Photophysics and bonding in neutral gold(I) organometallic complexes with an extended aurophilic supramolecular structure

Article abstract of DOI:10.1021/om060524t

This work represents a synergistic experimental/computational study of the molecular spectroscopy and bonding in Au(CO)Cl in solution and the solid state. The luminescence behavior for crystalline solids is similar for Au(CO)Cl and related (RNC)AuCl complexes that likewise stack in infinite chains, and both exhibit orange-red unstructured phosphorescence bands with extremely large Stokes shifts ((15-20) × 10³ cm^-1). The long aurophilic distances computed for the ground state (～3.2 A) are contracted in the phosphorescent excited state (～2.6 A), demonstrating excimeric Au-Au covalent bonds. The spectral data suggest phosphorescent species in which the excimeric Au-Au bonding is extended beyond two adjacent molecules in the solid state. Controlling the concentration in frozen solutions attains phosphorescent bands due to dimeric species for which the emission energies are higher (in the blue region) than those for crystalline solids and are reproduced by ab initio calculations. The spectral findings herein suggest that predictive information about the supramolecular structure may be obtained by the luminescnce behavior. This is exemplified by crystals of (1,1,3,3-Me ₄BuNC)AuCl, whose red-orange luminsecence anticipated an extended-chain supramolecular structure, which was later verified crystallographically, as the molecules were found to pack in zigzag chains with alternating short (3.418 A) and long (4.433 A) aurophilic distances. Solutions of Au(CO)Cl exhibit negative deviation from Beer's law for, higher-energy monomer bands with the appearance of lower-energy bands at high concentrations due to the dimerization of molecules. Time-dependent density-functional theory (TD-DFT) calculations for monomer and dimer models show good agreement with the experimental spectra and account for the major absorption bands. Ab initio calculations (CCSD(T)/cc-pVTZ) show a blue shift of ～23 cm^-1 in the v_c=o frequency upon complexation, thus providing the first computational evidence of this anomalous blue shift known experimentally for Au(CO)Cl.

Full text of DOI:10.1021/om060524t

2550
Organometallics 2007, 26, 2550-2560
Photophysics and Bonding in Neutral Gold(I) Organometallic
Complexes with an Extended Aurophilic Supramolecular Structure
Oussama Elbjeirami,† Scott Yockel,† Charles F. Campana,‡ Angela K. Wilson,*,† and
Mohammad A. Omary*,†
Department of Chemistry and Center for AdVanced Scientific Computing and Modeling (CASCaM),
UniVersity of North Texas, P.O. Box 305070, Denton, Texas 76203-5070, and Bruker AXS Inc.,
5465 East Cheryl Parkway, Madison, Wisconsin 53711-5373
ReceiVed June 15, 2006
This work represents a synergistic experimental/computational study of the molecular spectroscopy
and bonding in Au(CO)Cl in solution and the solid state. The luminescence behavior for crystalline
solids is similar for Au(CO)Cl and related (RNC)AuCl complexes that likewise stack in infinite chains,
and both exhibit orange-red unstructured phosphorescence bands with extremely large Stokes shifts ((15-
20) × 103 cm-1). The long aurophilic distances computed for the ground state (∼3.2 Å) are contracted
in the phosphorescent excited state (∼2.6 Å), demonstrating excimeric Au-Au covalent bonds. The
spectral data suggest phosphorescent species in which the excimeric Au-Au bonding is extended beyond
two adjacent molecules in the solid state. Controlling the concentration in frozen solutions attains
phosphorescent bands due to dimeric species for which the emission energies are higher (in the blue
region) than those for crystalline solids and are reproduced by ab initio calculations. The spectral findings
herein suggest that predictive information about the supramolecular structure may be obtained by the
luminescnce behavior. This is exemplified by crystals of (1,1,3,3-Me4BuNC)AuCl, whose red-orange
luminsecence anticipated an extended-chain supramolecular structure, which was later verified crystal-
lographically, as the molecules were found to pack in zigzag chains with alternating short (3.418 Å) and
long (4.433 Å) aurophilic distances. Solutions of Au(CO)Cl exhibit negative deviation from Beer’s law
for, higher-energy monomer bands with the appearance of lower-energy bands at high concentrations
due to the dimerization of molecules. Time-dependent density-functional theory (TD-DFT) calculations
for monomer and dimer models show good agreement with the experimental spectra and account for the
major absorption bands. Ab initio calculations (CCSD(T)/cc-pVTZ) show a blue shift of ∼23 cm-1 in
the νCtO frequency upon complexation, thus providing the first computational evidence of this anomalous
blue shift known experimentally for Au(CO)Cl.
solvent dependence), which has been attributed to only modest
π back-donation from the metal to the carbonyl group.6
Calderazzo and co-workers reported a seminal study on the
properties of Au(CO)X compounds, in which they surveyed a
number of ligands.7 In addition to gold, other d10 metal-
carbonyl complexes have been studied to a lesser extent. For
example, copper carbonyl species of the form CuI(CO)X (X )
Cl, Br, CF3COO) have been isolated or at least identified, while
structurally characterized carbonyl complexes of silver(I) are
rare and harder to isolate.8-15 It is clear from experimental
evidence that gold carbonyl species are the most stable
thermodynamically, in comparison to the analogous copper and
silver complexes, with an Au-CO bond energy of ∼200 kJ/
mol as estimated by Calderazzo.7
Introduction
Chlorocarbonylgold(I) was first prepared in 19251 and has
been used as a common starting material for AuI complexes,
since the carbonyl species can be readily displaced by other
ligands. This compound is extremely air and moisture sensitive
and forms a colorless crystalline solid.1 The crystal structure
of Au(CO)Cl reveals that the molecules are arranged in head-
to-tail antiparallel infinite chains with Au‚‚‚Au contacts of 3.38
Å,2 similar to their isoelectronic RNCAuICl neutral isonitrile
compounds, that have identical association of antiparallel
chains.3 The stability of the halogen carbonyl complexes of Au-
(I) follows the same trend established earlier for palladium(II)
and platinum(II), with the stability for all these complexes
decreasing along the sequence Cl > Br > I.4,5 One interesting
IR characteristic in the Au(CO)Cl complex is the high νCtO
stretching vibrational frequency (2152-2162 cm-1 with some
(6) DellAmico, D. B.; Calderazzo, F. Gazz. Chim. Ital. 1973, 103, 1099.
(7) Calderazzo, F. J. Organomet. Chem. 1990, 400, 303.
(8) Meyer. R. J.; Pietsch, E. H.; Kotowski, A. Gmelin Handbuch der
Anorganischchen Chemie; Springer: Berlin, 1975; Silver Part B5.
(9) Antes, I.; Dapprich, S.; Frenking, G.; Schwerdtfeger, P. Inorg. Chem.
1996, 35, 2089.
* To whom correspondence should be addressed. E-mail: omary@unt.edu
(M.A.O.); akwilson@unt.edu (A.K.W.).
† University of North Texas.
‡ Bruker AXS Inc.
(1) Manchot, W.; Gall, H. Ber. Bunsen-Ges. Phys. Chem. 1925, 58, 2175.
(2) Jones, P. G. Z. Naturforsch., B 1982, 37, 823.
(3) White-Morris, R. L.; Olmstead, M. M.; Balch, A. L.; Elbjeirami, O.;
Omary, M. A. Inorg. Chem. 2003, 42, 6741.
(4) Andreini, B. P.; DellAmico, D. B.; Calderazzo, F.; Venturi, M. G.;
Peizzi, G.; Serge, A. J. Organomet. Chem. 1988, 354, 357.
(5) DellAmico, D. B.; Calderazzo, F.; Robino, P.; Serge, A. J. Chem.
Soc., Dalton Trans. 1991, 3, 3017.
(10) Caulton, K. G.; Davies, G.; Holt, E. M. Polyhedron 1990, 9, 2319.
(11) Bruce, M. I.; Ostazewski, A. P. P. J. Chem. Soc., Dalton Trans.
1973, 2433.
(12) Dias, H. V. R.; Lu, H. L. Inorg. Chem. 1995, 34, 5380.
(13) Dias, H. V. R.; Jin, W. J. Am. Chem. Soc. 1995, 117, 11381.
(14) Hurlburt, P. K.; Rack, J. J.; Dec, S. F.; Anderson, O. P.; Strauss, S.
H. Inorg. Chem. 1993, 32, 373.
(15) Dias, H. V. R.; Wiechang, J. Inorg. Chem. 1996, 35, 3687.
10.1021/om060524t CCC: $37.00 © 2007 American Chemical Society
Publication on Web 03/31/2007

Gold(I) Organometallic Chain Complexes
Organometallics, Vol. 26, No. 10, 2007 2551
The organometallic literature suggests that metal isonitriles
have been strongly overshadowed by metal carbonyls, most
likely because of the repulsive smell of isonitriles and not
essentially due to scientific reasons. The electronic structure of
RNC has been shown to be rather similar to that of CO;16 thus,
similar reactivity is expected for analogous complexes of both
ligands. The chemistry of metal isonitriles is relevant to several
interesting applications. For example, gold(I) isonitrile com-
plexes have been utilized for the deposition of gold films17,18
and for the preparation of new types of liquid crystalline
phases.19,20 The versatility enabled by varying the R group in
RNC ligands makes them good candidates to contribute to the
current state of knowledge regarding the photochemistry and
photophysics of CuI, AgI, and AuI species.21,22
specific luminescence properties, which can be used in single
or multicolor electroluminescent display and solid-state light
devices in which triplet emitters are significant.27 We have
recently communicated that the distinction for the Au-centered
solid-state phosphorescence energy in such complexes is perhaps
better drawn on the basis of the differences in the supramolecular
structure instead of the shortest Au‚‚‚Au distances, a hypothesis
made after contrasting the photophysical properties of one
compound in each category.28 In an effort to assess the generality
of this hypothesis, the present study examines the photophysical
behavior of five LAuCl organometallic complexes (L ) CO,
RNC) that pack in infinite zigzag chains through aurophilic
bonding.
Experimental and Computational Methodologies
The present study investigates the spectral and bonding
properties for solids of both Au(CO)Cl and (RNC)AuCl species,
which have similar solid-state packing. Detailed studies of the
excited-state properties of carbonyl complexes of d10 metal ions
are scarce.23 When the excited-state properties of metal com-
plexes are studied, it is found that an important distinction exists
between d1-9 and d10 configurations, in that the electronic
transitions with partially filled d subshells create ligand-field
excited states, while these d-d transitions are absent in the
closed-shell complexes. On the basis of this concept, the
spectroscopy and photochemistry of such systems are reliant
on the properties of other types of electronic excited states (e.g.
ligand-to-metal charge transfer (LMCT) or metal-to-ligand
charge transfer (MLCT)).23 In 1997, the Vogler group reported
a primarily photochemical study for Au(CO)Cl, with some
preliminary photophysical results showing the compound to
exhibit a longest-wavelength weak absorption feature at ∼250
nm.24 They attributed this solution absorption to a metal-centered
d-s transition, which the authors related to the solid-state
emission band at λmax 660 nm at room temperature.24 Inspired
by this precedent, herein we report a detailed spectroscopic and
bonding characterization of Au(CO)Cl in the solid state and
solution utilizing complementary experimental and computa-
tional methods to investigate this classical organometallic
complex. In addition, we compare the properties of Au(CO)Cl
to those of the closely related (RNC)AuCl (R ) tert-butyl,
methyl, cyclohexyl, and 1,1,3,3-tetramethylbutyl). We believe
that this study contributes to a better understanding of the
structure-luminescence relationships in mononuclear two-
coordinate AuI complexes. Despite the strong attention given
to such complexes, a major challenge lies in trying to relate
the luminescence energy or color to the crystallographic Au‚‚
‚Au ground-state distances.25,26 Once this relationship is well-
understood, it will be possible to design compounds to display
Materials. Au(CO)Cl was purchased from Strem Chemicals,
while cyclohexyl isocyanide (CyNC) and tert-butyl isocyanide (t-
BuNC) were purchased from ACROS Organics. (THT)AuCl (THT
) tetrahydrothiophene) and methyl isocyanide (MeNC) were
synthesized by slight modifications of published procedures.29,30
HPLC-grade acetonitrile and dichloromethane were distilled with
conventional drying agents, degassed by the freeze-pump-thaw
method three times prior to use, and kept under argon at all time.
All absorption measurements were carried out under argon, and
luminescence spectra were collected while the sample was kept
under helium and in the dark unless otherwise indicated.
Syntheses. The preparation methods for (RNC)AuCl (R )
cyclohexyl, methyl, tert-butyl) were reported previously3,31,32 and
were generally followed here for these compounds. Standard
Schlenk line techniques under an argon atmosphere were followed
in which equimolar amounts of (THT)AuCl and cyclohexyl
isocyanide, methyl isocyanide, or tert-butyl isocyanide were dis-
solved in 10 mL of deaerated dichloromethane followed by stirring
the solutions for 4 h. The solvent was then evaporated under reduced
pressure, and a white solid formed, which was extracted with
pentane, filtered, and dried under vacuum. Crystals were grown
for each compound from CH2Cl2/pentane solutions of recrystallized
solids.
(Me4BuNC)AuCl (Me4Bu ) 1,1,3,3-Tetramethylbutyl). (TH-
T)AuCl (200 mg, 0.62 mmol) was dissolved in 10 mL of deaerated
dichloromethane. 1,1,3,3-Tetramethylbutyl isocyanide (122 µL, 0.70
mmol) was then added, and the solution was stirred for 4 h, after
which the solvent was evaporated under reduced pressure and a
white solid formed. The solid was washed several times with
pentane, collected by filtration, and vacuum-dried (90% yield).
Crystals were grown from a CH2Cl2/pentane solution of the
recrystallized solid.
Physical Measurements. The luminescence measurements were
carried out for crystalline materials. Steady-state luminescence
spectra were acquired with a PTI QuantaMaster Model QM-4
scanning spectrofluorometer equipped with a 75 W xenon lamp,
emission and excitation monochromators, an excitation correction
unit, and a PMT detector. The emission and excitation spectra were
corrected for the wavelength-dependent detector response and xenon
lamp intensity, respectively. Temperature-dependent studies were
acquired with an Oxford optical cryostat using liquid helium or
liquid nitrogen as coolant. Lifetime data were acquired using a
(16) Elschenbroich, C.; Salzer, A. Organometallics: A Concise Introduc-
tion, 2nd ed.; VCH: Weinheim, Germany, 1992.
(17) Norton, P. R.; Young, P. A.; Cheng, Q.; Dryden, N.; Puddephatt,
R. J. Surf. Sci. 1994, 307, 172.
(18) Vaughan, L. G.; U.S. Patent 3661959-19720509, 1972.
(19) (a) Bachman, R. E.; Fioritto, M. S.; Fetics, S. K.; Cocker, T. M. J.
Am. Chem. Soc. 2001, 123, 537. (b) Kaharu, T.; Ishii, R.; Takahashi, S. J.
Chem. Soc., Chem. Commun. 1994, 1349.
(20) Ishi, R.; Kaharu, T.; Pirio, N.; Zhang, S. W.; Takahashi, S. J. Chem.
Soc., Chem. Commun. 1995, 1215.
(21) (a) Kual, C. Coord. Chem. ReV. 1990, 99, 213. (b) Ford, P. C.;
Vogler, A. Acc. Chem. Res. 1993, 26, 220. (c) Hong, T. S.; Li, P.; Leong,
W. K.; Fan, W. Y. J. Organomet. Chem. 2005, 690, 4132.
(22) Timney, J. A. Organomet. Chem. 1999, 27, 132.
(23) Shao, L.; Zhang, L.; Zhou, M.; Qin, Q. Organometallics 2001, 20,
1173.
(24) Kunkely, H.; Vogler, A. J. Organomet. Chem. 1997, 541, 177.
(25) Forward, J. M.; Fackler, J. P., Jr.; Assefa, Z. Photophysical and
Photochemical Properties of Gold(I) Complexes. In Optoelectronic Proper-
ties of Inorganic Compounds; Roundhill, D. M., Fackler, J. P., Jr., Eds.;
Plenum: New York, 1999; Chapter 6.
(26) Yam, V. W.; Lee, W. K. J. Chem. Soc., Dalton Trans. 1993, 2097.
(27) Review: Yersin, H. Top. Curr. Chem. 2004, 241, 1.
(28) Elbjeirami, O.; Omary, M. A.; Stender, M.; Balch, A. L. Dalton
Trans. 2004, 3173.
(29) Uso´n, R.; Laguna, A. In Organometallic Syntheses; King, R. B.,
Eisch, J. J., Eds.; Elsevier: Amsterdam, 1986; Chapter 3.
(30) Schuster, R. E.; Scott, J. E.; Cazanova, J. Org. Synth. 1966, 46, 75.
(31) Eggleston, D. S.; Chodosh, D. F.; Webb, R. L.; Davis, L. L. Acta
Crystallogr., Sect. C 1986, 42, 36.
(32) Schneider, W.; Angermaier, K.; Sladek, A.; Schmidbaur, H. Z.
Naturforsch., B 1996, 51, 790.

2552 Organometallics, Vol. 26, No. 10, 2007
Elbjeirami et al.
nitrogen laser interfaced with a tunable dye laser and a frequency
doubler, as part of fluorescence and phosphorescence subsystem
add-ons to the PTI instrument. The 337.1 nm line of the N2 laser
was used to pump a freshly prepared 1 × 10-2 M solution of the
organic continuum laser dye Coumarin-540A in ethanol, the output
of which was tuned and frequency-doubled to attain the suitable
excitation wavelength (on the basis of the steady-state luminescence
excitation spectra) used to generate the time-resolved data. Absorp-
tion spectra were acquired with a Perkin-Elmer Lambda 900 double-
beam UV/vis/near-IR spectrophotometer. Luminescence and life-
time studies for frozen solutions were conducted for selected
samples by placing a 5 mm Suprasil quartz cylindrical tube
containing the appropriate degassed solution in a liquid-nitrogen
filled Dewar flask with a Suprasil quartz cold finger and then
inserting this setup in the sample compartment of the PTI
instrument. Electronic absorption spectra were collected using 1
cm quartz cuvettes for solutions of crystalline samples prepared in
HPLC-grade acetonitrile that was degassed and kept under argon.
Due to the sensitivity of the Au(CO)Cl compound, all manipulations
and experiments for this compound were carried out in the dark
and under an inert atmosphere.
Crystallographic Details. A clear, colorless, platelike specimen
of [(CH3)3CCH2C(CH3)2NC]AuCl, C9H17NClAu, approximate di-
mensions 0.10 mm × 0.36 mm × 0.44 mm, was used for the X-ray
crystallographic analysis. The X-ray intensity data were measured
at 100(2) K on a Bruker KAPPA APEX II CCD area detector
system equipped with a graphite monochromator and a Mo KR
fine-focus sealed tube (λ ) 0.710 73 Å) operated at 1.5 kW power
(50 kV, 30 mA) and a Bruker Kryo-Flex low-temperature device.
All data collection and structure determination operations were
carried out with the APEX2 (version 2.0-2) Software Suite.33 The
detector was placed at a distance of 4.980 cm from the crystal. A
total of 1435 frames were collected with a scan width of 0.5° in ω
or φ and an exposure time of 30.0 s/frame. The total data collection
time was 12.76 h. The frames were integrated with the Bruker
SAINT33 (version 3.23A) software package using a narrow-frame
integration algorithm. The integration of the data using a monoclinic
unit cell yielded a total of 20 025 reflections to a maximum θ angle
of 30.60° (0.70 Å resolution), of which 3620 were independent
(average redundancy 5.53, completeness 99.2%, Rint ) 3.14%, Rsig
) 2.07%) and 3177 (87.76%) were greater than 2σ(I). The final
cell constants of a ) 19.9362(9) Å, b ) 8.0691(4) Å, c ) 14.8652-
(7) Å, â ) 98.263(2)°, and V ) 2366.50(19) Å3 are based upon
the refinement of the XYZ centroids of 6915 reflections above 20σ-
(I) with 5.292° < 2θ < 60.993°. Analysis of the data showed
negligible decay during data collection. Data were corrected for
absorption effects using the multiscan technique (SADABS,33
version 2.05). The ratio of minimum to maximum apparent
transmission was 0.6655. The calculated minimum and maximum
transmission coefficients (based on crystal size) were 0.0721 and
0.3652. The structure was solved and refined using the Bruker
SHELXTL33 (version 6.14) software package, using the space group
C2/c (No. 15), with Z ) 8 for the formula unit, C9H17NClAu. The
final anisotropic full-matrix least-squares refinement on F2 with
130 variables converged at R1 ) 1.62% for the observed data and
wR2 ) 4.33% for all data. The goodness of fit was 1.053. The
largest peak on the final difference electron density synthesis was
1.928 e/Å3, and the largest hole was -1.072 e/Å3 with an RMS
deviation of 0.114 e/Å3. On the basis of the final model, the
calculated density was 2.086 g/cm3 and F(000) was 1392 e.
Computational Details. Møller-Plesset second-order perturba-
tion theory (MP2)34 was used as the primary method to calculate
optimized structures, frequencies, and energetics of the ground and
triplet excited states of [Au(CO)Cl]n (where n ) 1-3). To better
characterize the frequencies of the C-O stretch upon complexation,
coupled-cluster theory with single, double, and quasiperturbative
triple excitations (CCSD(T))35,36 was also used. Throughout this
study, the correlation consistent basis sets (cc-pVxZ, where x )
D(2), T(3), Q(4), 5)37 were used for C and O, while for Cl, the
tight-d augmented correlation consistent basis sets (cc-pV(x+d)Z)38
were used. For Au, the newly developed correlation consistent basis
sets (cc-pVxZ-PP) were used,39 which include a 60-electron Stuttgart
relativistic pseudopotential.40 The calculated absorption energies
of the monomer and dimer were determined as the vertical
transitions between the minimum singlet ground state S0 and the
T1 triplet excited state (at the optimized S0 geometry). In addition,
time-dependent density functional theory (TD-DFT) calculations
with B3PW9141,42 were performed at the triple-ú level to assess
other possible excitations that contribute to the absorption spectra.
The emission energies were computed from the optimized T1
electronic state geometry and the S0 state (at the optimized T1
geometry). The dissociation energy (De) values for the dimer and
trimer S0 electronic states were predicted relative to dissociated
complexes, either by calculating the energy difference between the
optimized oligomer and the optimized monomers or by increasing
the separation between the monomeric complexes in the optimized
oligomer structure to ∼10 Å; the two methods agreed within 0.0026
eV (21 cm-1) for the dimer ground state. The latter method was,
therefore, adopted to estimate the excimer De value. All of these
computations were performed with the Gaussian 03 software
package,43 except for CCSD(T) calculations, which were done using
MOLPRO.44
Results and Discussion
1. Photophysics and Bonding in Au(CO)Cl. 1.1. Solution
Absorption. The absorption spectrum of Au(CO)Cl in aceto-
nitrile displays bands at λmax 195, 207, 219, and 280 nm, as
shown in Figure 1. The extinction coefficients for all these bands
are concentration dependent, deviating from Beer’s law. The
strong absorption bands at 195, 207, and 219 nm exhibit a
negative deviation from Beer’s law (Table 1), while the weak
(35) Purvis, G. D. I.; Bartlett, R. J. J. Chem. Phys. 1982, 76, 1910.
(36) Raghavachari, K.; Trucks, G. W.; Pople, J. A.; Head-Gordon, M.
Chem. Phys. Lett. 1989, 157, 479.
(37) Dunning, T. H., Jr. J. Chem. Phys. 1989, 90, 1007.
(38) Dunning, T. H., Jr.; Peterson, K. A.; Wilson, A. K. J. Chem. Phys.
2001, 114, 9244.
(39) Peterson, K. A.; Puzzarini, C. Theor. Chem. Acc. 2005, 114, 283.
(40) Figgen, D.; Rauhut, G.; Dolg, M.; Stoll, H. Chem. Phys. 2005, 311,
227.
(41) Becke, A. D. J. Chem. Phys. 1993, 98, 5648.
(42) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.;
Pederson, M. R. Phys. ReV. B 1992, 46, 6671.
(43) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.; Kudin, K.
N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.; Barone, V.;
Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson, G. A.;
Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.;
Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Klene, M.; Li,
X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Bakken, V.; Adamo, C.;
Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.;
Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma, K.;
Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.; Dapprich,
S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.; Rabuck, A.
D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui, Q.; Baboul, A.
G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.; Liashenko, A.;
Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Gonzalez, C.; Pople, J. A. Gaussian
03, revision C.02; Gaussian, Inc.: Wallingford, CT, 2004.
(44) Amos, R. D.; Bernhardsson, A.; Berning, A.; Celani, P.; Cooper,
D. L.; Deegan, M. J. O.; Dobbyn, A. J.; Eckert, F.; Hampel, C.; Hetzer,
G.; Knowles, P. J.; Korona, T.; Lindh, R.; Lloyd, A. W.; McNicholas, S.
J.; Manby, F. R.; Meyer, W.; Mura, M. E.; Nicklass, A.; Palmieri, P.; Pitzer,
R.; Rauhut, G.; Schu¨tz, M.; Schumann, U.; Stoll, H.; Stone, A. J.; Tarroni,
R.; Thorsteinsson, T.; Werner, H.-J. MOLPRO, A Package of Ab Initio
Programs Designed by H. -J. Werner and P. J. Knowles.
(33) APEX2 Software Suite, Version 2.0-2; Bruker AXS Inc., Madison,
WI, 2005.
(34) Møller, C.; Plesset, M. S. Phys. ReV. B 1934, 46, 618.

Gold(I) Organometallic Chain Complexes
Organometallics, Vol. 26, No. 10, 2007 2553
Figure 1. Absorption spectra for solutions of Au(CO)Cl in
acetonitrile at various concentrations: (a) 2.56 × 10-3 M; (b, c)
5.10 × 10-4 M; (d, top to bottom, respectively) 5.10 × 10-5, 2.56
× 10-5, 1.27 × 10-5, and 6.30 × 10-6 M.
Table 1. Concentration-Dependent Absorption Data for
Au(CO)Cl in Acetonitrile Solutions at Ambient Temperature
ꢀ, M-1 cm-1
Figure 2. Photoluminescence excitation (left) and emission (right)
spectra of crystals of Au(CO)Cl at various temperatures: (a) 295
K; (b) 235 K; (c) 180 K; (d) 135 K; (e) 100 K; (f) 77 K. The
spectra at different temperatures are offset for the sake of clarity.
c0, M
ꢀ195
ꢀ207
ꢀ219
ꢀ 280
6.30 × 10-6
1.27 × 10-5
2.56 × 10-5
5.10 × 10-5
5.10 × 10-4
2.56 × 10-3
1.74 × 10-4 a
7.52 × 104 3.17 × 104 1.84 × 104 undetectable
4.68 × 104 2.15 × 104 1.27 × 104 undetectable
3.28 × 104 1.53 × 104 8.75 × 103 undetectable
2.27 × 104 8.69 × 103 6.27 × 103 undetectable
670 and 630 nm, respectively. The changes are quite progressive
with temperature, as the two peaks become nearly identical in
intensity around 100 K. The emission color change from red at
room temperature to orange at cryogenic temperatures amounts
to a modest case of “luminescence thermochromism” resulting
from changes in the relative intensities of two emission bands.
This is reminiscent of prior occurrences of this phenomenon in
other monovalent coinage metal complexes, as are most well-
known for tetranuclear [Cu(halide)(arylamine)]4 clusters.48 The
two emission bands of the Au(CO)Cl solid exhibit essentially
indistinguishable excitation profiles (Figure 2). Lifetimes of 28.7
( 2.0 and 9.80 ( 0.16 µs are obtained (seemingly counterin-
tuitive) for the peak maxima at 295 and 77 K, respectively.
The decrease in the lifetime upon cooling from room temper-
ature to 77 K is due to the fact that the peak maximum at each
temperature represents an emitting state different from that at
the other temperature. Biexponential fits were used to extract
the lifetime of the minor component at selected temperatures
in the 77-295 K range and revealed that the lifetime of each
component followed the expected trend vs temperature, such
that longer lifetimes were obtained upon cooling due to
annihilation of the nonradiative decay rate. These results indicate
that the luminescence thermochromism in Au(CO)Cl occurs as
a result of relaxation (internal conversion) between two phos-
phorescent states, as opposed to being due to thermal compres-
sion of M‚‚‚M distances, which is known for other closed-shell
coordination compounds that contain M‚‚‚M packing in the solid
state.49,50 The assignment is as such because thermal contraction
usually leads to gradual shifts in the excitation and emission
energies vs temperature,50 which is different from the behavior
seen for Au(CO)Cl, as well as other literature precedents (e.g.,
ref 49), wherein the changes are due to relaxation between two
opaque
opaque
3.96 × 103 2.51 × 103 7.84
opaque
opaque
34.1
4.9 × 103 a 3.3 × 103 a
a From ref 24 with extinction coefficients quoted at 208 and 220 nm.
peak near 280 nm appears only at high concentrations and
exhibits a positive deviation from Beer’s law. These data suggest
the oligomerization of Au(CO)Cl complexes in solution, for
which aurophilic bonding is at least one contributing factor in
addition to other possible intermolecular forces such as multi-
polar interactions.45 Precedents for oligomerization of AuI
complexes in solution are known, which include work by Lin
and co-workers for diphosphinedithiolate dinuclear complexes46
and Patterson and co-workers for dicyano complexes.47 The high
extinction coefficient and negative deviation from Beer’s law
for the major absorption bands of Au(CO)Cl at 195, 207, and
219 nm in acetonitrile are consistent with charge-transfer
transitions characteristic of monomeric complexes. In contrast
to the absorbance bands attributed to the monomer, a number
of factors suggest that the weak band at 280 nm should be
assigned to an oligomeric species such as a [Au(CO)Cl]2 dimer;
these factors include the low extinction coefficient known for
such weakly bound aggregates,46,47 the red shift in energy
compared to the two aforementioned monomer bands, and the
positive deviation from Beer’s law.
1.2. Temperature-Dependent Solid-State Luminescence.
Figure 2 shows the luminescence spectra for a crystalline solid
of Au(CO)Cl versus temperature. The room-temperature spec-
trum gives rise to a broad unstructured emission with a peak in
the red region and an unresolved shoulder in the orange region.
The latter can be resolved from the red emission at low
temperatures. The emission maxima at 295 and 77 K occur at
(48) See, for example: (a) Ryu, C. K.; Kyle, K. R.; Ford, P. C. Inorg.
Chem. 1991, 30, 3982. (b) Lindsay, E.; Ford, P. C. Inorg. Chim. Acta 1996,
242, 51. (c) Ford, P. C.; Cariati, E.; Bourassa, J. Chem. ReV. 1999, 99,
3625.
(49) Dias, H. V. R.; Diyabalanage, H. V. K.; Eldabaja, M. G.; Elbjeirami,
O.; Rawashdeh-Omary, M. A.; Omary, M. A. J. Am. Chem. Soc. 2005,
127, 7489.
(45) Liau, R.-Y.; Mathieson, T.; Schier, A.; Berger, R. J.; Runeberg,
N.; Schmidbaur, H. Z. Naturforsch., B 2002, 57, 881.
(46) Tang, S. S.; Chang, C. P.; Lin, I. J. B.; Liou, L. S.; Wang, J. C.
Inorg. Chem. 1997, 36, 2294.
(47) Rawashdeh-Omary, M. A.; Omary, M. A.; Patterson, H. H. J. Am.
Chem. Soc. 2000, 122, 10371.

2554 Organometallics, Vol. 26, No. 10, 2007
Elbjeirami et al.
Figure 3. Hartree-Fock contours for frontier molecular orbitals of the Au(CO)Cl monomer at its optimized S0 ground state (left), Franck-
Condon vertical T1 excited state at the S0 geometry (middle), and fully optimized T1 excited state (right). The data are based on the output
files of MP2 calculations with the following basis sets: cc-pVDZ for Au, cc-pVDZ for CO, and cc-pV(D+d)Z for Cl. Orbital energies
shown are in eV.
emitting states such that one band increases in intensity while
another decreases. Analysis of the changes in the relative
intensities of the two peaks in the 77-180 K temperature range
(relative peak areas from a two-Gaussian fit at each temperature)
gives rise to an activation energy of 170 cm-1 for the internal
conversion from the orange-emitting state to the red-emitting
state. Due to the small energy difference between the two
emission peaks (<1000 cm-1) and the microsecond lifetimes
for each, we hypothesize that the red and orange emissions
represent spin-orbit states correlating to the same T1 triplet
state, although two closely spaced triplets (T1 and T2) could
in the electronic absorption and emission transitions due to the
ground-state aurophilic bonding and excited-state excimeric
bonding by comparing the contours of the orbitals involved as
well as the transition energies in monomeric and dimeric models.
Figures 3 and 4 show the Hartree-Fock contours for selected
frontier molecular orbitals from MP2 calculations for the Au-
(CO)Cl monomer and the [Au(CO)Cl]2 antiparallel dimer,
respectively, in their optimized singlet ground state (S0),
Franck-Condon vertical T1 excited-state at the S0 geometry,
and the fully optimized T1 state. For the dimer in Figure 4, we
note that the contour of the lowest unoccupied molecular orbital
(LUMO) in the optimized S0 state is not necessarily the same
as that of the higher singly occupied molecular orbital (SOMO)
of the T1 states. Despite the extensive use of HOMO and LUMO
contours in the literature to make conclusions about absorption
and emission assignments, this use remains fundamentally
inaccurate. Rather, it is more appropriate to draw conclusions
about the assignment of the lowest energy excitation bands on
the basis of the contours of the SOMOs of the Franck-Condon
excited state and conceive the emission bands from the
optimized excited states, instead of basing them on the HOMO
and LUMO of the ground state. In Figure 4, for example, the
ground-state S0 LUMO contour suggests that an Au-Au
π-bonded excimer may form, while an Au-Au σ-bonded
excimer is actually suggested on the basis of the contour of the
higher SOMO of either the Franck-Condon T1 state involved
in the excitation process or the optimized T1 state involved in
the phosphorescence process. Hence, we assign the lowest
energy phosphorescence and excitation bands in the antiparallel
dimer to a Au-Au σ-bonded excimer. The assignment of the
luminescence bands to excimeric Au-Au bonding is consistent
with the experimental spectral data in Figure 1, because the
unstructured profile of the phosphorescence bands and the huge
3
not be ruled out. The large energy difference between the DJ
levels (15 040, 17 640, and 27 760 cm-1 for J ) 3-1,
respectively)51 of free Au(I) due to large spin-orbit coupling
(ê5d ≈ 5100 cm-1 52
)
makes it reasonable to make the afore-
mentioned hypothesis that two spin-orbit states may be
responsible for the red and orange emission peaks of the Au-
(CO)Cl solid.
1.3. Assignments of the Electronic Transitions. In an effort
to shed more light on the exact orbital nature involved in the
electronic transitions for Au(CO)Cl in solution and in the solid
state, we have carried out a series of ab initio and DFT
calculations. These calculations attempt to model the changes
(50) See, for example: (a) Gliemann, G.; Yersin, H. Struct. Bonding
1985, 62, 87. (b) Connick, W. B.; Henling, L. M.; Marsh, R. E.; Gray, H.
B. Inorg. Chem. 1996, 35, 6261. (c) Nagasundaram, N.; Roper, G.; Biscoe,
J.; Chai, J. W.; Patterson, H. H.; Blom, N.; Ludi, A. Inorg. Chem. 1986,
25, 2947. (d) Burini, A.; Bravi, R.; Fackler, J. P., Jr.; Galassi, R.; Grant, T.
A.; Omary, M. A.; Pietroni, B. R.; Staples, R. J. Inorg. Chem. 2000, 39,
3158.
(51) Moore, C. E. Atomic Energy LeVels, Vol. III; United States
Department of Commerce/National Bureau of Standards: Washington, DC,
1958.
(52) Griffith, J. S. Theory of Transition Metal Ions; Cambridge University
Press: Cambridge, U.K., 1964.

Gold(I) Organometallic Chain Complexes
Organometallics, Vol. 26, No. 10, 2007 2555
Figure 4. Restricted open-shell Hartree-Fock contours for frontier molecular orbitals from calculations for the [(CO)AuCl]2 antiparallel
dimer (with Cl-Au-CO at the top and OC-Au-Cl at the bottom). The contours are shown for the S0 ground state (left), the Franck-
Condon vertical T1 excited state at the S0 geometry (middle), and the fully optimized T1 state (right). The methods and basis sets are the
same as those in Figure 3. Orbital energies shown are in eV.
Stokes shift are typical signatures of metal-centered excimeric
phosphorescence in d10 systems.25,26 A recent study by Coppens,
Omary, and co-workers has verified such an assignment for a
trinuclear Cu(I) pyrazolate complex by pump-probe time-
resolved single-crystal X-ray diffraction, in which a startling
intermolecular motion was observed in the crystal structure of
the excimeric phosphorescent state manifested by compressions
of Cu‚‚‚Cu distances and interplanar separations between
adjacent trimers by 0.56 and 0.65 Å, respectively!53 Kunkely
and Vogler noted that the huge Stokes shift in the Au(CO)Cl
solid was indicative of large structural changes associated with
electronic excitation, but they suggested that the emission was
a “d-s transition” localized on the Au atom, which leads to
the population of a σ* orbital in a monomeric complex.24 These
authors also stated that “the emission may be slightly modified
by Au-Au interaction”. However, the additional experimental
and computational data herein allow us to conclude that the
emission is instead caused by the intermolecular Au-Au
excimeric covalent bonding in the phosphorescent state, despite
the weakness of the metallophilic Au‚‚‚Au interaction in the
ground state.
Figure 3 revisits the electronic structure of the Au(CO)Cl
monomer, including the assignment of the low-lying possible
electronic absorption and emission transitions. The ground-state
HOMO and other low-lying occupied orbitals show the highest
electron density on the Cl-Au “left” side of the Cl-Au-CO
molecule, particularly on the chlorine atom. This is hardly
surprising, since the chloride is conventionally classified as a
π-donor ligand. Indeed, the π interactions with the chloride
destabilize the Au dπ orbitals such that they become the HOMO
2
instead of the strongly antibonding dz ligand-field σ* orbital,
(53) Vorontsov, I. I.; Kovalevsky, A. Yu.; Chen, Y.-S.; Graber, T.;
Novozhilova, I. V.; Omary, M. A.; Coppens, P. Phys. ReV. Lett. 2005, 94,
193003.
which is the HOMO-1 according to Figure 3. In contrast, the
frontier virtual orbitals, which are the likely targets to be

2556 Organometallics, Vol. 26, No. 10, 2007
Elbjeirami et al.
parison with Au(PR3)Cl complexes.9,24,54 More insights about
the assignments of the electronic absorption bands in Au(CO)-
Cl, including association effects, are provided in the discussion
of the TD-DFT results below.
Table 2. MP2-Calculated Geometries for the Optimized S0
and T1 States of [(CO)AuCl]n (n ) 1-3 with Antiparallel
Isomers for n ) 2, 3)a
bond length
bond angle
To complete the qualitative description of the Au(CO)Cl
electronic structure, we now discuss the issue of the monomer
emission. Figure 3 shows that both the HOMO and LUMO are
doubly degenerate orbitals in which Au exhibits π bonding and
antibonding interactions with the CO ligand, respectively,
suggesting that photon absorption renders the excited molecule
Jahn-Teller unstable. Indeed, optimizing the T1 triplet state
gives rise to a drastic molecular distortion in which the molecule
is no longer linear, as the C-Au-Cl angle becomes severely
bent and the orbital degeneracy no longer exists (Figure 3). Due
to the severe Jahn-Teller distortion, the calculated phospho-
rescence energy of the excited monomer lies in the near-IR
region (vide infra). While interesting, this result, however, does
not have a strong bearing on the luminescence properties of
the compound and is not in agreement with the experimental
data in this paper. It is believed that the Au-centered lumines-
cence from the two-coordinate Au(I) compounds presented
herein is observed only in the presence of the Au‚‚‚Au
interaction by association of monomeric molecules, which is
consistent with what has been shown in previous literature.25,26
Since the phosphorescence by this compound is exhibited in
the solid state, whereby monomeric molecules pack adjacent
to one another via aurophilic bonding, and since the computa-
tional data show that dimeric molecules exhibit Au-Au
excimeric bond formation in the T1 state, the monomer is not a
good model for the luminescence behavior of the solid. The
monomer model is, however, sufficient to describe dilute
solutions of the Au(CO)Cl compound, while such solutions do
not exhibit luminescence.
1.4. Ground- and Excited-State Bonding in [Au(CO)Cl]n
Species. Table 2 shows the calculated geometric parameters for
[Au(CO)Cl]n species (n ) 1-3), while Table 3 shows the
corresponding dissociation energies and vibrational frequencies
for selected modes. The computed bond distances and angles
within each linear complex in the S0 ground state are comparable
with the experimental crystal structure determined for Au(CO)-
Cl,2 with slight improvements obtained with the larger triple-ú
basis sets. In Table 3, the calculated νCtO frequencies have been
predicted with both MP2 and CCSD(T). With CCSD(T) at the
triple-ú level, the free CO frequency changes from 2153 to 2176
cm-1 in the Au(CO)Cl species, which displays a slight blue
shift upon complexation and provides computational evidence
in line with the anomalous blue shift known experimentally for
this metal carbonyl.7 It must be noted that MP2 at the triple-ú
level is an insufficient level of theory to correctly predict this
trend. In fact, previous computations using MP2 with different
basis sets also did not attain this anomalous shift.9,55 (While
MP2 is not appropriate for the quantitative description of the
νCtO shift in the ground state, MP2 is traditionally considered
an adequate level of theory to compute vibrational frequencies
and is used later for the qualitative comparison of ground and
excited states.)
basis
state set x Cl-Au Au-C C-O Au-Au Cl-Au-C Au-C-O
Au(CO)Cl
S0
T1
D
T
D
T
2.22
2.20
2.34
2.31
1.82
1.85
1.94
1.93
1.15
1.14
1.16
1.15
[Au(CO)Cl]2
S0
T1
D
T
D
T
2.23
2.22
2.24
2.22
1.87
1.86
1.92
1.90
1.15
1.14
1.16
1.15
3.20
177.0
177.1
158.3
160.3
177.8
177.9
166.7
167.7
3.17
2.62
2.61
[(CO)AuCl]3
S0
D
2.23
2.26
1.87
1.93
1.15
3.13
179.5
180.0
180.0
180.0
exptlb
S0
1.11
3.38
a The basis set combinations are cc-pVxZ for C and O, cc-pV(x+d)Z
for Cl, and cc-pVnZ-PP for Au. Calculated bond lengths are in Å and bond
angles in deg. b Reference 2.
Table 3. Selected IR Frequencies (cm-1) and Dissociation
Energies (eV) Calculated with MP2a
molecule
CO
basis set x
ν(C-O)
ν(Au-Au)
De
D
T
D
T
D
T
D
T
D
T
D
2113 (2143)
2121 (2153)
2119 (2168)
2121 (2176)
2052
Au(CO)Cl
Au(CO)Cl*
[Au(CO)Cl]2
[(CO)AuCl]2*
[(CO)AuCl]3
2027
2128, 2129b
2130, 2132b
2033, 2594b
2034, 2733b
2125, 2123,b 2137c
50
47
153
160
44, 67b
0.44
0.43
2.49
2.27
0.91
a Additional values given in parentheses were computed with CCSD(T).
An asterisk indicates a molecule in its optimized T1 excited state.
b Symmetric and asymmetric stretches, respectively. c The mode at 2137
cm-1 is primarily from the CO on the central Au while the two others are
primarily from the carbonyls on the two peripheral Au atoms.
populated in the low-energy absorption transitions, generally
exhibit higher electron density on the Au-CO “right” side of
the Cl-Au-CO molecule than on the “left” side. The LUMO
exhibits a strong Au 6px,y character, where gold is engaged in
π* antibonding interactions with the ligands; interestingly, the
admixture with the ligands involves rather similar contributions
from a CO π* ligand orbital and a Cl 3px,y atomic orbital. This
is somewhat surprising, because CO is a better π acceptor than
Cl. The aforementioned orbital description is qualitatively the
same within each monomer in the dimer (Figure 4), in that the
frontier orbitals show delocalized character on all atoms. The
common practice in the literature is to assign electronic
transitions on the basis of the major compositions of the
molecular orbitals allegedly involved in the electronic transi-
tions. Thus, utilizing the data in Figures 3 and 4 suggests that
the lowest energy excitation bands in Au(CO)Cl are essentially
due to halide-to-metal charge transfer (XMCT) transitions from
filled orbitals with a strong chloride 3px,y character to vacant
orbitals with a strong gold 6px,y character. Clearly, chemical
intuition alone cannot be followed for the assignment of the
lowest energy absorption bands in this molecule, because one
would expect an opposite assignment (i.e., metal to ligand charge
transfer or MLCT) from a gold 5d filled orbital to a carbonyl
π* vacant orbital. Indeed, the latter assignment has been
commonly suggested in the literature for Au(CO)Cl in com-
The ground-state aurophilic bonding is accounted for in the
present calculations because MP2 is known to describe metal-
lophilic bonding.56 Indeed, the theoretical methodology em-
ployed overdescribes this closed-shell bonding interaction, as
(54) Savas, M. M.; Mason, W. R. Inorg. Chem. 1987, 26, 301.
(55) Fortunelli, A.; Germano, G. J. Phys. Chem. A 2000, 104, 10834.
(56) For reviews, see: (a) Pyykko¨, P. Chem. ReV. 1997, 97, 597. (b)
Pyykko¨, P. Angew. Chem., Int. Ed. 2004, 43, 4412. (c) Pyykko¨, P. Inorg.
Chim. Acta 2005, 358, 4113.

Gold(I) Organometallic Chain Complexes
Organometallics, Vol. 26, No. 10, 2007 2557
the calculated Au‚‚‚Au aurophilic bond distances in the dimer
and trimer (∼3.13-3.20 Å, as shown in Table 2) are shorter
than the crystallographic value of 3.38 Å.2 This is partly due to
inherent limitations in the MP2 method; see ref 57 for a detailed
discussion of the limitations of various methods and basis sets
for the description of metallophilic bonding. The De value for
the S0 states of the dimer and trimer, as shown in Table 3, of
∼0.45 eV per bond (∼10 kcal/mol) suggests aurophilic bonding
in the ground state. This value is commensurate with the typical
values reported for aurophilic bonds in the literature.56 The
trimer De value for two aurophilic bonds is slightly higher than
twice the dimer De value, suggesting a small degree of
cooperativity. Table 3 also shows that aurophilic association in
the antiparallel linear oligomers leads to a νAu-Au frequency of
50 cm-1 in the dimer and two frequencies in the trimer of 44
and 67 cm-1, representing symmetric and asymmetric Au-Au-
Au stretching, respectively. When the ground-state monomer
energies lower than those in dilute solutions. Although the dimer
absorptions are not as greatly shifted from the monomer bands
as is seen experimentally, they do demonstrate that aurophilic
bonding leads to a shift in the absorption spectra. It is likely
that the computed smaller-than-expected red shift upon dimer-
ization is a result of the known limitation of density functional
theory in describing dispersion forces, which are significant in
metallophilic bonding.57 Overall, the computed absorption
spectra for the monomer and dimer are in reasonable agreement
with the experimental absorption spectra, particularly
because our theoretical treatment does not include spin-orbit
coupling.
For the luminescence emission and excitation transitions, we
only modeled the lowest spin-forbidden transitions, since
phosphorescence usually occurs from the lowest triplet (T1)
excited state. Table 4 shows the computed photophysical
parameters for monomer and dimer models. The near-IR
emissions of the monomer are not comparable to experiment,
because monomers are not good models for describing the Au-
centered emissions in two-coordinate Au(I) complexes (vide
supra), as discussed earlier. As for the dimer, the MP2-calculated
phosphorescence energy lies in the violet region near 400 nm
with both basis set levels used, while the experimental solid-
state emission is in the red region. Although the calculated
Stokes shift is similar in magnitude to that experimentally
determined for the solid and is consistent with the drastic
excited-state distortion in the excimeric state, the discrepancy
between the calculated violet T1 f S0 excimer emission and
the red phosphorescence of solid Au(CO)Cl renders the dimer
model insufficient to describe the species responsible for the
solid-state emission. Given this discrepancy, we decided to
explore the hypothesis that the solid-state emission is caused
by an excited-state oligomer that is larger than a dimer due to
delocalized Au-Au excimeric bonding beyond two adjacent
units. Because modeling trimers and larger clusters with MP2
is rather demanding computationally, the problem proved easier
to solve experimentally. Hence, we have carried out photolu-
minescence studies for 77 K frozen solutions of Au(CO)Cl at
different concentrations in CH2Cl2 (which was chosen due to
solubility reasons, although it does not form a transparent glass
at 77 K, like 2-MeTHF or EPA). Figure 5 shows the results of
these studies versus concentration. Frozen solutions of Au(CO)-
Cl in dichloromethane exhibit strong luminescence, even with
concentrations as low as 5 × 10-5 M, at which the dominant
band maximum is at 410 nm along with a small shoulder around
560 nm, as shown in Figure 5. The longer wavelength peak is
due to larger oligomers that have an aggregation number close
to that of the luminescent species in the solid state, while the
shorter wavelength peak is due to a smaller oligomer, possibly
a dimer. A further increase in the concentration leads to a
corresponding increase in the relative intensity of the longer
wavelength band, which is red-shifted to 590 nm, as it becomes
the most pronounced emission band at 0.05 M. This red shift is
attributed to the formation of larger oligomers as the solution
concentration increases. The position of the shorter wavelength
band is unaffected by concentration. Our study clearly shows
the presence of two types of oligomers in frozen solutions of
Au(CO)Cl in dichloromethane and suggests the increasing
tendency for the formation of larger oligomers upon increasing
the concentration. Lifetime data obtained for both peaks are
consistent with Au-centered phosphorescent emissions from
ν
CtO frequency is compared with those of the associated dimer
and trimer, the aurophilic bonded systems lead to a further blue
shift of the CO frequency, as presumed since the aurophilic
bonding leads to less π back-bonding to the CO (π*), which
further stabilizes the CtO bond. On the basis of the dimer
electronic structure (Figure 4) discussed above, photoexcitation
leads to the formation of an excimer with a bona-fide Au-Au
covalent σ bond in the relaxed geometry of the T1 state of the
dimer. This is manifested by the 2.61 Å distance calculated for
this bond, representing a contraction by 0.56 Å from the ground-
state aurophilic bond of the dimer (Table 2, x ) T). Upon
excitation to the T1 state, the De value of the dimer is drastically
increased by over 5 times to 2.27 eV in the resulting excimer,
as shown in Table 3. The formation of this Au-Au bonded
excimer also leads to drastic changes in the vibrational frequen-
cies (Table 3). Thus, the Raman-active νAu-Au dimer frequency
calculated at the MP2 triple-ú level exhibits a drastic blue shift
to 160 cm-1 in the optimized T1 excimeric phosphorescent state
from the 47 cm-1 value in the optimized S0 aurophilic ground
state. Meanwhile, the νCtO frequency in the dimer S0 ground
state is centered at 2131 cm-1, which becomes two well-
separated bands at 2034 and 2733 cm-1 in the T1 state. The
average of these two νCtO values (2383 cm-1) is blue-shifted
from the ground-state value by 252 cm-1, which is likely
due to the diminished π back-bonding to CO in the excimeric
state caused by the now chemically bonded Au-Au
centers. This is further supported by a Au-C bond (by ∼0.05
Å) that is longer in the T1 excimer than in the S0 state dimer,
which is also consistent with a weaker π back-bonding in the
excimer.
1.5. Quantitative Evaluation of Photophysical Parameters.
This section discusses the quantum-mechanical calculation of
the electronic absorption and luminescence transitions for
monomer and dimer models in relation to the experimental
values. For the electronic absorption, TD-DFT is used because
of the convenience of this method to compute multiple spectral
transitions. The computed results are listed in Table 4 and can
be compared to the experimental solution absorption data in
Figure 2. The monomer TD-DFT calculations give rise to three
lowest energy absorption bands with nonzero (g0.01) oscillator
strengths at 198, 227, and 240 nm, all clearly within the bands
seen experimentally for dilute solutions of Au(CO)Cl in
acetonitrile (Figure 2). The TD-DFT calculations further predict
two lowest energy absorption bands for the dimer at 251 and
265 nm, clearly red-shifted from the monomer absorption bands,
consistent with the experimental finding that increasing the
concentration of Au(CO)Cl leads to the appearance of bands at
(57) Omary, M. A.; Sinha, P.; Bagus, P. S.; Wilson, A. K. J. Phys. Chem.
A. 2005, 109, 690.

2558 Organometallics, Vol. 26, No. 10, 2007
Elbjeirami et al.
distorted with bonding properties different from those of the
ground state, as in the case for the Au(CO)Cl complex. The
lifetimes at 77 K for (CyNC)AuCl, (t-BuNC)AuCl, and (MeN-
C)AuCl are 30, 40, and 79 µs, respectively. The broad/
unstructured profiles and the magnitudes of the lifetime values
for the emission bands of the gold isonitrile chloride and gold
carbonyl chloride compounds are consistent with Au-centered
phosphorescent emissions from excited states in which the Au-
Au excimeric bonding is delocalized beyond two adjacent
complexes. Also as in Au(CO)Cl, solutions of (RNC)AuCl in
dicholromethane show no luminescence at ambient temperature
but concentration-dependent studies for frozen solutions show
higher energy excimeric phosphorescent emissions significantly
blue-shifted from the solid-state phosphorescence. To demon-
strate with a representative example, Figure 7 shows that frozen
solutions of (t-BuNC)AuCl in dichloromethane give rise to at
least three phosphorescence bands that we assign to three *[(t-
BuNC)AuCl]n oligomers. At high concentration, the relative
intensities of the three major bands are dependent on the
excitation maxima (Figure 7). Upon excitation at short wave-
lengths, the band at 680 nm is the dominant feature, whereas
the band at 420 nm becomes the dominant feature upon
excitation at longer wavelength. At lower concentration, the
dominant bands are those at 420 and 525 nm when using long-
wavelength excitation maxima, while no peaks are observed
when using short-wavelength excitation, as shown in Figure 7.
Lifetime data for the three peaks are consistent with Au-centered
phosphorescent emissions (τ ) 0.97 ( 0.12, 2.6 ( 0.29, and
32.5 ( 1.97 µs are obtained for the bands with maxima at 420,
525, and 680 nm, respectively).
Figure 5. Photoluminescence excitation (left) and emission (right)
spectra of frozen solutions (77 K) of Au(CO)Cl in CH2Cl2 at
different concentrations: (a) 5.0 × 10-5 M; (b) 5.0 × 10-3 M; (c)
5.0 × 10-2 M.
largely distorted excited states (τ ) 1.21 ( 0.15 and 7.27 (
0.46 µs for the bands with maxima at 410 and 590 nm,
respectively). Remarkably, the most dilute solutions to exhibit
luminescence give rise to phosphorescence bands that lie in the
same violet region as that where the MP2 calculations in Table
4 predict the excimer T1 f S0 phosphorescence to take place.
As the concentration is increased, the violet emission near 400
nm is diminished while lower energy phosphorescence bands
are observed, ultimately approaching the solid-state behavior
in frozen solutions with the highest concentration studied. These
findings bear out the aforementioned hypothesis, because the
larger oligomers in the solid state will be broken down into
smaller oligomers upon dissolution, which would then emit at
higher energy than in the solid state.
2. Photophysics of Linear-Chain LAuX Complexes. Neu-
tral isonitrile compounds that have supramolecular structures
similar to the that of Au(CO)Cl (i.e. antiparallel chains) have
similar luminescence properties with an orange-red phospho-
rescent emission and an extremely large Stokes shift (∼15 000-
20 000 cm-1), as shown in Figure 6. Stokes shifts with such
large magnitudes indicate that the excited states are drastically
The spectral findings herein suggest that predictive informa-
tion about the supramolecular structure may be obtained by the
luminescence behavior. This is exemplified by crystals of (Me4-
BuNC)AuCl, whose red-orange luminescence (Figure 6) an-
ticipated an extended-chain supramolecular structure, which was
verified crystallographically at a later stage during the course
of this investigation. Figure 8 shows that this compound exhibits
a crystal structure in which the molecules are packed in zigzag
chains with alternating short (3.418 Å) and long (4.433 Å)
aurophilic distances. The crystal data for the complex are given
in Table 5, while selected interatomic distances and angles for
the neutral complex are given in Table 6. The Au-Cl distance
is 2.253(2) Å, and the C-Au distance is 1.929(2) Å. For a
comparison, the C-Au and Au-Cl distances in (CyNC)AuCl,
in which molecules associate in a similar head-to-tail fashion,
are 1.961(4) and 2.258(4) Å, respectively.3 The increase in the
Au-C bond length might be due to an increase in the
electrostatic effects of the ligand. Each gold atom is nearly
linearly coordinated by the two ligands, as indicated by the Cl-
Au-C angle of 176.37(6)°, which is consistent with the
Table 4. Computed Photophysical Parameters for the Monomer and Antiparallel Dimer of Au(CO)Cl in Comparison with
Experimental Data for the Solid and Frozen Solutions of Au(CO)Cla
model
Au(CO)Cl
basis set x
λ
exc(S0fT1) (nm)
λ
em(T1fS0) (nm)
Stokes shift (103 cm-1
)
λabs(S0 f Sn) (nm)
D
T
D
T
220
220
246
244
322
282
824
788
407
394
674
412
33.344
32.851
16.088
15.573
16.200
11.200
197 (0.01), 227 (0.27), 240 (0.03)
251 (0.01), 265 (0.06)
[Au(CO)Cl]2
exptl (solid)
exptl (frozen soln)
a The basis set combinations are cc-pVxZ for C and O, cc-pV(x+d)Z for Cl, and cc-pVnZ-PP for Au. The λexc and λem values for the spin-forbidden
transitions are calculated via MP2, while the λabs values for spin-allowed transitions are calculated via TD-DFT with the oscillator strength values given in
parentheses.

Gold(I) Organometallic Chain Complexes
Organometallics, Vol. 26, No. 10, 2007 2559
Au‚‚‚Au contacts of 3.3894(7) and 3.5816(7) Å for X ) Cl,
3.4864(9) and 3.7036(9) Å for X ) Br, and 3.7182(11) and
3.9304(12) Å for X ) I.3 The head-to-tail orientation along the
chains of (Me4BuNC)AuCl molecules is favored, since the
adjacent dipoles of the molecules alternates as the dipole-dipole
interactions are dominant.
Although the ground-state aurophilic bonding is essentially
nonexistent between adjacent dimers in crystals of (Me4BuNC)-
AuCl, the luminescence properties of these crystals are distinctly
different from those of dimeric species such as the lower
concentrations in frozen solutions of Au(CO)Cl and (t-BuNC)-
AuCl shown in Figures 5 and 7, respectively, or single crystals
of ((p-tosyl)CH2NC)AuCl.28 These species exhibit a dimeric
structure and blue-green emissions with Stokes shifts signifi-
cantly smaller than those observed for the crystalline solid state
of infinite-chain complexes, including crystals of (Me4BuNC)-
AuCl. Thus, the red-orange emission with a huge Stokes shift
for (Me4BuNC)AuCl crystals suggests the delocalization of the
excimeric bonding beyond two adjacent molecules in the solid
state, possibly due to cooperative M-M excimeric bonding, as
suggested for other systems exhibiting this bonding.57 For this
delocalization to take place, the 4.433 Å distances in (Me4-
BuNC)AuCl chains of dimers must shrink in the phosphorescent
state of these crystals. While this distance is clearly too long
for any significant ground-state interactions, it is not too long
for excimeric interactions involving d10 monovalent coinage
metals. A relevant comparison exists with the aforementioned
time-resolved X-ray diffraction study for crystals of the tri-
nuclear Cu(I) pyrazolate {[3,5-(CF3)2Pz]Cu}3.53 The emitting
species of that system has been identified as a hexanuclear
dimer- of-trimer unit, owing to intertrimer Cu-Cu contraction.
Remarkably, although the ground-state Cu‚‚‚Cu intertrimer
separations are 3.787 and 4.018 Å, it was the longer distance
that contracted in the phosphorescent state.53 If one considers
the difference in the van der Waals radii between the two metals
Figure 6. Photoluminescence excitation and emission spectra for
crystals of (a) (Me4BuNC)AuCl, (b) (t-BuNC)AuCl, (c) Au(CO)-
Cl, (d) (CyNC)AuCl, and (e) (MeNC)AuCl at 77 K.
analogous structure of (CyNC)AuCl, where the Cl-Au-C angle
is 178.2(4)°.3 The dihedral angles of C-Au-Au-C and Cl-
Au-Au-Cl are all 180.0°. Molecules of (Me4BuNC)AuCl self-
associate to from dimers with an Au‚‚‚Au distance of 3.418 Å,
but these dimers pack to form extended loose chains. The
neighboring Au‚‚‚Au contact between adjacent dimers in the
loose chain is 4.433 Å, which lies outside the accepted range
of significant aurophilic bonding (2.9-3.6 Å). For comparison,
crystals of (CyNC)AuX form infinite chains with alternating
Figure 7. Photoluminescence emission spectra of frozen solutions (77 K) of (t-BuNC)AuCl in CH2Cl2 at different excitation wavelengths:
(a) 335 nm; (b) 310 nm; (c) 275 nm. The spectra on the left side are for a 0.05 M solution, while the spectra on the right side are for a 5
× 10-5 M solution.

2560 Organometallics, Vol. 26, No. 10, 2007
Elbjeirami et al.
Table 5. Crystal Data for [(CH3)3CCH2C(CH3)2NC]AuCl
empirical formula
formula wt
C9H17NClAu
371.65
temp
100(2) K
savelength
0.710 73 Å
cryst size
cryst habit
cryst syst
0.44 × 0.36 × 0.10 mm
clear, colorless plate
monoclinic
space group
C2/c (No. 15)
unit cell dimens
a
b
c
19.9362(9) Å
8.0691(4) Å
14.8652(7) Å
90°
R
â
98.263(2)°
90°
γ
V
Z
2366.50(19) Å3
8
density (calcd)
abs coeff
F(000)
2.086 Mg/m3
12.613 mm-1
1392
Table 6. Selected Bond Lengths (Å) and Bond Angles (deg)
for the Structure of [(CH3)3CCH2C(CH3)2NC]AuCl
Au1-C1
N1-C1
C2-C4
C2-C5
1.929(2)
1.137(3)
1.524(3)
1.536(3)
Au1-Cl1
N1-C2
2.2532(5)
1.468(3)
1.529(3)
0.9800
C2-C3
C3-H3A
C1-Au1-Cl1
N1-C1-Au1
N1-C2-C3
N1-C2-C5
C3-C2-C5
176.37(6)
C1-N1-C2
N1-C2-C4
C4-C2-C3
C4-C2-C5
C2-C3-H3A
174.9(2)
177.06(18)
105.26(15)
108.38(16)
109.05(18)
107.69(17)
110.08(18)
115.84(16)
109.5
(rvdw ) 1.40 Å for CuI vs rvdw ) 1.70 Å for AuI),58 the CuI‚‚
‚CuI separation of 4.018 Å in {[3,5-(CF3)2Pz]Cu}3 is chemically
similar to an analogous AuI‚‚‚AuI separation of 4.618 Å, which
is even longer than the 4.433 Å intermetal distances in chains
of (Me4BuNC)AuCl complexes.
Acknowledgment. Partial support by the Robert A. Welch
Foundation (Grant B-1542 to M.A.O.), the National Science
Foundation (CAREER Award Grants CHE-0349313 and CHE-
0239555 to M.A.O. and A.K.W., respectively), the U.S.
Department of Energy (Award DE-FC26-06NT42856 to M.A.O.),
and the U.S. Department of Education (P116Z050070 to A.K.W.
for the Center for Advanced Scientific Computing and Modeling
(CASCaM)) are gratefully acknowledged. Computational re-
sources were provided by the NSF-CRIF (CHE-0342824) and
the National Computational Science Alliance under CHE010021,
which utilized the NCSA IBM p690. Additional computational
support was provided by the University of North Texas
Academic Computing Services.
Figure 8. Crystal structure of [(CH3)3CCH2C(CH3)2NC]AuCl. The
top figure shows anisotropic atomic displacement ellipsoids for the
non-hydrogen atoms at the 50% probability level, while hydrogen
atoms are displayed with an arbitrarily small radius. The bottom figure
shows the molecular packing with the view showing the projection
down the b axis.
Supporting Information Available: CIF file giving X-ray
crystallographic data for [(CH3)3CCH2C(CH3)2NC]AuCl. This
material is available free of charge via the Internet at http://
pubs.acs.org.
(58) Bondi, A. J. Phys. Chem. 1964, 68, 441.
OM060524T