Control of photochemical, photophysical, electrochemical, and photocatalytic properties of rhenium(I) complexes using intramolecular weak interactions between ligands

Article abstract of DOI:10.1021/ja053814u

Intramolecular interactions between ligands have been successfully applied as a novel tool for controlling various properties of a series of cis,trans-[Re(dmb)(CO)₂(PR3)(PR′3)]⁺-type complexes (dmb = 4,4′-dimethyl-2,2′-bipyridine), in the ground state and in the excited state and in the one-electron reduced form. For rhenium complexes with two triarylphosphine ligands, P(p-XPh)₃, the dmb ligand was sandwiched by four aryl rings having CH(aryl)-π(pyridine)-π(aryl) interactions. On the other hand, complexes with one triarylphosphine ligand and one trialkylphosphite ligand, P(OR)₃, had π-π and CH-π interactions between each pyridine ring in the dmb ligand and the aryl group in the P(p-XPh)₃. Various properties of these two series of rhenium complexes were compared with those of complexes having two trialkylphosphite ligands, which do not interact through space with the dmb ligand. Properties of the complexes associated mainly with the dmb ligand are strongly affected by the intramolecular interactions: (1) UV/vis absorptions to the π-π* and ¹MLCT excited states were both red-shifted, but (2) emission from the ³MLCT excited state was blue-shifted; (3) the lifetime of the ³MLCT excited state was prolonged up to 3-fold; (4) the reduction potential in the ground state was positively shifted by 110 mV with π-π and CH-π interactions and by 180-200 mV with the CH-π-π interactions. (5) In the excited states, the oxidation power of the complex was also enhanced by the intramolecular interactions. (6) In the corresponding one-electron-reduced species cis,trans-[Re(dmb^-.)(CO) ₂(PR₃)(PR′₃)], the intramolecular interactions are maintained and strongly affected their UV/vis spectra. (7) Photocatalysis for CO₂ reduction was significantly enhanced only by the CH-π-π interaction.

Full text of DOI:10.1021/ja053814u

Published on Web 10/15/2005
Control of Photochemical, Photophysical, Electrochemical,
and Photocatalytic Properties of Rhenium(I) Complexes Using
Intramolecular Weak Interactions between Ligands
Hideaki Tsubaki,† Akiko Sekine,‡ Yuji Ohashi,‡ Kazuhide Koike,§
Hiroyuki Takeda,† and Osamu Ishitani*,†
Contribution from the Department of Chemistry and Department of Chemistry and Materials
Science, Graduate School of Science and Engineering, Tokyo Institute of Technology,
Ookayama 2-12-1, Meguro-ku, Tokyo, 152-8551, Japan, and National Institute of AdVanced
Industrial Science and Technology, Onogawa 16-1, Tsukuba 305-8569, Japan
Received June 10, 2005; E-mail: ishitani@chem.titech.ac.jp
Abstract: Intramolecular interactions between ligands have been successfully applied as a novel tool for
controlling various properties of a series of cis,trans-[Re(dmb)(CO)2(PR3)(PR′3)]+-type complexes (dmb )
4,4′-dimethyl-2,2′-bipyridine), in the ground state and in the excited state and in the one-electron reduced
form. For rhenium complexes with two triarylphosphine ligands, P(p-XPh)3, the dmb ligand was sandwiched
by four aryl rings having CH(aryl)-π(pyridine)-π(aryl) interactions. On the other hand, complexes with
one triarylphosphine ligand and one trialkylphosphite ligand, P(OR)3, had π-π and CH-π interactions
between each pyridine ring in the dmb ligand and the aryl group in the P(p-XPh)3. Various properties of
these two series of rhenium complexes were compared with those of complexes having two trialkylphosphite
ligands, which do not interact through space with the dmb ligand. Properties of the complexes associated
mainly with the dmb ligand are strongly affected by the intramolecular interactions: (1) UV/vis absorptions
1
to the π-π* and MLCT excited states were both red-shifted, but (2) emission from the 3MLCT excited
state was blue-shifted; (3) the lifetime of the 3MLCT excited state was prolonged up to 3-fold; (4) the reduction
potential in the ground state was positively shifted by 110 mV with π-π and CH-π interactions and by
180-200 mV with the CH-π-π interactions. (5) In the excited states, the oxidation power of the complex
was also enhanced by the intramolecular interactions. (6) In the corresponding one-electron-reduced species
cis,trans-[Re(dmb-•)(CO)2(PR3)(PR′3)], the intramolecular interactions are maintained and strongly affected
their UV/vis spectra. (7) Photocatalysis for CO2 reduction was significantly enhanced only by the CH-π-π
interaction.
Introduction
complexes including Ru(II), Ir(III), Pt(II), and Re(I) have been
studied intensively in applications as a light absorber, which
has been sought to absorb a wider energy range of solar
light, and also as an emitter, which must have high quantum
yield of emission, in light-harvesting8 and electroluminescent
devices.6
Control of photochemical, photophysical, and electrochemical
properties of transition metal complexes is of great interest
because of the advantages of these complexes in fields of
chemistry as diverse as photo- and electrocatalysts,1-3 pho-
tonic sensors,4 photochemical energy and electron transfer,1,5
chemi- and electroluminescence,6 molecular electronics, and
photonics.5b,7 In the past two decades, d6 transition metal diimine
(5) (a) Wasielewski, M. R. Chem. ReV. 1992, 92, 435-461. (b) Balzani, V.;
Juris, A.; Venturi, M.; Campagna, S.; Serroni, S. Chem. ReV. 1996, 96,
759-833. (c) De Cola, L.; Belser, P. Coord. Chem. ReV. 1998, 177, 301-
346. (d) Vlcek, A. Coord. Chem. ReV. 2000, 200-202, 933-978. (e)
Balzani, V.; Juris, A. Coord. Chem. ReV. 2001, 211, 97-115.
(6) (a) Baldo, M. A.; O’Brien, D. F.; You, Y.; Shoustikov, A.; Sibley, S.;
Thompson, M. E.; Forrest, S. R. Nature 1998, 395, 151-154. (b) Handy,
E. S.; Pal, A. J.; Rubner, M. F. J. Am. Chem. Soc. 1999, 121, 3525. (c)
Gao, F. G.; Bard, A. J. J. Am. Chem. Soc. 2000, 122, 7426. (d) Lamansky,
S.; Djurovich, P.; Murphy, D.; Abdel-Razzaq, F.; Lee, H.-E.; Adachi, C.;
Burrows, P. E.; Forrest, S. R.; Thompson, M. E. J. Am. Chem. Soc. 2001,
123, 4304-4312. (e) Carlson, B.; Phelan, G. D.; Kaminsky, W.; Dalton,
L.; Jiang, X.; Liu, S.; Jen, A. K. Y. J. Am. Chem. Soc. 2002, 124, 14162-
14172. (f) Welter, S.; Brunner, K.; Hofstraat, J. W.; De Cola, L. Nature
2003, 421, 54-57.
(7) (a) Harriman, A.; Ziessel, R. Coord. Chem. ReV. 1998, 171, 331-339. (b)
Barigelletti, F.; Flamigni, L. Chem. Soc. ReV. 2000, 29, 1-12.
(8) (a) Anderson, P. A.; Keene, F. R.; Meyer, T. J.; Moss, J. A.; Strouse, G.
F.; Treadway, J. A. J. Chem. Soc., Dalton Trans. 2002, 3820-3831. (b)
Molnar, S. M.; Nallas, G.; Bridgewater, J. S.; Brewer, K. J. J. Am. Chem.
Soc. 1994, 116, 5206-5210. (c) Hagfeldt, A.; Gra¨etzel, M. Acc. Chem.
Res. 2000, 33, 269-277.
† Department of Chemistry, Tokyo Institute of Technology.
‡ Department of Chemistry and Materials Science, Tokyo Institute of
Technology.
§ National Institute of Advanced Industrial Science and Technology.
(1) Photosensitization and Photocatalysis Using Inorganic and Organometallic
Compounds; Kalyanasundaram, K., Graetzel, M., Eds; Kluwer Academic
Publishers: Boston, 1993.
(2) (a) Fujita, E.; Brunschwig, B. S.; Ogata, T.; Yanagida, S. Coord. Chem.
ReV. 1994, 132, 195-200. (b) Sutin, N.; Creutz, C.; Fujita, E. Comments
Inorg. Chem. 1997, 19, 67-92.
(3) (a) Johnson, F. P. A.; George, M. W.; Hartle, F.; Turner, J. J. Organome-
tallics 1996, 15, 3374. (b) Christensen, P.; Hamnett, A.; Muir, A. V. G.;
Timney, J. A. J. Chem. Soc., Dalton Trans. 1992, 1455.
(4) (a) Demas, J. N.; DeGraff, B. A. Coord. Chem. ReV. 2001, 211, 317-351.
(b) Kunugi, Y.; Mann, K. R.; Miller, L. L.; Exstrom, C. L. J. Am. Chem.
Soc. 1998, 120, 589-590. (c) Yam, V. W. W.; Wong, K. M.-C.; Zhu, N.
J. Am. Chem. Soc. 2002, 124, 6506-6507. (d) Zhan, W.; Alvarez, J.;
Crooks, R. M. J. Am. Chem. Soc. 2002, 124, 13265-13270.
9
15544
J. AM. CHEM. SOC. 2005, 127, 15544-15555
10.1021/ja053814u CCC: $30.25 © 2005 American Chemical Society

Ligand Interactions to Control Re(I) Complex Properties
A R T I C L E S
Table 1. Comparison of the Properties among 0+, 2+, 3+, 1e+,
1h+, and 1j+
Chart 1a
+
[Re(X2bpy)(CO)2LL
′]
a
b
c
redd
λabs
λem
τem
E1
/2
e
X
L
L
′
nm
nm
ns
V
TNCO
0+f
2+
H
CO
CF3 CO
py
P(OEt)3
P(OEt)3
P(OEt)3
351 542 1033 -1.59
5.9
0.1
363 623
397 651
162 -1.19
3+ g
H
12 (-1.64)h ∼0
1e+ Me P(p-FPh)3 P(p-FPh)3 396 595 1046 -1.73
17.3
0.5
0.2
1h+ Me PPh3
1j+ Me P(O-iPr)3 P(O-iPr)3 366 605
P(OEt)3
382 606
511 -1.81
381 -1.86
a 1MLCT absorption maximum measured in MeCN. b Emission maxi-
mum measured in MeCN. c Lifetime of emission measured in MeCN.
d Redox potential (X2bpy/X2bpy-•) vs Ag/AgNO3. See General Procedures
in Experimental Section. e Turnover number of CO formation by reduction
of CO2 using the rhenium complex as photocatalyst. See Photocatalytic
Reaction in Experimental Section. f Reference 19. g Reference 12. h Peak
potential for the irreversible process.
Two methods are available for changing the properties of
these complexes: (i) use of a ligand with a different ligand-
field strength, to adjust the energy levels of the dπ orbitals in
the central metal and (ii) introduction of functional groups with
different electro-donating or -accepting properties to the diimine
ligand, to adjust the π* orbital energy level, (these orbitals are
related to useful photochemical, photophysical, and electro-
chemical events).
As typical examples, rhenium diimine complexes, which are
useful as CO2-reduction photocatalysts,9,10 electroluminescent
devices11 and so on, are shown in Table 1. Although the complex
fac-[Re(bpy)(CO)3{P(OEt)3}]+ (0+; bpy ) 2,2′-bipyridine) emits
strongly with high quantum yield even in solution at room
temperature, and can act as an efficient photocatalyst for CO2
reduction,10a its absorption in the visible region is very limited,12
so that it must be modified to allow effective use of solar energy.
Introduction of either trifluoromethyl groups to the bpy ligands
(2+),10b or a pyridine ligand having a weaker ligand field than
that of the P(OEt)3 ligand (3+),13 red-shifts the 1MLCT
absorption band by 12-46 nm (see Table 1). However, these
changes also lessen their photocatalytic activities, because of
a Sum of Tolman’s ø values of each phosphorus ligand (from ref 23).
The ø represents electron-attracting ability of the phosphorus ligand.
3
blue-shift the emission from MLCT, increase the lifetime of
the 3MLCT excited state, and provide stronger oxidation power
in both the ground state and excited states, stronger reduction
power of the one-electron-reduced (OER) complex, and high
photocatalytic activity for CO2 reduction. A series of rhenium-
(I) biscarbonyl complexes with two phosphorus ligands at the
cis-positions to the 4,4′-dimethyl-2,2′-bipyridine(dmb) ligand
(shown in Chart 1) were synthesized: both phosphorus ligands
may have π-π and/or CH-π interaction(s) with the dmb ligand.
3
the shorter lifetimes of the MLCT excited state, which must
Experimental Section
be quenched by a reductant, and weaker reduction power of
the corresponding reduced complex as an important intermedi-
ate.13 How can the conflicting effects of these perturbations on
the properties of metal complexes be canceled and controlled?
We report here a new way to manipulate the properties of
metal diimine complexes using intramolecular interactions
General Procedures. IR spectra were recorded with a JEOL JIR-
6500 FTIR spectrophotometer using 1 cm-1 resolution. UV/vis spectra
were measured with a JASCO V-565 or Photal MCPD-2000 spectro-
photometer. Emission spectra were recorded at 25 °C using a JASCO
FP-6600 spectrofluorometer with correction for the detector sensitivity
determined using correction data supplied by JASCO. Emission
quantum yields were evaluated with quinine bisulfate (Φem ) 0.546)14
as a standard. Emission lifetimes were measured either with a Horiba
NAES-1100 time-correlated single-photon-counting system (the excita-
tion source was a nanosecond H2 lamp, NFL-111, and the instrumental
response time was less than 1 ns) or with a Continuum YG680-10
Nd:YAG pulse laser source (third-harmonic generation at 355 nm, 10
ns fwhm, 10 mJ/pulse) and Hamamatsu Photonics R926 photomultiplier
tube on a Jobin-Yvon HR-320 monochromator.15 Proton-NMR spectra
were measured in an acetone-d6 solution using a Bruker AC300P (300
MHz) system. Residual protons of acetone-d6 were used as an internal
standard for the measurements. Cyclic voltammograms of the complexes
were measured in an acetonitrile solution containing tetra-n-butylam-
1
between ligands, which red-shift the MLCT absorption but
(9) Hawecker, J.; Lehn, J. M.; Ziessel, R. HelV. Chim. Acta 1986, 69, 1990-
2012.
(10) (a) Hori, H.; Johnson, F. P. A.; Koike, K.; Ishitani, O.; Ibusuki, T. J.
Photochem. Photobiol., A: Chem. 1996, 96, 171-174. (b) Koike, K.; Hori,
H.; Ishizuka, M.; Westwell, J. R.; Takeuchi, K.; Ibusuki, T.; Enjouji, K.;
Konno, H.; Sakamoto, K.; Ishitani, O. Organometallics 1997, 16, 5724-
5729. (c) Hori, H.; Johnson, F. P. A.; Koike, K.; Takeuchi, K.; Ibusuki,
T.; Ishitani, O. J. Chem. Soc., Dalton Trans. 1997, 1019-1023.
(11) (a) Lundin, N. J.; Walsh, P. J.; Howell, S. L.; McGarvey, J. J.; Blackman,
A. G.; Gordon, K. C. Inorg. Chem. 2005, 44, 3551-3560. (b) Yam, V.
W.-W.; Li, B.; Yang, Y.; Chu, B. W.-K.; Wong, K. M.-C.; Cheung, K.-K.
Eur. J. Inorg. Chem. 2003, 4035-4042. (c) Ng, P. K.; Gong, X.; Chan, S.
K.; Lam, L. S. M.; Chan, W. K. Chem.sEur. J. 2001, 7, 4358-4367. (d)
Li, F.; Zhang, M.; Cheng, G.; Feng, J.; Zhao, Y.; Ma, Y.; Liu, S.; Shen, J.
Appl. Phys. Lett. 2004, 84, 148-150. (e) Li, F.; Zhang, M.; Feng, J.; Cheng,
G.; Wu, Z.; Ma, Y.; Liu, S.; Sheng, J.; Lee, S. T. Appl. Phys. Lett 2003,
83, 365-367.
(14) Melhuish, W. H. J. Phys. Chem. 1961, 65, 229.
(12) Hori, H.; Koike, K.; Ishizuka, M.; Takeuchi, K.; Ibusuki, T.; Ishitani, O.
J. Organomet. Chem. 1997, 530, 169-176.
(13) Koike, K.; Tanabe, J.; Toyama, S.; Tsubaki, H.; Sakamoto, K.; Westwell,
J. R.; Johnson, F. P. A.; Hori, H.; Saitoh, H.; Ishitani, O. Inorg. Chem.
2000, 39, 2777-2783.
(15) (a) Koike, K.; Okoshi, N.; Hori, H.; Takeuchi, K.; Ishitani, O.; Tsubaki,
H.; Clark, I. P.; George, M. W.; Johnson, F. P. A.; Turner, J. J. J. Am.
Chem. Soc. 2002, 124, 11448-11455. (b) Gholamkhass, B.; Koike, K.;
Negishi, N.; Hori, H.; Sano, T.; Takeuchi, K. Inorg. Chem. 2003, 42, 2919-
2932.
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15545

A R T I C L E S
Tsubaki et al.
-
monium tetrafluoroborate (0.1 M) as supporting electrolyte using an
ALS/CHI620 electrochemical analyzer with a glassy-carbon disk
working electrode (3 mm diameter), a Ag/AgNO3 (0.1 M) reference
electrode, and a Pt counter electrode. The supporting electrolyte was
dried in vacuo at 100 °C for 1 day prior to use.
give the CF3SO3 salts of 1d+ as a yellow solid, which was
recrystallized using ether-CH2Cl2. The PF6- salts of 1d+ were obtained
in a manner similar to that for 1a+.
Yield: 83%. Anal. Calcd for C50H39N2O2F9P3Re: C, 52.22; H, 3.42;
N, 2.44. Found: C, 52.18; H, 3.31; N, 2.38. H NMR (δ, 300 MHz,
1
Materials. Dimethylformamide (DMF) and triethanolamine (TEOA)
were distilled under a reduced pressure from 4Å molecular sieves and
KOH, respectively. Acetonitrile was dried three times over P2O5 and
then distilled from CaH2 prior to use. Tetra-n-butylammonium tet-
rafluoroborate was triply recrystallized from ethyl acetate/benzene.
Other reagents and solvents were purchased from Kanto Chemical Co.,
Junsei Chemical Co., Tokyo Kasei Co., Wako Pure Chemical Industries,
and Aldrich Chemical Co. and used without further purification.
Synthesis Procedures. The CF3SO3 and PF6 salts of fac-[Re-
(dmb)(CO)3(PR3)]+ (R ) p-MeOPh, Ph, p-FPh, p-ClPh, O-iPr, OEt)
and the PF6- salts of cis,trans-[Re(dmb)(CO)2(PPh3)2]+ (1c+), cis,trans-
[Re(dmb)(CO)2{P(p-FPh)3}2]+ (1e+), cis,trans-[Re(dmb)(CO)2(PPh3)-
{P(OEt)3}]+ (1h+), cis,trans-[Re(dmb)(CO)2{P(O-iPr)3}2]+ (1j+), and
cis,trans-[Re(dmb)(CO)2{P(OEt)3}2]+ (1k+) were synthesized according
to previously reported methods.12,13,15a Similar methods were used for
the synthesis of the following complexes. The reaction solutions were
cooled with water during irradiation in all cases.
CD3COCD3): 8.22 (s, 2H, bpy-3,3′), 7.94 (d, 2H, J ) 5.4 Hz, bpy-
6,6′), 6.91 (d, 2H, J ) 5.4 Hz, bpy-5,5′), 7.20-7.40 (m, 21H, Ph,
p-FPh), 7.00-7.20 (m, 6H, p-FPh), 2.45 (s, 6H, CH3-bpy). FT-IR
(CH3CN): νCO/cm-1 ) 1937, 1867. UV/vis (CH3CN): λmax/nm (ꢀ/103
M-1 cm-1) ) 400 (2.7), 313 (sh), 293 (17), 275 (17), 223 (2h).
cis,trans-[Re(dmb)(CO)2{P(p-ClPh)3}2]+PF6- (1f+PF6-) was syn-
thesized in a manner similar to that for the synthesis of 1d+PF6
.
-
Yield: 79%. Anal. Calcd for C50H36N2O2Cl6F6P3Re: C, 46.10; H,
-
-
1
2.79; N, 2.15. Found: C, 46.33; H, 2.90; N, 2.12. H NMR (δ, 300
MHz, CD3COCD3): 8.23 (s, 2H, bpy-3,3′), 8.07 (d, 2H, J ) 5.6 Hz,
bpy-6,6′), 7.03 (d, 2H, J ) 5.6 Hz, bpy-5,5′), 7.25-7.45 (m, 24H,
p-ClPh), 2.49 (s, 6H, CH3-bpy). FT-IR (CH3CN): νCO/cm-1 ) 1942,
1872. UV/vis (CH3CN): λmax/nm (ꢀ/103 M-1 cm-1) ) 395 (2.6), 315
(sh), 296 (19), 273 (19), 224 (sh), 232 (22).
cis,trans-[Re(dmb)(CO)2{P(p-MeOPh)3}{P(O-iPr)3}]+PF6
-
-
(1g+PF6-). A THF solution (30 mL) containing 105 mg (0.17 mmol)
of fac-[Re(dmb)(CO)3(CF3SO3)] and 1 mL of triisopropyl phosphite
was refluxed for 24 h under an argon atmosphere, and then 183 mg
(0.52 mmol) of tris(4-methoxyphenyl)phosphine was introduced into
the solution. This solution was irradiated for 1 h under an argon atmos-
phere. The light exposed solution was evaporated and then chromato-
graphed on silica gel using CH2Cl2 as eluent. The yellow layer was
cis,trans-[Re(dmb)(CO)2{P(p-MeOPh)3}2]+PF6- (1a+PF6-). A THF
solution (280 mL) containing 106 mg (0.11 mmol) of fac-[Re(dmb)-
(CO)3{P(p-MeOPh)3}]+CF3SO3 and 455 mg (1.30 mmol) of tris(4-
-
methoxyphenyl)phosphine was irradiated under an argon atmosphere
using a high-pressure mercury lamp with an uranyl glass filter (>330
nm) for 1 h. After irradiation the solvent was evaporated under reduced
pressure, and the residual orange solid was washed with ether and
recrystallized from CH2Cl2-ether. To a 2 mL methanolic solution of
-
collected and evaporated under reduced pressure to give the CF3SO3
salts of 1g+ as a yellow solid, which was recrystallized using ether-
CH2Cl2. The PF6- salt of 1g+ was obtained in a manner similar to that
for 1a+.
-
this solid, a concentrated NH4+PF6 methanolic solution was added
dropwise. The precipitated PF6- salts of 1a+ were collected by filtration,
washed with water, and then dried in vacuo.
Yield: 45%. Anal. Calcd for C44H54N2O8F6P3Re: C, 46.68; H, 4.81;
1
N, 2.48. Found: C, 46.89; H, 4.79; N, 2.45. H NMR (δ, 300 MHz,
Yield: 69%. Anal. Calcd for C56H54N2O8F6P3Re: C, 52.71; H, 4.27;
N, 2.20. Found: C, 52.64; H, 4.80; N, 2.05. H NMR (δ, 300 MHz,
CD3COCD3): 8.48 (d, 2H, J ) 5.9 Hz, bpy-6,6′), 8.37 (s, 2H, bpy-
3,3′), 7.33 (d, 2H, J ) 5.9 Hz, bpy-5,5′), 7.13-7.23 (m, 6H, p-MeOPh),
6.82-6.89 (m, 6H, p-MeOPh), 4.50-4.67 (m, 3H, POCH(CH3)2), 3.81
(s, 9H, CH3O-Ph), 2.55 (s, 6H, CH3-bpy), 1.03 (d, 18H, J ) 5.9 Hz,
POCH(CH3)2). FT-IR (CH3CN):
(CH3CN): λmax/nm (ꢀ/103 M-1 cm-1) ) 388 (3.3), 312 (sh), 285 (sh),
246 (41).
1
CD3COCD3): 8.16 (s, 2H, bpy-3,3′), 7.93 (d, 2H, J ) 5.8 Hz, bpy-
6,6′), 6.89 (d, 2H, J ) 5.8 Hz, bpy-5,5′), 7.00-7.25 (m, 12H, Ph-m),
6.66-6.90 (m, 12H, Ph-o), 3.79 (s, 18H, CH3O-Ph), 2.45 (s, 6H, CH3-
bpy). FT-IR (CH3CN): νCO/cm-1 ) 1930, 1859. UV/vis (CH3CN):
ν
CO/cm-1 ) 1939, 1865. UV/vis
λ
max/nm (ꢀ/103 M-1 cm-1) ) 415 (2.6), 291 (sh), 273 (sh), 244 (71).
-
cis,trans-[Re(dmb)(CO)2{P(p-MeOPh)3}(PPh3)]+PF6- (1b+PF6
)
cis,trans-[Re(dmb)(CO)2{P(p-FPh)3}{P(OMe)3}]+PF6- (1i+PF6-).
The acetonitrile complex cis,trans-[Re(dmb)(CO)2(MeCN){P(p-
FPh)3}]+PF6- was synthesized by irradiation to an acetonitrile solution
(250 mL) containing 180 mg (0.203 mmol) of fac-[Re(dmb)(CO)3{P(p-
FPh)3}]+PF6- under an argon atmosphere for 1 h, and then the solvent
was evaporated under reduced pressure. The resulting yellow solid was
dissolved into 50 mL of a THF solution containing 1 mL of trimethyl
phosphite and refluxed for 36 h under an argon atmosphere. After
evaporation of the solution, the residue was purified with column
chromatography on silica gel using CH2Cl2 as eluent. The orange layer
was obtained by irradiation of a THF solution (200 mL) containing 85
mg (0.098 mmol) of fac-[Re(dmb)(CO)3(PPh3)]+CF3SO3- and 151 mg
(0.43 mmol) of tris(4-methoxyphenyl)phosphine under an argon
atmosphere for 3 h. After evaporation of the solution, the residue was
purified with column chromatography on silica gel using CH2Cl2 as
eluent. The yellow layer was collected and evaporated under reduced
-
pressure to give the CF3SO3 salts of 1b+ as a yellow solid, which
-
were recrystallized using ether-CH2Cl2. The PF6 salts of 1b+ were
obtained using a procedure analogous to that given for 1a+. Elemental
analysis data were obtained using the CF3SO3 salts of 1b+.
-
was collected and evaporated under reduced pressure to give 1i+PF6
-
as yellow solid, which was recrystallized using ether-CH2Cl2.
Yield: 20%. Anal. Calcd for C35H33N2O5F9P3Re: C, 41.55; H, 3.29;
N, 2.77. Found: C, 41.98; H, 3.26; N, 2.85. H NMR (δ, 300 MHz,
CD3COCD3): 8.54 (d, 2H, J ) 5.9 Hz, bpy-6,6′), 8.44 (s, 2H, bpy-
3,3′), 7.27-7.38 (m, 2H, bpy-5,5′), 7.27-7.38 (m, 6H, p-FPh), 7.11-
7.18 (m, 6H, p-FPh), 3.50 (d, 9H, JH,P ) 11.0 Hz, POCH3), 2.55 (s,
6H, CH3-bpy). FT-IR (CH3CN): νCO/cm-1 ) 1950, 1877. UV/vis
(CH3CN): λmax/nm (ꢀ/103 M-1 cm-1) ) 379 (3.8), 311 (sh), 287 (18),
267 (20), 257 (sh), 232 (26).
Yield: 38%. Anal. Calcd for C52H48N2O6F3P2ReS: C, 55.07; H, 4.27;
N, 2.47. Found: C, 54.81; H, 4.51; N, 2.37. H NMR (δ, 300 MHz,
1
1
CD3COCD3): 8.18 (s, 2H, bpy-3,3′), 7.91 (d, 2H, J ) 5.9 Hz, bpy-
6,6′), 6.86 (d, 2H, J ) 5.9 Hz, bpy-5,5′), 7.25-7.40 (m, 15H, Ph),
7.10-7.25 (m, 6H, p-MeOPh), 6.75-6.90 (m, 6H, p-MeOPh), 3.79
(s, 9H, CH3O-Ph), 2.44 (s, 6H, CH3-bpy). FT-IR (CH3CN): νCO/cm-1
) 1933, 1861. UV/vis (CH3CN): λmax/nm (ꢀ/103 M-1 cm-1) ) 411
(2.4), 292 (sh), 271 (sh), 238 (44).
cis,trans-[Re(dmb)(CO)2(PPh3){P(p-FPh)3}2]+PF6- (1d+PF6-) was
obtained by irradiation of a THF solution (250 mL) containing 80 mg
cis,trans-[Re(dmb)(CO)2{P(OEt)3}{P(OMe)3}]+PF6- (1l+PF6-) was
obtained by irradiation to a THF solution (90 mL) containing 50 mg
(0.087 mmol) of fac-[Re(dmb)(CO)3{P(p-FPh)3}]+CF3SO3 and 115
-
-
mg (0.44 mmol) of triphenylphosphine under an argon atmosphere for
2 h. After evaporation of the solution, the residue was purified with
column chromatography on silica gel using CH2Cl2 as eluent. The
orange layer was collected and evaporated under reduced pressure to
(0.065 mmol) of fac-[Re(dmb)(CO)3{P(OMe)3}]+PF6 and 2 mL of
triethylphosphite under an argon atmosphere for 1 h. After evaporation
of the solution, the residual yellow solid was washed with ether and
then recrystallized from CH2Cl2-ether.
9
15546 J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005

Ligand Interactions to Control Re(I) Complex Properties
A R T I C L E S
-
Table 2. Crystallographic Data for 1e+CF3SO3
empirical formula
C51H36F9N2O5P2ReS
formula weight
crystal system
space group
a/Å
b/Å
c/Å
1208.0
triclinic
P1h
15.5782(16)
13.5247(11)
13.0531(9)
116.249(3)
79.374(6)
95.966(4)
2423.5(4)
2
R/deg
â/deg
γ/deg
V/Å3
Z
T/K
253
Rint
0.0435
23 566
642
0.0339
0.0824
1.066
number of total reflections
number of parameters
R1[I > 2σ(I)]a
wR2[I > 2σ(I)]b
GOFc on F2
a R1 ) Σ(||FO| - |FC||)/Σ|FO|. b wR2 ) [Σ[w(FO2 - FC2)2]/Σ[w(FO2)2]]1/2
.
c GOF ) [Σw(|FO | - |FC |)2/ (m - n)]1/2, where m ) number of reflections
2
2
Figure 1. UV/vis absorption (solid line) and emission spectra (dotted line)
of 1e+ measured in acetonitrile solution at 298 K.
and n ) number of parameters.
Yield: 87%. Anal. Calcd for C23H36N2O8F6P3Re: C, 32.06; H, 4.21;
N, 3.25. Found: C, 32.53; H, 4.11; N, 3.18. H NMR (δ, 300 MHz,
CD3COCD3): 8.92 (d, 2H, J ) 5.9 Hz, bpy-6,6′), 8.59 (s, 2H, bpy-
at 365 nm using a 500 W high-pressure mercury lamp with both band-
pass and ND filters (Asahi Bunko Co.). A gas sample was taken using
a gastight syringe, and the reaction products, i.e., CO and H2, were
measured using a Shimazu GC-9A gas chromatograph with a TCD
detector. The light intensity incident into the cell was 8.22 × 10-9
einstein s-1 for 1a+, 1c+, 1e+, 1f+, 1h+, and 1j+, and 5.24 × 10-9
einstein s-1 for 1b+, 1d+, 1g+, 1i+, 1k+, and 1l+, determined using a
K3Fe(C2O4)3 actinometer.
1
3,3′), 7.62 (d, 2H, J ) 5.9 Hz, bpy-5,5′), 3.86 (dq, 6H, JH,H ) JH,P
)
7.0 Hz, POCH2CH3), 3.48 (dd, 9H, JH,P ) 10.3, 0.7 Hz, POCH3), 2.63
(s, 6H, CH3-bpy), 1.02 (t, 9H, J ) 7.0 Hz, POCH2CH3). FT-IR
(CH3CN): νCO/cm-1 ) 1956, 1882. UV/vis (CH3CN): λmax/nm (ꢀ/103
M-1 cm-1) ) 363 (4.2), 309 (sh), 281 (16), 264 (18), 256 (17).
Crystal Structure Determination. The single crystals of 1e+CF3SO3-
were obtained by slow diffusion of diethyl ether into a dichloromethane
solution containing the complex. Diffraction data of 1e+CF3SO3- were
collected on a Rigaku RAPID imaging-plate diffractometer with
graphite-monochromated Mo KR radiation (λ ) 0.710 73 Å) using
yellow needle crystals, 0.4 × 0.1 × 0.1 mm3 at 253 K with a Rigaku
low-temperature apparatus (-20 °C). The intensity data were corrected
for Lorentz and polarization effects. The structure was solved by the
direct methods using the TEXSAN program16 and was refined on F2
by means of full-matrix least-squares procedures, using the SHELXL-
97 program.17 No secondary extinction corrections were applied. In
the least-squares refinements, all non-hydrogen atoms were refined with
anisotropic displacement parameters, but all hydrogen atoms were
refined using the riding model isotropic thermal parameters which are
1.2 times those of the attached carbon atoms. Details of the crystal-
lographic data are given in Table 2.
Results
UV/vis Absorption, Emission, and IR Spectra. Figure 1
shows typical UV/vis absorption and emission spectra of 1e+.
A metal-to-ligand (dmb) charge transfer (MLCT) absorption
band was observed with an absorption maximum at 396 nm;
the intense and relatively sharp band around 292 nm is
attributable to dmb-localized π-π* absorption.13,18a All com-
plexes strongly emit from their 3MLCT excited states in
acetonitrile solution at room temperature, and these emission
bands were unstructured and broad, similar to those in Figure
1. Table 3 summarizes the absorption and emission maxima,
the apparent stokes shifts (∆Eabs-em), the molar extinction
coefficients, the emission lifetimes (τem), and the emission
quantum yields (Φem) with radiative and nonradiative decay rate
constants kr and knr calculated using the following equations,
since all complexes investigated in this work were stable against
irradiation at 400 nm.
Quenching Experiments. Reductive-quenching experiments of the
excited complexes were carried out using TEOA as an electron donor.
At least four samples for each complex were prepared as DMF
solutions, with various TEOA concentrations between 0 and 0.6 M.
Stern-Volmer plots of the observed decrease in emission intensity with
increasing of TEOA concentration showed a good linear relation with
an intercept of 1, so that the reductive-quenching rate constant kq can
be calculated using eq 1 and the emission lifetime of the complex in
kr ) Φem/τem
(6)
(7)
knr ) (1 - Φem)/τem
the absence of TEOA τem
.
In IR spectra of all the complexes we observed two CO
stretching bands, attributable to symmetric and antisymmetric
vibrations in cis-biscarbonyl rhenium bipyridine complexes with
a C2V symmetry.13 In these cases, the principal force constant
kCO and the interaction force constant kCO,CO both can be
calculated using energy factored force field (EFFF) theory.18
The results are summarized in Table 4.
I0/I ) 1 + kqτem [TEOA]
(1)
Photocatalytic Reaction. A 4 mL aliquot of DMF-TEOA (5:1 v/v)
solution containing each of the rhenium complexes (0.5 mM) was
introduced into a quartz cell (7 mL volume), and CO2 was bubbled
through for 20 min before the cell was sealed using a rubber septum
(Aldrich). The sample solution was kept at 25 ( 1 °C and was irradiated
(16) Corporation, M. S. TEXSAN, Program for the Crystal Structure; The
Woodlands, TX 77381-5209, USA, 2000.
(17) Sheldrick, G. M. SHELXL97, Program for the Crystal Structure; University
of Go¨ttingen: Germany, 1997.
(18) (a) Ishitani, O.; George, M. W.; Ibusuki, T.; Johnson, F. P. A.; Koike, K.;
Nozaki, K.; Pac, C.; Turner, J. J.; Westwell, J. R. Inorg. Chem. 1994, 33,
4712-4717. (b) Braterman, P. S. Metal Carbonyl Spectra; Academic
Press: London, 1975.
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15547

A R T I C L E S
Tsubaki et al.
Table 3. Photophysical Properties of 1+ Measured in MeCN
ORTEP drawing of 1e+ is shown in Figure 3 together with that
of 1h+, which we have previously reported.13 Comparison of
the structure of 1e+, with two P(p-FPh)3 ligands, and that of
1h+, with one PPh3 and one P(OEt)3 ligand, is interesting,
because intramolecular interactions in these complexes are
different. The configurations of the atoms around the rhenium
metal are very similar in these two complexes except for the
Re-P bond lengths. For Re-PPh3 these are 2.41 Å -2.45 Å
in both complexes, whereas Re-P(OEt)3 in 1h+ was 2.35 Å,
similar to Re-P bonds reported for [Re(bpy)(CO)3(PR3)]+-type
complexes.19
Solution at 298 Ka
b
e
λ
abs(MLCT)
nm
λ
(
abs π−π*) λem
∆
Eabs
τem
ns
krd
104 s-
knr
-
em
1
c
1
1
complex
nm
nm
cm-
Φem
×
×
104 s-
1a+
1b+
1c+
1d+
1e+
1f+
1g+
1h+
1i+
415
411
402
400
396
395
388
382
379
366
365
363
291
292
294
293
292
296
285
288
287
284
283
281
618
608
600
598
595
592
612
606
602
7920 1224 0.07
7880 1101 0.09
8210 1074 0.08
8280 1003 0.07
8440 1046 0.10 10
8430 1220 0.13 11
9430 601 0.03
9680 511 0.05
9780 665 0.06
5.5
8.2
7.8
7.0
76
83
90
93
93
71
5.0
150
190
140
260
270
260
9.8
9.0
7.9
8.5
1j+
1k+
1l+
605 10 790 381 0.03
603 10 820 355 0.03
602 10 940 361 0.05 14
We have previously reported that intramolecular π-π and
CH-π interactions exist between the bpy ligand and the aryl
groups (Ar) of the PAr3 ligand in fac-[Re(bpy)(CO)3(PAr3)]+.19
Similar intramolecular π-π and CH-π interactions were
-
a All complexes are PF6 salts. b Apparent Stokes shift defined as
∆Eabs-em ) Eabs - Eem.
c Quantum yield of emission: experimental error
observed in 1h+PF6 and are respectively drawn using red or
-
was within 10%. d Radiative decay rate constant. e Nonradiative decay rate
constant.
blue dotted lines in Figure 3b; its pyridine rings were distorted
(see Supporting Information); such distortion was also observed
in fac-[Re(bpy)(CO)3(PAr3)]+. Two pyridine rings of the dmb
ligand are π-stacked with two phenyl groups on the triphen-
ylphosphine. The distances between the centers of π-stacked
rings are 3.85 Å and 3.95 Å, and the angles between the rings
are 21.53° and 35.02°, respectively.20 The distance between C(7)
in the pyridine ring and H(32) in the phenyl group is 2.94 Å,
which is less than the sum of the corresponding van der Waals
radii.
Table 4. νCO Frequencies and Force Constants for 1+ and the
Corresponding One-Electron-Reduced Species in MeCN Solution
at 298 Ka
OER speciesd
b
c
b
c
e
1
e
νCO
cm-
kCO
kCO,CO
νCO
cm-
kCO
kCO,CO
∆
kCO
∆
kCO,CO
1
1
1
1
1
1
complex
N m-
N m-
N m-
N m-
N m-
Å
1a+
1930 1451 54.3 1902 1404 58.0
-47
-3.7
-4.4
1859
1825
1b+
1c+
1d+
1e+
1f+
1g+
1h+
1i+
1933 1454 55.2
1861
1936 1460 54.5 1908 1412 58.9
Intramolecular interactions also exist among the two P(p-
-48
FPh)3 ligands and the dmb ligand in crystals of 1e+CF3SO3
.
-
1865
1830
In this case, each pyridine ring is sandwiched between two aryl
groups. One of these aryl groups interacts with the pyridine
ring in parallel fashion (red dotted lines in Figure 3a), where
the plane-plane distances are 3.55 Å and 4.18 Å and the plane-
plane angles are 12.94(22)° and 19.07(32)°, respectively. The
second aryl group is closely placed in the pyridine ring with a
larger angle, 54.20(19)° and 46.05(19)°, where the distances
between H(16) and C(8) and between H(28) and C(3), 2.89 Å
and 2.81 Å, both are less than the sums of the corresponding
van der Waals radii (blue dotted lines in Figure 3a). Conse-
quently, each of the two pyridine rings interacts with two sets
of two phenyl groups; two CH(phenyl)-π(pyridine)-π(phenyl)
1937 1462 53.8
1867
1939 1465 53.8 1910 1416 57.5
1869 1834
1942 1470 53.9 1912 1420 56.8
-49
-50
-3.7
-2.9
1872
1837
1939 1462 56.9
1865
1945 1472 56.3 1917 1426 58.4
-46
-44
-2.1
-2.2
1872
1840
1950 1480 56.4
1877
1946 1472 57.8 1919 1428 60.0
1j+
1k+
1l+
1871
1840
1953 1484 57.3
1879
1956 1488 57.4
1882
interactions exist among the dmb ligand and the two P(p-XPh)3
ligands in 1e+CF3SO3
.
-
1H NMR Spectra. All complexes displayed three signals due
to the dmb-aromatic-ring protons and one singlet due to the
methyl groups on the dmb ligand, as summarized in Table 7.
The data show that two pyridine rings of the dmb ligand are
equivalent; at 25 °C the phosphine ligands can rotate around
the Re-P bonds on the NMR time scale. Strong shielding effects
by the aryl groups of the phosphine ligand(s) are observed in
the dmb-aromatic-ring protons and even in the methyl protons
in 1a+-1i+, especially in 1a+-1f+, in contrast with 1j+-1l+
which have no aryl group in the phosphorus ligands. Similar
shielding effects have been reported for ruthenium(II)21 and
rhenium(I)19 bpy complexes, with π-π interaction between the
bpy ligand and a phenyl ring bounded to another ligand. The
strong shielding effects observed in 1a+-1i+ therefore suggest
-
a All complexes are PF6 salts. b Principal force constant. c Interaction
force constant. d The one-electron-reduced species were electrochemically
produced in acetonitrile solution containing 0.1 M Et4NBF4 under Ar
atmosphere. See ref 18a. e Differences between the OER species and their
parent complexes.
Electrochemical Behavior. Figure 2 shows cyclic voltam-
mograms of 1e+, 1h+, and 1j+ as typical examples. For all
complexes, one reversible wave was observed in the cathodic
scan, attributable to a dmb-based one-electron reduction. Re-
versibility of a one-electron oxidation wave attributable to metal-
based, e.g., ReI/ReII, was different; however, a reversible or
quasireversible wave was observed for complexes with two
trialkylphosphite ligands (1j+-1l+), but oxidation waves of
complexes with two triarylphosphine ligands (1a+-1f+) were
irreversible. Table 5 summarizes the electrochemical data.
Crystal Structures. X-ray crystallographic analysis of
(19) Tsubaki, H.; Tohyama, S.; Saitoh, H.; Sakamoto, K.; Ishitani, O. Dalton
Trans. 2005, 385-395.
(20) Janiak, C. J. Chem. Soc., Dalton Trans. 2000, 3885-3896.
(21) Bolger, J. A.; Feruguson, G.; James, J. P.; Long, C.; McArdle, P.; Vos, J.
G. J. Chem. Soc., Dalton Trans. 1993, 1577-1583.
-
1e+CF3SO3 was performed. Tables 2 and 6 summarize the
crystallographic data and selected bond lengths and angles. The
9
15548 J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005

Ligand Interactions to Control Re(I) Complex Properties
A R T I C L E S
Figure 2. Cyclic voltammograms of 1e+ (red) and 1h+ (blue) and 1j+ (black) measured in acetonitrile solution containing n-Bu4NBF4 (0.1 M) using a
glassy carbon working electrode, a Pt counter electrode, and a Ag/AgNO3 (0.1 M) reference electrode. The scan rate was 200 mV/s.
Table 5. Electrochemical Data for 1+a
Discussion
E (V vs Ag/AgNO3)
Intramolecular Interactions between Ligands. X-ray crys-
red b
ox b
ox c
E1
/
E1
/2
Ep
2
tallographic analysis data of 1h+ with both the PPh3 ligand and
the P(OEt)3 ligand (Figure 3b) demonstrate that two sets of
intramolecular π-π interactions are present between two of the
three phenyl groups on the PPh3 ligand and two pyridine rings
in the dmb ligand and that CH-π interaction also takes place
between one of the phenyl groups on PPh3 and the pyridine
ring on dmb. Intramolecular interactions were not observed be-
tween the triethyl phosphite and the dmb ligands. The crystal
structures of nine fac-[Re(bpy)(CO)3(PR3)]+-type complexes19
clearly indicate that, in complexes with a triarylphosphine lig-
and, similar π-π and CH-π interactions exist between the
complex
(dmb/dmb•-
)
(ReI/ReII)
(ReI/ReII)
1a+
1b+
1c+
1d+
1e+
1f+
1g+
1h+
1i+
-1.83 (74)
-1.81 (66)
-1.79 (65)
-1.75 (65)
-1.73 (62)
-1.71 (66)
-1.85 (66)
-1.81 (66)
-1.74 (63)
-1.86 (63)
-1.82 (65)
-1.79 (66)
+0.80
+0.89
+0.99
+1.04
+1.09
+1.11
+0.89
+1.04
+1.10
+1.05
1j+
1k+
1l+
+1.04(81)
+1.08(85)
1
aryl group and the bpy ligand in crystal. Their H NMR and
a All complexes are PF6 salts. Cyclic voltammograms were taken in
MeCN solution containing 0.1 M nBu4NBF4 at a 200 mV/s scan rate using
a glassy carbon working electrode, a Pt counter electrode, and a 0.1 M
Ag/AgNO3 reference electrode under Ar atmosphere. b Redox potential for
a reversible or quasireversible wave and peak potential difference in
parentheses. c Peak potential for irreversible oxidation.
-
UV-vis spectra and redox potentials strongly suggest that such
weak interactions are maintained even in solution, but these
weak interactions were not observed in the cases of complexes
with a trialkyl- and triphenyl phosphite. On the other hand, the
dmb ligand in 1e+ is sandwiched between four of the six
4-fluorophenyl groups on the two phosphine ligands, as
shown in Figure 3a. The two phenyl groups are parallel to each
of the pyridine rings, where π-π interactions were observed,
and the other two phenyls are sited above each pyridine ring
with dihedral angles 54.20° and 46.05°, where CH-π interac-
tions were observed; specifically, two sets of three-centered
CH(phenyl)-π(pyridine)-π(phenyl) interactions operated on
both pyridine rings of dmb.
Table 6. Selected Bond Lengths (Å) and Angles (deg) for
1e+CF3SO3
-
Bond Lengths
Re(1)-P(1)
Re(1)-N(1)
Re(1)-C(1)
C(1)-O(1)
2.4135(10)
2.182(3)
1.906(4)
1.158(5)
Re(1)-P(2)
Re(1)-N(2)
Re(1)-C(2)
C(2)-O(2)
2.4211(9)
2.182(3)
1.903(4)
1.153(5)
Bond Angles
P(1)-Re(1)-N(1)
P(1)-Re(1)-C(1)
P(2)-Re(1)-N(1)
P(2)-Re(1)-C(1)
N(1)-Re(1)-N(2)
N(1)-Re(1)-C(2)
N(2)-Re(1)-C(2)
P(1)-Re(1)-P(2)
89.99(10)
87.90(11)
91.04(10)
90.78(11)
74.58(11)
168.86(14)
94.28(14)
178.42(3)
P(1)-Re(1)-N(2)
P(1)-Re(1)-C(2)
P(2)-Re(1)-N(2)
P(2)-Re(1)-C(2)
N(1)-Re(1)-C(1)
N(2)-Re(1)-C(1)
C(1)-Re(1)-C(2)
92.36(9)
90.15(12)
89.07(9)
89.07(12)
97.65(14)
172.23(14)
93.48(16)
Proton-NMR data obtained in an acetone-d6 solution (Table
7) show that the chemical shifts for dmb-protons of the com-
plexes with one (1g+-1i+) and two (1a+-1f+) triarylphosphine
ligand(s) were shifted upfield by 0.2-0.4 ppm and 0.4-1.0 ppm,
respectively, compared to complexes with no triarylphosphine
ligand (1j+-1l+). This suggests that the weak interactions
observed in crystal remain even in solutions, and the similarities
in each group suggest that configurations among the dmb ligand
and the phosphorus ligands are similar in each group, so that
similar CH(phenyl)-π(pyridine)-π(phenyl) interactions exist
in 1a+-1d+ and 1f+, and π-π and CH-π interactions exist in
1g+ and 1i+.
that the intramolecular interactions between the dmb and the
triarylphosphine ligands remain even in solution. This hypothesis
is also supported by comparison of the spectroscopic and
electrochemical data of 1a+-1f+, 1g+-1l+, and 1j+-1l+; see
the Discussion section.
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15549

A R T I C L E S
Tsubaki et al.
-
Figure 3. ORTEP drawings of (a) 1e+CF3SO3 and (b) 1h+PF6-. Displacement ellipsoids are shown at the 50% probability level. Counteranions were
omitted for clarity.
Table 7. 1H NMR Chemical Shift Data for the dmb Ligand of 1+ in
strongly influences kCO.
19,22 Figure 4 shows the relation between
Acetone-d6 Solutiona
kCO and the sum of the Tolman ø values of each phosphorus
ligand (Σø); these are used as a measure of the electron accepting
ability of a phosphorus ligand.23 The good linear relation clearly
shows that the electron densities on the dπ orbitals of the central
rhenium are affected by the phosphorus ligands through the
P-Re bonds. Almost no effect of the intramolecular through-
space interactions between ligands is observed.
The dmb-based reduction potential E1/2(dmb/dmb•-) is strongly
affected by the intramolecular interactions between the dmb
and triarylphosphine ligands. As shown in Figure 5, three
lines can be drawn for the relation between E1/2(dmb/dmb•-)
and Σø, depending on the number of triarylphosphine ligands:
1a+-1f+ with two triarylphosphine ligands; 1g+-1i+ with one
triarylphosphine ligand; 1j+-1l+ with no triarylphosphine
ligands. If the phosphorus ligands affect the electron accepting
δ/ppm (J/Hz)
complex
3,3′
5,5′
6,6
′
CH3
1a+
1b+
1c+
1d+
1e+
1f+
1g+
1h+
1i+
8.16 (s)
8.18 (s)
8.12 (s)
8.22 (s)
8.27 (s)
8.23 (s)
8.37 (s)
8.36 (s)
8.44 (s)
8.62 (s)
8.60 (s)
8.59 (s)
6.89 (d, 5.8)
6.86 (d, 5.9)
6.81 (d, 5.9)
6.91 (d, 5.8)
7.01 (d, 5.9)
7.03 (d, 5.1)
7.33 (d, 5.6)
7.39 (d, 5.5)
7.27-7.38
7.93 (d, 5.8)
7.91 (d, 5.9)
7.87 (d, 5.9)
7.94 (d, 5.5)
8.01 (d, 5.9)
8.07 (d, 5.6)
8.48 (d, 5.9)
8.48 (d, 5.5)
8.54 (d, 5.9)
8.89 (d, 5.5)
8.91 (d, 5.7)
8.92 (d, 5.9)
2.45 (s)
2.44 (s)
2.44 (s)
2.45 (s)
2.47 (s)
2.49 (s)
2.55 (s)
2.54 (s)
2.55 (s)
2.64 (s)
2.63 (s)
2.63 (s)
1j+
1k+
1l+
7.68 (d, 5.9)
7.63 (d, 5.7)
7.62 (d, 5.9)
a All complexes are PF6 salts.
-
Effects on the Ground State. The principal force constant
kCO of the CO ligands is a reliable measure of the electron
density on the dπ orbitals of the central rhenium, since π-back-
donation from dπ orbitals to the antibonding π* orbital of CO
(22) The oxidation potential Ep.a.(ReI/ReII) cannot be used as a reliable measure
of the electron density on dπ orbitals of the central rhenium because of
the considerable difference in reversibility of the oxidation wave on the
cyclic voltammograms, as seen in Figure 2.
(23) Tolman, C. A. Chem. ReV. 1977, 77, 313-348.
9
15550 J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005

Ligand Interactions to Control Re(I) Complex Properties
A R T I C L E S
Figure 4. Plots of kCO against Σø values of the PR3 ligands.
Figure 6. Variations of MLCT absorption maxima λabs(MLCT) (in
acetonitrile solution at 298 K) with Σø values of the PR3 ligands.
Figure 5. Variations of E1/2 (dmb/dmb•-) with Σø values of the PR3 ligands
(in acetonitrile solution containing 0.1 M n-Bu4NBF4; potentials are vs a
Ag/AgNO3 reference electrode).
Figure 7. Variations of π-π*(dmb) absorption maxima λabs(π-π*) (in
acetonitrile solution at 298 K) with Σø values of the PR3 ligands.
ability of the dmb ligand mostly through the P-Re-N bonds,
so that there is no effect of the intramolecular interactions
between the dmb and phosphorus ligands, only one straight line
would exist in Figure 5. In each of the three groups of the
complexes, a good linear relation was observed between Σø and
10-43 nm (110-340 meV), in good agreement with the
differences in E1/2(dmb/dmb•-) described above. This is reason-
able because the energy of MLCT absorption is proportional to
the difference in energy levels between the π* orbital of the
dmb ligand and the dπ orbital of the central rhenium.
E
1/2(dmb/dmb•-). It follows that the effect of the intramolecular
Another band attributed to dmb-ligand centered π-π*
absorption was also red-shifted by the intramolecular interac-
tions. The energy of the absorption was only weakly affected
by Σø of the phosphorus ligands, probably because the energy
of the π orbital is also shifted by the P-Re-N through-bond
interaction, with the same direction of the energy shift of the
π* orbital; the average differences from 1j+-1l+ with no triaryl-
phosphine ligands were 1-7 nm (20-110 meV) for 1g+-1i+
and 7-15 nm (100-220 meV) for 1a+-1f+ (Figure 7).
In the series of tricarbonyl rhenium complexes fac-[Re(bpy)-
(CO)3(PAr3)]+ (bpy ) 2,2′-bipyridine and PAr3 ) triarylphos-
phine) with π-π and CH-π interactions between the bpy ligand
and the aryl groups,19 similar shifts of the π-π*(bpy) absorption
band and E1/2(bpy/bpy•-) were observed in those for 1g+-1i+.
It follows that both types of the diimine ligand, i.e., bpy and
through-space interactions between ligands can be distinguished
from the effect of the N-Re-P through-bond interactions. By
extrapolating the lines, we deduce that the intramolecular π-π
and CH-π interactions shift E1/2(dmb/dmb•-) positively by ca.
110 mV and the CH(phenyl)-π(pyridine)-π(phenyl) interac-
tions by 180-200 mV.
The intramolecular interactions also affect the UV/vis absorp-
tion spectra of the complexes. Figure 6 illustrates the relation
between the absorption maximum of the MLCT band and Σø,
where there are again three lines depending on the number of
triarylphosphine ligands. Differences between 1g+-1i+ with one
triarylphosphine ligand and 1j+-1l+ with no triarylphosphine
ligands were 11-19 nm (100-160 meV) and between
1a+-1f+ with two triarylphosphine ligands and 1j+-1l+ were
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15551

A R T I C L E S
Tsubaki et al.
Figure 9. Qualitative illustration of the potential energy curves of the
ground state (GS) and metal-to-ligand charge-transfer excited state (1MLCT
in the case of absorption and 3MLCT in the case of emission) of cis,trans-
[Re(dmb)(CO)2(PR3)(PR′3)]+ where (a) R, R′ ) alkoxy and (b) aryl groups.
In both cases the energies of the GS and 3MLCT are assumed to be identical
with each other.
Figure 8. Variations of MLCT emission maxima λem (in acetonitrile
solution at 298 K) with Σø values of the PR3 ligands.
dmb, and of the ligand in the trans position to the triarylphos-
phine ligand, i.e., CO and P(OR)3, should not strongly influence
the intramolecular interactions. However, the CH(phenyl)-
π(pyridine)-π(phenyl) interaction has much stronger effects on
properties correlated with the dmb ligand.
1
be explained using Figure 9 as follows. Both the MLCT and
3MLCT excited states of a metal complex usually have different
nuclear geometries and solvation from the corresponding ground
state, as shown in Figure 9a, because of the differing polarization
of the molecule between the ground and excited states.74 The
intramolecular interactions probably cause the electron located
Effects on the 3MLCT Excited State. All the rhenium
complexes reported here showed strong and unstructured
emission from the lowest 3MLCT excited state at room
temperature in solution; an example is shown in Figure 1. Figure
8 illustrates the relation between the emission maximum and
Σø, where dependence on the number of the triarylphosphine
ligand(s) in the complex was again observed, and each group
of the complexes shows a different relation with Σø. The pattern
is significantly different from that for MLCT absorption bands
described above (Figure 6): typically, 1c+-1f+, with two
triarylphosphine ligands, emit with higher energy than the other
complexes with one or no triarylphosphine ligand, while the
MLCT absorption bands of 1c+-1f were observed at much
lower energy. The CH(phenyl)-π(pyridine)-π(phenyl) interac-
tions consequently reduce the MLCT absorption energy but raise
the emission energy from the 3MLCT excited state. In the cases
of 1g+-1i+, although the emissions were observed at similar
wavelength as 1j+-1l+, examination of the lower energy MLCT
absorptions of 1g+-1i+ relative to 1j+-1l+ leads to the
conclusion that the π-π and CH-π interactions in 1g+-1i+
also cause higher-energy shifts of the emission from their
“ordinal” energies. The π-π and CH-π interactions and the
CH(phenyl)-π(pyridine)-π(phenyl) interaction consequently
decrease the Stokes shifts by ∼140 and ∼310 meV, respectively.
3
mainly on the dmb ligand in the MLCT excited state to be
partially dispersed to the aryl groups and the hydrogen on the
phosphorus ligands. This should lower the difference in the
polarization between the ground state and the MLCT excited
states, reducing the gap between their energy surfaces on the
abscissa as shown in Figure 9b. In this situation, the absorption,
i.e., the transition from the lowest vibrational level in the ground
state to the MLCT excited state, is observed at lower energy,
but the emission, i.e., the jump from the lowest vibrational level
of 3MLCT excited to the ground state, is at higher energy. The
1
3
smaller gap between the ground state and the MLCT excited
state may also give rise to smaller Franck-Condon factors, as
illustrated in Figure 9b, slowing nonradiative decay.
Another possible explanation for the phenomena involving
weak interactions is an “internal solvation” effect of the aryl
groups on the phosphine ligands, such that when the dmb ligand
is between two hydrophobic aryl groups, the local environment
around the bpy ligand has a lower regional dielectric constant
than others. This hydrophobic environment could cause the
MLCT excited state to be unstable, so that emission is blue-
shifted. This cannot, however, explain the positive shift of the
1
reduction potential and the red-shift of the MLCT absorption
3
The lifetime of the MLCT excited state was also strongly
band.
affected by the intramolecular interactions (Table 3): the π-π
and CH-π interactions and the CH(phenyl)-π(pyridine)-
π(phenyl) interaction increase the emission lifetime (τem) by
about twice and three times, respectively. This variation is
dominated by effects of the intramolecular interactions on the
nonradiative decay rate from 3MLCT (knr in Table 3), whereas
the radiative decay rate (kr) did not systematically change.
As described above, the intramolecular interactions red-shift
Figure 10 shows T-T absorption spectra of 1e+, 1h+, and
1k+ measured 20 ns after a laser flash. The broad absorption
band at 420-600 nm is attributable to the (dmb-•) ligand
localized excitation in the 3MLCT excited state. The absorption
of (1e+)* with two triarylphosphine ligands was broader, and
its λmax was red-shifted by 26 nm relative to that of (1k+)*
without triarylphosphine. This also suggests that the CH(phen-
yl)-π(pyridine)-π(phenyl) interaction is maintained in the
excited state of the complexes with two triarylphosphines. The
1
the MLCT absorption band but blue-shift the emission from
3
the MLCT excited state. These “discrepant” phenomena can
9
15552 J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005

Ligand Interactions to Control Re(I) Complex Properties
A R T I C L E S
Figure 10. T-T Absorption spectra of 1e+ (red), 1h+ (blue), and 1k+
(black) measured 20 ns after a laser flash (355 nm). The solvent was
acetonitrile, and the absorbance changes (∆absorbance) were corrected by
the relative absorption light quanta of the sample solutions.
Figure 12. UV/vis absorption spectra of the one-electron-reduced (OER)
species produced using flow electrolysis in Ar-saturated acetonitrile solution
containing 0.1 M Et4NBF4 at room temperature: 1e (red), 1h (blue), and
1j (black).
that a one-electron-reduced (OER) species, i.e., 1e, was produced
by the electrolysis and remained stable during spectrum mea-
surement. The νCO bands of 1e were shifted to lower frequency
relative to those of 1e+ (Table 4). Such shifts are usually ob-
served in one-electron reduction of rhenium(I) diimine com-
plexes, because stronger π-back-donation from the Re to the
antibonding π* orbital of the CO ligands, which was promoted
by higher electron density on the (dmb•-) ligand, weakens the
CO bonds.3,10b,18a Similar IR spectral changes were observed
for the electrochemical reduction of 1a+, 1c+, 1f+, 1h+, and
1j+. Table 4 summarizes the principal and interaction force
constants kCO and kCO,CO and their differences (∆kCO and
∆kCO,CO) between the OER species and their parent complexes.
The similarity in ∆kCO and ∆kCO,CO for all measured rhenium
complexes clearly indicates that the electron density on the dπ
orbitals of the central rhenium is predominantly affected by the
phosphorus ligands through the Re-P bonds in the OER species.
On the other hand, UV/vis absorption spectra of OER species
with one or two triarylphosphine ligands were dramatically
different from those with none. Figure 12 shows three typical
examples, i.e., 1j, 1h, and 1e, in which the phosphorus ligands
have similar ø values. The spectrum of 1j has common
characteristics with spectra of various fac-[Re(LL•-)(CO)3-
(L′)]n+- and cis,trans-[Re(LL•-)(CO)2{P(OR)3}(L′)]n+-type com-
plexes (LL ) bpy and its derivatives; L′ ) monodentate ligand;
n ) 0 or 1);10a,b,18 the vibrational bands were observed at 480-
540 nm. This electronic transition is attributed to ligand (dmb•-)-
centered excitation. The corresponding band of 1h with a
triphenylphosphine ligand was, on the other hand, nonvibrational
and broad. In the spectrum of 1e with two tri(4-fuluorophenyl)-
phosphine ligands, no absorption maximum was seen around
500 nm, and the absorption extended to 800 nm. The spectra
of 1a and 1c were very similar to those of 1e. These results
strongly suggest that the intramolecular interactions between
the dmb ligand and the triarylphosphine ligand(s) are maintained,
and probably enhanced, in the OER species. It is reasonable to
suppose that the extra electron residing in the π* orbital of the
dmb ligand is partially dispersed to the aryl groups on the
Figure 11. IR spectral change in the νCO region by flow electrolysis of
1e+ (0.5 mM) in Ar-saturated acetonitrile solution containing 0.1 M
Et4NBF4. The complex was reduced at 0 V to -1.96 V vs Ag/AgNO3 at a
constant flow rate of 0.5 mL/min.
T-T absorption of (1h+)* with one triarylphosphine ligand was
also red-shifted, but the difference was 6 nm compared with
(1k+)*.
Effects on One-Electron-Reduced Species of the Com-
plexes. In most systems using a rhenium(I) diimine complex
as photocatalyst or electrocatalyst, the one-electron-reduced
(OER) species of the complex, where the added electron is
mainly located in the diimine ligand, is an important intermedi-
ate produced by reductive quenching of the 3MLCT excited state
of the complex9,10,18a,24,25 or electrochemical reduction.3,26 We
investigated the effects of the intramolecular interactions on
properties of the OER species of the rhenium complexes.
Figure 11 shows the IR spectra in the νCO region of an
acetonitrile solution containing 1e+ and an electrolyte after flow
electrolysis18a at various potentials. A set of observed isosbestic
points and results of coulometric measurements clearly show
(24) Kutal, C.; Corbin, A. J.; Ferraudi, G. Organometallics 1987, 6, 553-557.
(25) Kalyanasundaram, K. J. Chem. Soc., Faraday Trans. 2 1986, 82, 2401-
2415.
(26) (a) Sullivan B. P.; Bolinger C. M.; Conrad, D.; Vining, W. J.; Meyer, T.
J. J. Chem. Soc., Chem. Commun. 1985, 1414-1416. (b) Hayashi, Y.; Kita,
S.; Brunschwig, B. S.; Fujita, E. J. Am. Chem. Soc. 2003, 125, 11976-
11987.
3
triarylphosphine ligand(s), as in the MLCT excited state.
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15553

A R T I C L E S
Tsubaki et al.
Figure 13. Relation between the quenching rate constant kq of the emission
from the excited 1+ by TEOA and Σø value of the PR3 ligands.
Figure 14. Relation between the quenched fraction ηq of the excited 1+
and Σø value of the PR3 ligands in 4 mL of TEOA-DMF (1:5 v/v) solution
at room temperature.
Effects on the Photochemical Electron Transfer and
Photocatalysis of the Rhenium Complexes. We expect that
the intramolecular interactions increase the oxidation power of
the rhenium complex in the 3MLCT excited state, as well as in
Table 8. Photocatalytic Reduction of CO2 Using 1+ as
Photocatalyst
a
b
complex
TNCO
ΦCO
ηqc/%
1
1a+
1b+
1c+
1d+
1e+
1f+
1g+
1h+
1i+
2.3
4.6
6.8
13.1
17.3
10.6
0.5
0.6
1.4
0.2
0.2
0.01
0.07
0.16
0.19
0.20
0.22
0.02
0.05
0.09
0.02
0.06
0.12
21
36
56
64
79
91
11
38
79
22
45
70
the ground state, even though the MLCT absorption band is
red-shifted. This is because of the positive shift in the reduction
potential of the complex in the ground state and of the blue
3
shift of the emission from the complex in the MLCT excited
state. Figure 13 shows the relation between Σø and the
quenching rate constant of the emission from the excited 1+ by
triethanolamine (TEOA), which has been often used as a
sacrificial electron donor in photocatalytic CO2 reduction using
various rhenium complexes.9,10 As expected, different lines can
be drawn for each group, including complexes with no tri-
arylphosphine ligand, with one, and with two. From Figure 13,
acceleration of the quenching is estimated at ∼5 times for the
CH-π and π-π interactions and ∼35 times for the CH(phen-
yl)-π(pyridine)-π(phenyl) interactions. In photocatalytic CO2
reduction using rhenium complexes, a mixed solution of 1:5
(v/v) TEOA-dimethylformamide has been often used. In this
solution the quenched fraction of the excited 1+, ηq obtained
using eq 8 where I and I0 are the respective emission intensities
in the presence and absence of TEOA, varied with both
electronic properties of the phosphorus ligands and the intramo-
lecular interactions, as shown in Figure 14. The intramolecular
interactions can cause a red-shift of the 1MLCT absorption band
of complexes with the same ηq value, up to 50 nm.
1j+
1k+
1l+
1.0
a Turnover number of CO formation based on the complex used (0.5
mM). b Quantum yield for CO formation. 4 mL of DMF-TEOA (1.26 M)
solution containing the complex (2.0 mM) was irradiated at 365 nm. Light
intensity was 8.22 × 10-9 einstein s-1 for 1a+, 1c+, 1e+, 1f+, 1h+, and
1j+, and 5.24 × 10-9 einstein s-1 for 1b+, 1d+, 1g+, 1i+, 1k+, and 1l+.
Error (10%. c Reductive-quenching efficiency of the 3MLCT excited states
of 1+ by TEOA (1.26 M).
clear, because the reaction mechanism of photocatalytic reduc-
tion of CO2 using rhenium complexes has not been settled.
However, there are some significant experimental results. It has
been reported that reaction of OER species produced by
reductive quenching with CO2 is a significant process in fac-
[ReI(X2bpy)(CO)3PR3]+-photocatalytic reduction of CO2, which
red
proceeds efficiently when E1/2 for the complex is more
negative than -1.4 V vs Ag/AgNO3.10b All of the complexes
1+ satisfy this requirement, as shown in Table 5. Stability of
the OER species is also important for CO2 reduction photo-
catalysts. For example, fac-[Re(bpy)(CO)3L]+ (L ) pyridine
and PPh3) cannot work as photocatalyst for CO2 reduction
because of instability of their OER species, from which loss of
L initiates a chain reaction of decomposition of the parent
complex.10c,27 During the photocatalytic reaction, much faster
decomposition of 1g+-1i+ with one triarylphosphine ligand was
observed than with the other complexes. Figure 15b shows a
I0 - I
I0
ηq )
× 100
(8)
Table 8 summarizes photocatalyses of 1+ for CO2 reduction.
Only those complexes with the CH(phenyl)-π(pyridine)-
π(phenyl) interactions 1a+-1f+ can photocatalyze CO2 reduc-
tion to give CO. Since the photocatalytic reactions are initiated
by the reductive quenching process of the 3MLCT excited state
of 1+ by TEOA, the intramolecular interaction should facilitate
this process, as described above.
The other complexes had low or no photocatalytic ability in
this reaction condition. The reason for this difference is not yet
(27) Hori, H.; Ishihari, J.; Koike, K.; Takeuchi, K.; Ibusuki, T.; Ishitani, O. J.
Photochem. Photobiol., A: Chem. 1999, 120, 119-124.
9
15554 J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005

Ligand Interactions to Control Re(I) Complex Properties
A R T I C L E S
1e and 1j (Figure 15a and c), and the presence of CO2 in the
reaction solution (closed circles) did not increase the decay rate,
implying that 1h has too short a lifetime to react with CO2
because it decomposes to give the corresponding solvento
complexes [Re(dmb)(CO)2{P(OEt)3}(S)]+ (S ) DMF and
TEOA) probably via eqs 9 and 10, where Ox ) [Re(dmb)-
(CO)2{P(OEt)3}(PPh3)]+ or other oxidants.
[Re(dmb•-)(CO)2{P(OEt)3}(PPh3)] + S f
[Re(dmb•-)(CO)2{P(OEt)3}(S)] + PPh3 (9)
[Re(dmb•-)(CO)2{P(OEt)3}(S)] + Ox f
[Re(dmb)(CO)2{P(OEt)3}(S)]+ + Ox-• (10)
This is likely to be a main reason for the low photocatalysis of
1h+, and probably also for 1g+ and 1i+. Both of the other OER
species 1e and 1j were relatively stable under Ar atmosphere
but decayed faster under CO2, where the pseudo-first-order rate
constants were, respectively, 0.016 and 0.013 s-1. We do not
have any experimental data about the properties of the CO2
adducts produced from the OER species, which might strongly
affect the photocatalytic reactions.
Conclusion
Intramolecular interactions, especially CH(phenyl)-π(pyri-
dine)-π(phenyl) interactions, between the dmb ligand and the
triarylphosphine ligand(s), have striking and unique effects on
the various properties associated with the dmb ligand in the
ground state, in the excited state, and in the one-eletron-reduced
state. Typical examples are summarized in Table 1, in which
the complexes with similar Σø values (1e+, 1h+, and 1j+) can
be compared with each other. The interactions cause the
UV/vis absorptions to both π-π* and 1MLCT to be red-shifted,
3
but the emission from MLCT is blue-shifted, the lifetime of
3
the MLCT excited state is prolonged, the oxidation power in
both the ground and excited states is enhanced, and photoca-
talysis for CO2 reduction is also enhanced. Moreover, properties
associated with the central rhenium, specifically the electron
densities on the dπ orbitals, are much less affected by the
interactions or are unaffected. These unique characteristics of
the intramolecular interactions give a novel and potentially
useful method to control the photophysical, photochemical, and
electrochemical functions of metal complexes.
Acknowledgment. We are grateful thanks to Dr. Koichi
Nozaki of the Osaka University for his helpful discussion. This
work was partly supported by a Grant-in-Aid for Scientific
Research No. 17036015 on Priority Areas (434:Chemistry of
Coordination Space) from the Ministry of Education, Culture,
Sports, Science and Technology of Japan, and CREST, Japan
Science and Technology Agency.
Figure 15. Decay curves of the OER species of (a) 1e+, (b) 1h+, and (c)
1j+, produced by the photochemical reduction by TEOA using 365-nm light
in 4 mL of TEOA-DMF (1:5 v/v) solution at room temperature, under Ar
atmosphere (O) or under CO2 atmosphere (b). The sample concentration
was 0.5 mM.
Supporting Information Available: X-ray crystallographic
files (CIF) of 1e+, deviations of the atoms constructing the
pyridine rings of the dmb ligand from their best planes in the
crystals of 1e+CF3SO3- and 1h+PF6-. This material is available
free of charge via the Internet at http://pubs.acs.org.
decay curve of 1h produced by photochemical reduction of 1h+
using TEOA. Decay under Ar atmosphere (open circles) was
much faster than for the other types of the complexes such as
JA053814U
9
J. AM. CHEM. SOC. VOL. 127, NO. 44, 2005 15555