Macrocyclic [Cu(I/II)(bite)](+/2+) (bite = biphenyldiimino dithioether): An example of fully-gated electron transfer and its biological relevance

Article abstract of DOI:10.1021/ja9644034

Template condensation of 2,2'-diaminobiphenyl, 1,4-bis(2-formylphenyl)-1,4-dithiabutane, and copper(II) tetrafluoroborate yields the new macrocyclic compound [Cu(I)(bite)l(BF₄) (bite = biphenyldiimino dithioether). [Cu(I)(bite)]BF₄ c

Full text of DOI:10.1021/ja9644034

J. Am. Chem. Soc. 1997, 119, 8857-8868
8857
Macrocyclic [CuI/II(bite)]+/2+ (bite ) biphenyldiimino
dithioether): An Example of Fully-Gated Electron Transfer and
Its Biological Relevance1
Scott Flanagan,† Jun Dong,‡ Kenneth Haller,§ Shengke Wang,‡
W. Robert Scheidt,*,§ Robert A. Scott,*,‡ Thomas R. Webb,*,|
David M. Stanbury,*,| and Lon J. Wilson*,†
Contribution from the Department of Chemistry and Laboratory for Biochemical and Genetic
Engineering MS-60, William Marsh Rice UniVersity, 6100 Main Street, Houston, Texas
77005-1892, the Department of Chemistry and Biochemistry, UniVersity of Notre Dame,
Notre Dame, Indiana 46556, the Center for Metalloenzyme Studies, UniVersity of Georgia,
Athens, Georgia 30602-2556, and the Department of Chemistry, Auburn UniVersity,
Auburn, Alabama 36849
ReceiVed December 23, 1996X
Abstract: Template condensation of 2,2′-diaminobiphenyl, 1,4-bis(2-formylphenyl)-1,4-dithiabutane, and copper-
(II) tetrafluoroborate yields the new macrocyclic compound [CuI(bite)](BF4) (bite ) biphenyldiimino dithioether).
[CuI(bite)]BF4 crystallizes in the orthorhombic space group P212121 with a ) 14.379(3) Å, b ) 21.370(3) Å, c )
8.046(2) Å, V ) 2534.7(7) Å3, Z ) 4, R1 ) 0.045, and R2 ) 0.048. The X-ray structure of [CuI(bite)](BF4) reveals
distorted tetrahedral N2S2 coordination about copper, with one unusually short Cu-S(thioether) bond of 2.194(2) Å.
Oxidation of [CuI(bite)](BF4) with nitrosyl tetrafluoroborate gives [CuII(bite)](BF4)2. [CuII(bite)](BF4)2 crystallizes
in the tetragonal space group I41/a with a ) 11.640(2) Å, c ) 39.527(3) Å, V ) 5355.6(7) Å3, Z ) 8, R1 ) 0.061,
and R2 ) 0.063. X-ray crystallography of [CuII(bite)](BF4)2 reveals an approximately square planar CuN2S2 structure
-
with two distant axial BF4 anions (Cu-F 2.546(4) Å) completing a “pseudo-octahedral” coordination sphere.
Comparative EXAFS studies of solid samples and acetonitrile solutions of [CuI(bite)](BF4) and [CuII(bite)](BF4)2
demonstrate that the primary coordination environments of both species are the same in solution as in the solid.
Copper(I/II) electron self-exchange kinetics measured by 1H NMR line broadening of [CuI(bite)]+ in the presence of
[CuII(bite)]2+ reveal an overall first-order process with a rate constant of 21.7(1.9) s-1 at 295 K in acetone-d6. This
result represents the first example of fully-gated electron transfer by small-molecule copper(I). The gating process
likely involves inversion at sulfur and the tetrahedral f square planar structural change coincident with electron
transfer. Variable-temperature 1H NMR coalescence temperatures for methylene ligand protons of [CuI(bite)](BF4)
(287 K) demonstrate possible correlation of fast electron transfer with high ligand mobility for this and related
small-molecule copper(I/II) couples. Comparison with other small-molecule copper systems also reveals that fast
electron transfer is not always observed with coordination-number-invariance and conserved geometry during redox
turnover, contrary to popular interpretations of the entatic state hypothesis for blue-copper protein active sites.
Introduction
protein electron-transfer dynamics should ideally demonstrate
coordination-number-invariance (CNI) and an outer-sphere
mechanism of electron transfer. Synthetic copper systems rarely
meet these two criteria, and the literature documents only a few
well-defined candidates.3-14 The high kinetic lability of copper
and its tendency to adopt different coordination numbers and
Unlike heme iron and iron-sulfur electron-transfer proteins,
cuproproteins have no extrudable coordination complex, since
the active-site structure exists only through chelation of the
copper ion with protein residues.2 Studies carried out on the
proteins themselves are severely limited due to factors such as
the restricted range of temperatures and solvents that the native
protein can endure. Furthermore, when results can be obtained
from large biomolecules, interpretation of the data is hampered
by difficulties in distinguishing properties of the active site from
properties of the protein as a whole. Thus, the study of small-
molecule copper complexes offers one of the few means to
evaluate the active-site contribution to electron transfer for
copper proteins. Small-molecule model compounds for copper
(3) Coggin, D. K.; Gonza´lez, J. A.; Kook, A. M.; Stanbury, D. M.;
Wilson, L. J. Inorg. Chem. 1991, 30, 1115-1125.
(4) Coggin, D. K.; Gonza´lez, J. A.; Kook, A. M.; Bergman, C.; Brennan,
T. D.; Scheidt, W. R.; Stanbury, D. M.; Wilson, L. J. Inorg. Chem. 1991,
30, 1125-1134.
(5) Knapp, S.; Keenan, T. P.; Zhang, X.; Fikar, R.; Potenza, J. A.;
Schugar, H. J. J. Am. Chem. Soc. 1987, 109, 1882.
(6) Knapp, S.; Keenan, T. P.; Liu, J.; Potenza, J. A.; Schugar, H. J. Inorg.
Chem. 1990, 29, 2191.
(7) Knapp, S.; Keenan, T. P.; Zhang, X.; Fikar, R.; Potenza, J. A.;
Schugar, H. J. J. Am. Chem. Soc. 1990, 112, 3452-3464.
(8) Groenveld, C. M.; van Rijn, J.; Reedijk, J.; Canters, G. W. J. Am.
Chem. Soc. 1988, 110, 4893-4900.
† Rice University.
‡ University of Georgia.
§ University of Notre Dame.
(9) Drew, M. G. B.; Cairns, C.; McFall, S. G.; Nelson, S. M. J. Chem.
Soc., Dalton Trans. 1980, 2020-2027.
| Auburn University.
X Abstract published in AdVance ACS Abstracts, August 15, 1997.
(1) Taken in part from the following: Flanagan, J. S. Ph.D. Dissertation,
Rice University, 1995.
(10) Drew, M. G. B.; Cairns, C.; Nelson, S. M.; Nelson, J. J. Chem.
Soc., Dalton Trans. 1981, 942-948.
(11) Goodwin, J. A.; Bodager, G. A.; Wilson, L. J.; Stanbury, D. M.;
Scheidt, W. R. Inorg. Chem. 1989, 28, 35.
(2) Solomon, E. I. In Copper Coordination Chemistry: Biological &
Inorganic PerspectiVes; Karlin, K. D., Zubieta, J., Eds.; Adenine Press:
Guilderland, New York, 1983; pp 1-22.
(12) Goodwin, J. A.; Stanbury, D. M.; Wilson, L. J.; Eigenbrot, C. W.;
Schiedt, W. R. Inorg. Chem. 1987, 109, 2979-2991.
S0002-7863(96)04403-4 CCC: $14.00 © 1997 American Chemical Society

8858 J. Am. Chem. Soc., Vol. 119, No. 38, 1997
Flanagan et al.
concentrated aqueous NH3. The now basic solution was brought to
neutrality by the addition of 20 mL of HCl. The mixture appeared as
a milk-white suspension. Dichloromethane (500 mL) was added, and
the resulting mixture was stirred for ∼30 min. The CH2Cl2 layer was
saved. The inorganic phase was washed twice more with 200 mL
portions of CH2Cl2. The combined organic fractions were washed four
times with 150 mL portions of deionized water, dried with Na2SO4
(anhydrous), and evaporated to dryness to yield an orange oil. The oil
was left under vacuum overnight, yielding 7.94 g of crude product. It
was then triturated with 1 mL of ethanol to yield a mass of yellow
solid. The collected solid was dissolved in ∼20 mL of hot ethanol.
After the solution was cooled in a freezer for ∼15 min, a batch of
yellow-orange crystals was collected and determined to be unreacted
2,2′-dinitrobiphenyl. The ethanol solution was then left to cool
overnight in the freezer, yielding pure, pale yellow crystals of 2,2′-
diaminobiphenyl. The final product was dried overnight under vacuum.
Yield: 5.47 g (72%), mp 77-79 °C (lit.15 80 °C). 1H NMR (90 MHz,
TMS internal reference at 0 ppm, CDCl3): δ 6.70-7.67 (8H, m,
biphenyl), 3.70 (4H, m, -NH2, br).
1,4-Bis(2-carboxyphenyl)-1,4-dithiabutane. 1,4-Bis(2-carboxy-
phenyl)-1,4-dithiabutane was prepared by the method of Livingstone,
with minor modification.16 A mixture of 8 g of KOH, 10 g of
thiosalicylic acid, 6.1 g of 1,2-dibromoethane, and 250 mL of ethanol
was brought to reflux under N2 for 1.5 h. The resultant cloudy brown
solution was poured into ∼200 mL of deionized water. The mixture
immediately turned clear brown. The addition of ∼50 mL of
concentrated aqueous HCl resulted in the precipitation of a fine white
powder. The mixture was allowed to stand overnight, whereupon the
solid was collected by Bu¨chner filtration and rinsed well with hot
deionized water, methanol, acetone, and finally diethyl ether. The filter
cake was dried in a vacuum desiccator (10-3 Torr) for 30 min. The
still-moist product was suspended in 250 mL of 50/50 methanol/CH2-
Cl2 and refiltered. The filter cake was placed back in the vacuum
desiccator and left overnight. The low solubility of the product in any
solvent except pyridine precluded characterization by conventional 1H
NMR. The success of all subsequent synthetic steps and eventual
characterization of the final products by X-ray crystallography serve
as proof of the identity of the compound. Yield: 18.3 g (84%), mp
291-295 °C (lit.16 297 °C).
1,4-Bis(2-(hydroxymethyl)phenyl)-1,4-dithiabutane. The proce-
dure of Lindoy and Smith17 for the synthesis of 1,4-bis(2-(hydroxy-
methyl)phenyl)-1,4-dithiabutane was modified to give improved yield.
1,4-Bis(2-carboxyphenyl)-1,4-dithiabutane (45.53 g) was added to 500
mL of dry tetrahydrofuran (THF, freshly distilled from lithium
aluminum hydride (LAH)). LAH (11.37 g) was added carefully in
small portions. The addition of LAH released enough heat to bring
the THF to boiling. At this point, the round-bottomed flask containing
the mixture was fitted with a reflux condenser, and the round-bottomed
flask was suspended in a heated ultrasonic bath (65 °C) overnight. The
next day, the gray suspension was cautiously poured into 500 mL of
methanol to produce vigorous bubbling. Deionized water was then
carefully added over a period of 45 min producing a white gel. A
portion of diatomaceous earth equal in size to approximately one-third
of the volume of the white gel was added and mixed thoroughly. The
gel was then filtered through a 1 in. bed of diatomaceous earth. The
solid product was treated by Soxhlet extraction overnight with methanol.
The filtrate and extract were combined and reduced in volume by
boiling. The volume was doubled with deionized water to aid
precipitation. After the solution was allowed to cool, the white
precipitate was collected by vacuum filtration and dried overnight in a
vacuum desiccator. Yield: 39.5 g (95%), mp 111 °C (lit.17 111 °C).
1H NMR (300 MHz, CDCl3 (residual CHCl3 assigned as reference at
7.25 ppm)): δ 2.30 (2H, s, -OH, br), 3.09 (4H, s, -SCH2CH2S-), 4.76
ppm (4H, s, benzyl), 7.18-7.42 ppm (8H, m, aromatic).
Figure 1. Ligands for the CNI copper compounds.
geometries in the +1 and +2 oxidation states pose formidable
obstacles in the design and synthesis of appropriate small-
molecule systems. Despite such difficulties, the present work
describes kinetics results for a new four-coordinate copper
couple, [CuI/II(bite)]+/2+, by employing a ligand carefully tailored
to control such problems (Figure 1). The present data supple-
ment those collected for several previously described open-chain
five-coordinate complexes,3,4,11-13 while being the first to be
obtained for a closed, macrocyclic ligand designed to enforce
an outer-sphere electron self-exchange reaction mechanism.
Experimental Section
Materials. All materials were used as received unless otherwise
noted. Acetone, dichloromethane, chloroform, diethyl ether, concen-
trated aqueous hydrochloric acid, potassium hydroxide, methanol,
sodium sulfate (anhydrous), concentrated aqueous ammonia, and
tetrahydrofuran were obtained from EM Science. Diatomaceous earth
(Celite), 2,2′-dinitrobiphenyl, 1,2-dibromoethane, lithium aluminum
hydride, manganese dioxide (activated), sodium tetrafluoroborate,
nitroethane, silica gel, tin(II) chloride (anhydrous), thiosalicylic acid,
and tetrabutylammonium tetrafluoroborate were obtained from Aldrich.
Argon and nitrogen were obtained from Trigas Industrial Gases.
Chloroform-d6, acetonitrile-d3, and acetone-d6 were obtained from
Cambridge Isotope Laboratories. Copper(II) tetrafluoroborate tetrahy-
drate was obtained from Alfa. Copper(II) sulfate, n-butanol, and
potassium chloride were obtained from J. T. Baker. Ethanol was
obtained from Quantum Chemical Corporation. Nitrosyl tetrafluo-
roborate was obtained from Strem.
2,2′-Diaminobiphenyl. 2,2′-Dinitrobiphenyl (10.05 g) was dissolved
in ∼500 mL of concentrated aqueous HCl. Anhydrous SnCl2 (64.83
g) was mixed with 5 mL of deionized water and added to the HCl
solution. Stirring the mixture for 30 min resulted in a cloudy pale
yellow solution. The temperature was then brought to between 65 and
70 °C and held there for 30 min. During heating, the solution became
darker yellow. The temperature was reduced to 5 °C by means of an
ice bath, and the mixture was slowly neutralized with 300 mL of cold
1,4-Bis(2-formylphenyl)-1,4-dithiabutane. A procedure similar to
that of Lindoy and Smith17 was followed to synthesize 1,4-bis(2-
formylphenyl)-1,4-dithiabutane. Dry 1,4-bis(2-(hydroxymethyl)phen-
yl)-1,4-dithiabutane (10 g) and 70 g of active MnO2 were suspended
in 1000 mL of chloroform. The stirred mixture was refluxed for 8.5
(13) Goodwin, J. A.; Wilson, L. J.; Stanbury, D. M.; Scott, R. A. Inorg.
Chem. 1989, 28, 42-50.
(14) Mu¨ller, E.; Bernardinelli, G.; Reedijk, J. Inorg. Chem. 1996, 35,
1952-1957.
(15) Farnum, D. G.; Rasheed, K. J. Org. Chem. 1977, 42, 573-574.
(16) Livingstone, S. E. J. Chem. Soc. 1956, 437-440.
(17) Lindoy, L. F.; Smith, R. J. Inorg. Chem. 1981, 20, 1314-1316.

Fully-Gated Electron Transfer
J. Am. Chem. Soc., Vol. 119, No. 38, 1997 8859
h. The MnO2 was removed by filtration through Celite and rinsed twice
with 200 mL portions of chloroform. The brilliant pink chloroform
solution so obtained was evaporated to dryness, yielding a pink solid
that was vacuum dried overnight. Recrystallization from hot ethanol
yielded a white solid. Yield: 5.5 g (56%), mp 135-137 °C (lit.17 137
°C). 1H NMR (250 MHz, CDCl3 (residual CHCl3 assigned as reference
at 7.25 ppm)): δ 3.19 (4H, s, -SCH2CH2S-), 7.34-7.88 (8H, m,
aromatic), 10.37 (2H, s, -CHO).
Table 1. Summary of Crystal Data and Intensity Collection
Parameters
[CuI(bite)](BF4)
[CuII(bite)](BF4)2
T, K
formula
formula wt, amu 600.9
294
C28H22BCuF4N2S2
123
C28H22B2CuF8N2S2
687.8
cryst dimens, mm 0.08 × 0.15 × 0.50
0.02 × 0.30 × 0.42
space group
cryst syst
a, Å
P212121
orthorhombic
14.739(3)
21.370(3)
8.046(2)
2534.4(7)
4
I41/a
[CuI(bite)](BF4). 1,4-Bis(2-formylphenyl)-1,4-dithiabutane (0.49 g)
and 0.55 g of CuII(BF4)‚4H2O were stirred with 70 mL of n-butanol
and brought to a boil in an Erlenmeyer flask (note the requirement for
a high-boiling solvent.) This resulted in the formation of a greenish-
blue suspension. 2,2′-Diaminobiphenyl (0.30 g) was dissolved in 20
mL of n-butanol with the aid of slight heating and added to the reaction
producing an immediate dark-green color that quickly turned reddish
brown. The volume of the solution was reduced to ∼10 mL by boiling.
The flask was stoppered and placed in the freezer overnight. The deep
orange-red product was collected by filtration, rinsed well with hot
toluene, and dried for several hours in a vacuum desiccator. The
product was then dissolved in CH2Cl2 and filtered. The CH2Cl2 was
removed by evaporation, and the solid was dried in a vacuum desiccator
for several minutes. At this point, the orange-red film was dissolved
in methanol and left in a covered beaker for several days until orange-
red crystals formed. The crystals were collected by filtration. Alterna-
tiVe Purification: Purification by column chromatography was achieved
by loading 1.0 g of compound dissolved in 10% more than the minimum
necessary volume of dichloromethane onto a 3 in. high × 2 in. diameter
column of silica gel eluted with dichloromethane. A yellow band eluted
while the product remained at the top of the column. Increasing the
polarity of the eluant by slowly introducing methanol caused the elution
of two additional yellow bands, followed by the product, which had a
leading green edge. This green edge was discarded. Evaporating the
product band to dryness, followed by the crystallization procedure
described above, resulted in the orange product. Yield: 0.30 g (28%),
mp 283-285 °C (dec.). 1H NMR (300 MHz, CDCl3 (residual CHCl3
assigned as reference at 7.25 ppm)): δ 3.43-3.57 (4H, m, -SCH2-
CH2S-), 7.13-7.84 (16H, m, aromatic), 8.38 (2H, s, -CHN-). Anal.
Calcd. for C28H22N2S2CuBF4: C, 55.96; H, 3.69; N, 4.66; S, 10.67.
Found: C, 56.23; H, 3.70; N, 4.45; S, 10.97.
tetragonal
11.640(2)
b, Å
c, Å
39.527(3)
5355.6(7)
8
V, Å3
Z
density, calcd
1.575
1.70
g/cm3
radiation
diffractometer
criterion for obsd Fo > 4.0σ(Fo)
obsd reflections 2004
graphite monochromated Mo KR (λ ) 0.710 73 Å)
Siemens R3m/V
Enraf-Nonius CAD4
Fo > 3σ(Fo)
1617
ind reflections
2566
1.071
R1 ) 0.045
R2 ) 0.048
1.52
2775
1.045
R1 ) 0.061
R2 ) 0.063
1.04
µ, mm-1
residuals
goodness of fit
data/parameter
6.1
7.1
temperature device operating at 123 ( 2 K. Cell constants and an
orientation matrix for data collection were obtained from a least-squares
refinement, based on the setting angles of 25 reflections in the range
of 11.8° < 2θ < 35.8°, determined by the computer-controlled diagonal
slit method of centering using the SET4 routine. The tetragonal cell
parameters and calculated volume are as follows: a ) 11.640(2) Å, c
) 39.527(3) Å, V ) 5355.6(7) Å3. For Z ) 8 and fw ) 687.78 Da,
the calculated density is 1.71 g/cm3. As a check on crystal quality, ω
scans of several intense reflections were measured; the width at half-
height was 0.43° with a take-off angle of 2.8°. The systematic absences
of hkl, h + k + l ) 2n + 1, hk0, h ) 2n + 1, k ) 2n + 1, and 00l,
l ) 2n + 1 uniquely determine the space group to be the centrosym-
metric I41/a (No. 88).
The data were collected at a temperature of 123 ( 2 K using the
θ/2θ scan technique. The scan rate was fixed at 4.12°/min (in ω).
The scan range (in degrees) was determined as a function of θ to correct
for the separation of the KR doublet; the peak width was estimated to
be 0.95 + 0.34(tan θ°), and an additional 25% above and below this
range was collected to ensure adequate background ranges. Data were
collected to a maximum 2θ of 52.86° (maximum sin θ/λ of 0.626).
The horizontal counter aperture was also adjusted as a function of θ,
varying from 2.5 to 2.7 mm; the vertical counter aperture was set at
4.0 mm. The diameter of the incident beam collimator was 1.3 mm,
and the crystal to detector distance was 21 cm. For intense reflections,
an attenuator was automatically inserted in front of the detector; the
attenuator factor was 19.3. A summary of data collection parameters
is given in Table 1.
[CuII(bite)](BF4)2. [CuI(bite)](BF4) (0.18 g) was dissolved in 50
mL of dichloromethane. [NO](BF4) (0.05 g) was added after which
the solution was stirred and heated. The color immediately began to
darken. After 10 min, another 0.04 g of [NO](BF4) was added to the
now brown solution. The solution was allowed to cool for 45 min
leading to the formation of a dark-green solid. The product was
collected by filtration and air dried. The product was then dissolved
in 40 mL of nitroethane and crystallized by vapor diffusion with diethyl
ether. The microcrystalline product was collected by filtration and dried
in a vacuum desiccator overnight. Yield: 0.04 g (20%). Anal. Calcd.
for C28H22N2S2CuB2F8: C, 48.90; H, 3.22; N, 4.07; S, 9.32. Found:
C, 48.25; H, 3.50; N, 4.44; S, 9.86.
Crystal Growth. Single crystals of [CuI(bite)](BF4) and [CuII(bite)]-
(BF4)2 were grown as described above.
X-ray Diffraction Data Collections. [CuI(bite)](BF4). A crystal
was mounted on the end of a glass fiber. The unit cell was determined
by centering 24 reflections in the 2θ range from 15 to 29°. Data were
collected using the 2θ - ω scan technique and processed on a computer-
controlled four-circle Nicolet R3m/V crystal diffractometer with Mo
KR radiation. Two reflections established as standards were checked
routinely for deviations of positions and intensities after every 100
reflections. Lorentz and polarization corrections were performed, as
was an absorption correction (based on Ψ scans). Systematic absences
uniquely identified the space group as P212121 (No. 19). A summary
of data collection parameters is given in Table 1.
X-ray Crystal Structure Determinations. [CuI(bite)](BF4). The
structure was solved by direct methods. All computations were
performed with SHELXTL PLUS VMX, Release 4.11, 1990 (Siemens
Analytical X-ray Instruments, Inc.) based on SHELX (Sheldrick, G.
M.). Other atoms were recovered from difference maps calculated after
least-squares refinement of the CuS2 core of the cation. Difference
-
maps indicated that the BF4 ion was disordered. We refined a
disordered model of two tetrahedrons about the B atom, with a common
refined B-F distance and constrained F-F distance (FVAR and DFIX
commands in SHELXTL-PLUS), with isotropic temperature factors on
each F held equal. The occupancy of the major tetrahedron [F(1)-
F(4)] refined to 0.71(1). All non-hydrogen atoms of the cation were
ultimately refined anisotropically; the hydrogen atoms were added in
calculated positions. The enantiomorph was determined using the
method of Rogers.18
[CuII(bite)](BF4)2. A green diamond-shaped plate single crystal of
[CuII(bite)](BF4)2, C28H22N2S2B2F8Cu, of approximate dimensions 0.02
× 0.30 × 0.42 mm was mounted on the end of a hollow glass fiber
with the plate face approximately perpendicular to the φ axis using
cyanoacrylate glue. Preliminary examination and data collection were
performed using Mo KR radiation (λh ) 0.710 73 Å) on an Enraf-Nonius
CAD4 diffractometer equipped with a graphite crystal incident beam
monochromator and a gas flow liquid nitrogen refrigerated low-
[CuII(bite)](BF4)2. A total of 5729 reflections were collected, of
which 2775 were unique. As a check on crystal and electronic stability
(18) Rogers, D. Acta Crystallogr. 1981, A37, 734.

8860 J. Am. Chem. Soc., Vol. 119, No. 38, 1997
Flanagan et al.
four representative reflections were measured every 60 min of X-ray
exposure. Plots of the intensities of these standard reflections versus
time showed no systematic changes over the time of the data collection,
and no correction for decay was applied.
[CuII(bite)](BF4)2 was obtained at -170 °C with a Varian E-Line EPR
spectrometer. Diphenylpicrylhydrazyl was employed as a field refer-
ence. Integration was standardized with respect to a solution of CuSO4
of known concentration. Copper X-ray absorption spectroscopic (XAS)
data were collected at the Stanford Synchrotron Radiation Laboratory
on beam line 7-3 with the SPEAR ring running at 3.0 GeV and
between 60 and 90 mA. Tables S-XIV and S-XV of the Supporting
Information detail the data collection and data reduction conditions and
parameters. Data reduction generally followed published methods.25
Conductivity. Solution conductivity data were collected in dichlo-
romethane using a YSI Model 31 conductivity bridge. A temperature
of 25 °C was maintained by a constant temperature bath. Dichlo-
romethane solutions 10-5 M in [CuI(bite)](BF4), [CuII(bite)](BF4)2, and
tetra-n-butylammonium tetrafluoroborate (TBABF4) were made up
volumetrically. The cell constant was determined to be θ ) 1.58 cm-1
by using a 0.01 M aqueous solution of potassium chloride to calibrate
the conductivity cell.
Electrochemistry. Acetonitrile was refluxed over CaH2 for 12 h
and distilled immediately before use. Solutions were prepared in a
solvent fume hood. A stock solution of 2.5 × 10-2 M NaBF4 was
prepared and used to create solutions of 10-3 M ferrocene and [CuI-
(bite)](BF4). Cyclic voltammograms were collected using a BAS-
100BW electrochemical analyzer. Scans were collected at a rate of
100 mV/s using a standard three-electrode configuration with a 1 mm
platinum disc working electrode, Ag/AgCl reference electrode, and
platinum wire auxiliary electrode.
Lorentz and polarization corrections were applied to the data. The
linear absorption coefficient is 10.45 cm-1 for Mo KR radiation. An
empirical absorption correction based on a series of ψ-scans was applied
to the data. Relative transmission coefficients ranged from 0.793 to
1.000 with an average value of 0.881. Intensities of equivalent
reflections were averaged. The agreement factors for the averaging of
the 2688 multiply measured reflections with their respective duplicates
were 0.084 based on intensity and 0.059 based on Fo. The relatively
high values are due to the large number of unobserved data in the data
set resulting from the small sample size.
The structure was solved with the MULTAN19 direct methods
program which revealed the positions of 11 of the 17 unique
non-hydrogen atoms of the cation. The other non-hydrogen atoms were
located from a subsequent difference electron density Fourier map. All
hydrogen atoms could be located from difference electron density
Fourier maps calculated after preliminary refinement of the non-
hydrogen atoms. The structure was refined by a full-matrix least-
squares process where the function minimized was ∑w(|Fo| - |Fc|)2,
2
2
and the weight w is defined as w ) 4Fo σ2(Fo ) ) 1/σ2Fo. The standard
2
deviation on intensities is defined as follows: σ2(Fo ) ) [S2(C + R2B)
2
+ p2Fo ]/Lp2, where S is the scan rate, C is the total integrated peak
count, R is the ratio of scan time to background counting time, B is the
total background count, Lp is the Lorentz-polarization factor, and the
parameter p (here set to 0.04) is a factor introduced to down-weight
intense reflections.
Variable-Temperature 1H NMR. 1H NMR spectra were collected
as a function of temperature on an IBM/Bru¨ker AF-300 NMR
spectrometer. Typically 16 scans were accumulated into 32 K of
memory and Fourier transformed. Data were collected from room
temperature down to approximately 5 °C above the freezing point of
the solvent. Spectra were accumulated at 10 °C intervals. The signals
due to residual solvent protons were used as internal chemical shift
references, with the CHD2CN quintet assigned a value of 1.93 ppm,
the CHD2COCD3 quintet a value of 2.04 ppm, and the CHDCl2 triplet
a value of 5.32 ppm.
Electron Self-Exchange Kinetics. Spectra were collected in a
manner similar to that used for the variable-temperature 1H NMR
experiments described above. Line widths were obtained from the
transformed spectra by using the Lorentzian curve-fitting routine
available on the IBM/Bru¨ker AF-300 spectrometer. Data were collected
as a function of temperature and concentration of the oxidized form of
the copper complex. Samples were prepared to maintain a constant
ionic strength of µ ) 25 mM in acetone-d6 with TBABF4 as the
supporting electrolyte. Ten milliliters of an acetone-d6 solution was
prepared from 60.1 mg of [CuI(bite)](BF4). This stock solution was
then used as solvent to prepare solutions A and B. Solution A was
prepared by dissolving 14.8 mg of TBABF4 to a volume of 3 mL.
Solution B was prepared by dissolving 3.4 mg of [CuII(bite)](BF4)2 to
a volume of 1 mL. NMR samples were prepared by mixing appropriate
aliquots of A and B to achieve the desired concentration of [CuII(bite)]-
(BF4)2. Solution B was used immediately after the [CuII(bite)](BF4)2
had completely dissolved, and the NMR samples were frozen in liquid
nitrogen until just prior to insertion into the spectrometer. This
procedure prevented significant decomposition of the somewhat unstable
oxidized form of the complex.
Scattering factors were taken from Cromer and Waber.20 Anomalous
dispersion effects were included in Fc;21 the values for ∆f ′ and ∆f ′′
were those of Cromer.22 Only the 1617 reflections having intensities
greater than 3.0 times their standard deviation (on Fo) were used in the
refinements. The model for the final cycles of refinement assigned
anisotropic displacement parameters to the non-hydrogen atoms and
included the hydrogen atoms in fixed idealized positions (d[C-H] )
0.95 Å, B[H] ) 1.1B[Cattached]), thus including 195 variable parameters
and resulting in a data/variable ratio of 7.1/1. Refinement converged
with unweighted and weighted agreement factors of R1 ) ∑|Fo - Fc|/
2
∑|Fo| ) 0.061 and R2 ) (∑w(Fo - Fc)2/∑wFo )1/2 ) 0.063, and an
estimated standard deviation of an observation of unit weight of 1.04.
The highest correlation coefficient was 0.38. The highest peak in the
final difference electron density Fourier map was 1.06 e-/Å3 with an
estimated error based on ∆F23 of 0.14 located near the copper atom.
The highest peak not associated with the heavy atoms was 0.56 e-/Å3.
Plots of ∑w(|Fo| - |Fc|)2 versus |Fo|, reflection order in data collection,
sin θ/λ and various classes of indices showed no unusual trends. All
calculations were performed on a VAXstation 3200 using the SDP/
VAX program system24 unless noted otherwise.
Spectroscopy. Electronic spectra were collected on a Hewlett
Packard 8452A Diode Array Spectrophotometer in 1 cm matched quartz
cuvettes. Solutions 1.0 × 10-5 M in [CuI(bite)](BF4) and 1.0 × 10-4
M in [CuII(bite)](BF4)2 were made up volumetrically. Scans were
collected in the range of 190-820 nm, with the shorter wavelength
limited by the solvent. Far IR spectra were collected for [CuI(bite)]-
(BF4) and [CuII(bite)](BF4)2 and compared to those for KBF4, NaBF4,
and [CuII(en)2](BF4)2. Data were collected on a Perkin-Elmer 1430
ratio recording infrared spectrophotometer. Compounds were run as
their Nujol mulls on CsI plates. The spectrometer was purged with
argon gas. The X-band EPR spectrum of a frozen acetonitrile glass of
Numerical Methods. Calculations were performed with Student
Matlab for Macintosh (The Math Works, Inc.) on a Macintosh Performa
400 desktop computer. Inverse relaxation times, 1/T2, were determined
from the line widths (∆ν1/2) simply by multiplying by the constant π:
1/T2 ) π∆ν1/2. First-order rate constants, k, were obtained as the inverse
paramagnetic contribution to relaxation time, 1/T2p, by subtracting the
inverse natural relaxation time at a given temperature from the inverse
average relaxation time observed, 1/T2, in the presence of copper(II):
k ) 1/T2p ) 1/T2 - 1/T2n.
(19) Main, P.; Fiske, S. J.; Hull, S. E.; Lessinger, L.; Germain, G.;
DeClercq, J.-P.; Woolfson, M. M. MULTAN 11/82; University of York:
York, England, July 1982.
(20) Cromer, D. T.; Waber, J. T. In International Tables for X-Ray
Crystallography; The Kynoch Press: Birmingham, England, 1974; Vol.
IV, Table 2.2B.
(21) Ibers, J. A.; Hamilton, W. C. Acta. Crystallogr. 1964, 17, 781.
(22) Cromer, D. T. In International Tables for X-Ray Crystallography;
The Kynoch Press: Birmingham, England, 1974; Vol. IV, Table 2.3.1.
(23) Cruickshank, D. W. J. Acta. Crystallogr. 1949, 2, 154.
(24) Frenz, B. A. In Computing in Crystallography; Schenk, H., Olthof-
Hazelkamp, R., vanKonigsveld, H., Bassi, G. C., Eds.; Delft University
Press: Delft, Holland, 1978; pp 64-71.
The activation parameters of the self-exchange reaction were
determined from Eyring plots. Since
q
q
k ) (κT/h)e-∆H /RTe∆S /R
(1)
the enthalpy, ∆Hq, and the entropy, ∆Sq, of activation may be obtained
(25) Scott, R. A. Methods Enzymol. 1985, 117, 414-459.

Fully-Gated Electron Transfer
J. Am. Chem. Soc., Vol. 119, No. 38, 1997 8861
from a plot of ln(k/T) versus 1/T, where T is temperature, R is the gas
constant, k is the Boltzmann constant, and h is Planck’s constant. Least-
squares calculations were weighted with the uncertainty of ln(k/T) for
each point.26
Results and Discussion
Both of the blue copper proteins plastocyanin (Pc) and Azurin
(Az) have been structurally characterized in their Cu(I) and Cu-
(II) oxidation states.27-29 In both instances, the major structural
difference between the +1 and +2 oxidation states is a slight
lengthening of the copper-ligand bonds in the reduced forms.
The geometry of the coordination sites in these proteins has
been described as a compromise between the preferred geom-
etries of Cu(I) and Cu(II). Vallee and Williams were the first
to coin the term entatic30 in describing the application of Lumry
and Eyring’s rack mechanism31 to metalloenzymes. This
sterically-imposed entatic state is now frequently cited as the
source of the fast self-exchange rates of these proteins, due to
the supposed lowering of the enthalpic component of the
activation free energy and reduced Franck-Condon barrier.28,32-34
A fact worth noting, however, is that for at least one measure-
ment of the activation parameters for azurin a favorable entropic
component of the activation free energy is implicated as being
far more influential than the enthalpic component.35-37 Fur-
thermore, the literature documents one rather startling example
of a four-coordinate small-molecule system ([CuI/II((imidH)2-
bp)2]+/2+), which is coordination-number-invariant and thus
rather closely mimics the entatic state, yet has one of the slowest
self-exchange rates (by NMR line broadening) of any copper
redox couple.5,7
Figure 2. ORTEP drawing of the [CuI(bite)]+ cation with the atom
labeling scheme. The opposite enantiomorph is shown in order to
facilitate comparison with the structure of the Cu(II) analogue.
Scheme 1
In a view that challenges the traditional interpretation of the
entatic state hypothesis, a recent examination of the electronic
structure of the blue copper sites attributes the rapid electron
transfer and high redox potential to the favorable configuration
of ligand orbitals. Specifically, Guckert et al.38 postulate that
the orbitals are arranged such that changes in the ligand-metal
orbital mixing are compensated by changes in the orbital
energies in a balancing act which results in fairly advantageous
electronic configurations in both oxidation states. This view is
somewhat at odds with the traditional entatic state hypothesis
which postulates that the rigid protein environment enforces an
unfavorable geometry on the copper(II) oxidation state. In fact,
the authors postulate that an entatic state is enforced upon the
copper(I) oxidation state.
+1 and +2 oxidation states. We report here for the first time
an N2S2* (S* designates thioether as opposed to S which
designates thiolate) macrocyclic system which is four-coordinate
and nearly tetrahedral for the Cu(I) state but which “semi-
coordinates” two BF4- counterions (in the solid state only; see
below) to form an axially elongated octahedral complex for the
Cu(II) state, with a surprisingly “flattened” N2S2* donor atom
set. The synthesis of these [CuI/II(bite)](BF4)1,2 salts is outlined
in Scheme 1. The compounds in Figure 1 are the first copper
compounds for which electron-transfer kinetics data have been
obtained on systems designed to guarantee both outer-sphere
mechanisms and coordination-number-invariance. Although the
previously studied five-coordinate open-chain complexes likely
operate through outer-sphere mechanisms,4 the ability of the
ligand arms to dissociate leaves some possibility of inner-sphere
electron transfer. Such dissociative mechanisms are not avail-
able to the present [CuI/II(bite)]+/2+ macrocyclic compounds,
virtually eliminating any pathways for inner-sphere electron
transfer.
Figure 1 displays the ligands for the four CNI copper
complexes for which self-exchange rate constant data are
available and X-ray structural data are available for both the
(26) Bevington, P. R.; Robinson, D. K. Data Reduction and Error
Analysis for the Physical Sciences; 2nd ed.; McGraw-Hill, Inc.: New York,
1992.
(27) Guss, J. M.; Harrowell, P. R.; Murata, M.; Norris, V. A.; Freeman,
H. C. J. Mol. Biol. 1986, 192, 361.
(28) Shepard, W. E. B.; Anderson, B. F.; Lewandoski, D. A.; Norris, G.
E.; Baker, E. N. J. Am. Chem. Soc. 1990, 112, 7817-7819.
(29) Guss, J. M.; Bartunik, H. D.; Freeman, H. C. Acta Crystallogr. 1992,
B48, 790-811.
(30) Vallee, B. L.; Williams, R. J. P. Proc. Nat. Acad. Sci. U.S.A. 1968,
59, 498-505.
(31) Lumry, R.; Eyring, H. J. Phys. Chem. 1954, 58, 110-120.
(32) Malmstro¨m, B. G. Eur. J. Biochem. 1994, 223, 711-718.
(33) Gray, H. B.; Malmstro¨m, B. G. Comments Inorg. Chem. 1983, 2,
203-209.
(34) Williams, R. J. P. J. Mol. Catal. 1985, 30, 1-26.
(35) Groenveld, C. M.; Canters, G. W. Eur. J. Biochem. 1985, 153, 559-
564.
(36) Groenveld, C. M.; Canters, G. W. ReV. Port. Quim. 1985, 27, 145.
(37) Groenveld, C. M.; Dahlin, S.; Reinhammar, B.; Canters, G. W. J.
Am. Chem. Soc. 1987, 109, 3247.
(38) Guckert, J. A.; Lowery, M. D.; Solomon, E. I. J. Am. Chem. Soc.
1995, 117, 2817-2844.
X-ray Crystal Structure Determinations. [CuI(bite)](BF4).
An ORTEP structural plot for the [CuI(bite)]+ cation is presented
in Figure 2, while the bond lengths and angles are given in
Tables 2 and 3. The structure consists of separate [CuI(bite)]+

8862 J. Am. Chem. Soc., Vol. 119, No. 38, 1997
Table 2. Bond Lengths (Å) for [CuI(bite)](BF4)a
Flanagan et al.
different times in two different locations! The N1-Cu-S1/
N2-Cu-S2, N1-Cu-N2/S1-Cu-S2, and N1-Cu-S2/N2-
Cu-S1 dihedral angles are all near the ideal for a tetrahedron
(90°) at 78.0(3)°, 76.7(3)°, and 88.9(3)° (av ) 81.20(3)°). The
interior angles of the tetrahedron range from 97.8(2)° to 132.5-
(2)° with the average value of 105.4° very near the ideal of
109.5°. The angle between the biphenyl rings is 64.8(3)°. The
N1-C9-C8-C3-S1-Cu chelate ring is essentially co-planar
with the C3-C8 phenyl ring, while the N2-C22-C23-C28-
S2-Cu chelate ring is puckered significantly.
atom-atom
distance (Å)
atom-atom
distance (Å)
Cu-S(1)
2.194(2)
1.953(5)
1.820(8)
1.828(7)
1.285(8)
1.436(8)
1.501(11)
1.412(10)
1.370(12)
1.407(10)
1.378(10)
1.377(11)
1.410(11)
1.497(10)
1.413(9)
1.384(11)
1.375(9)
1.404(9)
1.354(10)
1.399(10)
1.388(10)
1.370(10)
1.386(15)
1.386(15)
Cu-S(2)
2.323(2)
1.938(5)
1.807(7)
1.793(7)
1.426(9)
1.274(8)
1.363(10)
1.417(11)
1.369(11)
1.476(10)
1.415(9)
1.384(11)
1.389(11)
1.404(10)
1.372(10)
1.394(10)
1.468(10)
1.428(10)
1.364(12)
1.368(10)
1.346(10)
1.396(9)
1.368(13)
1.371(17)
Cu-N(1)
Cu-N(2)
S(1)-C(2)
S(2)-C(1)
N(1)-C(9)
N(2)-C(21)
C(1)-C(2)
C(3)-C(8)
C(5)-C(6)
C(7)-C(8)
C(10)-C(11)
C(11)-C(12)
C(13)-C(14)
C(15)-C(16)
C(16)-C(21)
C(18)-C(19)
C(20)-C(21)
C(23)-C(24)
C(24)-C(25)
C(26)-C(27)
B-F(1)
S(1)-C(3)
S(2)-C(28)
N(1)-C(10)
N(2)-C(22)
C(3)-C(4)
C(4)-C(5)
C(6)-C(7)
C(8)-C(9)
C(10)-C(15)
C(12)-C(13)
C(14)-C(15)
C(16)-C(17)
C(17)-C(18)
C(19)-C(20)
C(22)-C(23)
C(23)-C(28)
C(25)-C(26)
C(27)-C(28)
B-F(2)
The structure of [CuI(bite)](BF4) is unusual primarily for the
extraordinarily short CuI-S(thioether) bond length of 2.194(2)
Å, which is the shortest length to date for this type of bond
(typically ∼2.3 Å). This distance is much shorter than the Cu-
S*(methionine) distance of 2.90 Å found in plastocyanin and
is actually much closer to the Cu-S-(cysteine) distance of 2.13
Å.39 The extremely tight fit of the ligand about the Cu(I) center
may be due to the relatively small cavity of the macrocycle. In
fact, in order for Schiff base condensation of this compound to
succeed, the rather high boiling point solvent n-butanol (bp 118
°C) was required, as the typical solvents used in this kind of
reaction (i.e., methanol and ethanol) did not effect closure of
the macrocycle. This structure is also the first to be determined
for a macrocyclic ligand employing an N2S2*(thioether) donor
atom set. The molecule is distorted from perfect C2 symmetry,
most notably in the difference in Cu-S bond lengths, but also
in the difference in geometry of the N-C-C-C-S-Cu chelate
rings in that one is essentially planar while the other is puckered.
The source of the distortions from perfect C2 symmetry is
unknown, but may be related to the unusually high redox
potential of the [CuI/II(bite)]+/2+ couple (see below). The
distortions may also be an effect of crystal packing. A related,
nonmacrocyclic, compound has been reported with an N2S2
(thiolato) donor atom set, which employs a similar strategy of
using a biphenyl-twisted Schiff base;40 the Cu(II) form of that
compound has been reported, but the Cu(I) has not.
B-F(3)
B-F(4)
B-F(5)
B-F(6)
B-F(7)
B-F(8)
a Estimated standard deviations are given in parentheses.
Table 3. Bond Angles (deg) for [CuI(bite)](BF4)a
atoms
angle (deg)
atoms
angle (deg)
S(1)-Cu-S(2)
97.8(1) S(1)-Cu-N(1)
111.6(2) S(1)-Cu-N(2)
97.8(2) N(1)-Cu-N(2)
98.0(3) Cu-S(1)-C(3)
104.1(3) Cu-S(2)-C(1)
99.0(2) C(1)-S(2)-C(28)
123.5(5) Cu-N(1)-C(10)
118.8(6) Cu-N(2)-C(21)
104.6(2)
132.5(2)
110.6(2)
104.7(2)
94.8(2)
104.2(3)
114.6(4)
113.2(4)
S(2)-Cu-N(1)
S(2)-Cu-N(2)
Cu-S(1)-C(2)
C(2)-S(1)-C(3)
Cu-S(2)-C(28)
Cu-N(1)-C(9)
C(9)-N(1)-C(10)
Cu-N(2)-C(22)
S(2)-C(1)-C(2)
S(1)-C(3)-C(4)
C(4)-C(3)-C(8)
C(4)-C(5)-C(6)
C(6)-C(7)-C(8)
C(3)-C(8)-C(9)
N(1)-C(9)-C(8)
126.4(5) C(21)-N(2)-C(22) 119.9(6)
115.8(5) S(1)-C(2)-C(1)
113.8(5) S(1)-C(3)-C(8)
121.0(6) C(3)-C(4)-C(5)
120.1(7) C(5)-C(6)-C(7)
122.5(7) C(3)-C(8)-C(7)
130.4(6) C(7)-C(8)-C(9)
116.1(5)
124.6(5)
120.0(7)
119.5(7)
116.9(6)
112.7(6)
The enantiomer reported here is the opposite of the one
obtained in the first structure determination of this compound.
The compound spontaneously resolves under the conditions of
crystallization. The chirality of the present enantiomer with
respect to the twist about the C2 axis is of the ∆ configuration.41
The Cahn-Ingold-Prelog (R and S) configurations42 of the
biphenyl and copper chiral centers are R, while both of the sulfur
centers are S.
128.5(6) N(1)-C(10)-C(11) 120.2(6)
N(1)-C(10)-C(15) 120.3(6) C(11)-C(10)-C(15) 119.4(7)
C(10)-C(11)-C(12) 121.9(7) C(11)-C(12)-C(13) 120.7(7)
C(12)-C(13)-C(14) 117.6(7) C(13)-C(14)-C(15) 122.6(7)
C(10)-C(15)-C(14) 117.8(7) C(10)-C(15)-C(16) 125.4(7)
C(14)-C(15)-C(16) 116.7(6) C(15)-C(16)-C(17) 116.9(6)
C(15)-C(16)-C(21) 126.2(6) C(17)-C(16)-C(21) 116.4(6)
C(16)-C(17)-C(18) 121.7(7) C(17)-C(18)-C(19) 121.0(7)
C(18)-C(19)-C(20) 118.7(7) C(19)-C(20)-C(21) 120.5(7)
N(2)-C(21)-C(16) 118.8(6) N(2)-C(21)-C(20) 119.4(6)
C(16)-C(21)-C(20) 121.7(6) N(2)-C(22)-C(23) 126.3(7)
C(22)-C(23)-C(24) 114.5(7) C(22)-C(23)-C(28) 129.0(6)
C(24)-C(23)-C(28) 116.5(6) C(23)-C(24)-C(25) 123.0(7)
C(24)-C(25)-C(26) 120.8(7) C(25)-C(26)-C(27) 118.1(7)
[CuII(bite)](BF4)2. An ORTEP structural plot for [CuII(bite)]-
(BF4)2 is presented in Figure 3, while the bond lengths and bond
angles are given Tables 4 and 5. [CuII(bite)]2+ has crystallo-
graphically imposed C2 symmetry. The dihedral angle between
the two N-Cu-S planes of only 3.0° is an extreme degree of
flattening when compared to all other coordination compounds
employing the biphenyl moiety.5-7,40,43-45 The presence of two
C(26)-C(27)-C(28) 122.7(7) S(2)-C(28)-C(23)
122.6(5)
-
long axial CuII-F interactions of 2.546(4) Å with BF4 gives
S(2)-C(28)-C(27)
F(1)-B-F(2)
F(2)-B-F(3)
F(2)-B-F(4)
F(5)-B-F(6)
F(6)-B-F(7)
F(6)-B-F(8)
118.2(6) C(23)-C(28)-C(27) 118.9(6)
109.3(7) F(1)-B-F(3)
112.8(6) F(1)-B-F(4)
109.7(7) F(3)-B-F(4)
109.7(11) F(5)-B-F(7)
111.4(11) F(5)-B-F(8)
110.2(11) F(7)-B-F(8)
109.5(7)
108.2(6)
107.2(7)
108.2(11)
109.3(11)
107.9(10)
the structure a resemblance to that of [CuII(en)2](BF4)2, which
(39) Guss, J. M.; Freeman, H. C. J. Mol. Biol. 1983, 169, 521-563.
(40) Anderson, O. P.; Becher, J.; Frydendahl, H.; Taylor, L. F.; Toftlund,
H. J. Chem. Soc., Chem. Commun. 1986, 699-701.
(41) Purcell, K. F.; Kotz, J. C. Inorganic Chemistry; W. B. Saunders
Company: Philadelphia, 1977; pp 638-644.
(42) Cross, L. C.; Klyne, W. Pure Appl. Chem. 1976, 45, 11-30.
(43) Cheeseman, T. P.; Hall, D.; Waters, T. N. J. Chem. Soc. 1966, A,
1936.
a Estimated standard deviations are given in parentheses.
cations and BF4- anions. There are no Cu-F contacts less than
4.0 Å. The coordination about the copper atom is distorted
tetrahedral; the Cu-S distances are highly unequal (2.194(2)
and 2.323(2) Å). This latter feature of the structure has been
reproduced with data sets from three different crystals at widely
(44) Pignolet, L. H.; Taylor, R. P.; Horrocks, W. D., Jr. J. Am. Chem.
Soc. 1969, 90, 5457.
(45) Frydendahl, H.; Toftlund, H.; Becher, J.; Dutton, J. C.; Murray, K.
S.; Taylor, L. F.; Anderson, O. P.; Tiekink, R. T. Inorg. Chem. 1995, 34,
4467-4476.

Fully-Gated Electron Transfer
J. Am. Chem. Soc., Vol. 119, No. 38, 1997 8863
Figure 3. ORTEP drawing of [CuII(bite)](BF4)2 with the atom labeling
scheme.
Figure 4. EXAFS FT spectra of solid (s) (30 K) and 1.0 × 10-3
M
acetonitrile solution (- - -) (10 K): (a) [CuI(bite)](BF4); (b) [CuII(bite)]-
(BF4)2.
a
Table 4. Bond Lengths (Å) for [CuII(bite)](BF4)2
distance (Å)
distance (Å)
that the addition of methyl groups to the 6,6′ positions of the
ligand (i.e., [CuII(biteMe2)](BF4)2]) would force the copper(II)
center into a more tetrahedral configuration, perhaps eliminating
ambiguities about the “pseudo-coordinated” anions. However,
even for the present [CuII(bite)]+/2+ compound, solution con-
ductivity and EXAFS measurements together imply that the
BF4- anions are not strongly coordinated in solution (see below)
where the electron self-exchange studies have been performed.
The crystal is racemic, with both members of the enantiomeric
pair present in the unit cell. The configurations of the biphenyl,
copper, and sulfur centers for the enantiomer shown in the
ORTEP diagram of Figure 3 are all S. Comparison of the CuI
and CuII cations which have identical configurations at the
biphenyl moiety illustrates that both sulfur atoms have inverted
configuration in the CuII cation from that found in the CuI cation.
Examination of flexible molecular stick models with (R)-
biphenyl configuration suggests that square planar CuII can be
ligated fairly strainlessly for both R,R and S,S configurations
of the sulfur atoms. Conversely, similar models containing
tetrahedral CuI results in a highly strained configuration when
the sulfur atoms are R,R and a much less-strained configuration
when they are S,S.
Cu-S
2.286(2)
1.990(5)
2.546(4)
1.821(7)
1.783(7)
1.54(1)
1.381(9)
1.397(9)
1.38(1)
1.36(1)
1.38(1)
1.424(9)
1.451(9)
C9-N
1.273(8)
1.450(8)
1.40(1)
1.37(1)
1.38(1)
1.39(1)
1.35(1)
1.42(1)
1.48(2)
1.407(9)
1.345(9)
1.397(9)
1.406(9)
Cu-N
N-C10
C10-C11
C10-C15
C11-C12
C12-C13
C13-C14
C14-C15
C15-C15′
B-F1
Cu-F4
S-C2
S-C3
C2-C2′
C3-C4
C3-C8
C4-C5
C5-C6
C6-C7
C7-C8
C8-C9
B-F2
B-F3
B-F4
a Estimated standard deviations in the least significant digits are given
in parentheses. Coordinates of primed atoms are related to those of the
corresponding unprimed atoms by the transformation, 1.0 - x, 0.5 -
y, z.
a
Table 5. Bond Angles (deg) for [CuII(bite)](BF4)2
angle (deg)
angle (deg)
S-Cu-S′
S-Cu-N
89.77(9)
87.4(2)
176.5(2)
95.4(3)
91.7(1)
84.9(1)
C7-C8-C9
C8-C9-N
Cu-N-C9
114.9(6)
126.7(6)
125.2(5)
118.7(4)
116.1(6)
116.6(7)
120.6(7)
122.8(7)
118.5(8)
120.1(9)
120.1(8)
122.4(9)
116.0(8)
124.3(5)
119.5(6)
110.2(6)
108.3(6)
108.0(6)
110.3(6)
111.8(6)
108.2(6)
S-Cu-N′
N-Cu-N′
S-Cu-F4
S-Cu-F4′
N-Cu-F4
N-Cu-F4′
F4-Cu-F4′
Cu-S-C2
Cu-S-C3
C2-S-C3
S-C2-C2′
S-C3-C4
S-C3-C8
C4-C3-C8
C3-C4-C5
C4-C5-C6
C5-C6-C7
C6-C7-C8
C3-C8-C7
C3-C8-C9
Cu-N-C10
C9-N-C10
N-C10-C11
N-C10-C15
C11-C10-C15
C10-C11-C12
C11-C12-C13
C12-C13-C14
C13-C14-C15
C10-C15-C14
C10-C15-C15′
C14-C15-C15
Fl-B-F2
EXAFS. [CuI(bite)](BF4). The copper K edge spectrum of
[CuI(bite)](BF4) in the solid state (Figure S-5, Supporting
Information) differs slightly from the spectrum obtained in
frozen acetonitrile; however, the EXAFS spectra are quite
similar as shown in Figure 4a. The position of the K edge
confirms the assignment of a formal oxidation state of +1 to
the copper center (8984 eV for copper(I) versus 8988 eV for
copper(II)). The Fourier transforms (FTs) both exhibit a broad
peak with an R′ value of approximately 1.6 Å. (Note, R′ is not
necessarily equal to the radial distance of the scattering atom
because the phases are not known independently.) The source
of the peak is likely to be the two coordinated nitrogen atoms
and the nearest coordinated sulfur atom. The relatively close
spacing between the values of the three distances determined
by X-ray crystallography (Cu-S(1) 2.194(2) Å, Cu-N(1)
1.953(5) Å, Cu-N(2) 1.938(5) Å) offers an explanation for the
lack of resolution in the solid state and the only slight resolution
in the solution state. The farthest sulfur atom (Cu-S(2)
2.323(2) Å) is apparently not visible in the FT due to weaker
scattering at the larger distance. While the EXAFS and FT data
support a strong similarity of the solid and solution state
structures for the near coordination sphere, the slight differences
observed in the edge spectra suggest a structural difference more
removed from the copper atom. Possible explanations may be
93.1(2)
90.1(2)
175.3(2)
101.4(2)
100.3(2)
105.2(3)
108.4(4)
120.7(5)
117.8(5)
120.8(6)
118.9(6)
121.5(7)
121.1(7)
118.3(7)
119.3(6)
125.6(6)
F1-B-F3
Fl-B-F4
F2-B-F3
F2-B-F4
F3-B-F4
a Estimated standard deviations in the least significant digits are given
in parentheses. Coordinates of primed atoms are related to those of the
corresponding unprimed atoms by the transformation, 1.0 - x, 0.5 -
y, z.
has been described as “semi-coordinated” by the BF4- anions.46
Recent structures obtained for related nickel complexes45 suggest
(46) Brown, D. S.; Lee, J. D.; Melsom, B. G. A. Acta. Crystallogr. 1968,
B24, 730.

8864 J. Am. Chem. Soc., Vol. 119, No. 38, 1997
Flanagan et al.
the distant coordination of a fifth ligand such as acetonitrile or
a less dramatic structural change involving outer solvation of
the cation. Conductivity measurements reported below suggest
that [CuI(bite)](BF4) is a 1:1 electrolyte in CH2Cl2 solution
which is consistent with the lack of observed fluorine in the
solution state FT.
in dichloromethane, with the most notable difference being the
cutoff points of the respective solvents (Figure S-3, Supporting
Information). However, the spectrum obtained in acetonitrile
(λmax ≈ 350 nm; ꢀ ) 8000 M-1 cm-1) is significantly blue
shifted with respect to the other two (λmax ≈ 400 nm; ꢀ ) 6000
M-1 cm-1). When interpreted in light of the EXAFS and
variable-temperature 1H NMR data below, this spectral behavior
provides some evidence of an interaction between [CuI(bite)]-
(BF4) and acetonitrile.
[CuII(bite)](BF4)2. The UV-vis spectrum of [CuII(bite)]-
(BF4)2 in acetonitrile, as shown in Figure S-4 (Supporting
Information), displays a low-energy band at 646 nm not found
in the [CuI(bite)](BF4) spectrum. This moderately intense band
(ꢀ ) 1154 cm-1 M-1) looks quite like a “d-d” transition arising
from a distorted square-planar geometry.47 Other synthetic
copper(II) complexes with thioether donors also exhibit low-
energy bands of similar intensity.48 Most likely, these LF bands
achieve their enhanced intensity by stealing intensity from an
intense [CuII r S* (thioether)] charge-transfer (CT) band located
at higher energy via a vibronic coupling mechanism. In
comparison, the band in the spectrum of Pc(CuII) at 595 nm
with ꢀ = 4500 cm-1 M-1 is thought to arise from a [CuII r
S(cysteine)] charge transfer rather than a [CuII r S*(methionine)]
transition.49
The spectrum obtained in acetone is blue shifted only slightly
from that obtained in acetonitrile. In addition, the intensity of
each peak is slightly lower in acetone. Since [CuII(bite)](BF4)2
decomposes in acetone and acetonitrile if left for extended
periods of time (several days), the reduced intensity of the
absorptions in acetone may reflect a slightly more rapid
decomposition in that solvent. The resemblance of the two
spectra suggests that the structure of the compound is likely to
be very similar in both solvents, thus coordination by either
solvent is improbable.
[CuII(bite)](BF4)2. The K edge, EXAFS and FT spectra of
[CuII(bite)](BF4)2 all show very little change upon comparing
the solid state and frozen acetonitrile solution data (Figure S-6,
Supporting Information). The position of the K edge confirms
the assignment of a formal +2 oxidation state to the copper
center (8984 eV for copper(I) versus 8988 eV for copper(II)).
The absence of a near edge peak at approximately 8980 eV
corresponding to a 1s f 3d transition in the solution state
spectrum suggests that the complex does not undergo a geometry
change to tetrahedral upon dissolution, since tetrahedral com-
plexes of copper(II) generally display such a band. Since
octahedral and square-planar copper complexes are not known
to exhibit well-resolved 1s f 4pz transitions (as would be seen
for nickel), the spectrum is consistent with a “pseudo-octahedral”
coordination geometry as observed in the solid state and with
“pseudo-octahedral” or square-planar geometry in solution. The
splitting observed in the first shell FT peak (Figure 4b) is
expected, the lower R′ peak being mainly due to Cu-N and
the higher R′ peak being mainly due to Cu-S. The “pseudo-
coordinated” fluorine atoms observed at a distance of 2.546(4)
Å in the solid state X-ray crystallographic structure are not
observed in the FT, most likely due to the long distance and
poor scattering ability of fluorine. (Note that the sulfur atom
at 2.323(2) Å in [CuI(bite)](BF4) is not seen either, despite the
shorter distance and stronger scattering ability of sulfur relative
to that of fluorine.) The fluorine could well be displaced by
acetonitrile nitrogen in the solution state and still be consistent
with the observed spectra. Thus, the data are consistent with a
variety of solution-state species, including a square-planar
complex, a complex weakly bound to BF4-, or a complex
weakly bound to CH3CN.
Far IR Spectroscopy. The presence of two long axial
-
CuII-F interactions of 2.546(4) Å with BF4 (see X-ray data
above) gives the structure of [CuII(bite)](BF4)2 a striking
resemblance to that of [CuII(en)2](BF4)2, which has been
described as “semi-coordinated” by the BF4- anions.46 The far
infrared spectrum of [CuII(en)2](BF4)2 (nujol mull) contains a
Conductivity. A saturated solution of [CuII(bite)](BF4)2 in
CH2Cl2 (2 × 10-5 M) registered a molar conductivity of 69
cm2 equiv-1 Ω-1 as compared to 80 cm2 equiv-1 Ω-1 for [CuI-
(bite)](BF4) and 100 cm2 equiv-1 Ω-1 for [(n-Bu)4N](BF4) at
the same concentrations at 25 °C. [CuI(bite)](BF4) and [CuII-
(bite)](BF4)2 in dichloromethane both behave as near 1:1
electrolytes when compared to [(n-Bu)4N](BF4). The solution-
state conductivities indicate a substantial degree of dissociation
of the “semi-coordinated” BF4- counterions (as observed in the
solid state) of the Cu(II) molecule in CH2Cl2. This may be
represented by the equilibrium
-
weak, symmetry forbidden, BF4 absorption at 350 cm-1
indicating the “semi-coordination” of a BF4- anion.50 A similar
band is observed for [CuII(bite)](BF4)2 at 350 cm-1, while the
spectrum of [CuI(bite)](BF4) contains no such band. These
results are consistent with the relative specificity of the Cu-F
interactions seen in the X-ray crystal structures.
EPR Spectroscopy. [CuII(bite)](BF4)2. Past attention has
often focused upon the EPR spectrum of Pc(CuII) which displays
an unusually small A value (0.008 cm-1) compared to simple
||
-
[CuII(bite)](BF4)2 h [CuII(bite)](BF4)+ + BF4
(2)
copper(II) coordination compounds.51 In this regard, [CuII(bite)]-
(BF4)2 gives an undistinguished copper(II) EPR spectrum having
g ) 2.00, g ) 2.17, and A ) 0.015 cm-1, with integration
Conductivity measurements alone do not allow one to distin-
guish between a weak ion pair and a coordinated BF4- for the
species indicated as [CuII(bite)](BF4)+. Note that CH2Cl2 as a
||
||
of the signal demonstrating that all copper in the sample is EPR
active.
-
solvent would favor association of [CuII(bite)]2+ with BF4
Electrochemistry. [CuI(bite)](BF4). Although [NO](BF4)
oxidizes [CuI(bite)](BF4) by one electron chemically, the
compound displays poor electrochemical reversibility. Cyclic
voltammetry experiments in CH3CN show that at lower scan
much more strongly than would CH3CN, because of both the
difference in dielectric constants and the difference in their
nucleophilicities. Thus, CH3CN could displace weakly-
coordinated BF4-, while CH2Cl2 would not. The data are
consistent with the acetonitrile EXAFS data presented above
in that [CuII(bite)](BF4)2 may exist in CH3CN as a free four-
coordinate complex, as a loose ion pair with BF4-, as a complex
weakly coordinated by one BF4- anion, or as any of the above
with a coordinated CH3CN ligand.
(47) Styka, M. C.; Smierciak, R. C.; Blinn, E. L.; DeSimone, R. E.;
Passariello, J. V. Inorg. Chem. 1978, 17, 82.
(48) Martin, M. J.; Endicott, J. F.; Ochrymowycz, L. A.; Rorabacher,
D. B. Inorg. Chem. 1987, 26, 3012-3022.
(49) Gewirth, A. A.; Solomon, E. I. J. Am. Chem. Soc. 1988, 110, 3811.
(50) Brown, D. S.; Lee, J. D.; Melsom, B. G. A.; Hathaway, B. J.; Procter,
I. M.; Tomlinson, A. A. G. J. Chem. Soc., Chem. Commun. 1967, 369.
(51) Boas, J. F. In Copper Proteins and Copper Enzymes; Lontie, R.,
Ed.; CRC Press: Boca Raton, FL, 1984; Vol. 1, p 2.
Electronic Spectroscopy. [CuI(bite)](BF4). The UV-vis
spectrum obtained in acetone closely resembles that obtained

Fully-Gated Electron Transfer
J. Am. Chem. Soc., Vol. 119, No. 38, 1997 8865
Table 6. Redox Potentials for Pertinent CuI/II Couplesa
b
copper redox pair
reference electrode
solvent
DAS
N4/N4O
N2S2*/N2S2*(F2)
S4*/S4*O2
N2SS*
S4*/S5*
S3*O/S4*O2
N2SS*
N4
N2S(?)
N5
N5
N5
N5
E1/2
ref
60
TW
48
61,62
48
48
63-65
7
66
12
[CuI/II(2,9-Me2phen)2]+/2+
[CuI/II(bite)]+/2+
SCE
Ag/AgCl
NHE
CH3CN
CH3CN
H2O
H2O
H2O
H2O
H2O
CH3CN
H2O
CH3CN
CH3CN
CH3CN
CH3CN
+0.92
+0.91
+0.82
+0.78
+0.75
+0.60
+0.38
+0.35
+0.18
+0.10
-0.02
-0.03
-0.12
[CuI/II(Me2-2,3,2-S4)]+/2+
[CuI/II(fungal laccase)]+/2+
[CuI/II([15]aneS5)]+/2+
NHE
NHE
[CuI/II([14]aneS4)]+/2+
[CuI/II(plastocyanin)]+/2+
[CuI/II(2,2′-bis(2-imidH)bp)2]+/2+
[CuI/II(stellacyanin)]+/2+
[CuI/II(py)2DAP]+/2+
SCE
SCE
SCE
SCE
Ag/AgCl
[CuI/II(imidR)2DAP]+/2+
[CuI/II(imidH)2DAP]+/2+
[CuI/II((5-MeimidH)2DAP)]+/2+
12
12
67
a DAS ) donor atom set. TW ) this work. b Potentials are reported relative to NHE. Correction factors of 0.22 and 0.24 V were added to
potentials reported versus Ag/AgCl and SCE, respectively, to adjust the values to NHE. This correction ignores unknown solvent junction potentials
for results obtained in nonaqueous solvents but should yield qualitatively useful information. Potentials are compared in this manner due to the fact
that many of the compounds of interest are not reported versus Fc/Fc+.
Figure 5. Cyclic voltammograms in CH3CN (2.5 × 10-2 M NaBF4)
at 100 mV/s (Pt button working electrode, Pt wire auxiliary electrode,
and Ag/AgCl reference electrode) of (a) ferrocene (10-3 M), (b) [CuI-
(bite)](BF4) (10-3 M), and (c) solvent + electrolyte blank.
rates, anodic and cathodic peaks develop, but with large peak-
to-peak separations (Figure 5). The oxidative peak exhibits
pronounced broadening and reduced current intensity with
1
Figure 6. Variable-temperature H NMR (300 MHz) of [CuI(bite)]-
(BF4) in acetone-d6.
respect to the reductive peak. The [CuI/II(bite)](BF4)1,2 redox
couple shows only quasi-reversible behavior (∆Epp ) 0.36 V)
compared to the Fc0/+ couple (∆Epp ) 0.13 V) at a scan rate of
50 mV/s. The E1/2 of +0.69 V versus Ag/AgCl (E1/2 ) +0.34
for ferrocene under the same conditions) is quite high for a CuI/II
couple (see Table 6), although the large value of ∆Epp introduces
some uncertainty into this assessment. The wide peak-to-peak
separation and the improved behavior at low scan rates imply
sluggish electron transfer at the electrode.
The redox potentials observed for the compounds in this study
are compared to those from other systems in Table 6. A unique
aspect of the comparison made in the table is that the structures
of both the copper(I) and copper(II) forms of all species listed
are either known from X-ray crystallography or can be inferred
with a high degree of certainty from structures of related
complexes. Two trends can be identified in the table. One is
the shift toward higher potentials for those complexes having
Cu-S(thioether) linkages, which is as expected in view of the
strong stabilization of Cu(I) by this soft-binding donor atom.
The other trend is for those systems having the higher potentials
to have significant structural differences between the two
oxidation states, either through a change in coordination number
or through some other alteration such as inversion at sulfur (as
in [Cu(bite)]n+). This tendency emerges despite the rather varied
coordination geometries and donor atoms of the various
complexes. The trend may reflect the fact that copper(II) is
rather strongly destabilized in environments which favor
tetrahedral copper(I) and tends to force addition of a fifth ligand
upon oxidation. Thus, ligands such as 2,9-dimethyl-1,10-
phenanthroline, which strongly favor copper(I), have difficulty
accommodating the electronic requirements of copper(II), even
after undergoing large geometric distortion to accommodate an
additional ligand.
Variable-Temperature 1H NMR of [CuI(bite)](BF4). Vari-
able-temperature 1H NMR spectra have been collected in
acetone-d6 (Figure 6), dichloromethane-d2, and acetonitrile-d3
(Figures S-7 and S-8, Supporting Information). The spectra
obtained for acetone-d6 and dichloromethane-d2 are qualitatively
very similar. In both cases, very little change is observed in
the spectra as the temperature is lowered except for the ethylene
signal which appears as a doublet of doublets at low temperature.
At higher temperatures, these signals converge to a singlet,

8866 J. Am. Chem. Soc., Vol. 119, No. 38, 1997
Flanagan et al.
Table 7. 1H NMR Line Broadening Data for [CuI(bite)](BF4) as a
Function of Temperature and the Concentration of [CuII(bite)](BF4)2:
Line Widths (Hz), First-Order Electron Self-Exchange Rate
Constants, and Their Standard Deviationsa
concentration of [CuII(bite)](BF4)2 (M)
(K) 0.0 0.2 × 10-3 0.4 × 10-3 0.6 × 10-3 0.8 × 10-3 (s-1) (s-1
temp
kex σkex
)
235 1.56
255 1.07
275 1.04
295 1.18
315 1.39
325 1.15
1.84
2.71
4.17
7.72
14.35
17.86
2.40
2.25
3.97
7.84
13.99
18.94
1.93
2.25
4.21
7.87
1.89
2.10
4.00
8.99
1.4 0.8
4.0 0.8
9.6 0.4
21.7 1.9
40.1 0.8
54.2 2.4
a In acetone-d6 (µ ) 25 mM, TBABF4 electrolyte). Line width is of
the imine H+ resonance at ∼8.6 ppm.
although without significant line broadening. The only signifi-
cant difference of behavior between the two solvents is the
convergence temperature for this signal. In dichloromethane-
d2, the splitting pattern converges to a sharp singlet at 287 (
10 K, whereas in acetone-d6, convergence appears to lie above
296 K. (Note, the peaks at 3.85 and 4.50 ppm are present in
the acetone-d6 as obtained from the supplier and do not arise
from protons on the bite ligand; the peak at 1.50 ppm in
dichloromethane-d2 is due to water.) The behavior at low
temperature implies that the ethylene protons exist as two pairs
of equivalent nuclei, which indicates that the asymmetry seen
in the crystal structure is effectively removed in solution. This
could be easily achieved by a rapid oscillation in the two Cu-S
bond lengths, leading to effective C2 symmetry. The high-
temperature behavior evidently arises from a temperature-
dependent chemical shift, leading to accidental overlap of the
two resonances. This could arise from a temperature-dependent
conformational equilibrium within the ethylene linkage.
A significantly different situation is obtained in acetonitrile-
d3. The behavior of the ethylene signal appears to be fairly
similar to that observed in the other solvents in the range of
288-348 K, with convergence of the signal to a sharp singlet
imminent at 348 K. However, below 288 K, the spectrum is
dramatically different. All peaks begin to broaden significantly,
and the imine signal reaches a coalescence at 248 ( 10 K. By
238 K, the imine pattern has begun to resolve into two broad
signals of unequal intensity (Figure S-8, Supporting Informa-
tion). Once again, this behavior of [CuI(bite)](BF4) in CD3CN
is quite different from that in any other solvent. The splitting
of the imine signal at low temperature is most likely due to
either (1) coordination of acetonitrile or (2) an equilibrium
mixture of two isomers, related, perhaps, by inversion at sulfur.
Slow interconversion for either of these processes at low
temperature could lead to the observed behavior. Acetonitrile,
because of its relatively high dielectric constant, could uniquely
support such an equilibrium. Either of these equilibrium
explanations would also offer reasonable explanations for the
EXAFS and UV-vis spectra already discussed. The lack of
any such behavior in acetone and the similarity of the acetone
spectra to those obtained in noncoordinating dichloromethane
strongly suggest that [CuI(bite)]+ remains cleanly four-
coordinate in those solvents. For this reason, and because of
solubility considerations, acetone-d6 was selected as the solvent
of choice for the electron self-exchange kinetics.
Figure 7. Plot of T2-1 (s-1) versus concentration (mM) of [Cu(II)] for
the [CuI/II(bite)](BF4)1,2 system as function of temperature in acetone-
d6.
significant line broadening but subsequent increments in
[Cu(II)] lead to no further broadening. This behavior is
illustrated in Figure 7. From these results the overall rate law
for the electron-exchange process is given simply as
rate of exchange ) kex[CuI(bite)]+
(3)
Least-squares analysis of the results gives a value of 21.7 (1.9)
s-1 for kex at 295 K in acetone-d6. The activation parameters
have likewise been determined to be ∆Hq ) 23.3(0.7) kJ mol-1
and ∆Sq ) -140.8(2.3) J mol-1 K-1
.
In view of the unusual rate law for the electron-exchange
process and the X-ray structural evidence for the redox reaction
leading to inversion at both sulfur atoms, it is reasonable to
think that the mechanism consists of some of the following
steps:
[Cu(bite)]+ h *[Cu(bite)]+
[Cu(bite)]2+ h *[Cu(bite)]2+
(4)
(5)
[Cu(bite)]+ + *[Cu(bite)]2+ h *[Cu(bite)]2+ + [Cu(bite)]+
(6)
*[Cu(bite)]+ + [Cu(bite)]2+ h [Cu(bite)]2+ + *[Cu(bite)]+
(7)
*[Cu(bite)]+ + *[Cu(bite)]2+ h [Cu(bite)]2+ + [Cu(bite)]+
(8)
[Cu(bite)]+ + [Cu(bite)]2+ h *[Cu(bite)]2+ + *[Cu(bite)]+
(9)
Electron Self-Exchange Kinetics by 1H NMR Line Broad-
ening. The electron self-exchange kinetics in acetone-d6 was
investigated by use of 1H NMR line broadening measurements.
For this purpose, the line width of the resonance at ∼8.8 ppm
arising from the imine protons of [CuI(bite)]+ was measured as
a function of the concentration of [CuII(bite)]2+ over a range of
temperatures. The results are given in Table 7, and they show
a rather unusual trend: the first addition of Cu(II) leads to a
In this mechanism, steps 4 and 5 are intramolecular isomeriza-
tion reactions, while steps 6-9 are electron-transfer reactions
that preserve the approximate configurations of the reactants.
Process 6 can be ruled out, because it would not lead to the
observed rate law. Processes 7 and 8 could lead to the observed
rate law if prior isomerization of Cu(I) were rate limiting.
Process 9 could also be operative if relaxation of the product,

Fully-Gated Electron Transfer
J. Am. Chem. Soc., Vol. 119, No. 38, 1997 8867
*[Cu(bite)]2+, were slow. The molecular-modeling studies
described above suggest that isomerization of Cu(I) is unfavor-
able, which would be consistent with it as the rate-limiting
process. On the other hand, relaxation of *[CuII(bite)]2+ could
be slow, even if it were less energetically driven than relaxation
of *[CuI(bite)]+. Irrespective of whether process 7, 8, or 9 is
operative, the derived rate law identifies kex as k4.
[CuI/II([15]aneS5)]+/2+ exceeds those of all other small-molecule
copper systems published to date.54 Van de Linde et al. have
suggested that, in fact, the enthalpic contribution to the electron
transfer may be much smaller than expected for a bond breaking/
forming process. The authors of that study postulate that the
CuII-S bond rupture is a very low-energy process and occurs
at a rate of about 4 orders of magnitude faster than electron
transfer. Comparison with the activation parameters of [CuI/II(5-
MeimidH)2DAP]+/2+ reveals that the values for both enthalpy
and entropy of activation for the two systems (one CNI, the
other non-CNI) are essentially identical.54 Since the [CuI/II(5-
MeimidH)2DAP]+/2+ system undergoes no bond breaking or
forming, the Cu-S bond which breaks in [CuII[15]aneS5)]2+
must not introduce a large barrier to electron transfer.
Rorabacher and co-workers52,53 have examined numerous
cross-exchange reactions of copper systems in which two
isomers of copper can be formed in either oxidation state. When
cross reactions are performed with simple redox partners that
exhibit no such isomerism, the consequence is a mechanistic
“square scheme”. Note that the scheme in reactions 4-9 is
more complex than the square scheme. This arises because both
the oxidant and reductant can participate in isomerization
reactions. As the Rorabacher group has shown, the square
scheme can lead to a variety of complex kinetic behaviors,
including first-order kinetics. However, they have never
reported a self-exchange reaction that shows first-order kinetics.
In this regard, our observations on the [CuI/II(bite)]+/2+ system
are unique.
Finally, the activation entropy of [CuI/II(bite)](BF4)1,2 is
extremely negative at ∆Sq ) -140.8(2.3) J mol-1 K-1. An
explanation for this low entropy of activation in the [CuI,II
-
(bite)]+/2+ system will depend on the mechanism of isomeriza-
tion of Cu(I) (eq 4). Entropy of activation in blue-copper
proteins such as azurin lies at the other extreme, with ∆Sq )
+96 (29) J mol-1 K-1 35
This enormous entropy value likely
.
An interesting comparison can be made with the [CuI/II(trans-
cyhx[14]aneS4)]+/2+ redox couple.53 It has been shown by X-ray
crystallography that inversion at two of the S atoms accompanies
the redox process for this macrocyclic complex. The two atoms
undergoing the inversion process are trans to each other in the
macrocyclic ring, which thus differs from the [CuI/II(bite)]+/2+
system where the two inverting S atoms are cis to each other.
reflects the release of solvent molecules upon approach of two
protein molecules, a process for which there is no analogy in
small molecules. Thus, caution should be exercised in making
casual assessments about the active site (a “small-molecule”
CNI site) to overall protein electron-transfer rates.
Conclusions
Efforts to measure the self-exchange rate constant of the [CuI/II
-
A combination of results from EXAFS, X-ray crystallography,
electronic spectroscopy, NMR spectroscopy, conductivity, and
far IR spectroscopy has provided a fairly clear picture of the
solution state structures of [CuI(bite)](BF4) and [CuII(bite)]-
(BF4)2. In acetonitrile, [CuI(bite)](BF4) shows unusual behavior,
which could be due to a configurational equilibrium or to the
binding of an acetonitrile molecule as a fifth ligand, consistent
with the known affinity of acetonitrile for copper(I) (e.g.,
[CuI(CH3CN)4]+).57 In acetone and dichloromethane, however,
no evidence of solvent or anion coordination has been observed.
Since copper(I) prefers four-coordinate tetrahedral coordination
environments to any other, it is almost certain that a close
approximation of the tetrahedral, four-coordinate solid-state
structure is present in these two solvents. [CuII(bite)](BF4)2 may
weakly coordinate one BF4- anion in the solution state, but the
lack of any difference between the UV-vis spectra obtained in
strongly coordinating acetonitrile and weakly coordinating
acetone suggests that the complex adopts the same structure in
both solvents. Thus, a four-coordinate, square-planar geometry
is the most likely structure for this complex in solution.
Of the copper couples so far examined, two stand out as
especially informative with respect to the question of blue-
copper electron transfer, in part because of the macrocyclic
nature of their ligands which helps to insure outer-sphere
electron transfer. [CuI/II(bite)]+/2+ and [CuI/II([15]aneS5)]+/2+54
have both been characterized by X-ray crystallography in both
the copper(I) and copper(II) forms, and each system offers its
own unique insight into copper electron transfer.
[CuI/II([15]aneS5)]+/2+ offers an interesting look at a system in
which the coordination number changes from 5 to 4 upon
reduction of the copper(II) to copper(I). Despite this bond
breaking process, electron transfer remains fast (k ) 1.9(0.6)
× 105 M-1 s-1 at 298 K) and the enthalpy of activation is rather
low (∆Hq ) 14(4) kJ/mol). [CuI/II(bite)]+/2+ offers an informa-
tive comparison, in that, while this complex is almost certainly
(trans-cyhx[14]aneS4)]+/2+ redox couple by 1H NMR line
broadening were not successful because additions of Cu(II) led
to no observable line broadening. This behavior was attributed
to a low second-order self-exchange rate constant.53 We suggest
as an alternative explanation that the rate law is actually overall
first-order, as seen for [CuI/II(bite)]+/2+
.
The self-exchange rate constant for [CuI/II(bite)]+/2+, 21.7
(1.9) s-1 (295 K, acetone-d6), is difficult to compare directly to
rate constants observed for other copper systems, as it appears
to be the sole case for which electron self-exchange is truly
first order at copper(II) concentrations on the order of 1 mM.
The [CuI/II(bite)](BF4)1,2 system is only the sixth copper complex
for which activation parameters have been determined by the
variable-temperature line broadening technique. The enthalpy
of activation of ∆Hq ) 23.3(0.7) kJ/mol is fairly high and may
be indicative of a Cu-ligand bond breaking/forming process
that could occur before electron transfer. One system in which
such a coordination number change is known to occur
([CuI/II([15]aneS5)]+/2+ 54
) was determined to have an enthalpy
of activation of only ∆Hq ) 14(4) kJ/mol. The complex is
5-coordinate and square pyramidal in the Cu(II) state and
tetracoordinate and pseudotetrahedral in the Cu(I) state.55 On
the basis of the large enthalpic changes expected for a bond
breaking/bond forming process, one might predict the self-
exchange rate for this small-molecule system to be quite slow.
In fact, a measurement has shown that the self-exchange rate
constant of k ) 2.2(1.1) × 105 M-1 s-1 (25 °C, D2O) for
(52) Dunn, B. C.; Ochrymowycz, L. A.; Rorabacher, D. B. Inorg. Chem.
1995, 34, 1954-1956 and references therein.
(53) Salhi, C. A.; Yu, Q.; Heeg, M. J.; Villeneuve, N. M.; Juntunen, K.
L.; Schroeder, R. R.; Ochrymowycz, L. A.; Rorabacher, D. B. Inorg. Chem.
1995, 34, 6053-6064 and references therein.
(54) Vande Linde, A. M. Q.; Juntunen, K. L.; Mols, O.; Ksebati, M. B.;
Ochrymowycz, L. A.; Rorabacher, D. B. Inorg. Chem. 1991, 30, 5037.
(55) Corfield, P. W. R.; Ceccarelli, C.; Glick, M. D.; Moy, I. W.;
Ochrymowycz, L. A.; Rorabacher, D. B. J. Am. Chem. Soc. 1985, 107,
2399-2404.
(56) Rijn, J. v.; Driessen, W. L.; Reedijk, J.; Lehn, J. Inorg. Chem. 1984,
23, 3584-3588.
(57) Sharpe, A. G. Inorganic Chemistry; Longman: London, 1986; p
611.

8868 J. Am. Chem. Soc., Vol. 119, No. 38, 1997
Flanagan et al.
coordination-number-invariant in acetone, significant geometric
reorganization occurs upon electron transfer. Despite being
coordination-number-invariant, electron transfer is quite slow,
has a high enthalpy of activation (∆Hq ) 23.3(0.7) kJ/mol) and
is a first-order process limited by sulfur inversion. One other
system worth careful scrutiny, despite not having a macrocyclic
ligand, is [CuI/II((1-Meimid)2bp)2]+/2+ (Figure 1).7 This system
exhibits very little geometric change upon reduction of the
copper(II) form and is also coordination-number-invariant. It
also demonstrates an immeasurably small (by 1H NMR T1
measurements) electron transfer rate (k < 1 × 102 M-1 s-1 at
293 K).
sense, the entatic state view is likely valid since it explains the
site geometry as a compromise between factors which stabilize
both the +1 and +2 copper oxidation states. Finally, and
perhaps most importantly, these studies of coordination-number-
invariant (and non-coordination-number-invariant) copper co-
ordination compounds have emphasized that one must exercise
extreme care in ascribing properties of the cuproprotein as a
whole to the copper active site.
Acknowledgment. This work was supported by the NSF
(CHE-9412736 to L.J.W. and D.M.S.) and the Robert A. Welch
Foundation (C-0627 to L.J.W.). S.F. thanks the U.S. NIH for a
NIGMS Training Grant (GM-08362) at Rice University, and
W.R.S. also thanks the U.S. NIH (GM-38401).
Taken together, the emerging data suggest that, at least for
these particular small-molecule compounds, small reorganiza-
tional processes at the copper site do not necessarily promote
fast electron-transfer reactions, particularly when one considers
the [CuI/II((1-MeimidH)2bp)2]+/2+ system which has a small
structural reorganization process and yet possesses unmeasurably
Supporting Information Available: Figures S-1 and S-2,
stereo packing diagrams for the unit cells of [CuI(bite)](BF4),
[CuII(bite)](BF4)2; Table S-I, fractional coordinates and isotropic
displacement coefficients, Table S-II, anisotropic displacement
parameters, Table S-III, H-atom coordinates and isotropic
displacement parameters, Table S-IV, torsional angles, Table
S-V, least squares planes and dihedral angles, and Table S-VI,
configuration of chiral centers for [CuI(bite)](BF4); Table S-VII,
fractional coordinates and isotropic displacement coefficients,
Table S-VIII, general atomic displacement parameters, Table
S-IX, fractional tetragonal coordinates and isotropic thermal
parameters for the fixed hydrogen contributors, Table S-X, root-
mean-square amplitudes of atomic displacements of the atoms
refined anisotropically, Table S-XI, torsional angles, Table
S-XII, least squares planes and dihedral angles, and Table
S-XIII, configuration of chiral centers for [CuII(bite)](BF4)2;
Table S-XIV, X-ray absorption spectroscopy parameters for
solid-state [CuI(bite)](BF4) and [CuII(bite)](BF4)2; Table S-XV,
X-ray absorption spectroscopy parameters for solution-state [CuI-
(bite)](BF4) and [CuII(bite)](BF4)2; Table S-XVI, self-exchange
rate constants for CNI Cu(I/II) couples and coalescence tem-
peratures for their Cu(I) compounds; Figure S-3, electronic
spectra of [CuI(bite)](BF4) in acetone, acetonitrile, and CH2-
Cl2; Figure S-4, electronic spectra of [CuII(bite)](BF4)2 in
acetone and acetonitrile; Figure S-5, X-ray absorption spectra,
EXAFS, and FTs for solid [CuI(bite)](BF4) and in acetonitrile
solution; Figure S-6, X-ray absorption spectra, EXAFS, and FTs
for solid [CuII(bite)](BF4)2 and in acetonitrile solution; Figure
small electron self-exchange rates.5-7 Furthermore, the [CuI/II
-
([15]aneS5)]+/2+ couple clearly demonstrates that a cuproprotein
active site need not be coordination-number invariant to promote
a fast electron-transfer reaction.54 The [CuI/II(bite)]+/2+ couple
further demonstrates that CNI does not ensure rapid electron
transfer either. As a corollary to these latter observations, the
blue-copper active sites could exhibit larger reorganization and
still be fast. This conclusion is at odds with the traditional
entatic state view which typically attributes rapid electron
transfer at the blue-copper sites to small structural differences
between the copper(I) and copper(II) state.58 This latter view
has become so widely accepted as to even appear in textbook
explanations of blue-copper reactivity.59 Guckert et al.38 and
Knapp et al.7 have offered alternative explanations for fast
electron transfer in terms of favorable electronic pathways
provided by the blue-copper site. The unusual nature of the
coordination geometry and ligating atoms of the blue-copper
site may have more to do with achieving high redox potentials
than with achieving fast electron transfer. The tendency of
coordination-number-invariant small-molecule copper couples
to display high redox potentials certainly bolsters this point of
view. Thus, in a thermodynamic (but not necessarily kinetic)
(58) Williams, R. J. P. Eur. J. Biochem. 1995, 234, 363-381.
(59) Kaim, W.; Schwederski, B. Bioinorganic Chemistry: Inorganic
Elements in the Chemistry of Life; Wiley: Chichester, 1994; p 196.
(60) Doine, H.; Yano, Y.; Swaddle, T. W. Inorg. Chem. 1989, 28, 2319-
2322.
(61) Malmstro¨m, B. G. In New Trends in Bio-Inorganic Chemistry;
Williams, R. J. P., DaSilva, R. J. R. F., Eds.; Academic Press: London,
1978; pp 59-78.
(62) Reinhammar, B. R. M. Biochim. Biophys. Acta 1972, 275, 245-
259.
1
S-7, variable-temperature H NMR spectra of [CuI(bite)](BF4)
1
in CD2Cl2; Figure S-8, variable-temperature H NMR spectra
of [CuI(bite)](BF4) with 15 mM [(CH3)4N]BF4 in acetonitrile-
d3; Figure S-9, Eyring plot for electron exchange in the
[CuI/II(bite)]+/2+ system; Figure S-10, recalculated electron-
exchange dependence on [Cu(II)] for the [CuI/II((5-MeimidH)2-
DAP)]+/2+ system; Figure S-11, recalculated electron-exchange
Eyring plot for the [CuI/II((5-MeimidH)2DAP)]+/2+ system (43
pages). See any current masthead page for ordering and Internet
access instructions.
(63) Sykes, A. G. AdV. Inorg. Chem. 1991, 36, 377-408.
(64) Bu¨chi, F. N.; Bond, A. M.; Codd, R.; Huq, L. N.; Freeman, H. C.
Inorg. Chem. 1992, 31, 5007-5014.
(65) Armstrong, F. A.; Butt, J. N.; Govindaraju, K.; McGinnis, J.; Powls,
R.; Sykes, A. G. Inorg. Chem. 1990, 29., 4858-4862.
(66) Peisach, J.; Levine, W. G.; Blumberg, W. E. J. Biol. Chem. 1967,
242, 2847-2858.
(67) Coggin, D. K. Ph.D. Dissertation, Rice University, 1990.
JA9644034