W. Chem. Eur. J. 7, 1084 (2001). 51 Ledo-Suárez A.; Rivas J; Rodríguez-Abreu C. F.; Rodríguez M. J.; Pastor E.; Hernández-Creus A.; Oseroff S. B.; López-Quintela M. A. Angew. Chem., Int. Ed. 46, 8823 (2007). 52 Vázquez-Vázquez C.; Bañobre-López M.; Mitra A.; López-Quintela M.A., Rivas J. Langmuir 25, 8208 (2009). 53 a) Zheng J.; Petty J. T.; Dickson R. M. J. Am. Chem. Soc. 124, 13982 (2002). b) Zheng J.; Petty J. T.; Dickson R. M. J. Am. Chem. Soc. 125, 7780 (2003). 54 Zhang H.; Huang X.; Li L.; Zhang G.; Hussain I.; Li Z.; Tan B. Chem. Commun. 48, 567 (2012). 55 Kawasaki H.; Hamaguchi K.; Osaka I.; Arakawa R. Adv. Funct. Mater. 21, 3508 (2011). 56 Bootharaju M. S.; Pradeep T. Langmuir 27, 8134 (2011). 57 Huang C. C.; Yang Z.; Lee K. H; Chang H. T. Angew. Chem., Int. Ed. 46, 6824 (2007). 58 Xie J.; Zheng Y.; Ying J.Y. Chem. Commun. 46, 961 (2010). 59 Adhikari B.; Banerjee A. Chem. Mater. 22, 4364 (2010). 60 Shang L.; Dong S. Biosens. Bioelectron. 24, 1569 (2009). Chapter 1 23 61 Shiang Y.-C.; Huang C.-C.; Chang H.-T. Chem. Commun. 23, 3437 (2009). 62 Liu Y. L.; Ai K. L.; Cheng X. L.; Huo L. H.; Lu L. H. Adv. Funct. Mater. 20, 951 (2010). 63 Jao Y.-C.; Chen M.-K.; Lin S.-Y. Chem. Commun. 46, 2626 (2010). 64 Wu X.; He X.; Wang K.; Xie C.; Zhou B.; Qing Z. Nanoscale 2, 2244 (2010). 65 Harding C.; Habibpour V.; Kunz S.; Farnbacher A.N-S.; Heiz U.; Yoon B.; Landman U. J. Am. Chem. Soc 131, 538 (2009). 66 Vajda S.; Pellin M.J.; Greeley J.P.; Marshall C.L.; Curtiss L.A.; Ballentine G.A.; Elam J.W.; Catillon-Mucherie S.; Redfern P.C.; Mehmood F.; Zapol P. Nature Mater. 8, 213 (2009). 67 Lee S.; Molina L.M.; López M.J.; Alonso J.A.; Hammer B.; Lee B.; Seifert S.; Winans R.E.; Elam J.W.; Pellin M.J.; Vajda S. Angew. Chem. Int. Ed. 48, 1 (2009). 68 Selva J.; Martínez S.E.; Buceta D.; Rodríguez-Vázquez M.J.; Blanco M.C.; López-Quintela M.A.; Egea G. J. Am. Chem. Soc. 132, 6947 (2010). CHAPTER 2. Experimental Section Chapter 2 27 2.1. INTRODUCTION. This chapter is focused on the description of the synthesis technique and the posterior characterization techniques of the as-synthesized copper clusters during my research. First of all, characteristics of the reactants will be explained. Secondly, details of the general synthetic procedure used throughout this thesis will be given and finally experimental characterization methods will be briefly summarized. Experimental Section 28 2.2. REACTANTS. The different used reactants have been summarized in the Table 2.1. All the reactants were used without further purification. Table 2. 1| List of the used reactants. Name Chemical Formula Purity (%) Brand Copper electrode 99.99 GoodFellow Platinum electrode 99.95 GoodFellow Aluminium oxide Al2O3 99.99 Alfa Aesar Tetrabutylammonium nitrate (C4H9)4NNO3 >97 Fluka Tetrabutylammonium acetate (C4H9)4NAcO 99 Fluka Tetrabutylammonium Bromide (C4H9)4NBr 99 Fluka Acetone C3H6O 100 Prolabo Acetonitrile C2H3N 99.9 Sigma-Aldrich Sulfuric Acid H2SO4 97 Merck Pentane C5H12 99 Sigma-Aldrich Chloroform CHCl3 99 GPR-Rectapur Ethanol EtOH 99.9 GPR-Rectapur Copper nitrate Cu(NO3)2 98 Aldrich Copper acetate CuOAc 98 Aldrich Toluene C7H8 99.8 Aldrich 1-hexanethiol C7H14S 95 Sigma-Aldrich Hydrazine hydrate N2H4.H2O 98 Aldrich Sodium hydroxide NaOH Prolabo Dodecanethiol C12H26S 98 Aldrich Lead nitrate Pb(NO3)2 99 Sigma-Aldrich Aluminium nitrate Al(NO3)3.9H2O 98 Aldrich Potasium hydroxide KOH Sigma-Aldrich Cobalt Sulphate Co(SO4).7H2O 99 Sigma Ferric nitrate Fe(NO3)3.9H2O 98 Panreac Nickel nitrate Ni(NO3)2.6H2O Aldrich Zinc sulphate Zn(SO4).7H2O 99 Aldrich Sodium Nitrate Na(NO3) 99 Sigma-Aldrich Methylene Blue C16H18ClN3S · 3H2O 82 Sigma-Aldrich Heptane C7H16 Sigma-Aldrich Chapter 2 29 2.3. ELECTROCHEMICAL SYNTHESIS AND PURIFICATION OF COPPER CLUSTERS. 2.3.1. Synthesis Procedure. As we briefly explained in Chapter 1, electrochemical synthesis was first introduced by Reetz in 1994 1 and has been used by our group for many years in order to synthesize nanoparticles and clusters. 2 This synthesis displays different points: (1) cluster size can be controlled by tuning the current density (high current densities give smaller particles). This relationship 3 between the current density and the particle size from the free energy formation of a metal cluster, ΔG(N), has two terms, as follows: ΔG(N)=-ze|η|+Φ(N) [1] where N is the number of ions in the cluster, η is the overpotential, z is the ion charge, e is the charge of the electron, and Φ isthe electrostatic potential. The first term is related to the number, N, of metal ions that move from the solution to the crystal phase on the surface of the cathode and the second term is related to the increase in the surface energy associated with the creation of the metal cluster surface. This increase in the surface energy equals the difference between the binding energies of the N bulk metal ions and those of the N metal ions arranged on the surface of the metal crystal. Both terms in Eq. 1 are functions of the size of the metal cluster, N. The size of the critical nucleus in two dimensions is given by: Nc= bsε2/(zeN)2 [2] where Nc is the number of atoms in the cluster, b is the factor that relates the surface area S of the nucleus to its perimeter P (b= P2/4S; b = Π for a circular nucleus), s is the area occupied by one atom on the surface of the metal nucleus, and ε is the edge energy. Therefore, Nc strongly depends on the overpotential and is inversely proportional to η2. The critical radius of the surface nucleus rc is a function of the overpotential: rc=sε/ze η [3] Experimental Section 30 Therefore, rc is inversely proportional to the overpotential (η), and the electron transfer overpotential may be defined as the rate of change of electrode potential with the current associated with the limiting rate of electron transfer across the phase boundary between the electrically conducting electrode and the ionic conducting solution. The overpotential is directly related to the current density. (2) the isolation of the clusters is simple—they precipitate out of the solution when formed; (3) tetralkyl ammonium salts were used as supporting electrolyte and also as a capping agent, therefore, the R4 N+Xstabilized the clusters (4) the reaction proceeds with high yields (>95%) and can be performed on a moderate scale. The cluster formation mechanism consists of (1) dissolution of the sacrificial anode (Cu0  Cu2+), (2) transporting of the electroactive species (metal cations) to the cathode, (3) reductive metal adatom formation at the cathode interface, (4) aggregation of the metal adatoms, (5) stabilization by ‘self organization’ of the ammonium ions around the metal core (6) reaction product desorption from the cathode, (7) reaction product diffusion to the solution. A brief scheme of this mechanism can be observed in Scheme 2.1. Scheme 2. 1| Electrochemical mechanism formation of CuCLs stabilized by tetrabutylammonium salts. Chapter 2 31 2.3.2. Clusters stabilization. Metal clusters must be stabilized against the aggregation into larger particles. The stabilization of metal clusters can be classified as follows: (1) Electrostatic stabilization: results from the adsorption of ions to the electrophilic metal surface creating a electrical double layer, which results in a coulombic repulsion force between particles (Figure 2.1.A). 4 (2) Steric stabilization: achieved by surrounding the metal center by large organic molecules (such as poly (N-vinyl pyrrolidone) (PVP), which prevent close contact of the metal particle centers (Figure 2.1.B). (3) By ligands (P, N, S donors): phospines, thiols, etc have been very exploited throughout the years as suitable cluster ligands for the “metallic full shell clusters”. (4) Electrosteric stabilization: strong coordination of bulky molecules such as surfactants at the surface of the particles, which is achieved well using tetralkyl salts, where the negatively charged anion is binded to the metal surface and the alkyl chains shield the metallic core like an umbrella (Figure 2.1.C). (5) Solvent stabilization: solvents as tetrahydrofurane (THF) or THF/MeOH can act as stabilizers. 5 Figure 2. 1| Schematic image of different stabilized nanoparticles. A) Ions adsorbed onto the particle surface, creating an electrical double layer which provides Coulombic repulsion and thus stabilization against aggregation. B) Two polymer-protected particles interacting. The region between the two particles becomes crowded as a high local concentration of polymer builds up. C) Electrosteric stabilization of a Pd particle by tetra(octyl) ammonium stabilizer.Imagen taken from reference 4. A) B) C) Experimental Section 38 2.4.4. Transmission Electron Microscopy (TEM and HRTEM). Transmission electron microscopy (TEM) has traditionally been used to obtain information about the nanoparticles size and shape. Although it should be noted that there is around a 0.2nm size uncertainty for TEM measurements and melting effects can occur under electron beam heating during the imaging. 27 Due to this, the size of sub-nanometer clusters cannot be obtained accurately from TEM measurements, but it can be useful for rejecting the presence of large nanoparticles. On the other hand, High resolution transmission electron microscopy (HRTEM) can provide us structural information (with a highest resolution of ~0.1 nm) in despite of more pronounced heating effects can occur during the measurement. Low-resolution TEM images of our CuCLs were obtained on a Philips CM20 electron microscope operating at an acceleration voltage of 120 kV, whereas HRTEM images were performed on a JEOL JEM2010F electron microscope operating with an acceleration voltage of 200 kV. Samples were prepared on carbon-coated TEM grids by adding a drop of a diluted solution of CuCLs, onto the grid and evaporating the solvent in roomconditions. The measured lattice separations were compared against the standards for Cu, Cu2O, and CuO (Table 2.2). Table 2. 2| Theoretical Lattice separations of the different Cu compounds. Compound Cu Cu2O CuO Lattice separations/Å 1.81 3.02 2.51 <200> <110> <002> 2.4.5. X-ray photoelectron spectroscopy (XPS). Photoelectron spectroscopy is a powerful tool when investigating the composition and electronic state of metal cluster cores. The basis of XPS rely on the foton study by calculating the BE (atom binding energy or ionization energy for a certain level) of the inner atomic electron shells, which depends mainly of the atomic number. Atoms in a higher positive oxidation state will show higher binding energies due to the extra columbic interactions between the photoemitted electron and the ion core. This ability to discriminate different oxidation states and chemical environments is one of the major strengths of XPS. Here, the average oxidation states of copper atoms in our samples were recorded using a VG Escalab 250iXL spectrometer (VG Scientific) equipped with a Chapter 2 39 hemispherical analyzer and Al KR X-ray monochromated source. An X-ray spot of 500 μm was used to generate photoelectrons, which were collected from a take-off angle of 90°. The argon partial pressure in the analysis chamber was maintained at 3 × 10-8 mbar during data acquisition by turbomolecular differential pumping. The measurement was performed in a constant analyzer energy mode (CAE) with 100 eV pass energy for survey spectra and 20 eV pass energy for the high resolution spectra. The intensities were estimated by calculating the area under each peak after subtraction of the S-shaped background. Binding energies (BEs) of Cu 2p could be determined by referencing to the adventitious C 1s peak at 285.0 eV. Atomic ratios were computed from peak intensity ratios and Schofield atomic sensitivity factors. Samples for XPS analysis were prepared, leaving a drop of solution containing CuCLs on a piece of mirror finish polished silicon wafer, evaporating the solvent quickly, and immediately introducing the sample into the XPS prechamber under high vacuum conditions. Ar+ ions (1.5 keV, 60 s) were employed to remove adventitious carbon, which comes from the surface sample exposure to the atmosphere, and to remove also possible oxides at the sample surface. The amount of material removed is a function of the incident energy and sputtering time. Samples were kept rotating in order to avoid a shadow effect. The efficiency of the sputtering process was checked by the disappearance of the Si oxide native layer from the silicon wafer substrate. Bombardment was performed using an EXO5 ion gun incorporated into the equipment, provided with a scanning unit to raster the ion beam, operating at a voltage of 1.5 kV and a scan size of 2 mm, producing a sample current of 0.3 μA for 60 s of bombardment. The higher oxidation state of copper (II) can be identified by a shakeup satellite at ∼10 eV from the Cu 2p3/2 main peak. This satellite peak is not present for other chemical species of copper, such as Cu(0) or Cu(I). Cu 2p3/2 binding energy and an Auger parameter in the copper clusters samples were investigated. Experimental Section 40 Figure 2. 4| XPS database reference for Cu2p region of the different oxidation states of Cu. 2.4.6. Cyclic Voltammetry (CV). Cyclic voltammetry consists of cycling the potential of an electrode, which is immersed in an unstirred solution, and measuring the resulting current at the “working electrode” during the potential scan. The potential of the “working electrode” is controlled versus a reference electrode (Ag/AgCl). The important parameters of a cyclic voltammogram are the magnitudes of the anodic and cathodic peaks current and their position. CV´s of different CuCLs sizes were performed by David Buceta Fernández. The experimental conditions consist on the deposition of the CuCLs solution onto glassy carbon (GC) electrodes and covered with one monolayer of Nafion (5µL of a 0.025% solution). The electrode surface was cleaned before each deposition by abrasion by polishing with alumina 0.5 µm average particle size. The final cleaning of the GC electrodes was performed in water in an ultrasonic bath. The counter electrode was a platinum wire while an Ag/AgCl electrode was used as the reference. The CuCLs were examined in 0.1M KOH by cyclic voltammetry (at sweep rate of 10mV/s) in the potential range between -0.5 V and +1.3V vs Ag/AgCl in a deaerated electrolyte solution under nitrogen at room temperature. Chapter 2 41 2.4.7. Atomic Force Microscope (AFM-NC). An atomic force microscope can be used to investigate any surface, even poorly or non-conducting ones. The interaction force that dominates NC-AFM imaging is the short-range interaction force of the chemical bonding between the atoms of the AFM tip and the sample surface. 28 The resolution obtained by AFM is determined in large part by the size of the probe tip used for imaging. Attractive Van der Waals forces acting between the tip and the sample are detected, the tip scans above the surface and topographic images are created. The width of the particles depends on the probe shape; however, the particle height is independent of it. This fact is not a function of particle size but rather an intrinsic property of AFM. It is important to know if the probe is much smaller than the feature to be imaged; then the probe-generated artifacts will be minimal and the dimensional measurements will be accurate. Here AFM tip Single Crystal Silicon, N-type, 0.01 -0.025Ω-cm, Antimony doped, was used. CuCLs samples were prepared by drying 5 µL of copper clusters on a mica substrate and using a XE-100 instrument (Park Systems Corporation) in noncontact mode. Figure 2. 5| NC-AFM topography image of mica substrate with their corresponding cross-section. Experimental Section 42 REFERENCES. 1 Reetz M.T.; Helbig W. J. Am. Chem. Soc. 116, 1401 (1994). 2 a) Rodríguez-Sánchez M.L.; Rodríguez M.J.; Blanco M.C.; Rivas J.; LópezQuintela M.A. J. Phys. Chem. B 109, 1183 (2005). b) Rodríguez-Sánchez M.L.; Blanco M.C.; López-Quintela M.A. J. Phys. Chem. B 104, 9683 (2000). c) Ledo A.; Martínez F.; López-Quintela M.A.; Rivas J.; Physica B 398, 273 (2007). d) LedoSuárez A.; Rivas J.; Rodríguez-Abreu C.F.; Rodríguez M.J.; Pastor E.; HernándezCreus A.; Oseroff S.B.; López-Quintela M.A. Angew. Chem. Int. Ed. 46, 8823 (2007). e) Rodríguez-Vázquez M.J.; Blanco M.C.; Lourido R.; Vázquez-Vázquez C.; Pastor E.; Planes G.A.; Rivas J.; López-Quintela M.A. Langmuir 24, 12690 (2008). f) Ledo-Suárez A.; Rodríguez-Sánchez L.; Blanco M. C.; López-Quintela M. A. Phys. stat. sol. (a) 203, 1234 (2006). 3 Reetz M.T. Active Metals, VCH, Weinheim, A. Fürstner Ed., 279 (1996). 4 Aiken III J.D.; Finke R.G. Journal of Molecular Catalysis A: Chemical 145, 1 (1999). 5 Franke R.; Rothe J.; Pollmann J.; Hormes J.; Bönnemann H.; Brijoux W.; Hindenburg T. J. Am. Chem. Soc. 118, 12090 (1996). 6 Mayell J.S.; Bard A.J. J. Am. Chem. Soc. 85, 421 (1963). 7 Littlehailes J.D.; Woodhall B.J. Discuss. Faraday Soc., 45, 187 (1968). 8 Rodríguez-Sánchez L.; Blanco M. C., López-Quintela M. A. J. Phys. Chem. B, 104, 9683 (2000). 9 Fry A.J.; Krieger R.L. J. Org. Chem. 41, 54 (1976). 10 Santiago González B.; Rodríguez M.J.; Blanco C.; Rivas J.; López-Quintela M.A.; Gaspar M. J.M. Nano Lett. 10, 4217 (2010). 11 Vilar-Vidal, N. Master Thesis, “Synthesis, stabilization and characterization of atomic metal clusters”, University of Santiago de Compostela (2007). 12 Price R. C.; Whetten R. L. J. Am. Chem. Soc. 127, 13750 (2005). 13 Whetten R. L.; Khoury J. T.; Alvarez M. M.; Murthy S.; Vezmar I.; Wang Z. L.; Stephens P. W.; Cleveland C. L.; Luedtke W. D.; Landmann U. Adv. Mater. 8, 428 (1996). Chapter 2 43 14 (a) Siebrands T.; Giersig M.; Mulvaney P.; Fischer C.-H. Langmuir 9, 2297 (1993). b) Wilcoxon J. P.; Martin J. E.; Provencio P. Langmuir 16, 9912 (2000). c)Gautier C.; Taras R.; Gladiali S.; Beurgi T. Chirality 20, 483 (2008). d) Frein S.; Boudon J.; Vonlanthen M.; Scharf T.; Barber J.; Seuss-Fink G.; Beurgi T.; Dechenaux R. Helv. Chim. Acta 91, 2321 (2008). 15 (a) Jiménez V. L.; Leopold M. C.; Mazzitelli C.; Jorgenson J. W.; Murray R. W. Anal. Chem. 75, 199 (2003). (b) Song Y.; Jiménez V.; McKinney C.; Donckers R.; Murray R. W. Anal. Chem. 75, 5088 (2003). (c) Choi M. M. F.; Douglas A. D.; Murray R. W. Anal. Chem. 78, 2779 (2006). (d) Zhang Y.; Shuang S.; Dong C.; Lo C. K.; Paau M. C.; Choi M. M. F. Anal. Chem. 81, 1676 (2009). 16 Schaaff T. G.; Knight G.; Shafigullin M. N.; Borkman R. F.; Whetten R. L. J. Phys. Chem. B 102, 10643 (1998). 17 Kreibig U.; Zacharias P. Z. Phys A-Hadron Nucl. 231, 128 (1970). 18 Lisiecki I.; Pileni M.B. J. Am. Chem. Soc. 115, 3887 (1993). 19 a) Hussain I.; Graham S.; Wang Z.; Tan B.; Sherrington D. C.; Rannard S. P.; Cooper A. I.; Brust M. J. Am. Chem. Soc. 127, 16398 (2005). b) Robinson III, D.A., PhD Dissertation, “Synthesis and characterization of metal nanoclusters stabilizaed by dithiolates”. Georgia State University (2011). 20 Cottancinm E.; Celep G.; Lemé J.; Pellarin M.; Huntzinger J.R.; Vialle J.L.; Broyer M. Theor. Chem. Acc. 116, 514 (2006). 21 a) Kelly K. L.; Coronado E.; Zhao L. L.; Schatz G. C. J. Phys. Chem. B 107, 668 (2003). b) Rao T. U. B.; Pradeep T. Angew. Chem. Int. Ed. 49, 3925 (2010). 22 Whetten R. L.; Khoury J. T.; Alvarez M. M.; Murthy S.; Vezmar I.; Wang Z. L.; Stephens P. W.; Cleveland C. L.; Luedtke W. D.; Landman U. Adv. Mater. 8, 428 (1996). 23 Waters C. A.; Mills A. J.; Johnson K. A.; Schiffrin D. J. Chem. Commun. 4, 540 (2003). 24 Dass A.; Stevenson A.; Dubay G. R.; Tracy J. B.; Murray R. W. J. Am. Chem. Soc. 130, 5940 (2008). 25 Arnold R. J.; Reilly J. P. J. Am. Chem. Soc. 120, 1528 (1998). Experimental Section 44 26 Schaaff T. G. Anal. Chem. 76, 6187 (2004). 27 a) Zhang W.J.; Miser D.E. Journal of Nanoparticle Research 8, 1027 (2006). b) Ramasamy P.; Guha S.; Shibu E. S.; Sreeprasad T. S.; Bag S.; Banerjee A.; Pradeep T. J. Mater. Chem. 19, 8456 (2009). c) Rao T. U. B.; Pradeep T. Angew. Chem. 122, 4017(2010) Angew. Chem. Int. Ed. 49, 3925 (2010). 28 Perez R.; Payne M. C.; Stich I.; Terakura K. Phys. Rev. Lett. 78, 678 (1997). CHAPTER 3. Synthesis of Small Copper Clusters Chapter 3 47 3.1. INTRODUCTION. Fluorescent metal clusters have been generating significant interest nowadays; their syntheses have been optimized and their properties studied mostly due to their possible application as proteins or cellular labels (in biological imaging applications). Ideally, for biomedical applications, non-toxic aqueoussoluble clusters with a strong emission and good photostability are desirable. Besides, small metal clusters consisting of only in a few atoms are of considerable interest due to their greater stability. Several approaches have been reported for the synthesis of metal clusters in aqueous solution, 1 being Cu clusters less studied. In this research work, we first developed a new aqueous electrochemical synthesis of small copper clusters. The optimization of the synthesis has been studied by studying the different parameters which had an influence on the product. Purification of the obtained CuCLs by centrifugation was studied resulting in a control-size selective method. CuCLs should be optical, structural and morphological characterized by using UVVis, TEM/HRTEM, LDI-TOF and ESI-TOF, NC-AFM, Fluorescence, CV, XPS. Synthesis of Small Copper Clusters 54 3.3.2. Kinetic evolution during the synthesis. UV-Vis absortption spectrum (Figure 3.4.A) changes in the course of the synthesis at different synthesis times.The UV-Vis absorption at the beginning (t=40s) shows the first absorption band centered at 250nm. When the time is around 100s another two bands centered at 275 and 300nm appears, which will increase their intensity with the time. Surface plasmon band of copper CuSPB ≈560-570nm 3 is not observed nor 800 nm absorption band attributed to the aqueous Cu cations [Cu(H2O)6]+2 d-d transition in a tetragonally distorted octahedral or square-planar ligand field. The band at 250 nm was considered by Zhao et al. 4 as a ligand-to-metal charge transfer transition. But, studies by Ozin et al. 5 revealed CuCLs absorption at 247-256 nm (in solid Xenon) related with a Cu2 cluster and 272 nm with a Cu3. Therefore, these bands (250 and 275 nm) can be associated to the first “stable” copper clusters. De Waele et al. 6 and VazquezVazquez et al. 7 obtained similar results by radiolytic and microemulsions synthesis showing absorptions at 264 nm and 250 nm respectively on the earliest stages of the synthesis. Figure 3. 4| Absorbance versus time during the electrochemical synthesis of small copper clusters E31. Time between successive spectra is Δt =20 s.(A) Absorbance versus time obtained by fitting to Lorentzians for the peaks centered at 257, 275,302 and 360 nm during the synthesis.(B) In addition, for a more systematic study, we have fitted each absorption band to Lorentzians. In this way, one can see the increase of the different absorption intensity for each peak at different times (Figure 3.4.B). It can be seen that the absorbance at 275 and 300 nm began at 400s, which evolve more slowly than the 257 nm band; therefore, it can be ascribed to intermediate larger B) 300 400 500 600 700 800 0.0 0.4 0.8 1.2 1.6 2.0 2.4 0 s 800 s Absorbance/AU Wavelength/nm 200 300 400 500 600 700 800 0.0 0.5 1.0 1.5 1=257nm 2=275nm 3=302nm 4=360nm Absorbance Lorentzian/AU Wavelength/nm A) Chapter 3 55 clusters. The broad band located at 360 nm could be attributed to the aggregation of the particles at the end of the synthesis. These results agree with the literature. 8 3.3.3. Purification and PL characterization of small CuCLs. According to the Experimental Section, two different purification processes of the precipitated obtained in E31 have been done. We have obtained two different samples: CuCLs in water and CuCLs in EtOH. UV-Vis results showed on Figure 3.5.A display three main absorption bands located at (1) 5.85 eV (212 nm), (2) 5.36 eV (231 nm), and (3) 4.19 eV (296 nm) for CuCLs/H2O and Figure 3.5.B two main bands at (1) 5.99 eV(207 nm) and (2) 4.54 eV (273 nm) for CuCLs/EtOH. Figure 3. 5| UV-visible and fluorescence spectra of CuCLs/H2O . Excitation wavelength=296nm, emission wavelength = 408nm. Inset: Blue emission observed for an aqueous solution of CuCLs irradiating at λexcitation=296nm. (A) UV-Vis and emission and excitation spectra of CuCLs/EtOH. Excitation wavelength =275nm, emission wavelength=300nm.(B) The maximum emission peak of the as-purified CuCLs was located at 3.04 eV (408 nm) for CuCLs/H2O and at 4.13 eV (300 nm) for CuCLs/EtOH. It has to be noted that CuCLs/EtOH displays also some emission at 408 nm. The maximum excitation peak appears of both samples agree with the last absorption bands found in their UV-Vis spectra. Samples are colorless under day light, but are highly blue when irradiated with UV light (300nm) as it is shown in the inset of Figure 3.5.A. 200 300 400 500 600 700 0.0 0.8 1.6 2.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Absorbance/AU 3 2 1 Wavelength/nm Normalized PL Intensity/AU 200 300 400 500 600 700 0.0 0.8 1.6 2.4 0.0 0.2 0.4 0.6 0.8 1.0 1.2 2 1 Normalized PL Intensity/AU Absorbance/AU Wavelength/nm A) B) Synthesis of Small Copper Clusters 56 It has been shown that the simple spherical Jellium model seems to describe sufficiently well the emission energy in Au clusters1c through the expression: Eg= Efermi/N1/3, Eq. 1 where Efermi is the Fermi energy of bulk material 9 and N is the number of atoms per cluster. Assuming that a similar model can be applied to Cu clusters, one can get the number of atoms per cluster from the position of the emission band. The double emission peaks displayed by CuCLs/EtOH sample indicates the existence of two different size populations. Table 3.2 shows the energy emission values of the different samples and their metal core obtained by Eq.1. Table 3.2| Summary of the emission properties of the different as-purified CuCLs. Sample Eemi1/eV Eemi2/eV N1 N2 CuCLs/H2O 3.04 - 12 - CuCLs/EtOH 4.08 3.04 5 12 In order to check that the PL properties of the CuCLs are inherent of metal core and are not due to a possible solvatochromic effect, CuCLs/H2O was dried and redispersed in ethanol showing (Figure 3.6) the same emission band, which corroborates the assumption that the fluorescence come from the metal core. Although there are only a few reports about fluorescent CuCLs (Table 1.2Chapter 1), our results are consistent with other works. Vázquez-Vázquez et al.7 observed emission of CuCLs at 333nm at first stages of their synthesis, assigned to small CuNCLs with N<10. Emission at 408nm has also observed by Goswami et al. 10 assigned the emission to a Cu13 center. Thiol protected CuCLs studied by Wei et al. 11 display emission at 425 nm and they assigned to a Cu8 cluster. Therefore, different characterization techniques will be necessary in order to check the first approximation cluster size by the Jellium model. Chapter 3 57 Figure 3. 6| PL emission spectrum of CuCLs/H2O (black squares) initially (λexcitation=296 nm) and after being redispersed in EtOH (red open circles) (λexcitation=350 nm). Almost no emission maximum shift (408 nm) is observed. 3.3.4. Further characterization of small CuCLs. (A) Tauc approach. It is known that small metal clusters display a semiconductor-like behavior because of the appearance of a band gap at the Fermi level. For this reason, we have tried to determine the band gap by the traditional Tauc approach used for semiconductors, using the equation: (αhυ) = A (hν-Eg)n, Eq.2 where hυ is the photo energy (h is Plank constant and υ is the frequency radiation), α is the absorption coefficient, A is a constant, Eg is the band gap value and n equals ½ for a direct transition and 2 for an indirect transition. One can see (Figure 3.7) that it is difficult to fit the spectral data by this equation, which indicates that clusters display a molecule-like behavior rather than a classical semiconductor behavior. Nevertheless, the approximate band gap, which can be deduced by the Tauc approach. CuCLs/H2O would display 3.7eV for a direct transition, which are larger than the direct band gap of bulk Cu2O (2.17eV) and CuO (1.4eV) 12 , and also larger than other reported copper oxides nanostructures showing values in the range 2.24eV to 2.75eV. 13 CuCLs/EtOH direct transition would give us a mixture of band gaps between 5.29-4.15eV. The 250 300 350 400 450 500 550 600 650 0.00 0.25 0.50 0.75 1.00 Normalized PL /AU Wavelength/nm Synthesis of Small Copper Clusters 58 results here obtained agree with the Jellium model: ETauc (CuCLs/H2O) ≈ 3.7 eV (Jellium Eg≈ 3.04 eV) and ETauc (CuCLs/EtOH) ≈4.15 eV (Jellium Eg ≈ 4.15 eV). In conclusion, bigger energy gaps are obtained for CuCLs/EtOH indicating the presence of smaller clusters sizes. Figure 3.7| Tauc plots derived from the absorption spectra of CuCLs/H2O for a direct bandgap fitting (Eg = 3.7eV) (left) and for CuCLs/EtOH (5.29eV≤Eg≤ 4.15eV) (right). (B)TEM/HRTEM: TEM and HRTEM studies of CuCLs/H2O have been carried out. Small clusters are sensitive to electron beam irradiation and can produce larger nanoparticles upon irradiation for a prolonged time. 14 The isolated clusters of sub-nanometer dimensions are not expected to be seen clearly. Figure 3.8 shows the presence of small CuCLs forming aggregates of different sizes. The average size of the clusters inside the aggregates (Figure 3.8.C) is 0.61±0.13nm. This size is comparable to the size expected for the smallest 3D more stable closed-shell geometrical cluster, Cu13, (0.77nm). Therefore, one can conclude that the number of atoms per cluster is N ≈ < 13 atoms, which again agrees with previous estimations. Figure 3.9.A shows the HRTEM image of the sample after some time of irradiation with the strong electron beam. Crystal planes are now observed with a d space distance of 2.02±0.23Ǻ (corresponding to the (111) plane of the fcc phase of metal copper) indicating that clusters fuse to form crystal nanoparticles under the strong electron irradiation beam. 15 2.4 2.7 3.0 3.3 3.6 3.9 4.2 0.00 0.05 0.10 Eg Eg=3.7eV (h)1/2/(eVcm-1)1/2 h/eV 2.0 2.5 3.0 3.5 4.0 4.5 5.0 5.5 0.00 0.25 0.50 0.75 1.00 1.25 Eg4 Eg3 Eg2 Eg1 Eg1=5.29eV Eg2=5.05eV Eg3=4.65eV Eg4=4.15eV (h)1/2/(eVcm-1)1/2 h/eV Chapter 3 59 Figure 3. 8| TEM image of Cu samples showing CuCLs forming aggregates of different sizes (A). Enlargement of one aggregate of CuCLs (B). Size histogram of individual clusters inside of the aggregates shown in A (100 counts). Average cluster size: 0.61±0.13nm(C). On Figure 3.9.B the same area was examined during upon five minutes. As the irradiation time increase big particles started to appear on the grid. It is believed to be due to the coalescence of small CuCLs leading to the formation of large particles. B A 0,0 0,5 1,0 1,5 2,0 2,5 3,0 0 5 10 15 20 25 30 35 % particles Diameter(nm) C Synthesis of Small Copper Clusters 60 Figure 3. 9| HRTEM of Cu samples after being exposed to the electron beam. HRTEM image of the cluster’s evolution under the irradiation beam with time. HRTEM showing lattice spacing of 2.02±0.23Ǻ (average from 100 measurements) corresponding to the Cu (111) plane (A). TEM image of small copper clusters showing their conversion into bigger nanocrystals upon the irradiation of the electron beam, when a grid with clusters was irradiated during 5minutes (B). tirr d=2.02±0.23Ǻ Cu(111) A) B) Chapter 3 61 (C) NC-AFM: To further investigate the size of the copper clusters, non-contact atomic force microscopy (AFM-NC) studies were carried out (Experimental details on Chapter 2) deposited on mica substrates (with a rms ≈ 150pm). Small clusters (CuCLs/EtOH) NC-AFM image is displayed on Figure 3.10.A (average size 0.42±0.21 nm). Figure 3.10.A shows CuCLS/H2O with an average size (0.6±0.27 nm). It has to be noticed that the calculated size for the CuCLs/H2O clusters agrees with TEM results and also with the size estimated for a closed-shell cuboctahedra or icosahedra structure Cu13, (0.77nm). From theoretical studies, 16 it is known that small copper clusters CuNCLs (N≤6) have a tendency to be planar. Thus, the AFM images suggest than mainly planar CuN clusters with N≤6 are obtained in the CuCLs/EtOH sample and 3D clusters with N≈13 atoms are seen for CuCLs/H2O. The presence of a small amount of larger CuCLs on the Figure 3.10.B could be explained by the aggregation of smaller clusters when drying the drop deposited on the mica surface (e.g., sizes above 1.4 nm). Figure 3. 10| NC-AFM topography images of CuCLs/EtOH (A) and CuCLs/H2O (B) deposited on mica substrates (rms ≈150 pm) (left). Section analysis of the solid lines of AFM images (middle) and their corresponding histogram distribution heights (right). 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0 5 10 15 20 25 30 Average Height=0.42 ± 0.21 nm Relative Frequency(%) Height/nm 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0 5 10 15 20 Average Height=0.65 ± 0.27 nm Relative Frequency(%) Height/nm A) B) Synthesis of Small Copper Clusters 62 (D) Mass spectrometry Study: Mass spectrometry analysis was performed in order to determine the number of atoms of the clusters here studied. Two different techniques were used: laser desorption ionization time of flight (LDI-TOF) and electrospray ionization time of flight (ESI-TOF). (D.1) LDI-TOF results: The clusters will be identified by a comparison between the theoretical and the experimental mass spectra. Two different LDI analyses were carried out: one at low mass range from 100 to 600 amu (Figure 3.11) and high mass range from 900 to 2000 amu (Figure 3.12). For simplicity, assignments of CuCLs in the mass spectra are named as [N:X:Y] being N the number of copper atoms, X, buthyl groups and Y nitrogen atoms in the identified clusters. Eight different species CuN were found with N=2-5 in the low-mass range and N= 11-14 in the highmass range. Peaks found in the low-mass range marked with *, corresponding to Cu1 species, were not considered as clusters because of the associated difficulty to assign them either as clusters of Cu(0) or complexes of Cu(I), and will not be further discussed. Experimental and predicted isotopic patterns, as well as a summary of the ligands associated with different clusters are shown in the Annex C3. A summary of the main mass peaks will be showed in Tables 3.3 and 3.4 with their calculated theoretical m/z value. We should point out that the found divergence (1-2 m/z units) between theoretical and experimental data for some clusters could be attributed to the presence of adsorbed H+ ions. Besides no clusters peaks were found in the positive ion mode. The main conclusion we can derive from this mass spectra analysis is the presence of CuN clusters formed by N ≈ < 14 atoms which nicely agrees with the previous estimations made from the photoluminescent properties. Chapter 3 63 Figure 3. 11| Low m/z range LDI-TOF mass spectrum of copper clusters samples in water detected in negative ion mode. (A). Inset: Magnification of 350-700 amu range. Magnification showing one of the identified CuCLs together with its theoretical isotopic pattern: [Cu3Bu3N2]- (B). Figure 3. 12| High m/z range LDI-TOF mass spectrum of copper clusters samples in water detected in negative ion mode. (A). Magnification showing one of the identified CuCLs together with its theoretical isotopic pattern: [Cu14Bu2N(OH)2]- (B). A) B) EXPERIMENTAL THEORETICAL A) B) A EXPERIMENTAL THEORETICAL Synthesis of Small Copper Clusters 70 Figure 3. 15| Cyclic voltammetry of CuCLs/EtOH (Cu5) (black line) and CuCLs/H2O (Cu13) (gray line) in 0.1M KOH. Scan rate 10mV/s. 3.3.5. Stability of CuCLs. Fluorescence stability of metal clusters is an issue that should be explored for future applications. Shifts on the emission wavelengths can be associated to an oxidation process or aggregation of the clusters. Fluorescence of the assynthesized CuCLs was therefore studied. The CuCLs were found to be very stable, showing the same emission spectra after even one year of being stored at room temperature and also at 4ºC (Figure 3.16). The emission peak remained unchanged at 300 nm and 408 nm for CuCLs in EtOH and in H2O respectively. Figure 3.16| PL spectra of CuCLs/EtOH (A) and CuCLs/H2O (B) one year after the synthesis indicating the high stability of clusters. -1.5 -1.0 -0.5 0.0 0.5 -3 -2 -1 0 1 CuCLs/H2O CuCLs/EtOH I/ *10-6A E/V (Ag/AgCl) A) B) 300 400 500 600 0.0 0.5 1.0 1.5 PL 1 year after PL initial exc=297nm Normalized PL/AU Wavelength/nm 300 400 500 600 0.0 0.5 1.0 1.5 2.0 2.5 PL 1 year after PL initial exc=275nm Normalized PL/AU Wavelength/nm Chapter 3 71 3.4. SUMMARY. A direct one-step and environmentally friendly synthesis method was developed to produce fluorescent aqueous-soluble small copper clusters at 25ºC. Size and therefore photoluminescent properties of copper clusters can be optimized by adjusting the purification method. Different characterization techniques establish the small size of the copper clusters (CuNCLs, N<13 atoms). Stability over the time was studied observing no photoluminescence changes for CuCLs solutions stored at 4ºC and also at room Tª. Synthesis of Small Copper Clusters 72 REFERENCES. 1 a) Adhikari B.; Banerjee A. Chem. Mater. 22, 4364 (2010). b) Shang L.; Dong S. Chem. Commun. 9, 1088 (2008). c) Zheng J.; Nicovich P. R.; Dickson R.M. Annu Rev Phys Chem. 58, 409 (2007). d) Rao T. U. B.; Pradeep T. Angew. Chem. 122, 4017 (2010). 2 Rodríguez-Sánchez, L. PhD Dissertation: “Electrochemical synthesis of Ag and Co nanoparticles. Kinetics formation and physicochemical properties”. University of Santiago de Compostela (2003). 3 Lisiecki I.; Pileni M. B. J. Am. Chem. Soc. 115, 3887 (1993). 4 Zhao M.; Sun L.; Crooks R.M. J. Am. Chem. Soc. 120, 4877 (1998). 5 OzinG.A.; Mitchell S.A.; McIntosh D.F.; Mattar S. J. Phys. Chem. 87, 4651 (1983). 6 De Waele V.; Kecht J.; Tahri Z.; Mostafavi M.; Bein T.; Mintova S. Sensors and Actuators B 126, 338 (2007). 7 Vázquez-Vázquez C.; Bañobre-López M.; Mitra A.; López-Quintela M.A.; Rivas J. Langmuir 25, 8208 (2009). 8 a) Khatouri J.; Mostafavi M.; Amblard J.; Belloni J. Chem. Phys. Lett. 191, 351 (1992). b) Ershov B. G.; Janata E.; Michaelis M.; Henglein A. J. Phys. Chem. 95, 8996 (1991). 9 de Heer W.A. Rev. Mod. Phys. 65, 611 (1993). 10 Goswami N., Giri A., Bootharaju M.S., Xavier P.L., Pradeep T., Pal S.K. Anal. Chem. 83, 9676 (2011). 11 Wei W. T., Lu Y. Z., Chen W. , Chen S. W., J. Am. Chem. Soc. 133, 2060 (2011). 12 Ghijsen J.; Tjeng L.H.; van Elp J.; Eskes H.; Westerink J.; Sawatzky G.A.; Czyzyk M.T. Phys Rev B 38, 11322 (1988). 13 a) Zhang X.; Zhang D.; Ni X.; Zheng H. Solid-State Electronics 52, 245 (2008).b) Singh D. P.; Neti N. R.; Sinha A.S.K.; Srivastava O.N. J. Phys. Chem. C 111, 1638 (2007). c) Ng C.H.B.; Fan W.Y. J. Phys Chem B 110, 20801 (2006). Chapter 3 73 14 Muhammed M.A.H.; Pradeep T. Chem. Phys. Lett. 449, 186 (2007). 15 Shibu E.S.; Muhammed M.A.H.; Tsukuda T.; Pradeep T. J Phys Chem C 112, 12168 (2008). 16 a) Kabir M.; Mookerjee A.; Bhattacharya A.K. Eur. Phys. J. D 31, 477 (2004). b) Poater A.; Duran M.; Jaque P.; Toro-Labbe A.; Sola M. J. Phys. Chem. B 110, 6526 (2006). 17 a) DiCenzo S.B.; Berry S.D.; Hartford E.H. Jr. Phys. Rev. B 38, 8465 (1988). 18 Wertheim G.K.; DiCenzo S.B.; Buchanan D.N.E. Phys Rev B 33, 5384 (1986). b) Kaden W. E.; Wu T.; Kunkel W. A.; Anderson S.L. Science 326, 826 (2009). 19 Pan J.S.; Tao J.G.; Huan C.H.A.; Zhang Z.; Chai J.W.; Wanga S.J. Applied Surface Science 256, 1088 (2009). 20 a) Shin D.-W.; Dong C.; Mattesini M.; Augustsson A.; Mao S.; Chang C.L.; Persson C.; Ahuja R.; Nordgren J.; Wang S.X.; Guo J.-H. Chem. Phys. Lett. 422, 543 (2006). b) Shin D.-W.; Wang S.X.; Marshall A.F.; Kimura W.; Dong C.; Augustsson A.; Guo J. Thin Solid Films 473, 267 (2005). 21 a) Zhang P.; Sham T. K. Phys. Rev. Lett. 90, 245502 (2003). (b) Moriarty P. Comment on “X-Ray Studies of the Structure and Electronic Behavior of Alkanethiolate-Capped Gold Nanoparticles: The Interplay of Size and Surface Effects. Phys. Rev. Lett. 92, 109601 (2004). (c) Zhang P.; Sham T. K.; Zhang and Sham Reply, Phys. Rev. Lett. 92, 109602 (2004). 22 Nicoletta Ferreti PhD Dissertation: “X-ray photoelectron spectroscopy of size selected copper clusters on silicon”. Technische Universität Berlin, (2009). 23 Jirka I. Surf. Sci. 232, 307 (1990). 24 Akola J.; Walter M.; Whetten R.L.; Häkkinen H.; Grönbeck H. J. Am. Chem. Soc. 130, 3756 (2008). 25 a) Ambrose J.; Barradas R.G.; Shoesmith D.W. J. Electroanal. Chem. 47, 47, (1973). b) de Chialvo M.R.G.; Salvarezza R.C.; Vasquez D.; Arvia A.J. Electrochim. Acta 30, 1501 (1985). c) Sander U.; Strehblow H.H.; Dohrmann J.K. J. Phys. Chem. 85, 447(1981). CHAPTER 4. Thermal Growth of Small Copper Clusters Chapter 4 77 4.1. INTRODUCTION. Fluorescent metal clusters show great promise for biological imaging, sensing and photocatalytic applications. A wide palette of band gaps and therefore photoluminescent properties are desirable, which makes necessary the cluster size increase. Therefore, the synthesis control over the cluster size results in a big challenge in their future applications as well as the possibility of studying the evolution of their inherent properties with the size. As we explained on Chapter 1 advances on the Cu (0) clusters synthesis have been lately developed, but their size is smaller than 14 atoms. Consequently, the main purpose in this Chapter will be the obtaining and characterization of larger Cu (0) clusters. The strategy here proposed consists of a thermal control growth of the smaller clusters. Thermal treatments in solution of 2nm gold nanoparticles (AuNPs) stabilized with DDT (dodecanethiol) were studied by Zhong et al, 1 obtaining a size evolution until 5-6 nm diameter AuNPs. Also studies of solid AuNPs - DDT were developed by Teranishi et al. 2 and solid CuNPs by Dong et al. 3 The increasement of the clusters size by temperature annealing and therefore shifting their emission wavelength into the Vis region was developed by Chun-Sung et al. 4 They showed a clear PL shift in their Si clusters (emission wavelength 365nm) with 162ºC annealing obtaineing emissions at 420 nm, and finally higher temperatures led to Si nanoparticles with no PL properties. Thermal growth methods depend basically on the interaction between the capping agent (here tetrabutylammonium nitrate TBANO3) and the metal core. This interaction will be crucial on the desorption, coalescence and aggregation process during the temperature treatment, which will determine the growth or ripening process. Four main growth steps can are assumed on this thermal evolution (Scheme 4.1): (1) Desorption of the tetralkyl ammonium salts from the metal core of small Cu(0) CLs. (2) Approximation of metal cores by attractive forces. (3) Coalescence of the metal cores. (4) Reorganization of the larger clusters. As it is illustrated in the Scheme 4.1, the basic concept of our thermal strategy was to explore the viability of a thermally activated coalescence of the small clusters toward the formation of larger clusters by the mechanism: Cuxa+ + Cuyb- Cux+y(a-b)+ Thermal Growth of Small Copper Clusters 78 Scheme 4. 1| Evolution of small Cu(0) clusters into larger CuCLs with thermal control growth. Chapter 4 79 4.2. EXPERIMENTAL SECTION. Small purified CuCLs (CuN, N≤13) in powder form were subjected to a thermal growth. 1mg of these CuCLs were dried at 80ºC (Tª < 100ºC, water boiling point, and also Tª<117ºC, TBANO3 melting point) for 1 hour. A slight increase in temperature could also have a significant impact on the capping agent also. Therefore, thermal growth of the clusters must be performed with a temperature<temperature melting of the tetrabutylammonium salt. Then, 10 mL of acetonitrile (MeCN) was added and the mixture was vigorously stirring for three hours at ambient temperature to allow the clusters to redispersion. During the stirring process, the solvent changed from colorless to light yellow and finally to a light brown color. The product was then characterized by different techniques: MALDI MS, ESI MS, AFM, TEM, XPS, UV-Vis and fluorescence spectroscopy. High-performance liquid chromatography (HPLC) was used as a purification process in order to separate the different clusters sizes. The HPLC chromatographs employed was Waters 600 controller pump capable of gradient elution and equipped with a Waters 2998 PDA detection system and multi-λ Waters 2475 fluorescence. Analytical (4.6x150 mm) and preparative columns (21.2x250 mm) were both C18 columns (octadecyl carbon chain).The mobile phase consisted of acetonitrile: water (65:35) mixtures. The analytical and preparative total flow rate, 1 mL/min and 20mL/min, respectively with injection total volumes of 10 and 190 µL. Fractions for fluorescence analysis were collected over the time intervals on the specific peaks of the chromatogram. . Thermal Growth of Small Copper Clusters 86 Figure 4. 6| ESI-TOF mass spectrum (negative ion mode) of the CuCLs. Table 4. 2| Assignments of the clusters in the negative mode corresponding to LDI-TOF, MALDI-TOF and ESI –TOF spectra. LDI-TOF MALDI-TOF ESI-TOF m/zexp m/ztheo Cluster m/zexp m/ztheo Cluster m/zexp m/ztheo Cluster 349.34 349.73 [Cu5O2]- 365.41 365.74 [Cu5O3]- 579.06 578.40 [Cu8N5]- 693.07 699.01 [Cu11]- 679.22 679.48 [Cu10N2O]- 670.03 670.49 [Cu10(OH)2H]- 754.71 755.03 [Cu11N4]- 1040.01 1041.07 [Cu16CN]- 1143.75 1143.83 [Cu18]- 1222.05 1222.39 [Cu19NH]- 1279.25 1270.92 [Cu20]- 400 600 800 1000 1200 1400 [Cu20]- [Cu10(OH)2H]- [Cu5O3]- m/z Chapter 4 87 4.3.4. XPS results. XPS measurements were carried out in order to check the oxidation state of Cu in these CuCLs. Results (Figure 4.7) show a smaller shift of the Cu2p3/2 peaks BE than the smaller CuCLs, which is consistent with the size increase of the sample. 8 Identical bulk spin–orbit coupling (19.9 eV) was found showing the presence of Cu (0). Figure 4. 7| Cu 2p3/2 peak of small CuCLs (cyan line) and large CuCLs (red line).Arrow shows the BE value of bulk copper. Inset: Some BE,s of mass selected deposited CuCLs. A blue shift with respect to the bulk peak is observed for all clusters. (Image taken from reference 8) (A). Cu 2p3/2 BE’ s as function of cluster size. A general trend with increasing cluster size is obvious. (Image taken from reference [8]) (B). As it can be seen on Figure 4.7.A the copper clusters here synthesized show a similar binding energy (BE) shift throught smaller values that those observed by 930 931 932 933 934 935 936 937 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Cu 2p3/2 CuNCLs (N20) CuNCLs (N13) Cubulk Normalized Photoelectron Intensity/AU Binding Energy/eV A) B) Thermal Growth of Small Copper Clusters 88 Ferreti (Inset Figure 4.7.A). Figure 4.7.B shows the general trend found by Ferreti for the BE with the cluster size. It has to be said that for smaller clusters of CuN (N≤10) the values change without a clear tendency. But from 10 to more than 70 atoms the trend is clear observing the approaching to the bulk value when the cluster size increase. 4.3.5. Stability of large clusters. Stability of the as-synthesized larger clusters in aquous solutions and throughout the time have been studied. Aqueous solubilization of larger copper clusters is desirable for suitable in biological applications. Direct redispersion in water was unsuccesful and no emission was found, which indicates their instability in aqueous solutions. Hence, another aqueous solubilization method was tried, named “caustic extraction”. Due to the miscible character of water and acetonitrile, phase transfer is not possible. However, the used process consisted of a direct staged contact between the MeCN phase and a concentrated caustic solution. Smiley 9 discovered that by adding NaOH or KOH concentrated solutions (50% weight of hydroxide aqueous solution) to an acetonitrile solution in the range 1:1 to 1:10 volume, two inmiscible phases are formed. By stirring this two phase system acetonitrile can be purified removing the H2O and HCN impurities present on the MeCN. If we assume that larger CuCLs in MeCN can be surrounded by CNgroups, when this purification is carried out, the clusters will go to the aqueous phase. Briefly, 2mL of CuCLs were added to a mixture in volume 2:1 (NaOH: MeCN). Initially, CuCLs remained in the interphase observed by a yellow colour. Vigorous stirring was maintained for 72 hours. The PL intensity at 480 nm on the aqueous phase CuCLs increased with time. After 1 hour stirring the most emissive specie was found to be Cu7. After 20 hours Cu19 emission (477nm) appears (Figure 4.8.A). Time stability of CuCLs transferred to water was studied (Figure 4.8.B). No changes on the emission band were observed after one year. This fact point out the great stability of these clusters even in an aqueous media at room temperature. Chapter 4 89 Figure 4. 8| Emission spectra during the time of the aquous phase obtained by “caustic extraction” (A). Time stability emission spectra of Cu19 clusters. Black circles correspond to emission after synthesis and red circles to emission after 1 year (B). 4.3.6. HPLC separation of large CuCLs. The size-selective separation of polydisperse solutions of metal clusters has relevance in order to obtain monodisperse solutions required for applications such as catalysis or sensoring. It has been shown that monodisperse clusters can be obtained by different separation processes, such as fractionated precipitation, 10 size exclusion column chromatography (SEC) 11 , HPLC 12 , sequential size-selective precipitation 13 and gel electrophoresis. 14 200 300 400 500 600 0 50 100 150 200 200 300 400 500 600 exc=370nm exc=290nm PL Intensity/AU Wavelength/nm 1 hour 20 hours 72 hours 200 300 400 500 600 0.0 0.2 0.4 0.6 0.8 1.0 1.2 excitation=380nm Normalized PL/AU Wavelength/nm B) A) Thermal Growth of Small Copper Clusters 90 For this reason, Rodriguez Cobo E. has carried out an HPLC separation of the large Cu clusters synthesized in this Chapter (under the conditions summarized on Experimental Section) 15 . The method optimization was developed with an analytical column and then the sample was studied using a preparative column. Figure 4.9 shows the obtained chromatogram with a dominant set of peaks starting at 2.18 min. Figure 4. 9| Chromatogram (emission detected at 296 nm y exciting at 270nm) of asprepared larger CuCLs. The mobile phase was 65:35 acetonitrile/water ratio at a flow rate of 20 mL/min, and the column was a preparative C18, isocratic separation. Different fractions were collected: 1=2.18 min, 2=3.31 min, 3=5.75 min, 4=6.25 min, 5= 7.02 min, 6=7.60 min, 7=10.10 min, 8=11.53 min, 9=12.57 min, 10=14.40 min, 11=15.81 min, 12=28.50 min. The “ideal” separation would be some order of cluster elution related to core cluster size. This question was studied by fluorescence spectroscopy of the different collected fractions (previously concentrated, Figure 4.10). It can be clearly seen that emissions at lower wavelengths (smaller clusters) (λemi=300 nm Cu5 by Jellium ) appear at the first collected fractions, from 1 to 7, mixture of smaller emissions (λemi1=300 nm+ λemi2=408 nm Cu5 and Cu13 by Jellium) are observed from 8 to 10 fractions, and finally waste fraction spectra displays the emission at higher wavelengths, (λemi=460 nmCu18). 010 20 30 40 50 12 11 10 9 8 7 6 5 4 3 2 1 Emission/AU Time/min Chapter 4 91 Fraction 1-7 Fraction 8-12 Waste Figure 4. 10| Normalized PL spectra of the different obtained fractions and the waste after HPLC separation. The presence of larger clusters on the waste, can be explained by the detection PL method on the HPLC, the fixed excitation wavelength was not suitable for these clusters, and, therefore, they do not show any peak on the chromatogram. Additionally, this order of elution of the clusters is as if the 200 300 400 500 600 700 0.00 0.25 0.50 0.75 1.00 Cu5 exc=270nm emi=300nm Normalized PL Intensity/AU Wavelength/nm 200 300 400 500 600 700 0.00 0.25 0.50 0.75 1.00 Cu13 Cu5 exc=270nm emi=300nm Normalized PL Intensity/AU Wavelength/nm 200 300 400 500 600 700 0.00 0.25 0.50 0.75 1.00 Cu13 exc=296nm emi=408nm Normalized PL Intensity/AU Wavelength/nm 200 300 400 500 600 700 0.00 0.25 0.50 0.75 1.00 Cu8 Cu5 exc=270nm emi=300nm Normalized PL Intensity/AU Wavelength/nm 200 300 400 500 600 700 0.00 0.25 0.50 0.75 1.00 Cu8 exc=290nm emi=365nm Normalized PL Intensity/AU Wavelength/nm 200 300 400 500 600 700 0.00 0.25 0.50 0.75 1.00 Cu18 exc=370nm emi=470nm Normalized PL Intensity/AU Wavelength/nm Thermal Growth of Small Copper Clusters 92 chromatographic separation behaved as a size exclusion process. In conclusion, PL results show that copper clusters differ in core size and can be separated. HPLC has to be regarded as an extremely promising tool to use with polydisperse samples of metal clusters, paying careful attention to the different factors that can have an influence in the elution order. Chapter 4 93 4.4. SUMMARY. Thermal control growth of small copper clusters into larger clusters was successfully developed confirmed by the emission at higher wavelengths and the presence of different clusters sizes by mass spectrometric studies. The “Caustic extraction” method was satisfactory in order to obtain the as-synthesized large CuCLs in aqueous medium. Large CuCLs displayed high photoluminescence stability during one year stored at ambient temperature. The polydisperse large CuCLs sample can be separated into different monodisperse fractions using the HPLC technique. Thermal Growth of Small Copper Clusters 94 REFERENCES. 1 Zhong C.J.; Zhang W.X.; Leibowitz F.L.; Eichelberger H.H. Chem. Commun. 13, 1211 (1999). 2 Teranishi T.; Hasegawa S.; Shimizu T.; Miyake M. Adv. Mater. 13, 1699 (2001). 3 Dong T.Y.; Wu H.-H.; Lin M.-Ch. Langmuir 22, 6754 (2006). 4 Chun-Sung Y.; Kauzlarich S.M.; Wang Y.C.; Lee H.D.W.H. J. Clu. Sci. 11, 423 (2000). 5 a) Wu Z.; Lanni E.; Chen W.; Bier M. E.; Ly D.; Jin R. J. Am. Chem. Soc. 131, 16672 (2009). b) Wu Z.; Jiang D.; Lanni E.; Bier M. E.; Jin R. J. Phys. Chem. Lett. 1, 1423 (2010). 6 Ketch J.; Tahri Z.; De Waele V.; Mostafavi M.; Mintova S.; BeinT. Chem. Mater. 18, 3373 (2006). 7 Kawasaki H.; Kosaka Y.; Myoujin Y.; Narushima T.; Yonezawa T.; Arakawa R. Chem Comm 47, 7740 (2011). 8 Nicoletta Ferreti PhD Dissertation “X-ray photoelectron spectroscopy of size selected copper clusters on silicon”. Technische Universität Berlin, (2009). 9 Smiley R.A.; Wilmington, Del. United States Patent, (1981). 10 a) Price R. C.; Whetten R. L. J. Am. Chem. Soc. 127, 13750 (2005). b) Whetten R. L.; Khoury J. T.; Alvarez M. M.; Murthy S.; Vezmar I.; Wang Z. L.; Stephens P. W.; Cleveland C. L.; Luedtke W. D.; Landmann U. Adv. Mater. 8, 428 (1996). 11 (a) Siebrands T.; Giersig M.; Mulvaney P.; Fischer C.-H. Langmuir 9, 2297 (1993). (b) Wilcoxon J. P.; Martin J. E.; Provencio P. Langmuir 16, 9912 (2000). 12 (a) Jimenez V. L.; Leopold M. C.; Mazzitelli C.; Jorgenson J. W.; Murray R. W. Anal. Chem. 75, 199 (2003). (b) Song Y.; Jimenez V.; McKinney C.; Donckers R.; Murray R. W. Anal. Chem. 75, 5088 (2003). (c) Choi M. M. F.; Douglas A. D.; Murray R. W. Anal. Chem. 78, 2779 (2006). (d) Zhang Y.; Shuang S.; Dong C.; Lo C. K.; Paau M. C.; Choi M. M. F. Anal. Chem. 81, 1676 (2009). 13 Yang X.; Su Y.; Paau M.Ch.; Choi M.M.F. Anal. Chem. 84, 1765 (2012). Chapter 4 95 14 a) Schaaff T. G.; Knight G.; Shafigullin M. N.; Borkman R. F.; Whetten R. L. J. Phys. Chem. B 102, 10643 (1998). b) Gautier C.; Bürgi T. J. Am. Chem. Soc. 133, 11079 (2006). 15 Rodríguez-Cobo E., PhD Dissertation:“Synthesis, properties and functionalization of metal clusters”. University of Santiago de Compostela. (2012). Synthesis of Large Copper Clusters 102 Absorbance Photoluminescence Figure 5. 2| UV-Vis absorption and emission spectra of the different samples (E4 and E44) just after synthesis. A competition process can explain the synthesis of larger clusters. Our group 2 found that the reduction of the AcOTBA begins around -2.2 V, value that is achieved on the E44 sample. Therefore, the competition between the formation of small nuclei and the reduction of the counterion will happen at the same time giving less small Cu nuclei, and consequently a size increase of the initially formed clusters instead of the generation of smaller clusters. Thus, different characterization techniques would now be used in order to identify the Cu clusters present in the E44 sample. 200 300 400 500 600 700 800 0 200 400 600 800 exc=290nm exc=350nm Cu13 Cu7 PL intensity/AU Wavelength/nm 200 300 400 500 600 700 800 0.00 0.25 0.50 0.75 1.00 Supernatant E4/MeCN 300nm 275nm 244nm Normalized Absorbance/AU Wavelength/nm 200 300 400 500 600 700 800 0.00 0.25 0.50 0.75 1.00 Precipitate E4/Heptane 360nm 275nm 226nm Normalized Absorbance/AU Wavelength/nm 200 300 400 500 600 700 800 0.00 0.25 0.50 0.75 1.00 292nm E44/MeCN 430nm 345nm 257nm Normalized Absorbance/AU Wavelength/nm 200 300 400 500 600 700 800 0 200 400 600 exc=270nm exc=350nm Cu5 PL intensity/AU Wavelength/nm 200 300 400 500 600 700 800 0 50 100 150 200 Cu25 Cu12 exc=230nm exc=350nm exc=430nm Cu8 PL intensity/AU Wavelength/nm Chapter 5 103 5.3.2. Kinetic evolution during the synthesis. The UV-Vis spectra evolution (Figure 5.3.A) displayed similar bands to small CuCLs located at ≈260 nm, ≈280 nm, ≈ 297 nm, ≈360 nm, but new bands appeared ≈314 nm and ≈ 650 nm, besides the change in their intensities. Different absorbance bands were fitted with Lorentzians functions and plotted versus the time (Figure 5.3.B). Herein, the most intense bands were 260 and 280 nm (assigned to a Cu2 and Cu3 cluster respectively) until the 300 s, but then the most intense was the absorption at 297 nm (related with a Cu13 cluster). Figure 5. 3| Absorbance versus time during the electrochemical synthesis of CuCLs. Time between successive spectra is Δt =20 s . Inset shows a photograph of the E44 sample.(A). Absorbance versus time for the peaks centered at 260, 280, 297, 314, 366, 590 and 650 nm (fitted by Loretzians) during the synthesis. The comparison between kinetic evolution on small and larger copper clusters synthesis reveals differences. As we have observed on Chapter 3, the most intense absorption bands were located at 257 and 275 nm, but now on larger CuCLs the most intense are located at 257 and 297 nm. Besides, the band at 360 nm increase the intensity and a new band appear from the intial stages of the synthesis at 650 nm. Therefore, it can be concluded the presence of Cu13 on this sample also (absorption at 297 nm emission at 408 nm), but the band edge located near the 430 nm can be the responsible of the emission at 520 nm corresponding to CuN with N>13. A) B) 300 400 500 600 700 800 0.0 0.6 1.2 1.8 2.4 0 s 800 s Absorbance/AU Wavelength/nm -200 0 200 400 600 800 0.0 0.5 1.0 1.5 260nm 280nm 297nm 314nm 366nm 590nm 650nm Absorbance Intensity/AU Time/s Synthesis of Large Copper Clusters 104 5.3.3. Band gap determination. Energy band gaps (Eg) of large CuCLs were calculated by using the Jellium model: 3.42, 3.03 and 2.38 eV (emission energy) corresponding to Cu5, Cu8 and Cu25 metal cores. This result agreed with the Egap calculated by the direct transition Tauc approach (Figure 5.4.A) which also displayed four different Eg values between 2.32 eV and 4.26 eV. The values are summarized in Table 5.2. It should be noted that the Cu8 and Cu13 metal cores correspond to stable “magic numbers” (Cu8) and “geometric favorable structure” (Cu13 ). Figure 5. 4| Tauc plot for a direct band gap fitting (2.32 eV≤Eg≤4.26 eV) (A). Emission and excitation spectrum of fluorescent large CuCLs with different excitation wavelengths from 230 to 470 nm (B). Table 5. 2|Comparison between the Band gap values calculated from the Tauc approach and derived from the emission wavelengths. Eg/eV 3.42 3.05 2.32 Eemission /eV (nm) 3.42 3.03 2.38 Cluster core by Jellium Cu8 Cu13 Cu25 B) A) 2 3 4 5 0.0 0.8 1.6 2.4 Eg1=4.26 eV Eg2=3.42 eV Eg3=3.05 eV Eg4=2.32 eV Eg4 Eg3 Eg1 Eg2 (h)1/2(eVcm-1)1/2 (h)/eV 200 300 400 500 600 700 800 ex=470nm ex=450nm Fluorescence Intensity/AU Wavelength/AU ex=430nm ex=350nm ex=410nm ex=390nm em3=510nm em2=400nm ex=330nm ex=310nm ex=290nm ex=270nm ex=250nm em1=360nm ex=230nm Chapter 5 105 5.3.4. TEM and NC-AFM measurements. TEM and NC-AFM were performed in order to study the cluster size. Results obtained by TEM confirm the polydispersity of the sample showing an average diameter of 1.37±0.34nm (Figure 5.5.A). NC-AFM heights display a closer value: 1.12±0.5nm (Figure 5.5.C). Therefore, the larger size of the assynthesized clusters was confirmed. Figure 5. 5| NC-AFM topography image of CuCLs deposited on mica substrate (rms ≈150 pm).(A) Section analysis of the solid lines is given on (B) and the corresponding histogram distribution height (C). TEM image of CuCLs. Inset: histogram distribution (D). 0.0 0.5 1.0 1.5 2.0 2.5 3.0 0 5 10 15 20 25 30 Frequency(%) X Axis Title 0 1 2 3 4 0 5 10 15 20 25 Average Height=1.12±0.5 nm Frequency Count(%) Size/nm A) B) C) D) Synthesis of Large Copper Clusters 106 5.3.5. Aqueous solubility and mass spectrometry studies. The aqueous solubilization of larger clusters was carried out by a direct process: 3mL of CuCLs were dried on the rotary evaporator at 60ºC (lower temperature than acetonitrile boiling point: 80ºC and TBAAc melting point: 96.5ºC) giving a brown gel, which was subsequently redispersed in water (Aqueous Fraction). This gel was not very soluble in water and the insoluble precipitate was again redispersed in acetonitrile (MeCN Fraction). Both fractions were studied by UV-Vis absorption (Figure 5.6.A), fluorescence spectroscopy (Figure 5.6.B) and LDI-TOF mass spectrometry (Figure 5.7). UV-Vis of MeCN fraction shows a decrease on the 245 nm absorbance whereas the spectrum of the aqueous fraction displays an absorbance band at 245nm. Fluorescence emission at 408 nm (Cu13 clusters) indicated that smaller CuCLs are soluble in water while lager clusters are not. Figure 5. 6| UV-Vis spectra of large CuCLs initial (black line), aqueous transfer (inset, blue line) and insoluble precipitate redispersed in MeCN (red line) (A) Emission spectra of CuCLs in water (B). Taking into account PL and absorption results, CuCLs redispersed in water should mainly composed by Cu13 metal cores (emission at 408nm). LDITOF measurements were carried out on both fractions. Therefore, different sets of peaks are expected from aqueous fraction to acetonitrile aliquots. As it can be seen on Figure 5.7 the most intense group of peaks on both spectra corresponds to low m/z values: 609.20 and 649.52 amu corresponding to [Cu9NC2]- and [Cu10N]- for MeCN and aqueous fraction respectively. However, there are a set of peaks present only on the MeCN fraction: 1289.11 and 1627.24 amu which can be assigned to: [Cu19CN5]- and [Cu25NCN]- respectively. On the basis of the LDI200 300 400 500 600 700 800 0 50 100 150 Cu13 ex=296nm em=408nm ex=450nm PL intensity/AU Wavelength/nm 300 400 500 600 700 800 0.0 0.3 0.6 0.9 1.2 1.5 CuCLs initial CuCLs after aqueous transfer CuCLs in water 300 400 500 600 0.0 0.1 0.2 0.3 0.4 0.5 Absorbance/AU Wavelength/nm Absorbance/AU Wavelength/nm A) B) Chapter 5 107 mass spectra results, the final product in the aqueous fraction only comprises small CuCLs. Figure 5. 7| LDI-TOF mass spectra of CuCLs in MeCN (A). Magnification of the set of peaks from 1200 to 1400 range (B). LDI-TOF of CuCLs in water (D). All the spectra were collected in negative ion mode. Magnification of some peaks are shown as example with their theoretical isotopic patterns. 600 800 1000 1200 985 990 995 1000 Theoretical [Cu11Bu4N(OH)O]- m/z 985 990 995 1000 Experimental [Cu11Bu4N(OH)O]- m/z 600 605 610 615 620 Theoretical [Cu9NC2]- m/z 600 605 610 615 620 Experimental [Cu9NC2]- m/z m/z 1200 1400 1600 1800 1275 1280 1285 1290 1295 1300 1305 Experimental [Cu19CN5]- m/z 1280 1285 1290 1295 1300 1305 Theoretical [Cu19CN5]- m/z 1615 1620 1625 1630 1635 1640 1645 Theoretical [Cu25NCN]- m/z 1610 1620 1630 1640 1650 Experimental [Cu25NCN]- m/z m/z 600 900 1200 1500 1800 630 640 650 Experimental [Cu10N]- m/z 640 645 650 655 660 665 Theoretical [Cu10N]- m/z m/z 1200 1300 1400 1233.11=[Cu19CN]- 1260.94=[Cu19CN3]- 1289.11=[Cu19CN5]- 1317.3=[Cu19CN7]- 1344.7=[Cu19CN9]- 1344.7 1317.3 1289.11 1260.94 1233.11 m/z A) B) C) Synthesis of Large Copper Clusters 108 A comparison between the different m/z values found on the MeCN and the aqueous fractions is displayed in Table 5.3. It can be concluded that the initial sample is formed by a mixture of CuCLs with the presence until a Cu25 core. However, aquous fraction only shows the presence of small CuCLs with a Cu10 core. Table 5. 3| LDI-TOF results in negative ion mode for the MeCN and aqueous fraction at the optimized high mass range from 400 to 1800 amu with their experimental and theoretical m/z peaks and their corresponding assignation. MeCN Fraction Aqueous Fraction m/zexp m/ztheo Cluster m/zexp m/ztheo Cluster 609.20 609.94 [Cu9NC2]- 649.52 649.47 [Cu10N]- 990.69 991.48 [Cu11Bu4N(OH)2O]- 1233.11 1233.39 [Cu19CN]- 1260.94 1261.40 [Cu19CN3]- 1289.11 1289.41 [Cu19CN5]- 1317.30 1317.43 [Cu19CN7]- 1344.70 1345.44 [Cu19CN9]- 1627.24 1628.67 [Cu25NCN]- Chapter 5 109 5.4. SUMMARY. One step electrochemical synthesis of large CuCLs with CuN (8≤N≤25) were successful developed and their size were characterized by NC-AFM and mass spectrometry techniques. Only small copper clusters with a Cu10 core were aqueous soluble. These visible-light absorption copper clusters are a promising materials on future applications as visible photocatalysts. Synthesis of Large Copper Clusters 110 REFERENCES. 1 a) Reetz M. T.; Helbig W. J. Am. Chem. Soc. 116, 7401 (1994). b) RodríguezSánchez M.L.; Rodríguez M.J.; Blanco M.C.; Rivas J.; López-Quintela M.A. J. Phys. Chem. B 109, 1183 (2005). c) Reetz M. T.; Winter M.; Breinbauer R.; Thurn-Albrecht T.; Vogel W. Chem. Eur. J. 7, 1084 (2001). 2 Rodríguez-Sánchez, L. PhD Dissertation: “Electrochemical synthesis of Ag and Co nanoparticles. Kinetics formation and physicochemical properties”. University of Santiago de Compostela (2003). CHAPTER 6. Photoluminescence Stability of CuCLs Photoluminescence Stability of CuCLs 118 excitation wavelengths are observed, but not on the emission. Therefore, the stability of the Cu13 clusters on different enviroments can be checked. Figure 6. 3| Excitation (black squares)-Emission (hollow squares) spectra (maximum intensity) of copper clusters in different solvents. (Pentane) ( λexc = 355nm, λemi =406nm). (Chloroform) (λexc=350nm, λemi=418nm). (Ethanol) (λexc =350nm, λemi =420nm). (Water) ( λexc =297nm, λemi =406nm). (C.3) O2 quenching. Oxygen is known in fluorescence to be like a potent quencher of fluorophores. In the quenching reaction, ground state of oxygen may react with triplet state specie under the formation of highly reactive singlet oxygen: S0 (hν)S1 (ISC)T T+3O2S0+1O2 The singlet oxygen can react with surrounding molecules, including excited as well as ground state molecules. This reaction normally destroys the “fluorophore”. Here aqueous Cu13CLs solution was investigated under oxygenated conditions. The oxygen-saturated Cu13 clusters solution exhibited the same intensity emission at 408 nm with no evidence of oxygen quenching (Figure 6.4). These results may indicate that CuCLs fluorescence is oxygen independent. 250 300 350 400 450 500 550 600 0.5 1.0 0.5 1.0 0.5 1.0 0.5 1.0 Water Wavelength/nm Ethanol Chloroform Normalized PL Intensity/AU Pentane Chapter 6 119 Figure 6. 4| Fluorescence stability of O2 (gray line) and N2 (black line) saturated small Cu13 clusters in aqueous solutions. (C.4) Temperature effects: Temperature is a determinant factor on the emission intensity. Normally, metal clusters decrease their PL intensity with temperature increase 5 although, Pradeep et al found that the intensity increased with an increasing temperature. 6 Here, Cu13CLs were studied under different temperatures (Figure 6.5) observing a monotonous decrease on the PL intensity. However, it is worth noting that the fluorescence emission properties are reversible, when the sample returns to an ambient temperature they display the same PL intensity. The explanation of this PL decrease is a non-radiative relaxation. Two factors may cause this decrease: (1) thermal activation of non-radiative trapping or (2) thermal escape from the clusters, which becomes prominent by increasing the temperature with any change in emission peak position. There is a non-linear dependence of the fluorescence intensity on temperature (Figure 6.5, Inset) and it required an increase of 40K (from 278K to 318K) for a 50% decrease of the fluorescence emission intensity. 300 400 500 600 0 5 10 N2 O2 ex=296nm em=408nm PL Intensity/AU Wavelength/nm Photoluminescence Stability of CuCLs 120 Figure 6. 5| Fluorescence spectra as a function of the temperature from 278K to 300K of Cu13 clusters in aqueous solution (Inset: PL intensity vs Temperature in K). 300 400 500 600 0 30 60 90 em=408nm 278K 288K 298K 318K 343K ex=296nm 275 300 325 350 20 30 40 50 PL intensity/AU T/K PL intensity/AU Wavelength/nm Chapter 6 121 6.3.2. Capping molecules-ligand exchange. As described on Chapter 2, tetrabutylammonium salts were used as capping agent. To further investigate the effect of the capping agent on the PL stability, the ligand exchange with different thiols was studied for the small copper clusters: Cu13, as a convenient approach for producing functionalizated subnanometer copper clusters. Three different thiols were used: 1-hexanethiol, 3mercaptopropionic acid (3-MP) and L-glutathione (L-Glu). During the process, thiol is expected to replace TBANO3 because of the affinity of the sulfur to copper. On the two-phase method, the clusters are expected to be obtained in the organic phase. On the solid solubilization, the clusters would be completely solved, but the results have been not completely satisfactory. These thiols have a tendency to precipitate at ambient temperature, and consequently the clusters may precipitate with them. In spite of this, optical absorption and fluorescence spectra of the resulting supernatant solution obtained after the reaction was compared for both thiols in Figure 6.6. It can be observed that the supernatant absorption (gray line) for both thiols display an intensity increase (comparing with the initial thiol solution black line) around 285 and 270 nm for 3-MP and LGlu respectively (marked by arrows). However, it should be noted a blue emission shift from 408 to 300 nm (Figure 6.6). Figure 6. 6| UV-Vis and PL spectra of Cu13 solid clusters after thiol exposure (24 hours, 50ºC). 280 320 360 400 440 480 0.00 0.03 0.06 0.0 0.3 0.6 0 10 20 0 10 20 emi=300nm L-Glu PL intensity/AU Absorbance/AU Wavelength/nm Cu6 Cu5 exc=270nm emi=330nm exc=293nm 3-MP Photoluminescence Stability of CuCLs 122 This should be related with the Cu13CLs etching leading to the formation of smaller CuCLs, which by the Jellium model correspond to Cu6 and Cu5 centers. This study will involve a deeper investigation, but our preliminary results show a possible easy functionalization of these CuCLs needed in order to a specific bind into different molecules, substrates or surfaces. The two-phase exchange between 1-hexanethiol/toluene and Cu13CLs/H2O has proceed through seven hours vigorous stirring, observing an organic phase color change from colorless to light yellow. Figure 6.7.A shows the absorption spectra of the Cu13 clusters in the organic phase with absorption maximum band at 300 nm after the first hour of stirring and two absorption bands after 7 hours stirring located at 289 nm and 360 nm. The retention of the absorption profile of the aqueous copper clusters on 300nm after the phase transfer confirms the retention of the Cu13 core. An increase of the PL during the stirring time was also observed in the organic phase (Figure 6.7.B). The emission band of the phasetransferred clusters displayed a well-resolved vibronic structure with two main peaks at 397nm and 417nm (according to the polarity solvent decrease explained on the Figure 6.3). The phase transfer happen due to the great affinity of the thiol group for the Cu (0), therefore the electrostatic attracted tetrabutylammonium salt was replaced by the thiol. Figure 6. 7| UV-Vis spectra of the time dependent phase-transfer of the CuCLs (A) and excitation and emission spectra of the CuCLs at different stirring time (B). The obtained results clearly established that the emission is an inherent property of the copper core, and the same electronic transition can be accessed for both ligands (TBANO3 and C6SH). Although the excitation spectrum could be modified, emission always happen between the same energy levels, implying that they are metal derived. This phase transfer via electrostatic interactions is 300 400 500 600 0.00 0.02 0.04 0.06 7 hours . . 1 hour Absorbance/AU Wavelength/nm 300 400 500 600 0 20 40 60 80 100 120 em=420nm exc=350nm 7 hours . . 1 hour PL Intensity/AU Wavelength/nm A) B) Chapter 6 123 fast and versatile and can be carried out under ambient conditions. The previous results indicate again the high stability of our copper clusters which could be functionalizated by other thiols in order to introduce additional functionalities to the sub-nanometric fluorescent copper clusters. 6.3.3. Fluorescent CuCLs as ion nanosensors. Metal-selective fluorescent nanosensors are attracting extensive attention nowadays. Several metal ions have been studied by fluorescence quenching of the metal clusters, such as: Hg(II) 7 , Cr(III) 8 , Cu(II) 9 , Fe(III) 10 , Pb(II) 11 ,4a,4d. Fluorescent water soluble CuCLs was studied as potential nanosensors of different ions in solution. The CuCLs in water (Cu13) (Chapter 3) was very luminescent, however, in the presence of Pb+2, the fluorescence of the Cu13 clusters was found to be quenched by Pb+2, which can be used as a selective and reversible “turn on-off” indicator for Pb+2 (Scheme 6.2). Scheme 6. 2| Schematic illustration of the fluorescent Cu13 CLs for Pb+2 “off-on” detection. To further evaluate the analytical detection of Pb+2 different concentrations (0.048, 4.83, 12.1, 19.3, 28.96, 38.61, 57.92, 77.2 and 96.52 µM) were added to CuCLs sample. As it can be seen in Figure 6.8.A upon addition of increasing concentrations of Pb+2, a gradual decrease in the fluorescence was observed at 408nm. From Figure 6.8.B it can be seen that the fluorescence ratio (F0/F) is sensitive to the [Pb+2]. The Stern Volmer plot (typical for static or dynamic quenching) shows a modified form, which is associated with the presence of accessible and inaccessible fluorophores 12 , this result agrees with the data provided in Chapter 3, where the Cu13 CLs are identified as Cu13 but also exist smaller Cu cores in the sample. Hence, the Cu13 can be the “accessible” Photoluminescence Stability of CuCLs 124 fluorophore and the smaller clusters the inaccessible fluorophores. For a limited range of quencher concentrations, the modified Stern-Volmer plot appears to be linear (from 0.048 to 20 µM) (y=0.9136+0.0134[Pb+2], R2=0.9908). Figure 6. 8| Emission spectrum representing the quenching effect of different concentrations of Pb+2 on fluorescent Cu13CLs (A). Plot of fluorescence quenching (F0/F) of the CuCLs at 408 nm incubated with different concentrations of Pb+2(B). Next, we investigated the selectivity of our CuCLs toward Pb+2 ions. The data presented in Figure 6.9 were collected at 20 μM concentrations. The results confirm the high selectivity of the CuCLs toward Pb+2 ions. Other metal ions had minor or negligible quenching effects on the fluorescence intensity of the CuCLs. Larger CuCLs (Cu20) (Chapter 4) were also studied on the sensing of Pb+2, almost no PL quenching effect can be observed. This behavior is still under study, because the real mechanism of CuCLs PL quenching is not well understood. Some authors associated the quenching to possible interactions between the ion and the capping agent, but our case will need a deep study. Finally, reversibility and recyclability of the CuCLs as fluorescent sensors were evaluated. Fluorescence quenching also is also time dependent, and CuCLs with 500nM can be quenched within 72 hours. Subsequent irradiation by a UV lamp (λexc=254 nm) for 21 hours were able to recover the quenched fluorescence to its initial intensity (Figure 6.10). The recovery mechanism is believed to consist of the Pb+2 reduction by the etransfer from the excited cluster. However, other studies will be necessary in order to explain the quenching recovering cycles allowed by this system. A) B) 300 400 500 600 700 0 100 200 exc=312nm 0.048 µM 4.83µM 12.1µM 19.3 µM 28.96 µM 38.61 µM 57.92 µM 77.2 µM 96.52 µM PL Intensity/AU Wavelength/nm 020 40 60 80 100 0.8 1.0 1.2 1.4 1.6 0 5 10 15 20 0.9 1.0 1.1 1.2 y=0.9136+0.0134[Pb+2] R2=0.9908 F/F0 Pb+2/M y=1.711-0.792e-x/57.75 R2=0.9918 F0/F Pb+2/M Chapter 6 125 Figure 6. 9| Selectivity of the CuCLs towards 20 μM metal ions concentrations (λexc= 312 nm). . Figure 6. 10| PL spectra of CuCLs emission during the time and recovering of the PL intensity after exposure to UV radiation. ControlAl Fe Ni Pb Cu KZn -0.5 0.0 0.5 1.0 1.5 2.0 (I0-I)/I 300 400 500 600 700 0 20 40 60 exc=312 nm Cu13 CLs initial Cu13 CLs+Pb+2 (after 3 days) Cu13 CLs+Pb+2 (after 21h irradiation) PL Intensity/AU Wavelength/nm Photoluminescence Stability of CuCLs 126 6.4. SUMMARY. The bright emission, high photostability and chemical stability of CuCLs make them attractive alternatives to conventional fluorophores. CuCLs ligands exchange took place by using different thiols. Fluorescent small CuCLs have demonstrated satisfactory results on the detection of Pb+2 ions. Besides the inherent CuCLs photoluminescence can be recovered under UV light irradiation. This fact makes possible the use of the CuCLs as an “off-on recyclable nanosensors. Chapter 6 127 REFERENCES 1 a) Basché T.; Kummer S.; Bräuchle C. Nature 373, 132 (1995). b) Ha T.; Enderle T.; Chemla D. S.; Selvin P. R.; Weiss S. Chem. Phys. Lett. 271, 1 (1997). c) Yip W. T.; Hu D. H.; Yu J.; Vanden Bout D. A.; Barbara P. F. J Phys Chem. A 102, 7564 (1998). 2 Mooradian A. Phys. Rev. Lett. 22, 185 (1969). 3 Darugar Q.; Qian W.; El-Sayed A. M.; Pileni M.P. J.Phys.Chem B 110, 143 (2006). 4 a) Goswami N.; Giri A.; Bootharaju M. S.; Xavier P.L.; Pradeep T.; Pal S.K. Anal. Chem. 83, 9676 (2011). b)Xie J. P.; Zheng Y. G.; Ying J. Y. J. Am. Chem. Soc. 131, 888 (2009). c) Wei W. T.; Lu Y. Z.; Chen W.; Chen S. W. J. Am. Chem. Soc. 133, 2060 (2011). d) Kawasaki H.; Hamaguchi K.; Osaka I.; Arakawa R. Adv. Funct. Mater. 21, 3508 (2011). e) Zhang H.; Huang X.; Li L.; Zhang G.; Hussain I.; Li Z.; Tan B. Chem. Commun. 48, 567 (2012). 5 a) Yu P.; Wen X.; Toh Y.-R.; Tang J. J. Phys. Chem. C 116, 6567 (2012). b) van Wijngaarden J.T.; Toikkanen O.; Liljeroth P.; Quinn B.M.; Meijerink A. J. Phys. Chem. C 114, 16025 (2010). 6 a) Shibu E. S.; Muhammed M. A. H.; Tsukuda T.; Pradeep T. J. Phys. Chem. C 112, 12168 (2008). b) Shibu E. S.; Pradeep T. Int. J. Nanosci. 8, 223 (2009). 7 a) Adhikari B.; Banerjee A. Chem. Mater. 22, 4364 (2010). b) Xie J.; Zheng Y.; Ying J.Y. Chem. Commun. 46, 961 (2010). c) Wei H.; Wang Z.; Yang L.; Tian S.; Hou CH.; Lu Y. Analyst 135, 1406 (2010). d) Huang Ch-Ch.; Yang Z.; Lee K-H.; Chang H.-T. Angew. Chem. 119, 6948 (2007). 8 Liu S.; Lu F.; Zhu J.-J. Chem. Commun. 47, 2661 (2011). 9 a) Zhang M.; Ye B.-C. Analyst 136, 5139 (2011). b) Durgadas C. V.; Sharma C. P.; Sreenivasan K. Analyst 136, 933 (2011). c) Liu H.; Zhang X.; Wu X.; Jiang L.; Burda C. Zhu J.-J. Chem. Commun. 47, 4237 (2011). 10 Ho J.A.; Chang H.-C.; Su W.-T. Anal. Chem. 84, 3246 (2012). 11 Yuan Z.; Peng M.; He Y.; Yeung E.S. Chem. Commun. 47,11981 (2011). 12 Joseph R. Lakowicz, Principles of Fluorescence Spectroscopy, Springer, (2006).