«Kinetic Investigations of Thiolate Protected Gold Nanoparticles: Protein Interactions, Electron Transfer, and Precursor Formation By Brian N. Turner ...»
Using this approach, it was hypothesized that integrating a “molecular wire” into the protecting monolayer of alkanethiolate protected gold nanoparticles should increase the electron transfer rate at a biased platinum electrode, as conductance happens faster, or more often, than tunneling, as illustrated in Figure 37.
Figure 37: Proof of concept experiments to confirm whether “wired” AuDodecanethiol nanoparticles (left) exhibit faster electron transfer than their “unwired” precursors (right).
Materials HAuCl4- •3H2O was prepared as described previously.121 HAuCl4- •3H2O was synthesized according to standard methods from electrochemically purified Canadian gold maple leaf coins (99.99%).121 Reagent and optima grade solvents, alkanethiols (90%), sodium borohydride, lithium borohydride, tetrabutylammonium hexafluorphospate (TBAPF6) and S-[4-(2phenylethynyl)phenyl]ethynylphenyl]thioacetate (PEPEPSAc) were purchased from Sigma-Aldrich. Anhydrous dichloromethane (99%) from Sigma Aldrich was used from a sure-sealed bottle. Common laboratory salts were reagent grade and purchased from Fisher scientific. Concentrated sulfuric acid was purchased from EMD. Absolute ethanol was purchased from Pharmco-AAPER. Hydrogen peroxide (30% v/v) was purchased from Acros. Chemicals for peptide synthesis (f-moc protected and side-chain protected amino acids, coupling reagents, and resins) were provided generously by the David Wright research group. 18 MΩ Water was obtained from a U.S. Filter Modulab water system with a 0.2 µm external filter, or from a Barnstead NANOpure Diamond water purification system. Electrodes were purchased from CH instruments, or were previously fabricated in this research group and handed down. Diamond paste was purchased from. Alumina was purchased from. Deuterated solvents (99.9% D) were purchased from Cambridge Isotope Laboratories.
Synthesis of Au1120 Dodecanthiolate690
1.5 g of tetraoctylammonium bromide (a phase transfer agent) was dissolved in 80 mL of toluene. 0.31 g of HAuCl4 was dissolved in 25 mL of water. To the vigorously stirring tetraoctylammonium bromide solution, the gold solution was added slowly. The resulting orange gold-containing organic phase was isolated and chilled in an ice bath for 30 minutes and stirred vigorously. 0.38 g of NaBH4 was dissolved in 25 mL of water.
To the stirring organic solution, 260 L of 1-hexanethiol was added. Immediately, the NaBH4 solution was added over a period of 10 seconds. The resulting black solution was stirred for 30 minutes on ice, and then additionally overnight at room temperature. The organic phase was isolated and rotary evaporated to yield a black solid. The solid was dissolved in a minimum of toluene and precipitated in 400 mL of ethanol at -20 C overnight, which was then decanted and centrifuged. The precipitate was collected on the walls of a round bottom flask by dissolution and rotary evaporation and sonicated in 400 mL of acetonitrile and then ethanol. 1H NMR: 1.45 ppm, 0.98 ppm broad peaks.
TEM: 3.4 ± 1.1 nm. UV/Vis: slight surface plasmon band observed. TGA: 34.5 % organic material.
Synthesis of Au228Octanethiol92 HAuCl4 (300 mg) and 160 mg LiBH4 were dissolved separately in freshly distilled THF (30 mL total), and then purged with argon. The gold solution was chilled to near 0°C in an ice bath. Then, 1-octanethiol (280 µL) was added via syringe to the gold solution.
After a few seconds, the LiBH4 solution was added dropwise to the mixture. It is important to add the LiBH4 shortly after addition of the 1-octanethiol, or precipitation of oligomeric precursor material will decrease the yield and increase the polydispersity and size of the nanoparticles. CAUTION: LiBH4 reacts with gold very exothermically;
therefore caution should be observed during addition, and care should be taken to quench and clean the area of any excess LiBH4 as it tends to float during weighing. The resulting black to brown solution was stirred overnight. To this, 30 mL of water were added, and this was extracted into 60 mL diethyl ether and washed with brine 3 times. After rotoevaporation of the solvent, the nanoparticles were suspended in acetonitrile, centrifuged, suspended in ethanol, centrifuged, and washed by drying on a round bottom flask and sonicating in a large excess of acetonitrile and/or ethanol until spectroscopically clean. Characterization data: TEM: 2.0 ± 0.8 nm. TGA: 25.0% organic material. UVVis: slight SPR band observed at 520 nm. 1H NMR: 0.98 and 1.50 ppm broad resonances.
Deprotection of S-[4-(2-Phenylethynyl)phenyl]ethynylphenyl]thioacetate (PEPEPSAc) 100 mg of PEPEPSAc was dissolved in 8 mL of dichloromethane and 2 mL of methanol.
4 drops of concentrated sulfuric acid was added to this. The reaction was stirred overnight under a reflux condenser to slow solvent evaporation. The volume was reduced, if necessary, and poured onto a silica gel plug topped with a filter paper.
Fractions were collected in increasing gradients of dichloromethane in hexanes.
Fractions that yielded a white solid upon evaporation were collected first (in low dichloromethane). Yellow solids were collected in later fractions (increasingly higher dichloromethane). White fraction (S-[4-(2-phenylethynyl)phenyl]ethynylphenyl]thiol, PEPEPSH): 1H NMR 7.51 ppm, m, 7H (aromatic); 7.38 ppm, m, 6H (aromatic); 3.54 ppm, s, 1H (thiol). First yellow fraction (di-S-[4-(2Phenylethynyl)phenyl]ethynylphenyl]disulfide, PEPEPS 2): 7.52 ppm, m, 14H (aromatic);
7.36 ppm, m, 12H (aromatic). 1H NMR were taken on a Bruker AV 400 MHz instrument.
Place exchange reactions with PEPEPSAc and PEPEPS2 20 mg of Au1120 Dodecanthiolate690 was dissolved in 4 mL toluene with 8 mg PEPEPSH and stirred for 7 days. The product was dried onto a round bottom flask and sonicated in 100 mL acetonitrile for 20 minutes, 3 times, and with ethanol for 20 minutes, 3 times.
20 mg of Au228Octanethiol92 was dissolved in 4 mL toluene with and stirred for 72 hours.
These particles were suspended and then centrifuged from acetonitrile and then from ethanol. Further purification was accomplished by drying them in a round bottom flask, and then sonicating in an excess of acetonitrile and then ethanol, once each.
Scanning Electrochemical Microscopy Experiments Nanoparticle samples were dissolved in anhydrous dichloromethane (20 mg in 5 mL) with 200 mg TBAPF6. These solutions were pushed through nylon syring filters (0.45 µm). 2 mm Pt substrate electrode was treated by polishing with successively smaller grit diamond paste, and then finally polished on a polishing wheel (0.1 and 0.05 µm Micropolish II with 8” Microcloth on a Metaserv 2000 Grinder/Polisher (Buehler)). The polished tip was then sonicated in deionized water and then in ethanol. A smooth surface was observed using an Olympus BX4 optical microscope ( 1000x magnification).
Occasionally, it was cleaned by holding in piranha solution for 1 minute. Finally, it was cleaned by constant scanning from -0.4 to 1.4 V (vs. silver wire quasi-reference electrode, with a Pt counter wire) in 0.1 M H2SO4 with 0.1 M KCl until a lack of hysteresis in the cyclic voltammogram was observed. The tip was rinsed with water and ethanol. The 25 µm Pt UME was treated similarly. During polishing of the UME, care was taken not to create a tip surface with an RG 5 (usually around 2-3 in these studies).
Ag/Ag+ quasi-reference wire electrode was prepared by applying a potential of 1V across the wire for about 1 minute, resulting in a blackened layer of AgCl. The AgCl layer was lightly scraped with fine grit sand paper to reveal some uncoated surface.
The UME was fitted into the plastic holder of a piezoelectric motor and visually positioned over the center of the Pt substrate electrode, which is fitted into a plastic well with the reference electrode. The well was filled with the nanoparticle solution and approached at a faster rate until an increase in current was observed (UME @ 600 mV, substrate at 0 mV), at which point it was stopped and backed up a minimum 80 µm.
Approach curves were collected with the UME at 200-600 mV and the substrate always at 0 mV (vs. Ag/Ag+). Dichloromethane was replenished when necessary to a visually identical volume (close to the brim). In between sample changes, the chamber and electrodes were rinsed with anhydrous dichloromethane until the rinsate was not colored.
Nanoparticle Surface Quantitation For dodecanethiolate nanoparticles, samples were prepared in DCTB matrix in toluene at approximately 400:1 matrix/analyte ratio. All MALDI-IM-MS analyses were performed using a Synapt HDMS (Waters Corp., Manchester, UK), equipped with a frequencytripled Nd:YAG (355 nm) laser operated at a pulse repetition frequency of 200 Hz. All spectra were acquired in the positive ion mode at laser energy settings approximately 10% above threshold values.
Gold-containing ion signals were extracted and identified using the MassLynx 4.1 (Waters Corp.) software package. Processed spectra were exported to Microsoft Excel, in which a custom spreadsheet was constructed for the remainder of the data processing.
three stoichiometries identified previously: Aux+1Lx, AuxLx, and AuxLx+1.32 Within these stoichiometries, an appropriate list of permutations was constructed with every possible combination of ligands and modifications such as sodium coordination and methyl esterification. Protons were added or subtracted as necessary to achieve a +1 charge state. Each processed spectrum was filtered by abundance, with any signal below 1% relative abundance with respect to the base peak discarded to reduce false positive identification. The remaining peaks were compared to the expected gold-thiolate ion list.
Any observed peak within 30 ppm mass accuracy of an expected peak was matched and used for quantitation.
For octanethiolate nanoparticle samples, a concentrated solution in CDCl3 was prepared, and standard 1H experiments run on a 500 MHz Bruker NMR, 64+ scans. Non-solvent residual peaks in the aromatic region were calibrated to 13 for the 13 aromatic protons in PEPEPS, and compared to the combined peaks for the alkane chain, representing 17 protons for octanethiol. The stoichiometric ratio was matched to the original nanoparticle formula to yield an estimation of average nanoparticle composition (semi-quantitative).
A Note About Surface Ligand Quantitation In the research group of David Cliffel, a methodology for ion mobility-MALDI mass spectrometry (IM-MALDI-MS) to quantify the relative composition of the protecting monolayer of gold nanoparticles was developed.65 As the technique was new, and the instrument was often used for other applications, it was only applied to a couple of the samples referred to in this chapter, but was proven effective in the analysis of similar samples not discussed here. Example data is shown below in Figure 38 for Au1120 Dodecanthiolate690.
Figure 38: MALDI-IM mass spectrum of AuDodecanethiolate nanoparticles functionalized with the molecular wire. The intensity ratio of the tetrameric species with no wire and the tetrameric species with one wire molecule yields the ratio of dodecanethiolate to molecular wire in the protecting monolayer.
IM-MALDI-MS allows for the separation of organic impurities from gold complexes, improving the signal to noise ratio for species important to quantify the surface composition. It was shown that the relative ratios of tetrameric Au4 Ligand4 ratios yields accurate quantitation when compared to the standard 1H NMR analysis that has been used to accomplish this task in the past. In the particular sample in Figure 37, two ions were identified as tetramers: Au4Dodecanethiol4 and Au4Dodecanethiol3PEPEPS1. The ratio of the two peaks yields a result of 7 % PEPEPS was present in the nanoparticle monolayer. A second sample with a greater feed ratio of PEPESH (during place exchange) yielded a result of 20% PEPEPS. 1H NMR quantitation was obscured with this sample, as the solvent residual peak in the spectrum (CDCl3) interfere with accurate quantitation, so the values for the average molecular formula of place exchanged AuOctanethiol nanoparticles, discussed later, are treated as only good estimates when compared to the quantitation in the AuDodecanethiol samples that will be discussed.
SECM Qualitative Observation of AuDodecanethiol Electron Transfer Rates SECM approach curves for Au1120 Dodecanthiolate690 exhibited a constant negative current drift through most of the approach, and terminated in a negative feedback, similar to that of the approach to a non-conductive surface. The approach is displayed in Figure 39, below.
Figure 39: SECM approach curve with 1V Pt UME through larger AuDodecanethiolate nanoparticles, 0V Pt substrate electrode (vs. Ag/Ag+), in dichloromethane with 0.1M TBAPF6. Total tip current instead of normalized tip current is displayed as an accurate, stable value for it,∞ could not be determined due to the constant current drift.
The negative feedback was, in this case, attributed to a combination of slow electron transfer and slow diffusion. Essentially, when the tip was near the substrate electrode, there was not a higher local concentration of reduced species. It was thought that a layer of oxidized nanoparticle, which has continually diffused from the tip electrode over the course of the entire approach, has created an insulating thin layer cell over the substrate electrode.
When the solution is removed from the cell and replaced with a solution of Au1120Dodecanethiolate660Wire30 an entirely different result is obtained, as demonstrated in Figure 40.
Figure 40: SECM approach curve with 1V Pt UME through larger “wired” AuDodecanethiolate nanoparticles, 0V Pt substrate electrode (vs. Ag/Ag+), in dichloromethane with 0.1M TBAPF6.
While the constant drift was still observed in this experiment, a positive feedback finally arose under a half tip diameter away from the substrate electrode. In this instance, it was proposed that the wire imparted fast electron transfer kinetics upon the nanoparticles, and a feedback loop was established close to the surface. The positive feedback result was repeated three times in the same cell with the same electrode. Eventually, the results became irreproducible as absolute tip current was continually lower with each experiment. Constantly drifting current over the course of multiple experiments suggested that the electrodes were increasingly fouled by excess thiol impurities.