Photochemistry of neutral isonitrile gold(I) complexes: Modulation of photoreactivity by aurophilicity and π-acceptance ability

Article abstract of DOI:10.1021/ja0703933

This work demonstrates for the first time that aurophilicity and ligand π-acceptance ability sensitize the photoreactivity of Au(I) complexes. Photolysis of LAu^ICl (L = RNC or CO) complexes leads to free L, Au^III, and Au⁰ photoproducts. Solutions of (p-tosyl)CH₂NCAuCl in dichloromethane undergo significant oligomerization leading to dimers and trimers with formation constants of 1.61 × 10³ and 6.61 × 10³ M^-1, respectively, representing the highest values reported to date for complexes that exhibit aurophilic association in solution. The photoproduct quantum yield (Φ) varies with the LAu^ICl concentration in solution. For (p-tosyl)CH₂NCAuCl, metallic gold forms with Φ = 0.0065 and 0.032 in 4.0 × 10^-5 and 4.0 × 10^-3 M dichloromethane solutions, respectively. Meanwhile, irradiation of t-BuNCAuCl primarily produces t-BuNCAuCl₃ with Φ = 0.0045 and 0.013 for 5.0 × 10^-5 and 5.0 × 10^-3 M dichloromethane solutions, respectively. For Au(CO)Cl, metallic gold forms with Φ = 0.013 and 0.065 upon irradiation of 8.0 × 10^-5 and 8.0 × 10^-3 M dichloromethane solutions, respectively. Hence, *[LAuX]_n oligomeric species are more photoreactive than monomeric species. The results also demonstrate intuitive control of Φ via modulation of the π-acceptance ability of L, as both follow CO > (p-tosyl)CH₂NC > (alkyl)NC in LAuCl, a trend that is also commensurate with the relative long-term photosensitivity of the corresponding solids and solutions. A new method for preparing stable small gold nanoparticles is described based on the fundamental findings above. Thus, photolysis of different concentrations of LAuX in solutions containing a primary amine-terminated dendrimer leads to clear solutions exhibiting tunable visible plasmon absorptions of gold nanoparticles; these solutions maintain their colors and stability indefinitely. TEM measurements for representative samples prepared by photolysis of (p-tosyl)CH₂NCAuCl solutions give rise to spherical nanoparticles as small as 5 nm.

Full text of DOI:10.1021/ja0703933

Published on Web 08/23/2007
Photochemistry of Neutral Isonitrile Gold(I) Complexes:
Modulation of Photoreactivity by Aurophilicity and
π-Acceptance Ability
Oussama Elbjeirami and Mohammad A. Omary*
Contribution from the Department of Chemistry, UniVersity of North Texas,
Denton, Texas 76203
Received January 18, 2007; E-mail: omary@unt.edu
Abstract: This work demonstrates for the first time that aurophilicity and ligand π-acceptance ability sensitize
the photoreactivity of Au(I) complexes. Photolysis of LAuICl (L ) RNC or CO) complexes leads to free L,
AuIII, and Au0 photoproducts. Solutions of (p-tosyl)CH2NCAuCl in dichloromethane undergo significant
oligomerization leading to dimers and trimers with formation constants of 1.61 × 103 and 6.61 × 103 M-1
,
respectively, representing the highest values reported to date for complexes that exhibit aurophilic association
in solution. The photoproduct quantum yield (Φ) varies with the LAuICl concentration in solution. For (p-
tosyl)CH2NCAuCl, metallic gold forms with Φ ) 0.0065 and 0.032 in 4.0 × 10-5 and 4.0 × 10-3
M
dichloromethane solutions, respectively. Meanwhile, irradiation of t-BuNCAuCl primarily produces t-
BuNCAuCl3 with Φ ) 0.0045 and 0.013 for 5.0 × 10-5 and 5.0 × 10-3 M dichloromethane solutions,
respectively. For Au(CO)Cl, metallic gold forms with Φ ) 0.013 and 0.065 upon irradiation of 8.0 × 10-5
and 8.0 × 10-3 M dichloromethane solutions, respectively. Hence, *[LAuX]n oligomeric species are more
photoreactive than monomeric species. The results also demonstrate intuitive control of Φ via modulation
of the π-acceptance ability of L, as both follow CO > (p-tosyl)CH2NC > (alkyl)NC in LAuCl, a trend that is
also commensurate with the relative long-term photosensitivity of the corresponding solids and solutions.
A new method for preparing stable small gold nanoparticles is described based on the fundamental findings
above. Thus, photolysis of different concentrations of LAuX in solutions containing a primary amine-
terminated dendrimer leads to clear solutions exhibiting tunable visible plasmon absorptions of gold
nanoparticles; these solutions maintain their colors and stability indefinitely. TEM measurements for
representative samples prepared by photolysis of (p-tosyl)CH2NCAuCl solutions give rise to spherical
nanoparticles as small as 5 nm.
compounds.8 Thus, one can obtain dimers, larger oligomers, one-
dimensional extended-chain polymers, or two-dimensional
Introduction
Metal isonitriles have traditionally been strongly overshad-
owed by metal carbonyls, probably because of the repulsive
scent of isonitriles and not necessarily due to scientific reasons.1
Exploring the chemistry of metal isonitriles has led to new
advances in both science and technology. For example, Car-
diolite, which has been approved for use in cardiac imaging, is
an isonitrile complex.2 Gold(I) isonitrile complexes are used
for the deposition of gold films3,4 and preparation of new liquid
crystalline phases.5-7 Elegant work pioneered by Schmidbaur
and co-workers has demonstrated fascinatingly diverse su-
pramolecular structures exhibited by the Au(RNC)X class of
polymeric sheets by varying the R group and X.8-19
Gold(I) compounds can be active in homogeneous catalysis,
overcoming previous problems associated with using Au(III)
in catalysis due to its quick reduction to metallic gold.20 The
(8) Schneider, W.; Angermaier, K.; Sladek, A.; Schmidbaur, H. Z. Naturforsch.
B 1996, 51, 790.
(9) Mathieson, T. J.; Langdon, A. G.; Milestone, N. B.; Nicholson, B. K. J.
Chem. Soc., Dalton Trans. 1999, 201.
(10) Bonati, F.; Minghetti, G. Gazz. Chim. Ital. 1973, 103, 373.
(11) Lentz, D.; Willemsen, S. J. Organomet. Chem. 2000, 612, 96.
(12) Eggleston, D. S.; Chodosh, D. F.; Webb, R. L.; Davis, L. L. Acta
Crystallogr. 1986, C42, 36.
(13) Ecken, H.; Olmstead, M. M.; Noll, B. C.; Attar, S.; Schlyer, B.; Balch, A.
L. J. Chem. Soc., Dalton Trans. 1998, 3715.
(14) Perreault, D.; Drouin, M.; Michel, A.; Harvey, P. D. Inorg. Chem.1991,
30, 2.
(15) Siemeling, U.; Rother, D.; Bruhn, C.; Fink, H.; Weidner, T.; Traeger, F.;
Rothenberger, A.; Fenske, D.; Priebe, A.; Maurer, J.; Winter, R. J. Am.
Chem. Soc. 2005, 127, 1102.
(16) Elbjeirami, O.; Omary, M. A.; Stender, M.; Balch, A. L. Dalton Trans.
2004, 3173.
(17) White-Morris, R. L.; Olmstead, M. M.; Balch, A. L.; Elbjeirami, O.; Omary,
M. A. Inorg. Chem. 2003, 42, 6741.
(18) Humphrey, S. M.; Mack, H. G.; Redshaw, C.; Elsegood, M. R. J.; Young,
K. J. H.; Mayer, H. A.; Kaska, W. C. Dalton Trans. 2005, 439.
(19) Schneider, D.; Schuster, O.; Schmidbaur, H. Organometallics 2005, 24,
3547.
(1) A recent paper has shown that fragrant isonitriles can actually be prepared;
see: Pirrung, M. C.; Ghorai. S. J. Am. Chem. Soc. 2006, 128, 11772.
(2) Du Pont Merck Pharmac. Co. Chem. Eng. News 1991, Jan. 14, 24.
(3) Norton, P. R.; Young, P. A.; Cheng, Q.; Dryden, N.; Puddephatt, R. J.
Surf. Sci. 1994, 307, 172.
(4) Vaughan, L. G. U.S. Patent 3661959 19720509, 1972.
(5) Bachman, R. E.; Fioritto, M. S.; Fetics, S. K.; Cocker, T. M. J. Am. Chem.
Soc. 2001, 123, 5376.
(6) Ishi, R.; Kaharu, T.; Pirio, N.; Zhang, S. W.; Takahashi, S. J. Chem. Soc.,
Chem. Commun. 1995, 1215.
(7) Kaharu, T.; Ishii, R.; Takahashi, S. J. Chem. Soc., Chem. Commun. 1994,
1349.
9
11384
J. AM. CHEM. SOC. 2007, 129, 11384-11393
10.1021/ja0703933 CCC: $37.00 © 2007 American Chemical Society

Photochemistry of Neutral Isonitrile Au(I) Complexes
A R T I C L E S
photochemical deposition of metallic gold can be utilized for
the fabrication of electronic devices.21 Pioneering work by the
Vogler group investigated Au(CO)Cl and showed that this
complex exhibits a longest-wavelength absorption at λmax ) 250
nm that was assigned to a metal-centered ds transition.22 The
photolysis of Au(CO)Cl in CH2Cl2 leads to formation of AuCl
and CO; the release of the latter is similar to photoreactions
known for other metal carbonyl complexes in which ligand-
field excitation leads to the population of M-L σ* orbitals.22
The initial photoproduct AuCl was proposed to disproportionate
into metallic gold and gold(III) chloride.22 The electronic
structure of RNC is believed to be rather similar to that of CO;
thus, the reactivities should be similar for complexes of both
ligand types.23 By analogy to the aforementioned Au(CO)Cl
photochemical study, we have decided to investigate the
compounds RNCAuCl (R ) t-Bu, Me, (p-tosyl)CH2). To the
best of our knowledge, photochemical studies of isonitrile Au-
(I) complexes have never been reported prior to this effort.
on the photophysical properties,13-19,32 the present work rep-
resents the first study that examines the effect of aurophilicity
on the photochemical reactivity of Au(I) complexes. In addition,
we systematically examine the effect of the electronic factor
on the photoproduct quantum yield in RNCAuCl and Au(CO)-
Cl complexes because the neutral ligands offer a range of
σ-donation/π-acceptance ability.23,33,34 We also report the
utilization of these fundamental findings to prepare small gold
nanoparticles stabilized in solutions of primary amine dendrim-
ers. While gold is inert in its bulk form, size reduction to the
nanoscale is known to lead to immense applications that range
from catalysis and photonics to cancer imaging and therapy.35,36
Common methods for preparing stable gold nanoparticles are
based on chemical or electrochemical reduction35,36 while, to
our knowledge, this is the first example by which photochem-
istry of Au(I) complexes is used to prepare stable gold
nanopatricles without using reducing agents.
Experimental Section
Yersin and Gliemann demonstrated that the emissions of
various [Pt(CN)4]2- salts occur as a result of Pt‚‚‚Pt extended-
chain interactions in crystals.24 It was suggested that association
of such compounds might also occur in solution, thus affecting
their photophysical behavior.25 Related studies by Adamson and
co-workers indeed demonstrated a non-Beer’s law behavior of
the electronic absorption spectral bands for aqueous solutions
of K2Pt(CN)4 and BaPt(CN)4 due to oligomerization in solution,
whichconsequentlyaffectedthephotophysical26aandphotochemical26b
behavior. In a similar fashion, Rh‚‚‚Rh interactions in Rh(I)
isocyanide complexes were shown to persist in solution with a
non-Beer’s law behavior of the electronic absorption spectral
bands.27 Another study by Pfab and Gerhardt demonstrated that
even single crystals of the tetracyanoplatinates(II) exhibit
photochemical activity (photooxidation).28 Inspired by these
studies for d8 complexes, we have decided to investigate whether
Au‚‚‚Au interactions take place in solution and whether they
influence the photochemistry of the Au(I) complexes studied
herein. This is a particularly compelling study due to the
significance of such interactions in Au(I) complexes, which have
led to the “aurophilic attraction” concept originally coined by
Schmidbaur.29 This interaction type falls under the general
category of “metallophilic bonding”, which has been reviewed
by Pyykko.30,31 While multiple spectroscopic investigations have
been carried out for isonitrile and carbonyl Au(I) complexes
and illustrated the strong influence played by aurophilic bonding
Materials and Syntheses. The preparations of RNCAuCl (R ) (p-
tosyl)CH2, Me, and t-Bu) followed procedures reported previously.12,16
The isonitriles (p-tosyl)CH2NC and t-BuNC were purchased from
ACROS Organics while Au(CO)Cl was purchased from Strem Chemi-
cals. The dendrimer 16-cascade/1,4-diaminobutane[4-N,N,N′,N′]/(1-
azabutylidyne)2/aminopropane was purchased from Sigma-Aldrich
under the commercial name “DAB-Am-16, Polypropylenimine hexa-
decaamine”, G 3.0; henceforth we will refer to this dendrimer as “DAB-
Am-16”. Au(tetrahydrothiophene)Cl and MeNC were synthesized by
slight modifications of published procedures.37,38 HPLC grade solvents
(acetonitrile, methanol, and dichloromethane) were distilled from
conventional drying agents, degassed by the freeze-pump-thaw
method three times prior to use, and kept under argon. All electronic
absorption measurements were carried out under argon. Au(CO)Cl is
light and moisture sensitive so the solution preparation was carried
out in a glovebox and the samples were stored in the dark under argon.
CAUTION! Isonitrile compounds are toxic and need to be handled with
adequate Ventilation due to their stench; extra care should be taken to
aVoid contact with solids and solutions containing such compounds.
Photolysis Reactions. Solutions of compounds studied photochemi-
cally were freshly prepared from dried CH2Cl2 under argon. A 5-mL
or 25-mL aliquot of each solution was irradiated with UV light by
using a Hanovia 450 W medium-pressure Hg lamp at ambient
temperature with no heat detected. Changes in the electronic absorption
spectra during the photolysis reactions were monitored using a Perkin-
Elmer Lambda-900 double-beam UV/vis/NIR absorption spectropho-
tometer. HPLC data were collected using a Shimadzu SPD-6A
spectrophotometer coupled with an LC-600 Shimadzu liquid chromato-
gram detector. A continuous flow of 0.5 mL/min was maintained using
(20) Teles, J. H.; Brode, S.; Chabanas, M. Angew. Chem., Int. Ed. 1998, 37,
1415.
(21) (a) D’Amico, J. F.; De Angelo, M. A.; Henrickson, J. F.; Kenney, J. T.;
Sharp, D. J. J. Electrochem. Soc. 1970, 118, 1695. (b) D’Amico, J. F.;
Litt, F. A.; De Angelo, M. A. J. Electrochem. Soc. 1972, 119, 256.
(22) (a) Kunkely, H.; Vogler. A. J. Organomet. Chem. 1997, 541, 177 (b) Vogler,
A.; Kunkely, H. Coord. Chem. ReV. 2001, 219, 489.
(32) Elbjeirami, O.; Yockel, S; Campana, C. F.; Wilson, A. K.; Omary, M. A.
Organometallics 2007, 26, 2550.
(33) Cotton. A. F.; Zingalez, F. J. Am. Chem. Soc. 1961, 83, 351.
(34) Csonka, I. P.; Szepes, L.; Modelli, A. J. Mass Spectrom. 2004, 39, 1456.
(35) For some representative examples, see: (a) Huang, X.; El-Sayed, I. H.;
Qian, W.; El-Sayed, M. A. J. Am. Chem. Soc. 2006, 128, 2115. (b) Zheng,
J.; Zhang, C.; Dickson, R. M. Phys. ReV. Lett. 2004, 93, 077402. (c) Li, J.;
Li, X.; Zhai, H. J.; Wang, L. S. Science 2003, 299, 864. (d) Chen, S. W.;
Ingram, R. S.; Hostetler, M. J.; Pietron, J. J.; Murray, R. W.; Schaaff, T.
G.; Khoury, J. T.; Alvarez, M. M.; Whetten, R. L. Science 1998, 280, 2098.
(e) Zheng, N,; Stucky, G. J. Am. Chem. Soc. 2006, 128, 14278. (f) An, Z.;
Tang, W.; Hawker, C. J.; Stucky, G. J. Am. Chem. Soc. 2006, 128, 15054.
(g) Goodman, D. W.; Chen, M. S. Science 2004, 306, 252. (h) Goodman,
D. W.; Choudhary, T. V. Appl. Catal. 2005, 291, 32. (i) Loo, C.; Lowery,
A.; Halas, N.; West, J.; Drezek, R. Nano Lett. 2005, 5, 709.
(23) Malatesta, L.; Bonati, F. Isocyanide Complexes of Metals; Interscience:
New York, 1969.
(24) For an example, see: (a) Yersin, H. J. Chem. Phys. 1978, 68, 4707. For
reviews, see: (b) Gliemann, G.; Yersin, H. Struct. Bonding 1985, 62, 87.
(c) Yersin, H.; Gliemann, G. Ann. N.Y. Acad. Sci. 1978, 313, 539.
(25) Gliemann, G.; Lechner, A. J. Am. Chem. Soc. 1989, 111, 7469.
(26) (a) Schindler, J. W.; Fukuda, R. C.; Adamson, A. W. J. Am. Chem. Soc.
1982, 104, 3596. (b) Schindler, J. W.; Adamson, A. W. Inorg. Chem. 1982,
21, 4236.
(27) Mann. K. R.; Lewis, N. S.; Williams, R. M.; Gray, H. B.; Gordon, J. G.,
II. Inorg. Chem. 1978, 17, 828.
(36) For reviews, see: (a) El-Sayed, M. A. Acc. Chem. Res. 2001, 34, 257. (b)
Kreibig, U.; Vollmer, M. Optical Properties of Metal Clusters; Springer-
Verlag: Berlin, 1995.
(37) Uso´n, R.; Laguna, A. Organometallic Syntheses; King, R. B., Eisch, J. J.,
Eds.; Elsevier: Amsterdam, 1986.
(38) Schuster, R. E.; Scott, J. E.; Cazanova, J. Org. Synth. 1966, 46, 75.
(28) Pfab, W. A.; Gerhardt, V. J. Lumin. 1984, 31/32, 582.
(29) Schmidbaur, H. Chem. Soc. ReV. 1995, 24, 391.
(30) Pyykko¨, P. Chem. ReV. 1997, 97, 599.
(31) (a) Pyykko¨, P. Angew. Chem., Int Ed. 2004, 43, 4412. (b) Pyykko¨, P. Inorg.
Chim. Acta 2005, 358, 4113.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11385

A R T I C L E S
Elbjeirami and Omary
Table 1. Deviation From Beer’s Law in Dilute Solutions of
(p-tosyl)CH2NCAuCl and t-BuNCAuCl
(p-tosyl)CH2NCAuCl
t-BuNCAuCl
1
1
co, M
ꢀ227, M-1 cm-
co, M
ꢀ230, M-1 cm-
3.2 × 10-6
5.2 × 10-6
5.2 × 10-5
5.2 × 10-4
3.5 × 104
7.2 × 103
4.4 × 103
1.3 × 103
2.3 × 10-6
3.8 × 10-6
3.8 × 10-5
3.8 × 10-4
8.4 × 104
5.4 × 104
1.2 × 104
4.7 × 103
HPLC-grade methanol. The column used was Discovery C-18 (5 cm
× 4.6 mm × 5 µm particle size).
Dendrimer-stabilized gold nanoparticles were prepared in a two-
step process. First, a 1-mL aliquot of a 1 × 10-3 M solution of DAB-
Am-16 in methanol was mixed with 4-mL solutions having different
concentrations of (p-tosyl)CH2NCAuCl in dichloromethane. Second,
the resulting 5-mL solution was irradiated at ambient temperature with
the aforementioned Hg lamp. After 1-2 h of exposure, the solutions
were analyzed by electronic absorption spectroscopy. The morphology
and size of the gold nanoparticles were determined via transmission
electron microscopy (TEM) using a 200 kV high-resolution imaging
TEM (JEOL 2100F). The solution of each sample was dispersed and
dried on a carbon-coated Cu grid.
Figure 1. Electronic absorption spectra vs concentration of (p-tosyl)CH2-
NCAuCl in dichloromethane at room temperature using a 1-cm cuvette.
The total (p-tosyl)CH2NCAuCl concentration is (a) 0.0468 M, (b) 0.0234
M, (c) 0.0156 M, (d) 0.009 36 M, (e) 0.006 68 M, (f) 0.005 85 M, (g)
0.005 21 M, and (h) 5.21 × 10-4 M.
Results
constant given by eq 2, wherein c1 and cn represent the
concentration of the monomer and n-mer, respectively.
Spectroscopic Studies. Dilute solutions of (p-tosyl)CH2-
NCAuCl exhibit electronic absorption spectra with a major band
at 225-227 nm (ꢀmax ) 3.3 × 104 M-1 cm-1 in acetonitrile)
assignable to charge-transfer transitions in monomeric mol-
ecules.32,39 Similarly, dilute solutions of t-BuNCAuCl and
MeNCAuCl exhibit electronic absorption spectra with major
bands around 230-240 nm (ꢀmax ) 8.4 × 104 and 3.2 × 104
M-1 cm-1 in acetonitrile, respectively). The effective chro-
mophore in the three complexes is the CsNtCsAuICl unit
while slight absorption changes are expected upon varying the
R group on the ligand. On the other hand, dilute solutions of
Au(CO)Cl in acetonitrile exhibit electronic absorption spectra
with major bands in the 208-222 nm range (ꢀmax ) 1.95 ×
104 M-1 cm-1), similar to the absorption data reported by
Kunkely and Vogler for this compound.22 The effect of
association of Au(CO)Cl on the absorption and luminescence
of this compound in solution and the solid state has been
addressed in an experimental/theoretical study that we have
recently reported,32 whereas in the work herein we examine the
solution association for RNCAuCl complexes.
Intermolecular interactions lead to deviations from Beer’s law.
Table 1 shows some representative data for the neutral isonitrile
complexes studied herein and illustrates a clear negative
deviation from Beer’s law for the monomer bands. This is not
due to real limitations of Beer’s law, which are normally
encountered when the analyte concentration is 0.01 M or
higher,40 whereas we observe the negative deviation at much
lower concentrations (Table 1). Thus, we attribute the deviation
from Beer’s law in this study to oligomerization of monomeric
RNCAuCl complexes, for which aurophilic bonding in solution
is at least one contributing factor in addition to other possible
intermolecular forces such as multipolar interactions.41,17
The oligomerization process for LAuICl complexes can be
represented by eq 1, with a monomer-“n-mer” equilibrium
n[LAuICl ] a [LAuICl ]n
(1)
(2)
n
K1n ) cn/c1
At higher concetrations, the tendency for oligomer formation
increases. This is evident by the appearance of lower-energy
absorption bands (e.g., at 275, 320, and 335 nm for (p-tosyl)-
CH2NCAuCl; see Figure 1). Assuming that the peak at 275 nm
is due to an oligomer, a general quantitative formula that
describes the equilibrium shown in eq 2 can be derived as eq
3,26a,27,42 wherein co is the initial concentration of (p-tosyl)CH2-
NCAuCl, A is the maximum absorbance, ꢀn is the molar
absortiviy of the n-mer, and b is the light path.
coA-1/n ) (n/{ꢀnb})A(n-1)/n + (K1nꢀnb)-1/n
(3)
A reasonable fit (R2 ) 0.999) is obtained by plotting coA-1/2
vs A1/2 (Figure 2a), therefore suggesting that the peak at 275
nm is due to a [(p-tosyl)CH2NCAuCl]2 dimer. A value of (1.61
( 0.34) × 103 M-1 is obtained for the formation constant of
the dimer, K12, at ambient temperature. The oligomer responsible
for the 320 and 335 nm bands is determined to correspond to
a trimer according to our quantitative analysis. These trimer
bands grow along with the dimer band at 275 nm but with a
different concentration dependence. In the determination of K13,
we have subtracted the concentration of the dimer (using the
derived value of K12) from the initial concentration of (p-tosyl)-
CH2NCAuCl.42 Using the corrected co, plotting coA-1/3 vs A2/3
(Figure 2b) gives a linear equation with a reasonable fit (R2 )
0.997). A value of (6.61 ( 0.40) × 103 M-1 for the formation
constant of the trimer, K13, is obtained at ambient temperature.
We obtain further thermodynamic parameters for the solution
association by carrying out these measurements versus temper-
ature. Thus, Table 2 summarizes the formation constants, molar
(39) Chastain, S. K.; Mason, W. R. Inorg. Chem. 1982, 21, 3717.
(40) Skoog, D. A.; Holler, F. J.; Nieman, T. A. Principles of Instrumental
Analysis, 5th ed.; Harcourt Brace: Philadelphia, 1998.
(41) Liau, R.-Y.; Mathieson, T.; Schier, A.; Berger, R. J.; Runeberg, N.;
Schmidbaur, H. Z. Naturforsch. B 2002, 57, 881.
(42) Rawashdeh-Omary, M. A.; Omary, M. A.; Patterson, H. H. J. Am. Chem.
Soc. 2000, 122, 10371.
9
11386 J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007

Photochemistry of Neutral Isonitrile Au(I) Complexes
A R T I C L E S
Table 2. A Summary of the Formation Constants, Molar Absorptivities, Free Energies, Enthalpies, and Entropies for the Formation of a
[(p-tosyl)CH2NCAuCl]2 Dimer (λmax ) 275 nm) and a [(p-tosyl)CH2NCAuCl]3 Trimer (λmax ) 320 nm) in Solution
∆
G1n (298 K),
∆
H1n (298 K),
∆
S1n (298 K),
1
ꢀn, M-1 cm-
K1n
kJ/mol
kJ/mol
J/K−mol
dimer
trimer
176 ( 5
56.1 ( 2.0
(1.61 ( 0.34) × 103 M-1
-18.3 ( 0.5
-21.8 ( 0.2
-28.6 ( 2.0
-36.5 ( 0.8
-34.9 ( 0.7
-49.5 ( 0.3
(6.61 ( 0.40) × 103 M-2
Figure 3. Absorption spectral changes during the photolysis of a 4.0 ×
10-3 M solution of (p-tosyl)CH2NCAuCl in dichloromethane under argon
at room temperature using a 1-cm cuvette after (a) 0, (b) 3, (c) 10, (d) 16,
and (e) 30 min of irradiation with a 450-W Hg lamp.
Table 3. A Summary of the Photoproduct Quantum Yields Φ(Au0)
and Φ(LAuCl3) vs Concentrations of (p-tosyl)CH2NCAuCl,
t-BuNCAuCl, MeNCAuCl, and Au(CO)Cl
c0, M
Φ
(Au0)
Φ(LAuCl3)
(p-tosyl)CH2NCAuCl
t-BuNCAuCl
MeNCAuCl
4.0 × 10-3
4.0 × 10-5
5.0 × 10-3
5.0 × 10-5
5.0 × 10-3
5.0 × 10-5
8.0 × 10-3
8.0 × 10-5
0.032
0.0065
0.0087
0.0021
N/A
N/A
0.065
0.013
N/A
N/A
Figure 2. (a) A plot of coA-1/2 vs A1/2 with absorbance values taken at
275 nm, characteristics of a [(p-tosyl)CH2NCAuCl]2 dimer. (b) A plot of
coA-1/3 vs A2/3 with absorbance values taken at 320 nm, characteristics of
a [(p-tosyl)CH2NCAuCl]3 trimer. Absorption data in both plots have been
taken at ambient temperature.
0.013
0.0045
0.0013
N/A
0.041
0.0088
Au(CO)Cl
absorptivities, free energies, enthalpies, and entropies for the
formation of [(p-tosyl)CH2NCAuCl]2 and [(p-tosyl)CH2NCAuCl]3
oligomers in solution.
photolysis of (p-tosyl)CH2NCAuCl solutions. However, evi-
dence of photodissociation of the free (p-tosyl)CH2NC ligand
has been obtained by HPLC measurements. Thus, HPLC
chromatograms for solutions of (p-tosyl)CH2NCAuCl show no
peaks before photolysis while the photoproducts exhibit a peak
with a retention time of 3.76 min. This peak is consistent with
the free isonitrile ligand because a separate measurement for
(p-tosyl)CH2NC shows one peak with a retention time of 3.80
min; see Figure S2 in the Supporting Information.
Photochemistry of (p-tosyl)CH2NCAuCl. Irradiation of
dichloromethane solutions of this complex leads to photolysis
that proceeds slowly, as indicated by the spectral changes for a
4.0 × 10-3 M solution after the first few exposure periods
(Figure 3). The major near-UV band with λmax ) 320-350 nm
and the weaker shoulder that extends to the visible region
(∼440-470 nm) are signatures of charge transfer and ligand
field transitions, respectively, for square-planar Au(III) com-
plexes that contain chloride ligands (AuCl4-, Au2Cl6, or
LAuCl3).43,44 The dominant molecular form of the square-planar
Au(III) species in the photoproducts here is likely RNCAuIII-
Cl3 in equilibrium with the free ligand and other Au(III) forms;
see the Discussion section. The free (p-tosyl)CH2NC ligand has
an absorption band at ∼340 nm (see Figure S1 in the Supporting
Information). Although its energy coincides with the absorptions
Analysis of the photolysis data in Figure 3 gives rise to Φ )
0.056 for the formation of (p-tosyl)CH2NCAuCl3. The solution
color becomes darker upon continued irradiation, giving rise to
broad peaks with λmax ∼585 nm. These peaks are attributed to
plasmon absorptions of colloidal gold particles, which form with
Φ ) 0.032. The increase in the baseline absorbance is a known
consequence of the production of metallic gold particles.22,43
Disproportionation of [AuICl]n species to form AuIIIand Au0
products is proposed as a secondary reaction governed by the
thermodynamic instability22,43 of [AuICl]n following photodis-
sociation of the donor ligand. A 4.0 × 10-5 M solution exhibits
a similar photochemical behavior with a very low efficiency
(Φ ) 0.0065 for metallic gold, Table 3).
of the photoproucts, this band is too weak (ꢀ ) 8.70 M-1 cm-1
)
to deduce the presence of the free ligand in equilibrium with
an Au(III) complex based on absorption changes during
(43) Puddephatt, R. J. The Chemistry of Gold; Elsevier: New York, 1978.
(44) (a) Gangopadhayay, A. K.; Chakravorty, A. J. Chem. Phys 1961, 35, 2206.
(b) Potts, R. A. J. Inorg. Nucl. Chem. 1972, 34, 1749. (c) Nalbandia, L.;
Boghosian, S.; Papatheodorou, G. N. Inorg. Chem. 1992, 31, 1769.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11387

A R T I C L E S
Elbjeirami and Omary
Figure 5. Absorption spectral changes during the photolysis of a 5.0 ×
10-3 M solution of MeNCAuCl in dichloromethane under argon at room
temperature using a 1-cm cuvette after (a) 0, (b) 2, (c) 5, (d) 10, (e) 15, (f)
25, and (g) 40 min of irradiation with a 450-W Hg lamp.
Figure 4. Absorption spectral changes during the photolysis of a 5.0 ×
10-3 M solution of t-BuNCAuCl in dichloromethane under argon at room
temperature using a 10-cm cuvette after (a) 0, (b) 11, (c) 35, (d) 80, and
(e) 125 min of irradiation with a 450-W Hg lamp.
Photochemistry of RNCAuCl (R ) t-Bu and Me). Figure
4 shows photolysis results for a 5.0 × 10-3 M solution of
t-BuNCAuCl in dichloromethane. The solution exhibits slow
photooxidation after several light exposure periods as indicated
by the formation of a new band with λmax near 325 nm attributed
to t-BuNCAuCl3, which forms with Φ ) 0.013. This absorption
is sufficiently strong so as to mask the weak structured
absorption of the complex before irradiation, which likely
represents an S0 f T1 monomer transition due its low extinction
coefficient (20.8 M-1 cm-1) and negative deviation from Beer’s
law (ꢀ336.0 decreases to 7.3 M-1 cm-1 for a 9.0 × 10-3
M
solution). With continued irradiation, the solution color becomes
darker due to formation of colloidal gold particles with λmax
)
602 nm and a calculated Φ ) 0.0087. In acetonitrile, the reaction
is less efficient since the solvent can coordinate to stabilize the
AuCl transient species before it can disproportionate. At a lower
concentration (5.0 × 10-5 M) in dichloromethane, photolysis
proceeds with a lower efficiency, affording metallic gold and
t-BuNCAuCl3 with Φ ) 0.0021 and 0.0045, respectively. The
results suggest a smaller size for the gold nanoparticles formed
from this solution based on the value of λmax (565 nm compared
to 602 nm for the 5.0 × 10-3 M solution) since gold
nanoparticles are known to exhibit size-dependent absorption
maxima.45 HPLC data for a solution of the photoproducts shows
only one peak with a retention time of 3.57 min. This peak is
assigned to the free isonitrile ligand because an independent
measurement for a solution of the free t-BuNC ligand in
dichloromethane shows one peak with a retention time of 3.58
min.
The efficiency of the photolysis of MeNCAuCl in dichlo-
romethane to give metallic gold is very low (the lowest among
all compounds studied here) with MeNCAuCl3 forming as the
dominant photoproduct. After several exposures of a 5.0 × 10-3
M solution, photooxidation takes place for the Au(I) complex
as suggested by the appearance of the band at λmax ) 322 nm
due to MeNCAuCl3, which remains the dominant feature even
at longer exposure periods (Figure 5) but the efficiency is rather
Figure 6. Absorption spectral changes during the photolysis of a 8.0 ×
10-3 M solution of Au(CO)Cl in dichloromethane under argon at room
temperature using a 1-cm cuvette after (a) 0, (b) 1, (c) 2.5, (d) 4, (e) 5.5,
(f) 7, and (g) 9 min of irradiation with a 450-W Hg lamp.
low, Φ ) 0.0013. Metallic gold does not form efficiently as
this photoproduct can be obtained only after extended irradiation
(8 h or longer).
Photochemistry of Au(CO)Cl vs Concentration. Photolysis
data for a 8.0 × 10-3 M dichloromethane solution of Au(CO)-
Cl are shown in Figure 6. After several exposure periods, a new
band appears at λmax ) 318 nm attributed to square-planar Au-
(III) photoproduct species, which form with Φ ) 0.041. This
is accompanied by a darker solution associated with a visible
band ascribed to colloidal gold (λmax ) 565 nm; Φ ) 0.065).
At a lower concentration (8.0 × 10-5 M), the dominant feature
is metallic gold (λmax ) 552 nm; Φ ) 0.013) but a weaker
band assigned to Au(III) photoproducts also appears (λmax
)
316 nm; Φ ) 0.0088). At a higher concentration of Au(CO)Cl,
the photoreaction is more efficient, as indicated by the higher
photoproduct quantum yield. It is worth mentioning that we have
been able to reproduce the photochemical results of Kunkely
and Vogler22 for a 1.0 × 10-3 M solution of the Au(CO)Cl
complex in dichloromethane since we obtain Φ ) 0.019 for
the Au(III) photoproduct of this solution. This value is
(45) Fendler, J. H. Nanoparticles and Nanostructured Films; Wiley-VCH:
Weinheim, 1998.
9
11388 J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007

Photochemistry of Neutral Isonitrile Au(I) Complexes
A R T I C L E S
intermediate between those for the two solutions above so it
serves to verify the trend of the concentration dependence of
the photolysis quantum yield in addition to showing that the
above values are not due to variation in experimental conditions
between two laboratories. The efficiency of the photochemical
process for Au(CO)Cl is higher than that for the RNCAuCl
analogues. Table 3 compares the photochemical quantum yields
for the RNCAuCl compounds with those for Au(CO)Cl at
different concentrations.
Au(I) and Ag(I) complexes in aqueous solution progressively
red-shifts their excitation band energies from those for dilute
solutions, ultimately leading to excitation energies that are even
lower than values for the corresponding solids.42 Balch and co-
workers reported a remarkable variation in the emission color
of the Au(I) diaminocarbene complex [Au{C(NHMe)2}2](PF6)
in frozen solutions of different solvents, several of which
exhibited significantly red-shifted emissions from those for
crystals.48 The association of (p-tosyl)CH2NCAuCl is driven
by aurophilic bonding in the solid state and most likely also in
solution. Crystals grown from dichloromethane solutions of this
compound form a crossed-dimer structure with a nearly stag-
gered conformation (in which aurophilic bonding plays the major
stabilizing factor vs other intermolecular interactions).16 The
high formation constants and association free energies and
enthalpies (Table 2) are consistent with staggered oligomers in
solution, as the alternative parallel or antiparallel association
modes are known to lead to lower stabilization energies (and
longer aurophilic distances).8,16,17,41,42
Discussion
Aurophilic Bonding in Solution. Deviations from Beer’s law
have been studied for other d10 and d8 closed-shell systems.
Schindler and co-workers reported a negative deviation for a
monomer MLCT band in aqueous solutions of [Pt(CN)4]2-
,
which was attributed to the formation of oligomers (trimers),26
while Gray and co-workers reported a similar study for Rh-
(I)-isocyanide complexes.27 As for d10 systems, Patterson and
co-workers reported the association of [Au(CN)2]- and
[Ag(CN)2]- complexes in water and methanol, in which negative
deviations from Beer’s law were attributed to the formation of
dimers and trimers,42 while Lin and co-workers reported another
study for the association of dinuclear diphosphine-dithiolate
Au(I) complexes in acetonitrile.46 The ∆G12 value of -18.29
kJ/mol for the [(p-tosyl)CH2NCAuCl]2 dimer at ambient tem-
perature is higher than the reported values for both the
mononuclear [Au(CN)2-]2 complexes (∆G12 ) -7.08 kJ/mol)42
and dinuclear Au(diphosphine)(dithiolate) complexes (∆G12
varied between -8.9 kJ/mol to -12 kJ/mol depending on the
bridging ligands).46 The higher value of the dimer formation
constant and the corresponding more negative ∆G12 value for
[(p-tosyl)CH2NCAuCl]2 than these literature precedents suggest
a thermodynamically stronger association in solutions of this
neutral isonitrile compound (and perhaps stronger Au‚‚‚Au
aurophilic bonds; vide infra). The negative ∆S values for the
dimer and the trimer of the (p-tosyl)CH2NCAuCl complex are
consistent with an association process and are comparable to
those reported earlier by Lin and co-workers.46 The extent of
aurophilic bonding and other intermolecular interactions that
contribute to the association of LAuCl complexes (L ) CO or
RNC) in solution affect the photochemical behavior and
photoproduct quantum yields.
Photolysis of RNCAuCl vs Au(CO)Cl. The electronic
structure of gold(I) complexes with σ-donor/π-acceptor ligands
(e.g., CO, CN-) is strongly influenced by π back-bonding.
Isocyanides are believed to be better σ-donors and poorer
π-acceptors than carbonyl.49,50 The nature of the R group can
affect the degree of π-acceptance capability in the RNC family.
An electron-withdrawing moiety (such as aryl-SO2 in (p-
tosyl)CH2NC) renders the isonitrile a better π-acceptor. On the
other hand, electron-donating groups make the isonitriles better
σ-donors and weaker π-acceptors, which is the situation in
MeNC and t-BuNC in comparison with (p-tosyl)CH2NC.
Johnston and Cooper have shown that the π-acceptance ability
of substituted aryl isonitriles increases in the order o- or p-CH3-
OC6H4NC < o- or p-CH3C6H4NC < C6H5NC < o- or p-FC6H4-
NC < o- or p-CF3C6H4NC < o- or p-ClC6H4NC < o- or
p-NO2C6H4NC, while the σ-donation ability follows the opposite
order.51 Chemical intuition suggests that the stronger the π
acceptance ability in organometallic LAuX complexes, the
higher the absorption energy, the greater the M-C bonding
character of the originating filled orbital depopulated, and the
greater the M-C antibonding character in the destination virtual
orbital populated. This suggested model is shown in Figure 7.
The spectral and photochemical data herein suggest that this
model is borne out. The relative absorption energies for dilute
solutions that represent monomer forms of the complexes in
this study before irradiation are consistent with this model. Thus,
dilute solutions of Au(CO)Cl in acetonitrile exhibit absorption
bands at λmax ) 207 and 220 nm, which are blue-shifted from
the major absorption band at λmax ) 227 nm for analogous
solutions of (p-tosyl)CH2NCAuCl, which in turn is blue-shifted
from the 230-240 nm bands exhibited by analogous solutions
of t-BuNCAuCl and MeNCAuCl. Consequently, the observed
experimental λmax values for the LAuX complexes examined
herein support our intuitive premise that stronger π-acceptor
It is interesting that the absorption spectral data suggest that
the association of (p-tosyl)CH2NCAuCl in dichloromethane is
not limited to dimer formation as trimers also form at high
enough concentrations. This is particularly intriguing because
crystals of (p-tosyl)CH2NCAuCl have the molecules associating
as crossed dimers in the solid state.17 This result adds to rare
literature precedents of greater association in solution than in
the solid state for other d10 complexes. For example, Eisenberg
and co-workers reported that solvent-free colorless crystals of
a dimeric gold(I) dithiocarbamate complex are not emissive and
do not exhibit Au‚‚‚Au intermolecular interactions; yet, dis-
solving these crystals leads to fluid or glassy solutions that
exhibit visible color and bright luminescence indicative of
association of the molecules in solution.47 Patterson and co-
workers reported that increasing the concentration of dicyano
(48) White-Morris, R. L.; Olmstead, M. M.; Jiang, F.; Tinti, D. S.; Balch, A. L.
J. Am. Chem. Soc. 2002, 124, 2327.
(49) Elschenbroich, C.; Salzer, A. Organometallics: A Concise Introduction,
2nd ed.; VCH: Weinheim, 1992.
(46) Tang, S. S.; Chang, C. P.; Lin, I. J. B.; Liou, L. S.; Wang, J. C. Inorg.
Chem. 1997, 36, 2294.
(47) Mansour, M. A.; Connick, W. B.; Lachicotte, R. J.; Gysling, H. J.;
Eisenberg, R. J. Am. Chem. Soc. 1998, 120, 1329.
(50) Mathieson, T.; Schier, A.; Schmidbaur, H. J. Chem. Soc., Dalton Trans.
2001, 1196.
(51) Johnston, R. F.; Cooper, J. C. J. Mol. Struct. 1991, 236, 297.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11389

A R T I C L E S
Elbjeirami and Omary
Figure 7. A model for the frontier orbitals showing the effective transition that leads to photolysis upon modulation of the π-acceptance ability of the L
ligand in LAuX complexes. The ∆ parameter represents the extent of π acceptance since it shows the deviation from the nonbonding character of the 5dπ
orbitals of Au(I) predicted by the σ-only bonding scheme of the crystal-field theory.
ligands will have a higher absorption energy due to the
increasing ∆ parameter shown in the model suggested in Figure
7.
uct quantum efficiency will be obtained at higher concentrations
versus dilute solutions. Balzani and co-workers have reported
that dilute (10-2-10-4 M) aqueous solutions of [Pt(CN)4]2-
show no photochemistry to 254-nm irradiation.52 However,
subsequent studies by Adamson and co-workers showed that
the triplet excited states of the oligomers present at high
concentrations of [Pt(CN)4]2- are photoreactive by a two-
electron oxidation pathway in the presence of a good two-
electron oxidant.26b Later on, Pfab and Gerhardt demonstrated
that PtII chains in solid tetracyanoplatinates(II) exhibit photo-
chemical activity as a result of removing an electron from the
one-dimensional valence band and placing it into a lattice site
between chains, which are octahedrally coordinated by water
molecules that act as acceptors.28 We are unaware of literature
precedents for concentration-dependent photolysis studies of d10
complexes, but there is precedence for concentration-dependent
studies of reagents added to such complexes to sensitize their
photoreactivity. For example, Kunkely and Vogler observed
The effective absorption that leads to photolysis entails
removing an electron from 5dπ orbitals, which are nonbonding
in the σ-bonding scheme according to simple crystal-field theory
concepts. However, when π bonding is considered, an increase
in the π-acceptance ability of the L ligand in LAuX leads to a
greater M-C bonding character in the 5dπ orbitals and greater
M-C antibonding character in the neutral ligand’s π* orbitals.
Hence, depopulation of the bonding 5dπ orbitals and population
of the ligand π* orbitals on exposure of the linear LAuCl
complex to a broad-band UV light lead to dissociation of L
ligands with an efficiency that should increase on going from
L ) (alkyl)NC f (p-tosyl)CH2NC f CO since the π-ac-
ceptance ability increases in the same direction. This is exactly
the trend of the quantum yield values we observe for these
complexes, as summarized in Table 3. The model suggested in
Figure 7 is different from the one proposed by Kunkely and
Vogler for the Au(CO)Cl complex, which suggested that the
photolysis is caused by the weak “ds” absorption transition that
populates the Au(I) 6s orbital, arguing that the latter orbital is
destabilized by σ-interaction with the ligands in a similar manner
to that for the destabilization of the eg* ligand-field orbitals in
octahedral metal carbonyls.22 These authors suggested that, albeit
weak, the “ds” transition has a larger effect on the Au-CO
bonding than the much stronger higher-energy major absorption
bands that were assigned as “dp” or “MLCT” in ref 22. Adopting
the model in ref 22 would lead to an opposite trend of the
relative quantum yields in this study since the RNC ligands are
stronger σ-donors and weaker π-acceptors than CO.
-
-
higher photolysis quantum yields for AuCl2 and AuBr2 in
the presence of higher concentrations of electron acceptors such
as CH2Cl2 or O2 but not the Au(I) complex itself.53
Despite the fact that all LAuX complexes herein (L ) RNC
or CO) exhibit higher photolysis quantum yields at higher
concentrations, the electronic factor (π acceptance ability of L)
distinguishes them better from one another in regards to
photoreactivity. The difference among the compounds in the
supramolecular association mode in their crystalline solid state
does not seem to have any bearing on the trends of the quantum
yields, although this is the major factor that governs their solid-
state luminescence behavior.16,32 The solid-state association
mode for the most photoreactive Au(CO)Cl and least photore-
active (alkyl)NCAuCl compounds is the same (antiparallel
chains), while the photoreactivity was intermediate for the (p-
tosyl)CH2NCAuCl compound whose crystals exhibit a crossed-
dimer structure.
We now turn our attention to explaining the increased
photoreactivity upon increasing the concentration of the neutral
complexes. As dimers and higher oligomers form in more
concentrated solutions, aurophilic bonding red-shifts the absorp-
tion energies of the effective photolysis transitions and thus
lower-energy photons will also contribute to the photolysis
reactions along with the higher-energy photons that excite both
the monomers and oligomers. Consequently, higher photoprod-
(52) Moggi, L.; Bolletta, F.; Balzani, V.; Scandola, F. J. Inorg. Nucl. Chem.
1966, 28, 2589.
(53) Kunkely, H.; Vogler, A. Inorg. Chem. 1992, 31, 4539.
9
11390 J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007

Photochemistry of Neutral Isonitrile Au(I) Complexes
A R T I C L E S
We propose that the photolysis of LAuICl (L ) CO and RNC)
in solution will lead to the following photochemical reaction
scheme upon absorbing UV irradiation, with the efficiency of
each step and the “n” value depending on the concentration and
extent of oligomerization:
This mechanism provides an explanation for the sensitivity
of the photochemical reactivity to concentration (hence auro-
philicity) of the LAuX complex because the proposed reactive
species is an oligomer. Nevertheless, additional processes likely
compete with or complement the proposed mechanism. For
example, direct oxidation or reduction without necessarily going
through the intermediate AuCl may be possible. Solvent-initiated
photoreactions are particularly relevant in halogenated solvents,
and it is often difficult to distinguish such processes from metal-
centered photoreactions.54 A solvent-initiated photoreaction may
lead to oxidation of RNCAuICl to form RNCAuIIICl3 and
H[AuCl4] in a manner akin to that suggested by Serafimova
and Hoggard for Au(PPh3)Cl.54b Although this mechanism does
not account for the formation of Au nanoparticles, it cannot be
completely ruled out because we observe some variation in the
distribution of the photoproducts upon changing the neutral
ligand in the different LAuCl complexes studied here. For
example, photolysis of MeNCAuCl does not attain Au nano-
particles while other complexes do to different extents (see
Figures 3-6 and Table 3).
Among the multiple reasonable possibilities for the exact
molecular form of the Au(III) complex in the photoproducts,
we adopt an assignment to square-planar LAuIIICl3 complexes
although it is likely that this species is in equilibrium with the
free ligand and other Au(III) forms. The presence of the free
ligand in the photoproduct mixture is substantiated by HPLC
data (vide supra). The Au(III) species are likely aggregated and/
or solvated to form molecular units such as Au2Cl6, LAuCl3,
(Sol)AuCl3, and H[AuCl4] (L ) neutral ligand, Sol ) solvent
molecule). The absorption spectral changes in this work and in
similar photochemical studies reported for Au(I) complexes22,53-55
are insufficient to identify the exact identity of these Au(III)
species because they have rather similar absorption pro-
files.43,44,54 Because of some variation in peak maxima with
different ligands, we prefer the assignment to LAuCl3 although
some of the other aforementioned forms as well as the free
ligand are likely present. The Au2Cl6 form may exist as a result
of disproportionation of AuCl species before diffusion of and
collision with the photodissociated ligand. The H[AuCl4] form
is possible in the solvent-initiated competing mechanism, as
Figure 8. (a) Absorption spectra showing the formation of gold nano-
patricles after irradiating the following solutions for 1-2 h: (i) a mixture
of 4 mL of 8.0 × 10-5 M (p-tosyl)CH2NCAuCl in dichloromethane and 1
mL of 1 mM DAB-Am-16 in methanol (6.4 × 10-5 M final concentration
of (p-tosyl)CH2NCAuCl), (ii) a mixture of 4 mL of 8.0 × 10-3 M (p-
tosyl)CH2NCAuCl in dichloromethane and 1 mL of 1 mDAB-Am-16 in
methanol (6.4 × 10-3 M final concentration of (p-tosyl)CH2NCAuCl), and
(iii) 8.0 × 10-3 M solution of (p-tosyl)CH2NCAuCl in dichloromethane
with no dendrimer added. (b) Normalized plasmon absorption peaks (solid
lines) with Gaussian peak-fit analysis (dotted lines) for samples i-iii. Vial
iii′ contains dark-grayish metallic gold that precipitates when vial iii is left
standing for several hours.
discussed above, while the (Sol)AuCl3 form is particularly
relevant with ligating solvents such as acetonitrile (but even
dichloromethane is not completely innocent).
It is interesting to note that the relative long-term stability of
the solids and solutions of the compounds examined here is
commensurate with their photochemical reactivity. For example,
solutions of Au(CO)Cl readily form metallic gold even under
room light within an hour. On the other hand, solutions of (p-
tosyl)CH2NCAuCl are not extremely light sensitive and remain
clear and colorless for several days before they decompose,
while the solid compound can be stored in room light only for
several weeks before it starts decomposing. In contrast, t-
BuNCAuCl and MeNCAuCl are extremely stable to light, as
we have not observed any decomposition for solids and solutions
of both compounds upon extended storage under room light.
These observations clearly correlate with the quantum yield (Φ)
values shown in Table 3 for the formation of the metallic gold
and LAuCl3 photoproducts. Indeed, the trend of the relative
(54) (a) Hoggard, P. E. Coord. Chem. ReV. 1997, 159, 235. (b) Serafimova, I.
M.; Hoggard, P. E. Inorg. Chim. Acta. 2002, 338, 105.
(55) For representative examples of other Au(I) complexes studied photochemi-
cally, see: (a) Foley, J. B.; Bruce, A. E.; Bruce, M. R. M. J. Am. Chem.
Soc. 1995, 117, 9595. (b) Che, C.-M.; Kwong, H.-L.; Poon, C.-K.; Yam,
V. W.-W. J. Chem. Soc., Dalton Trans. 1990, 3215. (c) Yam, V. W.-W.;
Choi, S. W.-K.; Lo, K. K.-W.; Dung, W.-F.; Kong, R. Y.-C. J. Chem.
Soc., Chem. Commun. 1994, 2379. See ref 23b for a comprehensive review.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11391

A R T I C L E S
Elbjeirami and Omary
Figure 9. TEM micrographs of dendrimer-stabilized Au nanoparticles using a copper grid covered by carbon film. The particles were prepared by photolysis
of 6.4 × 10-3 M (a and b, red solution) and 6.4 × 10-5 M (c and d, brown solution) solutions of (p-tosyl)CH2NCAuCl as described in the caption of Figure
8. Further TEM images are available in the Supporting Information.
long-term light stability based on the π-acceptor ability of L
extrapolates to several other LAuX compounds that we have
synthesized but have yet to subject to photolysis studies. Thus,
we are encouraged to expand the photochemical study herein
to investigate the photolysis of multiple other classes of Au(I)
complexes which, despite the strong attention given to their
photophysical properties, have not been subjected to photo-
chemical studies.55
Can the Fundamental Findings Above Aid the Stabiliza-
tion of Gold Nanoparticles? Since metallic gold is released
as a result of the photochemical reactions, it occurred to us to
conduct the photolysis reactions in solutions that contain
dendrimers terminated with primary amines, which are known
to stabilize gold nanopatricles commonly generated via wet
chemical reduction of Au(III) precursors.56 Thus, photolysis of
(p-tosyl)CH2NCAuCl solutions that contain the DAB-Am-16
dendrimer leads to solutions that exhibit multiple visible colors
depending on the initial concentration of the photoreactant.
Figure 8 shows that such colors correspond to broad absorption
peaks with λmax ) 516 and 536 nm for two representative
photolized solutions with initial (p-tosyl)CH2NCAuCl concen-
trations of 6.4 × 10-5 and 6.4 × 10-3 M, respectively; see traces
i and ii in Figure 8. This result is a clear indication of the
formation of gold nanopatricles with different sizes.35,36,57 The
isolated plasmon absorptions of these two photolized solutions
give rise to brown and red colors, respectively, representing
evenly dispersed suspensions that look clear to the naked eye
due to the small particle size. These are contrasted with the
purple suspension that results from the photolysis of a 8.0 ×
10-3 M solution of (p-tosyl)CH2NCAuCl with no dendrimer
stabilizer added. The latter gives rise to a broader and more
red-shifted absorption peak (λmax ) 590 nm, trace iii in Figure
8), which is characteristic of surface plasmon absorption of large
particles and a large size distribution.35,36,57 In the absence of
dendrimer, the gold nanoparticles agglomerate even further, as
evidenced by a clear physical change from a purple evenly
dispersed suspension to dark grayish colloids that precipitate
after several hours (Figure 8b). In contrast, gold nanoparticle
suspensions that contain a stabilizing dendrimer are known to
maintain their colors and dispersion stability.56 This is indeed
the case for the gold nanoparticles we have prepared by
photolysis of (p-tosyl)CH2NCAuCl solutions that contain the
DAB-Am-16 stabilizing dendrimer, for which we have not seen
any evidence of agglomeration during the approximately
9-month period since their original preparation and the publica-
tion of this manuscript.
Figure 9 shows representative transmission electron micros-
copy (TEM) images for the gold nanoparticles corresponding
to the aforementioned brown and red solutions while more TEM
images are provided in the Supporting Information (Figures S3
and S4). These TEM images show well-organized spherical Au
nanoparticles with an average particle diameter of 5.3 nm for
the brown solution and 6.5 nm for the red solution, as calculated
by averaging the particle diameters in selected regions of the
TEM images that represent the most dominant particles in each
solution. The smaller-sized gold nanoparticles are more mono-
dispersed and have a somewhat more uniform spherical shape
in comparison to the larger particles. The particle size increase
(56) For some representative examples, see: (a) Balogh, L.; Valluzzi, R.;
Laverdure, K. S.; Gido, S. P.; Hagnauer, G. L.; Tomalia, D. A. J. Nanopart.
Res. 1999, 1, 353. (b) Crooks, R. M.; Zhao, M.; Sun, L.; Chechik, V.;
Yeung, L. K. Acc. Chem. Res. 2001, 34, 181. (c) Balogh, L.; Tomalia, D.
A. J. Am. Chem. Soc. 1998, 120, 7355. (d) Garcia, M. E.; Baker, L. A.;
Crooks, R. M. Anal. Chem. 1999, 71, 256. (e) Zhao, M.; Crooks, R. M.
AdV. Mater. 1999, 11, 217.
(57) Link, S.; El-Sayed, M. A. Int. ReV. Phys. Chem. 2000, 19, 409.
9
11392 J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007

Photochemistry of Neutral Isonitrile Au(I) Complexes
A R T I C L E S
obtained by TEM for the dominant Au nanoparticles in the red
versus brown solution is consistent with the observed red shift
of the electronic absorption maximum and color change. There
are other particles with different sizes seen in the TEM data
besides the sizes quoted above (Figures S3 and S4), but the
combination of the TEM and spectral data suggest that, in
general, the particles are much more uniform in these dendrimer-
stabilized samples than the particles formed in the absence of
dendrimers. Further data in the Supporting Information (Figures
S5-S7) illustrate multi-Gaussian peak fitting analysis for the
broad plasmon absorptions resulting from the photolysis of
various LAuICl complexes, leading to several components
sometimes with a very high-energy separation between the peak
maxima. The multiple Gaussian peaks for each sample are
characteristic of different Aum particle sizes, which is consistent
with our proposed model. The generation of more than one
dominant set of Aum particles and the possible relevance of other
mechanisms, as discussed above, are factors that contribute to
AuCl and affect the size variation of the gold nanoparticles.
This constitutes autocatalysis by the Au nanoparticles formed
from the growth of the nucleation centers.58 However, further
studies are warranted to investigate this effect systematically
as has been done in ref 58. Efforts are also ongoing for assessing
the suitability of the gold nanoparticles prepared photochemi-
cally as described herein for catalytic and biomedical applica-
tions.
Acknowledgment. Dedicated to Prof. Hubert Schmidbaur.
We thank the National Science Foundation (CAREER Award,
Grant CHE-0349313), the Robert A. Welch Foundation (Grant
B-1542), and the U.S. Department of Energy (Award DE-FC26-
06NT42859) for partial funding of this research. We thank Dr.
Jose´ Caldero´n for his extensive assistance with the acquisition
of the HPLC data. We also thank Mr. Taehon Lee and Prof.
Moon Kim for the acquisition of TEM micrographs at the
University of Texas-Dallas.
0
making the relative quantum yield of LAuCl3/Aum different
from the 1:2 stoichiometric ratio of the disproportionation step
in the proposed mechanism.
Supporting Information Available: Further spectral, chro-
matography, and TEM data referred to in the text above (PDF
format). This material is available free of charge via the Internet
at http://pubs.acs.org.
Finally, we wish to comment on the observation that
photochemical changes leading to gold nanoparticle formation
are slow initially and increase with further irradiation. The
photoinduced production of AuCl leads to incremental formation
of gold atoms, which agglomerate to form small metal clusters.
The clusters formed can potentially act as nucleation centers
that might also catalyze the disproportionation of the remaining
JA0703933
(58) Mallick, K.; Wang, Z. L.; Pal, T. J. Photochem. Photobiol. A 2001, 140,
75.
9
J. AM. CHEM. SOC. VOL. 129, NO. 37, 2007 11393