Published on Web 06/22/2010

New Insights into Brust-Schiffrin Metal Nanoparticle Synthesis

Paul J. G. Goulet and R. Bruce Lennox*
Department of Chemistry and Centre for Self-Assembled Chemical Structures (CSACS), McGill UniVersity,
801 Sherbrooke Street West, Montreal, Quebec, Canada, H3A 2K6
Received May 10, 2010; E-mail: bruce.lennox@mcgill.ca

Scheme 1. Revised View of the Two-Phase Brust-Schiffrin Au

Abstract: A revised view of Brust-Schiffrin metal nanoparticle Nanoparticle Synthesis
syntheses is presented here. Precursor species of these reactions
are identified and quantified for Au, Ag, and Cu systems. Contrary
Downloaded via COMSATS INST OF INFORMATION TECHLGY on September 27, 2018 at 09:12:31 (UTC).

to the assumptions of previous reports, tetraalkylammonium metal

complexes are shown to be precursors of the two-phase reac-
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

tions, whereas M(I) thiolates are shown to be precursors of the

one-phase reactions. A new scheme is outlined for the two-phase
synthesis, and the implications of this scheme are discussed. A
new synthetic strategy employing well-defined precursors is also
introduced. Finally, M(I) thiolate formation, and its impact on
nanoparticle synthesis, is discussed. It is expected that the results
presented here will lead to modifications in the manner in which
these important syntheses are conducted.
the two-phase Brust-Schiffrin reaction is outlined (Scheme 1), and
its implications are discussed. This general reaction motif is shown
The two-phase Brust-Schiffrin method1 for the synthesis of to also apply to two-phase syntheses employing Ag and Cu, as
thiolate-protected gold nanoparticles has been tremendously influ- well as to those where the phase-transfer agent or thiol have been
ential in nanomaterials chemistry. It has inspired a number of related substituted. This information is utilized toward the development
approaches and remains widely employed more than 15 years after of a new synthetic strategy employing well-defined precursor
its initial publication.2,3 It reliably produces large quantities of species. A discussion is presented regarding the formation of M(I)
relatively monodisperse samples of small nanocrystals (1-6 nm). thiolate species and the potential impact of this on nanoparticle
These readily undergo ligand exchange and can be conveniently syntheses.
dried, purified, and redissolved. Brust et al.4 also developed a related
[NR4][AuX4] + 2 R′SH f [NR4][AuX2] + R′SSR′ + 2 HX
one-phase method that is carried out in polar solvent and shares
the positive attributes of the two-phase approach. Both of these
(1)
approaches have since been successfully applied to other metals, [NR4][AuX4] + 3 R′SH f [AuSR′]n + R′SSR′ +
including Ag and Cu.5-11 Although these reactions have been NR4X + 3 HX (2)
studied extensively by many groups, and their products have been
thoroughly characterized,7,8,10,12,13 there remain significant ques- The stoichiometry of the reduction of [NR4][AuX4] to
tions regarding their detailed mechanisms. In particular, there is [NR4][AuX2] (Au3+ to Au1+) was confirmed by monitoring the
little definitively known regarding the precursor species present in disappearance of the 402 nm absorption band of the Au3+ upon
solution prior to reduction with NaBH4. When the precursor addition of alkanethiol (with the solution going from intense orange
question has been addressed, researchers have generally assumed to colorless). As in a typical two-phase Brust-Schiffrin reaction,
the formation of M(I) thiolate polymers in both the one- and two- tetraoctylammonium bromide (TOAB, 3 equiv) in toluene was
phase reactions (where M ) Au, Ag, or Cu). Given that the size rapidly stirred with hydrogen tetrachloroaurate (HAuCl4, 1 equiv)
and properties of the products of these syntheses are dependent in water until all Au3+ was transferred to the organic phase. The
upon the reaction conditions,12,14 and thus implicitly upon the [TOA][AuX4] complex formed, likely with a mix of Cl- and Br-
reaction intermediates, it is necessary to establish clearly the ions. After discarding the aqueous phase, aliquots of dodecanethiol
composition of precursor solutions. were titrated into the isolated toluene phase. After each addition,
We report here the identification and quantification of the the solution was stirred until reaction was complete, and its
precursor species of Au, Ag, and Cu Brust-Schiffrin nanoparticle UV-visible absorption spectrum was recorded. Complete reduction
syntheses. A particular focus is placed upon the two-phase Au from Au3+ to Au1+ occurs with the addition of 2 equiv of
synthesis, as it is the most widely employed variant. We conclude dodecanethiol (Figure S1). This is consistent with reaction 1 and
that, contrary to the assumptions of previous reports, Au(I) thiolate inconsistent with Au(I) thiolate formation proceeding via reaction
([AuSR]n) is not a measurable (1H NMR) precursor in this reaction 2. Further addition of alkanethiol produces no additional change
when it is performed under typical conditions.1,12 Rather, Au(I)- in color or apparent precipitation.
and Au(III)- tetraalkylammonium complexes are the relevant Au The species present during this reaction were quantitatively
species in solution prior to reduction with NaBH4. The intermediate monitored by 1H NMR spectroscopy. A solution of TOAX and
step in this method is shown to follow reaction 1 rather than the [TOA][AuX4] was prepared by phase transfer of AuCl4- (1 equiv)
oft-stated reaction 2.3,15 In accordance with this, a revised view of into deuterated toluene with TOAB (2.5 equiv). After sufficient
9582 9 J. AM. CHEM. SOC. 2010, 132, 9582–9584 10.1021/ja104011b  2010 American Chemical Society
 COMMUNICATIONS
With this information, the complications associated with the
phase transfer step of the Brust-Schiffrin synthesis can be avoided
by simply carrying out these reactions with pure samples of
[TOA][AuBr4] and [TOA][AuBr2]. These are readily prepared via
a modified literature approach.16 Addition of 3 equiv of dode-
canethiol to 1 equiv of [TOA][AuBr4] in deuterated toluene
generates a single equivalent of dodecyl disulfide (Figure S2). All
[TOA][AuBr4] is reduced to [TOA][AuBr2], and 1 equiv of thiol
remains unreacted. Furthermore, when 3 equiv of dodecanethiol
are added to 1 equiv of [TOA][AuBr2], no reaction occurs and 3
equiv of free thiol are observed. These experiments reinforce the
fact that the intermediate step in the two-phase Brust-Schiffrin
reaction indeed proceeds via reaction 1. Importantly, they also
suggest a new synthetic strategy with control over well-defined
precursor species.
When Ag and Cu are used in the Brust-Schiffrin reaction, as
[TOA][AgBr2] and [TOA]2[CuBr4] respectively, similar behavior
to the Au case is observed (Figures S3 and S4). [TOA][AgBr2] is
prepared by codissolving TOAB (5 equiv) and AgNO3 (1 equiv)
in toluene, without signs of AgBr precipitate formation. No reaction
occurs when this complex, with Ag already in a +1 oxidation state,
is mixed with dodecanethiol. On the other hand, the reaction of
[TOA]2[CuBr4], prepared according to a modified literature pro-
cedure,17 with dodecanethiol, leads to a one-electron reduction via
reaction 3. As in the Au case, the formation of Ag(I) and Cu(I)
Figure 1. 1
H NMR spectra of TOAX + [TOA][AuX4] solutions with (a)
0, (b) 1, (c) 2, (d) 3, and (e) 4 equiv of dodecanethiol added. Also shown, alkanethiolate polymers18-21 is prevented by the stabilization
for comparison, are the 1H NMR spectra of pure solutions of (f) TOAB, afforded by the TOA+ cation.
(g) dodecanethiol, and (h) dodecyl disulfide. * indicates water and solvent
peaks. [TOA]2[CuBr4] + RSH f 1/2 RSSR + [TOA][CuBr2] +
stirring, the phases were allowed to separate (ca. 2 min). The toluene TOAB + HBr (3)
phase was then transferred to a new reaction vessel using a Pasteur
pipet, and the water phase was discarded. Dodecanethiol was added Phosphonium salts demonstrate chemical behavior analogous to
to each of five equal aliquots of the toluene solution (0, 1, 2, 3, or that of ammonium salts. Tetraoctylphosphonium cation (TOP+) is
4 equiv relative to the Au in each aliquot). These samples were thus substituted for TOA+ in a modified Brust-Schiffrin reaction.
stirred until reaction was complete, and 1H NMR spectra were [TOP][AuBr4] is readily synthesized according to a modified
obtained (Figure 1). Comparison with spectra of pure samples of literature procedure,16 dissolved in deuterated toluene (1 equiv),
TOAB, dodecanethiol, and dodecyl disulfide reveals that the and reacted with 3 equiv of dodecanethiol. As expected, this leads
spectrum of the mixture of TOAX and [TOA][AuX4] (Figure 1a) to the production of 1 equiv of dodecyl disulfide, while 1 equiv
is similar to that of pure TOAB (Figure 1f). There are, however, of thiol remains unreacted (Figure S5). Correspondingly, all
significant shifts of all the resonance peaks due to the association [TOP][AuBr4] present is reduced to [TOP][AuBr2], leading to
with the AuX4- anion. TOA+ peaks are not observed for each of significant shifts of the TOP+ peaks and a complete loss of the
the anions, but rather a single averaged peak is observed for each solution’s intense orange color. Thus, TOP+ behavior clearly
of the different chemical environments of the cation. This is parallels that of TOA+ in modified Brust-Schiffrin reactions.
consistent with a fast anion exchange process. Finally, reaction 1 is also operative when phenylethanethiol
After the reaction of 1 equiv of dodecanethiol (Figure 1b), 0.5 (PET) is used in the Brust-Schiffrin reaction (Figure S6).22-24
equiv of dodecyl disulfide is generated through complete oxidation Reaction of [TOA][AuBr4] (1 equiv) in deuterated toluene with
of the added thiol. Correspondingly, 0.5 equiv of [TOA][AuX2] is PET (5 equiv) yields 1 equiv of disulfide, and 3 equiv of thiol
produced, leading to further shifts of the TOA+ peaks. Addition of remain unreacted. [TOA][AuBr4] is quantitatively reduced to
2 equiv of dodecanethiol to TOAX and [TOA][AuX4] (Figure 1c), [TOA][AuBr2], with a corresponding color loss. There is no
on the other hand, generates 1 equiv of dodecyl disulfide. All of evidence of Au(I) thiolate formation, even when this reaction is
the Au3+ is reduced to Au1+, again producing shifts in the peaks carried out following the low temperature procedure of Zhu et al.23
of the TOA+ cation and leaving a colorless solution. Finally, The method of Zhu et al.23 is reported to generate Au(I) thiolate
addition of 3 and 4 equiv of the thiol (Figure 1d and 1e, intermediates that, upon further reduction, produce Au25 clusters
respectively) results in all of the [TOA][AuX4] being converted to in high yield. This discrepancy may be explained by water
[TOA][AuX2] with the associated generation of 1 equiv of the remaining in the reaction, which promotes the generation of thiolate
disulfide. However, in these cases, 1 and 2 equiv of free thiol, aggregates (vide infra).
respectively, are also observed (R CH2 at ca. 2.18 ppm and SH at The identification and quantification of precursor species provide
ca. 1.06 ppm as per pure dodecanethiol). These results clearly show key information for understanding and controlling these reactions.
that polymeric Au(I) thiolate species are not generated in measurable The thiol-to-metal ratio influences the size and properties of the
quantities (i.e., reaction 2 is not operative). Rather, the intermediate products.8,12,14 However, the mechanistic origins of this effect
step of the Brust-Schiffrin synthesis proceeds via reaction 1, where remain unclear. Here, we have shown explicitly how altering the
Au(I) and Au(III) tetraalkylammonium complexes are the relevant thiol-to-metal ratio also alters the ratios of different metal precursors
Au precursors. and adsorbates. These ratios may be important determinants of the
J. AM. CHEM. SOC. 9 VOL. 132, NO. 28, 2010 9583
COMMUNICATIONS

kinetics of nanoparticle formation. The new synthetic strategy and its impact on nanoparticle synthesis, was also discussed.
presented here, with well-defined precursors, provides a key tool Strategies for avoiding the production of these insoluble species
for disentangling these two factors (i.e., the redox and passivation were suggested. It is expected that the results presented here will
kinetics). lead to modifications in the way that these important syntheses are
The intermediate step in two-phase Brust-Schiffrin reactions was conducted.
previously believed to involve the generation of soluble M(I)
Acknowledgment. The Natural Sciences and Engineering
alkanethiolate species. However, these materials have been studied
Research Council of Canada (NSERC) and the Centre for Self-
by several groups18,19,25,26 and have been shown to be insoluble
Assembled Chemical Structures (CSACS) are gratefully acknowl-
in common solvents, including toluene. In fact, the poor solubility
edged for financial support.
of M(I) thiolates, in general, is well-known, and has been attributed
to intermolecular interactions between adjacent polymer units, Supporting Information Available: Experimental and synthetic
including metallophilic interactions20,21,25,27-30 Accordingly, in details, UV-visible absorption, and 1H NMR data. This material is
these reactions, precipitation should be viewed as an important available free of charge via the Internet at http://pubs.acs.org.
indicator that polymeric M(I) thiolate species are forming. Cor-
respondingly, a lack of precipitate is an indicator that it is unlikely References
that they are forming. (1) Brust, M.; Walker, M.; Bethell, D.; Schiffrin, D. J.; Whyman, R. J. Chem.
M(I) thiolate species form in significant amounts in the inter- Soc., Chem. Commun. 1994, 801–802.
(2) Daniel, M. C.; Astruc, D. Chem. ReV. 2004, 104, 293–346.
mediate step of the two-phase Brust-Schiffrin reaction when larger (3) Sardar, R.; Funston, A. M.; Mulvaney, P.; Murray, R. W. Langmuir 2009,
quantities of water are present (though typical conditions, with small 25, 13840–13851.
(4) Brust, M.; Fink, J.; Bethell, D.; Schiffrin, D. J.; Kiely, C. J. Chem. Soc.,
amounts of water present, do not lead to measurable quantities of Chem. Commun. 1995, 1655–1656.
M(I) thiolates). This occurs both with and without a quaternary (5) Murthy, S.; Bigioni, T. P.; Wang, Z. L.; Khoury, J. T.; Whetten, R. L.
Mater. Lett. 1997, 30, 321–325.
cation being present. M(I) thiolates also form (as white precipitates) (6) Heath, J. R.; Knobler, C. M.; Leff, D. J. Phys. Chem. B 1997, 101, 189–
if the precursor M(I) solutions are exposed to water, or if toluene 197.
(7) Kang, S. Y.; Kim, K. Langmuir 1998, 14, 226–230.
is removed. Other polar solvents (methanol, ethanol) are also (8) Bauer, C. A.; Stellacci, F.; Perry, J. W. Top. Catal. 2008, 47, 32–41.
capable of promoting the formation of M(I) thiolate polymers. This (9) Garitaonandia, J. S.; Insausti, M.; Goikolea, E.; Suzuki, M.; Cashion, J. D.;
is attributed to the breaking or absence of the metal anion-quaternary Kawamura, N.; Ohsawa, H.; Gil de Muro, I.; Suzuki, K.; Plazaola, F.; Rojo,
T. Nano Lett. 2008, 8, 661–667.
cation ion-pairs that form in nonpolar solvents, including toluene. (10) Ang, T. P.; Wee, T. S. A.; Chin, W. S. J. Phys. Chem. B 2004, 108, 11001–
Drying toluene solutions prior to thiol addition could prove useful 11010.
(11) Dong, T.-Y.; Wu, H. H.; Song, J. M.; Chen, I. G.; Kao, T. H. Appl. Surf.
in minimizing M(I) thiolate formation in two-phase reactions. One- Sci. 2009, 255, 3891–3896.
phase Brust-Schiffrin reactions carried out in polar solvents (THF, (12) Hostetler, M. J.; Wingate, J. E.; Zhong, C.-J.; Harris, J. E.; Vachet, R. W.;
Clark, M. R.; Londono, J. D.; Green, S. J.; Stokes, J. J.; Wignall, G. D.;
ethanol, methanol) without phase transfer agents also generate M(I) Glish, G. L.; Porter, M. D.; Evans, N. D.; Murray, R. W. Langmuir 1998,
thiolate precipitates. These results suggest that M(I) thiolate 14, 17–30.
(13) Templeton, A. C.; Wuelfing, W. P.; Murray, R. W. Acc. Chem. Res. 2000,
formation is likely to occur when Au, Ag, or Cu salts are exposed 33, 27–36.
to thiols in the presence of highly polar solvents where ion-pairs (14) Leff, D.; Ohara, P. C.; Heath, J. R.; Gelbart, W. M. J. Phys. Chem. 1995,
99, 7036–7041.
are not prevalent. This general conclusion is supported by observa- (15) Alvarez, M. M.; Khoury, J. T.; Schaaff, T. G.; Shafigullin, M.; Vezmar, I.;
tions reported in the literature.25,31-34 Whetten, R. L. Chem. Phys. Lett. 1997, 266, 91–98.
(16) Braunstein, P.; Clark, J. H. J. Chem. Soc., Dalton Trans. 1973, 1845–
Some M(I) thiolate formation is likely to commonly occur when 1848.
excess water is not removed, or when aqueous NaBH4 solutions (17) Gill, N. S.; Taylor, F. B.; Hatfield, W. E.; Parker, W. E.; Fountain, C. S.;
Bunger, F. L. Inorg. Synth. 1967, 9, 136–142.
are added very slowly, or at low concentration, in two-phase (18) Dance, I. G.; Fisher, K. J.; Banda, R. M. H.; Scudder, M. L. Inorg. Chem.
Brust-Schiffrin syntheses. This also occurs when one-phase 1991, 30, 183–187.
nanoparticle syntheses are carried out with thiols.24,32,34 The (19) Voicu, R.; Badia, A.; Lennox, R. B.; Ellis, T. H. Chem. Mater. 2000, 12,
2646–2652.
consequences of this are not yet fully understood for the wide (20) Duncan, W. E.; Ott, E.; Reid, E. E. Ind. Eng. Chem. 1931, 23, 381–384.
variety of synthetic methods that exist. However, it can be (21) Eller, P. G.; Kubas, G. J. J. Am. Chem. Soc. 1977, 99, 4346–4351.
(22) Donkers, R. L.; Lee, D.; Murray, R. W. Langmuir 2004, 20, 1945–1952.
reasonably expected to produce poor synthetic outcomes in some (23) Zhu, M.; Lanni, E.; Garg, N.; Bier, M. E.; Jin, R. J. Am. Chem. Soc. 2008,
cases. Partial M(I) thiolate formation generates unknown mixtures 130, 1138–1139.
(24) Wu, Z.; Suhan, J.; Jin, R. J. Mater. Chem. 2009, 19, 622–626.
of multiple metal precursors with different redox potentials and (25) Cha, S.-H.; Kim, J.-U.; Kim, K.-H.; Lee, J.-C. Chem. Mater. 2007, 19,
states of aggregation. It also produces mixtures of ligands with 6297–6303.
(26) Sandhyarani, N.; Pradeep, T. J. Mater. Chem. 2001, 11, 1294–1299.
different adsorption properties. The heterogeneous reaction of (27) Al-Sa’ady, A. K. H.; Moss, K.; McAuliffe, C. A.; Parish, R. V. J. Chem.
insoluble M(I) thiolates with reductants can be very slow and, in Soc., Dalton Trans. 1984, 1609–1616.
(28) Corbierre, M. K.; Lennox, R. B. Chem. Mater. 2005, 17, 5691–5696.
some cases, may not occur at all. These factors are likely to lead (29) Schmidbaur, H. Gold Bull. 2000, 33, 3–10.
to higher polydispersity, reduced yields, and insoluble metal and (30) Hermann, H. L.; Boche, G.; Schwerdtfeger, P. Chem.sEur. J. 2001, 7,
5333–5342.
M(I) thiolate byproducts.22,24,34,35 (31) Al-Sa’ady, A. K. H.; McAuliffe, C. A.; Parish, R. V.; Sandbank, J. A.
In summary, a revised view of Brust-Schiffrin nanoparticle Inorg. Synth. 1985, 23, 191–194.
(32) Schaaff, T. G.; Knight, G.; Shafigullin, M. N.; Borkman, R. F.; Whetten,
syntheses has been detailed here. Precursor species of these reactions R. L. J. Phys. Chem. B 1998, 102, 10643–10646.
have been identified and quantified. Tetraalkylammonium metal (33) Susha, A. S.; Ringler, M.; Ohlinger, A.; Paderi, M.; LiPira, N.; Carotenuto,
complexes are shown to be precursors of two-phase reactions, while G.; Rogach, A. L.; Feldmann, J. Chem. Mater. 2008, 20, 6169–6175.
(34) Reilly, S. M.; Krick, T.; Dass, A. J. Phys. Chem. C 2010, 114, 741–745.
M(I) thiolates are shown to be precursors of one-phase reactions (35) Templeton, A. C.; Chen, S.; Gross, S. M.; Murray, R. W. Langmuir 1999,
conducted in polar solvents. A new synthetic strategy employing 15, 66–76.
well-defined precursors was introduced. M(I) thiolate formation, JA104011B

9584 J. AM. CHEM. SOC. 9 VOL. 132, NO. 28, 2010