Pyridyl gold(I) alkynyls: A synthetic, structural, spectroscopic, and computational study

Article abstract of DOI:10.1021/om1003978

Planar neutral organometallic complexes have the potential to be exploited as building blocks for the construction of self-assembled nanoscale materials. We report the synthesis and first structural characterization of planar pyridyl gold(I) alkynyl complexes. Stable isolable complexes are only obtained when electron-rich pyridine ligands such as 4-(dimethylamino)pyridine or 4-aminopyridine are incorporated into the complexes. X-ray crystallography confirms that the pyridyl gold(I) alkynyls are neutral, linear, two-coordinate complexes. Furthermore, these gold(I) molecules form dimers in the solid state. However, no aurophilic interactions are observed to stabilize the solid-state structures. Density functional theory (DFT) calculations have been employed to model the electronic structure and thermodynamics of formation for selected complexes.

Full text of DOI:10.1021/om1003978

6186 Organometallics 2010, 29, 6186–6195
DOI: 10.1021/om1003978
Pyridyl Gold(I) Alkynyls: A Synthetic, Structural, Spectroscopic, and
Computational Study
Kelly J. Kilpin,† Raphael Horvath,† Geoffrey B. Jameson,‡ Shane G. Telfer,‡,§
Keith C. Gordon,†,§ and James D. Crowley*,†
†Department of Chemistry, University of Otago, PO Box 56, Dunedin, New Zealand, ‡Institute of
Fundamental Sciences, Massey University, Private Bag 11 222, Palmerston North, New Zealand, and
§MacDiarmid Institute for Advanced Materials and Nanotechnology, New Zealand
Received April 30, 2010
Planar neutral organometallic complexes have the potential to be exploited as building blocks for
the construction of self-assembled nanoscale materials. We report the synthesis and first structural
characterization of planar pyridyl gold(I) alkynyl complexes. Stable isolable complexes are only
obtained when electron-rich pyridine ligands such as 4-(dimethylamino)pyridine or 4-aminopyridine
are incorporated into the complexes. X-ray crystallography confirms that the pyridyl gold(I) alkynyls
are neutral, linear, two-coordinate complexes. Furthermore, these gold(I) molecules form dimers in
the solid state. However, no aurophilic interactions are observed to stabilize the solid-state structures.
Density functional theory (DFT) calculations have been employed to model the electronic structure
and thermodynamics of formation for selected complexes.
Introduction
stacked molecular aggregates often demonstrate interesting
photophysical and photochemical properties due to orbital
interactions between the π systems and metal ions of adja-
cent complexes. Che3 and others4 have used these planar
platinum(II) building blocks to generate nanostructured
materials with luminescent, semiconducting, liquid crystal-
line, and gelating properties.
Like platinum(II), gold(I) is well-known to form metal-
metal interactions in solution and the solid state. These
aurophilic interactions can give rise to photophysical proper-
ties such as luminescence. Interest in these phenomena has
led to the development of a range of gold(I)-containing
materials.5 One class of complexes that has been extensively
studied in this regard (because they tend to form robust air-
and water-stable organometallic compounds) are the gold(I)
alkynyl complexes.6 The gold(I) alkyne fragment is most
often found complexed to phosphine, carbene, or isonitrile
Planar organometallic complexes are beginning to show
promise as building blocks for the construction of self-
assembled nanoscale materials.1 In particular, the synthesis
of planar platinum(II) complexes containing π-conjugated
ligands has received considerable attention, because these
molecules are air- and water-stable and have the ability to
self-assemble by exploiting a combination of metal-metal
(Pt-Pt) and π-π interactions.2 The resulting cofacially
*To whom correspondence should be addressed. E-mail: jcrowley@
chemistry.otago.ac.nz. Fax: þ64 3 479 7906. Tel: þ64 3 479 7731.
(1) (a) Elemans, J. A. A. W.; van Hameren, R.; Nolte, R. J. M.;
Rowan, A. E. Adv. Mater. 2006, 18 (10), 1251–1266. (b) Hoeben,
F. J. M.; Jonkheijm, P.; Meijer, E. W.; Schenning, A. P. H. J. Chem.
Rev. 2005, 105 (4), 1491–1546. (c) Schenning, A. P. H. J.; Meijer, E. W.
Chem. Commun. 2005, No. 26, 3245–3258. (d) Shimizu, T.; Masuda, M.;
Minamikawa, H. Chem. Rev. 2005, 105 (4), 1401–1443. (e) Simpson,
C. D.; Wu, J.; Watson, M. D.; Muellen, K. J. Mater. Chem. 2004, 14 (4),
494–504. (f) Yamamoto, T.; Fukushima, T.; Aida, T. Adv. Polym. Sci.
2008, 220 (Self-Assembled Nanomaterials II), 1–27.
(2) (a) Bremi, J.; Caseri, W.; Smith, P. J. Mater. Chem. 2001, 11 (10),
2593–2596. (b) Caseri, W. R.; Chanzy, H. D.; Feldman, K.; Fontana,
M.; Smith, P.; Tervoort, T. A.; Goossens, J. G. P.; Meijer, E. W.;
Schenning, A. P. H. J.; Dolbnya, I. P.; Debije, M. G.; de Haas, M. P.;
Warman, J. M.; van de Craats, A. M.; Friend, R. H.; Sirringhaus, H.;
Stutzmann, N. Adv. Mater. 2003, 15 (2), 125–129. (c) Debije, M. G.; de
Haas, M. P.; Savenije, T. J.; Warman, J. M.; Fontana, M.; Stutzmann,
N.; Caseri, W. R.; Smith, P. Adv. Mater. 2003, 15 (11), 896–899.
(d) Debije, M. G.; De Haas, M. P.; Warman, J. M.; Fontana, M.;
Stutzmann, N.; Kristiansen, M.; Caseri, W. R.; Smith, P.; Hoffmann, S.;
Solling, T. I. Adv. Funct. Mater. 2004, 14 (4), 323–328. (e) Fontana, M.;
Caseri, W.; Smith, P. Platinum Met. Rev. 2006, 50 (3), 112–117.
(f) Fontana, M.; Chanzy, H.; Caseri, W. R.; Smith, P.; Schenning, A.
P. H. J.; Meijer, E. W.; Groehn, F. Chem. Mater. 2002, 14 (4), 1730–
1735. (g) Kim, E.-G.; Schmidt, K.; Caseri, W. R.; Kreouzis, T.;
Stingelin-Stutzmann, N.; Bredas, J.-L. Adv. Mater. 2006, 18 (15), 2039–
2043. (h) Anderson, B. M.; Hurst, S. K. Eur. J. Inorg. Chem. 2009, No.
21, 3041–3054. (i) Hori, A.; Shinohe, A.; Yamasaki, M.; Nishibori, E.;
Aoyagi, S.; Sakata, M. Angew. Chem., Int. Ed. 2007, 46 (40), 7617–7620.
(3) (a) Chen, Y.; Li, K.; Lu, W.; Chui, S. S.-Y.; Ma, C.-W.; Che, C.-
M. Angew. Chem., Int. Ed. 2009, 48 (52), 9909–9913. (b) Kui, S. C. F.;
Chui, S. S.-Y.; Che, C.-M.; Zhu, N. J. Am. Chem. Soc. 2006, 128 (25),
8297–8309. (c) Lu, W.; Chen, Y.; Roy, V. A. L.; Chui, S. S.-Y.; Che, C.-
M. Angew. Chem., Int. Ed. 2009, 48 (41), 7621–7625. (d) Lu, W.; Chui, S.
S.-Y.; Ng, K.-M.; Che, C.-M. Angew. Chem., Int. Ed. 2008, 47 (24),
4568–4572. (e) Lu, W.; Law, Y.-C.; Han, J.; Chui, S. S.-Y.; Ma, D.-L.;
Zhu, N.; Che, C.-M. Chem. Asian J. 2008, 3 (1), 59–69. (f) Lu, W.; Ng,
K.-M.; Che, C.-M. Chem. Asian J. 2009, 4 (6), 830–834. (g) Lu, W.; Roy,
V. A. L.; Che, C.-M. Chem. Commun. 2006, No. 38, 3972–3974. (h) So,
M.-H.; Roy, V. A. L.; Xu, Z.-X.; Chui, S. S.-Y.; Yuen, M.-Y.; Ho,
C.-M.; Che, C.-M. Chem. Asian J. 2008, 3 (11), 1968–1978. (i) Sun, Y.;
Ye, K.; Zhang, H.; Zhang, J.; Zhao, L.; Li, B.; Yang, G.; Yang, B.;
Wang, Y.; Lai, S.-W.; Che, C.-M. Angew. Chem., Int. Ed. 2006, 45 (34),
5610–5613. (j) Tong, G. S.-M.; Che, C.-M. Chem. Eur. J. 2009, 15 (29),
7225–7237. (k) Yuen, M.-Y.; Roy, V. A. L.; Lu, W.; Kui, S. C. F.; Tong,
G. S. M.; So, M.-H.; Chui, S. S.-Y.; Muccini, M.; Ning, J. Q.; Xu, S. J.;
Che, C.-M. Angew. Chem., Int. Ed. 2008, 47 (51), 9895–9899. (l) Zhu,
M.-X.; Lu, W.; Zhu, N.; Che, C.-M. Chem. Eur. J. 2008, 14 (31), 9736–
9746. (m) Shieh, S.-J.; Hong, X.; Peng, S.-M.; Che, C.-M. J. Chem. Soc.,
Dalton Trans. 1994, No. 20, 3067–8.
r
pubs.acs.org/Organometallics
Published on Web 10/29/2010
2010 American Chemical Society

Article
Organometallics, Vol. 29, No. 23, 2010 6187
two-electron-donor ligands, forming neutral linear two-
coordinate complexes. However, the steric bulk associated
with the phosphine and carbene ligands can prevent close
contacts between adjacent Au(I) ions. Replacement of these
bulky ligands with aromatic pyridinyl donor groups should
lead to the formation of neutral planar pyridyl gold(I)
alkynyl complexes such as 1a. These pyridyl gold(I) alkynyl
complexes could potentially be useful building blocks for the
synthesis of nanoscale materials. The lack of steric hindrance
around the Au(I) center should enable the two aromatic
Figure 1. (a) Space-filling molecular model of 1a. (b) Space-
filling representation of a potential stacking arrangement of 1a
stabilized by Au-Au and π-π interactions.
(4) (a) Frischmann, P. D.; Guieu, S.; Tabeshi, R.; MacLachlan, M. J.
J. Am. Chem. Soc. 2010, 132 (22), 7668–7675. (b) Camerel, F.; Ziessel,
R.; Donnio, B.; Bourgogne, C.; Guillon, D.; Schmutz, M.; Iacovita, C.;
Bucher, J.-P. Angew. Chem., Int. Ed. 2007, 46 (15), 2659–2662. (c) Clark,
M. L.; Diring, S.; Retailleau, P.; McMillin, D. R.; Ziessel, R. Chem. Eur.
J. 2008, 14 (24), 7168–7179. (d) Tam, A. Y.-Y.; Wong, K. M.-C.; Wang,
G.; Yam, V. W.-W. Chem. Commun. 2007, 20, 2028–2030. (e) Tam,
A. Y.-Y.; Wong, K. M.-C.; Yam, V. W.-W. Chem.;Eur. J. 2009, 15 (19),
4775–4778. (f) Tam, A. Y.-Y.; Wong, K. M.-C.; Yam, V. W.-W. J. Am.
Chem. Soc. 2009, 131 (17), 6253–6260. (g) Kozhevnikov, V. N.; Donnio,
B.; Bruce, D. W. Angew. Chem., Int. Ed. 2008, 47 (33), 6286–6289.
(h) Santoro, A.; Whitwood, A. C.; Williams, J. A. G.; Kozhevnikov, V. N.;
Bruce, D. W. Chem. Mater. 2009, 21 (16), 3871–3882. (i) Cardolaccia, T.;
Li, Y.; Schanze, K. S. J. Am. Chem. Soc. 2008, 130 (8), 2535–2545.
(j) Damm, C.; Israel, G.; Hegmann, T.; Tschierske, C. J. Mater. Chem.
2006, 16 (19), 1808–1816.
(5) (a) Hutchings, G. J.; Brust, M.; Schmidbaur, H. Chem. Soc. Rev.
2008, 37 (9), 1759–1765. (b) Laguna, A., Ed. Modern Supramolecular
Gold Chemistry: Gold-Metal Interactions And Applications; Wiley: New
York, 2008; 505 pp. (c) Puddephatt, R. J. Chem. Soc. Rev. 2008, 37 (9),
2012–2027. (d) Schmidbaur, H. Gold Bull. 2000, 33 (1), 3–10. (e) Schmidbaur,
H. Nature 2001, 413 (6851), 31–33. (f) Schmidbaur, H.; Schier, A. Chem.
Soc. Rev. 2008, 37 (9), 1931–1951.
ligands to adopt a coplanar conformation, which would
allow these organometallic fragments to assemble, through
a combination of π-π and Au(I)-Au(I) interactions, into
linear supramolecular stacks (Figure 1). Recent literature
provides support for this postulate: Lin et al. have demon-
strated that the planar pyridine containing gold(I) cations
[Au(py)2]þ 7 and [(py)Au(carbene)]þ 8 form linear stacks in
the solid state, while Laguna et al. have shown that neutral
[(py)Au(C6Cl5)]9 complexes behave in an analogous manner.
We have a longstanding interest in metal-metal interactions10
and have been exploring ways to exploit them to construct
nanoscale materials. As part of that work, we report here the
synthesis, properties, and structural characterization of pyridyl
gold(I) alkynyl complexes. Additionally, density functional
theory (DFT) calculations have been carried out in order
to explain the observed thermodynamic and spectroscopic
properties.
(6) (a) Hogarth, G.; Alvarez-Falcon, M. M. Inorg. Chim. Acta 2005,
358 (5), 1386–1392. (b) Hunks, W. J.; MacDonald, M.-A.; Jennings,
M. C.; Puddephatt, R. J. Organometallics 2000, 19 (24), 5063–5070.
(c) Partyka, D. V.; Gao, L.; Teets, T. S.; Updegraff, J. B.; Deligonul, N.;
Gray, T. G. Organometallics 2009, 28 (21), 6171–6182. (d) Partyka,
D. V.; Updegraff, J. B., III; Zeller, M.; Hunter, A. D.; Gray, T. G.
Organometallics 2007, 26 (1), 183–186. (e) Schuster, O.; Liau, R.-Y.;
Schier, A.; Schmidbaur, H. Inorg. Chim. Acta 2005, 358 (5), 1429–1441.
(f) Yip, S.-K.; Lam, W. H.; Zhu, N.; Yam, V. W.-W. Inorg. Chim. Acta
2006, 359 (11), 3639–3648. (g) Habermehl, N. C.; Eisler, D. J.; Kirby,
C. W.; Yue, N. L. S.; Puddephatt, R. J. Organometallics 2006, 25 (12),
2921–2928. (h) Habermehl, N. C.; Jennings, M. C.; McArdle, C. P.;
Mohr, F.; Puddephatt, R. J. Organometallics 2005, 24 (21), 5004–5014.
(i) Hunks, W. J.; Jennings, M. C.; Puddephatt, R. J. Z. Naturforsch., B:
Chem. Sci. 2004, 59 (11/12), 1488–1496. (j) Hunks, W. J.; Lapierre, J.;
Jenkins, H. A.; Puddephatt, R. J. Dalton Trans. 2002, No. 14, 2885–2889.
(k) Irwin, M. J.; Vittal, J. J.; Puddephatt, R. J. Organometallics 1997, 16 (15),
3541–3547. (l) Jia, G.; Payne, N. C.; Vittal, J. J.; Puddephatt, R. J.
Organometallics 1993, 12 (12), 4771–8. (m) Jia, G.; Puddephatt, R. J.;
Scott, J. D.; Vittal, J. J. Organometallics 1993, 12 (9), 3565–74. (n) Jia,
G.; Puddephatt, R. J.; Vittal, J. J. J. Organomet. Chem. 1993, 449 (1-2),
211–17. (o) Jia, G.; Puddephatt, R. J.; Vittal, J. J.; Payne, N. C.
Organometallics 1993, 12 (2), 263–5. (p) McArdle, C. P.; Irwin, M. J.;
Jennings, M. C.; Puddephatt, R. J. Angew. Chem., Int. Ed. 1999, 38 (22),
3376–3378. (q) McArdle, C. P.; Irwin, M. J.; Jennings, M. C.; Vittal, J. J.;
Puddephatt, R. J. Chem. Eur. J. 2002, 8 (3), 723–734. (r) McArdle, C. P.;
Jennings, M. C.; Vittal, J. J.; Puddephatt, R. J. Chem. Eur. J. 2001, 7 (16),
3572–3583. (s) McArdle, C. P.; Van, S.; Jennings, M. C.; Puddephatt,
R. J. J. Am. Chem. Soc. 2002, 124 (15), 3959–3965. (t) McArdle, C. P.;
Vittal, J. J.; Puddephatt, R. J. Angew. Chem., Int. Ed. 2000, 39 (21),
3819–3822. (u) Mohr, F.; Eisler, D. J.; McArdle, C. P.; Atieh, K.;
Jennings, M. C.; Puddephatt, R. J. J. Organomet. Chem. 2003, 670
(1-2), 27–36. (v) Mohr, F.; Jennings, M. C.; Puddephatt, R. J. Eur. J.
Inorg. Chem. 2003, No. 2, 217–223. (w) Mohr, F.; Jennings, M. C.;
Puddephatt, R. J. Angew. Chem., Int. Ed. 2004, 43 (8), 969–971. (x) Payne,
N. C.; Ramachandran, R.; Puddephatt, R. J. Can. J. Chem. 1995, 73 (1),
6–11. (y) Qin, Z.; Jennings, M. C.; Puddephatt, R. J. Chem. Eur. J. 2002,
8 (3), 735–738. (z) Bruce, M. I.; Horn, E.; Matisons, J. G.; Snow, M. R.
Aust. J. Chem. 1984, 37, 1163–1170. (aa) Bruce, M. I.; Hall, B. C.; Skelton,
B. W.; Smith, M. E.; White, A. H. Dalton Trans. 2002, No. 6, 995–1001.
(ab) Bruce, M. I.; Jevric, M.; Skelton, B. W.; Smith, M. E.; White, A. H.;
Zaitseva, N. N. J. Organomet. Chem. 2006, 691 (3), 361–370. (ac) Khairul,
W. M.; Fox, M. A.; Zaitseva, N. N.; Gaudio, M.; Yufit, D. S.; Skelton,
B. W.; White, A. H.; Howard, J. A. K.; Bruce, M. I.; Low, P. J. Dalton
Trans. 2009, No. 4, 610–620.
Experimental Section
General Considerations. Ethynylbenzene, ethynylferrocene,
AuCl(SMe2), 4-(dimethylamino)pyridine (4-dmap), and 4-ami-
nopyridine (4-NH2Py) were purchased from commercial sources
(Aldrich) and were used as received. NEt3 was distilled over
KOH prior to use. Petrol refers to the fraction of petroleum ether
boiling in the range 40-60 °C. The compounds [(4-dmap)2Au]-
[AuCl2], [(4-NH2Py)2Au][AuCl2], [(4-pic)2Au][AuCl2], and
[(py)2Au][AuCl2] were synthesized by literature procedures.7
As a precautionary measure, all reactions were carried out in
the absence of light. NMR experiments were carried out on
either a Varian 400-MR or a Varian 500 MHz VNMRS spectro-
meter operating at 298 K with chemical shifts referenced to
1
residual solvent peaks (CDCl3, H 7.26 ppm, 13C 77.16 ppm;
CD3CN, 1H 1.94 ppm, 13C 1.32, 118.26 ppm; d6-DMSO, 1H 2.50
ppm). Chemical shifts are reported in parts per million (ppm)
and coupling constants (J) in hertz (Hz). Standard abbreviations
indicating multiplicity were used as follows: m = multiplet, t =
triplet, d = doublet, s = singlet, b = broad. ESI-MS analysis
was carried out on methanol/dichloromethane solutions of the
compound on a Bruker MicroTOF-Q spectrometer and IR
(7) Lin, J. C. Y.; Tang, S. S.; Vasam, C. S.; You, W. C.; Ho, T. W.;
Huang, C. H.; Sun, B. J.; Huang, C. Y.; Lee, C. S.; Hwang, W. S.; Chang,
A. H. H.; Lin, I. J. B. Inorg. Chem. 2008, 47 (7), 2543–2551.
(8) Chiou, J. Y. Z.; Luo, S. C.; You, W. C.; Bhattacharyya, A.;
Vasam, C. S.; Huang, C. H.; Lin, I. J. B. Eur. J. Inorg. Chem. 2009, No.
13, 1950–1959.
(9) Fernandez, E. J.; Laguna, A.; Lopez-de-Luzuriaga, J. M.; Monge,
M.; Montiel, M.; Olmos, M. E.; Perez, J.; Rodriguez-Castillo, M. Gold
Bull. 2007, 40 (3), 172–183.
(10) (a) Crowley, J. D.; Bosnich, B. Eur. J. Inorg. Chem. 2005, No. 11,
2015–2025. (b) Crowley, J. D.; Goshe, A. J.; Bosnich, B. Chem. Commun.
2003, No. 3, 392–393. (c) Crowley, J. D.; Goshe, A. J.; Steele, I. M.; Bosnich,
B. Chem. Eur. J. 2004, 10 (8), 1944–1955. (d) Crowley, J. D.; Steele, I. M.;
Bosnich, B. Inorg. Chem. 2005, 44 (9), 2989–2991. (e) Crowley, J. D.;
Steele, I. M.; Bosnich, B. Eur. J. Inorg. Chem. 2005, No. 19, 3907–3917.

6188 Organometallics, Vol. 29, No. 23, 2010
Kilpin et al.
spectroscopy (KBr disks) on a Perkin-Elmer Spectrum BX FT-
IR spectrometer. Melting points were determined on a Sanyo
Gallenkamp apparatus and are uncorrected. Conductivity mea-
interlocking spheres. All TD-DFT calculations were carried out
this way, as the absorption spectra were also measured in the
solution phase.
surements were carried out on approximately 1.0 ꢀ 10-3
M
In order to ensure that calculated quantities are accurate, we
calculated mean absolute deviations (MADs) between theoret-
ical and experimental vibrational spectra. These are obtained
by averaging the absolute difference in cm-1 of the 5-10 most
prominent experimental peaks to the corresponding calculated
peaks in the spectral range 400-1800 cm-1; to account for
dissimilar anharmonicities, peaks found in different regions of
the spectra have to be scaled differently.13 We calculated the
scale factors to give the lowest MAD for each basis set.
The IR and Raman spectra were generated using a Lorentzian
acetonitrile solutions of the complexes using a Suntex SC-170
conductivity meter. Microelemental analysis was carried out at
the Campbell Microanalytical Laboratory at the University of
Otago. FT-Raman spectra of PhCtCAu(4-dmap) (1b) and
PhCtCAu(4-NH2Py) (1c) were acquired in KBr and in the
solution phase (CH3CN). The former was carried out using a
Senterra dispersive Raman microscope (Bruker Optics, Ettlin-
gen, Germany) using radiation from a 785 nm solid-state laser
operating at 50 mW. An Olympus 20ꢀ objective with a 50 μm
confocal pinhole was used to collect the Raman signal. In order
to minimize irradiation time and thus sample degradation, 36
individual spectra were acquired in a 6 ꢀ 6 geometry spaced over
a 200 μm grid, which were then coadded. KBr was used to
absorb heating caused by the laser in order to minimize sample
degradation. The solution-phase FT-Raman spectra were ac-
quired using a Bruker Equinox 55 interferometer coupled with a
FRA-106 Raman module and a D418T liquid-nitrogen-cooled
germanium detector, controlled by the Bruker OPUS v5.5 soft-
ware package. The spectra were acquired as 1024 coadded scans
using 1064 nm radiation of a Nd:YAG laser operating at
360 mW. The resolution of all spectra is 4 cm-1. UV-visible
absorption spectra were acquired with a Jasco V-550 spectro-
photometer. Steady-state emission spectra were acquired with a
Perkin-Elmer LS-50 B luminescence spectrometer using an
excitation wavelength of 280 nm. Excited-state emission and
absorption transients were acquired using a LP920K transient
absorption (TA) system by Edinburgh Instruments. Excitation
was carried out using third-harmonic radiation (355 nm) from a
Brilliant (Quantel) Nd:YAG laser operating at 5 Hz, and in TA
mode, a Xe900 450W xenon arc lamp was used as the probe
source. The photons were dispersed using a TMS300-A mono-
chromator with a 1800 grooves/mm grating, recorded on a R928
(Hamamatsu) photomultiplier and transcribed on a TDS3012C
(Tektronix) oscilloscope. The samples were in CH3CN solution
and were thoroughly degassed with argon prior to the experi-
ment. Cyclic voltammetric experiments in CH2Cl2 were per-
formed at 20 °C on solutions degassed with nitrogen. A three-
electrode cell was used with a Cypress Systems 1.0 mm diameter
Pt working electrode, Ag/AgCl reference electrode, and plati-
num-wire auxiliary electrode. The solution was ∼10-3 M in
electroactive material and contained 0.1 M [Bu4N][PF6] as the
supporting electrolyte. Voltammograms were recorded with the
aid of a Powerlab/4sp computer-controlled potentiostat. Poten-
tials are referenced to the reversible formal potential (taken as
E° = 0.00 V) for the decamethylferrocenium/decamethylferro-
cene (Fc*þ/0) process,11 where E° was calculated from the
average of the oxidation and reduction peak potentials under
conditions of cyclic voltammetry. Under the same conditions,
line shape with a half-width at half-maximum (HWHM) of 3 cm-1
.
The vibrational modes and molecular orbitals were visualized
using Molden14 and GaussView v5.0 (Gaussian, Inc.), respec-
tively, after extraction from the output files using GaussSum.15
Unambiguous assignment of vibrational modes from visual
comparison of spectra was possible for most peaks.
General Procedure for the Preparation of Gold(I) Pyridyl
Alkynes. To a solution of the gold(I) alkyne (1 equiv) in
dichloromethane (10 mL) was added the substituted pyridine
(1 equiv) and the solution stirred in a foil-covered flask at room
temperature for 1.5 h. The solution was reduced in volume (ca.
5 mL) and, if required, filtered through a plug of Celite to
remove any insoluble Au(0) material. Diethyl ether was added to
the solution, resulting in the precipitation of the gold(I) alkynyl
pyridyl complex as either an off-white or orange solid (60-73%
yield). Full details and characterization data can be found in the
Supporting Information.
X-ray Data Collection and Refinement. X-ray data for 2b and
1b were recorded with a Bruker APEX II CCD diffractometer at
˚
89(2) K using Mo KR radiation (λ = 0.710 73 A). The structures
were solved by either direct methods (1b) using SIR9716 or
Patterson methods (2b) using SHELXL-97.17 The resulting
Fourier maps revealed the location of all other non-hydrogen
atoms. Weighted full-matrix least-squares refinement on F2 was
carried out using SHELXL-97.17 For 2b all non-hydrogen
atoms were refined anisotropically, but for 1b the poor data
set meant only the gold atoms were able to be modeled aniso-
tropically. In both cases, the hydrogen atoms were included in
calculated positions and were refined as riding atoms with
individual (or group, if appropriate) isotropic displacement
parameters. The residual electron density in both 1b and 2b is
due to Fourier ripples around the gold atom (the peaks are
˚
approximately 1.2 A from the gold).
X-ray data for 1c,d were collected at 123 K on a Rigaku
Spider diffractometer equipped with a copper rotating anode
˚
X-ray source (λ = 1.541 78 A) and a curved image plate
detector. Structures were solved by direct methods and refined
against F2 using anisotropic thermal displacement parameters
for all non-hydrogen atoms. Hydrogen atoms were placed in
calculated positions and refined using a riding model (except
where noted in the Supporting Information).
E° calculated for Fcþ/0 was 0.55 V versus Fc*þ/0
.
Computations. Density functional theory (DFT) calculations
were carried out using the Gaussian 09 package (Gaussian,
Inc.). Optimization and frequency calculations were both per-
formed using the B3LYP functional with a variety of effective
core potentials (ECPs) for the gold atom and the 6-31G(d) basis
set for all other atoms. Time-dependent (TD) calculations were
carried out on the optimized structures using the B3LYP and the
Coulomb attenuated method (CAM) B3LYP functional12 with
the same basis sets described above. Solvent-based calculations
were also performed, using the IEF-PCM method, which treats
the solvent as a continuum containing the solute as a series of
All ORTEP18 diagrams have been drawn with 50% probabil-
ity ellipsoids. Crystal data and collection parameters are given
in Table 1. The CIF files CCDC 774529 (1b), CCDC 774532
(1c), CCDC 774530 (1d), and CCDC 774531 (2b) contain the
(13) Howell, S. L.; Gordon, K. C.; McGarvey, J. J. J. Phys. Chem. A
2005, 109 (12), 2948–2956.
(14) Schaftenaar, G.; Noordik, J. H. J. Comput.-Aided Mol. Des.
2000, 14 (2), 123–134.
(15) O’Boyle, N. M.; Tenderholt, A. L.; Langner, K. M. J. Comput.
Chem. 2008, 29 (5), 839–845.
(16) Altomare, A.; Burla, M. C.; Camalli, M.; Cascarano, G. L.;
Giacovazzo, C.; Guagliardi, A.; Moliterni, A. G. G.; Polidori, G.;
Spagna, R. J. Appl. Crystallogr. 1999, 32 (1), 115–119.
(17) Sheldrick, G. M. Acta Crystallogr., Sect. A: Found. Crystallogr.
2008, A64 (1), 112–122.
(11) Noviandri, I.; Brown, K. N.; Fleming, D. S.; Gulyas, P. T.; Lay,
P. A.; Masters, A. F.; Phillips, L. J. Phys. Chem. B 1999, 103 (32), 6713–
6722.
(12) Yanai, T.; Tew, D. P.; Handy, N. C. Chem. Phys. Lett. 2004, 393
(1-3), 51–57.
(18) Farrugia, L. J. J. Appl. Crystallogr. 1997, 30 (5, Pt. 1), 565.

Article
Organometallics, Vol. 29, No. 23, 2010 6189
Table 1. Crystallographic Data for 1c,d and 2b
1c
1d
2b
CCDC no.
formula
formula wt
774532
13H11N2Au
392.21
774530
C14H12NAu
391.21
774531
C19H19N2FeAu
528.18
C
cryst syst, space group
˚
˚
b, A
˚
c, A
monoclinic, P21/n
6.15020(10)
13.2177(2)
14.3392(10)
90
98.22(2)
90
1153.68(8)
4
0.50 ꢀ 0.12 ꢀ 0.10;
pale yellow, block
2.258
23.723
3256
1302 (0.2232)
1302/91/158
0.597
monoclinic, P21/c
6.393(2)
7.2809(3)
25.1246(9)
90
97.161(12)
90
1160.4(4)
4
0.16 ꢀ 0.12 ꢀ 0.01;
colorless, plate
2.239
23.558
14 531
2193 (0.0538)
2193/12/146
1.369
monoclinic, P21/c
10.833(3)
14.922(4)
10.734(3)
90
107.90(2),
90
1651.3(8)
4
0.21 ꢀ 0.07 ꢀ 0.02;
orange, plate
2.125
9.752
14 272
2897 (0.1299)
2897/0/199
1.018
a, A
R, deg
β, deg
γ, deg
3
˚
V, A
Z
cryst size, mm; color, habit
F
calcd, mg/mm3
μ, mm-1
no. of rflns collected
no. of indep rflns (Rint
)
no. of data/restraints/params
goodness of fit on F2
final R1 and wR2 indexes (I > 2σ(I))
final R1 and wR2 indexes (all data)
largest difference in peak and hole (e A
0.0743, 0.1514
0.0834, 0.1628
1.407 and -1.992
0.0369, 0.1127
0.0425, 0.1191
1.719 and -1.691
0.0560, 0.1105
0.1036, 0.1300
1.945 and -1.721
-3
˚
)
supplementary crystallographic data for this paper. These data
can be obtained free of charge via www.ccdc.ac.uk/conts/
retrieving.html or from the Cambridge Crystallographic Data
Centre, 12 Union Road, Cambridge CB2 1EZ, U.K. (fax: (þ44)
1223-336-033).
Scheme 1. Synthesis of Mononuclear Au(I) Pyridyl Alkynyl
Complexesa
Results and Discussion
Synthesis. The synthesis of the pyridyl gold(I) alkynyl
complex 1a has been reported previously. The complex was
obtained by reacting the polymeric gold(I) alkynyl complex
[Au(CtCPh)]n (1) with an excess of pyridine.19 We set out to
exploit this methodology to synthesize a family of pyridyl
gold(I) alkynyl complexes 1a-fand 2a-d(Scheme 1, method a).
The pyridyl ligands (3a-f, 1 equiv) were added to a suspen-
6a,20
sion of [Au(CtCAr)]n
(1 or 2; 1 equiv) in dichloro-
methane at room temperature, and the resulting reaction
mixture was stirred for 1.5 h. With the two most electron-rich
pyridine ligands, 4-(dimethylamino)pyridine (4-dmap, 3b;
pKa = 9.4) and 4-aminopyridine (4-NH2Py, 3c; pKa = 9.11),
homogeneous solutions formed. After filtration to remove a
small amount of insoluble Au(0), which had formed during
the reaction, the addition of diethyl ether caused the pre-
cipitation of the pyridyl alkynyl complexes 1b,c and 2b,c in
moderate yields (60-73%). However, the reactions of
[Au(CtCAr)]n with the less basic amines 4-methylpyridine
(19) Coates, G. E.; Parkin, C. J. Chem. Soc. 1962, 3220–6.
ꢀ
(20) (a) Vicente, J.; Chicote, M. T.; Alvarez-Falcon, M. M.;
ꢀ
Abrisqueta, M.-D.; Hernandez, F. J.; Jones, P. G. Inorg. Chim. Acta 2003,
347, 67–74. (b) Siemeling, U.; Rother, D.; Bruhn, C. Organometallics 2008, 27
(24), 6419–6426. (c) Vicente, J.; Chicote, M.-T.; Alvarez-Falcon, M. M.;
Jones, P. G. Organometallics 2005, 24 (24), 5956–5963. (d) The gold(I)
alkynyl complexes used in this study were prepared by established
methodologies.20a-c AuCl(SMe)2 was reacted with the appropriate
terminal alkyne (either ethynylbenzene or ethynylferrocene) in the
presence of triethylamine, providing the gold(I) alkynyl complexes
[Au(CtCPh)]n and [Au(CtCFc)]n in good yields. With the exception
of the tetrameric tert-butyl complex [Au(CtCBut)]4, it is generally
accepted that Au(I) alkynes are polymeric in nature, with the linear
coordination geometry of the gold(I) center completed by π-electron
donation from the alkyl group of a neighboring molecule. As a result, the
gold alkynyl complexes tend to be insoluble in common organic sol-
vents;such is the case for the phenyl derivative [Au(CtCPh)]n. The
ferrocenyl derivative [Au(CtCFc)]n, however, shows much better solu-
bility and dissolves readily in solvents such as dichloromethane.
a Legend: (a) CH2Cl2; (b) 1 or 2 (1 equiv), NEt3, CH2Cl2.
(4-pic, 3d; pKa = 6.021), 2,6-dimethylpyridine (2,6-lut, 3e;
pKa = 6.721), pyridine (py, 3a; pKa = 5.2521), 4,40-bipyridine

6190 Organometallics, Vol. 29, No. 23, 2010
Kilpin et al.
(bipy, 3f; pKa = 4.921), and 4-cyanopyridine (4-CNPy, 3g;
pKa =1.8622) proceeded differently from those reported above.
An excess (∼6 equiv) of the pyridyl ligands (3a,d,e) was
required to dissolve the [Au(CtCAr)]n polymers and give
homogeneous solutions, while refluxing a 100-fold excess of
3f,g with [Au(CtCPh)]n in either dichloromethane or tol-
uene failed to dissolve the polymeric starting materials.
Addition of diethyl ether to the reaction mixtures containing
the pyridyl ligands 3a,d,e resulted in the formation of
microcrystalline materials. However, the attempted isolation
of these pyridyl alkynyl complexes23 by filtration and vacuum
drying caused the colorless or orange solids to revert back to
the initial polymeric materials, 1 (yellow) or 2 (red). Micro-
analysis of the isolated solids was consistent with [Au(Ct
CAr)]n. Attempts to synthesize the analogous digold complexes
from 1,3- or 1,4-substituted [AuCtCC6H4CtCAu]n20c poly-
mers failed even when an excess of the electron-rich pyridyl
ligands were used.
Due to the difficulties encountered using method a, we
examined an alternative approach for the synthesis of the
pyridyl gold(I) alkynyl complexes (Scheme 1, method b). The
known [Au(py0)2](AuCl2) complexes7 4a-d were added to a
dichloromethane solution containing one of the alkynes, 1 or
2, in the presence of triethylamine at room temperature.
However, as with method a, only the pyridyl gold(I) alkynyl
complexes that contained electron-rich pyridine ligands (1b,c
and 2b,c) could be isolated. When the 4-pic- or py-containing
[Au(py0)2](AuCl2) complexes were used, the polymeric [Au(Ct
CAr)]n materials 1 or 2 were isolated from the reaction
mixtures in quantitative yields.
are only formed when a large excess of these weaker donor
ligands is present. Coates and Parkin report the isolation of
the pyridine-containing complex 1a (which crystallized with
free pyridine),19 using a method similar to that which we have
employed; however, in our hands we have been unable to
isolate an appreciable quantity of 1a. In the presence of a
large excess of the pyridyl ligands, the pyridine-containing
complexes form in solution, as evidenced by positive ion
ESMS,23 and can be precipitated as white or orange solids,
but when the solids are removed from the mother liquor and
dried, the pyridyl ligands are lost and the polymeric
[Au(CtCAr)]n starting materials regenerated.24 This beha-
vior is not uncommon for alkynyl complexes of the coinage
metal group;for example, the complex PhCtCCuNH3 is
able to be isolated, but it is so unstable that the ammonia is
lost when the complex is dried under a stream of nitrogen,
affording the starting material [Cu(CtCPh)]n.25 Likewise,
iPrNH2 is lost from the unstable complex PhCtCAgNH2iPr
by pumping at room temperature.26 Fortunately, the slow
diffusion of diethyl ether into an acetone solution of
[Au(CtCPh)]n and an excess of 3d produced a small number
of crystals (in a 5-10% yield) that were stable enough for
single-crystal X-ray analysis (vide infra), but the poor yield
obtained using this variation means that it is not a synthe-
tically viable approach to these molecules. We postulate that
the strong σ donation from the alkynyl ligand trans to the
pyridyl ligand weakens the Au-py bond, rendering the
pyridyl donor ligands labile. Due to this, the weakly bonded,
less basic pyridyl ligands are readily lost from the complexes
during the attempted isolation of the complexes.
The isolable mononuclear pyridyl compounds 1b, 2b, 1c,
and 2c showed moderate stability in the solid state27 but
decompose in solutions of strongly coordinating solvents
(e.g., d6-DMSO). All the isolable compounds were fully char-
acterized by NMR (1H and 13C) and IR spectroscopy, HR-
electrospray mass spectrometry (ESMS), and elemental
analysis. In addition, 1c,d and 2b were characterized by
single-crystal X-ray diffraction (see below).
The results above suggest that the described synthetic
approaches only provide stable isolable pyridyl gold(I) alky-
nyl complexes when strongly electron donating substituents
(e.g., 4-dmap and 4-NH2Py) are present in the pyridyl ligands
and this is supported by DFT calculations (vide infra). These
very basic ligands are strong donor ligands and are able to
readily displace the π-coordinated alkyne from the coordi-
nation sphere of the gold. When weaker pyridyl bases are
employed as ligands (e.g., 3a,d-f) the N-donor strength is
not of sufficient magnitude to displace the π-coordinated
alkyne group. The pyridyl-containing gold(I) alkynyl complexes
The 1H NMR spectra of the gold complexes 1b, 2b, 1c, and
2c are in agreement with the proposed mononuclear structures.
1
As an illustrative example, the H NMR spectra (aromatic
region) of PhCtCAu(4-dmap) (1b) is shown in Figure 2.
Interestingly, the β-protons of the 4-dmap ligand are shifted
downfield, as expected upon coordination to the gold(I) ion;
however, the R-protons are shifted upfield (Figure 2b).
Similar behavior is observed in the gold(I) complex 4b, which
is known to aggregate in solution;7 as such, we postulate that
this may be due to aggregation of the monomeric complexes
in solution. Furthermore, the positions of the signals in the
1H NMR spectra are very solvent dependent and the chemi-
cal shift of H3 (the ortho proton on the phenyl ring) in
different solvents suggests that the phenylalkyne moiety may
be forming a weak C-H- - -π intermolecular interaction with
the pyridine ligand on an adjacent complex in solution, as is
observed in the solid-state structure of 1c (vide infra, Figure 4a).
The 13C NMR spectra of 1b,c and 2b,c exhibit resonances
due to the CtC moiety between 95 and 110 ppm, values that
are similar to those for other gold(I) alkynyl complexes.6
The CtC stretching band in the IR spectra (KBr) of the
pyridyl gold alkynyl complexes was observed to range from
(21) Weast, R. C., Ed. CRC Handbook of Chemistry and Physics, 68th
ed.; CRC Press: Boca Raton, FL, 1987; 2464 pp.
(22) Pham-Tran, N.-N.; Bouchoux, G.; Delaere, D.; Nguyen, M. T.
J. Phys. Chem. A 2005, 109 (12), 2957–2963.
(23) Positive ion ESMS of the reaction mixtures indicates that the
pyridyl gold(I) alkynyl complexes are formed in solution; a series of
peaks corresponding tþo [M þ H]þ, [M þ Na]þ, and [M þ K]þ, along with
the cationic [Au(Py)2] fragment ion, are observed.
(24) (a) At the request of a reviewer, we have attempted to obtain a 1H
NMR spectrum of 1a. As stated in the text, we have been unable to
isolate a solid sample of 1a, due to its instability; as such we have
attempted to generate the compound in situ. A CDCl3 solution of the
DMAP complexþ1b was treated with 1 equiv of pyridinium p-toluene-
sulfonate ([PyH ][OTs-]). This resulted in ligand exchange and the
immediate formation of a white precipitate. The 1H NMR spectrum of
the resulting mixture shows peaks due to [DMAPHþ][OTs-] and free
pyridine. The insoluble white solid was isolated by filtration, and
microanalysis confirmed that it was [Au(CtCPh)]n (1). These results
suggest that 1a is formed in solution via ligand exchange and then
rapidly decomposes to free pyridine and 1. This is completely consistent
with our other data on the stability of 1a. Other workers have previously
shown that treatment of DMAP-containing complexes with protonated
pyridine leads to protonation of the more basic DMAP ligand and
coordination of the Py donor group. See: (b) Leigh, D. A.; Lusby, P. J.;
McBurney, R. T.; Symes, M. D. Chem. Commun. 2010, 46 (14), 2382–
2384. (c) Crowley, J. D.; Leigh, D. A.; Lusby, P. J.; McBurney, R. T.;
Perret-Aebi, L.-E.; Petzold, C.; Slawin, A. M. Z.; Symes, M. D. J. Am.
Chem. Soc. 2007, 129 (48), 15085–15090.
(25) Nast, R.; Pfab, W. Chem. Ber. 1956, 89, 415–421.
(26) Blake, D.; Calvin, G.; Coates, G. E. Proc. Chem. Soc. 1959, 396–397.
(27) In the solid state the complexes darken to a purple color
(decompose) over a period of 2-3 months in the absence of light.

Article
Organometallics, Vol. 29, No. 23, 2010 6191
Figure 2. Partial 1H NMR spectra (400 MHz, CD3CN, 298 K)
of (a) the ligand 3b, (b) the gold(I) complex 1b, and (c) the gold(I)
complex 4b.7
2116 to 2125 cm-1, which is similar to values for analogous
primary and secondary amine containing gold(I) alkynyl com-
pounds (2120-2134 cm-1) reported by Coates and Parkin.19
Positive-ion ESMS data also supported the formation of
the neutral mononuclear structures, with the spectra domi-
nated by a series of peaks corresponding to [M þ H]þ,
[M þ Na]þ, and [M þ K]þ, along with the cationic
[Au(Py)2]þ fragment ion. The observation of the [Au(Py)2]þ
ion may indicate the presence of an equilibrium between the
neutral monomeric [ArCtCAu(Py)] and the dimeric ion-paired
species [Au(Py)2][Au(CtCAr)2] in solution, as is observed
for the related [Au(Py)Cl] system.7 However, the [Au(CtCAr)2]-
anion is not detected in negative ion ESMS experiments, and
molar conductivity (ΛM) values in acetonitrile are very low
(6-7 Ω-1 cm2 mol-1), suggesting that the complexes are
nonelectrolytes.91 In combination, these results indicate that
the major gold(I) species present in solution are the neutral
monomeric [ArCtCAu(Py)] complexes.
The electrochemistry of 2b in dichloromethane solution
was investigated using cyclic and square wave voltammetric
techniques, and the results are presented in the Supporting
Information. These show the predicted one-electron chemi-
cally reversible oxidation of the ferrocenyl group at E° =
0.51 V. This value compares well with values reported previ-
ously of 0.53 and 0.41 V for examples of platinum28 and
nickel29 ferrocenyl acetylides, respectively.
Molecular Structures. X-ray crystallography was used to
unequivocally determine the molecular structures of the pyridyl
gold(I) alkynyl complexes 1c,d and 2b (Figure 3, Table 1), at
least in the solid state. Each of the complexes crystallizes in a
monoclinic space group (P21/c or P21/n) with four molecules
in the unit cell. The Au(I) ions are bound to a pyridyl and an
alkynyl donor in the expected linear fashion (N-Au-CtC
bond angles range from 173.7 to 179.8°). The Au-N bond
Figure 3. Molecular structures of 1c (a, b), 1d (c, d), and 2b (e, f)
as ORTEP diagrams. The thermal ellipsoids are shown at the
˚
50% probability level. Selected bond lengths (A) and angles
(deg) for 1c: Au1-C1 = 1.952(11), Au1-N1 = 2.063(9), C1-
C2 = 1.179(14); C1-Au1-N1 = 173.7(4), C2-C1-Au1 =
170.8(10). Selected bond lengths (A) and angles (deg) for 1d:
˚
Au1-C1 = 1.955(12), Au1-N1 = 2.069(10), C1-C2 = 1.219(17);
C1-Au1-N1 = 179.8(4), C2-C1-Au1 = 178.9(11). Selected
˚
bond lengths (A) and angles (deg) for 2b: N1-Au1 = 2.044(11),
C8-Au1 = 1.922(18), C8-C9 = 1.22(2); C9-C8-Au1 =
176.2(11), C8-Au1-N1 = 178.9(4).
lengths (2.025-2.069 A)7-9 and Au-CtCAr (1.922-1.955
˚
A)5c,6c,6d,20b,20c,30 bond lengths are similar to those observed
˚
in other gold(I) complexes. Each of the molecules forms a
dimer in the solid state. However, surprisingly, the dimers are
not stabilized by gold(I)-gold(I) interactions.31
(28) McAdam, C. J.; Blackie, E. J.; Morgan, J. L.; Mole, S. A.;
Robinson, B. H.; Simpson, J. Dalton Trans. 2001, No. 16, 2362–2369.
(29) Butler, P.; Gallagher, J. F.; Manning, A. R.; Mueller-Bunz, H.;
McAdam, C. J.; Simpson, J.; Robinson, B. H. J. Organomet. Chem.
2005, 690 (21-22), 4545–4556.
(30) (a) Gao, L.; Partyka, D. V.; Updegraff, J. B., III; Deligonul, N.;
Gray, T. G. Eur. J. Inorg. Chem. 2009, No. 18, 2711–2719. (b) Siemeling,
U.; Rother, D.; Bruhn, C. Chem. Commun. 2007, No. 41, 4227–4229.
(c) Yip, S.-K.; Cheng, E. C.-C.; Yuan, L.-H.; Zhu, N.; Yam, V. W.-W. Angew.
Chem., Int. Ed. 2004, 43 (37), 4954–4957.
(31) We also obtained the crystal structure of 1b (see the Supporting
Information). While we were readily able to determine the connectivity
of 1b, we have been unable to satisfactorily refine the data, despite
collecting multiple data sets from several different crystals of reasonable
quality. In all cases twinning was observed, which led to poor-quality
diffraction data. As observed with the other structures, 1b is a neutral
linear two-coordinate complex that dimerizes in the solid state. In this
case the formation of the dimer is stabilized by an Au(I)-Au(I) inter-
˚
action (Au-Au distance 3.19 A).

6192 Organometallics, Vol. 29, No. 23, 2010
Kilpin et al.
Figure 4. Ball-and-stick and space-filling representations of the solid-state dimers of 1c (a), 1d (b), and 2b (c).

Article
The complex 1c forms a “head-to-tail” dimer that is
Organometallics, Vol. 29, No. 23, 2010 6193
stabilized by a C-H-π interaction between the phenylace-
tylide group on one molecule of 1c and the 4-aminopyridine
of the neighboring complex (Figure 4a). The phenylacetylide
and 4-aminopyridine ligands of 1c are not coplanar as
expected (Figure 3b), because the phenylacetylide ligand
has to twist out of the plane in order to form the C-H-π
interaction. Although the gold(I) ions of the adjacent mole-
˚
cules are aligned, the Au(I)-Au(I) distance (3.705 A) is
˚
outside of the range (2.5-3.5 A) usually considered to
constitute an aurophilic interaction (Figure 4a).5f In con-
trast, the phenylacetylide and 4-picoline ligands of 1d are
essentially coplanar (Figure 3d). Two molecules of 1d are
cofacially stacked in a “head-to-tail” orientation (Figure 4b),
forming a dimer that is stabilized by a weak offset face-to-
face π-π interaction32 (centroid (4-pic)-centroid (4-pic)
33
˚
distance 3.945 A) and a weak Au-π interaction (Au-
˚
centroid (4-pic) distance 4.074 A, R = 20.74°). In 2b the
planes of the ferrocenyl and 4-dmap ligands are almost
orthogonal (the angle between the calculated mean planes
is 77.12°, Figure 3e,f). Two molecules of 2b form a “head-to-
tail” dimer that is supported by a weak lone-pair-π inter-
Figure 5. Measured FT-Raman spectra in MeCN and KBr and
the calculated Raman spectrum scaled by 0.9750. Major peaks
and equivalent calculated peaks are marked.
˚
action (N-centroid distance 3.791 A) between the lone pair
Table 2. MAD Values for PhCtCAu(4-dmap) (1b) and
PhCtCAu(4-NH2Py) (1c) in Vacuo and in Solutiona
of the NMe2 group on one complex with the 4-dmap ligand
of the adjacent molecule (Figure 4c).34
MAD/cm-1
DFT and Spectroscopy. To provide further insight into the
electronic and thermodynamic properties of the pyridyl
gold(I) alkynyl complexes, we have carried out DFT calcula-
tions for compounds 1a-c,g as well as a tetrameric model35
of the [Au(CtCPh)]n precursor, both in solvent and in vacuo.
We have previously shown that vibrational spectroscopy
is an excellent validation tool for gauging the effectiveness
and accuracy of quantum-mechanical calculations, and for
that reason frequency calculations were performed.36 For
example, the Raman activity of a molecule is closely related
to the polarizability of the valence electrons, while correct
modeling of the energies and reactivities is crucially depen-
dent on the nature and energetics of these. As such, we
screened for an appropriate basis set by comparing the
calculated Raman spectra to the experimentally obtained
spectrum (Supporting Information). A qualitative compar-
ison of the calculated and measured Raman spectra of
PhCtCAu(4-dmap) (1b) is shown in Figure 5. The Raman
spectrum is dominated by the CtC stretch at 2124 cm-1
(not shown) and phenyl-based vibrations between 500 and
IR
6.5
Raman
CtC
CtC scaled
1b (in vacuo)
7.0
7.0
7.9
7.4
59
43
67
35
18
1
26
0
1b (in CH3CN)
1c (in vacuo)
1c (in CH3CN)
7.6
a The ECP used in all cases is LANL2DZ. The in vacuo calculations
are compared with measurements carried out in KBr. CtC refers to the
use of a scale factor optimized in the 400-1700 cm-1 region, while
“CtC scaled” uses an alkyne-specific scale factor of 0.9568.
in vacuo and in solution (CH3CN), and also includes CtC
stretches scaled by a factor of 0.9568, as recommended by the
NIST Computational Chemistry Comparison and Bench-
mark Database (CCCBD).37 It appears that solvent-based
calculations yield only a small improvement for peaks below
1700 cm-1; however, the CtC stretch is far better modeled
and shows an excellent fit of experimental to calculated
wavenumbers when alkyne-specific scale factors are used.
As the structure of PhCtCAu(4-dmap) (1b) is relatively
simple and there is a direct connection of the CtC bond to
the Au atom, comparison of the calculated alkyne stretching-
wavenumbers affords a useful test for the effectiveness of the
various ECPs employed. The results can be found in the
Supporting Information (Table S1 and Figure S11). While
LANL2DZ showed the best performance overall when peaks
below 1700 cm-1 are considered, the CtC stretch is approxi-
mated rather poorly. However, this peak is easily assigned
and thus perhaps less important in terms of identification of
normal modes. Looking at intensity patterns in Figure S11, it
is evident that most basis sets perform very similarly. The
exception is the LANL1 ECPs, which generally show greatly
diminished intensities of peaks with significant metal involve-
ment. Overall, the results for most ECPs can be considered
usable (i.e., MAD < 10 cm-1).38
1700 cm-1. 4-dmap peaks are observed at 767 and 1076 cm-1
.
It should be noted that while the solution-phase spectrum
shows a very poor signal-to-noise ratio, several major peaks
are observed at wavenumbers expected from the solid-state
spectrum. An exception is the peak at 1542 cm-1, which is
shifted to 1553 cm-1 in solution. Table 2 shows the calculated
mean absolute deviations (MADs) for compounds 1b,c, both
(32) (a) Janiak, C. Dalton Trans. 2000, No. 21, 3885–3896. (b) Hunter,
C. A.; Lawson, K. R.; Perkins, J.; Urch, C. J. J. Chem. Soc., Perkin Trans. 2
2001, No. 5, 651–669.
(33) Tiekink, E. R. T.; Zukerman-Schpector, J. CrystEngComm 2009,
11 (7), 1176–1186.
(34) Mooibroek, T. J.; Gamez, P.; Reedijk, J. CrystEngComm 2008,
10 (11), 1501–1515.
(35) To simplify the calculation, the polymeric gold precursor 1 was
approximated by a tetrameric complex where the gold atoms bind end-
on to the alkyne group of a neighboring complex to give a square
arrangement (see the Supporting Information).
(37) Johnson, R. D., III. NIST Computational Chemistry Compar-
ison and Benchmark Database; NIST Standard Reference Database Number
101 Release 15: http://cccbdb.nist.gov/, February 2010.
(36) Walsh, P. J.; Gordon, K. C.; Lundin, N. J.; Blackman, A. G. J.
Phys. Chem. A 2005, 109 (26), 5933–5942.

6194 Organometallics, Vol. 29, No. 23, 2010
Kilpin et al.
appears to be a manifestation of the modification of the
Hartree-Fock (HF) exchange with increasing distance in
CAM-B3LYP.12,39 The more localized unoccupied molecu-
lar orbitals are expected to be of higher energies than their
delocalized B3LYP counterparts, which is not inconsistent
with the observed increase in transition energies.
Both methods calculated the lowest energy transition as a
transfer of an electron from the highest occupied molecular
orbital (HOMO) to the lowest unoccupied molecular orbital
(LUMO); however, with B3LYP it appears that a HOMO/
HOMO-1 f LUMOþ1 transition also contributes. The
HOMO f LUMO transition can also be described as a
π f π* transition, where the π and π* orbitals are phenyl-
based and delocalized, respectively (Figure 7). This transi-
tion is measured in the absorption spectrum at 279 nm, with a
red shoulder at ca. 291 nm; this represents a good approx-
imation by the B3LYP calculation, which shows peaks at 276
and 280 nm. It is also comparable to the UV-visible spectra
of PhCtCAu(PPh3)40 and 4-pyCtCAu(PPh3).41 One might
expect that the intuitively predicted ground- and excited-
state orbitals would appear as Au-alkynyl-based and pyr-
idyl-based, respectively, and that orbital scrambling is re-
sponsible for the deviation observed.42 A single-point triplet-
state calculation carried out on the ground-state geometry of
the complex shows that the lowest energy singly occupied
molecular orbital is located across the Au-alkynyl-phenyl
portion of the complex, while several higher-energy MOs are
pyridyl-based. A higher energy band observed at 258 nm has
been calculated by B3LYP TD-DFT as a transition solely
originating from the HOMO-1 orbital and terminating on
the LUMO and LUMOþ1 orbitals, which is represented as a
π4-dmap f π*delocalized electron transfer by the molecular
orbital pictures (Figure 7). A weak feature is also observed
at ca. 360 nm. This could be due to a spin-forbidden direct
excitation of the triplet state; however, no corresponding
peaks could be observed in the photoluminescence spectrum.
The absorption spectrum (Supporting Information) of the
compound FcCtCAu(4-dmap) (2b) shows a peak at 281 nm
(ε = 36 ꢀ 103 L mol-1 cm-1) and shoulders at ca. 427 nm
(ε = 1.6 ꢀ 103 L mol-1 cm-1) and 350 nm (ε = 3.6 ꢀ 103 L
mol-1 cm-1). The energies of these are consistent with
ferrocene d-d transitions previously found at 416 and
325 nm43 and suggest limited interaction between Fc and
the residual molecule, as indicated by the electrochemical
measurements on 2b.
Figure 6. UV-visible absorption (black) and emission (green)
spectra of PhCtCAu(4-dmap) (1b) at 298 K in CH3CN. TD
transitions calculated using B3LYP (orange) and CAM-B3LYP
(blue) are overlaid. The emission units are arbitrary.
Table 3. Energy Change (ΔEform) for the Reaction of the Pyridyl
Compounds with the Gold Precursor, Listed in Descending Order
of Electron-Donating Ability
pyridyl gold(I) alkynyl
complex
ΔEform in vacuo/kJ
pyridyl ligand
pKa
mol-1
PhCtCAu(4-DMAP) (1b)
PhCtCAu(4-NH2Py) (1c)
PhCtCAu(Py) (1a)
-4.5
-3.7
5.8
9.40
9.11
5.25
1.86
PhCtCAu(4-CNPy) (1g)
27.1
Having established the usefulness of LANL2DZ as an
ECP for modeling the gold atoms in these compounds, we
then used the single-point energies of the previously opti-
mized structures to calculate the energies of formation
(ΔEform) for the pyridyl gold(I) alkynyl complexes 1a-c,g
(Table 3 and the Supporting Information). Consistent with
the experimental data, the calculated energies of formation
(ΔEform) indicate that the pyridyl gold(I) alkynyl complexes
1b,c, which contain the most electron-donating pyridyl
ligands, are thermodynamically stable with respect to the
reactants, whereas the complexes 1a,g, which contain less
electron donating pyridyl ligands, are unstable with respect
to the reactants (Table 3).
TD-DFT calculations were performed using the B3LYP
and CAM-B3LYP methods; the results for PhCtCAu(4-
dmap) (1b) are presented in Figure 6 and in the Supporting
Information (Table S2). For both methods we employ the
IEF-PCM solvent model. It should be noted that virtually
the same transitions were calculated whether the solvated
or in vacuo equilibrium geometry was considered. From
Figure 6 it is evident that CAM-B3LYP overestimates
transition energies for solvent calculations. While the occu-
pied frontier orbitals calculated with B3LYP and CAM-
B3LYP are almost identical, there are significant differences
in the unoccupied or virtual orbitals (see Figure 7). Here, the
B3LYP orbitals are calculated to be somewhat more delo-
calized over the entire molecule, whereas the CAM-B3LYP
orbitals tend to be confined to one side of the molecule. This
Solution-phase emission spectra were acquired for PhCt
CAu(4-dmap) (1b) and compound 2b at excitation wave-
lengths of 280 nm. The former is weakly emissive, showing a
peak at 421 nm with some fine structure (see Figure 6), while
the latter shows no detectable emission from 400 to 800 nm.
We assign the 421 nm band observed for PhCtCAu(4-
dmap) (1b) as a triplet excited-state emission. The absorption-
emission peak-to-peak separation is ca. 12 000 cm-1, which is
significantly larger than for singlet-state emission investi-
gated by Gao et al. on a similar compound.30a Furthermore,
(39) Vlcek, A.; Zalis, S. Coord. Chem. Rev. 2007, 251 (3 þ 4), 258–287.
(40) Li, D.; Hong, X.; Che, C. M.; Lo, W. C.; Peng, S. M. J. Chem.
Soc., Dalton Trans. 1993, 19, 2929–32.
(41) Ferrer, M.; Gutierrez, A.; Rodriguez, L.; Rossell, O.; Lima, J. C.;
Font-Bardia, M.; Solans, X. Eur. J. Inorg. Chem. 2008, 18, 2899–2909.
(42) (a) We thank a reviewer for pointing out this interpretation.
(b) Elbjeirami, O.; Yockel, S.; Campana, C. F.; Wilson, A. K.; Omary,
M. A. Organometallics 2007, 26 (10), 2550–2560.
(38) (a) McGoverin, C. M.; Walsh, T. J.; Gordon, K. C.; Kay, A. J.;
Woolhouse, A. D. Chem. Phys. Lett. 2007, 443 (4-6), 298–303.
(b) Horvath, R.; Gordon, K. C. Coord. Chem. Rev. 2010, 254 (21-22),
2505–2518.
(43) Boulet, P.; Chermette, H.; Daul, C.; Gilardoni, F.; Rogemond,
F.; Weber, J.; Zuber, G. J. Phys. Chem. A 2001, 105 (5), 885–894.

Article
Organometallics, Vol. 29, No. 23, 2010 6195
Figure 7. Frontier orbitals of PhCtCAu(4-dmap) (1b), as calculated by CAM-B3LYP and B3LYP DFT methods.
the gap between the lowest energy triplet state and the
ground state has been calculated at 2.913 eV in vacuo, which
corresponds to a wavelength of 426 nm and is not incon-
sistent with the observed data.44 The fine-structure spacing is
1200 cm-1, which is comparable to skeletal ring-stretching
modes observed in the FT-IR and FT-Raman spectra. The
lack of emission observed for compound 2b is due to intra-
molecular quenching by the ferrocenyl moiety, which is not
uncommon.45
When the samples were subjected to time-resolved emis-
sion measurements, no transients could be detected for either
PhCtCAu(4-NH2Py) (1c) or PhCtCAu(4-dmap) (1b). This
may be due to low absorbance at the pump wavelength
(355 nm) leading to very limited excited-state population,
coupled with weak emission; transient absorption data were
observed. These transient signals showed lifetimes of 45((7)
ns for PhCtCAu(4-NH2Py) (1c) and 22((5) ns for PhCt
CAu(4-dmap) (1b). Despite thorough degassing using argon,
sample degradation was observed for both compounds, but as
it was more severe for the former, only a small data set could
be collected. The photolability of the samples meant that only
a cursory transient absorption spectrum could be collected for
PhCtCAu(4-dmap) (1b). This revealed two transient absorp-
tion peaks, centered at ca. 500 and 680 nm, both showing a
22 ns lifetime. Unfortunately, sample degradation precluded
experiments with extended acquisition times, from which we
may have been able to observe weak phosphorescence decay.
structures are stabilized by aurophilic interactions. Electronic
absorption spectroscopy exhibited peaks at ca. 280 nm for
compounds 1b and 2b. It was found that DFT calculations
using B3LYP and the 6-31G(d) and LANL2DZ basis sets for
“light atoms” and gold, respectively, are adequate in model-
ing the vibrations and electronic transitions. Little improve-
ment could be observed upon varying the effective core
potential; similarly, using the CAM-B3LYP functional for
time-dependent calculations did not yield a better approx-
imation to the absorption spectra. Triplet emission observed
for compound 1b at 421 nm was quenched in the ferrocene-
containing compound 2b. DFT calculations support the
experimental finding that only the pyridyl gold(I) alkynyl
complexes that contain very electron rich pyridine ligands
are thermodynamically stable with respect to the reactants.
The limited structural diversity and stability issues asso-
ciated with these molecules along with the absence of Au(I)-
Au(I) interactions in the solid-state structures suggests that it
will be difficult to exploit these molecules as supramolecular
synthons for the construction of self-assembled nanoscale
materials.
Acknowledgment. We thank Prof. Lyall Hanton for
useful discussions, Dr. Mervyn Thomas for his assistance
collecting NMR data, and Dr. John McAdam for his
assistance obtaining electrochemical data. A University
of Otago Research Grant and the Department of Chem-
istry, University of Otago provided financial support for
this work. The MacDiarmid Institute for Advanced
Materials and Nanotechnology is thanked for providing
funding for the purchase of a X-ray optics and detector
and the Allan Wilson Centre for Molecular Ecology and
Evolution for the purchase of the microfocus rotating-
anode X-ray generator.
Conclusions
We have synthesized and structurally characterized a family
of pyridyl gold(I) alkynyl complexes. The described synthetic
approaches only provide stable isolable complexes when
electron-rich pyridyl ligands such as 4-(dimethylamino)pyri-
dine and 4-aminopyridine are incorporated into the complexes.
In the solid state the pyridyl gold(I) alkynyl complexes
form dimers; however, somewhat surprisingly none of the
Supporting Information Available: Text, figures, tables, and
CIF files giving full experimental descriptions and spectroscopic
data for 1b,c and 2b,c, crystallographic data for 1b,c and 2b,c
along with figures and packing diagrams, and additional com-
putation data and figures showing the optimized structures.
This material is available free of charge via the Internet at http://
pubs.acs.org. Crystallographic data for 1b,c and 2b,c are also
available from the Cambridge Crystallographic Database (CCDC
774529-774532).
(44) This quantity was obtained by comparison of TD-DFT calcula-
tions using the B3LYP method of the singlet and triplet states, using
singlet and triplet geometries, respectively.
(45) (a) Thornton, N. B.; Wojtowicz, H.; Netzel, T.; Dixon, D. W. J.
Phys. Chem. B 1998, 102 (11), 2101–2110. (b) Ho, S. Y.; Cheng, E. C.-C.;
Tiekink, E. R. T.; Yam, V. W.-W. Inorg. Chem. 2006, 45 (20), 8165–
8174.