Rare, hexatomic, boat-shaped, cross-linked bis(iminodiphenylphosphorano) methanediide pincer carbon bridged photoluminescent copper clusters capped with methyl or halide bridges

Article abstract of DOI:10.1021/om1002478

The dimeric dilithium methanediide salt [Li₂C(Ph ₂P=NSiMe₃)₂]₂ ([Li ₂L]₂, L = [C(Ph₂P=NSiMe₃) ₂]^2-) reacted with 6 equiv of [NEt₄CuCl ₂] and 2 equiv of LiMe in THF or ether to give exclusively the six-copper cluster complex [Cu₆L₂(CH₃) ₂] (1). Similarly, [Li₂L]₂ reacted with 3 equiv of the bimetallic copper halide complexes [(cod)₂Cu ₂X₂] (X = Cl, Br, I) in the same solvent to give good yields of three halide-capped six-copper clusters [Cu₆L ₂(X)₂] (X = Cl (2), Br (3), I (4)) with structures very similar to that of 1. These four hexacopper clusters (1-4) as a family show similar ³¹P NMR spectra with an AB pattern with slightly different chemical shifts but identical coupling constants in solution. The reactivity of 2 was explored, and it was demonstrated that 1 can be quantitatively generated by addition of 2 equiv of LiMe to 2. Moreover, the cluster [Cu₆L ₂(O^tBu)₂] (5) with a similar cage structure is quantitatively generated by reaction of 2 with 2 equiv of NaO^tBu in solution according to ³¹P NMR spectroscopy. An additional copper complex with a bicopper formulation, [Cu₂L(PPh₃) ₂] (6), was synthesized by reaction of [Li₂L]₂ with 2 equiv of [(NEt₄)Cu(PPh₃)Cl₂]. We also observed a monocopper iodide complex with the doubly protonated parent methylene bridged ligand H₂L, [CuI(H₂L)] (7), as a minor component (about 10%) during the preparation of complex 4. Complexes 1-4, 6, and 7 were characterized by X-ray crystal diffraction. All four clusters (1-4) show a rare boat-shaped conformation of hexacopper clusters assembled by cross-linking of two copper atoms via geminal substitution on one bis(iminophosphorano) methanediide ligand. Two of these units combine along with two additional copper atoms to form the boat cluster. There is extensive direct Cu-Cu bonding. The cluster is doubly capped, top and bottom, by methyl groups (1) or halides (2-4, X = Cl, Br, I). All of these six-copper clusters absorb and emit in the UV-vis region. The absorption bands and photoluminescent wavelengths are substituent dependent; thus, optical properties of the complex are tunable by means of substitution of the capping ligands. The spectral and bonding characteristics were explored by means of Gaussian DFT calculations and NBO analysis.

Full text of DOI:10.1021/om1002478

Organometallics 2010, 29, 4251–4264 4251
DOI: 10.1021/om1002478
Rare, Hexatomic, Boat-Shaped, Cross-Linked
Bis(iminodiphenylphosphorano)methanediide Pincer Carbon Bridged
Photoluminescent Copper Clusters Capped with Methyl or Halide Bridges
Guibin Ma,† Michael J. Ferguson,‡ Robert McDonald,‡ and Ronald G. Cavell*,†
†Department of Chemistry and ‡X-ray Structure Determination Laboratory,
University of Alberta, Edmonton, Alberta T6G 2G2, Canada
Received March 30, 2010
The dimeric dilithium methanediide salt [Li2C(Ph2PdNSiMe3)2]2 ([Li2L]2, L=[C(Ph2P=NSiMe3)2]2-
)
reacted with 6 equiv of [NEt4CuCl2] and 2 equiv of LiMe in THF or ether to give exclusively the six-copper
cluster complex [Cu6L2(CH3)2] (1). Similarly, [Li2L]2 reacted with 3 equiv of the bimetallic copper halide
complexes [(cod)2Cu2X2] (X = Cl, Br, I) in the same solvent to give good yields of three halide-capped six-
copper clusters [Cu6L2(X)2] (X=Cl (2), Br (3), I (4)) with structures very similar to that of 1. These four
hexacopper clusters (1-4) as a family show similar 31P NMR spectra with an AB pattern with slightly
different chemical shifts but identical coupling constants in solution. The reactivity of 2 was explored, and it
was demonstrated that 1 can be quantitatively generated by addition of 2 equiv of LiMe to 2. Moreover, the
cluster [Cu6L2(OtBu)2] (5) with a similar cage structure is quantitatively generated by reaction of 2 with
2 equiv of NaOtBu in solution according to 31P NMR spectroscopy. An additional copper complex with a
bicopper formulation, [Cu2L(PPh3)2] (6), was synthesized by reaction of [Li2L]2 with 2 equiv of
[(NEt4)Cu(PPh3)Cl2]. We also observed a monocopper iodide complex with the doubly protonated parent
methylene bridged ligand H2L, [CuI(H2L)] (7), as a minor component (about 10%) during the preparation
of complex 4. Complexes 1-4, 6, and 7 were characterized by X-ray crystal diffraction. All four clusters
(1-4) show a rare boat-shaped conformation of hexacopper clusters assembled by cross-linking of two
copper atoms via geminal substitution on one bis(iminophosphorano)methanediide ligand. Two of these
units combine along with two additional copper atoms to form the boat cluster. There is extensive direct
Cu-Cu bonding. The cluster is doubly capped, top and bottom, by methyl groups (1) or halides (2-4,
X = Cl, Br, I). All of these six-copper clusters absorb and emit in the UV-vis region. The absorption bands
and photoluminescent wavelengths are substituent dependent; thus, optical properties of the complex are
tunable by means of substitution of the capping ligands. The spectral and bonding characteristics were
explored by means of Gaussian DFT calculations and NBO analysis.
Introduction
polynuclear Cu(I) carbon bound compounds with a variety of
bridged ligands have been extensively reported.2 Current inter-
est centers on photoluminescence properties which offer poten-
tial for applications in photovoltaic and optoelectronic devices.
Several previous studies have reported polynuclear copper(I)
clusters constructed with tetrahedral cores, octahedral cores, or
bridging ligands based on group 15 or 16 elements.3 A few
examples of copper(I) clusters with alkynyl ligands are also
known.4 To the best of our knowledge, there are no known
copper clusters supported by bridging carbon centers or carbon-
bound pincer chelate ligands.
The design and synthesis of unique d10 metal copper(I)
complexes in either monomeric or polynuclear forms with
subsequent exploration of their photophysical properties
and applications for selected catalysis functions continues
to be an active research area.1 A large variety of possible
metal-metal and metal-ligand bonding combinations
can be assembled, often creating interesting and sometimes
unique chemical properties. In addition, some luminescent
We have used the dilithiated bis(diphenyltrimethylsilyl-
iminophosphorano)methanediide salt [Li2L]2 (or the parent
methylene ligand H2L) in our laboratory to synthesize both
monomeric and bimetallic bridged carbon metal complexes.5
Particularly successful were the unique spirocyclic bimetallics6a,b
of Pd(II) and Rh(I) (Scheme 1). In addition, we recently reported
a spirocyclic dimetalated carbon Tl(I) complex of this same
ligand which dimerized to a tetrathallated cluster through the
formation of Tl(I)-Tl(I) bonds.6c Herein we report a set of
copper(I) hexatomic metal clusters with a unique, irregular
*To whom correspondence should be addressed. Tel: 780-492-5310.
Fax: 780-492-8231. E-mail: ron.cavell@ualberta.ca.
(1) (a) Ohishi, T.; Nishiura, M.; Hou, Z.-M. Angew. Chem., Int. Ed.
2008, 47, 5792–5795. (b) Badiei, Y. M.; Dinescu, A.; Dai, X.-L.; Palomino,
R. M.; Heinemann, F. W.; Cundari, T. R.; Warren, T. H. Angew. Chem., Int.
Ed. 2008, 47, 9961–9964. (c) Badiei, Y. M.; Krishnaswamy, A.; Melzer,
M. M.; Warren, T. H. J. Am. Chem. Soc. 2006, 128, 15056–15057.
(d) Zhang, X.-M.; Fang, R.-Q.; Wu, H.-S. J. Am. Chem. Soc. 2005, 127,
7670–7671. (e) Xie, H.-Y.; Kinoshita, I.; Karasawa, T.; Kimura, K.; Nishioka,
T.; Akai, I.; Kanemoto, K. J. Phys. Chem. B 2005, 109, 9339–9345. (f) Che,
C.-M.; Mao, Z.; Miskowski, V. M.; Tse, M.-C.; Chan, C.-K.; Cheung, K.-K.;
Phillips, D. L.; Leung, K.-H. Angew. Chem., Int. Ed. 2000, 39, 4084–4088.
r
2010 American Chemical Society
Published on Web 09/07/2010
pubs.acs.org/Organometallics

4252 Organometallics, Vol. 29, No. 19, 2010
Ma et al.
Scheme 1. Hetero- and Homobimetallic Spirocyclic Complexes
of Bis(diphenyltrimethylsilyliminophosphorano)methanediide
and a Tl(I) Dimeric Cluster of the Spirocyclic Complex of
the Same Ligand
Scheme 2. Preparation of Complexes 1-4
structural arrangement formed from an unusual ligating
structure. These complexes also show unusual and potentially
useful optical behavior.
Results and Discussion
We prepared the copper(I) clusters [Cu6L2X2] (L = C(Ph2Pd
NSiMe3)22-; X = CH3- (1), Cl- (2), Br- (3), I- (4)) by react-
ing the appropriate Cu(I) precursors (either CuCl2- salts plus
MeLi or Cu2(cod)X2 (X = Cl, Br, I) with the organodilithium
dimer [Li2C(Ph2PdNSiMe3)2]2, ([Li2L]2) in either THF or
ether solutions (Scheme 2). 31P NMR spectra of all these
Table 1. 31P NMR Chemical Shifts and Spin-Spin Coupling
Constants for Compounds 1-5
(2) (a) Yam, V. W.-W.; Lo, K. K.-W. Chem. Soc. Rev. 1999, 323–334.
(b) Yam, V. W.-W.; Lo, K. K.-W. J. Chem. Soc., Dalton Trans. 1995, 499–
500. (c) Yam, V. W.-W.; Chan, L.-P.; Lai, T.-F. J. Chem. Soc., Dalton Trans.
1993, 2075–2077. (d) Yam, V. W.-W.; Lee, W.-K.; Cheung, K.-K. J. Chem.
Soc., Dalton Trans. 1996, 2335–2339. (e) Jarvis, J. A. J.; Pearce, R.;
Lappert, M. F. J. Chem. Soc., Dalton Trans. 1977, 999–1003. (f) Yam, V.
W.-W.; Choi, S. W.-K.; Chan, C.-L.; Cheung, K.-K. Chem. Commun. 1996,
2067–2068. (g) Yam, V. W.-W.; Cheng, E. C.-C.; Zhu, N.-Y. Chem.
Commun. 2001, 1028–1029. (h) Yam, V. W.-W.; Lam, C.-H.; Cheung,
K.-K. Chem. Commun. 2001, 545–546. (i) Yam, V. W.-W.; Fung, W. K.-M.
Chem. Commun. 1998, 777–778. (j) Yam, V. W.-W.; Fung, W. K.-M.;
Cheung, K.-K. Chem. Commun. 1997, 963–964. (k) Knotter, D. M.; Smeets,
W. J. J.; Spek, A. L.; Van Koten, G. J. Am. Chem. Soc. 1990, 112, 5895–
5896. (l) He, X.-M.; Ruhlandt-Senge, K.; Power, P. P.; Bertz, S. H. J. Am.
Chem. Soc. 1994, 116, 6963–6964. (m) Olmstead, M. M.; Power, P. P. J. Am.
Chem. Soc. 1990, 112, 8008–8014. (n) Knotter, D. M.; Spek, A. L.; van
Koten, G. J. Chem. Soc., Chem. Commun. 1989, 1738–1740. (o) Wehman,
E.; van Koten, G.; Erkamp, C. J. M.; Knotter, D. M.; Jastrzebski, J. T. B. H.;
Stam, C. H. Organometallics 1989, 8, 94–99. (p) Lenders, B.; Grove, D. M.;
van Koten, G.; Smeets, W. J. J.; van der Sluis, P.; Spek, A. L. Organo-
metallics 1991, 10, 786–791. (q) Aalten, H. L.; van Koten, G.; Riethorst, E.;
Stam, C. H. Inorg. Chem. 1989, 28, 4140–4146. (r) Guss, M.; Mason, R.;
Søtofte, I.; van Koten, G.; Noltes, J. G. J. Chem. Soc., Chem. Commun.
1972, 446–447. (s) Chan, W.-H.; Zhang, Z.-Z.; Mak, T. C. W.; Che, C.-M.
J. Organomet. Chem. 1998, 556, 169–172. (t) Koshevoy, I. O.; Karttunen,
A. J.; Tunik, S. P.; Haukka, M.; Selivanov, S. I.; Melnikov, A. S.; Serdo-
bintsev, P. Y.; Khodorkovskiy, M. A.; Pakkanen, T. A. Inorg. Chem. 2008,
47, 9478–9488.
compda
PA (ppm)
PB (ppm)
JA-B (Hz)
solvent
Cu6L2(CH3)2 (1)
Cu6L2(Cl)2 (2)
Cu6L2(Br)2 (3)
Cu6L2(I)2 (4)
25.14
28.24
28.10
28.95
27.58
24.33
26.17
26.59
28.08
23.95
58
56
57
56
62
toluene
THF
THF
THF
THF
Cu6L2(OtBu)2 (5)
a L = [C{Ph2PdNSi(CH3)3}22-].
copper clusters in solution show an AB coupling pattern with
slightly different 31P chemical shifts, indicating nearly chemi-
cally identical but magnetically inequivalent P atoms within
the pincer ligand. The AB coupling constants were deduced by
simulation of the experimental spectra; the parameters for all
are almost identical (Table 1). All four complexes were struc-
turally characterized.
A monocopper complex of the (iminophosphorano)methine
ligand system, [{CH(PPh2NdSiMe3)2}CuPPh3], has recently
been reported.7 It was synthesized by the reaction of potassium
bis(iminophosphono)methanide with [(Ph3P)2CuI]. The crystal
structure shows that the carbon is not directly bonded to the
copper atom in this complex but, rather, shows the existence of a
weak ionic pair interaction. We have successfully prepared a
related bimetallic copper complex of the methanediide [Cu2L-
(PPh3)2] (L = C(Ph2PdNSiMe3)22-) (6) in situ by reacting
[NEt4][Cu(PPh3)Cl2] with [Li2L]2 in THF. The 31P NMR
spectrum of 6 shows one broad peak at 4.06 ppm and one
doublet peak at 19.34 ppm with a P-P coupling constant of 5
Hz, differing from the data for the previously reported com-
plex, indicating that a new compound has formed in the solution.
This new complex was isolated and structurally characterized
(see below).
(3) (a) Drew, M. G. B.; Lavery, A.; McKee, V.; Nelson, S. M.
J. Chem. Soc., Dalton Trans. 1985, 1771–1774. (b) Liu, C. W.; Hung,
C.-M.; Santra, B. K.; Wang, J.-C.; Kao, H.-M.; Lin, Z.-Y. Inorg. Chem.
2003, 42, 8551–8556. (c) Ohlmann, D.; Pritzkow, H.; Grutzmacher, H.;
Anthamatten, M.; Glaser, R. J. Chem. Soc., Dalton Trans. 1995, 1011–
€
1012. (d) Eichhofer, A.; Fenske, D.; Holstein, W. Angew. Chem., Int. Ed.
1993, 32, 242–245. (e) Fenske, D.; Steck, J.-C. Angew. Chem., Int. Ed.
1993, 32, 238–242. (f) Vogler, A.; Kunkely, H. J. Am. Chem. Soc. 1986, 108,
7211–7212. (g) Kutal, C. Coord. Chem. Rev. 1990, 99, 213–252.
(4) (a) Olbrich, F.; Kopf, J.; Weiss, E. Angew. Chem., Int. Ed. 1993,
32, 1077–1079. (b) ten Hoedt, R. W. M.; Noltes, J. G.; van Koten, G.; Spek,
A. L. J. Chem. Soc., Dalton Trans. 1978, 1800–1806.
(5) (a) Cavell, R. G.; Babu, R. P. K.; Aparna, K. J. Organomet. Chem.
2001, 617, 158–169. (b) Jones, N. D.; Cavell, R. G. J. Organomet. Chem
2005, 690, 5485–5496. (c) Cavell, R. G. In The Chemistry of Pincer
Compounds; Morales-Morales, D., Jensen. C. M., Eds.; Elsevier: Amsterdam,
2007; Chapter 14, pp311-346 and references therein.
(6) (a) Fang, M.; Jones, N. D.; Friesen, K.; Lin, G.-Y.; Ferguson,
M. J.; McDonald, R.; Lukowski, R.; Cavell, R. G. Organometallics
2009, 28, 1652–1665. (b) Ma, G.-B.; McDonald, R.; Cavell, R. G. Organo-
metallics 2010, 29, 52–60. (c) Ma, G.-B.; Ferguson, M. J.; Cavell, R. G.
Chem. Commun. 2010, 5370–5372.
It is interesting to note that when we prepared cluster 4
with a reaction stoichiometry of six Cu metals to 1 mol of the
ligand salt, [Li2L]2, complex [CuI(H2L)] (7) was formed in
(7) Panda, T. K.; Roesky, P. W.; Larsen, P.; Zhang, S.; Wickleder, C.
Inorg. Chem. 2006, 45, 7503–7508.

Article
Organometallics, Vol. 29, No. 19, 2010 4253
“bottom” of the cluster. The ligand does not form a complete
“pincer” structure about an individual metal. Two addi-
tional CuX units, each coordinated to the second, available,
nitrogen atom of the ligand, are incorporated by binding
directly, via a Cu(I)-Cu(I) bond, to the carbon substituted
geminal Cu(I) atom, and the appended X substituent (Me or
halide) on this inserted unit then coordinates to another
geminal Cu(I) atom to form Cu-X-Cu bridges spanning
the open face of the initial framework, thereby assembling a
capped boat-shaped Cu6 cluster (Figure 1). Each PCP pincer
ligand is thus bound to two “bottom” copper atoms (on different
units) via carbon and nitrogen atoms, and these copper atoms
bind to each other directly to form a C-Cu-Cu-C edge. Each
of the component (NPCCu2) units is relatively flat. The ligand
carbon bridged group of four Cu(I) atoms thus forms the body
of a boat with the other two Cu(I) atoms of the CuX units
appearing as the “stem” and “stern” (Figures 1 and 2) of the
boat. As shown in Figure 1, the pincer ligands surrounding
the Cu(I)6 core structure of three halide complexes (2-4) have
a similar configuration. Although similarly constructed, the
methyl-bridged complex 1 is different in that the conformation
has a mirror symmetry relative to that of 2-4.
Figure 1. Perspective view of complex 1 showing the atom-
labeling scheme. Only the ipso carbons of the phenyl groups
are shown. Hydrogen atoms (except for those of the bridging
methyl groups) are omitted. Non-hydrogen atoms are repre-
sented by Gaussian ellipsoids at the 20% probability level. Bond
lengths and angles are given in Table 2.
Within the unit cell of the chloride-bridged complex 2,
there are two different cluster molecules showing slightly
different bond distances and angles. These are labeled CuA-
(I)6 and CuB(I)6. In 3 and 4 there is only one cluster
arrangement. When comparing the three halide clusters, we
see that the Cu-X-Cu bond angles irregularly change
through the series Cl to I (average Cu-Cl-Cu=77.57°,
Cu-Br-Cu = 72.83° and Cu-I-Cu = 65.20°) so that the
˚
length trend provides a medium value for the Cl case (2.704 A;2),
˚
the longest value for the Br case (2.718 A; 3), and the shortest
˚
value for the I case (2.656 A; 4). The Cu-C-Cu bond angles
in both the bridged methyl complex 1 and the bromide
complex 3 are similar (Cu-C-Cu = 73.00°); however, the
˚
shorter Cu-C (2.018(3) A) bond distances in 1 compared to
˚
those of the Cu-Br complex 3 (2.2893(6) A) lead to the result
˚
that the shortest Cu-Cu bond distance (2.403(3) A) is found
in structure 1. It has also been found throughout that the
bond distances between the bridging atom (C, Cl, Br, and I)
and the two Cu(I) atoms are not equal; one is always slightly
longer than the other. The average differences are 0.067 (1),
Figure 2. Core structures of copper(I) clusters 1 (top left), 2 (top
right), 3 (bottom left), and 4 (bottom right). For selected bond
˚
distances (A) and angles (deg), see Table 2.
˚
0.021 (2), 0.048 (3) and 0.020 A (4), with the largest difference
situ in approximately 10% concentration along with 4 on the
basis of solution 31P NMR. That solution produced two
differently colored crystals, light green ([CuI(H2L)] (7)) and
red ([Cu6I2L2] (4)). Increasing the copper ratio to 10:1 yielded
7 as the major product, which was readily isolated for crystal
structure analysis (see below).
occurring in 1. Close examination of the structure of 1 shows
that there are dissimilar, close approach bond distances
between the methyl protons and nearby copper(I) atoms;
˚
˚
Cu(5)-H(3A)=1.957 A and Cu(6)-H(4A)=1.850 A,
which indicates the presence of additional agnostic bonding
interactions which may contribute to the differences between
the carbon-copper bonds (see the Supporting Information).
Crystals of all four cluster complexes (1-4) were obtained
either from ether or mixed ether/THF solutions. All are
bright yellow or golden orange. Crystal structures have been
determined for all, and they show a unique structural geometry.
The structure of 1 (which is representative of 1-4) is shown
in Figure 1, and the central cluster core structures of all
four are illustrated in Figure 2. The common feature of the
structures is provided by geminally binding two copper
The intramolecular Cu Cu separations fall in the range
3 3 3
˚
2.4000(4)-2.9458(5) A, which are shorter than or close to the
sum of the van der Waals radii of Cu(I) centers. Separations
increase in the following order: 1 (average separations
˚
2.6493(4) A), 4 (average 2.6956(8) A), 3 (average 2.7251(7)
˚
˚
A), and 2 (average 2.7342(6) for molecule A and average
˚
2.7250(6) A for molecule B). These average Cu Cu separa-
3 3 3
atoms to the central carbon of a methanediide (NdP-C2-
-
tions are all slightly longer than that previously reported for
2-
PdN) ligand. One of these geminal Cu(I) atoms in each unit
is then coordinated to a nitrogen atom from the other
complex unit, and these nitrogen-coordinated geminal Cu(I)
atoms of one unit then bind directly to the similarly nitrogen
coordinated geminal Cu(I) of the other ligand unit by means
of a direct Cu-Cu linkage. This connection forms the
the octahedron of copper atoms in Cu6Fe4(CO)16 anions
8
˚
in the complex [(diphos)2Cu]2Cu6Fe4(CO)16 (average 2.616 A),
but all are substantially shorter than those in the complex
(8) Doyle, D.; Eriksen, K. A.; Van Engen, D. J. Am. Chem. Soc. 1985,
107, 7914–7920.

4254 Organometallics, Vol. 29, No. 19, 2010
Ma et al.
Cu-C = 1.931(5) A in [Cu(CN)(PCy3)]613). However, the dis-
tances in 6 are compatible with four-coordinate copper for
˚
˚
tetrahedral structures (e.g., Cu-C = 2.01(2) A of [Cu(endP2)-
(t-BuNC)](ClO4) (endP2 = N,N0-bis[o-(diphenylphosphino)-
benzylidene]ethylenediamine)14a and Cu-C = 1.988(2) A of
˚
[Cu(CN)(PPh3)2]614b), which indicates that structure 6 has a
four-coordinate copper center. The Cu-Cu distance in 6(Cu(1)-
Cu(2)=2.7217(6) A) is compatible with the observed bond
˚
distances in 2-4 and also shows a weakly bonding interaction.
As we mentioned above, an unexpected single-copper
complex of the parent (iminophosphorano)methane ligand,
7, was detected as a minor component during the preparation
of cluster 4. Green crystals were obtained, and its structure is
illustrated in Figure 4 (selected bond lengths and angles and
data collection parameters are given in Tables 2 and 3). The
PCP bridging carbon is doubly protonated; thus, here the
H2L ligand behaves simply as a bidentate ligand, binding to
copper through two nitrogen atoms forming, along with the
bound I, a three-coordinate copper center. The Cu-N and
Figure 3. Perspective view of complex 6 showing the atom-
labeling scheme. Only the ipso carbons of the phenyl groups
are shown. Hydrogen atoms are omitted. Non-hydrogen atoms
are represented by Gaussian ellipsoids at the 20% probability
level. Bond lengths and angles are given in Table 2.
[Cu6(6-mpyt)6] (6-mpyt=6-methyl-2-pyridinethiolato;
1e
0
˚
Cu-I bond distances (Cu-N(N ) = 2.0480(19) A, Cu-I =
˚
2.831(1)-3.326(2) A). The Cu-Cu separations here are
˚
2.5071(5) A) in 7 are remarkably shorter than those in the
very similar to those in the Cu6 cluster assembly, wherein
a tridentate nitrogen ligand (N,N0,N00-trimethyl-1,3,5-
triazacyclohexane) is supported with a bridging Cl.9 The
two C-Cu bond distances to the PCP-bridged carbon connect-
ing two copper atoms in clusters 1-4are all different, and of note
is that one is slightly shorter than the other. Similarly, the P-N
and P-C bond distances within the chelate ligand are slightly
different, which creates the nonidentical P environments
consistent with the observed AB 31P NMR spectra.
four-coordinate structure15 but are compatible with those of
other reported three-coordinate copper centers of the type
(CuN2I).16 The six-membered ring (Cu-N-P-CH2-P0-N0)
is in a chair configuration, and the four atoms of N, P, P0, and N0
are located exactly in a plane. The mirror plane is preserved
through the three-atom sets, the methylene C, Cu, and I.
All four isolated cluster complexes 1-4 are quite stable in
the solid state under an inert atmosphere at room tempera-
ture, and they seem to be reasonably stable in the open
atmosphere. They can be kept for a few weeks in the air
without any observable color or weight change. In solution,
all complexes are quite stable in the glovebox atmosphere,
but upon exposure to air the complexes in solution are slowly
oxidized to materials with a blue color. The electrospray
mass spectra shows a molecular mass peak of all clusters with
the complete isotopic pattern for six copper atoms at 1524.0
(1), 1565.9 (2), 1653.8 (3), and 1747.8 amu (4) respectively
(see the Supporting Information). The spectral pattern is
identical with that simulated from the molecular formula.
Thus, these clusters are stable in the gas phase in the spectro-
meter environment. Thermogravimetric analysis (TGA)
indicated a decomposition temperature range for clusters 2-4
between 200 and 500 °C, whereas 1 the decomposition range
is 300-500 °C. Overall the copper clusters are quite ther-
mally stable, with complex 1 being the least stable member.
Some preliminary reactivity of the complex [Cu6L2Cl2] (2;
L = C(Ph2PdNSiMe3)22-) has been investigated. Quanti-
tative addition of 2 mol of LiCH3 to a THF solution of 2
The solid-state structure of the bimetallic complex 6 was
investigated by means of single-crystal X-ray diffraction
with the result shown in Figure 3. Selected bond lengths
and angles and data collection parameters are given in
Tables 2 and 3. The core structure of 6 (Figure 3) shows
two copper atoms bridged by a pincer carbon structure. Each
copper center is thus coordinated to three different donor
atoms (the pincer bridging carbon, a N, and a P) and bound
˚
to the other copper (the Cu(1)-Cu(2) distance is 2.7217(6) A),
making a four-coordinated metal center with a distorted-tetra-
hedral geometry. Two heterocyclic four-membered rings (Cu(1)-
C(1)P(1)N(1) and Cu(2)C(1)P(2)N(2)) are connected through
the two side borders of the three-membered ring (Cu(1)-
Cu(2)C(1)), and one points up and one points down relative
to the plane of the three-membered ring, forming a chair
conformation. The Cu-N (Cu(1)-N(1) = 2.142(3), Cu(2)-
˚
N(2) = 2.126(3) A) and Cu-P lengths (Cu(1)-P(3) =
˚
2.1608(9), Cu(2)-P(4) = 2.1582(9) A) are in the expected
˚
range (e.g., Cu-N=2.058(2)-2.074(2) A and Cu-P =
7
˚
2.1756(7) A in [{CH(PPh2NdSiMe3)2}CuPPh3] and Cu-
˚
˚
N = 2.116(7) and 2.121(6) A and Cu-P = 2.244(2) A in
(13) Lin, Y.-Y.; Cai, S.-W.; Che, C.-M.; Fu, W.-F.; Zhou, Z.-Y.; Zhu,
N-Y Inorg. Chem. 2005, 44, 1511–1524.
(14) (a) Jeffery, J. C.; Rauchfuss, T. B.; Tucker, P. A. Inorg. Chem.
1980, 19, 3306–3315. (b) Zhao, Y.; Hong, M.; Su, W.; Cao, R.; Zhou, Z.
Dalton Trans. 2000, 1685–1686.
(15) Niu, Y.-Y.; Zhang, N.; Hou, H.-W.; Zhu, Y.; Tang, M.-S.; Ng,
S.-W. J. Mol. Struct. 2007, 827, 195–200.
(16) (a) Muhle, J.; Sheldrick, W. S. Z. Anorg. Allg. Chem. 2003, 629,
2097–2102. (b) Healy, P. C.; Pakawatchai, C.; White, A. H. J. Chem. Soc.,
Dalton Trans. 1983, 1917–1927. (c) Walsdorff, C.; Park, S.; Kim, J.; Heo, J.;
Park, K.; Oh, J.; Kim, K. J. Chem. Soc., Dalton Trans. 1999, 923–930. (d)
Toth, A.; Floriani, C.; Chiesi-Villa, A.; Guastini, C. Inorg. Chem. 1987, 26,
3897–3902. (e) Dyason, J. C.; Engdhardt, L. M.; Healy, P. C.; Kildea, J. D.;
White, A. H. Aust. J. Chem. 1988, 41, 335–340. (f) Pohl, S.; Harmjanz, M.;
Schneider, J.; Saak, W.; Henkel, G. Dalton Trans. 2000, 3473–3479. (g)
Zhao, S.-B.; Wang, R.-Y.; Wang, S.-N. Inorg. Chem. 2006, 45, 5830–5840.
(h) Domaservitch, K. V.; Solntsev, P. V.; Sieler, J. Zh. Strukt. Khim.
(J. Struct. Chem.) 2005, 46 (suppl), S145–S151.
cis-[Cu2(μ-CN)(Phen)(PPh3)2]2[C(CN)3] BF4 2CH3CN10).
3
3
˚
The Cu-C bond distances (Cu(1)-C(1) = 2.031(3) A and
˚
Cu(2)-C(1) = 2.045(3) A) in 6 are remarkably longer than
that in the dicoordinate copper structure (e.g., 1.916(2) A in
˚
[Cy3PCuCH3])11 and even longer than these in the tricoordinate
structures (e.g., Cu-C = 1.921(3) A in [Cu(CN)(dppm)]12 and
˚
€
€
(9) Kohn, R. D.; Pan, Z.; Mahon, M. F.; Kociok-Kohn, G. Chem.
Commun. 2003, 1272–1273.
(10) Kamte, M. F.; Baumeister, U.; Schafer, W. Z. Anorg. Allg.
€
Chem. 2003, 629, 1919–1924.
(11) Schaper, F.; Foley, S. R.; Jordan, R. F. J. Am. Chem. Soc. 2004,
126, 2115–2124.
(12) Di Nicola, C.; Effendy; Pettinari, C.; Skelton, B. W.; Somers, N.;
White, A. H. Inorg. Chim. Acta 2006, 359, 53–63.

Article
Organometallics, Vol. 29, No. 19, 2010 4255
˚
Table 2. Selected Bond Lengths (A) and Angles (deg) for Compounds 1-4, 6, and 7
Compound 1
Cu(4)-Cu(6)
Cu(1)-Cu(6)
Cu(1)-Cu(5)
Cu(4)-Cu(5)
Cu(4)-Cu(3)
Cu(2)-Cu(6)
Cu(1)-Cu(2)
Cu(3)-Cu(5)
2.4000(4)
2.6128(4)
2.4047(4)
2.6512(4)
2.7496(4)
2.7732(4)
2.7546(4)
2.7785(4)
Cu(2)-Cu(3)
Cu(2)-C(1)
Cu(6)-C(1)
Cu(3)-C(2)
Cu(5)-C(2)
Cu(1)-C(3)
Cu(5)-C(3)
2.5649(4)
1.961(2)
1.993(2)
1.953(2)
1.992(2)
1.979(2)
2.057(3)
Cu(2)-Cu(3)-Cu(4)
Cu(2)-Cu(3)-Cu(5)
Cu(1)-Cu(2)-Cu(3)
Cu(3)-Cu(2)-Cu(6)
Cu(5)-Cu(1)-Cu(6)
Cu(1)-Cu(5)-Cu(4)
Cu(1)-Cu(6)-Cu(4)
104.647(12)
59.307(10)
105.540(12)
60.733(10)
66.244(12)
104.475(14)
105.788(14)
Cu(2)-Cu(5)-Cu(3)
P(1)-C(1)-P(2)
P(3)-C(2)-P(4)
Cu(4)-C(4)-Cu(6)
N(3)-Cu(4)-C(4)
N(2)-Cu(3)-C(2)
C(1)-Cu(6)-C(4)
56.330(10)
129.63(12)
129.51(12)
72.89(8)
158.11(9)
174.12(9)
159.50(9)
Compound 2
Cu(4B)-Cu(6B)
Cu(4B)-Cu(5B)
Cu(1B)-Cu(5B)
Cu(1B)-Cu(6B)
Cu(2B)-Cu(6B)
Cu(3B)-Cu(4B)
Cu(3B)-Cu(5B)
Cu(1B)-Cu(2B)
2.6624(6)
2.6510(6)
2.7429(6)
2.6264(6)
2.7301(5)
2.7688(5)
2.7397(5)
2.8317(6)
Cu(2B)-Cu(3B)
Cu(2B)-C(1B)
Cu(6B)-C(1B)
Cu(3B)-C(2B)
Cu(5B)-C(2B)
Cu(5B)-C(2B)
Cu(1B)-Cl(1B)
Cu(5B)-Cl(1B)
2.5982(5)
1.954(3)
1.966(3)
1.945(3)
1.967(3)
1.967(3)
2.1447(10)
2.1659(10)
Cu(2B)-Cu(3B)-Cu(4B)
Cu(2B)-Cu(3B)-Cu(5B)
Cu(1B)-Cu(2B)-Cu(3B)
Cu(3B)-Cu(2B)-Cu(6B)
Cu(5B)-Cu(1B)-Cu(6B)
Cu(1B)-Cu(5B)-Cu(4B)
Cu(1B)-Cu(6B)-Cu(4B)
107.619(18)
62.188(14)
106.360(17)
60.980(14)
64.933(16)
102.173(17)
106.452(19)
Cu(2B)-Cu(5B)-Cu(3B)
P(1B)-C(1B)-P(2B)
P(3B)-C(2B)-P(4B)
Cu(1B)-Cl(1B)-Cu(5B)
Cl(1B)-Cu(1B)-N(1B)
N(2B)-Cu(3B)-C(2B)
Cl(1B)-Cu(5B)-C(2B)
56.390(13)
128.91(18)
129.81(18)
79.04(3)
164.54(8)
174.55(12)
163.19(9)
Compound 3
Cu(4)-Cu(6)
Cu(4)-Cu(5)
Cu(1)-Cu(5)
Cu(1)-Cu(6)
Cu(2)-Cu(6)
Cu(3)-Cu(4)
Cu(3)-Cu(5)
Cu(1)-Cu(2)
2.7109(7)
2.6472(7)
2.7254(7)
2.6288(7)
2.7629(6)
2.7894(7)
2.7300(6)
2.8031(7)
Cu(2)-Cu(3)
Cu(2)-C(1)
Cu(6)-C(1)
Cu(3)-C(2)
Cu(5)-C(2)
Br(1)-Cu(5)
Br(1)-Cu(1)
2.5853(6)
1.944(4)
1.981(4)
1.948(4)
1.974(4)
2.3077(6)
2.2658(6)
Cu(2)-Cu(3)-Cu(4)
Cu(2)-Cu(3)-Cu(5)
Cu(1)-Cu(2)-Cu(3)
Cu(3)-Cu(2)-Cu(6)
Cu(5)-Cu(1)-Cu(6)
Cu(1)-Cu(5)-Cu(4)
Cu(1)-Cu(6)-Cu(4)
108.64(2)
62.025(17)
107.01(2)
61.065(17)
64.436(18)
106.39(2)
107.34(2)
Cu(2)-Cu(5)-Cu(3)
P(1)-C(1)-P(2)
P(3)-C(2)-P(4)
Cu(4)-Br(2)-Cu(6)
N(3)-Cu(4)-Br(2)
N(2)-Cu(3)-C(2)
C(1)-Cu(6)-Br(2)
56.396(16)
128.0(2)
128.0(2)
72.51(2)
161.56(10)
174.73(14)
159.62(11)
Compound 4
Cu(4)-Cu(6)
Cu(4)-Cu(5)
Cu(1)-Cu(5)
Cu(1)-Cu(6)
Cu(2)-Cu(6)
Cu(3)-Cu(4)
Cu(3)-Cu(5)
Cu(1)-Cu(2)
2.6480(7)
2.6391(9)
2.6630(8)
2.6308(8)
2.7373(7)
2.6805(7)
2.7583(7)
2.6983(7)
Cu(2)-Cu(3)
Cu(2)-C(10)
Cu(6)-C(10)
Cu(3)-C(20)
Cu(5)-C(20)
I(1)-Cu(5)
2.6052(6)
1.942(4)
1.989(4)
1.944(4)
1.983(4)
2.4716(6)
2.4531(7)
I(1)-Cu(1)
Cu(2)-Cu(3)-Cu(4)
Cu(2)-Cu(3)-Cu(5)
Cu(1)-Cu(2)-Cu(3)
Cu(3)-Cu(2)-Cu(6)
106.56(2)
59.639(18)
107.29(2)
60.278(18)
Cu(2)-Cu(5)-Cu(3)
P(1)-C(10)-P(2)
P(3)-C(20)-P(4)
Cu(4)-I(1)-Cu(6)
57.328(17)
130.1(2)
130.6(2)
64.933(19)

4256 Organometallics, Vol. 29, No. 19, 2010
Ma et al.
Table 2. Continued
67.26(2)
I(2)-Cu(6)-C(10)
N(2)-Cu(3)-C(20)
I(2)-Cu(4)-N(3)
159.13(11)
175.22(17)
155.58(11)
Cu(5)-Cu(1)-Cu(6)
Cu(1)-Cu(5)-Cu(4)
Cu(1)-Cu(6)-Cu(4)
104.01(3)
104.67(3)
Compound 6
Cu(1)-Cu(2)
Cu(1)-P(3)
Cu(1)-N(1)
Cu(1)-C(1)
2.7217(6)
2.1608(9)
2.142(3)
2.031(3)
Cu(2)-P(4)
Cu(2)-N(2)
Cu(2)-C(1)
2.1582(9)
2.126(3)
2.045(3)
Cu(2)-Cu(1)-P(3)
Cu(1)-Cu(2)-P(4)
C(1)-Cu(1)-P(3)
C(1)-Cu(2)-P(4)
116.60(3)
119.40(3)
152.45(9)
145.90(8)
C(1)-Cu(2)-N(2)
C(1)-Cu(1)-Cu(2)
N(1)-Cu(1)-P(3)
P(1)-C(1)-P(2)
77.81(11)
48.32(8)
128.46(7)
134.75(18)
Compound 7
Cu-I
2.5071(5)
P-N
1.5694(19)
1.8249(16)
1.8249(16)
Cu-N
2.0480(19)
2.0480(19)
P-C(10)
Cu-N0
P0-C(10)
I-Cu-N
I-Cu-N0
N-Cu-N0
127.64(5)
127.64(5)
104.57(11)
N-P-C(10)
P0-C(10)-P
111.78(12)
118.36(16)
resulted in an immediate color change from bright yellow to
green. The 31P NMR measurement of the resultant solution
showed clearly that the product was complex 1, which was
subsequently isolated from the solution and the identity
confirmed. A solution containing the isolated complex 2
was treated with 2 mol of NaOtBu in THF in slight excess.
Monitoring the reaction by 31P NMR showed that the
reaction proceeded very slowly, taking several days at room
temperature. Throughout the reaction period we observed
the initial buildup of 12 peaks in 3 pairs of AB coupling
patterns, which could be assigned to the cluster species
[Cu6L2Cl2] (the starting material), [Cu6L2(OtBu)Cl] (an
intermediate), and [Cu6L2(OtBu)2] (5) (the final product).
Eventually, the 31P NMR spectrum showed only one final
set of four peaks with an AB coupling pattern (Scheme 3),
which can be assigned to 5. It is interesting that these new
species can be generated by simple substitution reactions
without disrupting the six-copper core cluster structure. The
new cluster 5 may show some novel reactivity or catalytic
properties and is presently under study.
calculated HOMO and LUMO energy gaps are 300, 282, 290,
and 330 nm for 1-4, respectively). The DFT molecular orbital
coefficients show that the HOMO component is dominated by
copper d orbitals and the LUMO is mainly composed of copper
p orbitals (see the Supporting Information), thereby providing
support for the absorption band assignments to 3d f 4p transi-
tions. The weak shoulders on the three halide complexes (2-4)
could be due to either halide to metal charge transfer (XMCT) or
halide to ligand charge transfer (XLCT) because complex 1,
which is halide-free, does not show this shoulder. Additional
information from the spectrum of 7 (see below and Figure 5a)
supports the former interpretation (XMCT), as the latter
(XLCT) is unlikely in this simple chelate. Similar absorption
bands were previously observed in the dimeric copper complex
1f
[Cu2(dcpm)2]X2 (dcpm = bis(dicyclohexylphosphanyl)-
methane) and the hexanuclear octahedral copper complex
Cu6(mtc)18 (mtc- = di-n-propylmonothiocarbamate), where
assignments were similar.17 The absorption spectra of Cu(I)
halides have also been attributed to mixed d f s/LMCT
transitions.
Absorption spectra of 1-4 in solution are shown in
Figure 5a. All compounds display strong absorption peaks
in the UV (∼300 nm, ε ≈ 104 L cm-1 M-1) and near-visible
regions (∼425 nm, ε ≈ 5 ꢀ 103 L cm-1 M-1), in addition to
weaker bands or shoulders in between. Spectral data are
summarized in Table 4. The spectral patterns for 2-4 were
very similar to each other, with the shorter wavelength
absorption appearing as peaks at 299, 302, and 301 nm and
longer wavelength absorption appearing as peaks at 409,
413, and 423 nm in the case of 2-4, respectively. Much
weaker shoulders were present at 360 and 362 nm for 2 and 3,
and an even weaker shoulder appears at 364 nm in the case of
4. Complex 1 showed, in contrast, only two peaks at 294 and
427 nm. Following the literature,17 we assign the strong peak
in the UV region to metal-centered metal-metal 3d f 4p
transitions, while the peak in the near-visible region is most
likely due to ligand to metal charge transfer (LMCT)
and metal-centered d f s transitions. The strong peak
in the UV region nicely matches the energy gap between the
HOMO and LUMO, as shown by the DFT calculations (the
The Cu-Cu bond distances shown in the molecular
structures 1-4 decrease in the order 1, 2, 4, and 3, which is
in the same order as shown by the UV absorption peak
energies (294 (1), 299 (2), 301 (4), and 302 nm (3)), suggesting
a correlation between bond lengths and absorption peak
energy; thus, complexes 2-4 show a halide effect on the
transition energies of the absorption peaks (Figure 5a).
Similar to the case for 1, the monocopper iodide complex
[CuI(H2L)] (7) displays two strong peaks at 296 and 365 nm,
respectively (Figure 5a). Because there is only one copper
atom present in this structure, the absorption peak is shifted
to a high-energy wavelength compared to those for the six-
copper clusters. We note that the second peak at 365 nm in 7
is very close to the shoulder peaks which appear in the three
halide clusters. As halide to metal charge transfer (XMCT) is
the most likely process for 7, we suggest that the absorption
spectrum of 7 provides reasonable evidence for halide to
metal charge transfer (XMCT) in the six-copper clusters
(1-4).
(18) Sabin, F.; Ryu, C. K.; Ford, P. C.; Vogler, A. Inorg. Chem. 1992,
31, 1941–1945.
(17) Ford, P. C.; Vogler, A. Acc. Chem. Res. 1993, 26, 220–226.

Article
Organometallics, Vol. 29, No. 19, 2010 4257

4258 Organometallics, Vol. 29, No. 19, 2010
Ma et al.
Figure 4. Perspective view of the complex [CuI{CH2(PPh2d
NSiMe3)2}] (7) showing the atom-labeling scheme. Primed
atoms are related to unprimed ones by the mirror plane located
1
at x, /4, z. Non-hydrogen atoms are represented by Gaussian
ellipsoids at the 20% probability level. Hydrogen atoms are
shown with arbitrarily small thermal parameters for the methy-
lene group and are not shown for the methyl and phenyl groups.
Bond lengths and angles are given in Table 2.
Scheme 3. Reaction of 2 with MeLi To Form 1 and with NaOtBu
To Form 5
Figure 5. (a) UV-vis solution absorption spectra of 1-4 and 7
and (b) solid-state emission spectra of the same complexes
excited (ex) with 325 nm radiation provided by a He-Cd laser
source. Note that the absorption and emission maxima of 7 are
substantially different from those of 1-4 because of a difference
in molecular structure.
be assigned to the Cu6 cluster d r s transition.18,19 Relative
to 1, the emission bands of 2-4 show a red shift trend and the
presence of weaker HE bands suggests some halogen partic-
ipation in the emission spectra. This emissive state has
mixed d-s/XMCT character delocalized over the Cu6X2 core
with roughly equal contributions from each component.17,20
The mixed d-s/XLCT excited state accounts for the HE
emission band in each system.19 The very large Stokes shift
for the LE emission band of the Cu6X2 core clusters is con-
sistent with this assignment for the transitions. These assign-
ments are also consistent with early studies of the Cu4I4(Py)2
(Py = pyridine) core system.17-20 There is a clear trend connect-
ing absorption wavelength and emission energies which follows
the nature of the bridging substituents, suggesting that sub-
stitution of the capping ligand on the cluster could be used to
control the values of the optical properties.
In the solid state, the hexanuclear copper complexes (1-4)
show intense photoluminescence spectra (Figure 5b). At
ambient temperatures one intense low-energy emission at
λmaxem 579 nm (or 17 271 cm-1) was detected for complex 1
and 2-4 show two emission bands; respectively intense low
energy (LE) bands at 631 (15 848 cm-1), 655 (15 267 cm-1),
and 659 nm (15 175 cm-1) and much weaker, higher energy
(HE) bands at 528 (18 940 cm-1), 566 (17 668 cm-1), and
570 nm (17 544 cm-1). The bright LE emission band of 1 can
The monocopper iodide complex 7 also shows a symmetric
photoluminescence peak at λmaxem 496 nm (or 20 160 cm-1
)
similar to that of 1 without any additional shoulder or weak
bands; thus, these features likely arise from within the cluster
unit.
The bright emission colors of the four copper clusters 1-4
can be seen both in the solid state and in solution (with colors
(19) Kyle, K. R.; Ryu, C. K.; DiBenedetto, J. A.; Ford, P. C. J. Am.
Chem. Soc. 1991, 113, 2954–2965.
(20) Vitale, M.; Palke, W. E.; Ford, P. C. J. Phys. Chem. 1992, 96,
8329–8336.

Article
Organometallics, Vol. 29, No. 19, 2010 4259
Table 4. Summary of Absorption, Emission, and Excitation Spectral Data for Compounds 1-4 and 7a
compd
absorbance
excitation/toluene
emission/toluene
emission/solid
QY
1
2
3
4
7
427 (9.9 ꢀ 103), 294 (2.1 ꢀ 104)
409 (7.9 ꢀ 103), 299 (2.3 ꢀ 104)
419 (6.5 ꢀ 103), 302 (2.0 ꢀ 104)
423 (9.9 ꢀ 103), 301 (3.8 ꢀ 104)
365 (6.8 ꢀ 103), 296 (1.3 ꢀ 104)
326, 420
342, 421
605, 701 (sh)
633, 703 (sh)
579 (b)
0.75
0.71
631 (b), 528 (s)
655 (b), 566 (s)
659 (b), 570 (s)
496 (b)
365, 451
648, 714 (sh)
0.62
a All wavelengths are given in nm and extinction coefficients (in parentheses) in L cm-1 M-1. Legend: sh = shoulder, b = big, s = small.
Figure 7. Relationship between emission and absorbance
energies of the hexacopper clusters 1-4. The connecting lines
are not linear fits but rather association connections.
The intensities of the solution emission spectra excited by
two different energies, 325 or 420 nm, 342 or 420 nm, and
365 or 450 nm for compounds 1, 2, and 4, respectively
(corresponding to the two main excitation peaks observed
in all solution cases, as shown in Figure 6a with data given in
Table 4) were compared in order to evaluate which value of
excitation energy yielded the more intense emission. In all
cases the lower energy excitation (420 nm in the case of 1 and
2 and 450 nm in the case of 4) yielded the highest intensity
emission compared to its higher energy excitation at 325,
342, and 365 nm. In each case the same emission peak shape
(or spectral pattern) was detected (Figure 6b) for each
compound regardless of the excitation energy. The solution
emission spectra are also very similar to the spectra recorded
for the solids, except for slight differences in peak position
probably caused by solvent effects.
That different excitation energies, for either the solids or
the solutions, yield the same emission spectrum indicates that
the emission energy gap is a constant defined by the molec-
ular cluster. The quantum yields of these compounds were
determined in highly dilute concentrations in toluene
through comparison to a similarly dilute solution of a
Rhodamine 590 standard using the same measurement con-
ditions (Table 4). The measured quantum yields were 0.75,
0.71, and 0.62, respectively, for 1, 2, and 4, which are higher
than those reported for the carbon-bridged tetraatomic
copper cluster (QY = 0.42)2s but are lower than the highest
value reported for a mixed copper-gold cluster (QY =
0.92).2t Overall, these copper clusters are highly photo-
efficient compounds. Moreover, their quantum yields are
comparable through the hexaatomic series, indicating that
the emission is generated by the boat-shaped hexaatomic
copper cluster core.
Additional insight into the character of the transitions was
provided by a TD-DFT calculation on the core structure of
cluster 2, a representative example of the six-copper cluster
series. The results showed two allowed singlet excited state
transitions: ES1 (=S0 f S1) at 344.81 nm ( f = 0.0116) with
an energy gap of 3.5957 eV and ES2 (=S0 f S2) at 321.37 nm
( f = 0.0286) with an energy gap of 3.8580 eV. One forbidden
Figure 6. (a) Excitation spectra of 1, 2, and 4 in toluene. Note
that maximum excitation for 1 and 2 occurs at 420 nm and that
for 4 at 450 nm. (b) Emission spectra for complexes obtained by
excitation (ex) at the maximum indicated by the excitation
spectra: 420 nm in the case of 1 and 2 and 450 nm for 4. The
inset shows the color of emission under broad-band UV light for
these three complexes.
ranging from yellow to orange to red). The photo-
luminescent spectra of the solid materials are shown in Figure 5b,
and data are given in Table 4. The emission peaks shift
toward lower energy through the series 1, 2, and 4. The
emission peaks of 3 and 4 are very close to each other. In
solution, the emission and excitation spectra of compounds
1, 2, and 4 are very similar, as indicated in Figure 6 (with
numerical data given in Table 4). The excitations of these
three compounds in solution show very similar spectral
patterns, with two large excitation peaks (Figure 6b and
Table 4). In solution again there is a shift of the excitation
peaks to lower energy through the series 1 to 2 and 4. The
qualitative relationships between emission and absorbance
energies is illustrated in Figure 7.

4260 Organometallics, Vol. 29, No. 19, 2010
Ma et al.
transition at 331.85 nm ( f = 0.0000) with an energy gap
of 3.7362 eV was also calculated. ES1 is derived from the
three molecular orbital combinations HOMO-2 f LUMO
(coeff = 0.1806), HOMO f LUMO (coeff = 0.6550), and
HOMO f LUMOþ1 (coeff = -0.1115), while ES2 involves
thethree combinationsHOMO-2f LUMO (coeff= 0.6388),
HOMO-1 f LUMOþ2 (coeff = 0.1123), and HOMO f
LUMO (coeff = -0.1708), which correspond to the observed
onset of the absorption as shown in Figure 5a (409 and
299 nm) and to the two peaks which appear in the excited
spectra as shown in Figure 6 (red line: 421 and 342 nm).
The TD-DFT calculation thus qualitatively predicts the
Franck-Condon absorption and singlet manifold excita-
tion energy gaps based on the ground-state geometry of the
example cluster. The transition characteristics of the two
excited states ES1 (HOMO-2 f LUMO, HOMO f LUMO,
and HOMO f LUMOþ1) and ES2 (HOMO-2 f LUMO,
HOMO-1 f LUMOþ2, and HOMO f LUMO) (calculated
details are given in the Supporting Information) show that
the occupied molecular orbitals (HOMO-2, HOMO-1, and
the HOMO) possess prominent metal d character from six
copper atoms combined with p character from the carbene
carbon atoms, while the unoccupied molecular orbitals
(LUMO, LUMOþ1, and LUMOþ2) possess mainly p and
s character from six copper atoms and s character from the
carbene carbon and p and s character from the bridged
chlorine atoms. Thus, the two excitation or emission transi-
tion bands of cluster 2 can reasonably be assigned to the
transitions from copper 3d to copper 4p or copper 3d, to
copper 4s (3d f 4p or 3d f4s: metal-metal charge transfer
MMCT) and copper 3d to carbene 2p or halide 3s and 3p
(3d f 2p or 3d f 3s: metal to ligand charge transfer MLCT)
(for the details of contribution from the molecular orbitals to
the transition states see the Supporting Information). The
TD-DFT calculation results show that the photophysical
properties of these six-copper bridged carbene halide clusters
arise from the core structure.
All previously reported copper cluster systems have highly
symmetric geometries; Cu4 is a tetrahedron,21 and a pre-
viously reported Cu6 case is an octahedron.18 Herein we have
found a rare example of a carbon-bridged, six-copper com-
plex with an irregular boat-shaped core structure. In order to
understand the assembly bonding details of these clusters,
DFT calculations were performed with Gaussian-03.22 The
geometry optimization gave a minimum energy structure
with Cu-Cu distances longer than those determined in the
solid (see the Supporting Information); however, this is not
too surprising, because the crystal structures are deter-
mined at higher temperatures and are also subject to inter-
molecular forces, whereas almost all of the calculations use
isolated model structures. The principal structural feature is
described as two seven-membered rings (C(1)-Cu(6)-Cl(2)-
Cu(4)-N(4)-P(4)-C(2) or C(1)-P(3)-N(1)-Cu(1)-Cl(1)-
Cu(5)-C(2)) and two five-membered rings (C(1)-P(2)-N(3)-
Cu(3)-C(2) or C(1)-Cu(2)-N(2)-P(1)-C(2)) joined through
the axis formed by two carbene carbons (C(1)- - -C(2)). Each
copper(I) atom is coordinated with two atoms or a combina-
tion of three different atoms, combinations being C, X, N, C,
Figure 8. (top) Geometry of the six-copper cluster cores (1-4),
with the core structure of complex 4 on the left and two views of
the central core of the Cu6 cluster on the right, the uppermost
member being the symmetry of the cluster and the lower member
being the boat shape of the cluster. (bottom) Optimized struc-
ture of cluster 2.
or C, N (such as N(1)-Cu(1)-Cl(1), C(1)-Cu(2)-N(2),
Cl(1)-Cu(5)-C(2), C(2)-Cu(3)-N(3), Cl(2)-Cu(4)-N(4),
and Cl(2)-Cu(6)-C(1) in 2) showing a nearly lineargeometry
with the bond angles of C-Cu-N, C-Cu-X and N-Cu-X
in the range 169.21-174.22 which are close to 180° (see, as
an example, the arrangement for 2 shown in Figure 8).
The methyl or halide bridging and the PCP pincer ligand
(NdP-C-PdN) cross-linking interactions bring three
copper(I)-copper(I) pairs close with relatively short dis-
tances, indicating the existence of weak bonding, such as
Cu(2)-Cu(3), Cu(1)-Cu(6), and Cu(4)-Cu(5) for 2 as
shown in Figure 8.
The calculated molecular orbitals of 1 and 4, which are
representative of the methyl and one of the halide bridged
complexes out of the group of the four clusters, are shown in
Figures 9 and 10 (for 2 and 3 see the Supporting Information).
As indicated by the calculated molecular coefficients, the
strongest bonds between copper atoms within the cluster are
mainly formed from s orbital overlap with very little d con-
tribution. The methanediide bridging carbon atom bonds to the
two different copper(I) atoms through its py and pz atomic
orbitals, forming a Cu-C-Cu bond angle close to 95° (see
Figure 9, MO-184 and MO-185). The carbon to two copper
bonds are 12.1% copper s0.91d0.09 and 87.9% carbon s0.16p0.84
and 11.8% copper s0.93p0.01d0.06 and 88.2% carbon s0.13p0.87, as
shown in NBO-148 and NBO-149, respectively, by the NBO
analysis (Figure 11). The geometry optimized structure of 1
shows that the bond distances Cu(1)-Cu(6), Cu(4)-Cu(5),
Cu(2)-Cu(3), Cu(1)-Cu(5), and Cu(4)-Cu(6) all decrease
(21) Raston, C. L.; White, A. H. J. Chem. Soc., Dalton Trans. 1976,
2153–2156.
(22) (a) Frisch, M. J. et al. Gaussian-03, Revision C02; Gaussian, Inc.,
Pittsburgh, PA, 2004 (the full citation is given in the Supporting Information).
(b) NBO component included in the Gaussian-03 program suite: Glendening,
E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F. NBO version 3.1.

Article
Organometallics, Vol. 29, No. 19, 2010 4261
Figure 9. MO of cluster 1: (top left) MO-184; (top right) MO-185; (bottom left) HOMO (MO-186); (bottom right) LUMO (MO-187).
Figure 10. MO of cluster 4: (top left) LUMO (MO-185); (top right) HOMO (MO-184); (bottom left) MO-183; (bottom right) MO-182.
Figure 11. NBO of the key structure of the methanediide bridged carbon to two copper(I) atoms in cluster 1: (left) NBO-148; (right)
NBO-149.
gradually but fall within the usual Cu-Cu bonding distance,
which is consistent with the crystal structure data. NBO
analysis shows that the net bonding order of the three types
of Cu-Cu bonding distances are, first, about 0.059 for both

4262 Organometallics, Vol. 29, No. 19, 2010
Ma et al.
Cu(1)-Cu(6) andCu(4)-Cu(5) (calcd 2.632 A) withas0.16d0.84
-
techniques or in an argon-filled glovebox. The solvents used
(special THF, toluene, and ether) were dried over appropriate
drying agents and degassed by three freeze-pump-thaw cycles
prior to use. The 1H, 13C, and 31P NMR spectra were recorded
on a Varian I400 spectrometer operating at 400.13, 100.6, and
˚
(93.0%) and s0.23d0.77(7.0%) atomic orbital construction, sec-
ond, 0.124 for Cu(2)-Cu(3) (calcd 2.565 A) with s0.95p0.01d0.04
˚
(49.7%) and s0.95p0.01d0.04 (50.3%) (approximately equal per-
centages of an almost pure s atomic orbital linkage) and, last,
1
161.9 MHz, respectively. The H and 13C NMR spectra were
˚
0.180 for Cu(1)-Cu(5) and Cu(4)-Cu(6) (calcd 2.402 A) with
referenced internally using the residual proton solvent reso-
nances, which were referenced to SiMe4 (δ 0). 31P NMR
chemical shifts are given relative to an 85% H3PO4 external refer-
ence. UV-vis spectra were recorded with a Hewlett-Packard
8453 UV-vis DAD spectrophotometer. Photoluminescence
(PL) spectra of single crystals in a sealed tube (1-4, 7) were
measured at room temperature using the 325 nm line of a He-
Cd laser excitation source, and emission was detected with a
fiber optic digital charge-coupled device (CCD) spectrometer
whose spectral response was normalized using a standard black-
body radiator. Emission and excitation spectra in toluene were
recorded using a Varian-Cary fluorescence spectrophotometer.
Peak positions were determined by spectral deconvolution with
ORIGIN. Quantum yields were obtained by comparing with
Rhodamine-590 (QY = 0.95 in ethanol) dye in solution. Elemental
analyses and IR spectra were carried out at the Analytical and
Instrumentation Laboratory, Department of Chemistry, Univer-
sity of Alberta. The organolithium compounds [Li2L]2 (1) were
prepared according to our published procedures.23 The copper
halides were purchased from either Strem or Aldrich; cod and all
other precursors used were prepared by following or using proce-
dures similar to those described in the literature.24
Synthetic Procedures. Synthesis of [Cu6{C(Ph2PdNSiMe3)2}2-
(CH3)2] (1). To a stirred THF (15 mL) suspension of [NEt4CuCl2]
(0.159 g, 0.6 mmol) was added, as a white solid, [Li2C(Ph2Pd
NSiMe3)2]2 (0.115 g, 0.1 mmol). Then 0.125 mL (0.2 mmol) of
1.6 M LiCH3, as an ether solution, was injected at room tempera-
ture. The reaction mixture turned bright yellow-green immediately
after the injection. When the reaction was finished, the THF solvent
was removed under vacuum and the remaining powder was washed
with hexane and then dissolved in 10 mL of ether. The insoluble
solid was coalesced by centrifugation. The clear yellow resultant
solution was put inside a -20 °C freezer for a few hours. A yellow-
green powder was collected and dried under vacuum to yield
0.134 g of yellow-green material (yield 88.0%). Crystals for X-ray
diffraction were grown from THF solvent. IR data (Nujol mull):
3076 m, 3054 m, 2973 s, 2949 s, 2892 m, 2862 m, 1480 s, 1436 s, 1381
w, 1350 w, 1256 m, 1246 s, 1101 s, 1065 s, 1035 s, 915 w, 835 s, 780 s,
750 s, 739 s, 722 s, 713 m, 692 s, 672 m. 1H NMR (toluene-d8):δ7.90
(m, phenyl, 4H), 7.78 (m, phenyl, 4H), 7.59 (m, phenyl, 4H), 7.50
(m, phenyl, 4H), 7.09 (m, phenyl, 8H), 6.98 (m, phenyl, 4H), 6.90
(m, phenyl, 4H), 6.69 (m, phenyl, 8H), 1.04 (s, -CH3, 6H), 0.01 (s,
-Si(CH3)3, 18H), -0.11 (s, -Si(CH3)3, 18H). 13C{1H} NMR
(toluene-d8): δ 135.92 (m, phenyl), 135.10 (m, phenyl), 132.94 (d,
phenyl), 132.21 (d, phenyl), 131.95 (d, phenyl), 131.47 (d, phenyl),
129.69 (s, phenyl), 129.31 (s, phenyl), 4.86 (d, Si(CH3)3), 4.19 (d,
Si(CH3)3), -18.99 (t, bridged -CH3). 31P{1H} NMR (toluene-d8):
δ 25.14 (PA) and 24.33 (PB) (2JA-B = 58.0 Hz). ESI-MS (THF þ
toluene): m/z 1524.0 [M þ H]þ. Anal. Calcd for C64H82N4P4Si4-
Cu6(C4H10O)2: C, 51.68; H, 6.15; N, 3.35. Found: C, 51.74; H, 6.18;
N, 3.36. 1H NMR in toluene-d8 also detected the two ether solvent
molecules quantitatively.
s
0.84d0.16 (34.0%) and s0.81d0.19 (66.0%). The bond orders of all
the connected copper pairs are relatively low, indicating that the
interactions between these three types of copper pairs are
relatively weak, and it is clear that the extensive additional bond-
ing provided by the ligand contributes greatly to the stability of
the system.
DFT calculations for the bimetallic complex 6 and the mono-
metallic complex 7were also performed with Gaussian-03.22 The
calculated molecular orbitals of these two complexes are
presented in the Supporting Information. As indicated by
the calculated molecular coefficients, the bimetallic copper
complex showed multiple-bonding interactions between two
copper atoms, the carbene center, and two P and two N atoms,
which is very similar to our recently reported6b bimetallic
palladium structure. The molecular coefficients and molecular
orbitals for 6 and 7 are given in the Supporting Information.
NBO analysis for 6 shows that the two carbon to copper bonds
are 7.46% copper s0.93p0.02d0.05 and 92.54% carbon s0.18p0.82
and 8.02% copper s0.93p0.02d0.05 and 91.98% carbon s0.19p0.81
(as shown in NBO-77 and NBO-78, respectively, in the
Supporting Information); hence, the bonding uses the same
orbital components as cluster 1 but with lower copper and
higher carbon contributions.
Conclusions
In conclusion, we have assembled a family of capped
hexaatomic copper clusters with a unique nonsymmetric
boat-type geometry supported by a pair of pincer methane-
diide (NPCPN) ligands which geminally bind to two copper
atoms. The clusters are capped with methyl or halide ligands.
Substitution reactions can be conducted on the capping
ligands. This coordination geometry has not previously been
reported. All clusters are fully characterized and quite stable
in the solution, solid, and gas phases. These clusters are
strongly photoluminescent in the solid state (with the solu-
tions showing the same bands), and the absorption and
emission wavelength values are dependent on the nature of
the capping group ligand, indicating that the optical proper-
ties can be controlled by means of chemical alteration of the
bridging groups. Two other complexes were also identified
and characterized; a monocopper iodide complex and a
bimetallic copper complex bridged by a spirocyclic carbon
center. Both were fully characterized, including structural
characterization. By comparisons of the copper cluster emis-
sion spectral properties to those of the monocopper com-
plexes, we can verify that the tuning of the emission bands of
the copper clusters to low energy, toward the visible region,
with halide and methyl substitution is due to delocalized
electrons within the clusters. DFT and NBO analysis has been
conducted, as well as some preliminary reactivity studies.
Further investigations on the catalytic reactivity and additional
applications are proceeding.
Synthesis of [Cu6{C(Ph2PdNSiMe3)2}2Cl2] (2). To an ether
(10 mL) suspension of [(cod)2Cu2Cl2] (0.124 g, 0.3 mmol) was
added, as a white solid, [Li2C(Ph2PdNSiMe3)2]2 (0.115 g, 0.1
mmol), and the mixture was stirred at room temperature for a
few hours. The reaction mixture turned dark brown after 1 h and
(23) Kasani, A.; Babu, R. P. K.; McDonald, R.; Cavell, R. G. Angew.
Chem., Int. Ed. 1999, 38, 1483–1484.
(24) (a) van den Hende, J. H.; Baird, W. C., Jr. J. Am. Chem. Soc.
1963, 85, 1009–1010. (b) Kok, J. M.; Skelton, B. W.; White, A. H. J. Cluster
Sci. 2004, 15 (3), 365–376. (c) Bowmaker, G. A.; Wang, J.-R.; Hart,
R. D.; White, A. H.; Healy, P. C. J. Chem. Soc., Dalton Trans. 1992,
787–795.
Experimental Section
General Methods. All experimental manipulations were per-
formed under rigorously anaerobic conditions using Schlenk

Article
Organometallics, Vol. 29, No. 19, 2010 4263
finally become a bright yellow-green solution. When the reac-
tion was finished, insoluble solids were decanted by centrifuga-
tion. The clear yellow solution was kept inside a vial for a few
days. The resultant bright yellow crystals were collected and
dried under vacuum (yield 0.141 g (90.0%)). Crystals for X-ray
diffraction were grown from ether solvent inside the NMR tube.
IR data (Nujol mull): 3074 w, 3055 w, 2952 s, 2925 s, 2855 s, 1480
m, 1464 s, 1437 m, 1377 w, 1367 m, 1258 m, 1246 s, 1103 s, 1057 s,
1029 m, 1016 m, 841 s, 783 s, 740 s, 723 m, 714 w, 692 s, 674 m. 1H
NMR (THF-d8): δ 8.00 (b, phenyl, 4H), 7.59 (q, phenyl, 4H),
7.44 (q, phenyl, 4H), 7.30 (m, phenyl, 8H), 7.12 (t, phenyl, 4H),
7.01 (m, phenyl, 8H), 6.88 (m, phenyl, 4H), 6.68 (b, phenyl, 4H),
-0.36 (s, -Si(CH3)3, 18H), -0.41 (s, -Si(CH3)3, 18H). 13C-
{1H,31P} NMR (THF-d8): δ 137.02 (s, phenyl), 136.15 (s,
phenyl), 135.53 (s, phenyl), 134.29 (s, phenyl), 133.87 (s, phenyl),
133.21 (s, phenyl), 132.91 (s, phenyl), 132.45 (s, phenyl), 130.78 (s,
phenyl), 130.73 (s, phenyl), 130.33 (s, phenyl), 130.20 (s, phenyl),
128.12 (s, phenyl), 128.03 (s, phenyl), 5.40 (s, -Si(CH3)3), 4.21 (s,
-Si(CH3)3). 31P{1H} NMR (THF-d8): δ 28.24 (PA) and 26.17 (PB)
(2JA-B = 56.0 Hz). ESI-MS (THF þ toluene): m/z 1565.9
[M þ H]þ. Anal. Calcd for C62H76N4P4Si4Cl2Cu6(C4H10O)2: C,
49.05; H, 5.61; N, 3.27. Found: C, 49.29; H, 5.68; N, 3.33. 1H NMR
in THF-d8 also detected the two ether solvent molecules
quantitatively.
Synthesis of [Cu6{C(Ph2PdNSiMe3)2}2Br2] (3). To an ether
(10 mL) suspension of [(cod)2Cu2Br2] (0.150 g, 0.3 mmol) was
added, as a white solid, [Li2C(Ph2PdNSiMe3)2]2 (0.115 g, 0.1
mmol). The mixture was stirred at room temperature for 2 h.
The reaction mixture turned bright yellow through time as the
lithium salt was added. Insoluble solids were decanted by
centrifugation. The clear yellow solution was kept in a vial for
a few hours, and some white precipitate deposited on the bottom
of the vial. The remaining clear solution was transferred to a new
vial. After a few days, bright yellow crystals formed and they
were collected and dried under vacuum (yield 0.119 g (72.0%)).
Crystals for X-ray diffraction were grown from ether solvent
inside the NMR tube. IR data (Nujol mull): 3073 m, 3048 m,
2951 s, 2895 s, 1481 m, 1436 s, 1398 w, 1307 w, 1246 s, 1180 w,
1101 s, 1061 s, 1028 m, 1010 m, 832 s, 785 s, 769 s, 741 m, 722 w,
692 s, 673 m. 1H NMR (THF-d8): δ 7.72 (m, phenyl, 8H), 7.58 (q,
phenyl, 4H), 7.48(q, phenyl, 4H), 7.39 (m, phenyl, 4H), 7.32 (t,
phenyl, 8H), 7.12 (m, phenyl, 4H), 7.01 (m, phenyl, 4H), 6.91 (m,
phenyl, 4H), -0.26 (s, -Si(CH3)3, 18H), -0.30 (s, -Si(CH3)3,
18H). 13C{1H} NMR (THF-d8): δ 136.82 (d, phenyl), 135.82 (d,
phenyl), 133.83 (d, phenyl), 132.85 (d, phenyl), 132.54 (t,
phenyl), 131.65 (s, phenyl), 130.80 (s, phenyl), 130.33 (d,
phenyl), 129.20 (s, phenyl), 128.83 (t, phenyl), 128.09 (t, phenyl),
5.43 (d, Si(CH3)3), 4.26 (d, Si(CH3)3). 31P{1H} NMR (THF-d8):
δ 28.10 (PA) and 26.59 (PB) (2JA-B = 57.0 Hz). ESI-MS (THF þ
toluene): m/z 1653.8 [M þ H]þ. Anal. Calcd for C62H76N4P4Si4-
Br2Cu6: C, 45.01; H, 4.60; N, 3.38. Found: C, 45.10; H, 4.63;
N, 3.36.
(d, phenyl), 137.22 (d, phenyl), 136.84 (d, phenyl), 136.06 (d,
phenyl), 135.62 (d, phenyl), 134.73 (d, phenyl), 134.06 (d, phenyl),
133.76 (d, phenyl), 133.26 (d, phenyl), 132.75 (q, phenyl), 130.84 (d,
phenyl), 130.35 (d, phenyl), 129.13 (br, t, phenyl), 128.15 (m,
phenyl), 5.43 (d, -Si(CH3)3), 4.34 (d, -Si(CH3)3). 31P{1H} NMR
(THF-d8): δ 28.95 (PA) and 28.08 (PB) (2JA-B = 56.0 Hz). ESI-MS
(THFþtoluene): m/z 1747.8 [MþH]þ. Anal. Calcd for C62H76-
N4P4Si4I2Cu6 1/2C4H10O: C, 43.05; H, 4.57; N, 3.14. Found:
3
C, 43.03; H, 4.55; N, 3.18.
Synthesis of [{(PPh3)Cu}2C(Ph2PdNSiMe3)2] (6). To a yellow
THF (10 mL) solution of [(NEt4)Cu(PPh3)Cl2] (0.211 g, 0.4 mmol)
was added solid [Li2C(Ph2PdNSiMe3)2]2 (0.115 g, 0.1 mmol) with
stirring at room temperature. The reaction mixture was then stirred
at room temperature for 4 h to provide a light green solution. The
THF solvent was removed under vacuum, and the remaining
powder was dissolved in 10 mL of ether. The resultant insoluble
solid was washed with toluene twice and decanted by centrifuga-
tion. The remaining combined solution was reduced to half of the
volume under vacuum and put inside a -20 °C freezer for 2 days.
A yellow-green powder was collected and dried under vacuum:
0.142 g of a yellow-green powder was obtained (yield 59.0%). IR
data (Nujol mull): 3056 m, 3008 m, 2949 s, 2897 m, 2836 m, 1593 s,
1568 s, 1498 s, 1459 s, 1437 s, 1402 m, 1286 s, 1246 s, 1179 s, 1100 s,
1058 s, 1029 s, 935 s, 917 s, 830 s, 798 s, 784 m, 741 m, 729 m, 713 m,
693 s, 674 m. 1H NMR (THF-d8): δ 7.70-6.70 (m, phenyl,
50H), -0.578 (s, -Si(CH3)3, 18H). 13C{1H} NMR (THF-d8):
δ 141.71 (m, phenyl) 139.59 (d, phenyl), 135.10 (d, phenyl),
132.73 (t, phenyl), 130.40 (s, phenyl), 129.65 (s, phenyl),
129.35 (d, phenyl), 128.88 (d, phenyl), 127.40 (t, phenyl),
4.28 (s, Si(CH3)3). 31P{1H} NMR (THF-d8): δ 4.06 (b), 19.34
3
(t, JP-P = 5.0 Hz). Anal. Calcd for C67H68N2P4Si2Cu2: C,
66.66; H, 5.63; N, 2.32. Found: C, 67.00; H, 5.70; N, 2.32.
Synthesis of [CuCH2(Ph2PdNSiMe3)2I] (7). Similar to the
reaction conditions used for the preparation of 4 (Cu metal to
ligand molar ratio at 6:1), about 10.0% of compound 7 was
formed on the basis of 31P NMR in solution. Two different
colored crystals (a light green compound which proved to be
[CuCH2(Ph2PdNSiMe3)2I2] (7) and a red complex eventually
identified as [Cu6I2C(Ph2PdNSiMe3)2] (4)) were obtained from
the reaction solution. When the copper to ligand molar ratio was
increased to 10:1, the principal final isolated product was 7.
Thus, to an ether (10 mL) suspension of [(cod)2Cu2I2] (0.300 g, 0.5
mmol) was added [Li2C(Ph2PdNSiMe3)2]2 (0.115 g, 0.1 mmol) as a
white solid, and the mixture was stirred at room temperature for a
few hours. The reaction mixture eventually yielded a yellow-orange
solution. When the reaction was finished, insoluble solids were
decanted by centrifugation. The clear yellow-orange solution was
kept inside the vial for a few days. Slightly yellow-green crystals
were collected and dried under vacuum (yield 0.117 g (78.0%)).
Crystals for X-ray diffraction were grown from ether solvent
inside an NMR tube. IR data (Nujol mull): 3054 w, 2965 s, 2950
s, 2885 s, 2854 s, 2812 w, 1620 m, 1487 s, 1435 m, 1363 m, 1338 m,
1242 m, 1180 s, 1113 s, 1045 w, 1025 w, 999 m, 841 s, 791 s, 779 s,
759 m, 749 m, 709s, 692 m, 662 m. 1H NMR (THF-d8): δ 7.72 (t,
phenyl, 8H), 7.38 (m, phenyl, 4H), 7.29 (m, phenyl, 8H), 3.8 (b,
-CH2-, 2H), -0.25 (s, -Si(CH3)3, 18H). 13C{1H,13P} NMR
(THF-d8): δ 132.08 (s, phenyl), 131.51 (s, phenyl), 127.43 (s,
phenyl), 50.00 (br, -CH2-), 3.33 (s, -Si(CH3)3). 31P{1H} NMR
(THF-d8): δ 12.7 (broad peak). Anal. Calcd for C31H40N2P2Si2I-
Cu(C4H10O): C, 51.06; H, 6.08; N, 3.40. Found: C, 51.22; H, 5.95;
N, 3.47. 1H NMR in THF-d8 also detected the ether solvent
molecule quantitatively.
Synthesis of [Cu6{C(Ph2PdNSiMe3)2}2I2] (4). To an ether
(10 mL) suspension of [(cod)2Cu2I2] (0.178 g, 0.3 mmol) was
added white solid [Li2C(Ph2PdNSiMe3)2]2 (0.115 g, 0.1 mmol),
and the mixture was stirred at room temperature for a few hours.
The reaction mixture eventually turned yellow-orange. When
the reaction was finished, the insoluble solid was decanted by
centrifugation. The resultant clear yellow-orange solution was
kept in a vial for a few days. Bright orange crystals deposited and
were collected and dried under vacuum (yield 0.152 g (87.0%)).
Crystals for X-ray diffraction were grown from ether solvent
inside the NMR tube. IR data (Nujol mull): 3047 w, 2930 s, 2856
s, 1465 s, 1437 m, 1377 m, 1367 m, 1257 m, 1245 s, 1102 s, 1078 s,
1061 s, 1029 m, 832 s, 777 s, 751 s, 737 m, 713 m, 692 s, 672 m. 1H
NMR (THF-d8): δ 7.77 (q, phenyl, 4H), 7.61 (q, phenyl, 4H),
7.52 (q, phenyl, 4H), 7.44 (m, phenyl, 4H), 7.36 (t, phenyl, 4H), 7.28
(t, phenyl, 4H), 7.13 (m, phenyl, 4H), 7.05 (m, phenyl, 4H), 6.98
(m, phenyl, 4H), 6.92 (m, phenyl, 8H), -0.22 (s, -Si(CH3)3, 18H),
-0.29 (s, -Si(CH3)3, 18H). 13C{1H} NMR (THF-d8): δ 138.05
Crystal Structure Determination. Suitable crystals of 1-4, 6,
and 7 were mounted on glass fibers by means of mineral oil, and
data were collected using graphite-monochromated Mo KR
˚
radiation (0.710 73 A) on a Bruker D8/APEX II CCD diffrac-
tometer. The structures were solved by direct methods using
SHELXL-9725 and refined using full-matrix least squares on
(25) Sheldrick, G. M. Acta Crystallogr. 2008, A64, 112–122.

4264 Organometallics, Vol. 29, No. 19, 2010
Ma et al.
F2 (SHELXL-97).26 All of the non-hydrogen atoms in the struc-
ture were refined with anisotropic displacement parameters.
Selected crystal data and structure refinement details for all
the compounds are given in Table 3.
Acknowledgment. We thank the Natural Sciences and
Engineering Research Council of Canada and the Uni-
versity of Alberta for financial support. We also thank the
University of Alberta for supporting the X-ray Crystallo-
graphy, NMR and Analytical and Instrumentation
laboratories at the Department of Chemistry. We thank
Computational Details. The geometric optimizations of some
selected core structures were performed with the Gaussian03
program22 using the density functional theory (DFT) method.
The Becke three-parameter hybrid functional with the Lee-Yang-
Parr27 correlation functional (B3LYP)28 was employed for all
calculations. Gaussian NBO analysis was carried out for the final
optimized structures.22 The LanL2DZ basis set on Cu and halides
and the 6-31G(d) basis set on Si, P, N, C, and H were used for all the
calculations. In order to save computing time, models were used in
which the outside phenyl ring was replaced with methyl groups and
original methyl groups were replaced with H, except in the calcula-
tion of 1, where the methyl groups on Si were retained. Also, one
calculation for 4 was done with the complete molecular structure
including phenyl substituents (see the Supporting Information).
ꢀ
Jose Rodriguez-Nunez and Lydia Glover (University of
Alberta) for assistance with measurements of the photo-
luminescence spectra (PL) and Profs. Al Meldrum (Physics,
University of Alberta) and Jonathan G. C. Veinot
(Chemistry, University of Alberta) for access to the
solid and solution PL spectrometers.
Supporting Information Available: For 1-4, 6, and 7, CIF
files giving atomic positions and displacement parameters,
anisotropic displacement parameters, bond lengths, and bond
angles, figures giving mass spectra, NMR spectra, and TGA
analyses, and tables of crystal data, computed OPT structures,
molecular orbitals, and NBO, as well as details of TD-DFT.
This material is available free of charge via the Internet at http://
pubs.acs.org.
(26) Sheldrick, G. M. SHELXL-97: Program for Crystal Structure
€
€
Determination; University of Gottingen, Gottingen, Germany, 1997.
(27) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
(28) Becke, A. D. J. Chem. Phys. 1993, 98, 5648–5642.