Synthesis, structural characterization, and theoretical studies of goid(I) and gold(I)-gold(III) thiolate complexes: Quenching of gold(I) thiolate luminescence

Article abstract of DOI:10.1021/ic051168u

The gold(I) thiolate complexes [Au(2-SC₆H₄NH ₂)(PPh₃)] (1), [PPN][Au(2-SC₆H ₄NH₂)₂] (2) (PPN = PPh₃=N=PPh ₃), and [{Au(2-SC₆H₄NH₂)} ₂(μ-dppm)] (3) (dppm = PPh₂CH₂PPh ₂) have been prepared by reaction of acetylacetonato gold(I) precursors with 2-aminobenzenethiol in the appropriate molar ratio. All products are intensely photoluminescent at 77 K. The molecular structure of the dinuclear derivative 3 displays a gold-gold intramolecular contact of 3.1346-(4) A. Further reaction with the organometallic gold(III) complex [Au(C ₆F₅)₃(tht)] affords dinuclear or tetranuclear mixed gold(I)-gold(III) derivatives with a thiolate bridge, namely, [(AuPPh ₃){Au(C₆F₅)₃}(μ₂-2- SC₆H₄NH₂)] (4) and [(C₆F ₅)₃Au(μ₂-2-SC₆H ₄NH₂)(AudppmAu)(μ₂-2-SC₆H ₄NH₂)Au(C₆F₅)₃] (5). X-ray diffraction studies of the latter show a shortening of the intramolecular gold(I)-gold(I) contact [2.9353(7) or 2.9332(7) A for a second independent molecule], and short gold(I)-gold(III) distances of 3.2812(7) and 3.3822(7) A [or 3.2923(7) and 3.4052(7) A] are also displayed. Despite the gold-gold interactions, the mixed derivatives are nonemissive compounds. Therefore, the complexes were studied by DFT methods. The HOMOs and LUMOs for gold(I) derivatives 1 and 3 are mainly centered on the thiolate and phosphine (or the second thiolate for complex 2), respectively, with some gold contributions, whereas the LUMO for derivative 4 is more centered on the gold(III) fragment. TD-DFT results show a good agreement with the experimental UV-vis absorption and excitation spectra. The excitations can be assigned as a S → Au-P charge transfer with some mixture of LLCT for derivative 1, an LLCT mixed with ILCT for derivative 2, and a S → Au...Au-P charge transfer with LLCT and MC for derivative 3. An LMCT (thiolate → Au^III mixed with thiolate → Au-P) excitation was found for derivative 4. The differing nature of the excited states [participation of the gold(III) fragment and the small contribution of sulfur] is proposed to be responsible for quenching the luminescence.

Full text of DOI:10.1021/ic051168u

Inorg. Chem. 2006, 45, 1059−1068
Synthesis, Structural Characterization, and Theoretical Studies of
Gold(I) and Gold(I) Gold(III) Thiolate Complexes: Quenching of Gold(I)
Thiolate Luminescence
−
Manuel Bardaj´ı,*,†,‡,§ Maria Jose´ Calhorda,| Paulo J. Costa,| Peter G. Jones, Antonio Laguna,*,‡
M. Reyes Pe´rez,‡ and M. D. Villacampa‡
Qu´ımica Inorga´nica, Facultad de Ciencias, UniVersidad de Valladolid, E-47005 Valladolid, Spain,
Departamento de Qu´ımica Inorga´nica, Instituto de Ciencia de Materiales de Arago´n, UniVersidad
de Zaragoza-CSIC, E-50009 Zaragoza, Spain, Departamento de Qu´ımica e Bioqu´ımica, Faculdade
de Cieˆncias, UniVersidade de Lisboa, Campo Grande 1749-016 Lisboa, Portugal, and ITQB,
AVenida da Repu´blica, EAN, Apartado 127, 2781-901, Oeiras, Portugal, and Institut fu¨r
Anorganische und Analytische Chemie der Technischen UniVersita¨t, Postfach 3329,
D-38023 Braunschweig, Germany
Received July 13, 2005
The gold(I) thiolate complexes [Au(2-SC6H4NH2)(PPh3)] (1), [PPN][Au(2-SC6H4NH2)2] (2) (PPN
and [ Au(2-SC6H4NH2) 2 µ-dppm)] (3) (dppm PPh2CH2PPh2) have been prepared by reaction of acetylacetonato
gold(I) precursors with 2-aminobenzenethiol in the appropriate molar ratio. All products are intensely photoluminescent
at 77 K. The molecular structure of the dinuclear derivative 3 displays a gold gold intramolecular contact of 3.1346-
(4) Å. Further reaction with the organometallic gold(III) complex [Au(C6F5)3(tht)] affords dinuclear or tetranuclear
mixed gold(I) gold(III) derivatives with a thiolate bridge, namely, [(AuPPh3) Au(C6F5)3 2-2-SC6H4NH2)] (4) and
[(C6F5)3Au( 2-2-SC6H4NH2)(AudppmAu)( 2-2-SC6H4NH2)Au(C6F5)3] (5). X-ray diffraction studies of the latter show
a shortening of the intramolecular gold(I) gold(I) contact [2.9353(7) or 2.9332(7) Å for a second independent
molecule], and short gold(I) gold(III) distances of 3.2812(7) and 3.3822(7) Å [or 3.2923(7) and 3.4052(7) Å] are
also displayed. Despite the gold gold interactions, the mixed derivatives are nonemissive compounds. Therefore,
) PPh3dNdPPh3),
{
}
(
)
−
−
{
}(µ
µ
µ
−
−
−
the complexes were studied by DFT methods. The HOMOs and LUMOs for gold(I) derivatives 1 and 3 are mainly
centered on the thiolate and phosphine (or the second thiolate for complex 2), respectively, with some gold
contributions, whereas the LUMO for derivative 4 is more centered on the gold(III) fragment. TD-DFT results show
a good agreement with the experimental UV−vis absorption and excitation spectra. The excitations can be assigned
as a S
2, and a S
with thiolate
f
Au
−
f
f
P charge transfer with some mixture of LLCT for derivative 1, an LLCT mixed with ILCT for derivative
Au‚‚‚Au
Au P) excitation was found for derivative 4. The differing nature of the excited states [participation
−P charge transfer with LLCT and MC for derivative 3. An LMCT (thiolate f
AuIII mixed
−
of the gold(III) fragment and the small contribution of sulfur] is proposed to be responsible for quenching the
luminescence.
Introduction
related to Auronafin, a drug used against rheumatoid
arthritis.1 Moreover, closed-shell gold(I) complexes usually
display gold(I)-gold(I) interactions, which have been at-
tributed to correlation effects reinforced by relativistic effects
and electrostatic contributions.2,3 In contrast, gold(I)-gold-
(III) interactions are rare and have been reported only for
There has long been an interest in the synthesis of
derivatives containing the P-Au(I)-S unit because they are
* To whom correspondence should be addressed. E-mail: bardaji@qi.uva.es
(M.B.), alaguna@unizar.es (A.L.). Fax: + 34 983423013 (M.B.), + 34
976761187 (A.L.).
† Universidad de Valladolid.
‡ Universidad de Zaragoza-CSIC.
(1) Shaw, C. F., III. Gold: Progress in Chemistry, Biochemistry, and
Technology; Schmidbaur, H., Ed.; John Wiley & Sons: Chichester,
U.K., 1999; p 259.
(2) Pyykko¨, P. Chem. ReV. 1997, 97, 597.
(3) Runeberg, N.; Schu¨tz, M.; Werner, H. J. J. Chem. Phys. 1999, 110,
7210.
§ Present and permanent address: Qu´ımica Inorga´nica, Facultad de
Ciencias, Universidad de Valladolid, E-47005 Valladolid, Spain.
| Universidade de Lisboa and ITQB.
Institut fu¨r Anorganische und Analytische Chemie der Technischen
Universita¨t.
10.1021/ic051168u CCC: $33.50
Published on Web 01/12/2006
© 2006 American Chemical Society
Inorganic Chemistry, Vol. 45, No. 3, 2006 1059

Bardaj´ı et al.
some doubly bridged ylide complexes4-7 and sulfide,8,9
thiolate,10 or selenide11 mixed derivatives.
thiolate derivatives are obtained by reaction with an orga-
nometallic gold(III) complex, affording nonemissive deriva-
tives, although the molecular structure of the tetranuclear
complex maintains the short gold(I)-gold(I) distances and
additionally displays short gold(I)-gold(III) contacts. To
understand the photophysical properties in the gold(I) and
gold(I)-gold(III) complexes, DFT calculations were carried
out to determine the frontier orbitals and to assign the
electronic transitions.
Recently, luminescent gold(I) complexes have been re-
ported in which the P-donor ligand is a tertiary mono-, di-,
or triphosphine and the S-donor ligand is a thiolate or
dithiolate.12-20 The emission occurs over a wide range (ca.
400-700 nm) and has usually been assigned to a thiolate-
to-gold charge-transfer excited state, sometimes complicated
by the presence of gold-gold interactions and ascribed to a
thiolate-to-gold-gold charge transfer. Thiolate-to-phosphine
charge-transfer, gold-to-thiolate charge-transfer,21 or even
thiolate-centered transitions have also been suggested.15
Moreover, Yam and co-workers have designed diphosphine-
thiolate gold(I) complexes able to bind specifically to various
metal cations.15,22,23
Experimental Section
General. All reactions were carried out under an argon atmo-
sphere at room temperature. IR spectra were recorded on a Perkin-
Elmer 883 spectrophotometer, over the range 4000-200 cm-1 using
1
Nujol mulls between polyethylene sheets. H, 19F, and 31P{1H}
Zhang and co-workers have modeled several dinuclear
complexes with a dithiolate and a mono- or diphosphine
ligand to study both the luminescent properties and the
aurophilicity of the parent compounds. Their results, based
on correlated methods, revealed that both gold(I)-gold(I)
interactions and the dithiolate ligand are clearly involved in
the emission. They therefore assigned the emission to a
thiolate-to-gold-gold charge-transfer excited state.24,25
In this article, we report the synthesis of mono- and
dinuclear gold(I) thiolate complexes that are intensely
emissive at low temperature in the solid state. The dinuclear
derivative with a bridging diphosphine exhibits a short gold-
gold contact. Di- or tetranuclear mixed gold(I)-gold(III)
NMR spectra were recorded on a Bruker ARX‚300 or GEMINI
2000 apparatus in CDCl3 solution (if no other solvent is stated);
1
chemical shifts are quoted relative to SiMe4 (external, H), CFCl3
(external, 19F) and 85% H3PO4 (external, 31P). C, H, N, and S
analyses were performed with a Perkin-Elmer 2400 microanalyzer.
Mass spectra were recorded on a VG Autospec instrument using
the FAB technique (with a Cs gun) and 3-nitrobenzyl alcohol as
the matrix. Emission and excitation spectra were measured in the
solid state as finely pulverized KBr mixtures at room temperature
and 77 K with a Perkin-Elmer LS-55 spectrofluorometer. UV-vis
absorption spectra in dichloromethane solution were recorded at
298 K on a Shimadzu UV-1603 spectrometer.
Preparation of [Au(2-SC6H4NH2)(PPh3)] (1). To a dichlo-
romethane solution (20 mL) of 2-aminobenzenethiol (29 µL, 0.27
mmol) was added [Au(acac)(PPh3)]26 (150 mg, 0.27 mmol), and
the mixture was stirred for about 1 h. The solution was concentrated
to ca. 2 mL, and the addition of hexane afforded 1 as a white solid.
(4) Mazany, A. M.; Fackler, J. P., Jr. J. Am. Chem. Soc. 1984, 106, 801.
(5) Fackler, J. P., Jr.; Trzcinska-Bancroft, B. Organometallics 1985, 4,
1891.
(6) Schmidbaur, H.; Hartmann, C.; Reber, G.; Mu¨ller, G. Angew. Chem.,
Int. Ed. Engl. 1987, 26, 1146.
(7) Raptis, R. G.; Porter, L. C.; Emrich, R. J.; Murray, H. H.; Fackler, J.
P., Jr. Inorg. Chem. 1990, 29, 4408.
(8) Canales, F.; Gimeno, M. C.; Laguna, A.; Jones, P. G. Organometallics
1996, 15, 3412.
(9) Calhorda, M. J.; Canales, F.; Gimeno, M. C.; Jime´nez, J.; Jones, P.
G.; Laguna, A.; Veiros, L. F. Organometallics 1997, 16, 3837.
(10) Crespo, O.; Canales, F.; Gimeno, M. C.; Jones, P. G.; Laguna, A.
Organometallics 1999, 18, 3142.
(11) Canales, S.; Crespo, O.; Gimeno, M. C.; Jones, P. G.; Laguna, A.;
Mendizabal, F. Organometallics 2001, 20, 4812.
(12) Jones, W. B.; Yuan, J.; Narayanaswamy, R.; Young, M. A.; Elder, R.
C.; Bruce, A. E.; Bruce M. R. M. Inorg. Chem. 1995, 34, 1996.
(13) Forward, J. M.; Bohmann, D.; Fackler, J. P. Jr.; Staples, R. J. Inorg.
Chem. 1995, 34, 6330.
(14) Tang, S. S.; Chang, C.; Lin, I. J. B.; Liou, L.; Wang, J. Inorg. Chem.
1997, 36, 2294.
(15) Yam, V. W. W.; Chan, C. L.; Li, C. K.; Wong, K. M. C. Coord.
Chem. ReV. 2001, 216-217, 173.
(16) Bardaj´ı, M.; Laguna, A.; Vicente, J.; Jones, P. G. Inorg. Chem. 2001,
40, 2675.
(17) Bardaj´ı, M.; Laguna, A.; Pe´rez, M. R.; Jones, P. G. Organometallics
2002, 21, 1877.
(18) Vicente, J.; Gonza´lez-Herrero, P.; Garc´ıa-Sa´nchez, Y.; Jones, P. G.;
Bardaj´ı, M. Inorg. Chem. 2004, 43, 7516.
(19) Bardaj´ı, M.; de la Cruz, M. T.; Jones, P. G.; Laguna, A.; Mart´ınez, J.;
Villacampa, M. D. Inorg. Chim. Acta 2005, 358, 1365.
(20) Chen J. H.; Mohamed A. A.; Abdou H. E.; Bauer J. A. K.; Fackler J.
P.; Bruce A. E.; Bruce M. R. M. Chem. Commun. 2005, 1575.
(21) Watase, S.; Nakamoto, M.; Kitamura, T.; Kanehisa, N.; Kai, Y.;
Yanagida, S. J. Chem. Soc., Dalton Trans. 2000, 3585.
(22) Yam, V. W. W.; Li, C. K.; Chan, C. L. Angew. Chem., Int. Ed. 1998,
37, 2857.
1
Yield of 1: 130 mg, 83%. H NMR: δ 4.40 (s, 2H, NH2), 6.59
(td, 1H, JHH ) 7.4 and 1.3 Hz, C6H4), 6.70 (dd, 1H, JHH ) 7.7 and
1.3 Hz, C6H4), 6.90 (td, 1H, JHH ) 7.5 and 1.3 Hz, C6H4), 7.4-7.5
(m, 15H, Ph), 7.61 (dd, 1H, JHH ) 7.7 and 1.5 Hz, C6H4). 31P{1H}
NMR: δ 38.6 (s). IR: 3442 [m, ν(N-H)], 3340 [m, ν(N-H)],
1602 [s, δ(N-H)] cm-1. Anal. Calcd for C24H21AuNPS: C, 49.4;
H, 3.65; N, 2.4; S, 5.5. Found: C, 49.7; H, 3.8; N, 2.45; S, 5.2.
FAB+ (m/z, %, assignment): 459 (100, [AuPPh3]+), 583 (20, [M]+),
1042 (47 [M + AuPPh3]+).
Preparation of [PPN][Au(2-SC6H4NH2)2] (2). To 20 mL of a
dichloromethane solution of 2-aminobenzenethiol (22 µL, 0.2
mmol) was added [PPN][Au(acac)2]27 (93 mg, 1 mmol), and the
mixture was stirred for about 2 h. The solvent was removed, and
the white residue was washed with petroleum ether. Yield of 2:
1
69 mg, 70%. H NMR: δ 4.33 (s, 4H, NH2), 6.39 (t, 2H, JHH
)
7.5 Hz, C6H4), 6.50 (d, 2H, JHH ) 7.8 Hz, C6H4), 6.68 (t, 2H, JHH
) 7.2 Hz, C6H4), 7.4-7.6 (m, 32H, Ph + C6H4). 31P{1H} NMR:
δ 21.8 (s, PPh3dNdPPh3). IR: 3446 [w, ν(N-H)], 3319 [w, ν-
(N-H)], 1596 [s, δ(N-H)] cm-1. Anal. Calcd for C48H42-
AuN3P2S2: C, 58.6; H, 4.3; N, 4.25; S, 6.5. Found: C, 58.3; H,
4.2; N, 4.35; S, 6.2. FAB- (m/z, %, assignment): 445 (100,
[M-PPN]-).
Preparation of [{Au(2-SC6H4NH2)}2(µ-dppm)] (3). To a
dichloromethane solution (20 mL) of 2-aminobenzenethiol (11 µL,
0.1 mmol) was added [{Au(acac)}2(µ-dppm)] (98 mg, 0.1 mmol;
prepared as for PPh3 in ref 26), and the mixture was stirred for
(23) Li, C. K.; Lu, X. X.; Wong, K. M. C.; Chan, C. L.; Zhu, N.; Yam, V.
W. W. Inorg. Chem. 2004, 43, 7421.
(24) Pan, Q. J.; Zhang, H. X. Eur. J. Inorg. Chem. 2003, 23, 4202.
(25) Pan, Q. J.; Zhang, H. X. Organometallics 2004, 23, 5198.
(26) Vicente, J.; Chicote, M. T. Inorg. Synth. 1998, 32, 175.
(27) Vicente, J.; Chicote, M. T.; Saura-Llamas, I.; Lagunas, M. C. Chem.
Commun. 1992, 915.
1060 Inorganic Chemistry, Vol. 45, No. 3, 2006

Quenching of Gold(I) Thiolate Luminescence
Table 1. Details of Crystal Data and Structure Refinement for Complexes 3 and 5
3
5
empirical formula
formula weight
temperature
C37H34Au2N2P2S2
1026.66
100(2) K
C73H34Au4F30N2P2S2
2422.95
133(2) K
wavelength
0.71073 Å
0.71073 Å
crystal system
space group
monoclinic
P21/c
orthorhombic
Pca21
unit cell dimensions
a ) 11.5662(9) Å
b ) 11.2001(8) Å
c ) 26.757(2) Å
R ) 90°
a ) 22.864(2) Å
b ) 30.371(2) Å
c ) 24.697(2) Å
R ) 90°
â ) 91.870(1)°
γ ) 90°
â ) 90°
γ ) 90°
volume
3464.3(4) Å3
17150(2) Å3
Z
4
8
density (calcd)
absorption coefficient
F(000)
1.968 Mg/m3
1.877 Mg/m3
8.703 mm-1
1960
7.012 mm-1
8976
crystal habit
crystal size
θ range for data collection
index ranges
colorless needle
0.2 × 0.08 × 0.08 mm
1.76-28.78°
-12 e h e 15, -15 e k e 14,
-35 e l e 33
22494
8133 [R(int) ) 0.0870]
SADABS
1.000 and 0.565
full-matrix least-squares on F2
8133/0/406
colorless tablet
0.31 × 0.22 × 0.08 mm
0.67-28.30°
-30 e h e 30, -40 e k e 40,
-32 e l e 32
275207
42596 [R(int) ) 0.0953]
SADABS
0.746 and 0.483
full-matrix least-squares on F2
42596/337/1318
1.057
reflections collected
independent reflections
absorption correction
max and min transmission
refinement method
data/restraints/parameters
GOF on F2
0.880
final R indices [I > 2σ(I)]
R indices (all data)
largest diff peak and hole
R1 ) 0.0446, wR2 ) 0.0812
R1 ) 0.0675, wR2 ) 0.0873
2.324 and -1.421 e.Å-3
R1 ) 0.0446, wR2 ) 0.1134
R1 ) 0.1001, wR2 ) 0.1390
3.067 and -1.732 e. Å-3
about 1 h. The solution was concentrated to ca. 2 mL, and the
) 7.6 Hz, C6H4), 6.74 (t, 2H, JHH ) 7.3 Hz, C6H4), 7.5-7.7 (m,
20H, Ph). 19F NMR δ -121.05 (m, 4Fo), -122.93 (m, 2Fo),
addition of hexane afforded 3 as a yellow solid. Yield of 3: 85
1
2
3
3
mg, 83%. H NMR: δ 3.69 (t, 2H, J ) 11.2 Hz, CH2P), 4.50 (s,
4H, NH2), 6.55 (t, 2H, JHH ) 7.3 Hz, C6H4), 6.68 (d, 2H, JHH
-158.73 (t, JFF ) 20.1 Hz, 1Fp), -158.93 (t, JFF ) 20.1 Hz,
2Fp), -162.27 (m, 4Fm), -162.87 (m, 2Fm). 31P{1H} NMR: δ 33.8
(s). IR: 3460 [w, ν(N-H)], 3360 [w, ν(N-H)], 1620 [m, δ(N-
H)], 970 (s), 807 (s), 797 (s) cm-1 from the C6F5 group. Anal.
Calcd for C73H34Au4F30N2P2S2: C, 36.2; H, 1.4; N, 1.15; S, 2.65.
Found: C, 35.9; H, 1.45; N, 1.1; S, 2.3. FAB+ (m/z, %,
assignment): 902 (100, [M - 2(Au(C6F5)3) - SC6H4NH2]+), 1432
(22, [M - Au(C6F5)3 - C6F5 - H]+), 1600 (19, [M - Au(C6F5)3
- SC6H4NH2]+).
Crystal Structure Determination of Derivatives 3 and 5. The
crystals were mounted in inert oil on glass fibers and transferred
to the cold gas stream of a Bruker SMART CCD diffractometer.
Crystal data and details of data collection and structure refinement
are given in Table 1. Absorption corrections were based on multiple
scans (program SADABS). The structures were refined anisotro-
pically on F2 using the program SHELXL-97.29 For complex 3, all
non-hydrogen atoms were refined anisotropically, and hydrogen
atoms were included using a riding model. Special details of
refinement for 5: The structure is pseudosymmetric30 (the maximum
deviations from the corresponding centrosymmetric space group
are only ca. 0.7 Å) and was refined as an enantiomorphic twin.
The twinning parameter refined to 0.378(7). Note that a value of
0.5 would be expected if the structure were actually centrosym-
metric. The C and N atoms were refined isotropically. Several
significant difference peaks were observed, presumably correspond-
ing to solvent of crystallization, but could not be connected to form
any sensible molecule. Arbitrarily, two were refined as chlorine
and five as carbon. None of these atoms was taken into account in
)
8.2 Hz, C6H4), 6.92 (t, 2H, JHH ) 7.3 Hz, C6H4), 7.2-7.6 (m, 22H,
Ph + C6H4). 31P{1H} NMR: δ 29.2 (s). IR: 3400 [m, br, ν(N-
H)], 1605 [s, δ(N-H)] cm-1. Anal. Calcd for C37H34Au2N2P2S2:
C, 43.3; H, 3.35; N, 2.75; S, 6.25. Found: C, 43.4; H, 3.35; N,
2.45; S, 6.1. FAB+ (m/z, %, assignment): 902 (100, [M - SC6H4-
NH2]+), 1223 (37, [M + Au]+).
Preparation of [(AuPPh3){Au(C6F5)3}(µ2-2-SC6H4NH2)] (4)
and [(C6F5)3Au(µ2-2-SC6H4NH2)(AudppmAu)(µ2-2-SC6H4NH2)-
Au(C6F5)3] (5). To a dichloromethane solution (20 mL) of 1 (58
mg, 0.1 mmol) or 3 (51 mg, 0.05 mmol) was added [Au(C6F5)3-
(tht)]28 (tht ) tetrahydrothiophene; 79 mg, 0.1 mmol). After being
stirred for 2 h, the solution was concentrated to ca. 3 mL. Addition
of hexane afforded complex 4 or 5, respectively, as a white (4) or
1
yellow (5) solid. Yield of 4: 96 mg, 75%. H NMR: δ 4.23 (s,
2H, NH2), 6.42 (t, 1H, JHH ) 7.5 Hz, C6H4), 6.49 (d, 1H, JHH
)
8.0 Hz, C6H4), 6.93 (t, 1H, JHH ) 7.6 Hz, C6H4), 7.36 (d, 1H, JHH
) 7.8 Hz, C6H4), 7.5-7.6 (m, 15H, Ph). 19F NMR δ -120.72 (m,
4Fo), -122.83 (m, 2Fo), -158.85 (t, 3JFF ) 20.0 Hz, 1Fp), -159.20
3
(t, JFF ) 20.1 Hz, 2Fp), -162.46 (m, 4Fm), -162.93 (m, 2Fm).
31P{1H} NMR: δ 37.1 (s). IR: 3477 [w, ν(N-H)], 3390 [w, ν-
(N-H)], 1638 [w, δ(N-H)], 1615 [s, δ(N-H)], 967 (s), 805 (s),
794 (s) cm-1 from the C6F5 group. Anal. Calcd for C42H21Au2F15-
NPS: C, 39.35; H, 1.65; N, 1.1; S, 2.5. Found: C, 39.7; H, 1.95;
N, 1.15; S, 2.2. FAB+ (m/z, %, assignment): 459 (100, [Au-
(PPh3)]+), 583 (85, [M - Au(C6F5)3]+), 1114 (27, [M - C6F5]+),
1
1281 (11, [M]+). Yield of 5: 103 mg, 85%. H NMR: δ 3.74 (t,
2
2H, J ) 11.2 Hz, CH2P), 4.09 (s, 4H, NH2), 5.63 (t, 2H, JHH
)
7.2 Hz, C6H4), 6.46 (d, 2H, JHH ) 8.0 Hz, C6H4), 6.51 (d, 2H, JHH
(29) Sheldrick, G. M. SHELXL 97: A Program for Crystal Structure
Refinement; University of Go¨ttingen: Go¨ttingen, Germany, 1997.
(28) Uso´n, R.; Laguna, A.; Laguna, M. Inorg. Synth. 1989, 26, 87.
Inorganic Chemistry, Vol. 45, No. 3, 2006 1061

Bardaj´ı et al.
Scheme 1a
calculating the molecular weight and other composition-related
parameters. The aminic hydrogens were not located.
DFT Calculations. Density functional theory (DFT) calcula-
tions31 were carried out with the Amsterdam Density Functional
(ADF 2004.01) program.32 The Vosko, Wilk, and Nusair local
exchange correlation potential was used.33 Gradient-corrected
geometry optimizations34 were performed, using the generalized
gradient approximation (Perdew-Wang nonlocal exchange and
correlation corrections, PW9135). Relativistic effects were treated
with the ZORA approximation.36 The core orbitals were frozen for
Au [(1-4)s, (2-4)p, (3-4)d]; P; S [(1-2)s, 2p]; and C, N, and F
(1s). Triple-ú Slater-type orbitals (STOs) were used to describe the
valence shells of C, O (2s and 2p), P (3s, 3p), and Au (4f, 5d, 6s).
A set of two polarization functions was added to each C, O (single-
ú, 3d, 4f), P, S (single-ú, 3d, 4p), and Au (single-ú, 6p, 5f). Triple-ú
Slater-type orbitals (STOs) were used to describe the valence shells
of H (1s) with one polarization function (single-ú, 2p). Full
geometry optimizations were performed without any symmetry
constraints on complexes based on the crystal structures described
below. Time-dependent DFT calculations (TD-DFT)37 in the ADF
implementation were used to determine the excitation energies.
Mayer indices38 were calculated with the ADF densities using the
MAYER program39 and were used as bond strength indicators.
Three-dimensional representations of orbitals were obtained with
Molekel.40
a (i) [Au(acac)(PPh3)], (ii) [PPN][Au(acac)2], (iii) [{Au(acac)}2(µ-dppm)],
(iv) [Au(C6F5)3(tht)], (v) 2[Au(C6F5)3(tht)].
rivatives of gold(I), which are able to deprotonate the thiol.
The resulting thiolate coordinates to the corresponding gold-
(I) fragment, as a result of the loss of the protonated acac
ligand; this is known as the “acac method”.41 The corre-
sponding gold thiolate derivatives 1-3 were regioselectively
obtained, according to Scheme 1. Schmidbaur and co-
workers reported the regioselective double auration to afford
the sulfonium complex [(2-H2NC6H4)S(AuPPh3)2]BF4.42
These derivatives are air- and moisture-stable white (1 and
2) or yellow (3) solids at room temperature. They were
readily characterized by elemental analysis and mass, IR,
and NMR spectroscopies. The addition of an excess of [Au-
(acac)(PPh3)] does not produce further reaction, and the
monosubstituted derivative is always obtained. More gold
fragments can be introduced into such derivatives using the
N-donor atom, as found in the mononuclear gold(III)
derivative [Au(κC1,κN,2-NMe2CH2C6H4)(κN,κS,2-SC6H4-
NH2)](ClO4),43 by forming unsupported gold-gold interac-
tions, as found for [{(pta)3AuI}2AuI2(i-mnt)2] (i-mnt )
isomalononitriledithiolate, pta ) phosphatriazaadamantane)44
and [{(C6F5)3AuIII}AuI2(CH2PPh2CH2)2],45 or, more typically,
the S-donor atom to give derivatives as reported for other
thiolato derivatives such as [RS(AuL)n](n-1)+ (L ) phosphine,
n ) 1-3)46 and [(C6F5)3Au(µ2-SC6F5)(AudppfAu)(µ2-SC6F5)-
Au(C6F5)3] [dppf ) 1,1′-bis(diphenylphosphino)ferrocene]10
and sulfido derivatives such as [S(AuL)n](n-2)+ (L ) phos-
phine, diphosphine; n ) 2-6)47,48 and [S(Au2L2){Au-
(C6F5)3}n] (n ) 2, L ) PPh3; n ) 1, 2, L2 ) dppf).8,9
Results and Discussion
Synthesis and Characterization. We have carried out
reactions of 2-aminobenzenethiol with acetylacetonato de-
(30) We considered refining the structure of 5 in the centrosymmetric space
group Pbcn. We were, of course, aware of the problem with space
group ambiguity. We found the following. (i) In Pbcn, there are 987
violations of the systematic absences for the n glide plane. (ii) The
structure can indeed be solved in Pbcn, but the refinement stalls at
wR2 ) 50%, with a large number of difference peaks of 2-4 e Å-3
.
(iii) The basic routine ADDSYMM of the program PLATON (A.L.
Spek, University of Untrecht, Netherlands) indeed suggests the higher
symmetry (as we had already ascertained) but its EXACT counterpart
does not, confirming the significant departure from centrosymmetry.
Other symmetry-checking programs come to the same conclusion. We
accept that a disordered centrosymmetric model might, in general, be
better than an ordered noncentrosymmetric model but nevertheless
prefer the latter in this case. Clearly, final proof becomes more difficult
as the deviation from centrosymmetry becomes smaller (and least-
squares matrices tend to singularity).
(31) Parr, R. G.; Yang, W. Density Functional Theory of Atoms and
Molecules; Oxford University Press: New York, 1989.
(32) (a) ADF2004.01; SCM, Theoretical Chemistry, Vrije Universiteit:
Amsterdam, The Netherlands; available at http://www.scm.com. (b)
Te Velde, G.; Bickelhaupt, F. M.; van Gisbergen, S. J. A.; Fonseca
Guerra, C.; Baerends, E. J.; Snijders, J. G.; Ziegler, T. J. Comput.
Chem. 2001, 22, 931. (c) Fonseca Guerra, C.; Snijders, J. G.; Te Velde,
G.; Baerends, E. J. Theor. Chem. Acc. 1998, 99, 391.
(33) Vosko, S. H.; Wilk, L.; Nusair, M. Can. J. Phys. 1980, 58, 1200.
(34) (a) Fan, L.; Ziegler, T.; J. Chem. Phys. 1991, 95, 7401. (b) Versluis,
L.; Ziegler, T. J. Chem. Phys. 1988, 88, 322.
(35) Perdew, J. P.; Chevary, J. A.; Vosko, S. H.; Jackson, K. A.; Pederson,
M. R.; Singh, D. J.; Fiolhais, C. Phys. ReV. 1992, B46, 6671.
(36) van Lenthe, E.; Ehlers, A.; Baerends, E. J. J. Chem. Phys. 1999, 110,
8943.
(37) (a) van Gisbergen, S. J. A.; Groeneveld, J. A.; Rosa, A.; Snijders, J.
G.; Baerends, E. J.; J. Phys. Chem. A 1999, 103, 6835. (b) Rosa, A.;
Baerends, E. J.; van Gisbergen, S. J. A.; van Lenthe, E.; Groeneveld,
J. A.; Snijders, J. G.; J. Am. Chem. Soc. 1999, 121, 10356. (c) van
Gisbergen, S. J. A.; Rosa, A.; Ricciardi, G.; Baerends, E. J. J. Chem.
Phys. 1999, 111, 2499.
(41) Vicente, J.; Chicote, M. T. Coord. Chem. ReV. 1999, 193-195, 1143.
(42) Lo´pez-de-Luzuriaga, J. M.; Sladek, A.; Schneider, W.; Schmidbaur,
H. Chem. Ber. 1997, 130, 641.
(43) Vicente, J.; Chicote, M. T.; Bermudez, M. D.; Jones, P. G.; Fittschen,
C.; Sheldrick, G. M. J. Chem. Soc., Dalton Trans. 1986, 2361.
(44) Fackler, J. P.; Staples, R. J.; Assefa, Z. Chem. Commun. 1994, 431.
(45) Uso´n, R.; Laguna, A.; Laguna, M.; Tarton, M. T.; Jones, P. G. Chem.
Commun. 1988, 740.
(46) Sladek, A.; Angermaier, K.; Schmidbaur, H. Chem. Commun. 1996,
1959.
(47) Canales, F.; Gimeno, M. C.; Jones, P. G.; Laguna, A. Angew. Chem.,
Int. Ed. Engl. 1994, 33, 769.
(38) (a) I. Mayer. Chem. Phys. Lett. 1983, 97, 270. (b) Mayer, I. Int. J.
Quantum Chem. 1984, 26, 151.
(39) Bridgeman, A. J.; Empson, C. J. MAYER, version 1.2.3; The University
of Hull: Hull, U.K., 2004.
(40) Portmann, S.; Lu¨thi, H. P. Chimia 2000, 54, 766.
1062 Inorganic Chemistry, Vol. 45, No. 3, 2006

Quenching of Gold(I) Thiolate Luminescence
Table 3. Hydrogen Bonds for Complex 3 (Å and deg)
D-H‚‚‚Aa
d(D-H) d(H‚‚‚A) d(D‚‚‚A) <(DHA)
N(1)-H(1B)‚‚‚S(2) #1
N(2)-H(2B)‚‚‚S(1) #2
0.89
0.89
2.67
2.65
3.507(7)
3.537(7)
158.3
178.8
a Symmetry transformations used to generate equivalent atoms: #1 -x,
-y + 2, -z; #2 -x, -y + 1, -z.
Figure 1. Structure of 3, showing displacement ellipsoids at the 50%
probability level and the atom-numbering scheme. H atoms have been
omitted for clarity.
Table 2. Selected Bond Lengths (Å) and Angles (deg) for Complex 3
Figure 2. Structure of one of the two independent molecules of complex
5 in the crystal. H and F atoms have been omitted for clarity. Radii are
arbitrary.
Au(1)-P(1)
Au(1)-S(1)
Au(1)-Au(2)
Au(2)-P(2)
Au(2)-S(2)
2.2519(17)
2.2967(17)
3.1346(4)
2.2496(18)
2.2986(19)
S(1)-C(51)
S(2)-C(61)
N(1)-C(56)
N(2)-C(62)
1.767(7)
1.773(8)
1.369(9)
1.374(9)
cases, e.g., [{Au(Fmes)}2(µ-dppm)] (Fmes ) 2,4,6-tris-
(trifluoromethyl)phenyl].50 The Au(1)-Au(2) distance is
shorter than that found in the related compound [{Au(2-
SPym-4-NH2)}2(µ-dppm)]51 [3.3317(2) Å; 2-SPym-4-NH2 )
4-amino-2-pyrimidinethiol], but longer than those in [{Au-
(SSNH2)}2(µ-dppm)]52 [3.0987(10) Å; SSNH2- ) 2-amino-
5-mercapto-1,3,4-thiadiazolate] and [{Au(dpt)}2(µ-dppm)]53
[3.0353(3) Å; dpt ) S2P(p-C6H4OMe)(O-c-C5H9)], all of
which contain a single dppm bridge. There are no short
gold-gold intermolecular contacts.
The Au-P [2.2496(18), 2.2519(17) Å] and Au-S [2.2967-
(17), 2.2986(19) Å] bond distances are typical for this type
of complex. Complex 3 also shows two N-H‚‚‚S contacts
that can be interpreted as hydrogen bonds (Table 3).
(ii) Crystal Structure of Complex 5. The tetranuclear
structure of 5 was confirmed by X-ray diffraction studies; it
crystallizes with two independent molecules and disordered
solvent in the asymmetric unit. One molecule is shown in
Figure 2, and a selection of bond lengths and angles is
collected in Table 4.
As expected, the coordination geometries of the AuI and
AuIII centers are linear and square-planar, respectively, with
only slight distortions. After complexation of the gold(III)
fragments, there is a shortening of the gold(I)-gold(I)
distance to 2.9353(7) Å [2.9332(7) Å in the second mol-
ecule], which has not been observed in other similar mixed
complexes. To our knowledge, this AuI-AuI bond length is
the shortest reported for a monobridged dinuclear dppm gold
complex. The Au(2)-P(1)‚‚‚P(2)-Au(3) torsion angle of
-35.4(1)° [34.0(1)°], therefore shows the usual cis confor-
P(1)-Au(1)-S(1)
P(1)-Au(1)-Au(2)
S(1)-Au(1)-Au(2)
P(2)-Au(2)-S(2)
179.18(7)
82.21(5)
97.16(5)
172.72(7)
P(2)-Au(2)-Au(1)
S(2)-Au(2)-Au(1)
C(51)-S(1)-Au(1)
C(61)-S(2)-Au(2)
91.63(5)
95.37(5)
105.1(2)
100.7(2)
Therefore, we carried out reactions with a gold(III) derivative
that led to mixed di- or tetranuclear gold(I)-gold(III)
derivatives 4 and 5 according to Scheme 1.
Complexes 4 and 5 are air- and moisture-stable white (4)
or yellow (5) solids at room temperature. They were readily
characterized by elemental analysis and mass, IR, and NMR
spectroscopies. The main features are as follows: (a) the IR
spectra show absorptions around 970, 805, and 795 cm-1
due to the pentafluorophenyl groups;49 (b) the 19F NMR
spectra show the presence of one type of tris(pentafluo-
rophenyl)gold(III) unit; and (c) a singlet is observed in the
phosphorus NMR spectra. Features b and c point to a
symmetric structure for complex 5.
X-ray Crystal Structure Determination of Derivatives
3 and 5. (i) Crystal Structure of Complex 3. The structure
of the complex 3 was established by X-ray diffraction and
is shown in Figure 1, with selected bond lengths and angles
reported in Table 2. The angles P(1)-Au(1)-S(1) [179.18-
(7)°] and P(2)-Au(2)-S(2) [172.72(7)°] indicate a linear
geometry around both gold atoms, although slightly distorted
for Au(2). The intramolecular Au-Au distance, 3.1346(4)
Å, indicates metal-metal interaction; the observed Au(1)-
P(1)‚‚‚P(2)-Au(2) torsion angle of -24.1° shows that the
two P-Au-S moieties are slightly twisted with respect to
each other; this cis conformation is usually preferred to the
trans conformation, which has been observed in only a few
(50) Bardaj´ı, M.; Jones, P. G.; Laguna, A.; Moracho, A.; Fischer, A. K. J.
Organomet. Chem. 2002, 648, 1.
(51) Wilton-Ely, J. D. E. T.; Schier, A.; Mitzel, N. W.; Nogai, S.;
Schmidbaur, H. J. Organomet. Chem. 2002, 643-644, 313.
(52) Tzeng, B.; Schier, A.; Schmidbaur, H. Inorg. Chem. 1999, 38, 3978.
(53) Maspero, A.; Kani, I.; Mohamed, A. A.; Omary, M. A.; Staples, R.
J.; Fackler, J. P., Jr. Inorg. Chem. 2003, 42, 5311.
(48) Canales, F.; Gimeno, M. C.; Laguna, A.; Jones, P. G. J. Am. Chem.
Soc. 1996, 118, 4839.
(49) Uso´n, R.; Laguna, A.; Garc´ıa, J.; Laguna, M. Inorg. Chim. Acta 1979,
37, 201.
Inorganic Chemistry, Vol. 45, No. 3, 2006 1063

Bardaj´ı et al.
Table 4. Selected Bond Lengths (Å) and Angles (deg) for Complex 5a
Table 5. Electronic Absorption Data for the Thiol and Complexes
1-5a
Au(1)-C(31)
Au(1)-C(11)
Au(1)-C(21)
Au(1)-S(1)
Au(1)-Au(2)
Au(2)-P(1)
Au(2)-S(1)
Au(2)-Au(3)
Au(3)-P(2)
Au(3)-S(2)
Au(3)-Au(4)
Au(4)-C(111)
Au(4)-C(101)
Au(4)-C(121)
Au(4)-S(2)
2.001(14)
2.050(13)
2.071(13)
2.375(3)
3.2812(7)
2.269(3)
2.341(3)
2.9353(7)
2.270(3)
2.332(3)
3.3822(7)
2.017(14)
2.039(15)
2.105(15)
2.372(3)
Au(1′)-C(131)
Au(1′)-C(151)
Au(1′)-C(141)
Au(1′)-S(1′)
Au(1′)-Au(2′)
Au(2′)-P(1′)
Au(2′)-S(1′)
Au(2′)-Au(3′)
Au(3′)-P(2′)
Au(3′)-S(2′)
Au(3′)-Au(4′)
Au(4′)-C(231)
Au(4′)-C(221)
Au(4′)-C(241)
Au(4′)-S(2′)
2.036(13)
2.064(13)
2.091(15)
2.372(3)
3.2923(7)
2.268(3)
2.342(3)
2.9332(7)
2.271(3)
2.337(3)
3.4052(7)
2.036(14)
2.044(14)
2.081(15)
2.375(3)
compd
λ/nm (ꢀ × 10-3/M-1 cm-1
)
HSC6H4NH2
247 sh (6.5), 300 (2.3), 349 sh (0.3)
1
2
3
4
5
245 sh (32), 269 sh (11), 277 sh (8.2), 311 (6.2)
268 (34), 274 (30), 319 (21)
266 sh (16), 277 sh (13), 308 sh (5.5)
244 sh (33), 269 (18), 276 (17), 345 (5.2)
268 sh (41), 277 (42), 342 (9.4)
a In CH2Cl2 solution, 5.0 × 10-5 Μ.
C(31)-Au(1)-C(11)
C(31)-Au(1)-C(21)
C(11)-Au(1)-C(21)
C(31)-Au(1)-S(1)
C(11)-Au(1)-S(1)
C(21)-Au(1)-S(1)
C(31)-Au(1)-Au(2)
C(11)-Au(1)-Au(2)
C(21)-Au(1)-Au(2)
S(1)-Au(1)-Au(2)
P(1)-Au(2)-S(1)
P(1)-Au(2)-Au(3)
S(1)-Au(2)-Au(3)
P(1)-Au(2)-Au(1)
S(1)-Au(2)-Au(1)
Au(3)-Au(2)-Au(1)
P(2)-Au(3)-S(2)
91.2(4)
177.6(4)
90.1(4)
93.2(3)
175.0(3)
85.6(3)
102.5(3)
131.0(3)
78.1(3)
C(131)-Au(1′)-C(151)
C(131)-Au(1′)-C(141)
89.8(4)
91.4(4)
C(151)-Au(1′)-C(141) 177.0(4)
C(131)-Au(1′)-S(1′)
C(151)-Au(1′)-S(1′)
C(141)-Au(1′)-S(1′)
174.0(3)
85.3(3)
93.7(3)
C(131)-Au(1′)-Au(2′) 130.2(3)
C(151)-Au(1′)-Au(2′) 78.2(3)
C(141)-Au(1′)-Au(2′) 103.0(3)
45.48(7)
S(1′)-Au(1′)-Au(2′)
45.33(7)
174.28(14)
87.24(9)
93.53(8)
133.34(9)
46.09(8)
139.41(2)
177.66(13)
89.56(9)
92.40(8)
91.9(4)
174.47(13) P(1′)-Au(2′)-S(1′)
87.89(9)
91.92(8)
133.79(10) P(1′)-Au(2′)-Au(1′)
46.33(8)
138.22(2)
176.23(13) P(2′)-Au(3′)-S(2′)
88.00(9)
94.30(8)
91.6(5)
P(1′)-Au(2′)-Au(3′)
S(1′)-Au(2′)-Au(3′)
Figure 3. UV-vis absorption spectra of complexes 1 and 3-5 (ꢀ in M-1
cm-1 versus wavelength in nm).
S(1′)-Au(2′)-Au(1′)
Au(3′)-Au(2′)-Au(1′)
spectra of complexes 1 and 3-5 are shown in Figure 3. The
absorptions below 300 nm are related to the pentafluorophe-
nyl, phosphine, or thiol ligands; the addition of the gold(III)
fragments to complexes 1 or 3 increases the intensity of the
absorptions around 270 and 280 nm. The lowest-energy band,
absent in the chloro analogues [AuCl(PPh3)] and [(AuCl)2-
(µ-dppm)], has usually been related to sulfur-to-gold charge-
transfer transitions and shifts from ca. 310 to ca. 340 nm
after coordination of gold(III) to sulfur.18,23,54 Accordingly,
these peaks are tentatively assigned to ligand-to-metal [gold-
(III) and/or gold(I)] charge-transfer transitions, although the
thiol ligand shows a weak absorption in this region. In
addition, there are absorption tails that extend to ca. 360 nm
(derivatives 1 and 2), ca. 380 nm (derivative 3), ca. 410 nm
(derivative 4), and ca. 425 nm (derivative 5). Further
discussion of these assignments is provided in conjunction
with the discussion of the computational results below.
The solid-state emission and excitation spectra at 77 K
for the isolated complexes were obtained and are shown in
Figure 4. The compounds do not emit at room temperature.
In this case, the most striking feature is the intense photo-
luminescence at 77 K, with emission maxima at 452 (1),
465 (2), and 538 (3) nm for the gold(I) thiolate complexes,
whereas the mixed gold(I)-gold(III) thiolate derivatives are
not emissive. The excitation spectra are complex, with
maxima at 330 (1), 353 (2), and 391 (3) nm. Compound 3
shows the broadest emission peak, and a second, less intense
emission peak centered at 514 nm can be obtained, with an
excitation maximum at 359 nm.
P(2)-Au(3)-Au(2)
S(2)-Au(3)-Au(2)
C(111)-Au(4)-C(101)
P(2′)-Au(3′)-Au(2′)
S(2′)-Au(3′)-Au(2′)
C(231)-Au(4′)-C(221)
C(111)-Au(4)-C(121) 176.4(4)
C(231)-Au(4′)-C(241) 177.6(4)
C(101)-Au(4)-C(121)
C(111)-Au(4)-S(2)
C(101)-Au(4)-S(2)
C(121)-Au(4)-S(2)
Au(2)-S(1)-Au(1)
Au(3)-S(2)-Au(4)
90.8(5)
93.1(3)
175.0(4)
84.6(3)
C(221)-Au(4′)-C(241)
C(231)-Au(4′)-S(2′)
C(221)-Au(4′)-S(2′)
C(241)-Au(4′)-S(2′)
89.6(5)
84.9(3)
176.4(3)
93.7(3)
88.58(10)
92.53(11)
88.19(10) Au(2′)-S(1′)-Au(1′)
91.93(11) Au(3′)-S(2′)-Au(4′)
a Primes indicate atoms of the second independent molecule.
mation. The AuI-AuIII distances, 3.2812(7) and 3.3822(7)
Å [3.2923(7) and 3.4052(7) Å], are comparable to those
obtained for the complex [{S(Au2dppf)}2{Au(C6F5)2}][CF3-
SO3] [3.2195(8), 3.3661(10) Å]9 with a bridging sulfide
ligand, but shorter than those reported in other sulfide,
thiolate, or selenide mixed derivatives [3.404(1) - 4.011(1)
Å].8-11 The shortest of these AuI-AuIII distances are similar
to the AuI-AuI contacts in many molecules. The Au-S-
Au angles differ slightly, with the narrower, 88.19(10)° [Au-
(1)-S(1)-Au(2)] [88.58(10)°], corresponding to the shortest
AuI-AuIII distance.
A lengthening of the AuI-S distances should be expected
as a consequence of the coordination of the Au(C6F5)3
moieties; in fact, the AuI-S distances in 5 are 2.332(3) and
2.341(3) Å [2.337(3) and 2.342(3) Å], longer than the
corresponding distances in 3, 2.2967(17) and 2.2986(19) Å.
The AuIII-S distances of 2.372(3) and 2.375(3) Å [2.372-
(3) and 2.375(3) Å] compare well with those found in the
mixed AuI-AuIII derivative with a bridging thiolate ligand
[(C6F5)3Au(µ2-SC6F5)(AudppfAu)(µ2-SC6F5)Au(C6F5)3].10 The
Au-P lengths are similar to those observed in 3.
The emission observed in gold(I) thiolate complexes has
usually been assigned as arising from a S (π)-to-Au (6p)
LMCT excited state and is therefore affected by the substit-
uents of the thiolate ligand.12,54 When substituents on the
thiolate do not produce significant electronic perturbations,
Photophysical Studies. The absorption spectra of these
complexes and the ligands were measured in the range 200-
600 nm, and the results are summarized in Table 5. The
1064 Inorganic Chemistry, Vol. 45, No. 3, 2006

Quenching of Gold(I) Thiolate Luminescence
Table 6. Relevant Calculated Distances (Å) and Angles (deg) for
Models of Complexes 1-3, Compared with Experimental Values for 3
distance/angle
3a
3h
3m
2
1
1h
1m
Au‚‚‚Au
Au-S
3.135 3.142 3.197
2.297 2.418 2.317 2.315 2.319 2.314 2.320
2.299 2.428 2.319 2.330
-
-
-
-
Au-P
2.250 2.424 2.271
2.253 2.431 2.277
-
2.298 2.265 2.278
Au‚‚‚N
3.497 3.597 3.609 3.258 3.601 3.537 3.530
4.792 4.872 4.836 3.865
P-C (dppm)
S-Au-P
S-Au-S
1.841 1.858 1.855
-
-
-
-
1.846
173
179
-
1.856
172
179
-
172
178
-
-
175
178
177
174
-
-
-
-
-
-
-
Au-P‚‚‚P-Au 24.1
27.0
25.9
a X-ray structure.
selected ligands are used.15,56 Finally, multiple state emissions
(MC, LMCT, and LMMCT) have been proposed to explain
the white emission and the thermochromism of dialkyldithio-
phosphonate dinuclear derivatives.57
X-ray diffraction studies showed that the nonemissive
tetranuclear gold(I)-gold(III) thiolate complex 5 exhibits a
shorter gold(I)-gold(I) distance than the strongly emissive
parent dinuclear gold(I) thiolate complex 3. A purely metal-
metal-centered transition should thus be ruled out. The
spectrum of complex 3 is red-shifted compared to that of
complex 1, a typical behavior induced by the gold(I)-gold-
(I) interactions, which points to an LMCT transition (or an
LMMCT transition for the emission of complex 3). The same
assignment could explain why complexes 4 and 5 are
nonemissive: After coordination of the gold(III) fragment,
the latter transition becomes less likely, and other transitions
involving the gold(III) center could take place. TD-DFT
calculations should help to clarify the nature of these
transitions.
DFT Calculations. DFT calculations (ADF program; see
details in the Experimental Section) were performed to
elucidate the nature of the bonding in these complexes and
to interpret the optical properties of the gold(I) and gold-
(I)-gold(III) complexes described above. The phenyl groups
of the phosphine ligands are modeled by hydrogen (h
models) or by a methyl group (m models), and C6F5 is
modeled by CF3, owing to the size of some of the complexes,
especially complex 5 with four gold atoms and several
ligands. The models of each complex are denoted as 1h, 1m,
3h, 3m, 4h, and 5h. Complexes 1 and 2 were also fully
calculated.
(i) Au(I) Complexes. The quality of the models was
checked by comparing the geometries of models 3h and 3m
with the experimental data for complex 3. Some relevant
distances and angles are given in Table 6.
DFT methods provide good agreement between experi-
mental and calculated structures, except when weak interac-
tions, such as the Au(I)‚‚‚Au(I) interaction in complex 3,
exist. Again, the size of the systems prevented us from using
Figure 4. Excitation and emission spectra at 77 K in the solid state: (a)
complex 1, (b) complex 2, (c) complex 3.
the presence of gold-gold interactions strongly influences
the emission bands, which become LMMCT in nature.13-15,55
The shift of the emission maxima in a series of substituted
benzenethiolate derivatives has suggested an MLCT transi-
tion.21 Moreover, the existence of π electrons in the phos-
phine ligand can complicate the assignment because an
LLCT transition [S (π) to Ph (π*)] is possible.21 LC
transitions can also contribute to the emission spectra when
(54) Narayanaswamy, R.; Young, M. A.; Parkhurst, E.; Ouellette, M.; Kerr,
M. E.; Ho, D. M.; Elder, R. C.; Bruce, A. E.; Bruce, M. R. M. Inorg.
Chem. 1993, 32, 2506.
(55) Roundhill, D. M.; Fackler, J. P., Jr. Optoelectronic Properties of
Inorganic Compounds; Plenum Press: New York, 1998; p 195.
(56) Van Zyl, W. E.; Lo´pez-de-Luzuriaga, J. M.; Mohamed, A. A.; J. M.;
Staples, R. J.; Fackler, J. P., Jr. Inorg. Chem. 2002, 41, 4579.
(57) Lee, Y. A.; McGarrah, J. E.; Lachicotte, R. J.; Eisenberg, R. J. Am.
Chem. Soc. 2002, 124, 10662.
Inorganic Chemistry, Vol. 45, No. 3, 2006 1065

Bardaj´ı et al.
The transition at 359 nm results from mixed excitations
from the HOMO to LUMO+1 (22%), LUMO+2 (19%), and
LUMO+3 (32%). The HOMO is 80% located on the thiolate
(49% on sulfur), whereas the three LUMOs are located on
one of the thiolates (Figure 6). Gold does not contribute more
than 10% to each orbital. Therefore, the excitations can be
assigned as interligand, from S in one ligand to the π* levels
of the other (LUMO+2, LUMO+3), or intraligand, from S
to π* in the same ligand (LUMO+1). The excitation can be
assigned to LLCT (S f π*) mixed with ILCT. The
introduction of the second thiolate changes the character of
the low-energy transitions.
Figure 5. Representation of the orbitals of model 1m of complex [Au-
(2-SC6H4NH2)(PPh3)] (1) involved in the lowest-energy transition.
3m was found to be a better model for complex 3 than
3h. The details on the frontier orbitals are given in the
Supporting Information. These complexes have two gold
atoms, being close to a dimer of complex 1. The same trends
are observed, that is, the occupied levels are strongly
localized on the thiolate ligand (and mostly on sulfur)
whereas the empty ones show participation of the metal and
the phosphine ligand.
The agreement between the TD results (3m) and those
experimentally obtained for complex 3 is quite good. The
most intense transitions are calculated at 322 and 345 nm
for 3m, the experimental value for absorption being 308 nm.
These transitions start in HOMO-5 and HOMO-4, and both
end in the LUMO. There is no relevant mixing of other
transitions. HOMO-5 and HOMO-4, as well as the HOMO
and other orbitals between them, are essentially located on
one or both of the thiolate ligands, with different contribu-
tions of sulfur and the π levels of the phenyl group (Figure
7). The LUMO is located on the two gold(I) atoms and the
phosphorus (Figure 7), representing the Au‚‚‚Au interaction
and the Au-P bond.
more suitable correlated methods. The calculated values
3.142 and 3.197 Å for models 3h and 3m, respectively,
compare surprisingly well with the experimental Au-Au
distance of 3.135 Å, as do the P-C distances and the angles.
On the other hand, the Au-S and Au-P distances in 3h are
too long. The introduction of the methyl groups leads to a
better agreement between calculated and experimental values
in the Au-S and Au-P distances. Mayer indices (MIs),
which measure the bond strengths independently of the
distance, indicate a stronger Au‚‚‚Au bond (MI ) 0.187)
for 3m than for 3h (0.138). In complex 1, the introduction
of phenyl groups lengthens the Au-P bond by ca. 0.03 Å,
the other distances and angles remaining the same. Although
both 1h and 1m provide a good model for complex 1, methyl
groups must be included (3m) to obtain a good model of
complex 3.
The energy, composition, and nature of the frontier orbitals
of models 1h and 1m and complex 1 are given in the
Supporting Information. TD-DFT calculations were per-
formed in order to assign the absorption spectra (see
Supporting Information). The lower-energy absorptions
should correspond to the excitation maxima observed in the
solid state at 77 K and be responsible for the luminescent
properties.
These transitions can therefore be assigned to LMCT (S
f Au‚‚‚Au-P charge transfer), mixed with some LLCT and
MC character. A similar result was obtained for the complex
[Au2(µ-i-mnt)(µ-dppm)], in which the low-energy absorption
For complex 1, the agreement between the calculated and
experimental (311 nm; UV-vis spectrum) values is better
when PMe3 (316 nm, 1m), rather than PH3 (293 nm, 1h), is
used in the model. This band is assigned to a mixed transition
from the HOMO to LUMO+2 (62%) and LUMO+3 (17%),
the HOMO being a thiolate-based orbital (88% thiolate, with
25% in sulfur) and the LUMOs being strongly located on
the phosphine, with a large contribution of gold. The
transition can thus be assigned as LMCT (S f Au-P charge
transfer) with some mixture of LLCT. The three orbitals
involved are shown in Figure 5.
is a combination of S (py) f dppm, S (py) f Au (pz) and
25
2
2
Au(dx -y ) f Au (pz). This assignment is similar to that
obtained for the model of 1 (1m), LMCT (S f Au-P), but
the gold atom is also involved in the Au-Au interaction.
The transition energies are shifted, and the nature of the
transition is modified to include the Au‚‚‚Au interaction.
(ii) Au(III) Complexes. The mixed Au(I)-Au(III) com-
plexes 4 and 5 contain one or two additional Au(C6F5)3 units
bound to the sulfur atoms, compared to 1 and 3. The crystal
structure of 5 revealed a significant contraction of the
Au‚‚‚Au distance to 2.935 Å, compared to that in 3, which
was the most interesting structural aspect introduced by
binding of the gold(III) fragments. Although the calculated
gold(I)-gold(I) distance (3.294 Å) is longer than in 3h (3.142
Å) and much longer that the experimental value (2.935 Å),
the MIs indicate that the Au‚‚‚Au interaction is stronger in
5h (0.142) than in 3h (0.138). We already pointed out the
limitations of model 3h, so model 5h should also provide
only moderate agreement with experiment.
In complex 2, the phosphine ligand of 1 has been replaced
by another thiolate, so that the complex becomes anionic.
The most intense calculated excitations correspond well with
the experimental excitation maximum, that is, the band at
353 nm is calculated to appear at 359 nm, with the shoulder
at 335 nm being reproduced by the two bands at 343 and
330 nm. The experimental band at 319 nm in the UV-vis
absorption spectrum is also close to the calculated value (359
nm), especially taking into account that the species is anionic
and solvent effects should be more important in this case.
The composition of the frontier orbitals of model 4h (see
the Supporting Information) differs from what was found
1066 Inorganic Chemistry, Vol. 45, No. 3, 2006

Quenching of Gold(I) Thiolate Luminescence
Figure 6. Representation of the orbitals of complex [Au(2-SC6H4NH2)2]- (2) involved in the lowest-energy transition.
Figure 7. Representation of the orbitals of model 3m of complex [{Au(2-SC6H4NH2)}2(µ-dppm)] (3) involved in the lowest-energy transition.
Figure 8. Representation of the orbitals of model 4h of complex [(AuPPh3){Au(C6F5)3}(µ2-2-SC6H4NH2)] (4) involved in the lowest-energy transition.
for complexes 1-3, as the Au(III) fragment participates to
a significant extent in many of them. The two most intense
excitations were calculated at 345 and 319 nm, in agreement
with the experimental values (345 nm for absorption). The
first transition has two components, both starting at
HOMO-1 and ending at the LUMO (61%) or LUMO+1
(27%); the two components of the second transition also start
at HOMO-1, ending at LUMO+2 (75%) or the LUMO
(11%). HOMO-1 is 82% located on the thiolate ligand, but
the sulfur atom contributes much less to it than in the related
1m model complex, the π levels of the phenyl group being
more important. The LUMO is 69% localized in the Au-
(III)-(CF3)3 fragment (26% in gold), whereas LUMO+1 and
LUMO+2 are Au(I)-P σ* levels (ca. 60% on the phosphine
and ca. 20% on gold). These four levels are shown in Figure
8.
The experimental absorption around 350 nm in the
absorption spectrum should arise from a mixed LMCT
(thiolate f AuIII and thiolate f Au-P). The strong involve-
ment of Au(III) in the LUMO changes the character of the
excited state and most probably influences the emissive state.
Notice that the sulfur atom is absent in the levels associated
with the excitations, whereas it played an important role in
the mononuclear Au(I) complex.
for 4h, but their quality is limited by the same factors as for
model complex 3h. It can still be stated, however, that the
lowest excitations also start on π* orbitals of the phenyl
group in the thiolate ligand and end on combinations of Au-
(III) d and Au(I)-P σ* orbitals. Significant shifts are found
between the calculated (448 nm) and experimental (342 nm)
values. The use of H instead of Ph influences the calculated
energies because it also affects the Au‚‚‚Au interactions,
which seems to be important for these transitions, as seen
for model complex 3h. Also, all of the calculations are
limited by the quality of the DFT method.
Conclusions
The acac method is well suited to the preparation of
phosphino gold(I) derivatives 1-3 of 2-aminobenzene-
thiolate, which emit intensely in the solid state at 77 K. The
sulfur atoms are able to coordinate to Au(C6F5)3 moieties to
give di- and tetranuclear mixed gold(I)-gold(III) derivatives
4 and 5; the latter shows a shortening of the intramolecular
gold(I)-gold(I) interaction (already observed in 3) and short
gold(I)-gold(IIII) distances. This coordination to the thiolate
ligand of a strongly electron-withdrawing gold(III) fragment
quenches the emission. DFT studies allowed us to determine
the electronic levels of these derivatives, and TD-DFT
calculations allowed us to determine the electronic excita-
tions. The excitation transitions could thus be mainly
The most intense excitations calculated for complex 5h
(Supporting Information) are similar in nature to those found
Inorganic Chemistry, Vol. 45, No. 3, 2006 1067

Bardaj´ı et al.
assigned (all are mixtures) as LMCT (Sf Au-P charge
transfer) for derivative 1, LLCT for derivative 2, and
LMMCT (Sf Au‚‚‚Au-P charge transfer) for derivative 3.
The introduction of Au(III) in the structure results in the
appearance of thiolate f AuIII transitions for derivatives 4
and 5. The change in the character of the absorbing state
probably affects the way the complex reaches the emissive
state and leads to the quenching of luminescence.
mischen Industrie for financial support. We thank Acc¸a˜o
Integrada Luso-Espanhola E31/04. P.J.C. acknowledges FCT
for a grant (SFRH/BD/10535/2002).
Supporting Information Available: Structural data for 3 and
5 (CIF format); tables of population analyses of 1h, 1m, 1, 2, 3h,
3m, 4h, and 5h; tables of calculated excitations and oscillator
strengths of 1h, 1m, 1, 2, 3h, 3m, 4h, and 5h; relevant calculated
distances and angles for models of complexes 4h and 5h, compared
to experimental values for 5; representations of the frontier orbitals
of 1. This material is available free of charge via the Internet at
http://pubs.acs.org.
Acknowledgment. This work is dedicated to Dr. Jose´
Antonio Abad on the occasion of his retirement. We thank
the Direccio´n General de Investigacio´n Cient´ıfica y Te´cnica
(Project BQU2001-2409-C02-01) and the Fonds der Che-
IC051168U
1068 Inorganic Chemistry, Vol. 45, No. 3, 2006