ABSTRACT
FUNCTIONALIZING Au25 NANOCLUSTERS WITH CROWN ETHER LIGANDS FOR THE DETECTION OF DISSOLVED HEAVY METALS
Gold nanoclusters are an intermediate form between molecular and bulk gold. In some respects, they retain a physical appearance to gold nanoparticles but possess molecular-like properties unique to them. One notable example is the highly stable and symmetrical Au25(SCH2CH2Ph)18 (Au25) nanocluster, which has captivated many owing to its rich electronics and optical properties. Like their nanoparticle counterparts, gold nanoclusters are capable of undergoing surface modifications by ligand exchange. Here, we attempt to tailor Au25 through ligand exchange with crown ether ligands for the detection of bismuth(III), cadmium(II), lead(II), mercury(II), and thallium(III) by ion recognition methods. Crown ethers are well known for their chelating properties, in some cases generating “sandwich” complexes with appropriate ions. Here we attempt to exploit this property on the crown ether functionalized nanoclusters to induce aggregations, detectable by UV- Vis. Instead, we report the unusual outcomes of the ligand exchange reactions and unexpected reactions between the Au25 clusters and the active metal ions. In addition to these studies, we explore diffusion-ordered 1H-NMR spectroscopy (DOSY) as an alternative to transmission electron microscopy (TEM), an important technique used for determining particle diameter. DOSY addresses some of the limitations of TEM while providing equally, if not more, precise measurements.
Randy Espinoza May 2018
FUNCTIONALIZING Au25 NANOCLUSTERS WITH CROWN ETHER LIGANDS FOR THE DETECTION OF DISSOLVED HEAVY METALS
by Randy Espinoza
A thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Chemistry in the College of Science and Mathematics California State University, Fresno May 2018
© 2018 Randy Espinoza APPROVED For the Department of Chemistry:
We, the undersigned, certify that the thesis of the following student meets the required standards of scholarship, format, and style of the university and the student's graduate degree program for the awarding of the master's degree.
Randy Espinoza Thesis Author
Jai-Pil Choi (Chair) Chemistry
Alam Hasson Chemistry
Joy J. Goto Chemistry
For the University Graduate Committee:
Dean, Division of Graduate Studies AUTHORIZATION FOR REPRODUCTION OF MASTER’S THESIS
X I grant permission for the reproduction of this thesis in part or in its entirety without further authorization from me, on the condition that the person or agency requesting reproduction absorbs the cost and provides proper acknowledgment of authorship.
Permission to reproduce this thesis in part or in its entirety must be obtained from me.
Signature of thesis author: ACKNOWLEDGMENTS This work was not possible on my own. First, I would like to thank my mother and stepfather, family, and friends for providing support from home. In the lab, I’d like to thank the Der, Garret, Misk, and other undergraduate students for providing assistance and Monika and Logan for constant feedback and new insight related to research. Many mistakes were made along the way, and many experiments did not always go as planned, but in all cases, I had an advisor with great patience to help fix them, thank you, Dr. Choi. Along the journey there were many troughs along the way, some of which I discouraged me, thank you, Dr. Krishnan, for providing the much needed encouragement during those tough times. I’d also like to thank Dr. Ghosh and her group in UC Merced for much inspiration, the Bridges to Doctorate program and everyone in it for providing the tools necessary to get through this, my fellow peers who provided assistance, small or large, and the whole Chemistry’s staff, you are the best. TABLE OF CONTENTS Page
LIST OF TABLES ...... vii
LIST OF FIGURES ...... viii
INTRODUCTION ...... 1
Gold Nanoparticles ...... 1
Diffusion-Ordered 1H-NMR Spectroscopy Theory ...... 11
METHODS AND MATERIALS ...... 15
Synthesis of Crown Ether-Based Ligands ...... 15
Synthesis of Au25(SCH2CH2Ph)18 ...... 17
Ligand Exchange Reactions ...... 19
Metal Reactions with Au25 ...... 19
Differential Pulse Voltammetry ...... 20
Diffusion-Ordered 1H-NMR Spectroscopy ...... 20
RESULTS AND DISCUSSION ...... 22
- Synthesis of Au25(SCH2CH2Ph)18 and Crown Ether Ligands ...... 22
- Ligand Exchange Reactions of Au25(SCH2CH2Ph)18 with Crown Ether Ligands ...... 26
- Metal Reactions with Au25(SCH2CH2Ph)18 ...... 31 Diffusion Constants and Size Estimations by Diffusion-Ordered NMR Spectroscopy ...... 42
CONCLUSION ...... 52
REFERENCES ...... 55
LIST OF TABLES
Page
Table 1. Halfway potentials (V) of the Au25 products obtained using DPV...... 41 Table 2. Average diffusion coefficients (10-9 m2/s) from DOSY-NMR for Ferrocene (1.8 mM) and Au25 (70 µM) at different temperatures...... 49 Table 3. Average diffusion coefficients (10-9 m2/s) from DOSY-NMR for Ferrocene (1.8 mM) and Au25 (140 µM) at different temperatures ...... 49 Table 4. Average diffusion coefficients (10-9 m2/s) from DOSY-NMR for Ferrocene (1.8 mM) and Au25 (270 µM) at different temperatures...... 50
Table 5. The diameter of Au25 (nm) for each concentration and temperature calculated from DOSY-NMR using Equation 2...... 50
LIST OF FIGURES
Page
Figure 1. TEM image of gold nanoparticles ranging from 10 - 50 nm (left) and 20 nm gold nanoparticles (right)...... 2
Figure 2. The MALDI mass spectra (A) and UV-Vis (B) of various clusters...... 3
Figure 3. The SWV of various clusters...... 3
Figure 4. Reaction scheme of synthesis of gold clusters...... 5
- Figure 5. On the left in blue is the absorbance spectrum of Au25 and in red 0 is the absorbance of Au25 ...... 6
Figure 6. Complete ionic structure of [TOA][Au25(SCH2CH2Ph)18] (left.) ...... 6
Figure 7. Ligand exchange can occur at all three possible site...... 7
Figure 8. The energy barriers corresponding to the ligand exchange at site A...... 8 Figure 9. From left to right are 2-Hydroxymethyl-12-crown-4 (12C4), 2- Hydroxymethyl-15-crown-5 (15C5), 2-Hydroxymethyl-18-crown- 6(18C6), and 1-Aza-18-crown-6 (A18C6)...... 10
Figure 10. Scheme of the reversible metal-crown ether complex...... 10
Figure 11. A TEM time-lapse of AuNP aggregating...... 12
+ - Figure 12. UV-Vis absorption spectra of [TOA] [Au25(SCH2CH3)18] ...... 23 Figure 13. 1H-NMR of 2-[(6-Mercaptohexyl)oxy]methyl-12-crown-4 in CDCl3...... 24 Figure 14. 1H-NMR of 2-[(6-Mercaptohexyl)oxy]methyl-15-crown-5 in CDCl3...... 24 Figure 15. 1H-NMR of 2-[(6-Mercaptohexyl)oxy]methyl-18-crown-6 in CDCl3...... 25
1 Figure 16. H-NMR of N-(6-Mercaptohexyl) Aza-18-Crown-6 in CDCl3...... 25
- Figure 17. (a) The H-NMR spectrum of [Au25(SCH2CH2Ph)18] in acetone-d6. .. 27
1 Figure 18. H-NMR spectrum of Au25 after ligand exchange reaction with 2-[(6-Mercaptohexyl)oxy]methyl-12-crown-4 in CDCl3...... 27 ix ix Page
1 Figure 19. The H-NMR spectrum of Au25 after ligand exchange reaction with 2-[(6-Mercaptohexyl)oxy]methyl-15-crown-5 in CDCl3...... 28
1 Figure 20. The H-NMR spectrum of Au25 after ligand exchange reaction with 2-[(6-Mercaptohexyl)oxy]methyl-18-crown-6 in CDCl3...... 28
1 Figure 21. The H-NMR spectrum of Au25 after ligand exchange reaction with N-(6-Mercaptohexyl) Aza-18-Crown-6 in CDCl3...... 29
Figure 22. UV-Vis of products from the ligand exchange reactions of Au25 and (A) 12-crown-4, (B) 15-crown-5, (C) 18-crown-6, and (D) aza-18- crown-6 ligands...... 30
Figure 23. UV-Vis spectrum of Au24Hg. The spectrum of the cluster was obtained in DCM, and the absorbance was normalized...... 32
Figure 24. UV-Vis spectrum of Au24Cd. The spectrum of the cluster was obtained in DCM, and the absorbance was normalized...... 34
0 Figure 25. UV-Vis spectrum of Au25 in DCM oxidized by lead(II)...... 35
0 Figure 26. UV-Vis spectrum of Au25 DCM after reaction with thallium(III)...... 35
+ Figure 27. UV-Vis spectrum of Au25 in DCM oxidized by bismuth...... 36
Figure 28. DPV of Au25 in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode...... 37
Figure 29. DPV of Au24Hg in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode .... 38
Figure 30. DPV of Au24Cd in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode...... 39
Figure 31. DPV of Au25 reacted with lead(II) in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode...... 39
Figure 32. DPV of Au25 reacted with thallium(III) in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode...... 40
Figure 33. DPV of Au25 reacted with bismuth(III) in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode...... 40
Figure 34. The DOSY-NMR spectrum of Au25 and ferrocene plotted as chemical shift (ppm) vs. gradient pulse amplitude (gz) from 1-24 or 8.72 to 410. mT/m...... 42 x x Page
Figure 35. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars...... 44
Figure 36. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars...... 44
Figure 37. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars...... 45
Figure 38. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars...... 45
Figure 39. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars...... 46
Figure 40. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars...... 46
Figure 41. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars...... 47
Figure 42. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars...... 47
Figure 43. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars...... 48
INTRODUCTION
Gold has been a valued noble metal throughout early human history for its luster, malleability, and resistance to oxidation and has found use as a form of decorative ornaments, jewelry, and as currency for centuries. Its niche use as a vibrant ruby-red stain in glass making has been used in many cultures, an unusual property of gold not fully understood until the dawn of the 20th century. Michael Faraday’s work on colloidal gold and gold leaves helped shed light on these peculiar properties. He observed a relationship between the size and structure of the gold leaves and particles and their optical and electronic properties, but the physical mechanisms behind those observations were not well understood until later in the mid-20th century.1 These early findings paved the way to what is today is the study of nanomaterials.
Gold Nanoparticles Right around the early 1950s, two decades after the invention of the early electron microscopes, Turkevich and coworkers examined and characterized popular gold colloids (see Figure 1) through transmission electron microscopy (TEM).2 Most gold colloids are beyond the resolving limits of light microscopes, but TEM uses electrons rather than light to image particulates below 500 nm. The techniques used became fundamental for the study of all nanomaterials and allowed the field to expand further. Throughout the second half of the century, the study of metal nanoparticles exploded with many incredible advancements. Improvements in the synthesis of gold nanoparticles (AuNP) of different shapes and sizes with unique characteristics is one such example.3-4 2 2
Figure 1. TEM image of gold nanoparticles ranging from 10 - 50 nm (left) and 20 nm gold nanoparticles (right). Note: A mixture of spherical, flat triangular, and hexagonal particles (left) produces a dark blue- purples solution while the spherical particles (right) give a red solution. Reproduced in part from reference ref. 2 with permission of The Royal Society of Chemistry.
The general direction steered towards synthesizing smaller and smaller gold nanoparticles, eventually leading to the discovery of gold nanoclusters (AuNC), a form of gold smaller than AuNPs (< 3nm in diameter).5 These AuNCs possess both molecular-like and bulk-like properties, therefore are said to be the bridge between the two forms of gold. Bulk like properties manifest at diameters closer to the upper limit (ca. 3 nm), but at 1 nm molecular-like properties become dominant.5-6 At the lower limits, their optical absorption is a result of the discrete energy gaps found in AuNC rather than surface plasmon resonance. This is the dominant mechanism behind AuNP’s optical absorption.7 This trend is seen in
Figure 2 where small clusters like Au25 show a different absorption “fingerprint,”
8 but more massive clusters Au333 begin to manifest the iconic 540 nm plasmonic. The electrochemical energy gaps of the clusters decrease as the clusters become more massive as seen in their voltammograms in Figure 3, further supporting the molecular-bulk bridging notion. Furthermore, AuNCs have shown unique properties not found in the larger AuNPs like the catalytic capability for alcohols, carbon monoxide, and other small organic compounds.9-13 Their simplicity has 3 3 allowed the determination of their exact structure by density functional theory calculations (DTF) shedding some light into mechanisms of cluster growth, surface passivation, and ligand exchange.14-15 In turn, this has helped improve experimental yields.
Figure 2. The MALDI mass spectra (A) and UV-Vis (B) of various clusters. Note: The absorption spectra were transformed from wavelength to energy regarding eV. Adapted with permission from ref. 8. Copyright 2017 American Chemical Society.
Figure 3. The SWV of various clusters. Note: The focus in this illustration is the number of current peaks and the redox potentials. Therefore the intensity of the current peaks is made arbitrary. Adapted with permission from ref. 8. Copyright 2017 American Chemical Society. 4 4 Gold Nanoclusters The trend over the past few decades has steered towards the synthesis of small, atomically-precise nanoparticles, also known as nanoclusters. Gold nanoclusters are distinguished by their exact chemical formula, Aum(SR)n, where m is the number of atoms per cluster and n is the number of thiolate-ligands protecting it. Nanoclusters are unique as they offer critical structural information their nanoparticle counterparts cannot provide such as exact molar composition, molecular formula, and an exact molar mass. Brust and coworkers developed a more straightforward and safer one- and two-phase method for synthesizing AuNCs with appreciable yields.4 Since then, many have made improvements to the method with yields up to 50% of the desired cluster.16-19 The synthesis of AuNCs by the Brust method (synthetic scheme in Figure 4) typically begins with solubilized Au(III) reduced to Au(I) by a thiol-ligand.4 This leads to the formation of Au(I)-ligand polymers, which are further reduced by a strong reducing agent forming Au(0) aggregates. These aggregates are the seeds for the formation of AuNCs and AuNPs. Remaining Au(I)-complexes serve as protective capping agents for the Au(0) aggregates, i.e., passivation. The size of AuNCs is determined by taking control over conditions like the molar ratio between chloroauric acid and the thiolate-ligand, the concentration of the reducing agent, the temperature of the reaction, and the duration of the synthesis. Fine control over these variables is critical for the precise yield of desired clusters.16 Atomically precise gold nanoclusters come in many geometrical configurations and with an exact atomic count. The most common of which are those with a ‘magic number’ like Au10, Au18, Au25, Au38, Au102, and Au140, the subscript denoting the number of gold atoms per cluster.6 These clusters are the products of the etching of unstable clusters during the AuNC synthesis, the Au25 5 5
Figure 4. Reaction scheme of synthesis of gold clusters. Note: After Au(III) has been transferred over into the organic phase, it is reduced by thiolate ligands to Au(I) (ii). In this vital step, both stir speed and temperature are carefully controlled as it will determine the cluster size and yield. Finally, the Au(I)-thiolate polymers are reduced by sodium borohydride to yield the thiolate-passivated gold nanoclusters (ii). Adapted with permission from ref. 16. Copyright 2017 American Chemical Society.
21 - + cluster being one of the most stable. The cluster Au25(SCH2CH2Ph)18 (TOA being the counter ion), Au25 for short, is a highly symmetrical nanocluster available in the oxidation states z = 1-, 0, and 1+, the anionic form being most
22 common. The charge state of the Au25 cluster has a direct influence on its absorption spectrum evident in Figure 5. The Au25 nanocluster can be dissected structurally into two groups as seen in Figure 6: the gold core and -S-Au-S-Au-S- staples.7, 23 The gold core consists of thirteen gold atoms arranged into an icosahedron structure with one central gold atom. The twelve icosahedral gold atoms are bound to six -S-Au-S-Au-S- “staples” along the x-y-z coordinates. Although these staples function to stabilize the cluster, like AuNPs, the ligands can be replaced by ligand exchange in the presences of free thiol-ligands, which alter the cluster’s overall chemical and physical behavior.14, 21, 24-27 The ligand exchange mechanisms described here were solved by various methods through a combination of DFT methods, crystallography, and extended X-ray absorption fine structure (EXAFS).14-15, 28 Ligand exchange can be initiated at three different reaction sites (see Figure 7). For clarity, the gold atoms directly bonded to the -S-Au-S-Au-S- staple motif are referred to as “core gold,” the 6 6
- Figure 5. On the left in blue is the absorbance spectrum of Au25 and in red is the 0 absorbance of Au25 . + 20 Note: On the right is the absorbance of Au25 . Reproduced in part from ref. 20 (right figure) with permission of The Royal Society of Chemistry. Adapted with permission from ref. 21 (left figure). Copyright 2017 American Chemical Society.
Figure 6. Complete ionic structure of [TOA][Au25(SCH2CH2Ph)18] (left.) Note: In orange are the thirteen icosahedral gold atoms stapled by six Au2S3 motifs, the gold atoms in light orange and sulfur atoms in yellow. On the left, both the TOA cation and the ligands (gray) are shown as stick models while on the right they have been omitted for clarity.
7 7 two gold atoms in the staple as “staple Au,” the two outer sulfur atoms as “terminal SR,” and the central sulfur as “central SR.” The incoming thiol is denoted as SHR’ to highlight the presence of a hydrogen as it too participates in the ligand exchange.
Figure 7. Ligand exchange can occur at all three possible site. Note: The incoming thiol (SHR’) may initiate the exchange by interacting with the core Au and terminal SR (A), the staple Au and terminal SR (B), or the staple Au and central SR (C). For simplicity, the bonds between the staple Au and core gold atoms have been omitted for clarity.
The ligand exchange mechanisms can be challenging to follow as many steps are involved, but they can be simplified into three critical steps. First, a bond between either the central/terminal SR and staple Au/gold core atom is broken to make room for the incoming thiol. New bonds are formed generating an intermediate staple motif, which consists of both the incoming thiol and the central/terminal SR. Finally, the central/terminal SR is liberated as a thiol (SHR). Fernando and Aikens’ computational work discuss three reaction pathways in great detail. The most favorable reaction pathway is B (shown in Figure 8) due to a lower second energy barrier of 0.76 eV. In comparison, the least favorable reaction pathway (not shown here) has the highest second energy barrier of 1.15 8 8 eV. This explains why not all the ligands on the cluster completely exchange in most reactions. Figure 8 illustrates the energy barrier of site B, reactants, the transition states, the intermediate, and the products.
Figure 8. The energy barriers corresponding to the ligand exchange at site A. Note: A simple model of the Au25 and the incoming thiol are shown to illustrate bond breaking, bond formation, transition states, and the intermediate. The ligand exchange energies and mechanisms of sites B and C are not shown here, but both follow a similar pathway with different energies. Adapted with permission from ref. 15. Copyright 2017 American Chemical Society.
In the past, metal nanoparticles have been tailored for chemical sensing by the replacement of the surface ligands with other ligands modified with interacting motifs.29-31 As previously reported, the interactions between the surface modified nanoparticle have been observed as a change in their optical absorption. Considering that both gold nanoclusters and nanoparticles share similar physical 9 9 attributes such as the ability to undergo ligand exchange reactions, Au25 could be considered as suitable candidate for chemical sensing of heavy metals.32-33 The major objective is to modify the cluster with metal-chelating ligands like crown ethers and investigate the interactions between the target analyte and the cluster. Sensing selectivity is determined by the size and composition of the crown
34 ether motif. The goal is to develop highly selective and sensitive Au25-based chemical sensors modified with 12-crown-4, 15-crown-5, 18-crown-6, and aza-18- crown-6. These sensors exploit crown ether’s chelating properties for the detection of three heavy metals, cadmium, mercury and lead. Crown ethers have attracted attention for their phase transfer capabilities. Their highly electronegative cavities allow the ability to selectively and reversibly chelate a wide range of cations while remaining soluble in non-aqueous mediums. Small crown ethers such as 2-hydroxymethyl-12-crown-4 (see Figure 9) possess a smaller cavity, which preferably chelates cations like sodium, while 2- hydroxymethyl-18-crown-6 with larger cavities chelates larger cations such as potassium and is reflected their corresponding dissociation constants. For 1-aza- 18-crown-6 one would expect similar chelation like its 18-crown-6 counterpart. Instead, it possesses a higher affinity for protons due to the favorable formation of a protonated tertiary ammine.32 As shown in Figure 10, a higher affinity for protons can be useful as it can be used to separate the chelated metal under acidic conditions. The formation of “crown-cation-crown” sandwich complexes are also possible when the size of the cavity and size of the cation are appropriate. Others have exploited this property in crown ether functionalized gold nanoparticles to induce semi-permanent aggregations of AuNPs. Close proximities between
10 10
Figure 9. From left to right are 2-Hydroxymethyl-12-crown-4 (12C4), 2- Hydroxymethyl-15-crown-5 (15C5), 2-Hydroxymethyl-18-crown-6(18C6), and 1- Aza-18-crown-6 (A18C6). Note: These crown ethers will serve as the sensing site of Au25 for the metals mercury, cadmium, and lead.
Figure 10. Scheme of the reversible metal-crown ether complex. Note: First, a metal cation forms a complex with the crown ether, but under acidic conditions, the protonated crown ether strips the metal away. The protonated crown ether can be deprotonated with a base). Adapted with permission from ref. 32. Copyright 2017 American Chemical Society.
11 11
AuNPs leads to an observable and measurable optical response. In a similar sense, the goal is to functionalize Au25 with 12C4, 15C5, 18C6 and A18C6 ligands for the sensing of heavy metals. The advantage of using AuNCs over AuNPs is the additional information AuNCs provide (e.g., exact molecular formula), which could lead to detection limit improvements.
Diffusion-Ordered 1H-NMR Spectroscopy Theory The lack of immediate access to a TEM during these investigations delayed key observations. This instrument is an invaluable tool for size determination, elemental analysis, and structure determination. Unfortunately, not every academic institution has access to one, let alone a high-resolution TEM for analysis of ultra-small particles (~1 nm). If available, users are required to undergo proper training before operating the instrument, sometimes inaccessible to undergraduates. Additionally, AuNCs are sensitive and tend to form aggregates with neighboring clusters under extended exposure to the TEM electron beam,35 sometimes resulting in skewed data as shown in Figure 11. A solution to overcome this challenge is diffusion-ordered nuclear magnetic resonance (DOSY-NMR), which could be used as an alternative technique for size determination of ultra- small particles such as Au25. This NMR technique is appealing to many as it is widely available, does not require intense training, and the integrity of Au25 is preserved. Diffusion-ordered NMR spectroscopy is a powerful technique used to measure the diffusion coefficients of small molecules in solution.36 DOSY-NMR has been used to calculate their diffusion coefficients and approximate the
12 12
Figure 11. A TEM time-lapse of AuNP aggregating. Note: The image on the right was taken immediately after the one on the left. Courtesy of Amaral and Ghosh for the assistance. hydrodynamic radius of molecules, including AuNPs,37 by the Stokes-Einstein equations, where diffusion is calculated from
퐷 = 푘푏푇 (1) 6휋휂푟ℎ
2 -1 Where D is the diffusion constant (m s ), kb is the Boltzmann constant (1.38064852 × 10-23 m2 kg s-2 K-1), T is temperature in Kelvin, 6 is a coefficient obtained as the molecular radius approaches 1.0 nm, η is the viscosity of the solvent, and rh is the hydrodynamic radius of the molecule. Equation 1 can be rearranged when a standard, such as a ferrocene, with a known hydrodynamic radius, is used to give
푟퐹푐퐷퐹푐 푟푁퐶 = (2) 퐷푁퐶
Where rNC is the radius of the nanocluster, rFc is the radius of ferrocene, DFc is the diffusion of ferrocene, and DNC is the diffusion of the nanocluster in nanometers. The radius is obtained assuming the molecule or particles are nearly 13 13 spherical. Note that the other parameters can be ignored as both the AuNC and ferrocene diffusion coefficients are obtained simultaneously in the exact environmental conditions. In DOSY-NMR, a series of 1D spectra with decreasing signal strength is obtained and plotted to give the diffusion coefficients. This decay is a result when a sequence of radio-frequency pulses is applied to an increasing gradient magnetic field. This mode, available in most modern NMRs, is known as pulsed field gradients (PFG) and is integral to DOSY-NMR. Because diffusion is an inherent property of molecules in solution phase, PFGs allow diffusion to influence the measurements, seen as a decrease of the signal with increasing gradients. This is due to the improper homonuclear alignment of the molecules within the applied magnetic field, where some regions experience a lesser influence. The decay of the signal with increasing gradient strength is exponentially plotted as
푆 = 푆° 푒−퐷푞2훥′ (3)
Known as the Stejskal−Tanner Equation where D is the diffusion coefficient, S is the NMR signal adjusted for the diffusion delay, S° is the NMR signal, D is the diffusion of the molecule or particle (m2s-1), q is the area of the gradient pulse, and Δ’ is the corrected diffusion delay (s). The value q is obtained from γgzδ, where γ is the gyromagnetic ration of hydrogen, gz is the gradient amplitude in tesla (Tm-1) along the z-axis, and δ is the gradient pulse width (10-3s). The corrected diffusion delay can be calculated by
훿(훼2+3훼−2) 휏(훼2+2훼−1) 훥′ = 훥 + + (4) 6 2
14 14
Where Δ is the diffusion delay (seconds), α is 0.2, and τ is the spin echo (1 × 10-3s). The diffusion delay is a value that may be different per experiment. Once the diffusion coefficient is determined, then the size of the AuNC can be determined relative to the ferrocene standard.
METHODS AND MATERIALS
Synthesis of Crown Ether-Based Ligands This was the protocol used for 12-crown-4 and 15-crown-5 analogs. First, 1.0 g of 2-hydroxymethyl-18-crown-6 (2.4 mmol) and 0.345 g of NaH (60% in mineral oil; 14.6 mmol) were allowed to stir for 30 minutes in 20 mL of 1,2- dimethylformamide (DMF). Then, 2.72 mL of 1,2-dibromohexane (17.7 mmol) was added to the solution and allowed to stir overnight at room temperature. The reaction is quenched with methanol, and the DMF is evaporated under reduced pressure. The crude product is dissolved in 100 mL of dichloromethane (DCM), then washed thrice with 100 mL of DI water. The solvent is dried with MgSO4 followed by evaporation under reduced pressure. The remaining oily product is purified by column chromatography. The intermediate, 2-[(6- bromohexyl)oxy]methyl-18-crown-6, and 0.8073 g of thiourea (10.61 mmol) were both dissolved in 50 mL of ethanol and heated to reflux for 16 hours. The solution is removed from heat and the solvent is evaporated under reduced pressure. The crude and 0.6556 g of NaOH (16.39 mmol) are both dissolved in nanopure water and allowed to reflux for an additional 2 hours. Again, the solution the solution is removed from heat then acidified with 1M HCl. The crude is extracted from the aqueous solution using 100 mL of DCM, which is then washed thrice with 100 mL of DI water. The organic solvent is dried once more with MgSO4 then the solvent is evaporated under reduced pressure. The final product (clear oily residue) is purified by column chromatography to yield 2-[(6-Mercaptohexyl)oxy]methyl-18- crown-6. The final yield was 0.4607 g (33% yield). 16 16 Synthesis of N-(6-Mercaptohexyl) Aza-18-Crown-6 Briefly, 0.4878 g of 1-aza-18-crown-6 (1.85 mmol), 3.2 g of 1,6- dibromohexane (13 mmol), and 10.2009 g of sodium carbonate (96.2451 mmol) were dissolved in 35 mL of acetonitrile. The mixture was heated to 80 °C and refluxed for 72 hr. The solution was then filtered and the solvent evaporated under reduced pressure. The crude was taken up with 100 mL of nanopure water and extracted thrice with 50 mL of dichloromethane. The organic solution was dried with MgSO4 and evaporated under reduced pressure to yield an oily product. Like the other crowns, the thiol was introduced by thiourea. Next,1.4010 g of thiourea (18.405 mmol) and 1.566 g of the intermediate was dissolved in 50 mL of ethanol and heated to reflux for 18 hours; note that the yield of this intermediate is much higher than calculated possibly due to the presence of unreacted 1,6- dibromohexane therefore an excess amount of thiourea was used. The mixture was removed from heat and the ethanol was evaporated under reduced pressure. The crude was brought up with 100 mL of DI water and 1.0605 g of NaOH was added to the mixture and allowed to reflux for 2 hours. Then, the mixture was neutralized using concentrated HCl. DCM was used to extract the product but was not successful. Most of the 1,6-hexanedithiol byproduct was removed during the extraction and the product remained in the aqueous phase. Unexpectedly, the product remained in the aqueous phase was acidified once again down to 2 pH and extracted three times using DCM with some success. The solvent was dried under vacuum leaving behind 31.80 mg of the product, a yield of 5%. This incredibly low yield is due to loss of the product during the DCM extractions; limiting starting reagent prevented further optimization. The crude product was purified using a silica gel column with the mobile phase being 1:1 hexane and ethyl acetate. 17 17
Synthesis of Au25(SCH2CH2Ph)18 To date, there is an abundance of methods available for synthesizing a wide variety of AuNCs. Here we describe the methods developed by Brust et al. which yield the phenylethylthiolate-protected clusters.
- Synthesis of Au25(SCH2CH2Ph)18 Nanocluster: Two-phase Method Of many methods available for synthesizing AuNPs, the Brust two-phase method was initially chosen for its selective yield of Au25, Au140 and Au309 nanoclusters. In this protocol chloroauric acid (HAuCl3, 2.9945g) dissolved in nanopure water and tetraoctylammonium bromide (TOAB, 5.0077g) dissolved in toluene is allowed to stir together for 30 minutes. During this step, the gold ions are transferred over to the organic phase by the TOAB, a phase transfer catalyst. When complete, the yellow aqueous layer turns clear and the organic phase acquires a dark orange color. At this step, the aqueous layer is removed and stirring is slowed down. The following steps are critical for the formation of Au25 clusters and careful attention must be taken. 2-Phenylethanethiol (PET, 3.5mL) is added and allowed to continue stirring for another hour at 0 °C in the absence of light. During this period, gold(III) is reduced to gold(I) forming an intermediate complex with the thiol which is observed as a change in color from orange to colorless. After an hour, stirring is increased (1100 rpm) and an ice-cold solution of sodium borohydride (NaBH4, ~4g) is added and allowed stir overnight. This strong reducing agent drives the formation of thiol-protected gold clusters by the reduction of the thiol-gold(I) complex to gold(0). It is critical to maintaining the reaction at 0 °C otherwise the formation of larger gold clusters will occur. After completion, AuNPs formed are separated and purified by size through fractional precipitation. First, toluene is removed under reduced pressure and the slurry left 18 18 behind is washed with 50 mL of methanol. This is repeated several times removes unreacted free thiols, TOAB, and any other impurities formed during the reaction.
Next, Au25 is extracted first from the crude product with acetonitrile (ACN).
Clusters of Au25 are soluble in more polar solvents while larger clusters like Au140 are not and are extracted with less polar solvents such as dichloromethane or toluene. This step is repeated however many times as necessary.
Synthesis of Au25(SCH2CH2Ph)18 Nanocluster: One-phase Method This one-phase method was later implemented as it was found to have good yields of Au25 nanoclusters. All glassware should be thoroughly washed with aqua regia and rinsed with nanopure water. First, 0.2044 g of HAuCl4•3H2O (0.5190 mmol) and 0.3253 g of tetraoctylammonium bromide (TOAB, 0.5949 mmol) were dissolved in 20 mL of N2-purged tetrahydrofuran (THF) and stirred at 900 rpm for
15 mins under N2 atmosphere. The solution turns red indicating successful dissolution of gold chloride ions into THF. Then, stirring speed is reduced to 30 rpm and 0.3472 g (2.512 mmol) of phenylethanethiol in 1 mL of THF is added to the reaction dropwise. In this key step, the Au(I)-thiolate polymers formed dictate the formation of Au25 clusters. Therefore it is important to allow slow stirring throughout this step for one hour until the solution turns clear. Next, 0.1953 g of
NaBH4 (5.163 mmol) dissolved in 10 mL ice-cold, nanopure water is added all at once and stirring speed is increased to 900 rpm, the solution acquiring a dark- brown color. In this step, Au(I)-thiolate complexes are reduced to Au(0), and the formation of nanoclusters begins. The formation of Au25 clusters is monitored by
UV-Vis spectroscopy; Au25 spectrum development slows after 4 hours. After 9 hours, stirring is stopped, and the two-phase solution is brought up with 15 mL of DCM. This allows easy separation of the two phases. The organic phase is 19 19 collected and washed three times with 10 mL of nanopure water. The DCM is dried by rotary evaporation, and the oily-crude product is washed 5 times with 10 mL of methanol to remove excess thiols and TOAB; TOAB can be difficult to remove. Alternatively diethyl ether can be used to separate the phase transfer agent from the clusters effectively. The Au25 clusters are extracted with ACN 2-3 times for the recovery of 60 mg Au25 (40% yield by gold atomic basis).
Ligand Exchange Reactions This method applies to all ligands used in this study. First, a 1:3 molar ratio of Au25 (0.0103 g) and 2-[(6-Mercaptohexyl)oxy]methyl-18-crown-6 (0.0438 g) were weighed into a clean vial. The mixture is dissolved in 10 mL of N2-purged DCM and allowed to stir rapidly for 3 days, at room temperature. The dead spaces above the solvent are blanketed with N2 to limit the oxygen from interfering. The solvent is evaporated under reduced pressure, and the crude is washed three times with 10 mL of methanol or hexane to remove any free thiols. Methanol was used for the 12-crown-4 and hexane was used for 15-crown-5 and 18-crown-6 ligand exchange reactions because the nanoclusters with the latter two ligands were now soluble in polar solvents. The final product was dried under vacuum for further analysis. The yield was 17.1 mg (yield of 83.1%).
Metal Reactions with Au25 The following protocol was used for all the reactions (except for cadmium).
1:1 molar ratio between Au25 (0.0059 g) and 0.218 mL of HgCl2 solution were measured into a clean pre-weighed vial. The metals were previously prepared as a solution of 1 mg/mL. The mixtures were dissolved in 10 mL of degassed ACN and stirred rapidly at room temperature at different times: 15 mins (thallium), 30 minutes (mercury), and 24 hours (bismuth and lead). The reaction is complete 20 20 once all the product precipitates from solution. At this point, the precipitate can be collected by centrifugation and washed three times with 10 mL ACN to remove unreacted reagents. The final product is dried under vacuum for further analysis. For cadmium, the reaction was performed like before, except a 1:10 molar ratio was used. The reaction was allowed to proceed at 60 °C for 30 minutes. The crude was then rinsed three times with 10 mL of ACN, then extracted with hot (ca. 60 °C) ACN at least three times or until colorless. The product was dried under vacuum and saved for further analysis.
Differential Pulse Voltammetry Roughly 5–12 mg of sample was used per analysis; reaction yields were a limiting factor. Samples were dissolved in 5-7 mL of DCM (N2 purged for 15 mins) containing 0.1 M tetrabutylammonium hexafluorophosphate (TBAPF6; working electrolyte). A blanket of N2 was added over the solvent to prevent oxygen gas from dissolving back into the solution. The cell consisted of the following electrodes: 1 mm platinum disk (working electrode), Ag/AgCl (reference electrode), and a fine platinum wire-mesh (auxiliary electrode). The working electrode was thoroughly polished prior to use using 1.0 µM, 0.3 µM, and 0.05 µM alumina solutions. The parameters used were: scan range from -0.4 V to 1.2 V, increments of 0.01 V, a sample width of 0.0167 s, pulse period of 0.2 s, a pulse amplitude of 0.05 V amplitude, and pulse width of 0.05 s.
Diffusion-Ordered 1H-NMR Spectroscopy All measurements were made using the Varian 400 MHz spectrometer with a double-resonance pulse-field resonance pulsed-field-gradient probe, a shielded z gradient coil, and a gradient amplifier. The sample concentrations used for Au25 were 0.5 mg/mL (70 µM), 1.0 mg/mL (140µM), and 2.0 mg/mL (270 µM), the 21 21 volume being 600 µL, each prepared in replicates (nine in total). All the samples were prepared in acetone-d6 containing 0.33 mg/mL (1.8 mM) of ferrocene standard. Measurements were conducted at temperatures of 293.15 K, 298.15 K, and 303.15 K. Before every measurement the 90 ° pulse width was calibrated and the diffusion delay (Δ) optimized for concentration or temperature; note changes such as this apply to all corresponding replicates. The gradients used ranged from 1.395 G/cm to 33.49 G/cm in increments of 1.395 G/cm; 800 to 19,200 in increments of 800 in terms of gzlvl1. Other parameters used include a gradient pulse duration of 2 ms, the spectral width 4036 Hz, and the acquisition time was 2.02 s, a spectral width of 4,058 Hz, an imbalance factor of 0.2 (α), and the time period in between midpoints of the gradient pulses (τ) was 1 ms.
RESULTS AND DISCUSSION
- Synthesis of Au25(SCH2CH2Ph)18 and Crown Ether Ligands
Synthesis of Au25(SCH2CH3)18
Two methods for synthesizing Au25 nanoclusters follow the Brust two- phase method or the one-phase method recently reported by Kwak et al. The first method yields appreciable amounts of gold nanoclusters (~10% yield) and the second method yields up to 40%. By yields alone, the latter method is the preferred method for synthesizing Au25 only. The advantage of the two-phase route is the production of a wider range of clusters such as Au309, Au140, Au25, and trace amounts of other uncommon clusters. Because here we focus on Au25, the latter is preferred. Not only does the latter method yield greater quantities of Au25, but the method also proves to be simpler and faster. The reaction can be performed at room temperature, comes to completion within one day’s work, and extraction and purification of the clusters require many times fewer steps. In summary, the one phase method is simple, requires over less than half the amount of time, and cost-effective.
The Au25 nanoclusters can easily be separated from the other clusters by fractional separation. Almost all the nanoclusters produced by the Brust method and its variations are all soluble in nonpolar solvents such as dichloromethane and toluene, but almost all are sparingly or not soluble in polar solvents such as water, methanol, and acetonitrile. One of the few exceptions to this rule is Au25. The Au25 clusters are moderately soluble in acetonitrile, which allows for easy separation by fraction precipitation with excellent purity. The UV-Vis spectrum of pure Au25 is determined by the well-defined absorption bands at 680, 450, and 400 nm, as 23 23 shown in Figure 12. Another subtle band at 790 nm and a shoulder at 320 nm are reliable indicators of high purity Au25. The concentration of Au25 can be determined by calculating the molar extinction coefficient if necessary. While not observed by UV-Vis, trace amounts of minor nanoclusters are present but have not been observed to interfere with any of the experiments significantly, therefore are ignored.
+ - Figure 12. UV-Vis absorption spectra of [TOA] [Au25(SCH2CH3)18]
Synthesis of Crown Ether Ligands The NMR spectrum for the synthesized ligands is shown in Figures 13-16. For all the thiol ligands prepared, the crown ether signals appear as multiplets between 3.6 ppm to 3.8 ppm corresponding to the cyclic ether. Size of the crown does not appear to induce a chemical shift nor alter the line shape significantly to discuss here. The central protons within the alkane linker for all the ligands appear as multiplets between 1.7 ppm and 1.3 ppm. At 2.5 ppm, the presence of a triplet should confirm the presence of a thiol group at the end of the linker; this signal corresponds to the hydrocarbon directly bonded to the thiol. Note that Figure 16 shows intense alkane peaks, an indication of lower purity. 24 24
1 Figure 13. H-NMR of 2-[(6-Mercaptohexyl)oxy]methyl-12-crown-4 in CDCl3.
1 Figure 14. H-NMR of 2-[(6-Mercaptohexyl)oxy]methyl-15-crown-5 in CDCl3.
25 25
1 Figure 15. H-NMR of 2-[(6-Mercaptohexyl)oxy]methyl-18-crown-6 in CDCl3.
1 Figure 16. H-NMR of N-(6-Mercaptohexyl) Aza-18-Crown-6 in CDCl3.
26 26 - Ligand Exchange Reactions of Au25(SCH2CH2Ph)18 with Crown Ether Ligands Crown ethers are well-known phase transfer catalysts due to their excellent chelating capabilities. The size of the crown and its atomic composition are contributing factors that determine the selectivity of the crown, e.g., 18-crown-6 is an excellent chelator of potassium ions while 12-crown-4 is better suited for chelating sodium ions.34, 38 Thiol modified crown ethers have been successfully used in ligand exchange reactions AuNPs for the detection of alkali ions and
30 melamine in solution. Like AuNPs, Au25 are capable of undergoing ligand exchange reactions; our group has previously studied ligand exchange using an
39-40 assortment of alkanethiols. Here we attempt to tailor Au25 with crown ether ligands by ligand exchange reactions for the detection of dissolved heavy metals. The ligand exchange reactions were allowed to proceed for a period of three nights to allow enough time for the ligands to exchange to completion. Ligand exchange yield for lengthy ligands is a little over 60% as reported by Alawdi.40 The ligand exchange was monitored by 1H-NMR as shown in Figures
1 18-21. Figure 17 is the H-NMR spectrum of Au25 before ligand exchange where the phenyl is observed from 7.1 ppm to 7.3 ppm (Figure 17A), the α-CH2 peak at
+ 3.0 ppm and the β-CH2 at 3.3 ppm (Figure 17B). For the TOA counterion, is observed at 0.9 ppm (t, CH3), 1.3 ppm (m, CH2), 1.5 ppm (m, CH3), and 3.5 to 3.6 ppm (m, α-CH2) in Figure 17C. The signal intensity for the PET ligand is reduced after ligand exchange as evident in figures 18-21, where the aromatic signals broadened and suppressed beyond recognition. For all ligand exchange reactions, the crown ether peaks emerged from 3.6 ppm to 3.8 ppm, an indication of a successful ligand exchange. 27 27
- Figure 17. (a) The H-NMR spectrum of [Au25(SCH2CH2Ph)18] in acetone-d6. Note: For clarity, the lower two spectra (b and c) are expanded regions of the upper spectrum.
1 Figure 18. H-NMR spectrum of Au25 after ligand exchange reaction with 2-[(6- Mercaptohexyl)oxy]methyl-12-crown-4 in CDCl3. 28 28
1 Figure 19. The H-NMR spectrum of Au25 after ligand exchange reaction with 2- [(6-Mercaptohexyl)oxy]methyl-15-crown-5 in CDCl3.
1 Figure 20. The H-NMR spectrum of Au25 after ligand exchange reaction with 2- [(6-Mercaptohexyl)oxy]methyl-18-crown-6 in CDCl3.
29 29
1 Figure 21. The H-NMR spectrum of Au25 after ligand exchange reaction with N- (6-Mercaptohexyl) Aza-18-Crown-6 in CDCl3.
The 1H-NMR spectra support successful ligand exchange in all the cluster, but it does not provide us any information regarding the integrity of the cluster. UV-Vis measurements of the products were showing signs of decomposition as seen in Figure 22. We repeated the experiment while taking additional precautions such as purging the solution with N2, stirring in complete darkness and reduce reaction time. to prevent decomposition. Furthermore, we performed several controls such as attempting to react plain 18-crown-6 with Au25 and repeat prior ligand exchange reactions with alkanethiols in the conditions described before. To our demise, nothing we did appear to work. Not shown here, a crude ligand exchange reaction appeared to be successful for 2 nm silver nanoparticles without signs of decomposition or aggregation, adding to our frustrations. The UV-vis of all the products is almost identical between one another, where the Au25 features vanish beyond oblivion, resembling the spectrum of larger clusters like Au140. We can also make the assumption that the products formed cannot be larger than 3 nm in diameter since the 540 nm surface plasmon band is 30 30
Figure 22. UV-Vis of products from the ligand exchange reactions of Au25 and (A) 12-crown-4, (B) 15-crown-5, (C) 18-crown-6, and (D) aza-18-crown-6 ligands. Note: All absorption bands characteristic of Au25 has gone extinct. The spectra of the clusters were obtained in DCM and absorbance were normalized. not observed, typically exclusive for gold nanoparticles larger than ca. 3 nm (see Figure 2B in p. 6).3, 8 In all the products a new, unrecognized set of broad shoulder peaks can be seen from 500-600 nm, not to be mistaken for the 540 nm plasmonic band. This absorbance behavior is similar to the ligand exchange reactions involving the achiral ligands N-isobutyryl-L-cysteine and N-isobutyryl-D-cysteine reported by Si et al.25 These peaks may correspond to a single unidentified cluster, but this is speculation, and further evidence is a need (e.g., mass spectroscopy and crystallographic analysis). These findings alone were not enough to end further investigations, but the next set of experiments reported in the following section is the final nail in the coffin. 31 31 - Metal Reactions with Au25(SCH2CH2Ph)18
Initially, the goal was to develop Au25 chemical sensors by surface modifications with cation-selective ligands. During the control studies, we discovered an unexpected reaction between Au25 nanoclusters and the three target metals. When an excess amount of mercury(II), cadmium(II), or lead(II) was added to a solution of Au25 nanoclusters, key characteristic UV-vis bands were either completely or partially suppressed. Figures 23-27 illustrate a change in the absorption patterns after reacting Au25 with these three metals. Unlike in metallic nanoparticles, absorption in the UV-Vis region of AuNCs is due to discrete electronic transitions, therefore any changes to the core (e.g., oxidation state or elemental composition) affects these properties.9, 11, 19, 41-45 Therefore, the optical change seen after the reaction between Au25 and the active metals suggests there is chemical interaction occurring leading to either the formation of a bimetallic nanocluster or the oxidation of anionic Au25 to its neutral state. This led us to conclude that anionic Au25 nanoclusters are not suitable for the detection of dissolved cadmium, lead, and mercury. Recent findings investigated the reactions between mercury or cadmium salts with Au25 and concluded the formation of bimetallic clusters of Au24Hg and
11, 46-47 Au24Cd, respectively. The UV-Vis spectra of Au24Hg and Au24Cd were an identical match to our results, thus confirming the formation of bimetallic nanoclusters without the need for further evidence. The mechanisms and reasons why, in some cases, alloying occurs has not been thoroughly studied. Some speculate that metals with similar atomic radii are capable of exchanging with one or more gold atoms of the cluster.46 Earlier studies looked into the reaction between silver(I) and Au25, which also produced a
43, 48-50 bimetallic nanocluster of Au24Ag. Interestingly, these reports determined the 32 32 clusters to remain, in some cases, physically similar to Au25, but their optical and electrochemical properties were significantly different.
To further investigate whether Au25 is reactive towards other metals, we expanded our analyte list to include metals of the p-block, which would help us determine possible target analytes. The expanded list includes thallium(III) and bismuth(III) salts to investigate the reactivity of Au25 as discussed in the following.
Figure 23. UV-Vis spectrum of Au24Hg. The spectrum of the cluster was obtained in DCM, and the absorbance was normalized.
The first reaction studied was between Au25 and mercury(II). The optical absorption in Figure 23 revealed a familiar spectrum resembling Au25 in some ways where the 690 nm and 400 nm (see Figure 10) have remained almost unchanged. One major change is the loss of the 450 nm peak and the loss of a 33 33 minor broad peak around 790 nm, a behavior consistent with that of literature.47 The addition of a foreign metal to the cluster typically leads to the suppression of
48 the 450 nm as seen in Au24Ag clusters, while the oxidation of the cluster leads to
0 + the loss of the 790 nm broad shoulder as seen in Au25 and Au25 clusters. Without elemental analysis and mass spectroscopy, we cannot determine whether the foreign metal has integrated itself into the cluster by UV-Vis alone, but these subtle absorption patterns can be used to make early guesses of the outcome of
Au25.
The reaction between Au25 and cadmium(II) is significantly slower than mercury(II), and as a result, the procedures for this particular reaction was modified. The only adjustment was to allow the reaction to proceed at an elevated temperature to achieve better yields and shorten the reaction time. Using a temperature higher than 60 °C is strongly discouraged as the risk of aggregation is likely to occur; the UV-Vis of the unreacted clusters (not shown) was reminiscent of clusters larger than Au140. We speculate the Au24Cd cluster to be thermally stable at 60 °C. Surprising is the similarity between the absorption patterns between Au24Hg and Au24Cd as shown in Figures 23 and 24, respectively. It is not understood why the two are almost identical, but we can speculate that it is because elements within the same group exhibit similar reaction behavior onto
Au25.
When Au25 reacts with lead(II), the UV-Vis spectrum of the product
0 (Figure 25) appears similar to the neutral form of Au25 . We suspect that the
0 product is, in fact, Au25 due to matching absorption bands, where the peak and shoulder at 445 nm and 790 nm, respectively, are both suppressed, but the peak at 395 nm becomes significantly sharper as shown in Figure 5 (p. 6). Like the 34 34
Figure 24. UV-Vis spectrum of Au24Cd. The spectrum of the cluster was obtained in DCM, and the absorbance was normalized.
reaction with cadmium, the reaction between lead and Au25 is on the slower side, but a higher temperature is not required. At first glance, it may appear as all the lead has oxidized the cluster, but this is not the case as trace amounts of O2 dissolved in the solvent competes to oxidize the cluster. Thorough degassing helped remove most of the oxygen and significantly slow the oxidation by O2
(blanks not shown). Note that it is difficult to determine the amount of Au25 oxidized by lead oxidized how much of the oxidation was from oxygen through UV-Vis.
We determined the reaction between Au25 and thallium(III) lead to the oxidation of the nanocluster to its neutral state by UV-Vis. The absorbance of the
0 product as shown in Figure 26 is identical to the spectrum of Au25 . If both spectra
0 of this product and of Au25 were superimposed on top of each other, they would match and be almost indistinguishable. This observation is similar to the reaction 35 35 between lead(II) and Au25, but the difference is the time it takes for the reaction to complete. At room temperature, thallium(III) reacts completely with Au25 in about 15 minutes while with lead(II) requires several hours, typically overnight. One possible explanation is the potential reduction difference of the two, thallium(III) being a stronger oxidizer than lead(II).
0 Figure 25. UV-Vis spectrum of Au25 in DCM oxidized by lead(II). Note: The spectrum of the cluster was obtained in DCM, and the absorbance was normalized.
0 Figure 26. UV-Vis spectrum of Au25 DCM after reaction with thallium(III). Note: The spectrum of the cluster was obtained in DCM, and the absorbance was normalized. 36 36
The last reaction involving bismuth(III) and Au25 led the unexpected
+ formation of Au25 . This unstable cluster was identified by its unique line shape of the 400 nm and 445 nm peaks, both appearing to be suppressed, and the 790 nm shoulder is no longer present. This absorption pattern is indistinguishable from the
+ UV-Vis spectrum of Au25 (comparing Figures 5 and 27). Based on this
+ observation, we speculate the end product to be the cationic Au25 cluster.
+ Figure 27. UV-Vis spectrum of Au25 in DCM oxidized by bismuth. Note: The spectrum of the cluster was obtained in DCM, and the absorbance was normalized.
Differential pulse voltammetry (DPV) is a sensitive voltammetry technique used to calculate the halfway potentials of ionic species in solution and is sensitive
z enough for working in small quantities. The cluster Au25(SCH2CH2Ph)18 may exist in multiple oxidations states, where z= 1-, 0, and 1+. Whether Au25 is anionic, cationic, or neutral, three halfway potential peaks can be observed from -
51 0.4 V to 1.2 V potentials. Figure 28 is the voltammogram of Au25 at room 37 37 temperature in 0.1 M TBAPF6 (CH2Cl2, N2 purged) where first oxidation potential is observed at -0.03 V, the second oxidation potential at 0.26 V, and the third oxidation potential at 1.01 V. The first reduction peak is not shown in Figure 28 as it lies beyond the scan range in our measurements (ca. -1.1 V).
0.6
µA 0.1 I/ -0.4
-0.9 1 0.5 0 -0.5 + E (V vs. Ag/Ag )
Figure 28. DPV of Au25 in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode.
The reaction of Au25 and mercury(II) does not go to completion as trace amount of dissolved oxygen is enough to compete with mercury(II) which results
0 in the production of both Au24Hg and Au25 , the latter in small quantities not observed by UV-Vis, but enough to be visible in the DPV seen as minor peaks.
The two new peaks are observed at 0.51 V and 0.77 V in addition to the three Au25 peaks as seen in Figure 29. The Au24Hg cluster has been reported to be neutral
+ therefore its first oxidation leads to the formation of Au24Hg and the second
2+ oxidation state to Au24Hg . The third oxidation potential peak is not observed in our voltammograms but has been reported to appear at a more positive potential around 1.2 V.47 38 38
Figure 29. DPV of Au24Hg in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode. Note: *These current peaks correspond to the presence of Au25.
As stated previously, the reaction rate and yield between cadmium(II) and
Au25 are low. This is reflected in the DPV of the product where the major peaks three peaks correspond to the unreacted Au25 clusters and the minor peaks at 0.49 V and 0.69 V to correspond to the first and second oxidation potentials of the
Au24Cd cluster, respectively. Like Au24Hg, the halfway potential peaks shifted more positive, the third oxidation potential lying at a more positive potential reported to be around 1.2 V. A better voltammogram than the one shown in Figure 30 can be obtained from a pure sample, which has been a challenge to obtain. For the three other reactions between bismuth(III), lead(II), and thallium(III) no new halfway potential peaks were observed (Figures 31-33). The three peaks that are observed in these cases appear to be almost identical to the
Au25 indicating the preservation of the cluster. Note that the position or number of peaks observed does not change significantly between the different oxidation states of Au25. In the case for the reactions involving lead(II), thallium(III), and
0 + bismuth(III), the final oxidation state of the cluster is either Au25 or Au25 evident by UV-Vis but impossible to determine through DPV alone. 39 39
Figure 30. DPV of Au24Cd in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode. Note: *These current peaks correspond to the presence of Au25.
0.5
0 I/µA -0.5
-1 1 0.5 0 -0.5 + E (V vs. Ag/Ag )
Figure 31. DPV of Au25 reacted with lead(II) in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode.
40 40
0.7 0.5 0.3 0.1
I/µA -0.1 -0.3 -0.5 -0.7 1 0.5 0 -0.5 + E (V vs. Ag/Ag )
Figure 32. DPV of Au25 reacted with thallium(III) in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode.
0.7 0.5 0.3 0.1
I/µA -0.1 -0.3 -0.5 -0.7 1 0.5 0 -0.5 + E (V vs. Ag/Ag )
Figure 33. DPV of Au25 reacted with bismuth(III) in 0.1M TBAPF6 (DCM) using a Pt working electrode, Ag/AgCl reference electrode, and a Pt auxiliary electrode.
41 41
Table 1 compares the peak potentials between the bimetallic products and the oxidized Au25 products. The first, second, and third oxidation peaks for Au25 and the products of bismuth, lead, and thallium are labeled as 2+/1+, 1+/0, and
0/1-, respectively. Note that only the two halfway potentials of the Au24Hg and
Au24Cd clusters are listed, while the potentials belonging to Au25 within the same voltammograms were omitted. The 0/1- the potential for both Au24Hg and Au24Cd lie beyond the scan range used in our measurements.
Table 1. Halfway potentials (V) of the Au25 products obtained using DPV.
The likely reason why in some cases an introduced metal reacts with Au25 to form a bimetallic cluster compatibility such as similar size and electronic properties. Here that trend holds true as both mercury and cadmium are both similar in size and alloy with Au25, but for the smaller metals like bismuth, lead, and thallium, a bimetallic cluster is not formed. Both the UV-Vis and DPV results provide enough evidence for this speculation, but further evidence, e.g. mass 42 42 spectroscopy, elemental analysis, and crystallographic is required to solidify this conclusion.
Diffusion Constants and Size Estimations by Diffusion-Ordered NMR Spectroscopy The two peaks examined correspond to the phenyl moieties on the ligands of Au25 and the ferrocene. Ferrocene was used as the standard because it does not overlap with the PET nor the TOA peaks, does not react with Au25, and its size is known to be 0.3 nm.52 Figure 34 shows the two peaks decaying over increasing gradient strength. The cluster Au25 is significantly heavier and larger than ferrocene. Therefore its peaks decay much slower.
Figure 34. The DOSY-NMR spectrum of Au25 and ferrocene plotted as chemical shift (ppm) vs. gradient pulse amplitude (gz) from 1-24 or 8.72 to 410. mT/m. Note: In the NMR software, these parameters are set from 500 – 12,000 in increments of 500, the As the gradient strength is increased, both the Au25 phenyl peaks (7.19 ppm) and ferrocene peak (4.16 ppm) decay at different rates. The region between 3.5 ppm to 8.0 ppm is shown to highlight these two peaks. 43 43
The target peaks in each DOSY-NMR spectrum were individually integrated to obtain the area. The area under each separate peak is plotted versus the increasing gradient strength, which results in a Gaussian curve, but can be fitted into exponential form. Before plotting, the NMR gradients (in gz) must be converted to the area of the gradient pulse (q), where q is γgzδ, and the corrected diffusion delay (Δ’) must be calculated from Equation 4. The area, which can be normalized to the largest peak, is plotted versus q2Δ’ and fitted exponentially.
Using Equation 3, the diffusion coefficients were calculated for Au25 concentrations of 70 µM, 140 µM, and 270 µM each at 293.15 K, 298.15K, and 303.15 K. The diffusion coefficients calculated are under the assumption that the trace amounts of differently sized nanoclusters are present, but are not in sufficient amounts to skew the results. All measurements were made in replicates of three; nine samples in total. The diffusion coefficients were also measured simultaneously for ferrocene, but the concentration in each sample remained at 1.8 mM. From Figures 35-43, the normalized areas were averaged and plotted as described previously to give, as expected, a decay which could be fitted exponentially for a total of nine plots. Depending on the concentration or the temperature, the Δ parameter was adjusted to allow proper signal decay, resulting in different q2Δ’ ranges. The deviation between replicates appears to vary substantially between different temperatures and concentrations, especially at 293.15 K, but in all cases, the regression (r2) was no less than 0.98. All curves
(Figures 35-43) were fitted exponentially (Equation 3) using 24 points, where S0 is the normalized area of the peak signal and D is the diffusion coefficient. Alternatively, the curve can be transformed (−ln S vs. q2Δ′) and fitted linearly with less accuracy. 44 44
Figure 35. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars. Note: The measurements were made in acetone-d6 at 293.15 K and the concentration of Au25 and Fc at 70 µM and 1.8 mM, respectively. The normalized area of the signal (S) was plotted against the gradient pulse area with respect to the corrected diffusion delay (q2Δ’). The error bars were obtained from the standard deviation of three separate replicates.
Figure 36. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars. Note: The measurements were made in acetone-d6 at 298.15 K and the concentration of Au25 and Fc at 70 µM and 1.8 mM, respectively. The normalized area of the signal (S) was plotted against the gradient pulse area with respect to the corrected diffusion delay (q2Δ’). The error bars were obtained from the standard deviation of three separate replicates. 45 45
Figure 37. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars. Note: The measurements were made in acetone-d6 at 303.15 K and the concentration of Au25 and Fc at 70 µM and 1.8 mM, respectively. The normalized area of the signal (S) was plotted against the gradient pulse area with respect to the corrected diffusion delay (q2Δ’). The error bars were obtained from the standard deviation of three separate replicates.
Figure 38. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars. Note: The measurements were made in acetone-d6 at 293.15 K and the concentration of Au25 and Fc at 140 µM and 1.8 mM, respectively. The normalized area of the signal (S) was plotted against the gradient pulse area with respect to the corrected diffusion delay (q2Δ’). The error bars were obtained from the standard deviation of three separate replicates. 46 46
Figure 39. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars. Note: The measurements were made in acetone-d6 at 298.15 K and the concentration of Au25 and Fc at 140 µM and 1.8 mM, respectively. The normalized area of the signal (S) was plotted against the gradient pulse area with respect to the corrected diffusion delay (q2Δ’). The error bars were obtained from the standard deviation of three separate replicates.
Figure 40. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars. Note: The measurements were made in acetone-d6 at 303.15 K and the concentration of Au25 and Fc at 140 µM and 1.8 mM, respectively. The normalized area of the signal (S) was plotted against the gradient pulse area with respect to the corrected diffusion delay (q2Δ’). The error bars were obtained from the standard deviation of three separate replicates. 47 47
Figure 41. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars. Note: The measurements were made in acetone-d6 at 293.15 K and the concentration of Au25 and Fc at 270 µM and 1.8 mM, respectively. The normalized area of the signal (S) was plotted against the gradient pulse area with respect to the corrected diffusion delay (q2Δ’). The error bars were obtained from the standard deviation of three separate replicates.
Figure 42. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars. Note: The measurements were made in acetone-d6 at 298.15 K and the concentration of Au25 and Fc at 270 µM and 1.8 mM, respectively. The normalized area of the signal (S) was plotted against the gradient pulse area with respect to the corrected diffusion delay (q2Δ’). The error bars were obtained from the standard deviation of three separate replicates. 48 48
Figure 43. The average signal decay of Au25 (triangles) and Ferrocene (circles) with error bars. Note: The measurements were made in acetone-d6 at 303.15 K and the concentration of Au25 and Fc at 270 µM and 1.8 mM, respectively. The normalized area of the signal (S) was plotted against the gradient pulse area with respect to the corrected diffusion delay (q2Δ’). The error bars were obtained from the standard deviation of three separate replicates.
The average diffusion coefficients obtained are plotted in tables 1-5 below for both ferrocene and Au25. Each table represents a different concertation of Au25 at the three different temperatures. In comparison to other diffusion experiments conducted by Alawdi using cyclic voltammetry, the measured diffusion coefficients of Au25 by DOSY-NMR are greater by orders of magnitude. These results are high, but not unusual considering the mechanisms for measuring diffusion between the two techniques are different. Referencing the work of Canzi et al., the diffusion coefficient calculated for ferrocene in chloroform-d was
-9 2 1.62±0.02 ·10 m /s compared to our highest diffusion coefficient in acetone-d6 of 11.±6. ·10-9m2/s. Comparing the two results, they are not within the 95% confidence level, therefore, are different results.37 We suspect variables such as temperature, concentration, and solvent differences from our experiments may have contributed to this outcome. 49 49
The diffusion coefficients at different temperatures were expected to, more or less, remain the same. If it were the case in which the diffusion coefficients had changed, it would be observed at elevated temperatures. Surprisingly this was not the case as the diffusion coefficients for both Au25 and ferrocene drastically decreased from 293.15 to 298.15 K, but then rose back up at 303.15 K, with the exception of the sample containing 270 µM Au25. This is odd in many ways, and a definitive trend cannot be determined based on such a narrow temperature range with only three data point. Nevertheless, we proceeded to calculate the diameter of
Au25 from the diffusion coefficients from Tables 2-4.
Table 2. Average diffusion coefficients (10-9 m2/s) from DOSY-NMR for Ferrocene (1.8 mM) and Au25 (70 µM) at different temperatures. The diffusion coefficients from three trials were averaged and the standard deviation calculated for each temperature.
Table 3. Average diffusion coefficients (10-9 m2/s) from DOSY-NMR for Ferrocene (1.8 mM) and Au25 (140 µM) at different temperatures. The diffusion coefficients from three trials were averaged and the standard deviation calculated for each temperature.
293.15 K 298.15 K 303.15 K
Ferrocene 11. ±6. 6.9 ±1.0 11. ±2.
Au25 8. ±5. 4.3 ±0.7 8. ±2.
50 50 Table 4. Average diffusion coefficients (10-9 m2/s) from DOSY-NMR for Ferrocene (1.8 mM) and Au25 (270 µM) at different temperatures. The diffusion coefficients from three trials were averaged and the standard deviation calculated for each temperature.
- Through reported TEM studies, the size of Au25(SCH2CH2Ph)18 has been accepted to be 1.1±0.4 nm.5 Our results to range between 0.68 nm to 1.1 nm (Table 5) which is near the reported TEM value. Using Equation 2, we solve for rNC (hydrodynamic radius) of Au25 at different concentrations where rFc is the hydrodynamic radius of ferrocene (0.3 nm), DFc is the calculated diffusion coefficient of ferrocene, and DNC is calculated diffusion coefficient of Au25. Here, the reported diameters appear to increase with temperature, but it is difficult to make this correlation without further investigations. Further experiments at a wider temperature range may provide conclusive information regarding this correlation.
Table 5. The diameter of Au25 (nm) for each concentration and temperature calculated from DOSY-NMR using Equation 2. The propagation of error for each value was calculated from their respective diffusion coefficient averages.
51 51
Although the diffusion constants of both Au25 and ferrocene carry substantial error, the ratio, which is critical for the size estimations, between the two constants remain consistent throughout the experiments. This consistency yields a low error in the size estimations, meaning that environmental conditions
(e.g., temperature and concentration) affect both Au25 and ferrocene equally.
CONCLUSION
- In conclusion, the negatively charged Au25 cluster is not suitable for detecting bismuth(III), cadmium(II), lead(II), mercury(II), and thallium(III) ions by chelation through the use of crown ethers. We successfully synthesized crown ether ligands suitable for ligand exchange onto the surfaces of metal nanoclusters and nanoparticles, but for reasons not yet understood was shown to be incompatible with Au25. Repeated ligand exchange attempts under strict conditions have always led to, perhaps unidentified cluster or aggregation of Au25 as observed by UV-Vis. It was discovered that unmodified Au25 prior responded to the presence of all the target analytes, observed by UV-Vis and DPV data.
In the presence of mercury(II) or cadmium(II), Au25’s UV-Vis fingerprint regions had changed, but still reminiscent of the original absorption pattern. Their DPV data showed a positive shift of the first and second oxidation potentials, an indication of the presence of a new species. We later confirmed our results with findings reported by other groups who had characterized the products to be
Au24Hg and Au24Cd bimetallic nanoclusters. These unforeseen reactions were an inconvenience, but a fascinating as it presents a different method of tailoring the properties of Au25. In contrast, Au25 did not appear to alloy with bismuth(III), lead(II), and thallium(III) metal ions. However, oxidation of Au25 to its neutral and cationic states by these metals was observed by UV-Vis with supporting DPV analysis. The UV-Vis spectra of Au25 after reacting with lead(II) or thallium(III) is
0 indistinguishable from Au25 and reaction with bismuth(III) leads to a spectrum
+ almost identical to that of Au . The DPV of the oxidized Au25 clusters had not changed significantly, a well-documented observation. This further supports the 53 53 notion that bimetallic clusters had not been formed in the presence of bismuth(III), lead(II), or thallium(III) ions. It has been speculated that for alloying to occur, the incoming metal must be similar in size and electronic character, which was the case for the p-block metals in this study. Others have reported similar observations with cobalt, zinc, and nickel. Here, we come to two conclusions: one, Au25 incapable of undergoing ligand exchange with crown ether ligands without influencing the integrity of the whole cluster and two, Au25 irrelevant of oxidation state is not suitable for the detection of mercury(II) nor cadmium(II). However, we speculate that the neutral form of Au25 may be better suitable for the detection of some metals because it is less redox active as compared to its anionic form. In contrast, we had a greater degree of success determining the hydrodynamic radius of Au25 clusters by DOSY-NMR. The common method of determining the size of metallic nanoclusters and nanoparticles has been by TEM and other electron microscopy methods as it provides accurate information such as size, shape, and crystal structure. To characterize Au25, a high-resolution TEM is required as its small size challenges the limits of most TEM instruments. To date, this institution does not possess such instrument. For our research goals, determining the size of our gold nanoclusters is most important. Here, we employed DOSY-NMR as an indirect method of calculating the size of the Au25 nanoclusters in solution at various temperatures and concentrations. Our calculations determined the size range of Au25 from 0.68±0.17 nm to 1.1±0.4, in agreement with the current literature (1.1±0.4 nm). The size estimates were all slightly below than the reported TEM measurements and appear to fluctuate with temperature, which could be a result of the ligands. Separate studies comparing the 54 54 free ligands to a mixture of Au25 and free ligands could reveal how ligand length affects the calculated diffusion coefficients. We do not suspect other clusters are the source of error because the two other major clusters produced by our synthesis are Au140 and Au309, which would skew our values to be greater. Our group is currently studying Au140 using DOSY-
NMR which would allow us to compare the results to Au25. A separate study of a mixture of both Au25 and Au140 nanoclusters would help determine how the presence of other clusters affect the calculated diffusion coefficients and diameters. No direct trend was observed regarding temperature nor concentration, but we plan to further pursue these investigations and determine if there is a relationship. Here we show an alternative, reliable method of indirectly determining the size of Au25 by DOSY-NMR. This method can be used for other nanoclusters e.g. identify the products of the crown ether ligand exchange reactions with Au25.
REFERENCES
1. Faraday, M., The Bakerian Lecture: Experimental Relations of Gold (and Other Metals) to Light. Philosophical Transactions of the Royal Society of London 1857, 147, 145-181.
2. Turkevich, J.; Stevenson, P. C.; Hillier, J., A study of the nucleation and growth processes in the synthesis of colloidal gold. Discussions of the Faraday Society 1951, 11 (0), 55-75.
3. Duff, D. G.; Baiker, A.; Edwards, P. P., A new hydrosol of gold clusters. Journal of the Chemical Society, Chemical Communications 1993, (1), 96- 98.
4. Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R., Synthesis of thiol-derivatised gold nanoparticles in a two-phase Liquid-Liquid system. Journal of the Chemical Society, Chemical Communications 1994, (7), 801- 802.
5. Donkers, R. L.; Lee, D.; Murray, R. W., Synthesis and Isolation of the Molecule-like Cluster Au38(PhCH2CH2S)24. Langmuir 2004, 20 (5), 1945- 1952.
6. Negishi, Y.; Nobusada, K.; Tsukuda, T., Glutathione-protected gold clusters revisited: Bridging the gap between gold(I)-thiolate complexes and thiolate- protected gold nanocrystals. Journal of the American Chemical Society 2005, 127 (14), 5261-5270.
7. Zhu, M.; Aikens, C. M.; Hollander, F. J.; Schatz, G. C.; Jin, R., Correlating the Crystal Structure of A Thiol-Protected Au25 Cluster and Optical Properties. Journal of the American Chemical Society 2008, 130 (18), 5883- 5885.
8. Kwak, K.; Thanthirige, V. D.; Pyo, K.; Lee, D.; Ramakrishna, G., Energy Gap Law for Exciton Dynamics in Gold Cluster Molecules. The Journal of Physical Chemistry Letters 2017, 8 (19), 4898-4905.
9. Xie, S. H.; Tsunoyama, H.; Kurashige, W.; Negishi, Y.; Tsukuda, T., Enhancement in Aerobic Alcohol Oxidation Catalysis of Au-25 Clusters by Single Pd Atom Doping. ACS Catal. 2012, 2 (7), 1519-1523. 56 56 10. Wu, Z. L.; Jiang, D. E.; Mann, A. K. P.; Mullins, D. R.; Qiao, Z. A.; Allard, L. F.; Zeng, C. J.; Jin, R. C.; Overbury, S. H., Thiolate Ligands as a Double- Edged Sword for CO Oxidation on CeO2 Supported Au- 25(SCH2CH2Ph)(18) Nanoclusters. Journal of the American Chemical Society 2014, 136 (16), 6111-6122.
11. Deng, H.; Wang, S.; Jin, S.; Yang, S.; Xu, Y.; Liu, L.; Xiang, J.; Hu, D.; Zhu, M., Active metal (cadmium) doping enhanced the stability of inert metal (gold) nanocluster under O2 atmosphere and the catalysis activity of benzyl alcohol oxidation. Gold Bulletin 2015, 48 (3), 161-167.
12. Prakash, P.; Gravel, E.; Li, H.; Miserque, F.; Habert, A.; den Hertog, M.; Ling, W. L.; Namboothiri, I. N. N.; Doris, E., Direct and co-catalytic oxidative aromatization of 1,4-dihydropyridines and related substrates using gold nanoparticles supported on carbon nanotubes. Catalysis Science & Technology 2016, 6 (17), 6476-6479.
13. Yan, N.; Liao, L. W.; Yuan, J. Y.; Lin, Y. J.; Weng, L. H.; Yang, J. L.; Wu, Z. K., Bimetal Doping in Nanoclusters: Synergistic or Counteractive? Chemistry of Materials 2016, 28 (22), 8240-8247.
14. Heinecke, C. L.; Ni, T. W.; Malola, S.; Mäkinen, V.; Wong, O. A.; Häkkinen, H.; Ackerson, C. J., Structural and Theoretical Basis for Ligand Exchange on Thiolate Monolayer Protected Gold Nanoclusters. Journal of the American Chemical Society 2012, 134 (32), 13316-13322.
15. Fernando, A.; Aikens, C. M., Ligand Exchange Mechanism on Thiolate Monolayer Protected Au25(SR)18 Nanoclusters. The Journal of Physical Chemistry C 2015, 119 (34), 20179-20187.
16. Zhu, M.; Lanni, E.; Garg, N.; Bier, M. E.; Jin, R., Kinetically Controlled, High-Yield Synthesis of Au25 Clusters. J. Am. Chem. Soc. 2008, 130 (4), 1138-1139.
17. Wu, Z.; MacDonald, M. A.; Chen, J.; Zhang, P.; Jin, R., Kinetic Control and Thermodynamic Selection in the Synthesis of Atomically Precise Gold Nanoclusters. J. Am. Chem. Soc. 2011, 133 (25), 9670-9673.
18. Chen, T.; Yao, Q.; Yuan, X.; Nasaruddin, R. R.; Xie, J., Heating or Cooling: Temperature Effects on the Synthesis of Atomically Precise Gold Nanoclusters. J. Phys. Chem. C 2016, Ahead of Print. 57 57 19. Kwak, K.; Tang, Q.; Kim, M.; Jiang, D. E.; Lee, D., Interconversion between Superatomic 6-Electron and 8-Electron Configurations of M@Au- 24(SR)(18) Clusters (M = Pd, Pt). Journal of the American Chemical Society 2015, 137 (33), 10833-10840.
20. Hesari, M.; Workentin, M. S.; Ding, Z. F., Near-infrared electrochemiluminescence from Au-25(SC2H4Ph)(18)(+) clusters co-reacted with tri-n-propylamine. RSC Adv. 2014, 4 (56), 29559-29562.
21. Shichibu, Y.; Negishi, Y.; Tsunoyama, H.; Kanehara, M.; Teranishi, T.; Tsukuda, T., Extremely high stability of glutathionate-protected Au-25 clusters against core etching. Small 2007, 3 (5), 835-839.
22. Zhu, M.; Eckenhoff, W. T.; Pintauer, T.; Jin, R., Conversion of Anionic [Au25(SCH2CH2Ph)18]− Cluster to Charge Neutral Cluster via Air Oxidation. The Journal of Physical Chemistry C 2008, 112 (37), 14221- 14224.
23. Heaven, M. W.; Dass, A.; White, P. S.; Holt, K. M.; Murray, R. W., Crystal Structure of the Gold Nanoparticle [N(C8H17)4][Au25(SCH2CH2Ph)18]. Journal of the American Chemical Society 2008, 130 (12), 3754-3755.
24. Guo, R.; Song, Y.; Wang, G.; Murray, R. W., Does Core Size Matter in the Kinetics of Ligand Exchanges of Monolayer-Protected Au Clusters? J. Am. Chem. Soc. 2005, 127 (8), 2752-2757.
25. Si, S.; Gautier, C.; Boudon, J.; Taras, R.; Gladiali, S.; Bürgi, T., Ligand Exchange on Au25 Cluster with Chiral Thiols. The Journal of Physical Chemistry C 2009, 113 (30), 12966-12969.
26. Knoppe, S.; Burgi, T., The fate of Au-25(SR)(18) clusters upon ligand exchange with binaphthyl-dithiol: interstaple binding vs. decomposition. Phys. Chem. Chem. Phys. 2013, 15 (38), 15816-15820.
27. Li, G.; Abroshan, H.; Liu, C.; Zhuo, S.; Li, Z. M.; Xie, Y.; Kim, H. J.; Ros, N. L.; Jin, R. C., Tailoring the Electronic and Catalytic Properties of Au-25 Nanoclusters via Ligand Engineering. Acs Nano 2016, 10 (8), 7998-8005.
28. MacDonald, M. A.; Chevrier, D. M.; Zhang, P.; Qian, H.; Jin, R., The Structure and Bonding of Au25(SR)18 Nanoclusters from EXAFS: The Interplay of Metallic and Molecular Behavior. J. Phys. Chem. C 2011, 115 (31), 15282-15287. 58 58 29. Ho, M.-L.; Hsieh, J.-M.; Lai, C.-W.; Peng, H.-C.; Kang, C.-C.; Wu, I. C.; Lai, C.-H.; Chen, Y.-C.; Chou, P.-T., 15-Crown-5 functionalized Au nanoparticles synthesized via single molecule exchange on silica nanoparticles: Its application to probe 15-Crown-5/K+/15-Crown-5 "sandwiches" as linking mechanisms. J. Phys. Chem. C 2009, 113 (5), 1686- 1693.
30. Kuang, H.; Chen, W.; Yan, W.; Xu, L.; Zhu, Y.; Liu, L.; Chu, H.; Peng, C.; Wang, L.; Kotov, N. A.; Xu, C., Crown ether assembly of gold nanoparticles: melamine sensor. Biosensors & bioelectronics 2011, 26 (5), 2032-7.
31. Nakashima, H.; Furukawa, K.; Kashimura, Y.; Torimitsu, K., Anisotropic assembly of gold nanorods assisted by selective ion recognition of surface- anchored crown ether derivatives. Chemical communications (Cambridge, England) 2007, (10), 1080-2.
32. Luo, H.; Dai, S.; Bonnesen, P. V., Solvent Extraction of Sr2+ and Cs+ Based on Room-Temperature Ionic Liquids Containing Monoaza-Substituted Crown Ethers. Analytical Chemistry 2004, 76 (10), 2773-2779.
33. Alizadeh, A.; Khodaei, M. M.; Karami, C.; Workentin, M. S.; Shamsipur, M.; Sadeghi, M., Rapid and selective lead (II) colorimetric sensor based on azacrown ether-functionalized gold nanoparticles. Nanotechnology 2010, 21 (31), 315503/1-315503/8.
34. Izatt, R. M.; Pawlak, K.; Bradshaw, J. S.; Bruening, R. L., Thermodynamic and Kinetic Data for Macrocycle Interaction with Cations and Anions. Chemical Reviews 1991, 91 (8), 1721-2085.
35. Rodriguez-Vazquez, M. J.; Blanco, M. C.; Lourido, R.; Vazquez-Vazquez, C.; Pastor, E.; Planes, G. A.; Rivas, J.; Lopez-Quintela, M. A., Synthesis of Atomic Gold Clusters with Strong Electrocatalytic Activities. Langmuir 2008, 24 (21), 12690-12694.
36. Harmon, J.; Coffman, C.; Villarrial, S.; Chabolla, S.; Heisel, K. A.; Krishnan, V. V., Determination of Molecular Self-Diffusion Coefficients Using Pulsed- Field-Gradient NMR: An Experiment for Undergraduate Physical Chemistry Laboratory. J. Chem. Educ. 2012, 89 (6), 780-783.
37. Canzi, G.; Mrse, A. A.; Kubiak, C. P., Diffusion-Ordered NMR Spectroscopy as a Reliable Alternative to TEM for Determining the Size of Gold Nanoparticles in Organic Solutions. The Journal of Physical Chemistry C 2011, 115 (16), 7972-7978. 59 59 38. Flink, S.; Boukamp, B. A.; van den Berg, A.; van Veggel, F. C. J. M.; Reinhoudt, D. N., Electrochemical Detection of Electrochemically Inactive Cations by Self-Assembled Monolayers of Crown Ethers. Journal of the American Chemical Society 1998, 120 (19), 4652-4657.
39. Santiago, N. Fluorescence Energy Transfer from Laser Dye to Gold Nanoparticles. California State University, Fresno, 2014.
40. Alawdi, H. Electrochemical Heterogeneous Electron Transfer Rate Constants and Thermodynamic Parameters of Gold Nanoparticle Assemblies with Different Alkanethiolate Ligands. Fresno State, 2015.
41. Niihori, Y.; Matsuzaki, M.; Pradeep, T.; Negishi, Y., Separation of Precise Compositions of Noble Metal Clusters Protected with Mixed Ligands. J. Am. Chem. Soc. 2013, 135 (13), 4946-4949.
42. Yao, C.; Lin, Y.-j.; Yuan, J.; Liao, L.; Zhu, M.; Weng, L.-h.; Yang, J.; Wu, Z., Mono-cadmium vs Mono-mercury Doping of Au25 Nanoclusters. J. Am. Chem. Soc. 2015, 137 (49), 15350-15353.
43. Wang, S. X.; Song, Y. B.; Jin, S.; Liu, X.; Zhang, J.; Pei, Y.; Meng, X. M.; Chen, M.; Li, P.; Zhu, M. Z., Metal Exchange Method Using Au-25 Nanoclusters as Templates for Alloy Nanoclusters with Atomic Precision. Journal of the American Chemical Society 2015, 137 (12), 4018-4021.
44. Thanthirige, V. D.; Kim, M.; Choi, W.; Kwak, K.; Lee, D.; Ramakrishna, G., Temperature-Dependent Absorption and Ultrafast Exciton Relaxation Dynamics in MAu24(SR)18 Clusters (M = Pt, Hg): Role of the Central Metal Atom. The Journal of Physical Chemistry C 2016, 120 (40), 23180-23188.
45. Alkan, F.; Munoz-Castro, A.; Aikens, C. M., Relativistic DFT investigation of electronic structure effects arising from doping the Au-25 nanocluster with transition metals. Nanoscale 2017, 9 (41), 15825-15834.
46. Yao, C.; Lin, Y.-j.; Yuan, J.; Liao, L.; Zhu, M.; Weng, L.-h.; Yang, J.; Wu, Z., Mono-cadmium vs Mono-mercury Doping of Au25 Nanoclusters. Journal of the American Chemical Society 2015, 137 (49), 15350-15353.
47. Liao, L.; Zhou, S.; Dai, Y.; Liu, L.; Yao, C.; Fu, C.; Yang, J.; Wu, Z., Mono- Mercury Doping of Au25 and the HOMO/LUMO Energies Evaluation Employing Differential Pulse Voltammetry. Journal of the American Chemical Society 2015, 137 (30), 9511-9514. 60 60 48. Choi, J.-P.; Fields-Zinna, C. A.; Stiles, R. L.; Balasubramanian, R.; Douglas, A. D.; Crowe, M. C.; Murray, R. W., Reactivity of [Au25(SCH2CH2Ph)18]1− Nanoparticles with Metal Ions. The Journal of Physical Chemistry C 2010, 114 (38), 15890-15896.
49. Guidez, E. B.; Mäkinen, V.; Häkkinen, H.; Aikens, C. M., Effects of Silver Doping on the Geometric and Electronic Structure and Optical Absorption Spectra of the Au25–nAgn(SH)18– (n = 1, 2, 4, 6, 8, 10, 12) Bimetallic Nanoclusters. The Journal of Physical Chemistry C 2012, 116 (38), 20617- 20624.
50. Li, Q.; Wang, S.; Kirschbaum, K.; Lambright, K. J.; Das, A.; Jin, R., Heavily doped Au25-xAgx(SC6H11)18- nanoclusters: silver goes from the core to the surface. Chemical Communications 2016, 52 (29), 5194-5197.
51. Lee, D.; Donkers, R. L.; Wang, G.; Harper, A. S.; Murray, R. W., Electrochemistry and Optical Absorbance and Luminescence of Molecule- like Au38 Nanoparticles. Journal of the American Chemical Society 2004, 126 (19), 6193-6199.
52. Bond, A. M.; Colton, R.; Harvey, J.; Hutton, R. S., Voltammetric studies of ferrocene and the mercury dithiophosphate system at mercury electrodes over a temperature range encompassing the mercury liquid-solid state transition. Journal of Electroanalytical Chemistry 1997, 426 (1), 145-155.