Ion-Pair Charge Transfer Photochemistry in Rhenium(I) Borate Salts

Article abstract of DOI:10.1021/ic9605074

The photophysics and photochemistry of the salt [(bpy)Re(CO)₃(py)⁺][BzBPh₃^-] (ReBo, where bpy = 2 2′-bipyridine, py = pyridine, Bz = C₆H₅CH₂ and Ph = C₆H₅) has been investigated in THF and CH₃CN solutions. UV-visible absorption and steady-state emission spectroscopy indicates that in THF ReBo exists primairly as an ion-pair. A weak absorption band is observed for the salt in THF solution that is assigned to an optical ion-pair charge transfer transition. Stern-Volmer emission quenching studies indicate that BzBPh₃^- quenches the luminescent dπ (Re) → * (bpy) metal-to-ligand charge transfer excited state of the (bpy)Re(CO)₃(py)⁺ chromophore. The quenching is attributed to electron transfer from the benzylborate anion to the photoexcited Re(I) complex, (bpy^-*)Re^II(CO)₃(py)⁺* + BzBPh₃^- → (bpy^-*)Re^I(CO)₃(Py) + BzBPh₃*. Laser flash photolysis studies reveal that electron transfer quenching leads to irreversible reduction of the Re(I) cation to (bpy^-*)Re^I(CO)₃(py). Photoinduced electron transfer is irreversible owing to rapid C-B bond fragmentation in the benzylboranyl radical, PhCH₂BPh₃* → PhCH₂* + BPh₃*. Quantitative laser flash photolysis experiments show that the quantum efficiency for production of the reduced complex (bPy^-*)Re^I(CO)₃(py) is unity, suggesting that C-B bond fragmentation in the benzylboranyl radical occurs more rapidly than return electron transfer within the geminate radical pair that is formed by photoinduced electron transfer.

Full text of DOI:10.1021/ic9605074

6800
Inorg. Chem. 1996, 35, 6800-6808
Ion-Pair Charge Transfer Photochemistry in Rhenium(I) Borate Salts
Bruce H. McCosar and Kirk S. Schanze*,†
Department of Chemistry, University of Florida, P.O. Box 117200, Gainesville, Florida 32611-7200
ReceiVed May 9, 1996X
The photophysics and photochemistry of the salt [(bpy)Re(CO)3(py)+][BzBPh3-] (ReBo, where bpy ) 2,2′-
bipyridine, py ) pyridine, Bz ) C6H5CH2 and Ph ) C6H5) has been investigated in THF and CH3CN solutions.
UV-visible absorption and steady-state emission spectroscopy indicates that in THF ReBo exists primairly as an
ion-pair. A weak absorption band is observed for the salt in THF solution that is assigned to an optical ion-pair
-
charge transfer transition. Stern-Volmer emission quenching studies indicate that BzBPh3 quenches the
luminescent dπ (Re) f π* (bpy) metal-to-ligand charge transfer excited state of the (bpy)Re(CO)3(py)+
chromophore. The quenching is attributed to electron transfer from the benzylborate anion to the photoexcited
•
Re(I) complex, (bpy-•)ReII(CO)3(py)+* + BzBPh3- f (bpy-•)ReI(CO)3(py) + BzBPh3 . Laser flash photolysis
studies reveal that electron transfer quenching leads to irreversible reduction of the Re(I) cation to (bpy-•)ReI(CO)3(py).
Photoinduced electron transfer is irreversible owing to rapid C-B bond fragmentation in the benzylboranyl radical,
•
•
•
PhCH2BPh3 f PhCH2 + BPh3 . Quantitative laser flash photolysis experiments show that the quantum efficiency
for production of the reduced complex (bpy-•)ReI(CO)3(py) is unity, suggesting that C-B bond fragmentation in
the benzylboranyl radical occurs more rapidly than return electron transfer within the geminate radical pair that
is formed by photoinduced electron transfer.
Introduction
While a substantial body of the experimental work concerning
charge transfer processes in metal-organic systems has centered
Photoinduced charge transfer processes in transition metal
complexes have been extensively investigated.1-4 Research in
this area is motivated by a number of factors. First, a significant
amount of information concerning the electronic structure of
inorganic and metal-organic complexes has been provided by
studies of charge transfer spectroscopy;5-8 furthermore, studies
of the spectroscopy and dynamics of charge transfer excited
states have provided a wealth of experimental data that has
guided theoretical work on electron transfer processes.2c,5-7,9-11
Second, metal complexes that exhibit charge transfer absorption
bands in the visible region have provided the paradigm for
chemical systems that convert light to chemical energy.12
Finally, there has been recent interest in the application of metal
complexes that feature charge transfer absorption and lumines-
cence to the development of molecular probes for biology and
chemical sensing.13,14
on inner-sphere transitions such as metal-to-ligand charge
transfer (MLCT),2 ligand-to-metal charge transfer (LMCT),3 and
ligand-to-ligand charge transfer (LLCT),15,16 there have been a
number of investigations that have focused on the properties of
“outer-sphere” transitions that involve charge transfer to or from
a species that is not directly coordinated to the metal center.16-18
In general, such outer-sphere charge transfer transitions involve
an ion-pair comprising an anionic electron donor and a cationic
electron acceptor and are termed ion-pair charge transfer (IPCT)
transitions.16
Our group has been developing (photo)chemical probes that
can be applied to study electronic structure and dynamics of
charge transfer excited states in metal complexes.19 Toward
this objective, we have applied a number of “reactive” organic
donors that undergo rapid, irreversible chemical processes
triggered by single electron oxidation. To date, these reactive
donors have been used mainly to probe ligand-to-ligand charge
transfer states in transition metal complexes.19
The study presented herein describes the application of
“reactive electron donors” to probe ion-pair charge transfer in
salts comprising a cationic metal complex acceptor and an
anionic organic “reactive donor”. The objective of this study
† E-mail: kschanze@chem.ufl.edu.
X Abstract published in AdVance ACS Abstracts, October 1, 1996.
(1) Gafney, H.; Adamson, A. W. J. Am. Chem. Soc. 1972, 94, 8238.
(2) (a) Bock, C. R.; Meyer, T. J.; Whitten, D. G. J. Am. Chem. Soc. 1974,
96, 4710. (b) Bock, C. R.; Connor, J. A.; Guitierrez, A. R.; Meyer, T.
J.; Whitten, D. G.; Sullivan, B. P.; Nagle, J. K. J. Am. Chem. Soc.
1979, 101, 4815. (c) Meyer, T. J. Prog. Inorg. Chem. 1983, 30, 389.
(3) Endicott, J. F. In Concepts of Inorganic Chemistry; Adamson, A. W.,
Fleischauer, P. D., Eds.; Wiley-Interscience: New York, 1975; p 81.
(4) Balzani, V.; Scandola, F. In Photoinduced Electron Transfer, Part D;
Fox, M. A., Chanon, M. D., Eds.; Elsevier: Amsterdam, 1988; p 148.
(5) (a) Creutz, C.; Taube, H. J. Am. Chem. Soc. 1973, 95, 1086. (b) Tom,
G. M.; Creutz, C.; Taube, H. J. Am. Chem. Soc. 1974, 96, 7827. (c)
Richardson, D. E.; Taube, H. J. Am. Chem. Soc. 1983, 105, 40.
(6) (a) Powers, M. J.; Meyer, T. J. J. Am. Chem. Soc. 1978, 100, 4393.
(b) Powers, M. J.; Meyer, T. J. J. Am. Chem. Soc. 1980, 102, 1289.
(7) (a) Doorn, S. K.; Hupp, J. T. J. Am. Chem. Soc. 1989, 111, 4704. (b)
Lu, H.; Petrov, V.; Hupp, J. T. Chem. Phys. Lett. 1995, 235, 521.
(8) Zink, J. I.; Shin, K.-S. AdV. Photochem. 1991, 16, 119.
(9) Walker, G. C.; Barbara, P. F.; Doorn, S. K.; Hupp, J. T. J. Phys. Chem.
1991, 95, 5712.
(13) Pyle, A. M.; Barton, J. K. Prog. Inorg. Chem. 1990, 38, 412.
(14) Demas, J. N.; DeGraff, B. A. Anal. Chem. 1991, 63, 829A.
(15) Vogler, A.; Kunkeley, H. Commun. Inorg. Chem. 1990, 9, 201.
(16) For reviews of work in the area of ion-pair charge transfer see: (a)
Vogler, A.; Kunkely, H. In Photoinduced Electron Transfer II; Mattay,
J., Ed.; Topics in Current Chemistry 158; Springer-Verlag: Berlin,
1990; p 1. (b) Billing, R.; Rehorek, D.; Hennig, H., Ibid. p 151.
(17) (a) Kochi, J. K.; Bockman, T. M. AdV. Organomet. Chem. 1991, 33,
51. (b) Bockman, T. M.; Kochi, J. K. J. Am. Chem. Soc. 1989, 111,
4669.
(18) (a) Wang, Z.; Kutal, C. Inorg. Chim. Acta 1994, 226, 285. (b)
Sugimoto, H.; Hataoka, H.; Mori, M. J. Chem. Soc., Chem. Commun.
1982, 1301. (c) Sabbatini, N.; Bonazzi, A.; Ciano, M.; Balzani, V. J.
Am. Chem. Soc. 1984, 106, 4055.
(10) Spears, K. G.; Wen, X.; Arrivo, S. M. J. Phys. Chem. 1994, 98, 9693.
(11) (a) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265.
(b) Newton, M. D.; Sutin, N. Annu. ReV. Phys. Chem. 1984, 35, 437.
(12) (a) Energy Resources Through Photochemisry and Catalysis; Gra¨tzel,
M., Ed.; Academic Press: New York, 1983. (b) Hagfeldt, A.; Gra¨tzel,
M. Chem. ReV. 1995, 95, 49.
(19) (a) Wang, Y.; Hauser, B. T.; Rooney, M. M.; Schanze, K. S. J. Am.
Chem. Soc. 1993, 115, 5675. (b) Wang, Y.; Lucia, L. A.; Schanze, K.
S. J. Phys. Chem. 1995, 99, 1961. (c) Wang, Y.; Schanze, K. S. J.
Phys. Chem. 1996, 100, 5408.
S0020-1669(96)00507-1 CCC: $12.00 © 1996 American Chemical Society

Rhenium(I) Borate Salts
Inorganic Chemistry, Vol. 35, No. 23, 1996 6801
vacuum until all the THF solvent had evaporated. Following this, 10
mL of argon-degassed water was added via a cannula. Under a mild
flow of argon, the flask was shaken and stirred to dissolve the MgCl+
salt of the product. Another Schlenk flask containing an argon-degassed
solution of 1.5 g of tetramethylammonium chloride (13.6 mmol) in 30
mL of water was prepared, and the aqueous borate solution was
transferred into this Schlenk flask by filter cannula, whereupon a cloudy-
white precipitate formed immediately. After thorough mixing of the
solution, the precipitate was allowed to settle and crystallize further
before being collected on a Bu¨chner funnel in air. The product was
recrystallized twice by dissolving the solid in a minimum amount of
dry acetonitrile and dropping the solution into a rapidly stirring beaker
of diethyl ether (typically 6- to 10-fold of the amount of the acetonitrile).
The product was obtained as a white powdery solid, yield 1.06 g (80%).
Anal. Calcd for C29H34BN: C, 85.50; H, 8.41; N, 3.44. Found: C,
85.64; H, 8.70; N, 3.46. 1H NMR (300 MHz, CD3CN): δ 2.48 (q,
2H, benzyl CH2 split by spin +3/2 boron nucleus), 3.08 (s, 12H, 4 ×
methyls), 6.56 (d, 2H), 6.65-6.85 (m, 6H), 6.94 (t, 6H), and 7.21 (m,
6H). 13C NMR (75 MHz, CD3CN): δ 54.90 (methyls), 121.41, 125.19,
125.22, 125.25, 125.68, 129.23, and 134.75
fac-(2,2′-Bipyridyl)tricarbonylpyridylrhenium(I) Benzyltriphen-
ylborate (ReBo). The following procedure was performed under red
light. In a Schlenk tube, 306 mg of RePF6 (470 µmol) and 193 mg of
TBo (473 µmol) were dissolved in 25 mL of argon-degassed acetoni-
trile. Degassed water was added to the acetonitrile solution via cannula
until the stirred solution became cloudy; at this point, a minimum
amount of degassed acetonitrile was added to clear it again. The
acetonitrile was then evaporated under reduced pressure, leaving a
brownish oil in water. This oil was extracted by cannula addition of
degassed methylene chloride (triple or greater the volume of the
remaining water). After vigorous mixing, no trace of the oil remained
and the methylene chloride layer was yellow. The methylene chloride
solution was carefully transferred via a cannula to a dry Schlenk flask
containing anhydrous MgSO4. On the following day, the methylene
chloride solution was decanted from the MgSO4 and transferred by
cannula to another dry, argon-degassed Schlenk flask, whereupon the
solution was concentrated to 10 mL volume under reduced pressure.
The methylene chloride solution was then transferred directly into a
500 mL flask containing 300 mL of air-saturated diethyl ether that was
being stirred. At this point the yellow ReBo product began to
precipitate slowly. The powdered precipitate became less sticky by
allowing it to stir in the diethyl ether for 20-30 min. The product
was obtained as an air-stable yellow powder, yield 315 mg (80%). Solid
ReBo was air-stable; however, in solution the salt decomposes on
exposure to light or air. Anal. Calcd for C43H35BN3O3Re: C, 61.57;
H, 4.21; N, 5.01. Found: C, 61.19; H, 4.04; N, 4.82. 1H NMR (300
MHz, CD3CN): δ 2.48 (q, 2H, Bo benzyl CH2 split by spin +3/2 boron
nucleus), 6.54 (d, 2H, Bo aromatics), 6.61-6.84 (m, 6H, Bo aromatics),
6.92 (t, 6H, Bo aromatics), 7.16-7.30 (m, 8H, Bo aromatics overlapping
with Re 2 × m-pyridyl), 7.74 (t, 2H, Re 2 × 5-bpy), 7.82 (t, 1H, Re
p-pyridyl), 8.20 (t, 2H, Re 2 × 4-bpy), 8.28 (d, 2H, Re 2 × o-pyridyl),
8.31 (d, 2H, Re 2 × 6-bpy), and 9.20 (d, 2H, Re 2 × 3-bpy).
was to design a system in which photoexcitation of an ion-pair
could be coupled to an exceedingly fast, irreversible chemical
process. We rationalized that by monitoring the efficiency of
the overall photochemistry it would be possible to gain
information concerning the electronic structure and dynamics
of the IPCT state produced by photoexcitation.
In order to design an effective system, an anionic electron
donor is required which undergoes a rapid chemical process
triggered by single electron oxidation. Our studies were guided
by Schuster and co-workers’ recent work which indicates that
alkyltriphenylborates (R-BPh3-) undergo an exceedingly rapid
C-B bond fragmentation upon single electron oxidation, i.e.,20
-e
-
•
R-BPh3
-8 R-BPh3 fast8R• + BPh3
(1)
Thus, in the present study we have focused attention on salts
which pair the cationic metal complex acceptor (bpy)ReI(CO)3-
(py)+ (bpy ) 2,2′-bipyridine and py ) pyridine) with the anionic
“reactive” donor benzyltriphenylborate, BzBPh3- (Bz ) C6H5-
CH2- and Ph ) C6H5-). This system is of interest for several
reasons. First, the (bpy)ReI(CO)3(py)+ cation exhibits a com-
paratively long-lived dπ (Re) f π* (bpy) MLCT excited state
which is strongly luminescent. This excited state is a relatively
strong oxidant and has been well characterized with respect to
its photoredox reactions with a variety of neutral electron
donors.21 Second, in nonpolar solvents at relatively low
concentration [(bpy)ReI(CO)3(py)+][BzBPh3-] exists predomi-
nantly as an ion-pair. In contrast, in polar solvents at relatively
low concentration, this salt exists primarily as solvent-separated
or free ions. Therefore, comparison of the overall efficiency
of the photochemistry observed for the salts in nonpolar and
polar solvents allows evaluation of the effect on ion-pairing on
the rates of the charge transfer processes.
Experimental Section
General Synthetic Methods. Anhydrous THF was prepared by
distilling over NaK/benzophenone prior to use. Triphenylboron was
prepared by acidifying an aqueous solution of the NaOH adduct with
CO2, filtering out the solid, drying it in Vacuo, and purifying it by
vacuum sublimation. All other solvents and chemicals were of reagent
grade and used without purification. NMR spectra were taken on a
GE QE 300-MHz instrument.
fac-(2,2′-Bipyridyl)tricarbonylpyridylrhenium(I)Hexafluorophos-
phate (RePF6). This compound was prepared and purified by using a
literature method19a,22 and was isolated as a bright yellow powder in
88% yield. Anal. Calcd for C18H13F6N3O3PRe: C, 33.24; H, 2.01; N,
6.46. Found: C, 32.85; H, 1.73; N, 6.08. 1H NMR (300 MHz, CD3-
CN): δ 7.28 (t, 2H, 2 × m-pyridyl), 7.78 (t, 2H, 2 × 5-bpy), 7.84 (t,
1H, p-pyridyl), 8.25 (t, 2H, 2 × 4-bpy), 8.29 (d, 2H, 2 × o-pyridyl),
8.38 (d, 2H, 2 × 6-bpy), and 9.22 (d, 2H, 2 × 3-bpy). 13C NMR (75
MHz, CD3CN): δ 124.5, 126.4, 128.6 (m-, p-, and o- pyridyl, respec-
tively), 139.7, 140.9, 151.7, and 153.6 (5-, 4-, 6-, and 3-bpy, respec-
tively).
Photophysical and Photochemical Experiments. Except for UV-
visible absorbance measurements, all photochemical experiments were
carried out in solvents that were thoroughly degassed with argon. UV-
visible spectra were recorded using an HP 8452A diode array
spectrophotometer; spectra shown in this work are composites of many
measurements, taken over a wide range of dilutions, with absorbances
discarded if greater than 1 or less than 0.01. Luminescence lifetimes
and steady-state emission data were determined using methods and
equipment described in previous publications.21,22 Luminescence
Tetramethylammonium Benzyltriphenylborate (TBo).20a Tri-
phenylboron (788 mg, 3.25 mmol) was dissolved in 20 mL of anhydrous
THF in a 250-mL Schlenk flask. After degassing with argon, 10 mL
of 1.0 M benzylmagnesium chloride in diethyl ether (10 mmol) was
added to the flask via a syringe. Purged a final time with argon, the
mixture was allowed to stir and react for 1 h at 0 °C and 2 h more at
room temperature. The Schlenk flask was then maintained under
2+
quantum yields were determined relative to Ru(bpy)3 in degassed
water (Φem ) 0.057) and were corrected for differences in solvent
refractive index. Transient absorption experiments were performed
using a nanosecond flash photolysis system also described previously,22b
exciting the samples with the third harmonic of a Nd:YAG laser (355
nm, 6 ns fwhm).
Relative actinometry transient absorption experiments were carried
out by using a modified version of the technique described by
Carmichael and Hug23 and by us in a previous publication.21b Briefly,
(20) (a) Chatterjee, S.; Davis, P. D.; Gottschalk, P.; Kurz, M. E.; Sauerwein,
B.; Yang, X.; Schuster, G. B. J. Am. Chem. Soc. 1990, 112, 6329. (b)
Murphy, S. T.; Zou, C.; Miers, J. B.; Ballew, R. M.; Dlott, D. D.;
Schuster, G. B. J. Phys. Chem. 1993, 97, 13152.
(21) (a) Schanze, K. S.; MacQueen, D. B.; Perkins, T. A.; Cabana, L. A.
Coord. Chem. ReV. 1993, 122, 63. (b) Lucia, L. A.; Schanze, K. S.
Inorg. Chim. Acta 1994, 225, 41.
(22) (a) MacQueen, D. B.; Schanze, K. S. J. Am. Chem. Soc. 1991, 113,
7470. (b) Wang, Y.; Schanze, K. S. Chem. Phys. 1993, 176, 305.
(23) Carmichael, I.; Hug, G. L. J. Phys. Chem. Ref. Data 1986, 15, 1.

6802 Inorganic Chemistry, Vol. 35, No. 23, 1996
McCosar and Schanze
For convenience these spectra are separated into two spectral
regions, with the first extending from 300 to 330 nm in the UV
and the second from 330 nm in the UV to 450 nm in the visible.
In the first region, both ReBo and RePF6 exhibit two com-
paratively strong transitions (ꢀ g 10 000 M-1 cm-1) which are
ascribed primarily to ligand localized (bpy and py) π,π*
transitions.19c Although for each salt these ligand localized π,π*
transitions are blue-shifted slightly in CH3CN compared to THF,
the significant feature is that the spectra of ReBo and RePF6
in the same solvent are superimposable in this region. Thus,
we conclude that the benzylborate anion has no effect on the
absorption of the complex in the 300-330 nm region in THF
where ion-pairing is expected to be significant.
In the second region, both ReBo and RePF6 exhibit a broad,
featureless absorption with ꢀ ≈ 4 × 103 M-1 cm-1. This
absorption is due primarily to the dπ (Re) f π* (bpy) MLCT
transition. However, close inspection of the 330-450 nm region
reveals an interesting feature: the spectra of RePF6/THF,
RePF6/CH3CN, and ReBo/CH3CN are closely similar, while
ReBo/THF shows slightly greater absorptivity between 330 and
450 nm. It is important that the enhanced absorptivity of ReBo
in THF is reproducible and quite clearly is a property unique
to ReBo in the nonpolar solvent. A difference spectrum of
ReBo relative to RePF6 in THF indicates that the enhanced
absorptivity is due to a broad, featureless transition with λmax
≈ 350 nm and ꢀmax ≈ 300 M-1 cm-1. We attribute this broad
absorption band observed for ReBo in THF to an ion-pair charge
transfer (IPCT) transition from the benzylborate anion to the
Re(I) cation.
Figure 1. UV-visible absorption spectra of RePF6 and ReBo. (a)
ReBo/THF; (b) ReBo/CH3CN; (c) RePF6/THF; (d) RePF6/CH3CN.
Spectra were obtained on solutions with 1.0-4 M concentrations.
a solution of RePF6 (c ) 120 µM) in THF served as an actinometer.
The transient absorption of the Re f bpy MLCT excited state (ReMLCT
)
was monitored at 370 nm and the concentration produced by the laser
pulse was determined by using the value of ∆ꢀM37L0CT for ReMLCT at 370
21b
nm (11 300 M-1 cm-1
)
t)0
∆A
∆ꢀ
370 nm
concentration of ReMLCT
)
(2)
MLCT
370
Emission Spectroscopy of RePF6 and ReBo. As noted
above, the (bpy)ReI(CO)3(py)+ chromophore displays a mod-
erately intense luminescence that emanates from the dπ (Re)
f π* (bpy) MLCT excited state manifold.21a In order to probe
the effect of the benzylborate anion and ion-pairing on the
MLCT state, steady state and time resolved luminescence
experiments were carried out on RePF6 and ReBo in CH3CN
and THF solutions.
t)0
where ∆A
is the transient absorption signal extrapolated to zero
370 nm
time. Sample solutions contained RePF6 (c ) 120 µM) along with
TBo added in a concentration sufficient to quench 90% of the Re-
bpy MLCT emission. The transient absorption of the neutral Re
complex (bpy-•)ReI(CO)3(py) (Re-0) produced in the sample solutions
by 355 laser excitation was monitored at 350 nm and its concentration
was determined by using the value of ∆ꢀ
Re-0
at 350 nm (5800 M-1
350
cm-1
)
21b
Both complexes display a broad, featureless MLCT lumi-
nescence in CH3CN and THF solution; emission band maxima
(λem) and luminescence quantum efficiencies (Φem) are listed
in Table 1. First, Φem for ReBo is lower compared to RePF6
in the same solvent. The extent of quenching of the steady-
state luminescence by the benzylborate is quantified by the
∆A350 nm
concentration of Re-0 )
(3)
Re-0
350
∆ꢀ
where ∆A350 nm is the transient absorption signal at 350 nm for the
sample solution. Actinometer and sample solutions were run consecu-
tively and repeatedly at laser powers ranging from 0.1 to 2.0 mJ/pulse
(exposure area ca. 0.25 cm2). The RePF6 actinometer solutions were
contained in a 3.0 mL quartz cuvette and were stirred vigorously during
data acquisition. ReBo sample solutions were contained an a recir-
culating cell with a total volume of 50 mL to minimize the effects of
irreversible photochemistry upon the yield determinations.
parameter, ΦReBo/ΦRePF (Table 1) which indicates that lumi-
6
nescence from ReBo is suppressed (compared to RePF6) by
approximately a factor of 2 in CH3CN and a factor of 10 in
THF. Emission lifetimes (τem) were also determined for the
two complexes in CH3CN and THF (Table 1). Consistent with
the steady-state luminescence data, τem is suppressed for ReBo
compared to RePF6 and the extent of lifetime quenching in each
Results
solvent is quantified by the parameter τReBo/τRePF (Table 1). In
6
Salts and Solvents. The focus of this study is on the
CH3CN solution, τem and Φem are quenched to the same extent
by the benzylborate anion; however, in THF solution Φem is
quenched to a much greater extent than τem by the borate.
Several conclusions can be drawn from the luminescence data
presented thus far. First, it is clear that the benzylborate
quenches the MLCT excited state. Given that the singlet (and
triplet) states of the borate are at significantly higher energies
relative to the Re f bpy MLCT state, it is unlikely that
quenching is due to energy transfer.18a,20 Rather, we believe
that the quenching is due to electron transfer from the ben-
zylborate anion to the photoexcited Re(I) complex, i.e.
-
properties of the ion-pair [(bpy)ReI(CO)3(py)+,BzBPh3 ] (ReBo),
which contains the photoactive Re(I) cation paired with the
benzylborate electron donor. The properties of [(bpy)ReI(CO)3-
(py)+,PF6-] (RePF6) were examined in parallel to allow assess-
ment of the photophysics of the Re(I) cation in the absence of
the benzylborate donor. Both salts were examined in THF and
CH3CN, which were selected as examples of nonpolar and mod-
erately polar solvents, respectively. As outlined below, under
the conditions of the experiments carried out herein, ReBo and
RePF6 exist primarily as ion-pairs in THF and as free ions in
CH3CN.
(bpy-•)ReII(CO)3(py)+* + BzBPh3- f
Absorption Spectroscopy. The absorption spectra of ReBo
and RePF6 in the near-UV and visible regions as dilute solutions
in THF and CH3CN are presented for comparison in Figure 1.
•
(bpy-•)ReI(CO)3(py) + BzBPh3 (4)

Rhenium(I) Borate Salts
Inorganic Chemistry, Vol. 35, No. 23, 1996 6803
Table 1. Luminescence Properties of Re(I) Saltsa
λmaxem/nm
Φem
τ
em/ns
Φ
ReBo/ΦRePF
τ
ReBo/τRePF
1010 kqSS/M-1 s-1
1010 kqLT/M-1 s-1
6
6
salt
solvent
RePF6
RePF6
ReBo
ReBo
CH3CN
THF
CH3CN
THF
588
586
588
584
0.055
0.067
0.035
0.0081
234
238
160
102
1.9
13
1.8
3.4
0.64
0.12
0.68
0.43
a
em
SS
LT
λmax is emission maximum; Φem is emission quantum yield; τem is emission lifetime; kq and kq are apparent Stern-Volmer quenching
constants for TBo computed from steady-state and emission decay data, respectively.
Electron transfer quenching is supported by the fact that the
free energy change for eq 4 is estimated to be -0.1 eV in CH3-
CN based on the oxidation potential of BzBPh3- (Eox ≈ +1.09
V)20a and the excited state reduction potential of (bpy)ReI(CO)3-
(py)+ (Ered ≈ +1.22 V).21a
Second, the consistency of the steady-state and lifetime
quenching in CH3CN implies that electron transfer quenching
occurs only by a dynamic (i.e. diffusional) pathway in this
solvent. However, the disparity between the steady-state and
lifetime quenching in THF solution strongly implies that in this
solvent quenching occurs by both static and dynamic path-
ways.24-26 These qualitative results provide evidence that under
the conditions of the emission studies (c ≈ 120 µM), in CH3-
CN the Re(I) cation and benzylborate anion exist predominantly,
if not exclusively, as free ions, while in THF solution at 120
µM the salts exist predominantly as ion-pairs wherein electron
transfer quenching is very rapid, i.e.
[(bpy-•)ReII(CO)3(py)+*,BzBPh3 ] fast8
-
•
[(bpy-•)ReI(CO)3(py),BzBPh3 ] (5)
Stern-Volmer Luminescence Quenching. In order to care-
fully probe the effect of ion-pairing on the MLCT emission
Stern-Volmer luminescence quenching studies were carried out.
The concentration of the (bpy)ReI(CO)3(py)+ lumiphore was
fixed at 120 µM by addition of RePF6 and the concentration of
Figure 2. (A) Stern-Volmer plots for quenching of RePF6 by TBo.
Polygons represent experimental data and solid lines are linear least
squares fits of the data. (B) Comparison of experimental I0/I data points
in THF to various theoretical fits: I (dotted line), linear least-squares
fit of τ0/τ data; II (solid line), linear least-squares fit of I0/I data; III
(dashed line), “best” fit of experimental data to eq 9 (see text).
-
the BzBPh3 quencher was varied by addition of [NMe4+]-
[BzBPh3-] (TBo). Total luminescence quenching (i.e., static
and dynamic) was assessed by analysis of steady state intensity
quenching according to eq 6 while the contribution of dynamic
quenching only was evaluated by analysis of lifetime quenching
data according to eq 7. In eqs 6 and 7, I°em and τ°em refer to
for CH3CN, which qualitatively indicates that quenching is more
efficient in the less polar solvent. Another distinguishing feature
of the THF data is that the slope of the I°/I plot is significantly
greater than that of the τ°/τ plot. The latter feature is a clear
indication that static quenching occurs in THF and further
indicates that ReBo exists predominantly as ion-pairs in this
solvent.24-26
Before discussing the quantitative analysis of the Stern-
Volmer quenching data, it is useful to note that the rate of
dynamic quenching cannot exceed the diffusion limit. The
Debye-Stokes equation can be used to estimate the rate constant
for a diffusion-limited process27
I°
em ) 1 + kqSSτ°em[BzBPh3 ]
(6)
(7)
-
Iem
τ°
em ) 1 + kqLTτ°em[BzBPh3 ]
-
τem
the steady-state emission intensity and lifetime, respectively,
of RePF6 in the solvent of interest and kq represents the second-
order rate constant for quenching as determined from the slope
of the Stern-Volmer plots, with the superscripts SS and LT
referring to steady-state and lifetime, respectively.
Figure 2A illustrates the intensity and lifetime Stern-Volmer
plots for THF and CH3CN solutions. First, the plots of I°/I
and τ°/τ for CH3CN solutions are exactly co-linear. This clearly
indicates that in CH3CN dynamic quenching occurs exclusively
and ReBo exists as free ions. The Stern-Volmer data for
solutions in THF contrasts with that in CH3CN. Plots of both
I°/I and τ°/τ are more steeply sloped for THF compared to those
2NAkBT (rA + rD)2
b
diff
kq
)
(8)
(
)(
)
eb - 1
3η
rArD
where b ) (ZDZAe2)/(4πꢀoꢀrDAkBT), NA is Avogadro’s number,
η is the solvent viscosity, e is the electron charge, ꢀo is the
permittivity of free space, ꢀ is the static dielectric constant of
the solvent, rDA is the reaction encounter distance, ZD and ZA
are the charges on the donor (D) and acceptor (A) ions,
respectively, and rD and rA are the radii of D and A, respectively.
By using ZA ) +1 and rA ) 4.2 Å for the (bpy)Re(CO)3(py)+
acceptor and ZD ) -1 and rD ) 4.7 Å for the BzBPh3 donor
(24) Bolletta, F.; Maestri, M.; Moggi, L.; Balzani, V. J. Phys. Chem. 1974,
78, 1374.
(25) Rybak, W.; Haim, A.; Netzel, T. L.; Sutin, N. J. Phys. Chem. 1981,
85, 2856.
(26) Frank, R.; Rau, H. J. Phys. Chem. 1983, 87, 5181.
(27) Laidler, K. J. Chemical Kinetics, 3rd ed.; Harper and Row: New York,
1987; pp 212-220.

6804 Inorganic Chemistry, Vol. 35, No. 23, 1996
McCosar and Schanze
and rDA ) rD + rA, eq 8 predicts that for CH3CN and THF,
diff
respectively, kq is 4.0 × 1010 M-1 s-1 and 1.0 × 1011 M-1
s-1
.
The I°/I and τ°/τ data for both CH3CN and THF solutions
were subjected to linear-least squares analysis according to eqs
6 and 7, and the kq values are listed in Table 1.28 As expected,
for CH3CN solutions the kSS and kLqT values are the same
q
within experimental error and are close to the diffusion limit
calculated by the Debye-Stokes equation. This is reasonable,
-
given that electron transfer from BzBPh3 to (bpy-•)ReII-
(CO)3(py)+* is exothermic by ≈ 0.1 eV.1c,20 The kLqT value
determined from the THF lifetime data is also in accord with
an electron transfer reaction that occurs at slightly less than the
diffusion controlled limit.1c,20 However, a linear fit of the I°/I
data for THF solution provides an apparent quenching rate
constant that is substantially greater than that obtained by
lifetime quenching and slightly greater that the diffusion limit
estimated by Debye-Stokes equation. This observation clearly
indicates that static quenching occurs in this solvent.
Under conditions where static quenching occurs, the steady-
state emission intensity of the (bpy)Re(CO)3(py)+ chromophore
is expected to follow a dependence of the form,24,26
Figure 3. Transient absorption-difference spectra: (A) 120 mM RePF6
in THF solution, delay times ranging from 0 to 400 ns after laser pulse;
(B) 58 mM RePF6 and 100 mM triethylamine in THF solution, delay
times ranging from 0 to 40 µs after laser pulse; (C) 120 mM RePF6
and 360 mM TBo in THF solution, delay times ranging from 0 to 3.2
µs after laser pulse.
I°
em ) (1 + kqτ°[Q])(1 + âKIP[Q])
(9)
Iem
where kq is the rate constant for diffusional quenching, â is the
ratio of the extinction coefficients for the (bpy)Re(CO)3(py)+
chromophore as a free ion and in the ion-pair (ꢀion-pair / ꢀfree ion),
KIP is the assocation constant for the [(bpy)Re(CO)3-
(py)+,BzBPh3-] ion pair, and [Q] is the concentration of the
benzylborate quencher. This equation predicts that the plot of
I°/I will display upward curvature. Inspection of Figure 2A
reveals that over the concentration range of TBo that was
examined, curvature in the I°/I plot is not readily apparent.
However, the Stern-Volmer experiments in THF were limited
because TBo was not soluble above 1 mM. Furthermore, the
experiments were not carried out at constant ionic strength, and
previous studies demonstrate that under these conditions linear
plots of I°/I are observed for systems in which ion-pairing is
important.25 Nonetheless, in an effort to provide an estimate
for KIP, eq 9 was used to fit the plot of I°/I for the THF solution
data. The dashed line in Figure 2B was calculated by using eq
9 with kq ) 3.4 × 1010 M-1 s-1, â ) 1.0, and KIP ) 8 × 103
M-1. While the value of KIP arrived at in this manner is clearly
only an approximation, it is in accord with values determined
by Kochi and Bockman for ion-pairs comprised of monovalent
organic cations paired with monovalent organometallic anions
in THF solution.17
In summary, the Stern-Volmer luminescence quenching
experiments lead to the following conclusions. (1) In CH3CN
solution for concentrations e5 mM quenching of the Re f bpy
MLCT state occurs exclusively by dynamic quenching. No ion-
pairing is apparent within this concentration range. (2) In THF
solution MLCT quenching occurs by static and dynamic
pathways. In this solvent ion-pairing is significant even at
concentrations e0.1 mM. (3) The luminescence data is
consistent with an association constant KIP ) 8 × 103 M-1 for
ReBo in THF.
(2) the quantum efficiency for formation of the electron transfer
products. These studies were carried out in THF and CH3CN
with the objective of determining whether ion-pairing has an
effect on the nature and/or yields of the electron transfer
products. All experiments were carried out with the concentra-
tion of (bpy)Re(CO)3(py)+ held at 120 µM and the concentration
of BzBPh3 was adjusted by addition of TBo. The optical
density of solutions at the laser excitation wavelength (355 nm)
was matched in all samples.
Figure 3A illustrates transient absorption difference spectra
for a solution of RePF6 in THF at delay times ranging from 0
to 400 ns following laser excitation. The strong absorption at
370 nm and moderate bleaching at 320 nm are characteristic of
the Re f bpy MLCT excited state. In accord with this
assignment the transient decays with τ ) 240 ns, in good
agreement with the emission lifetime. The difference molar
absorptivity for the MLCT state of RePF6 in THF at 370 nm
(∆ꢀ370) was determined to be 11 300 M-1 cm-1 by the relative
actinometry method relative to the MLCT state of (bpy)Re-
(CO)3(4-benzylpyridine)+ in CH3CN (∆ꢀ370 ) 11 600 M-1
cm-1).21b
Figure 3B illustrates transient absorption spectra for a solution
of RePF6 in THF with 100 mM of triethylamine electron donor
at delay times ranging from 0 to 40 µs following laser excitation.
The difference absorption spectra are assigned to the neutral
complex, (bpy-•)ReI(CO)3(py), that is produced by electron
transfer quenching of the MLCT excited state by triethylamine,
i.e.,21b
-
(bpy-•)ReII(CO)3(py)+* + NEt3 f
+•
(bpy-•)ReI(CO)3(py) + NEt3 (10)
Transient Absorption Spectroscopy. Nanosecond laser
flash photolysis experiments were carried out in order to identify
the following: (1) the primary products of excited state electron
transfer between the Re(I) cation and the benzylborate anion;
The (bpy-•)ReI(CO)3(py) chromophore displays a moderate
(difference) absorption at 350 nm and a weak band at 490 nm
in the visible. The neutral complex persists for long times
+•
following laser excitation because NEt3 irreversibly decom-
(28) A priori the I°/I data for THF solutions should not be linear. This
point is discussed more thoroughly below.
poses and therefore cannot undergo return electron transfer with

Rhenium(I) Borate Salts
Inorganic Chemistry, Vol. 35, No. 23, 1996 6805
the reduced complex.29 In previous work we have determined
that the molar absorptivity of the (bpy-•)ReI(CO)3(py) chro-
mophore at 350 nm is ∆ꢀ350 ) 5800 M-1 cm-1 21b
.
Figure 3C illustrates transient absorption spectra for a THF
solution of RePF6 in the presence of 360 µM TBo for delay
times ranging from 0 to 3.2 µs following laser excitation. The
spectrum at the earliest delay time features a band at 370 nm
which is characteristic of the Re f bpy MLCT state; however,
the spectra at later delay times are clearly due to the neutral
complex (bpy-•)ReI(CO)3(py). The transient absorption at-
tributed to (bpy-•)ReI(CO)3(py) does not decay substantially
within the timescale accessible by the nanosecond flash system
(200 µs), indicating that return electron transfer to the oxidized
benzylborate donor is infeasible. The luminescence quenching
studies indicate that under the conditions used for the data shown
in Figure 3c, >90% of the Re f bpy MLCT state is quenched
and approximately (75%) of the Re(I) is present in ion-pairs
with the benzylborate anion. Thus, the data in Figure 3c are
consistent with the prompt formation of (bpy-•)ReI(CO)3(py)
via photoinduced electron transfer from the benzylborate anion
(eqs 4 and 5) followed by rapid decomposition of the benzylbo-
ranyl radical via C-B bond fragmentation
Figure 4. Plot of concentration of (bpy-•)Re(CO)3(py) vs concentration
of MLCT excited state, (bpy-•)Re(CO)3(py)+*, produced at laser powers
ranging from 0.1 to 2.0 mJ/pulse. Circles and squares represent data
for THF and CH3CN solutions, respectively. The solid line represents
the linear regression fit to the two data sets. The slope of the line is
0.92.
plot of the concentration of (bpy-•)ReI(CO)3(py) produced in
the sample solution vs the concentration of the Re f bpy MLCT
state produced in the actinometer at corresponding laser powers.
Several features are of interest with respect to this data. First,
the slope of the correlation in this figure is equal to the ratio of
the concentration of (bpy-•)ReI(CO)3(py) produced in the sample
to the concentration of the MLCT state produced in the
actinometer, averaged over the range of laser powers used for
the experiment. Assuming that the quantum efficiency for
formation of the MLCT state is unity, this ratio provides the
quantum yield for formation of the (bpy-•)ReI(CO)3(py) by
photoinduced electron transfer. The interesting feature is that,
for the experiments in THF and CH3CN, the data follow the
same correlation and the slope is approximately 0.92. In view
of the fact that in both solvents the transient absorption
experiments were performed under conditions where the MLCT
state was quenched with 90% efficiency (i.e., φ ) 0.90), this
result indicates that in both CH3CN and in THF (bpy-•)-
ReI(CO)3(py) is produced with unit efficiency following BzBPh3-
quenching of the Re f bpy MLCT state.
Steady State Photochemical Products. Although the work
presented herein is primarily concerned with the initial products
produced by photoinduced electron transfer in ReBo (i.e.,
products present within 50 µs of excitation), several qualitative
experiments were carried out to determine the identity of the
ultimate product(s) produced by continuous photolysis. Before
presenting these results, however, it is useful to note that several
groups have studied the reactivity of 19 e- complexes of type
(bpy-•)Re(CO)3(L), where L is either a neutral ligand such as
pyridine or R-CN or an anion such as halide or triflate.31,32
These studies suggest that the 19 e- complexes are unstable
and dissociate into a 17 e- radical via loss of L (eq 12).
•
•
PhCH2BPh3 f PhCH2 + BPh3
(11)
The C-B bond fragmentation is expected to produce the
benzyl radical which absorbs moderately at 318 nm (∆ꢀ ≈ 8000
M-1 cm-1);30 however, the absorption of this transient is is not
apparent in Figure 3C. This may be due to strong ground state
bleaching which is observed between 300 and 330 nm due to
depletion of the ground state Re(I) chromophore.
A transient absorption experiment carried out on a CH3CN
solution of RePF6 that contained TBo at a concentration of 1.6
mM produced time resolved spectra that are virtually the same
as those shown for the THF solution in Figure 3c. Thus, the
-
quenching of the Re f bpy MLCT state by BzBPh3 in the
more polar solvent environment also produces the neutral
complex (bpy-•)ReI(CO)3(py).
A series of quantitative transient absorption studies were
effected to determine the absolute quantum yield of (bpy-•)-
ReI(CO)3(py) that is produced by photoinduced electron transfer.
The objective of these experiments was to compare the
efficiency for formation of the neutral complex: (1) in THF
solution where (bpy)ReI(CO)3(py)+ is present largely as ion-
-
pairs with BzBPh3 and static quenching predominates; (2) in
CH3CN solution where (bpy)ReI(CO)3(py)+ is present predomi-
natly as the free ion and only diffusional quenching occurs. In
order to allow direct comparision of the yields in the two
solvents, the experiments were carried out with the same
concentration of (bpy)ReI(CO)3(py)+ (to insure that the samples
had the same optical density); however, the TBo quencher
concentration was adjusted such that 90% MLCT quenching
was achieved in both solvents (CH3CN, [RePF6] ) 120 µM,
[TBo] ) 1.6 mM; THF, [RePF6] ) 120 µM, [TBo] ) 300
µM). In both cases the yields of (bpy-•)ReI(CO)3(py) were
determined relative to the yield of the Re f bpy MLCT excited
state produced in a THF solution of RePF6 for laser powers
ranging from 0.1 to 2.0 mJ/pulse (see Experimental Section for
details).
•
(bpy-•)Re(CO)3(L) f (bpy)Re(CO)3 + L
(12)
Electrochemical studies suggest that this process takes place
on timescales ranging from milliseconds to seconds after
reduction.31 Thus, the 19 e- species (bpy-•)Re(CO)3(py) is
anticipated to be stable on the time scale of the transient
absorption experiments presented above, while permanent
photochemistry observed for ReBo is expected to derive from
Figure 4 illustrates a compilation of the quantitative transient
absorption data for THF and CH3CN solutions presented as a
(29) Monserrat, K.; Foreman, T. K.; Gra¨tzel, M.; Whitten, D. G. J. Am.
Chem. Soc. 1981, 103, 6667.
(30) Chatgilialoglu, C. In Handbook of Organic Photochemistry; Scaiano,
J. C., Ed.; CRC Press: Boca Raton, FL, 1989; Vol. II, p 1.
(31) O’Toole, T. R.; Younathan, J. N.; Sullivan, B. P.; Meyer, T. J. Inorg.
Chem. 1989, 28, 3923.
(32) Stor, G. J.; Hartl, F.; van Outersterp, J. W. M.; Stufkens, D. J.
Organometallics 1995, 14, 1115.

6806 Inorganic Chemistry, Vol. 35, No. 23, 1996
McCosar and Schanze
secondary reactions of the 17 e- radical, (bpy)Re(CO)3• which
forms by dissociation of (bpy-•)Re(CO)3(py).31,32 In view of
these facts, the photoproducts of continuous photolysis of ReBo
are expected to arise either by dimerization of (bpy)Re(CO)3
or coupling of the metal and benzyl radicals (eqs 13 and 14).
the orbital basis for the IPCT absorption in ReBo bears a close
analogy to IPCT bands previously observed in ion-pairs
comprising organic pyridinium and bipyridinium cations with
organic or organometallic anions. For example, Sullivan et al.
observed an IPCT absorption band at λmax ) 360 (ꢀmax ) 150
M-1 s-1) for CH3CN solutions of the salt 1,1′-dimethyl-4,4′-
bipyridinium tetraphenylborate, [MV2+][BPh4-]2.34 IPCT bands
in salts comprising pyridinium cations paired with organome-
tallic anions have also been reported by by Kochi and Bock-
man.17
•
•
2(bpy)Re(CO)3 f (bpy)(CO)3Re-Re(CO)3(bpy) (13)
•
•
(bpy)Re(CO)3 + PhCH2 f (bpy) (CO)3Re-CH2Ph (14)
The energy of an IPCT absorption band for a charge transfer
salt in solution is given by16
Our first experimental investigations examined the changes
in the UV-visible absorption spectrum induced by photolysis
of a degassed CH3CN solution of ReBo (c ) 1.2 mM) at 436
nm. Irradiation induces the appearance of a broad visible
absorption band with λmax ≈ 450 nm and a concomitant decrease
in the near-UV absorption of ReBo. An isosbestic point at λ
≈ 400 nm is maintained during the early stages of photolysis
(to ca. 25% conversion). Next, FTIR studies were carried out
∆EIPCT ) ∆E1/2 - ∆GIP + ø
(15)
where ∆EIPCT is the absorption band maximum, ∆E1/2 is the
difference between the half-wave potentials for electrochemical
oxidation and reduction of the anionic donor and cationic
acceptor, respectively, ∆GIP is the free energy for formation of
the ion-pair from the free ions, and ø is the reorganization energy
associated with the optical ion-pair charge transfer transition.
to monitor changes in the carbonyl stretching bands (νCO
induced by photolysis. In CH2Cl2 solution ReBo exhibits
characteristic νCO bands at 1936 cm-1 (broad) and 2037 cm-1
)
.
For the ReBo system, it is possible to estimate ø ≈ 1.0 eV
The FTIR spectrum of the photoproduct(s) in CH2Cl2 (produced
by irradiation of ReBo at 436 nm in degassed CH3CN solution)
is characterized by νCO bands at 1897 cm-1 (broad) and 2023
cm-1. Finally, the mass spectrum of the photoproduct mixture
(low resolution, positive ion FAB) shows prominent ions at m/e
517 and m/e 427.
-
from experimental values of ∆EIPCT ) 3.54 eV, E1/2(BzBPh3
BzBPh3) ) 1.09 V, E1/2(bpy/bpy-•) ) -1.16 V and ∆GIP
/
)
-0.23 eV. As noted above, Kochi and Bockman have
investigated the spectroscopic properties ion-pair charge transfer
in salts of the form A+Co(CO)4-, where A is a cationic
pyridinium or quinolinium cation.17 Extrapolation of a plot of
∆EIPCT vs ∆E1/2 for the A+Co(CO)4- series to ∆E1/2 ) 0 leads
to an estimate of ø ) 0.9 eV after correction for ∆GIP. Thus,
there is good qualitative agreement between the values of ø
calculated for the ReBo and A+Co(CO)4- systems, a fact which
supports the IPCT assignment for the weak band seen for ReBo
in THF.
The product studies are consistent with the overall photo-
chemistry occurring as suggested by eq 14, with (bpy)Re(CO)3-
CH2Ph being the major photoproduct. The UV-visible ab-
sorption data indicate that the major photoproduct has a single
visible absorption band which is at a lower energy compared
to the MLCT absorption of (bpy)Re(CO)3(py)+. This is
inconsistent with the visible absorption of the dimer (bpy)-
(CO)3Re-Re(CO)3(bpy), which displays three visible absorption
bands (λmax ) 470, 600, and 805 nm).32 On the other hand,
the absorption data is consistent with that of (bpy)Re(CO)3-
CH2Ph, which we have previously demonstrated to feature a
MLCT absorption band with λmax ) 450 nm.33 The FTIR data
also rule out the dimer as a possible photoproduct, for this
species is reported to display four νCO bands.32 Unfortunately,
the IR spectrum of an authentic sample of (bpy)Re(CO)3-CH2-
Ph is unavailable; however, the fact that the νCO bands for the
photoproduct mixture appear at lower frequency compared to
those for (bpy)Re(CO)3(py)+ is consistent with -CH2Ph being
a stronger σ-donor ligand than py. Finally, the mass spectral
data are in accord with (bpy)Re(CO)3-CH2Ph being a major
photoproduct, since the prominent peak observed at m/e 517 is
clearly the parent ion of the benzyl complex and the m/e 427
peak corresponds to (bpy)Re(CO)3+, which is the most likely
daughter ion which would arise from fragmentation of (bpy)-
Re(CO)3-CH2Ph+•.
It is also useful to draw an analogy between the IPCT salts
examined herein and charge transfer complexes formed between
neutral organic electron/donor acceptor pairs in solution. Recent
detailed studies by Farid, Gould, and co-workers provide
remarkable detail concerning the structure and electronic proper-
ties of organic charge transfer complexes.35 From bandshape
analysis of the charge transfer absorption and (exciplex)
fluorescence of a series of complexes they have been able to
extract ø values for the charge transfer processes in a variety
of solvents. For charge transfer complexes formed from
cyanoaromatic acceptors paired with alkylbenzene donors in
moderately polar solvents such as CHCl3, ø values ranging from
0.5 to 0.8 eV are typical.32 Of interest is the fact that the ø
-
estimated from the IPCT absorption of ReBo and A+Co(CO)4
salts in THF is larger than that observed for the organic
complexes. A possible reason for the larger ø values observed
for the organometallic ion-pairs is that the anion/cation pairs in
the organometallic salts cannot approach as closely as the
aromatic donor/acceptor complexes which are likely to exist in
“sandwich” geometries to optimize the donor/acceptor electronic
interaction.
Discussion
Ion-Pair Charge Transfer Absorption. To a first ap-
proximation, the IPCT absorption band observed for THF
solutions of ReBo derives from a transition involving the
HOMO of the benzylborate donor and the LUMO of the Re(I)
acceptor. Since the HOMO of the borate is the boron-benzyl
carbon σ bonding orbital (i.e., σC-B),20 and the LUMO of the
Re(I) complex is a π antibonding orbital localized predominately
on bpy (i.e., π*bpy),21a in a one electron approximation the IPCT
transition is σC-B f π*bpy. Given the σC-B f π*bpy assignment,
Photochemical Mechanism. The luminescence and transient
absorption data reveal that photoexcitation of ReBo under
conditions where the salt exists as free ions or as ion-pairs leads
to photoreduction of the (bpy)ReI(CO)3(py)+ acceptor with unit
(34) Sullivan, B. P.; Dressick, W. J.; Meyer, T. J. J. Phys. Chem. 1982,
86, 1473.
(35) (a) Gould, I. R.; Moody, R.; Farid, S. J. Am. Chem. Soc. 1988, 110,
7242. (b) Gould, I. R.; Farid, S. J. Phys. Chem. 1992, 96, 7635. (c)
Gould, I. R.; Young, R. H.; Mueller, L. J.; Farid, S. J. Am. Chem.
Soc. 1994, 116, 8176. (d) Gould, I. R.; Young, R. H.; Mueller, L. J.;
Albrecht, A. C.; Farid, S. J. Am. Chem. Soc. 1994, 116, 8188.
(33) Lucia, L. A. M.S. Thesis, University of Florida, 1993.

Rhenium(I) Borate Salts
Inorganic Chemistry, Vol. 35, No. 23, 1996 6807
Scheme 1
quantum efficiency. Scheme 1 illustrates a number of possible
mechanistic pathways that could explain the experimental
observations. In this scheme a variety of “states” are labeled
with boldface numerals; for example the ground state free ions
and ion-pair are labeled 1 and 2, respectively.
lifetime apparatus used for these studies, it is only possible to
estimate a lower limit for k4f5 g 1010 s-1. However, it is
important to note that even with k4f5 ≈ 1010 s-1 the rate of
electron transfer is sufficiently fast such that geminate pair 5 is
formed with unit efficiency upon photoexcitation of ion-pair 2
(i.e., relaxation of 4 by radiative and/or nonradiative decay is
not competitive with electron transfer).
It is important to note that some intermediate “states” in the
photochemical mechanism may exist in one or more “forms”.
Specifically, two pathways are envisaged for formation of
encounter complex 4, which consists of the MLCT excited state
Re(I) cation in close proximity to the benzylborate anion. Thus,
4 forms either by diffusion of the free ions together, 3 f 4, or
by direct MLCT photoexcitation of the ground state ion-pair, 2
f 4. Although 4 has the same electronic configuration
regardless of the pathway, details such as solvation state, ion-
pair geometry, average separation distance, and donor-acceptor
electronic coupling may be path dependent. In a like manner
it is possible that geminate radical pair 5, which is produced
by electron transfer from the borate to the Re(I) complex, may
exist in different solvation states and average inter-radical
separation distance. For example, geminate pair 5, which is
formed by diffusional encounter followed by electron transfer
(3 f 4 f 5), very likely has a greater average separation
distance and weaker electronic coupling between the Re and
boranyl radicals compared to the geminate pair which is formed
by direct IPCT excitation (2 + hν f 5).35a
On the basis of the experimental data obtained for solutions
of ReBo in CH3CN and THF, some conclusions can be drawn
with regard to the pathways for formation of geminate radical
pair 5. First, in CH3CN static quenching of the Re f bpy
MLCT emission was not observed. This result indicates that
in this solvent the predominant path to the geminate radical pair
is diffusional encounter followed by electron transfer (3 f 4
f 5). By contrast, in THF solution ReBo exists predominantly
as ion pair 2. In this case, geminate radical pair 5 can be
produced either via MLCT excitation followed by electron
transfer (2 + hν f 4 f 5) or by direct IPCT excitation (2 +
hν f 5). Unfortunately, due to overlap of the IPCT and MLCT
transitions in ReBo, it is not possible to photoselect one of these
two pathways. However, since the oscillator strength of the
MLCT transition is 20- to 30-fold larger than that for the IPCT
transition, it is likely that for ReBo in THF the geminate pair
is formed almost exclusively via the MLCT state.
Three channels are available for decay of geminate radical
pair 5. The first is return electron transfer (5 f 2) and leads
back to starting materials without net photochemistry. The other
two channels both lead to irreversible photoreduction of (bpy)-
ReI(CO)3(py)+. One channel involves C-B bond fragmentation
within geminate radical pair 5 followed by cage escape (5 f 6
f 8), while the other is cage escape to form free radicals
followed by C-B bond fragmentation (5 f 7 f 8). The
quantitative transient absorption experiments indicate that
(within experimental uncertainty) irreversible photoreduction of
(bpy)ReI(CO)3(py)+ occurs with unit quantum efficiency when
corrected for the efficiency of MLCT quenching. This indicates
that return electron transfer (5 f 2) does not occur and that 5
decays exclusively by one (or both) of the other channels. The
fact that return electron transfer does not take place to a
measurable extent in 5 indicates that the competing processes
occur with (a sum of) rates that exceed the rate of return electron
transfer by at least a factor of 20. Previous studies of the cage
escape efficiency for bimolecular photoinduced electron transfer
in the (bpy)ReI(CO)3(py)+ system suggest that return electron
transfer in 5 will occur with a rate in excess of 5 × 108 s-1 21b
.
If this is correct, then the experimental observation that return
electron transfer cannot compete with bond fragmentation and/
or cage escape implies that the latter processes occur with a
sum of rates that exceed 1010 s-1
. We conclude that the
dominant pathway for decay of geminate pair 5 is C-B bond
fragmentation within geminate radical pair 5 followed by cage
escape (5 f 6 f 8). This conclusion is based on observations
by Schuster and co-workers which indicate that C-B bond
•
fragmentation in boranyl radicals of the type R-BPh3 (R )
alkyl or benzyl) occurs at a rate in excess of 1011 s-1 20
The
.
results of the present study are consistent with these findings,
in that return electron transfer within geminate pair 5, although
expected to be competitive with cage escape, clearly cannot
compete with the exceedingly fast bond fragmentation process.
A significant point concerns the rate of electron transfer within
encounter complex 4 (i.e., k4f5). In the luminesence lifetime
studies of ReBo in THF a short-lived decay component that
could be ascribed to 4 was not observed. Given the relatively
Summary and Conclusions
The luminescence and transient absorption studies of ReBo
indicate that the Re f bpy MLCT excited state is efficiently
quenched by electron transfer from the benzylborate anion.
Regardless of whether MLCT quenching occurs via a diffusional
low radiative rate of the Re f bpy MLCT state (kr ≈ 105 s-1 21a
)
and the 500 ps time resolution of the single photon counting

6808 Inorganic Chemistry, Vol. 35, No. 23, 1996
McCosar and Schanze
pathway or within a preformed ion-pair, return electron transfer
within the geminate radical pair formed by electron transfer does
not occur and therefore irreversible photoreduction of the
acceptor cation occurs with unit quantum efficiency. Return
electron transfer is precluded by ultrafast dissociation of the
C-B bond in the boranyl radical.
Unfortunately, with the ReBo system it is not possible to
use photoselection to compare the quantum efficency for
photoreduction for IPCT and MLCT excitation of the ion-pair
in THF, because the two optical transitions occur in the same
wavelength range. Interest in this point is motivated by our
recent observations concerning cage escape yields in ion-pairs
of the type [(bpy)ReI(CO)3(py)+][Co(CO)4-],36 which imply that
return electron transfer is much more rapid in geminate pairs
formed by IPCT excitation compared with those formed by
excitation of the Re f bpy MLCT state in the ion-pair. Thus,
current work on Re-borate ion-pairs is directed to preparation
of ion-pairs in which the IPCT absorption is clearly red-shifted
from the Re f bpy MLCT transition.
Acknowledgment. We gratefully acknowledge support for
this project from the National Science Foundation (Grant No.
CHE 94-01620). B.H.M. holds an NSF Minority Fellowship
(1994-1997).
IC9605074
(36) Lucia, L. A. Ph.D. Dissertation, University of Florida, 1996.