Probing the electronic structures of [CU2(μ-XR 2)]n+ diamond cores as a function of the bridging X atom (X = N or P) and charge (n = 0, 1, 2)

Article abstract of DOI:10.1021/ja076537v

A series of dicopper diamond core complexes that can be isolated in three different oxidation states ([Cu₂(μ-XR₂)]ⁿ⁺, where n = 0, 1, 2 and X = N or P) is described. Of particular interest is the relative degree of oxidation of the respective copper centers and the bridging XR₂ units, upon successive oxidations. These dicopper complexes feature terminal phosphine and either bridging amido or phosphido donors, and as such their metal-ligand bonds are highly covalent. Cu K-edge, Cu L-edge, and P K-edge spectroscopies, in combination with solid-state X-ray structures and DFT calculations, provides a complementary electronic structure picture for the entire set of complexes that tracks the involvement of a majority of ligand-based redox chemistry. The electronic structure picture that emerges for these inorganic dicopper diamond cores shares similarities with the Cu ₂(μ-SR)₂ Cu_A sites of cytochrome c oxidases and nitrous oxide reductases.

Full text of DOI:10.1021/ja076537v

Published on Web 02/26/2008
Probing the Electronic Structures of [Cu2(µ-XR2)]n+ Diamond
Cores as a Function of the Bridging X Atom (X ) N or P) and
Charge (n ) 0, 1, 2)
Seth B. Harkins,† Neal P. Mankad,†,§ Alexander J. M. Miller,†
Robert K. Szilagyi,*,‡ and Jonas C. Peters*,†,§
DiVision of Chemistry and Chemical Engineering, Arnold and Mabel Beckman Laboratories of
Chemical Synthesis, California Institute of Technology, Pasadena, California 91125, and
Department of Chemistry and Biochemistry, Montana State UniVersity, Bozeman,
Montana 59717
Received September 1, 2007; E-mail: Szilagyi@Montana.edu; JPeters@MIT.edu
Abstract: A series of dicopper diamond core complexes that can be isolated in three different oxidation
states ([Cu2(µ-XR2)]n+, where n ) 0, 1, 2 and X ) N or P) is described. Of particular interest is the relative
degree of oxidation of the respective copper centers and the bridging XR2 units, upon successive oxidations.
These dicopper complexes feature terminal phosphine and either bridging amido or phosphido donors,
and as such their metal-ligand bonds are highly covalent. Cu K-edge, Cu L-edge, and P K-edge
spectroscopies, in combination with solid-state X-ray structures and DFT calculations, provides a
complementary electronic structure picture for the entire set of complexes that tracks the involvement of a
majority of ligand-based redox chemistry. The electronic structure picture that emerges for these inorganic
dicopper diamond cores shares similarities with the Cu2(µ-SR)2 CuA sites of cytochrome c oxidases and
nitrous oxide reductases.
Introduction
presence of highly covalent bonding between the copper centers
and the bridging thiolate (Cys) ligands.3e In the mixed-valence
An emerging theme in transition-metal coordination chemistry
is to consider the role of ligands in the stabilization of different
oxidation states and to view such ligands as redox-active.1 While
this view is distinct from the traditional model where oxidation
state changes are predominantly attributed to redox-active
transition-metal centers, it is in accord with a covalent bonding
description due to ligand-to-metal charge donation. Strong
covalency is apparently exploited within the active sites of
certain metalloenzymes.2 Relevant to the present study, the
bimetallic diamond core CuA sites found in cytochrome c
oxidases and nitrous oxide reductases serve as electron-transfer
relays with highly covalent thiolate bridging units that are key
to their function.3 A relevant investigation of the CuA center
engineered into Pseudomonas aeruginosa azurin revealed the
form of the CuA site ([Cu2(µ-SR)2]), spectroscopic and theoreti-
cal work have provided a picture of the electronic structure
originating from four major contributions: total sulfur (46%)
and â-methylene (3%) groups of the bridging cysteine ligands,
total nitrogen (6%) of the terminal histidine ligands, and the
two Cu centers (44%).
A biomimetic mixed-valence model complex that is structur-
ally faithful by virtue of its thiolate bridges, [(LiPrdacoSCu)2]+,4
was suggested to have even more sulfur character: 54% µ-S
and 38% Cu 3d contributions. This model system features
copper centers at a distance (2.93 Å)4 that is nonetheless
substantially larger than that found in CuA sites (∼2.5 Å).5 The
numerical assignments for the relative degree of copper versus
thiolate character within the redox-active molecular orbital/s
(RAMOs) of Cu2(µ-SR)2 mixed-valence systems suggest that
strong covalency contributes to the unique electron transfer (ET)
properties of CuA sites. However, the [(LiPrdacoSCu)2]+ model
complex does not serve as a functional model of the CuA site
as it lacks the requisite, reversible one-electron redox couple
between the CuICuI and the Cu1.5Cu1.5 states. To gain a better
appreciation of how covalency gives rise to the desirable ET
properties of CuA
† California Institute of Technology.
‡ Montana State University.
§ Present address: Massachusetts Institute of Technology.
(1) (a) Ray, K.; Bill, E.; Weyhermuller, T.; Wieghardt, K. J. Am. Chem. Soc.
2005, 127, 5641. (b) Bart, S. C.; Lobkovsky, E.; Bill, E.; Wieghardt, K.;
Chirik, P. J. Inorg. Chem. 2007, 46, 7055. (c) Szilagyi, R. K.; Lim, B. S.;
Glaser, T.; Holm, R. H.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. J.
Am. Chem. Soc. 2003, 125, 9158.
(2) Holm, R. H.; Kennepohl, R.; Solomon, E. I. Chem. ReV. 1996, 96, 2239.
(3) (a) Ferguson-Miller, S.; Babcock, G. T. Chem. ReV. 1996, 96, 2889. (b)
Gamelin, D. R.; Randall, D. W.; Hay, M. T.; Houser, R. P.; Mulder, T. C.;
Canters, G. W.; de Vries, S.; Tolman, W. B.; Lu, Y.; Solomon, E. I. J.
Am. Chem. Soc. 1998, 120, 5246. (c) Ramirez, B. E.; Malmstrom, B. G.;
Winkler, J. R.; Gray, H. B. Proc. Natl. Acad. Sci. U.S.A. 1995, 92, 11949.
(d) Antholine, W. E.; Kastrau, D. H. W.; Steffens, G. C. M.; Buse, G.;
Sumft, W. G.; Kroneck, P. M. H. Eur. J. Biochem. 1992, 209, 875. (e)
DeBeer-George, S.; Metz, M.; Szilagyi, R. K.; Wang, J.; Cramer, S. P.;
Lu, Y.; Tolman, W. B.; Hedman, B.; Hodgson, K. O.; Solomon, E. I. J.
Am. Chem. Soc. 2001, 123, 5757.
(4) Houser, R. P.; Young, V. G., Jr.; Tolman, W. B. J. Am. Chem. Soc. 1996,
118, 2101.
(5) (a) Blackburn, N. J.; Barr, M. E.; Woodruff, W. H.; van der Oost, J.; de
Vries, S. Biochemistry 1994, 33, 10401. (b) Williams, M.; Lapplalainen,
P.; Kelly, M.; Sauer-Eriksson, E.; Saraste, M. Proc. Natl. Acad. Sci. U.S.A.
1995, 92, 11955.
9
3478
J. AM. CHEM. SOC. 2008, 130, 3478-3485
10.1021/ja076537v CCC: $40.75 © 2008 American Chemical Society

Dicopper Diamond Core Complexes
A R T I C L E S
Scheme 1
sites, it is of interest to probe in detail Cu2(µ-X)2 sites that can
be studied as a function of various formal oxidation states.
In this paper we describe two unique sets of dicopper
complexes in which the Cu2(µ-XR2)2 diamond core topologies
are preserved across three formal oxidation states, providing a
unique opportunity to explore factors affecting the properties
of Cu2(µ-XR2)2 systems. Synthetic, X-ray diffraction (XRD),
and X-ray absorption spectroscopy (XAS) studies have been
used to probe the relationship between the redox behavior, geo-
metric structures, and electronic structures of these Cu2(µ-XR2)2
cores on the bridging X unit (X ) N vs P) and the overall
molecular charge (n ) 0, 1, 2). While the nature of the bridging
ligands employed in this study is distinct from the thiolate units
found in biological systems, striking electronic structural
similarities are evident between the redox-active molecular
orbitals of these inorganic model complexes and those that have
been elucidated for the CuA site.
oxidant such as [NO][BF4] were not fruitful. Instead, an inky
violet dicopper(I) degradation product, denoted herein as [Cu2-
(PNP-2H)2][BF4]2, was isolated in which C-C coupling of the
[PNP] ligands had occurred at the aryl ring in a position para
to the N atom (Scheme 1). The solid-state crystal structure of
this degradation product, which features two isolated, three-
coordinate copper(I) centers, is provided in the Supporting
Information. Loss of H2 or some byproduct derived from formal
loss of 2 H‚ can be inferred based upon the reaction product,
though no attempt has yet been made to study this transforma-
tion more carefully.
For the purposes of this study, it was necessary to add stability
to both the one- and two-electron oxidized species of the (PNP)
ligand system. To do this, a tert-butyl group para to the amido
N atom has been incorporated into the (PNP) ligand, and this
latter derivative is abbreviated herein as (tBu2-PNP) (Figure
1). (tBu2-PNP)H can be prepared by literature methods,9 and
subsequent reaction with mesitylcopper affords yellow and
highly emissive {(tBu2-PNP)Cu}2 (4) in good yield. The cyclic
voltammogram of 4 exhibits two reversible redox processes at
-0.49 and +0.30 V vs Fc+/Fc. Both processes are anodically
shifted relative to their corresponding couples in the {(PPP)-
Cu}2n+ system (-1.02 and -0.42 V).6 The oxidation of 4 with
Results and Discussion
For the purposes of this study, we required access to
synthetically well-defined dicopper systems for which amido
- 6,7
(NR2
)
or phosphido (PR2-)8 units occupied the bridging
positions and for which each series was stable across three
oxidation states. The synthesis and X-ray crystal structures of
the series of phosphide-bridged dicopper complexes {(PPP)-
1 equiv of either [FeCp2][BArF ] (ArF ) 3,5-(CF3)2-C6H3) or
4
AgSbF6 produces [{(tBu2-PNP)Cu}2]+ (5) as red-brown crys-
tals. The addition of 2 equiv of [NO][SbF6] to 4 in CH2Cl2
solution results in a deep blue color with an associated
appearance of a sharp optical transition at 16 100 cm-1 (ꢀ )
10 700 M-1 cm-1). The dicationic product of this reaction,
[{(tBu2-PNP)Cu}2][SbF6]2 (6), is sufficiently stable to be
isolated in crystalline form. The X-band electron paramagnetic
resonance (EPR) signal for monocation 5 features a very broad,
featureless isotropic signal with gx ) 1.987, gy ) 2.025, and gz
) 2.098. The absorption spectrum of 5 has a characteristic
charge-transfer band at 5330 cm-1 (ꢀ ) 3200 M-1 cm-1) that
is believed to be predominantly ligand-to-ligand in character
+
2+
Cu}2 (1), {(PPP)Cu}2 (2), and {(PPP)Cu}2 (3) ([PPP]-
)
bis(2-diisopropylphosphinophenyl)phosphide) were communi-
cated recently.8 A structurally analogous, highly luminescent
amido-bridged species {(PNP)Cu}2 ((PNP) ) bis(2-diisobu-
tylphosphinophenyl)amide) has also been described.7 Problem-
atic for this latter system, however, is that only {(PNP)Cu}2
and its monocation [{(PNP)Cu}2]+ can be isolated, despite the
presence of two independent and reversible redox couples in
the cyclic voltammetry data. Attempts to synthetically generate
the dication, [{(PNP)Cu}2]2+, by the addition of 2 equiv of an
(6) Harkins, S. B.; Peters, J. C. J. Am. Chem. Soc. 2004, 126, 2885.
(7) Harkins, S. B.; Peters, J. C. J. Am. Chem. Soc. 2005, 127, 2030.
(8) Mankad, N. P.; Rivard, E.; Harkins, S. B.; Peters, J. C. J. Am. Chem. Soc.
2005, 127, 16032.
(9) Ozerov, O. V.; Guo, C.; Papkov, V. A.; Foxman, B. M. J. Am. Chem. Soc.
2004, 126, 4792.
9
J. AM. CHEM. SOC. VOL. 130, NO. 11, 2008 3479

A R T I C L E S
Harkins et al.
Figure 1. Thermal ellipsoid representations of the core atoms of neutral 4, cation 5, and dication 6. See selected bond lengths (Å) above. Selected angles
(deg) follow. For 4: Cu1-N1-Cu2, 78.77(1); Cu1-N2-Cu2, 78.52(1); N1-Cu1-N2, 101.35(1); N1-Cu2-N2, 101.19(1). For 5: Cu1-N1-Cu2, 73.75-
(8); Cu1-N2-Cu2, 72.56(8); N1-Cu1-N2, 102.03(1); N1-Cu2-N2, 110.5(1). For 6: Cu1-N1-Cu2, 70.83(8); Cu1-N2-Cu2, 68.60(8); N1-Cu1-N2,
95.19(8); N1-Cu2-N2, 125.31(1).
(vide infra). Dication 6 is paramagnetic, contrasting its diamag-
netic phosphide-bridged relative 3.6 X-band EPR data for 6 in
a frozen dichloromethane/toluene (20:1) glass at 3.7 K show a
broad signal at g ) 1.998 and a weak half-field signal at g )
4.02, consistent with an S ) 1 spin system. No hyperfine
coupling information is resolvable over the temperature range
of 3.7-154 K.
experimentally determined by a change in the effective nuclear
charges of a system can differ substantially from formal
oxidation state assignments due to considerable donation/
removal of electron density by the metal’s donor ligands. The
effective nuclear charges and relative differences in effective
oxidation states can be generally obtained by examining the
rising K-edge inflection points of the metal center in question
and the K-edges of the donor atoms coordinated to the metal.
As shown in Figures 2A and 3A, the rising-edge inflection points
of the Cu K-edge spectra of 1 and 4 are between the reference
Figure 1 depicts the core ellipsoids obtained from high-quality
X-ray crystal structures for 4, cation 5 as its [BArF ] salt, and
4
dication 6 as its [SbF6] salt. These three solid-state structures
spectra for CuI(pyridine)4 (8986.1 eV) and CuIICl2 (8991.6
+
are to be compared with the previously reported set of crystal
structures for the {(PPP)Cu}2 series 1-3.6 Immediately
eV).11 Upon oxidation, the inflection points shift to higher
energy by less than 0.5 eV for 1-5. The small variation in the
energy of the inflection points lies approximately within the
resolution of the XAS technique at ∼9000 eV. Therefore, by
this measure the Cu centers in complexes 1-5 appear to have
similar effective oxidation states. The Cu K-edge spectrum of
the dication 6 (red trace) shows a 1.7 eV shift relative to neutral
4 (blue trace), which could be a result of substantial differences
in their effective oxidation states. To evaluate this possibility,
we estimated the Cu K-edge spectra of complexes with formally
four- and two-coordinate CuI centers (Figure 4A) and also a
complex with four-coordinate CuI and CuIII centers (Figure 4B).
Comparison of the measured spectrum of 6 with these estimated
spectra shows a close similarity between that for 6 and the four-
and two-coordinate copper centers. We therefore conclude that
the 1.7 eV shift that is observed in the Cu K-edge of 6 results
from the superposition of the K-edges of pseudo four- and two-
coordinate copper(I) centers, rather than from substantial
oxidation of the copper centers.
Figures 2B and 3B show two groups of spectral features at
the Cu L3-edge. The pre-edge features at around 932 eV arise
due to unoccupied Cu 3d orbital character. For the (PPP) series,
the neutral complex 1 shows no preedge at 931 eV (purple trace)
due to its filled 3d-shell. Pre-edge transitions arise at 933.3 eV
upon single and double oxidation steps to 2 (olive) and 3
(orange). For the (tBu2-PNP) series, the neutral complex 4
shows essentially no preedge feature (blue) due to the 3d10
configuration. As this complex is oxidized (green), a small pre-
edge feature emerges at 931.7 eV indicating limited metal
character in the RAMO. For the reference CuIICl2 spectrum the
pre-edge feature corresponds to 71% Cu 3d character.12 By
comparing the area of the pre-edge feature of 5 with that for
n+
evident for 4-6 is a common topology that is also seen in 1-3.6
The Cu‚‚‚Cu distance in 4 is 2.7283(3) Å, and it contracts by
ca. 0.2 Å upon oxidation to 5. While there is neglible change
in the Cu‚‚‚Cu distance between 5 and 6, there is a dramatic
asymmetric structural distortion of the Cu2(µ-NR2)2 core result-
ing in two elongated Cu-N distances (2.34 and 2.45 Å) and
two typical Cu-N distances (2.00 and 1.98 Å) for dicationic 6.
The unexpected asymmetry in the structure of 6 brings up
the possibility that the copper centers have different oxidation
states. A simplistic analysis suggests a pseudo-two-coordinate
CuI center that is very weakly coordinated to the bridging amido
units, and a four-coordinate CuII or CuIII center much more
tightly coordinated to the bridging amido units. The related,
but diamagnetic phosphide-bridged dication, {(PPP)Cu}22+ 3,6
has a symmetric Cu2(µ-PR2)2 core structure that can mostly
simply be described by two antiferromagnetically coupled CuII
centers. Hence, the relative oxidation states between the
monocations 2 and 5 and the dications 4 and 6 appear to depend
on the chemical nature of the bridging XR2 unit. But due to the
expected highly covalent bonding, redox chemistry likely affects
redox active molecular orbitals (RAMOs) with significant Cu
and bridging XR2 unit contributions. Such an orbital was initially
+
identified by a DFT study of {(SNS)Cu}2 ((SNS) ) bis(2-
tert-butylsulfanylphenyl)amido).6 Hence, a more careful analysis
is needed to establish the relative degree of oxidation of the
copper centers and the bridging ligands as electrons are removed
from these Cu2(µ-XR2)2 systems.
To experimentally investigate the RAMOs of 1-6, we
undertook a multi-edge X-ray absorption spectroscopic study
at the Cu K-, Cu L-, and P K-edges. These edges directly probe
the orbital composition (metal and ligand covalency) of the
frontier unoccupied orbitals and the effective oxidation states
of the Cu and P atoms.10 Effective oxidation states that are
(11) Kau, L. S.; Spira-Solomon, D. J.; Pennerhahn, J. E.; Hodgson, K. O.;
Solomon, E. I. J. Am. Chem. Soc. 1987, 109, 6433.
(12) (a) Gewirth, A. A.; Cohen, S. L.; Schugar, H. J.; Solomon, E. I. Inorg.
Chem. 1987, 26, 1133. (b) Didziulis, S. V.; Cohen, S. L.; Gewirth, A. A.;
Solomon, E. I. J. Am. Chem. Soc. 1988, 110, 250.
(10) Solomon, E. I.; Hedman, B.; Hodgson, K. O.; Dey, A.; Szilagyi, R. K.
Coord. Chem. ReV. 2005, 249, 97.
9
3480 J. AM. CHEM. SOC. VOL. 130, NO. 11, 2008

Dicopper Diamond Core Complexes
A R T I C L E S
Figure 2. Multi-edge (Cu K-edge, A; Cu L-edges, B; P K-edge, C) X-ray
absorption spectra for {(PPP)Cu}2n+, n ) 0 (1), 1 (2), 2 (3). Key to notes:
(a) Taken from ref 11. (b) The pre-edge features of {(PPP)Cu}2+ at ∼931
and ∼932 eV are due to the presence of endogeneous CuII species and
partially oxidized mesitylcopper(I) impurity. (c) A similar group of features
are also present at the Cu L2-edge, which indicates that they originate from
Cu-based transitions. (d) The (tBu2-PNP) ligand spectrum is renormalized
by 0.5 relative to the spectra of the other complexes shown, as it contains
only half as many phosphorus absorbers.
Figure 3. Multi-edge (Cu K-edge, A; Cu L-edges, B; P K-edge, C) X-ray
absorption spectra for {(tBu2-PNP)Cu}2n+, n ) 0 (4), 1 (5), 2 (6). Key to
notes: (a) Taken from ref 11. (b) A similar group of features is also present
at the Cu L2-edge, which indicates that they originate from Cu-based
transitions. (c) The (tBu2-PNP) ligand spectrum is renormalized by 0.5
relative to the spectra of the other complexes shown, as it contains only
half as many phosphorus absorbers.
Fe L-edge XAS for [Fe(CN)6]3/4- 13
In addition, we have
.
CuCl2, approximately 24% Cu 3d character can be assigned to
the RAMO of 5. For comparison, the Cu 3d character in the
thiolate-bridged, mixed-valence model complex [(LiPrdacoSCu)2]+
was estimated at 38%.3e Upon further oxidation of 5, another
electron hole is opened in the RAMO-1 orbital (red, 931 eV).
The comparatively small amount of Cu orbital character
indicated by these Cu L3-edge features supports the assignment
of a substantial degree of ligand-based redox chemistry upon
one- and two-electron oxidation steps in both the (PPP) and
(tBu2-PNP) series.
recently suggested that donation from an amido lone pair,
admixed with Cu 3d-character, into an orbital of aryl π*
character is likely to be one key element that gives rise to the
unusual emission characteristics of copper(I) arylamidophos-
phine complexes.14 The observation of strong backbonding
contributions in these Cu L3-edge spectra is consistent with such
a suggestion. Moreover, the presence of these features at similar
energies and intensities in all of the spectra suggests that the
copper centers maintain considerable CuI character as electrons
are removed from the dinuclear systems. Note also that the
single and double pre-edge features at 933.3 and 931-931.7
Transitions at 934-937 eV suggest a considerable amount
of Cu f L backdonation, presumably into the π* manifold of
the aryl rings. Solomon et al. have recently observed related
spectral features that result from backbonding contributions by
(13) Hocking, R. K.; Wasinger, E. C.; de Groot, F. M. F.; Hodgson, K. O.;
Hedman, B.; Solomon, E. I. J. Am. Chem. Soc. 2006, 128, 10442.
(14) Miller, A. J.; Dempsey, J. L.; Peters, J. C. Inorg. Chem. 2007, 46, 7244.
9
J. AM. CHEM. SOC. VOL. 130, NO. 11, 2008 3481

A R T I C L E S
Harkins et al.
is the 50% scaled spectrum of the {(tBu2-PNP)Cu}2 (green
+
spectrum). It matches quite well with the rising edge features
of 6 up to 2147 eV. This suggests that the Cu-P interactions
of the four-coordinate site in 6 are similar to those in the one-
electron more reduced complex 5. The violet spectrum in Figure
3C is the difference spectrum between the red and pink spectra
and thus corresponds to the pseudo-two-coordinate Cu site. The
three shoulders at 2148.0, 2148.8, and 2149.6 eV can be
correlated with the Cu-P bonding interaction and the C-P σ*
interactions. The shortening of the Cu-P distances in the
pseudo-two-coordinate site relative to the pseudo-four-coordinate
site, as observed in their respective X-ray crystal structures, are
in accord with these spectral observations.
The Cu-P bond distances are symmetric in the (PPP) series
1-3, and the rising edges in Figure 2C are rather featureless.
The small pre-edge features closer to the rising edge at 2144
eV reflect a weak terminal P(3p)-Cu(3d) covalent interaction.
However, the P K-edge spectra of the (PPP) series show a
second, more intense pre-edge feature at 2142.5 eV that
corresponds to the µ-PR2 f Cu 3d bonding interaction. As the
(PPP) complexes are oxidized, this pre-edge feature gradually
increases, confirming considerable µ-PR2 character in the
RAMO. An independent theoretical study of 1-3 by Head-
Gordon and co-workers provides further support for the non-
innocent nature of the (PPP) ligand system and strongly suggests
that oxidation at the covalent bridging µ-PR2 unit dominates,
rather than oxidation of the aryl rings.15 As they point out, our
observation6 of a strongly downfield-shifted 31P NMR µ-P signal
(266.1 ppm) for dication 3, compared to that for neutral 1 (-21.0
ppm), provides independent experimental evidence that the
phosphide units are being oxidized upon electron removal.
Figure 4. Predicted Cu K-edge spectra of a Cu2L2 complex from (A) two-
coordinate and four-coordinate CuI complexes and (B) from four-coordinate
CuI and CuIII complexes. Key to notes: (a) Data for B were taken from ref
11. (b) The four-coordinate CuIII spectrum was predicted from an averaged
+2 eV shifted CuCl2 and a +4 eV shifted {Cu(pyridine)4}{ClO4}. These
relative energy changes were estimated from edge shift positions given in
ref 26. The approximately 2 eV blue shift corresponds to the effective
nuclear charge increase upon one electron oxidation.
eV for 3 (orange trace) and 6 (red trace) reflect their singlet vs
triplet ground states, respectively. The inflection points for the
(tBu2-PNP) series are ca. 1 eV lower in energy than for the
(PPP) series, indicative of more cuprous character in the amido-
bridged set.
The experimentally probed RAMOs for 1-6 were also
examined by electronic structure calculations. Orbital plots
clearly show large bridging atom contributions. For example,
for complexes 1 and 4 the bridging atom contributions are ∼44%
and ∼35%, respectively, contrasting the modest terminal
phosphorus contributions of 14% and 11% (Figure 5). Further-
more, for 1 and 4 about 22% and 38% C and H characters are
delocalized into the aromatic π systems of the (PPP) and (PNP)
ligands, respectively. This leaves approximately 18% and 16%
Cu 3d character equally shared between the two metal sites, to
be compared with the experimental estimate of Cu 3d character
of approximately 24% per orbital from the Cu L-edge data
(Figures 2B and 3B; see also discussion above). It is important
to emphasize that the direct correlation between the experimental
metal-ligand covalencies and calculated molecular orbital
coefficients is limited by the lack of a transition dipole integral
for the P 1s f 3p transition. Nonetheless, the orbital composi-
tions of the Cu2(µ-XR2)2 cores presented herein are qualitatively
similar to those described for the CuA site.3e
The P K-edge spectra of 1-3 and 4-6 in Figures 2C and
3C, respectively, reflect the orbital contributions from phos-
phorus ligands. The weak pre-edge features at 2143.5 eV in
Figure 3C are indicative of modest terminal P(3p)-Cu(3d)
covalent interactions. These pre-edge features appear to be
slightly weaker in complexes 4-6; however, this is due to their
larger separation at 2144 eV (Figure 2C) from the rising-edge
compared to 1-3. The smaller PfCu donation is likely
compensated by an increased bridging amido NfCu donation
in 4-6 relative to bridging phosphido PfCu donation in 1-3.
The increased donation is already reflected by the ca. 0.2 Å
shorter Cu-N bonds, relative to the Cu-P bonds, even
considering the greater delocalization of the µ-N (Figure 5B,C)
than the µ-P (Figure 5A) into the aromatic carbon rings. The
rising edge shifts between the free (tBu2-PNP)H ligand and
complexes 4-6 are due to the terminal PfCu 4s/4p interactions.
The larger shift and different rising-edge structure for the most
oxidized complex 6 is due to the presence of two distinct
coordination sites. The four-coordinate site has smaller P
covalency relative to the pseudo-two-coordinate site due to
stronger interaction with the µ-amido group. The pseudo-two-
coordinate Cu site receives electron density dominantly from
the terminal phosphines, which increases the effective nuclear
charge of these P atoms and shifts their spectral features to
higher energy, by about 3 eV. In Figure 3C, the pink spectrum
The spin density plot for triplet 6 is shown in Figure 5D also
indicates that the electron density corresponding to the two
unpaired electrons is predominantly located (1.65 electrons) on
the ligands. Furthermore, the spin density contour plot clearly
shows the large difference between the two Cu coordination
environments, a difference that is also reflected by the drastic
changes in the Cu K- and P K-edge spectra (parts A and B of
Figure 3, respectively) relative to 4 and 5.
(15) Rhee, Y. M.; Head-Gordon, M. , submitted for publication.
9
3482 J. AM. CHEM. SOC. VOL. 130, NO. 11, 2008

Dicopper Diamond Core Complexes
A R T I C L E S
the loss of π-bonding interactions between the µ-P 3p orbitals,
mediated by overlap with Cu 3d orbitals (see Figure 5A).
In summation, the redox chemistry in these (PPP) and (tBu2-
PNP) dicopper complexes features a strong ligand contribution,
with a majority of the redox chemistry occurring at the bridging
N or P atoms. The effective oxidation states of the Cu centers
remain relatively constant throughout the two-electron-transfer
series, lying closest to the cuprous state. The more reduced
effective Cu oxidation states relative to the formal oxidation
states originate from the combined effect of metal-to-ligand
backdonation and substantial ligand-based redox due to the
highly covalent nature of the Cu2(µ-NR2)2 and Cu2(µ-PR2)2 core
units. These inorganic model diamond cores thus represent close
analogues of the Cu2(µ-SR)2 diamond core unit of the CuA site.
Experimental Section
General. All manipulations were carried out using standard Schlenk
or glovebox techniques under a dinitrogen atmosphere. Unless otherwise
noted, solvents were deoxygenated and dried by thorough sparging with
N2 gas followed by passage through an activated alumina column.
Nonhalogenated solvents were tested with a standard purple solution
of sodium benzophenone ketyl in tetrahydrofuran to confirm effective
oxygen and moisture removal. All reagents were purchased from
commercial vendors and used without further purification unless
otherwise stated. Mesitylcopper(I),16 diethylphosphoramidous dichlo-
ride,17 4,4′-di-tert-butyldiphenylamine,18 and [Cp2Fe][B(C6H3(CF3)2)4]19
were prepared according to literature procedures. Elemental analyses
were performed by Desert Analytics, Tucson, AZ. Deuterated solvents
were purchased from Cambridge Isotope Laboratories, Inc., and
degassed and dried over activated 3 Å molecular sieves prior to use. A
Varian Mercury-300 or INOVA-500 NMR spectrometer was used to
record H, 13C, 19F, and 31P NMR spectra at ambient temperature. H
chemical shifts were referenced to residual solvent. 19F chemical shifts
were referenced to external C6F6 (δ ) -165 ppm), and 31P chemical
shifts were referenced to external phosphoric acid (δ ) 0 ppm). GC-
MS data was obtained by injecting a dichloromethane solution into an
Agilent 6890 GC equipped with an Agilent 5973 mass selective detector
(EI). High-resolution EI mass spectroscopy was carried out by the
Caltech Chemistry Mass Spectral Facility using a JEOL JMS600
instrument. UV-vis measurements were taken on Cary 500 UV-vis-
NIR or Cary 50 UV-vis spectrophotometers using 1.0 or 0.1 cm quartz
cells with a Teflon stopper. XRD studies were carried out in the
Beckman Institute Crystallographic Facility on a Bruker Smart 1000
CCD diffractometer, and special refinement details are listed in the
comments sections of the CIF files (Supporting Information). Magnetic
measurements were undertaken at the Molecular Materials Research
Center at the California Institute of Technology. Photophysical data
were gathered at the Beckman Institute Laser Resource Center.
1
1
Figure 5. Molecular orbital plots of the redox-active orbitals (green and
yellow) of 1 (A, HOMO) and 4 (B, HOMO; C, HOMO - 1); spin density
EPR Measurements. X-band EPR spectra were obtained on a
Bruker EMX spectrometer (controlled by Bruker Win EPR software,
version 3.0) equipped with a rectangular cavity working in the TE102
mode. Variable-temperature measurements were conducted with an
Oxford continuous-flow helium cryostat (temperature range 3.6-300
K). Accurate frequency values were provided by a frequency counter
built into the microwave bridge. Solution spectra were acquired in
2-methyltetrahydrofuran or 20:1 dichloromethane/toluene solution.
Sample preparation was performed under a dinitrogen atmosphere in
an EPR tube equipped with a ground glass joint.
2+
plot (red and blue) for the triplet {(tBu2-PNP)Cu}2 complex (D).
Conclusions
The combined XAS and DFT data appear to rule out the
possibility of 6 being a valence-trapped complex with CuIII and
CuI sites. The small shift of the rising-edge positions at the Cu
K-edge, the lack of intense pre-edge features at the Cu L-edge,
and the increasing pre-edge intensities at the P K-edge indicate
in sum similar CuI character in each Cu2(µ-XR2)2 core inde-
pendent of the overall charge and the bridging group.
Electrochemistry. Electrochemical measurements were carried out
in a glovebox under a dinitrogen atmosphere in a one-compartment
The dramatic contraction of the Cu‚‚‚Cu distances (0.2-0.5
Å) upon oxidation is one fascinating structural feature of these
dicopper systems. A contributing electronic factor for the (PPP)
series, where the contraction is most dramatic, is likely to be
(16) Eriksson, H.; Håkansson, M. Organometallics 1997, 16, 4243.
(17) Perich, J. W.; Johns, R. B. Synthesis 1998, 2, 142.
(18) Zhao, H.; Tanjutco, C.; Thayumanavan, S. Tetrahedron Lett. 2001, 42, 4421.
(19) Chavez, I. J. Organomet. Chem. 2000, 601, 126.
9
J. AM. CHEM. SOC. VOL. 130, NO. 11, 2008 3483

A R T I C L E S
Harkins et al.
cell using a CH Instruments 600B electrochemical analyzer. Platinum
wire was used as both the working and auxiliary electrodes. The
reference electrode was Ag/AgNO3 in THF. The ferrocene couple Fc+/
Fc was used as an external reference. THF solutions of electrolyte (0.35
M tetra-n-butylammonium hexafluorophosphate) and analyte were also
prepared under an inert atmosphere.
The molecular orbital diagrams were generated using the Cube utility
in Gaussian and the GOpenMol25 visualizer program.
Synthesis of [Cu2(PNP-2H)2][BF4]2. NOBF4 (22.5 g, 0.192 mmol)
and {(PNP)Cu}2 (100 mg, 0.096 mmol) were combined in chloroben-
zene and stirred for 18 h, after which the solution had become inky
purple in color. The solution was filtered through a glass microfilter
and layered with petroleum ether. Upon standing for 4 days, analytically
pure, X-ray-quality crystals were obtained (86 mg, 74%). The room-
temperature NMR of the crystalline material is poorly resolved, likely
due to a fluxional interaction of the BF4 counterion and the Cu centers.
This is evident from very broad 19F NMR resonances rather than the
sharp signal that is expected for an uncoordinated anion. 1H NMR (300
MHz, CD2Cl2): δ 8.7 (br d), 8.51 (br d), 8.2 (br d), 7.8-7.3 (m), 6.93
(s), 2.45-1.55 (m), 1.11 (d), 1.03 (d), 0.98-0.92 (m). 19F NMR (282
MHz, CD2Cl2): δ -108.0 (br s), -125.5 (br s). 31P{1H} NMR (121
MHz, CD2Cl2): δ -27--31 (m), -32.8 (s). Anal. Calcd for C56H86B2-
Cu2F8N2P4: C, 55.41; H, 7.31; N, 2.31. Found: C, 55.01; H, 7.38; N,
2.26. UV-vis (CD2Cl2, nm(M-1 cm-1)): 293 (7370), 327 (sh), 570
(sh), 613 (16250).
Synthesis of Diisobutylchlorophosphine. Neat diethylphosphora-
midous dichloride (58.0 g, 0.333 mol) was added to a 2.0 M solution
of isobutyl magnesium chloride in diethyl ether (350 mL) at 0 °C over
a period of 30 min. Following addition, the solution was stirred at
ambient temperature for 1 h and the crude 31P NMR results indicated
that the starting material had been consumed (31P NMR: 47.9 ppm,
(Et)2NP(iBu)2) and diisobutylchlorophosphine was the only phosphorus-
containing product. The solution was again cooled to 0 °C, and a 2.0
M solution of anhydrous HCl in ether (350 mL) was added via cannula
while stirring vigorously. A considerable amount of solid precipitated
over a 2 h period, and the supernatant was isolated by cannula filtration.
The solids were extracted with 200 mL of diethyl ether, and the solvent
was removed by fractional distillation. The crude viscous oil was
purified by fractional vacuum distillation (26-29 °C at 0.005 Torr)
followed by removal of the residual diethyl ether by prolonged exposure
to vacuum at -10 °C with stirring. The product was isolated as a
spectroscopically pure, colorless oil (38.6 g, 65%) which exhibited a
single resonance by 31P NMR spectroscopy consistent with previously
published results.18 1H NMR (300 MHz, C6D6): δ 1.87 (m, 2H), 1.78
(br m, 2H), 1.24 (br m, 2H), 0.90 (br s, 12H). 31P{1H} NMR (121
MHz, C6D6): δ 109.9.
Synthesis of Di(2-bromo-4-tert-butylphenyl)amine. In air, neat Br2
(3.6 mL, 0.071 mol) at ∼5 °C was added dropwise to a slurry of 4,4′-
di-tert-butylphenylamine (10.0 g, 0.0356 mol) in acetic acid at
∼16 °C. Following addition, the solution was stirred at ambient
temperature for 2 h, a dilute solution of Na2S2O4 (500 mL) was added,
and the resulting solution was stirred for 15 min. The solids were
collected on a frit and washed with H2O (3 × 200 mL). The crude
product was purified by crystallization at -20 °C from methanol/
chloroform as a white solid (12.35 g, 80%). 1H NMR (300 MHz,
CDCl3): δ 7.59 (m, 2H), 7.23 (m, 4H), 6.30 (br s, 1H) 1.32 (s, 18H).
13C{1H} NMR (75.5 MHz, CDCl3): 145.9, 137.9, 130.2, 125.3, 117.8,
114.1, 34.5, 31.5. GC-MS(ES): 439 (M), 424 (M - CH3).
Synthesis of Bis(2-diisobutylphoshino-4-tert-butylphenyl)amine,
(tBu2-PNP)H. A 1.6 M solution of n-butyl lithium in hexane (26 mL)
was added dropwise to a solution of di(2-bromo-4-tert-butylphenyl)-
amine (6.0 g, 13.7 mmol) in diethyl ether (100 mL) at -70 °C with
stirring. The solution immediately became yellow in color and was
stirred at ambient temperature for 4 h at which time a precipitate formed.
The reaction mixture was again cooled to -70 °C, at which time
diisobutylchlorophosphine (7.67 g, 41.7 mmol) was added as a 1:1
solution with diethyl ether and the reaction was allowed to warm to
room temperature. After 36 h at ambient temperature, a large amount
Details of XAS Data Collection. The copper K-, phosphorus K-,
and copper L-edge XAS measurements were carried out at BL2-3, BL6-
2, and BL10-1, respectively, of Stanford Synchrotron Radiation
Laboratory under storage ring (SPEAR 3) conditions of 3 GeV energy
and 100-80 mA current. BL2-3 is a 1.3 T bend magnet beam line and
equipped with a Si(220) downward reflecting, double-crystal mono-
chromator. Data was collected in the energy range of 8660-9690 eV
using an unfocused beam and a 13-element Ge fluorescence detector
array for samples placed in a liquid He cryostat. The beam intensity
was maximized at 9685 eV. BL6-2 is a 56-pole, 0.9 T Wiggler beam
line with a liquid-nitrogen-cooled, Si(111) double-crystal monochro-
mator. P K-edge spectra were collected in the energy range of 2120-
2250 eV using an unfocused beam in a He-purged fly path at room
temperature and a Lytle fluorescence detector. The beam line was
optimized at 2320 eV. BL10-1 has a 30-pole 1.45 T Wiggler insertion
device with 6 m spherical grating monochromator. The samples were
placed in a vacuum chamber with typical pressures of 10-6 to 10-8
torr, and the energy was scanned between 925 and 955 eV. The incident
beam intensity and beam line optics were optimized at 920 eV. Data
collection was carried out in electron yield mode by a Channeltron
detector with 1.5 kV accelerating potential.
The solid samples were ground and pasted onto a contaminant-free
Kapton tape from Shercon or carbon tape from Specs CertiPrep in a
glovebox with sub parts per million oxygen and moisture levels.
Samples for hard X-ray transmission measurements were diluted in
and ground together with boronitride to minimize incident beam
absorption. Samples were protected by a thin polypropylene window
(Specs CertiPrep) from exposure to air during sample mounting and
change. Sample holders for Cu L-edge measurements were mounted
in a He-purged glovebag tightly wrapped around the vacuum chamber.
The incident photon energy was scanned in 0.5 eV steps outside the
rising edge region where the step-size was 0.1 eV. At least five scans
were averaged to obtain a good signal-to-noise ratio. The incident
photon energy was calibrated to the spectra of copper foil at the Cu
K-edge (first inflection point at 8979 eV), difluorocopper(II) at the Cu
L3- and L2-edges (white line positions at 930.5 and 950.5 eV,
respectively), and triphenylphosphineoxide at the P K-edge (maximum
of pre-edge feature at 2147.5 eV).
Electronic Structure Calculations. Density functional calculations
were carried out using the Gaussian03 suite.20 The electronic structures
of the computational models [{(PXP)Cu}2]n+, where n ) 0, 1, 2 and
X ) N or P in the bis(2-dimethylphosphinophenyl)amide or phosphide,
respectively, which is a slightly truncated version of the (tBu2-PXP)
ligand described in the manuscript; their ionic fragments were calculated
by employing a gradient-corrected density functional composed of
Becke nonlocal and Slater local density functional exchange21 and
Perdew nonlocal and Vosko-Wilk-Nussair local density functional
correlation22 functions (BP or BP86). Calculations were carried out
using Stuttgart-Dresden effective core potential and corresponding
valence triple-ú basis set (ECP2).23 We have utilized an ionic-fragment-
based approach24 to achieve rapid wave function convergence and to
compare and contrast the possibility of various formal oxidation states.
(20) Frisch, M. J. et al. Gaussian03, Gaussian, Inc.: Pittsburgh PA, 2004.
(21) Becke, A. D. Phys. ReV. A: Gen. Phys. 1988, 38, 3098.
(22) Perdew, J. P. Phys. ReV. B: Condens. Matter Mater. Phys. 1986, 33, 8822.
(23) Dolg, M.; Wedig, U.; Stoll, H.; Preuss, H. J. Chem. Phys. 1987, 86, 866.
Wedig, U.; Dolg, M.; Stoll, H.; Preuss, H. In Quantum Chemistry: The
Challenge of Transition Metals and Coordination Chemistry; Veillard, A.,
Ed.; D Reidel Publishing Co.: Dordrecht: The Netherlands, 1986; Vol.
176, pp 79-89.
(25) (a) Laaksonen, L. J. Mol. Graph. 1992, 10, 33. (b) Bergman, D. L.;
Laaksonen, L.; Laaksonen, A. J. Mol. Graph. Model. 1997, 15, 301.
(26) DuBois, J. L.; Mukherjee, P.; Stack, T. D. P.; Hedman, B.; Solomon, E. I.;
Hodgson, K. O. J. Am. Chem. Soc., 2000, 122, 5775.
(24) Szilagyi, R. K.; Winslow, M. J. Comput. Chem. 2006, 27, 1385.
9
3484 J. AM. CHEM. SOC. VOL. 130, NO. 11, 2008

Dicopper Diamond Core Complexes
A R T I C L E S
of precipitate had formed and was removed by filtration through Celite
after diluting the reaction solution with diethyl ether (100 mL). The
filtrate was treated with a 1.0 M solution of anhydrous HCl in ether (2
× 50 mL) which immediately resulted in the precipitation of the product
as an HCl salt. The solids were collected on a frit and washed with
petroleum ether (3 × 50 mL). These solids were suspended in a 4:1
mixture of THF/CH3CN, excess (>4 equiv) NaOMe was added, and
the mixture was stirred for 2 h. The solvent was removed in vacuo.
The sticky solids were extracted with petroleum ether (50 mL) and
filtered through Celite. The solvent was removed from the filtrate under
reduced pressure. The resultant tacky translucent solid was stirred in
CH3CN for 1 h, affording a tractable white solid (5.35 g, 68%) which
was pure by NMR spectroscopy. 1H NMR (300 MHz, CD2Cl2): δ 7.66
(t, 1H), 7.46 (m, 2H), 7.16 (m, 4H), 1.34 (s, 18H), 1.00 (d, 6H), 0.94
(d, 6H). 13C{1H} NMR (75.5 MHz, CD2Cl2): δ 145.9, 143.5, 128.8,
127.4, 126.7, 116.6, 39.6, 34.9, 31.9, 27.1, 24.9, 24.6. 31P{1H} NMR
(121 MHz, CD2Cl2): δ -53.1. Anal. Calcd for C36H61NP2: C, 75.88;
H, 10.79; N, 2.46; Found: C, 75.35; H, 10.20; N, 2.81. UV-vis (THF,
nm(M-1 cm-1)): 303 (20 700), 344 sh.
Synthesis of {(tBu2-PNP)Cu}2 (4). Mesityl-Cu (128 mg, 0.702
mmol) and (tBu2-PNP)H (400 mg, 0.702 mmol) were dissolved in
petroleum ether and stirred for 12 h at ambient temperature. The
emissive yellow solution was filtered through a glass microfilter, and
the solvent was removed in vacuo. The bright yellow solid was stirred
in CH3CN for 1 h in the dark and collected on a frit. Drying of the
solids in vacuo afforded a finely divided analytically pure yellow solid
(400 mg, 90%). Crystals suitable for XRD were obtained by slow
evaporation of the solvent from a solution of 4 in THF. 1H NMR (300
MHz, C6D6): δ 7.43 (m, 4H), 7.02 (m, 8H), 2.12 (n, 8H), 1.84-1.50
(m, 16H) 1.34 (s, 36H), 1.15 (d, 12), 0.89 (dd, 24), 0.76 (d, 12H).
13C{1H} NMR (75.5 MHz, C6D6): δ 168.1, 139.5, 128.5, 128.2, 127.1,
125.4, 40.0, 36.3, 34.6, 32.3, 26.7, 26.5, 25.9, 25.8. 31P{1H} NMR (121
MHz, C6D6): δ -35.3. Anal. Calcd for C72H120Cu2N2P4: C, 68.38; H,
9.56; N, 2.21. Found: C, 68.55; H, 9.38; N, 2.61. UV-vis (THF,
nm(M-1 cm-1)): 295 (20 500), 318 (21 100), 357 (33 300), 383 sh,
433 (4600), 459 (3800).
20 mL reaction vial and dissolved in THF (10 mL). The reaction
immediately turned red-brown in color, and stirring was continued for
12 h, after which time the solution was filtered through a glass
microfilter to remove the Ag0 byproduct. The solvent was removed
from the filtrate under reduced pressure, and the resultant red solids
were extracted into CH2Cl2 (4 mL), filtered, and layered with petroleum
ether. After 2 days, analytically pure red crystals (159 mg, 81%) were
obtained. Anal. Calcd for C72H120Cu2F6N2P4Sb: C, 57.63; H, 8.06; N,
1.87. Found: C, 57.64; H, 8.03; N, 2.17. Evans’ method (CD2Cl2, 298
K): 1.63 µB. UV-vis (CD2Cl2, nm(M-1 cm-1)): 260 (sh), 312 (22
360), 377 (sh), 430 (3810), 510 (2280), 586 (sh), 639 (sh), 804 (1510),
941 (940), 1788 (5490).
Synthesis of [{(tBu2-PNP)Cu}2][SbF6]2 (6). Complex 4 (100 mg,
0.079 mmol) and NOSbF6 (42 mg, 0.18 mmol) were combined in a 20
mL reaction vial and dissolved in CH2Cl2 (10 mL). Upon dissolution,
the solution immediately became red-brown in color then became blue
after 10 min. After an additional 3 h of stirring, the solution was
concentrated to 3 mL, filtered through a glass microfilter, and layered
with petroleum ether (15 mL). Cooling the mixture to -35 °C for 4
days caused a deep blue solid to precipitate. Two successive recrys-
tallizations of 6 from petroleum ether/CH2Cl2 at -35 °C afforded an
analytically pure, blue microcrystalline solid (65 mg, 47%) suitable
for XRD studies. Modest yields result from multiple recrystallizations
to ensure purity and are not reflective of a low conversion to product.
Anal. Calcd for C72H120Cu2F12N2P4Sb2: C, 49.81; H, 6.97; N, 1.61.
Found: C, 49.41; H, 6.96; N, 1.76. Magnetic susceptibility (CD2Cl2,
298 K): 1.69 µB. UV-vis (CD2Cl2, nm(M-1 cm-1)): 294 (16 500),
321 (sh), 346 (21 700), 426 (sh), 510 (sh), 574 (sh), 620 (12 900), 1297
(300).
Acknowledgment. Financial support provided by the DOE
(PECASE; J.C.P.); the ONR (N00014-06-1016; R.K.S.); and a
NSF Graduate Research Fellowship (N.P.M.). Larry Henling
and Dr. Mike Day provided crystallographic assistance. Portions
of this research were carried out at the Stanford Synchrotron
Radiation Laboratory, a national user facility operated by
Stanford University on behalf of the U.S. Department of Energy,
Office of Basic Energy Sciences. The SSRL Structural Molec-
ular Biology Program is supported by the Department of Energy,
Office of Biological and Environmental Research, and by the
National Institutes of Health, National Center for Research
Resources, Biomedical Technology Program.
Synthesis of [{(tBu2-PNP)Cu}2][B(C6H3(CF3)2)4] (5). Complex 4
(150 mg, 0.144 mmol) and [Cp2Fe][B(C6H3(CF3)2)4] (144 mg, 0.137
mmol) were combined in a 20 mL reaction vial and dissolved in CH2-
Cl2 (10 mL). The reaction immediately turned red-brown in color, and
stirring was continued for 30 min at which time the solvent was
removed in vacuo. The resultant solids were washed with petroleum
ether (2 × 10 mL), extracted into diethyl ether (7 mL), filtered through
a glass microfilter, and layered with petroleum ether. Analytically pure
red crystals (203 mg, 78%) suitable for XRD were obtained after two
successive recrystallizations from diethyl ether/petroleum ether at -35
°C. Anal. Calcd for C88H100BCu2F24N2P4: C, 55.53; H, 5.30; N, 1.47.
Found: C, 55.73; H, 5.21; N, 1.39. UV-vis (CD2Cl2, nm(M-1 cm-1)):
260 (sh), 309 (14 900), 359 (sh), 416 (2930), 529 (1090), 583 (sh),
661 (sh), 818 (1520), 981 (2053), 1684 (4500).
Supporting Information Available: Synthesis and crystal-
lography; DFT calculations; supplemental Cu K-edge spectra
of two- and four-coordinate CuI complexes; complete ref 20.
This material is available free of charge via the Internet at
http://pubs.acs.org.
Synthesis of [{(tBu2-PNP)Cu}2][SbF6] (5). Complex 4 (165 mg,
0.131 mmol) and AgSbF6 (44.8 mg, 0.131 mmol) were combined in a
JA076537V
9
J. AM. CHEM. SOC. VOL. 130, NO. 11, 2008 3485