Synthesis and characterization of ruthenium bis(β-diketonato) pyridine-imidazole complexes for hydrogen atom transfer

Article abstract of DOI:10.1021/ic7015726

Ruthenium bis(β-diketonato) complexes have been prepared at both the Ru^II and Ru^III oxidation levels and with protonated and deprotonated pyridine-imidazole ligands. Ru^II(acac) ₂(py-imH) (1), [Ru^III(acac)₂(py-imH)]OTf (2), Ru^III-(acac)₂(py-im) (3), Ru^II(hfac) ₂(py-imH) (4), and [DBU-H][Ru^II(hfac)₂(py-im)] (5) have been fully characterized, including X-ray crystal structures (acac = 2,4-pentanedionato, hfac = 1,1,1,5,5,5-hexafluoro-2,4-pentanedionato, py-imH = 2-(2′-pyridyl)imidazole, DBU = 1,8-diazabicyclo[5.4.0]undec-7-ene). For the acac-imidazole complexes 1 and 2, cyclic voltammetry in MeCN shows the Ru^III/II reduction potential (E_1/2) to be -0.64 V versus Cp₂Fe^+/0. E_1/2 for the deprotonated imidazolate complex 3 (-1.00 V) is 0.36 V more negative. The Ru^II bis-hfac analogues 4 and 5 show the same ΔE_1/2 = 0.36 V but are 0.93 V harder to oxidize than the acac derivatives (0.29 and -0.07 V). The difference in acidity between the acac and hfac derivatives is much smaller, with pK _a values of 22.1 and 19.3 in MeCN for 1 and 4, respectively. From the E_1/2 and pK_a values, the bond dissociation free energies (BDFEs) of the N-H bonds in 1 and 4 are calculated to be 62.0 and 79.6 kcal mol^-1 in MeCN - a remarkable difference of 17.6 kcal mol^-1 for such structurally similar compounds. Consistent with these values, there is a facile net hydrogen atom transfer from 1 to TEMPO^? (2,2,6,6-tetramethylpiperidine-1-oxyl radical) to give 3 and TEMPO-H. The ΔG° for this reaction is -4.5 kcal mol^-1. 4 is not oxidized by TEMPO^? (ΔG° = +13.1 kcal mol^-1), but in the reverse direction TEMPO-H readily reduces in situ generated Ru ^III(hfac)₂(py-im) (6). A Ru^II-imidazoline analogue of 1, Ru^II-(acac)₂(py-imnH) (7), reacts with 3 equiv of TEMPO^? to give the imidazolate 3 and TEMPO-H, with dehydrogenation of the imidazoline ring.

Full text of DOI:10.1021/ic7015726

Inorg. Chem. 2007, 46, 11190−11201
Synthesis and Characterization of Ruthenium Bis(â-diketonato)
Pyridine-Imidazole Complexes for Hydrogen Atom Transfer
Adam Wu, Joshua Masland, Rodney D. Swartz,† Werner Kaminsky,† and James M. Mayer*
Department of Chemistry, UniVersity of Washington, Campus Box 351700, Seattle, Washington
98195-1700
Received August 7, 2007
Ruthenium bis(
protonated and deprotonated pyridine
(acac)2(py-im) (3), RuII(hfac)2(py-imH) (4), and [DBU
X-ray crystal structures (acac 2,4-pentanedionato, hfac
2-(2 -pyridyl)imidazole, DBU 1,8-diazabicyclo[5.4.0]undec-7-ene). For the acac-imidazole complexes 1 and 2,
cyclic voltammetry in MeCN shows the RuIII reduction potential (E1 2) to be
â
-diketonato) complexes have been prepared at both the RuII and RuIII oxidation levels and with
imidazole ligands. RuII(acac)2(py-imH) (1), [RuIII(acac)2(py-imH)]OTf (2), RuIII-
H][RuII(hfac)2(py-im)] (5) have been fully characterized, including
1,1,1,5,5,5-hexafluoro-2,4-pentanedionato, py-imH
−
−
)
)
)
′
)
/
II
−
0.64 V versus Cp2Fe+/0. E1 for the
/
/2
deprotonated imidazolate complex 3 (
the same E1 0.36 V but are 0.93 V harder to oxidize than the acac derivatives (0.29 and −
−
1.00 V) is 0.36 V more negative. The RuII bis-hfac analogues 4 and 5 show
0.07 V). The
∆
)
/
2
difference in acidity between the acac and hfac derivatives is much smaller, with pKa values of 22.1 and 19.3 in
MeCN for 1 and 4, respectively. From the E1 and pKa values, the bond dissociation free energies (BDFEs) of the
/
2
1
N
−
H bonds in 1 and 4 are calculated to be 62.0 and 79.6 kcal mol- in MeCN
−
a remarkable difference of
17.6 kcal mol- for such structurally similar compounds. Consistent with these values, there is a facile net hydrogen
1
atom transfer from 1 to TEMPO• (2,2,6,6-tetramethylpiperidine-1-oxyl radical) to give 3 and TEMPO-H. The
for this reaction is ∆G
4.5 kcal mol-1. 4 is not oxidized by TEMPO• ( ° ) +13.1 kcal mol-1), but in the reverse
∆G°
−
direction TEMPO-H readily reduces in situ generated RuIII(hfac)2(py-im) (6). A RuII-imidazoline analogue of 1, RuII-
(acac)2(py-imnH) (7), reacts with 3 equiv of TEMPO• to give the imidazolate 3 and TEMPO-H, with dehydrogenation
of the imidazoline ring.
Introduction
cite just one example, the peroxidation of polyunsaturated
fatty acids is catalyzed by lipoxygenases using HAT from
the substrate to the active site FeIIIOH center, forming the
FeIIOH2 moiety and the substrate pentadienyl radical.4 As
this example illustrates, many metal-mediated HAT reactions
involve a redox change at the metal, coupled to a change in
the protonation state of the ligand.1-3,5 These systems can
be described by the square scheme in Scheme 1, with electron
Proton-coupled electron transfer (PCET) is a fundamental
process in chemistry and biology.1,2 Hydrogen atom transfer
(HAT) reactions are one class of PCET processes, in which
a hydrogen atom (H• ) H+ + e-) concertedly transfers from
one reagent to another in a single kinetic step. HAT reactions
of transition-metal species are receiving much attention
because of their role in metal-catalyzed oxidations, ranging
from metal-oxide surfaces to various metalloenzymes.1-5 To
(2) (a) Mayer, J. M. Annu. ReV. Phys. Chem. 2004, 55, 363. (b) Mayer,
J. M.; Rhile, I. J. Biochim. Biophys. Acta 2004, 1655, 51. (c) Mayer,
J. M.; Rhile, I. J.; Larsen, F. B.; Mader, E. A.; Markle, T. F.;
DiPasquale, A. G. Photosynth. Res. 2006, 87, 3. (d) Mayer, J. M.;
Mader, E. A.; Roth, J. P.; Bryant, J. R.; Matsuo, T.; Dehestani, A.;
Bales, B. C.; Watson, E. J.; Osako. T.; Valliant-Saunders, K.; Lam,
W.-H.; Hrovat, D. A.; Borden, W. T.; Davidson, E. R. J. Mol. Catal.
A: Chem. 2006, 251, 24.
* To whom correspondence should be addressed. E-mail: mayer@chem.
washington.edu.
† UW Chemistry crystallographic facility.
(1) (a) Cukier, R. I.; Nocera, D. G. Annu. ReV. Phys. Chem. 1998, 49,
337. (b) Hodgkiss, J. M.; Rosenthal, J.; Nocera, D. G. In Hydrogen-
Transfer Reactions; Hynes, J. T., Klinman, J. P., Limbach, H.-H.,
Schowen, R. L., Eds.; Wiley-VCH: Weinheim, Germany, 2007; Vol.
2, pp 503-562. (c) Stubbe, J.; Nocera, D. G.; Yee, C. S.; Chang, M.
C. Y. Chem. ReV. 2003, 103, 2167. (d) Hammes-Schiffer, S. In
Hydrogen-Transfer Reactions; Hynes, J. T., Klinman, J. P., Limbach,
H.-H., Schowen, R. L. Eds.; Wiley-VCH: Weinheim, Germany, 2007;
Vol. 2, pp 479-502. (e) Meyer, T. J.; Huynh, M. H. V. Inorg. Chem.
2003, 42, 8140.
(3) Mayer, J. M. Acc. Chem. Res. 1998, 31, 441.
(4) (a) Knapp, M. J.; Meyer, M.; Klinman, J. P. In Hydrogen-Transfer
Reactions; Hynes, J. T., Klinman, J. P., Limbach, H.-H., Schowen,
R. L., Eds.; Wiley-VCH: Weinheim, Germany, 2007; Vol. 4, pp
1241-1284. (b) Knapp, M. J.; Rickert, K.; Klinman, J. P. J. Am. Chem.
Soc. 2002, 124, 3865. (c) Glickman, M. H.; Klinman, J. P. Biochem-
istry 1996, 35, 12882.
11190 Inorganic Chemistry, Vol. 46, No. 26, 2007
10.1021/ic7015726 CCC: $37.00
© 2007 American Chemical Society
Published on Web 12/01/2007

Ruthenium Bis(â-diketonato) Pyridine-Imidazole Complexes
Scheme 1. Square Scheme for Hydrogen Atom Transfer
and RuIII species could be isolated and (ii) the ability to
prepare protonated and mono-deprotonated derivatives. After
some initial efforts, which are described below, we have
developed a system with acac (2,4-pentanedionato) and 2-(2′-
pyridyl)imidazole (py-imH) ligands, such as RuII(acac)2(py-
imH) (1), in which three corners of the square have been
isolated. Py-imH is a well-known chelating ligand with a
single ionizable proton,12,13 and stable RuII/RuIII pairs with
two acac ligands have been reported, including cis-RuII-
(acac)2(MeCN)2/cis-[RuIII(acac)2(MeCN)2]OTf14,15 and RuII-
(acac)2(bpy)/[RuIII(acac)2(bpy)]OTf16 (bpy ) 2,2′-bipyridine;
transfer (ET) and proton transfer (PT) reactions as the edges
and HAT as the diagonal.2
Our group has been building an understanding of metal-
mediated HAT reactions by developing chemical systems
in which both the thermodynamics and kinetics of HAT can
be determined. Isolation of at least three corners of the square
greatly facilitates these measurements. Among the systems
we have studied, iron-tris(biimidazoline) complexes have
been particularly informative in part because the FeII-
protonated, FeIII-protonated, and FeIII-deprotonated com-
plexes are all readily isolated.6 We have examined in detail
the thermochemistry of this system, including its large
entropy for HAT reactions, oxidations of C-H and O-H
bonds, rate constants for cross and self-exchange rates, and
the agreement with the Marcus cross relation.6a,b,7
A ruthenium system is of interest to test the generality of
our HAT conclusions and to explore the analogies between
HAT and ET. ET processes of ruthenium complexes have
been studied in great detail,8 in part because the substitution-
inert nature of low-spin RuII complexes provides valuable
stability and synthetic flexibility. The groups of Hammar-
stro¨m,9 Kramer,10 Nocera,11a,b and Meyer11c have each
developed elegant ruthenium-polypyridyl systems for PCET
studies. These systems involve stable RuII complexes and
the photolytic generation of the corresponding RuIII com-
plexes. Whereas photolytic initiation of reactions is of great
value for certain measurements, these systems are also
limited by difficulties in isolating the RuIII species because
of their high reduction potentials.
-
OTf- ) triflate, CF3SO3 ). The related pyridine-imidazole
complexes with hexafluoro-acac (1,1,1,5,5,5-hexafluoro-2,4-
pentanedionato, hfac) ligand have also been prepared in this
work. Described here are the syntheses and characterization
of the compounds that comprise new ruthenium PCET
systems, together with thermochemical measurements and
preliminary studies of their HAT reactions.
Results
Syntheses. Attempts to develop a ruthenium system
analogous to the iron-tris(biimidazoline) complexes started
with the known tris(bibenzimidazole) complex [RuII(H2-
bibzim)3](ClO4)2 (H2bibzim ) 2,2′-bibenzimidazole).17 Un-
like the iron system, [RuII(H2bibzim)3](ClO4)2 appears to
doubly deprotonate upon titration with base, which made the
study of HAT reactions problematic.18 To circumvent this
issue, [RuII(bpy)2(2-(2′-pyridyl)benzimidazole)](ClO4)2, which
contains only one protonation site, was synthesized.13
However, this complex has a high RuIII/II reduction potential
(E1/2 ) 0.86 V vs Cp2Fe+/0 in MeCN19), and in our hands
isolation of the RuIII derivative was not possible.18
Following these initial efforts, we turned our attention to
Ru(acac)2 complexes of 2-(2′-pyridyl)imidazole (py-imH)
and 2-(2′-pyridyl)imidazoline (py-imnH). The latter com-
plexes have more complex HAT chemistry, as described
below, so we start here with the aromatic py-imH com-
Our design criteria for a ruthenium PCET system were
(i) suitable one-electron reduction potentials so that both RuII
14
pounds. Treatment of cis-RuII(acac)2(MeCN)2 with 1.2
(5) There are also many examples of HAT-involving metal hydride
complexes, for instance: (a) Song, J.-S.; Bullock, R. M.; Creutz, C.
J. Am. Chem. Soc. 1991, 113, 9862. (b) Edidin, R. T.; Sullivan, J. M.;
Norton, J. R. J. Am. Chem. Soc. 1987, 109, 3945.
(6) (a) Roth, J. P.; Lovell, S.; Mayer, J. M. J. Am. Chem. Soc. 2000, 122,
5486. (b) Roth, J. P.; Mayer, J. M. Inorg. Chem. 1999, 38, 2760. (c)
Burnett, M. G.; McKee, V.; Nelson, S. M. J. Chem. Soc., Dalton Trans.
1981, 1492. (d) Wang, J. C.; Bauman, J. E., Jr. Inorg. Chem. 1965, 4,
1613.
equiv of py-imH12 in C6H6 for 5 h at 80 °C under N2 forms
RuII(acac)2(py-imH) (1) as a light-brown precipitate, which
was isolated by filtration in 78% yield (eq 1). 1 is very air-
(7) (a) Roth, J. P.; Yoder, J. C.; Won, T.-J.; Mayer, J. M. Science 2001,
294, 2524. (b) Mader, E. A.; Davidson, E. R.; Mayer, J. M. J. Am.
Chem. Soc. 2007, 129, 5153. (c) Mader, E. A.; Larsen, A. S.; Mayer,
J. M. J. Am. Chem. Soc. 2004, 126, 8066. (d) Yoder, J. C.; Roth, J.
P.; Gussenhoven, E. M.; Larsen, A. S.; Mayer, J. M. J. Am. Chem.
Soc. 2003, 125, 2629.
(8) (a) Wherland, S. Coord. Chem. ReV. 1993, 123, 169. (b) Meyer, T. J.;
Taube, H. In ComprehensiVe Coordination Chemistry; Wilkinson, G.
Ed.; Pergamon: New York, 1987; Vol. 1, pp 331-384.
(9) (a) Lomoth, R.; Magnuson, A.; Sjo¨din, M.; Huang, P.; Styring, S.;
Hammarstro¨m, L. Photosynth. Res. 2006, 87, 25. (b) Sjo¨din, M.;
Styring, S.; Wolpher, H.; Xu, Y.; Sun, L.; Hammarstro¨m, L. J. Am.
Chem. Soc. 2005, 127, 3855.
(12) Hughey, J. L.; IV;, Knapp, S.; Schugar, H. Synthesis 1980, 6, 489.
(13) (a) Haga, M. Inorg. Chim. Acta 1983, 75, 29. (b) Haga, M.;
Tsunemitsu, A. Inorg. Chim. Acta 1989, 164, 137.
(14) Baird, I. R.; Cameron, B. R.; Skerlj, R. T. Inorg. Chim. Acta 2003,
353, 107.
(15) Baird, I. R.; Rettig, S. J.; James, B. R.; Skov, K. A. Can. J. Chem.
1999, 77, 1821.
(10) Cape, J. L.; Bowman, M. K.; Kramer, D. M. J. Am. Chem. Soc. 2005,
127, 4208.
(11) (a) Roberts, J. A.; Kirby, J. P.; Nocera, D. G. J. Am. Chem. Soc. 1995,
117, 8051. (b) Kirby, J. P.; Roberts, J. A.; Nocera, D. G. J. Am. Chem.
Soc. 1997, 119, 9230. (c) Facenko, C. J.; Meyer, T. J.; Thorp, H. H.
J. Am. Chem. Soc. 2006, 128, 11020.
(16) El-Hendawy, A. M.; Alqaradawi, S. Y.; Al-Madfa, H. A. Transition
Met. Chem. 2000, 25, 572.
(17) Haga, M. Inorg. Chim. Acta 1983, 77, L39.
(18) Carter, E.; Kanzelberger, M. A. Unpublished data, 2003.
(19) Potential converted from RuIII/II E1/2 ) 1.17 V versus SCE using
E
1/2(Cp2Fe+/0) ) 0.31 V versus SCE.13a
Inorganic Chemistry, Vol. 46, No. 26, 2007 11191

Wu et al.
Figure 1. ORTEP drawings of (a) RuII(acac)2(py-imH) (1), (b) [RuIII(acac)2(py-imH)]+ (2), and (c) RuIII(acac)2(py-im) (3). Hydrogen atoms are omitted
for clarity except for the N-H atom.
sensitive in solution, but less so in the solid state. The RuIII
analogue [RuIII(acac)2(py-imH)]OTf (2) was prepared simi-
larly by reacting cis-[RuIII(acac)2(MeCN)2]OTf15 with 1.2
equiv of py-imH to give a brick-red solid in 74% yield
(eq 2). The deprotonated RuIII derivative RuIII(acac)2(py-im)
Unreacted cis-RuII(hfac)2(MeCN)2 (21%) was also eluted
from the column, but extending the reaction time or
increasing the amount of py-imH (up to 10 equiv) did not
improve the yield of 4. The addition of 1 equiv of DBU
immediately deprotonates 4 in MeCN to generate [DBU-
H][RuII(hfac)2(py-im)] (5), which was isolated as a black-
purple solid in 76% yield (eq 5). Spectroscopic and X-ray
characterizations of these compounds are presented in the
next sections, and all of the structures and compound
numbers are shown in Scheme 2 below.
(3) is most easily prepared by the removal of a hydrogen
atom from 1, using 1.2 equiv of TEMPO• in MeCN at room
temperature for 10 min (eq 3). The TEMPO-H byproduct is
X-ray Structures. X-ray crystal structures of 1-5 have
been solved. ORTEP drawings of each ruthenium complex
(with 50% probability ellipsoids) are shown in Figures 1 and
2, and the crystallographic and metrical data are given in
Tables 1 and 2. The ruthenium complexes all have very
similar distorted octahedral geometries, with trans angles
>170°. The py-imH ligands form five-membered chelate
rings with small bite angles of 78.4(3)-79.6(3)°. The Ru-O
bond lengths are all quite similar for the three RuII complexes
1, 4, and 5 (2.026(7)-2.056(6) Å), independent of whether
the ligand is acac or hfac. The oxidized compounds 2 and 3
have slightly shorter Ru-O distances of 1.995(4)-2.017(4)
Å. These values are typical of related compounds.20,21 In 1,
the RuII-N(imidazole) bond is 0.029(10) Å longer than the
RuII-N(py) bond, but this is reversed in the RuIII complexes
2 and 3, where the bonds to the imidazole or imidazolate
are 0.036(7) and 0.045(9) Å shorter. This presumably reflects
the greater π-backbonding for RuII f py. This is also evident
with the more electron-deficient hfac complex 4, which has
less π-backbonding and thus has similar Ru-N(py) and Ru-
N(imidazole) distances. Deprotonation of the imidazole
removed by sublimation, and 3 is isolated as a dark-brown
solid in 65% yield. 3 can also be generated by the reaction
of 2 with 1 equiv of the base DBU (1,8-diazabicyclo[5.4.0]-
undec-7-ene) in MeCN (eq 4), but this is less convenient
for preparative scale reactions. The addition of HOTf
reprotonates 3 to form 2 (eq 4).
Related RuII-hexafluoro-acac derivatives are accessible
starting from cis-RuII(hfac)2(MeCN)2.15 Refluxing this com-
pound with 1.7 equiv of py-imH in C6H6 for 16 h yields
RuII(hfac)2(py-imH) (4), analogous to eq 1. Red-brown 4 was
obtained in 27% yield after silica gel chromatography.
11192 Inorganic Chemistry, Vol. 46, No. 26, 2007

Ruthenium Bis(â-diketonato) Pyridine-Imidazole Complexes
Figure 2. ORTEP drawings of (a) RuII(hfac)2(py-imH) (4), (b) [RuII(hfac)2(py-im)]- (5), and (c) RuII(hfac)2(py-imnH) (8) (see below), showing one of the
two independent molecules in the unit cell for each of 4 and 8. Hydrogen atoms are omitted for clarity except for the N-H atom.
Table 1. X-ray Crystallographic Data for 1-5 and 8
1‚0.5C6H6
[2]2‚CH2Cl2
3
4
5
8
empirical formula
fw
cryst syst
space group
a (Å)
b (Å)
c (Å)
R (deg)
â (deg)
γ (deg)
V (Å3)
C21H24N3O4Ru
483.50
monoclinic
P21/c
7.5358(6)
15.9357(14)
17.595(2)
90
106.260(3)
90
2028.5(3)
4
1.583
0.806
0.71073
C39H44N6O14Cl2F6S2Ru2 C18H20N3O4Ru
C18H9N3O4F12Ru
660.35
monoclinic
P21/c
9.8590(4)
19.4590(10)
23.8400(13)
90
90.390(2)
90
4573.5(4)
8
1.918
0.815
0.71073
C27H25N5O4F12Ru C18H11N3O4F12Ru
1271.96
triclinic
P1h
443.44
monoclinic
P21/c
10.2674(5)
11.4128(6)
15.8465(10)
90
90.087(3)
90
1856.89(18)
4
812.40
662.37
triclinic
P1h
monoclinic
P21/c
9.87150(10)
19.7664(3)
23.8995(4)
90
90.4656(5)
90
4663.21(12)
8
7.6240(6)
11.7300(7)
14.9360(11)
112.531(3)
95.165(3)
94.249(3)
1220.17(15)
1
10.9792(3)
12.3770(4)
12.6990(5)
89.131(1)
80.044(2)
65.729(2)
1546.43(9)
2
Z
d
calcd (g/cm3)
1.731
1.586
1.745
1.887
µ (mm-1
λ (Å)
)
0.906
0.872
0.622
0.800
0.71073
0.71073
0.71073
0.71073
cryst size (mm3)
T (K)
0.24 × 0.17 × 0.12 0.29 × 0.20 × 0.01
0.59 × 0.59 × 0.48 0.24 × 0.10 × 0.08 0.40 × 0.40 × 0.20 0.48 × 0.26 × 0.24
130(2)
130(2)
130(2)
130(2)
130(2)
130(2)
θ range (deg)
2.41-25.49
-9 e h e 9
-19 e k e 19
-21 e l e 21
6807
2.70-25.45
-9 e h e 8
-14 e k e 14
-18 e l e 18
7785
2.20-28.32
-12 e h e 13
-14 e k e 14
-17 e l e 21
11 848
2.09-28.31
-13 e h e 13
-23 e k e 25
-31 e l e 31
18 527
2.07-30.02
-15 e h e 13
-16 e k e 13
-17 e l e 14
10 219
2.06-28.29
-13 e h e 13
-26 e k e 24
-31 e l e 31
19 246
index ranges
reflns collected
unique reflns
Rint
3652
0.0992
4479
0.0520
4276
0.0688
10 787
0.1274
6954
0.0713
10 926
0.0325
params refined
267
330
239
685
470
685
R1, wR2 (I > 2σ(I)) 0.0649, 0.1649
GOF 0.984
0.0580, 0.1669
1.051
0.0653, 0.1939
1.021
0.0768, 0.1993
0.917
0.0609, 0.1440
1.041
0.0601, 0.1736
1.010
Table 2. Selected Bond Lengths (Angstroms) and Angles (Degrees) in the X-ray Structures of 1-5 and 8
1
2
3
4a
5
8a
Ru1-O1
Ru1-O2
2.031(6)
2.046(6)
2.056(6)
2.044(6)
2.004(7)
2.033(7)
94.0(2)
92.7(2)
79.2(3)
178.9(3)
175.7(2)
176.3(3)
1.997(4)
2.006(4)
1.997(4)
2.005(4)
2.063(5)
2.002(4)
2.012(4)
1.995(4)
2.017(4)
2.039(6)
1.994(6)
91.92(18)
92.97(18)
78.4(3)
2.026(7)
2.041(6)
2.042(6)
2.028(7)
2.023(8)
2.024(8)
93.4(3)
93.5(3)
79.6(3)
179.5(3)
175.9(3)
174.1(3)
2.026(3)
2.042(3)
2.039(3)
2.036(3)
2.053(4)
2.035(4)
2.036(4)
2.047(3)
2.032(3)
2.039(4)
Ru1-O3
Ru1-O4
Ru1-N1
Ru1-N2
2.027(5)
2.014(4)
2.015(5)
O1-Ru1-O2
O3-Ru1-O4
N1-Ru1-N2
O1-Ru1-O3
O4-Ru1-N1
O2-Ru1-N2
92.98(16)
90.73(17)
79.2(2)
178.87(16)
173.78(18)
175.73(17)
94.33(12)
93.50(12)
79.14(14)
178.54(12)
171.13(12)
175.34(14)
93.09(16)
92.99(13)
78.77(18)
179.71(14)
174.70(16)
173.89(16)
179.15(17)
173.4(2)
174.2(2)
a Data are for one of the two independent molecules in the unit cell.
ligand shortens its bond to ruthenium, in both 3 and 5, as is
typical for transition-metal imidazole complexes.22 The
imidazole ligands of 1, 2, 4, and 5 engage in various
hydrogen bonding interactions in the crystals. In 1 and 4,
there are intermolecular hydrogen bonds between imidazole
N3-H and acac-oxygens (dN-O ) 2.806-2.913 Å) while in
2, the imidazole N3-H bonds with the OTf- counterion
(dN-O ) 2.747 Å). The deprotonated imidazolate N3 in 5
Inorganic Chemistry, Vol. 46, No. 26, 2007 11193

Wu et al.
hydrogen bonds the acidic proton of DBU-H+ with dN-N
) 2.746 Å. These hydrogen bonding distances are within
the typical range of 2.5-3.2 Å.23
due to the paramagnetic broadening, which lowers the signal
intensity and renders even the COSY diagonal peaks
unobservable. The fourth pyridine resonances are tentatively
assigned based on their proximity in chemical shift with the
other three pyridine signals, but the other signals for 2 and
3 could not be assigned.
The correspondences between the resonances of 2 and 3
are shown by 1H NMR titration in CD3CN. The addition of
1 equiv of HOTf (pKa ) 2.60)27 in 0.1 equiv increments to
a solution of 3 gradually changes the spectrum of 3 into that
of 2, with the growth of the imidazole-NH signal (δ 5.71).
2 can be reversibly titrated back to 3 with 1 equiv of DBU
(pKa(DBU-H+) ) 24.3228). Proton exchange between 2 and
3 is thus fast on the NMR time scale, so that solutions
containing both complexes show an averaged spectrum. This
and related self-exchange reactions will be discussed in a
future publication.29
UV-vis spectra of 1, 4, and 5 all show strong MLCT
bands in the visible region (ꢀ ) 6700-11 000 M-1 cm-1,
Figure 4), as is typical of RuII-pyridyl complexes.16,20b,30,31
As expected, the trend in the lowest MLCT energies, 1 < 5
< 4, follows the ease of oxidation (below). However, the
energies for 1 and 5 are quite close despite their 0.57 V
difference in reduction potentials, and the two MLCT bands
for 1 are much more widely spaced than those for the hfac
analogue 4. The RuIII complexes 2 and 3 have much weaker
charge-transfer transitions (ꢀ ) 2000, 1600 M-1 cm-1).
Electron impact mass spectra (EI/MS) of 1, 2, and 3 are
indistinguishable in the positive-ion mode, each showing a
mass cluster peaked at 444 m/z, which matches the simulated
isotopic pattern for [Ru(acac)2(py-im)]+. Thus, 1 and 2 are
deprotonated in the process of obtaining EI/MS, and 1 is
oxidized. With electrospray ionization (ESI) in MeCN, 2 and
3 show the protonated ion [Ru(acac)2(py-imH)]+, centered
at 445 m/z. (The air-sensitivity of 1 precludes its analysis
by ESI/MS.) Positive-ion ESI/MS of 4 and 5 similarly show
an isotopic pattern, which matches the oxidized protonated
species [Ru(hfac)2(py-imH)]+ (661 m/z). 5 also shows, in
the negative-ion ESI/MS, a cluster at 660 m/z for the parent
anion [Ru(hfac)2(py-im)]-.
1
Spectroscopic Characterization. The H NMR spectra
of 1-5 in CD3CN are consistent with the solid-state
structures. For instance, the spectrum of diamagnetic 1 shows
two inequivalent acac ligands [δ(CH3) 2.05, 2.00, and 1.51
(2CH3); δ(CH) 5.32, 5.29], six pyridine-imidazole-CH
signals (δ 7.09-8.75), and an imidazole-NH peak at δ 11.31
(which was confirmed by the exchange with CD3OD). The
RuII bis-hfac complexes 4 and 5 show similar proton
resonances except for the absence of CH3 peaks and a N-H
signal for 5. The 19F NMR spectra of 4 and 5 show four
singlets between δ -74.74 and -75.06 (referenced to CF3C-
(O)OH at δ -78.50),24 consistent with the four inequivalent
CF3 groups. The 13C{1H} NMR spectra of 4 and 5 in CD3-
OD25 show resonances for pyridine-imidazole (δ 118-168)
and hfac-CH (δ 92-94) and four quartets for each of hfac-
1
CF3 (δ 117-120, JCF ) 282 Hz) and hfac-C(O) (δ 165-
2
173, JCF ) 33 Hz), again consistent with molecular C1
symmetry.
The 1H NMR spectra of paramagnetic complexes 2 and 3
(low-spin d5) in CD3CN span a wide range, from δ 6 to -65
for 2 and δ 9 to -48 for 3 (Figure 3). The four acac-methyl
resonances for 2 (δ -22 to -17) and 3 (δ -18 to -5) are
assigned on the basis of integration. The imidazole-NH
signal of 2 at δ 5.71 was identified by its exchange with
1
added CD3OD. H 2D COSY NMR spectra (Figure S1 for
2 and Figure S2 for 3, in the Supporting Information), which
have previously been useful for paramagnetic assignments,26
show cross-peaks for three of the four pyridine resonances
in each spectrum of 2 and 3 (peak 1 couples to peaks 2 and
3 in Figure 3). The other couplings were not observed, likely
(20) (a) RuII(acac)2(3-amino-6-(3,5-dimethylpyrazol-1-yl)-1,2,4,5-tetra-
zine): Nayak, A.; Patra, S.; Sarkar, B.; Ghumaan, S.; Puranik, V. G.;
Kaim, W.; Lahiri, G. K. Polyhedron, 2005, 24, 333. (b) cis- and trans-
RuII(acac)2L2, [RuIII(acac)2(L)](ClO4), and trans-[RuIII(acac)2(L)2]-
(ClO4) (L ) 2,2′-dipyridylamine): Kar, S.; Chanda, N.; Mobin, S.
M.; Urbanos, F. A.; Niemeyer, M.; Puranik, V. G.; Jimenez-Aparicio,
R.; Lahiri, G. K. Inorg. Chem. 2005, 44, 1571. (c) RuII(acac)2(o-
benzoquinonediimine) and RuII(acac)2(N-phenyl-1,2-benzoquinone-
diimine): Mitra, K. N.; Choudhury, S.; Castio`eiras, A.; Goswami, S.
J. Chem. Soc., Dalton Trans. 1998, 2901.
Thermochemical Measurements. (i) Cyclic Voltamme-
try. Cyclic voltammograms (CVs) of 1-5 in MeCN all show
chemically reversible waves. In each case, the anodic and
cathodic currents (ia and ic) are equal within 10%. The peak
separations (Ep,a - Ep,c) at a scan rate of 100 mV s-1 are
close to those of ferrocene in the same solution (80-
(21) Reported structures for Ru-hfac complexes: (a) cis-RuII(hfac)2-
(MeCN)2, cis-RuII(acac)(hfac)(MeCN)2, RuIII(hfac)3: ref 15. (b) cis-
RuII(hfac)2(CO)2: Lee, F.-J.; Chi, Y.; Liu, C.-S.; Hsu, P.-F.; Chou,
T.-Y.; Peng, S.-M.; Lee, G.-H. Chem. Vap. Deposition 2001, 7, 99
(which reports slightly longer Ru-O distances, 2.050(2)-2.081(2) Å,
compared to those in 4 and 5). (c) cis- and trans-RuIIICl2(hfac)-
(PPh3)2: Colson, S. F.; Robinson, S. D.; Robinson, P. D.; Hinckley,
C. C. Acta Crystallogr. C 1989, 45, 715.
(22) (a) Tadokoro, M.; Kanno, H.; Kitajima, T.; Shimada-Umemoto, H.;
Nakanishi, N.; Isobe, K.; Nakasuji, K. Proc. Natl. Acad. Sci. U.S.A.
2002, 99, 4950. (b) Mandon, D.; Ott-Woelfel, F.; Fischer, J.; Weiss,
R.; Bill, E.; Trautwein, A. X. Inorg. Chem. 1990, 29, 2442. (c)
Mayboroda, A.; Comba, P.; Pritzkow, H.; Rheinwald, G.; Lang, H.;
van Koten, G. Eur. J. Inorg. Chem. 2003, 1703.
(26) (a) Kennedy, D. C.; Wu, A.; Patrick, B. O.; James, B. R. Inorg. Chem.
2005, 44, 6529. (b) Belle, C.; Bougault, C.; Averbuch, M.-T., Durif,
A.; Pierre, J.-L.; Latour, J.-M. Le Pape, L. J. Am. Chem. Soc. 2001,
123, 8053. (c) Bertini, I.; Capozzi, F.; Luchinat, C.; Turano, P. J. Magn.
Reson. 1991, 95, 244.
(27) Izutsu, K. Acid-Base Dissociation Constants in Dipolar Aprotic
SolVents; Blackwell Scientific: Boston, 1990.
(28) Schwesinger, R.; Schlemper, H. Angew. Chem., Int. Ed. Engl. 1987,
26, 1167.
(23) Jeffrey, G. A. An Introduction to Hydrogen Bonding; Oxford University
Press: New York, 1997.
(24) Walstrom, A.; Pink, M.; Tsvetkov, N. P.; Fan, H.; Ingleson, M.;
Caulton, K. G. J. Am. Chem. Soc. 2005, 127, 16780: supporting
information.
(29) Wu, A.; Mayer, J. M., to be submitted.
(30) Ghumaan, S.; Sarkar, B.; Patra, S.; Parimal, K.; van Slageren, J.;
Fiedler, J.; Kaim, W.; Lahiri, G. K. J. Chem. Soc., Dalton Trans. 2005,
706.
(31) Seddon, E. A.; Seddon, K. R. The Chemistry of Ruthenium; Elsevier:
Amsterdam, The Netherlands, 1984; p 474.
(25) CD3OD is used in 13C{1H} NMR instead of CD3CN because the latter
solvent has a resonance at δ 118 that significantly overlaps with the
low-intensity quartets of hfac-CF3.
11194 Inorganic Chemistry, Vol. 46, No. 26, 2007

Ruthenium Bis(â-diketonato) Pyridine-Imidazole Complexes
Figure 3. 1H NMR spectra of (a) [RuIII(acac)2(py-imH)]OTf (2) and (b) RuIII(acac)2(py-im) (3) in CD3CN. Peaks A to D are assigned as acac-CH3
protons, peaks 1, 2, 3, and 5 as pyridine protons, and 4, 6, 7, and 8 as acac- or imidazole-CH protons. The letters and numbers show the corresponding
signals between 2 and 3, as determined by reversible NMR titration by DBU/HOTf. Solvent and impurity peaks are denoted by asterisks (*).
100 mV), but at higher scan rates the ruthenium complexes
show larger separations (up to 40 mV larger). The waves
correspond to the RuIII/II redox couples. E1/2 for 1 and 2 is at
-0.64 V, which shifts to -1.00 V upon deprotonation to
form 3 (potentials (0.01 V in MeCN referenced to internal
Cp2Fe+/0). The hfac compounds are 0.93 V more difficult to
oxidize: E1/2 ) 0.29 (4) and -0.07 V (5). In both the acac
and hfac compounds, the protonated form (1, 2, or 4) has a
higher reduction potential than the deprotonated species (3
or 5) by 0.36 V.
hfac derivatives 4 and 5 with DBU and HOTf (eq 5).
Titration of 4 with Et3N (pKa(Et3NH+) ) 18.46)27 gives a
pKa of 19.3 ( 0.1 for 4.
Hydrogen Atom Transfer Reactions of the Imidazole/
Imidazolate Complexes with TEMPO•/TEMPO-H. Com-
plex 1 in CD3CN is rapidly oxidized by 1 equiv of TEMPO•
at ambient temperatures to produce 3 and TEMPO-H32 (eq
3 above). This reaction has been monitored by optical and
1H NMR spectroscopies and is evident by the solution color
changing from the red-purple of 1 to the pale-brown of 3.
This and related hydrogen atom transfer (HAT) reactions in
the Ru(acac)2 system will be described in detail in a future
publication, including HAT self-exchange reactions and
kinetic isotope effects.29
(ii) pKa Values. The interconversion of protonated 2 and
deprotonated 3 in MeCN by 1 equiv of DBU or HOTf
(eq 4 above) was monitored by optical spectroscopy,
1
confirming the H NMR results described above. Titration
of 2 with an excess of the weak base 2,4-lutidine (pKa(2,4-
lutidine-H+) ) 14.05)27 forms 3 in an equilibrium. With
concentrations of 2 and 3 determined from optical spectra
(Figure 4), the equilibrium constant for 2 + 2,4-lutidine a
3 + (2,4-lutidine-H)OTf was determined to be 0.011 ( 0.001
from the slope of the linear plot: [3][2,4-lutidine-H+]/[2]
vs [2,4-lutidine] (Figure S3 in the Supporting Information
and in the Experimental Section). This Keq and the pKa of
2,4-lutidine-H+ give pKa(2) ) 16.0 ( 0.1. Similarly, UV-
vis monitoring shows quantitative interconversion of the RuII
The hfac analogue 4, however, does not react with 36
equiv of TEMPO• in CD3CN at room temperature under N2
after 1 d (eq 6). To understand this lack of reactivity, the
expected product, RuIII(hfac)2(py-im) (6), was generated in
Inorganic Chemistry, Vol. 46, No. 26, 2007 11195

Wu et al.
Scheme 2. Square Schemes for (a) the Ru-acac-py-imH (1-3) and
(b) the Ru-hfac-py-imH (4-6) Systems (in MeCN at 298 K, E1/2
Values vs Cp2Fe+/0).
Figure 4. UV-vis spectra of RuII(acac)2(py-imH) (1), [RuIII(acac)2(py-
imH)]OTf (2), RuIII(acac)2(py-im) (3), RuII(hfac)2(py-imH) (4), and [DBU-
H][RuII(hfac)2(py-im)] (5) in MeCN.
situ by the oxidation of 5 with 1 equiv of tri-p-tolylaminium
hexafluorophosphate ([N(tol)3]PF6, E1/2 ) 0.38 V vs
Cp2Fe+/0 33
) (eq 7). Monitoring reaction 7 by UV-vis spec-
troscopy shows an isosbestic point at 450 nm up to 1 equiv
of [N(tol)3]PF6 (Figure S4). Beyond 1 equiv, the absorbance
due to [N(tol)3]PF6 at λmax ) 668 nm (ꢀ ) 26 200 M-1
Imidazoline Complexes and Their Reactions with
TEMPO•. Prior to studying the pyridine-imidazole com-
plexes above, we explored complexes of the partially
saturated analogue, 2-(2′-pyridyl)imidazoline (py-imnH).35
Analogous to the procedures used for 1 and 4, RuII(acac)2-
(py-imnH) (7) and RuII(hfac)2(py-imnH) (8) were synthesized
from the bis(acetonitrile) derivatives (eq 8).
1
cm-1)34 grows in. By H NMR, the addition of 1 equiv of
[N(tol)3]PF6 in CD3CN causes the disappearance of the
resonances of 5, but no resonances for paramagnetic 6 are
observed. The CV of this in situ-generated 6 shows a
reversible wave with E1/2 ) -0.07 V, identical to that of 5.
Complex 6 appears to slowly decay in solution, as small
amounts of the RuII protonated complex 4 are observed by
NMR after ∼20 min at room temperature under N2. Attempts
to isolate 6 by reprecipitation with CH2Cl2/n-pentane under
N2 lead to the isolation of 4. In situ-prepared 6 reacts rapidly
with 1 equiv of TEMPO-H to quantitatively form 4 (eq 6),
1
as monitored by H NMR and UV-vis (Figure S4, in the
Supporting Information) spectroscopies. The lack of reaction
of 4 with TEMPO• thus has a thermochemical, rather than a
kinetic origin (see below).
The X-ray structure of the hfac-imidazoline complex 8
(part c of Figure 2) is similar to that of the imidazole
analogue 4, but the saturated imidazoline C-C bond (1.591-
(10) Å) is longer than the imidazole CdC bond (1.443(14)
Å). 7 has a 1H NMR spectrum analogous to that of 1 except
(32) Ref 7c: supporting information.
(33) (a) Bandlish, B. K.; Shine, H. J. J. Org. Chem. 1977, 42, 561. (b)
Eberson, L.; Larsson, B. Acta Chem. Scand., Ser. B 1986, 40, 210.
(c) Rhile, I. J.; Markle, T. F.; Nagao, H.; DiPasquale, A. G.; Lam, O.
P.; Lockwood, M. A.; Rotter, K.; Mayer, J. M. J. Am. Chem. Soc.
2006, 128, 6075 and references therein.
(34) Gould, I. R.; Ege, D.; Moser, J. E.; Farid, S. J. Am. Chem. Soc. 1990,
112, 4290.
(35) Mohammadpoor-Baltork, I.; Abdollahi-Alibeik, M. Bull. Korean Chem.
Soc. 2003, 24, 1354.
11196 Inorganic Chemistry, Vol. 46, No. 26, 2007

Ruthenium Bis(â-diketonato) Pyridine-Imidazole Complexes
that the imidazoline CH2 multiplets (δ 3.6-4.0) and NH
singlet (δ 6.12) are shifted more upfield than the aromatic
imidazole CH (δ 7-8) and NH signals (δ 11.31), as
expected. The hfac complexes 8 and 4 show the same pattern.
The 13C{1H} NMR spectrum of 8 in CD3OD shows
resonances similar to those of 4 and 5, except that the
imidazoline C-H peaks (δ 46.12, 55.58) are more upfield
than those of the imidazole (δ 121-133). Cyclic voltammetry
of the imidazoline complexes gives RuIII/II E1/2 values of
-0.68 for 7 and 0.14 V for 8. Complexes 7 and 8 are slightly
easier to oxidize than their imidazole analogues 1 and 4, by
0.04 and 0.15 V, respectively.
BDFEs are derived from the pKa and E1/2 values using eq
11, where R is the gas constant, T is temperature, and F is
BDFE ) 2.3RTpKa + FE1/2 + CG
)
[1.37pKa + 23.1E1/2 + 54.9] kcal mol-1 (11)
the Faraday constant.37 CG is the free energy for H+
+
MeCN
e- f H•MeCN. It has been given by Tilset37a as the sum of
F[E°(Cp2Fe+/0) - E°(H+/H2)] (equal to 1.2 kcal mol-1),38
the free energy of formation of H• in the gas phase [∆G°f-
(H•)g ) 48.6 kcal mol-1],39 and the free energy of solvation
of H• (∆G°solv(H•)MeCN ) 5.1 kcal mol-1).40 Thus, CG in
MeCN with potentials referenced to Cp2Fe+/0 is equal to 54.9
kcal mol-1.37 Using eq 11, this value of CG, the pKa of 2
and the E1/2 for 3 give the BDFE of the N-H bond in 1 to
be 62.0 ( 1.0 kcal mol-1 in MeCN at 298 K.41 Similarly,
the BDFE of 4 is calculated to be 79.6 ( 1.0 kcal mol-1,
using the pKa of 4 and E1/2 of 5.
The RuII-acac-imidazoline complex 7 in CD3CN reacts
slowly with 3 equiv of TEMPO• at room temperature under
N2 to give the imidazolate complex 3 and TEMPO-H
1
(eq 9), as monitored by H NMR. The formation of 3
The four outside edges of the square scheme also form a
thermochemical cycle, so the sum of these four terms (in
free-energy terms) must equal to zero (eq 12). For both the
1.37[pKa(1) - pKa(2)] + 23.1[E1/2(3) - E1/2(2)] ) 0 (12)
involves removal of three hydrogen atoms from the imida-
zoline ring, one from the NH, and one from each of the
methylene groups. Such dehydrogenation of imidazoline has
not been observed in any of the iron chemistry we have
explored,6a,b,7 but oxidation of coordinated amines is well-
known for ruthenium complexes.36 This dichotomy may be
a result of the dehydrogenation requiring MIV intermediate-
(s), which are accessible for RuIV but too high in energy for
FeIV.36 Reaction 9 does not proceed quantitatively, but with
10 equiv of TEMPO•, a yield of 72% of 3 is observed by 1H
NMR after 1 d. The hfac analogue 8 also reacts slowly with
10 equiv of TEMPO• in CD3CN at room temperature under
N2 to aromatize the imidazoline ligand, but in this case the
RuII-protonated complex 4 is formed in 50% yield after 4 d
(eq 10), with some starting 8 (14%) still remaining.
acac and hfac systems, two reduction potentials and one pKa
have been measured, so eq 12 enables calculation of the
second pKa: 22.1 ( 0.3 for 1 and 13.2 ( 0.3 for [RuIII-
(hfac)2(py-imH)]+.
II. Thermochemistry and Reactivity. The thermochemi-
cal measurements are consistent with the observed reactivity
of the ruthenium complexes with TEMPO• and TEMPO-H.
The O-H BDFE of TEMPO-H is 66.5 ( 1.1 kcal mol-1.7b
The reaction of 1 plus TEMPO• to give 3 and TEMPO-H
therefore has ∆G°3 ) -4.5 ( 0.9 kcal mol-1 [ ) BDFE(1)
- BDFE(TEMPO-H)]42 and Keq = 2 × 103. This agrees with
the experimental observation that 1 + TEMPO• proceeds to
completion, as monitored by 1H NMR (eq 3). The calculated
free energy for the reaction of hfac complex 4 with TEMPO•
(eq 6) is strongly unfavorable, ∆G°6 ) +13.1 ( 0.9 kcal
mol-1. This is consistent with the lack of observed reactivity
in the forward direction and the facile reaction in the opposite
direction: 6 + TEMPO-H f 4 + TEMPO•.
The RuII acac complexes 1 and 7 are air-sensitive because
they are reducing and have relatively weak N-H bonds.
Stirring a solution of 1 in MeCN under air produces mainly
the RuIII deprotonated complex 3. The mechanism of reaction
(37) (a) Tilset, M. In Electron Transfer in Chemistry; Balzani, V., Ed.;
Wiley-VCH: Weinheim, Germany, 2001; Vol. 2, pp 677-713. (b)
Parker, V. D.; Handoo, K. L.; Roness, F.; Tilset, M. J. Am. Chem.
Soc. 1991, 113, 7493. (c) Ref 37a gives CG ) 54.9 kcal mol-1 for
cycles in MeCN with reduction potentials referenced to Cp2Fe+/0 (p.
681), whereas ref 37b gives CG ) 53.7 kcal mol-1 for cycles in MeCN
with reduction potentials referenced to H+/H2 in MeCN.
Discussion
I. Bond Dissociation Free Energies (BDFEs) of 1 and
4. The thermochemical data above can be assembled into
square schemes2 that are thermochemical maps of the
ruthenium acac-imidazole and hfac-imidazole systems
(Scheme 2). The horizontal equilibrium arrows give the pKa
values, the verticals give the E1/2 potentials, and the diagonals
are the bond dissociation free energies (BDFEs) for HAT.
(38) Kolthoff, I. M.; Chantooni, M. K., Jr. J. Phys. Chem. 1972, 76, 2024.
(39) CRC Handbook of Chemistry and Physics, 67th ed.; Weast, R. C.,
Ed.; CRC Press: Boca Raton, FL, 1986-1987; p D-69.
(40) Brunner, E. J. Chem. Eng. Data 1985, 30, 269.
(41) The estimated error in the BDFE is predominantly from the uncertainty
in CG, which we estimate to be (1 kcal mol-1 37
.
(42) The relative error in ∆G° calculated from the difference between two
BDFEs (determined with the same CG) is less than the error of each
BDFE because the uncertainty in CG cancels itself out.
(36) Keene, F. R. Coord. Chem. ReV. 1999, 187, 121 and references therein.
Inorganic Chemistry, Vol. 46, No. 26, 2007 11197

Wu et al.
∆BDE = ∆BDFE7b). For anilines and phenols, the differ-
ences are somewhat larger, for instance ∆BDE ) 5.2 kcal
mol-1 for p-CH3C6H4NH-H versus p-CF3C6H4NH-H.48 In
general, changes that make a compound less electron rich
will raise the reduction potential and lower the pKa and
therefore balance each other in terms of the BDFE (eq 11).
Thus, the BDFE (and BDE) are less sensitive to substituent
effects than either the E1/2 or pKa. Electron or proton transfer
involves changes in charge and charge distribution, whereas
homolytic X-H bond scission is to a first approximation a
nonpolar process. This has been beautifully illustrated by
DuBois et al. for nickel and palladium hydride complexes.49
For [Pd(H)(diphosphine)2]+ complexes, varying the ligands
shifts the redox potentials by 0.30 V (equivalent to 7 kcal
mol-1), whereas the pKa values shift by 4.7 units in the
opposite direction, so that the BDFEs vary by only 0.7 kcal
mol-1.49a In the related nickel system, the shift of 0.32 V in
E1/2 is more than offset by the 7 pKa unit shift, so that the
more basic compounds have higher BDFEs by 2.0 kcal
mol-1.49b
These examples illustrate that the 17.6 ( 0.4 kcal mol-1
shift between acac and hfac ruthenium complexes described
here is particularly large. It occurs because the reduction
potentials are much more affected than the pKa values: the
∆E1/2 of 0.93 V corresponds to ∆∆G° ) 21.4 kcal mol-1
() F∆E1/2), whereas the ∆pKa of 2.8 units is only ∆∆G° )
3.8 kcal mol-1 ( ) 2.3RT∆pKa). Whereas for toluene C-H48
and [HPd(diphosphine)2]+ Pd-H49a bond strengths, F∆E1/2
and 2.3RT∆pKa are equal in magnitude; for the acac versus
Scheme 3
of 1 with O2 could proceed by an initial electron transfer to
give 2 and O2-• (E ) -0.46 V, ∆G° ) +11 kcal mol-1), by
•
initial HAT to give 3 + HO2 (∆G° = +2 kcal mol-1),43 or
by chain or base-catalyzed processes.44 The overall reaction
of 1 with O2 is favorable by roughly 18 kcal per mole of
ruthenium (Scheme 3). This is only an estimate because it
uses the gas-phase value for 1/4 O2 + H• f 1/2 H2O;43 a proper
analysis would use the value in the MeCN solution. The hfac
complexes 4 and 8, in contrast, are not air-sensitive at least
in part because their reactions with O2 are significantly less
favorable: ∆G° ) ∼0 kcal mol-1 for 4 + 1/4 O2 f 6 + 1/2
H2O.
The RuIII hfac deprotonated complex 6 has eluded isolation
because of its ease of reduction, whereas the acac analogue
3 is quite stable. 6 appears to decay at least in part due to
reactions with trace impurities in the solvents used, despite
various purification attempts. The sensitivity of 6 is appar-
ently not due to its reduction potential, which at E1/2 ) -0.07
V versus Cp2Fe+/0 is relatively modest, but rather seems to
result from its ability to form a strong N-H bond (BDFE )
79.6 kcal mol-1). We and others have been working with a
variety of hydrogen atom abstractors,45 and as a general rule
of thumb, it is often difficult to isolate species that add H•
to form a bond with a BDFE above ca. 80 kcal mol-1.
Converting this to the more commonly used bond dissocia-
tion enthalpy (BDE),7b the borderline is ca. 85 kcal mol-1.
(For a given X-H bond, the BDE in MeCN is roughly 4.6
kcal mol-1 larger than the BDFE, using the not-always-
accurate assumption that S°(X) ) S°(XH).7b)
(45) Examples of isolated hydrogen atom abstractors (X•) with relative high
X-H BDEs (ignoring entropy differences between X and HX, BDFE-
(X-H) = BDE(X-H) - 5 kcal mol-1 in organic solvents7b): (a) [FeIII-
(PY5)(OMe)]2+ (PY5 ) 2,6-bis(bis(2-pyridyl)methoxymethane)-
pyridine), BDE ) 83.5 kcal mol-1 in MeOH: Goldsmith, C. R.; Jonas,
R. T.; Stack, T. D. P. J. Am. Chem. Soc. 2002, 124, 83. (b) [MnIII-
(PY5)(OH)]2+, BDE ) 82 kcal mol-1 in MeCN: Goldsmith, C. R.;
Cole, A. P.; Stack, T. D. P. J. Am. Chem. Soc. 2005, 127, 9904. (c)
[MnIII(H31)O]2- (H31 ) tris[(N′-tert-butylureaylato)-N-ethyl)]aminato),
III. Thermochemical Comparisons. Replacing two acac
ligands with less-donating hfac ligands makes the metal less
electron rich and raises the RuIII/II reduction potential. The
difference is 0.93 V for both the protonated (1, 2 vs 4) and
deprotonated imidazole complexes (3 vs 5) and 0.82 V for
the protonated imidazoline complexes (7 vs 8). Similar
differences in E1/2 have been reported for related acac/hfac
pairs: ∆E1/2 ) 0.88 V16 for [Ru(hfac/acac)2(bpy)]+/0 and 0.99
BDE ) 77 kcal mol-1 in DMSO ([MnIV(H31)O]2- and [FeIV(H31)O]2-
,
BDEs ) 110 and 115 kcal mol-1, were not isolated.): Gupta, R.;
Borovik, A. S. J. Am. Chem. Soc. 2003, 125, 13234. (d) [RuIV(bpy)2-
(py)O]2+, BDE ) 84 kcal mol-1 in MeCN: ref. 1e and Bryant, J. R.;
Mayer, J. M. J. Am. Chem. Soc. 2003, 125, 10351. (e) MnVIIO4-, BDE
) 80 kcal mol-1 in H2O; [Mn2(µ-O)2(phen)4]3+ (phen ) 1,10-
phenanthroline), BDE ) 79 kcal mol-1 in MeCN: ref. 3. (f) [TptBu,Me
-
CrIVO(py′H)]+ (TptBu,Me ) hydrotris(3-tert-butyl-5-methylpyrazolyl)-
borate, py′H ) 3-tert-butyl-5-methylpyrazole), BDE > 75.3 kcal mol-1
in CD2Cl2: Qin, K.; Incarvito, C. D.; Rheingold, A. L.; Theopold, K.
H. J. Am. Chem. Soc. 2002, 124, 14008. (g) [FeIV(O)(N4Py)]2+ and
[FeIV(O)(Bn-tpen)]2+ (N4Py ) N,N-bis(2-pyridylmethyl)-N-bis(2-
pyridyl)methylamine, Bn-tpen ) N-benzyl-N,N′,N′-tris(2-pyridylm-
ethyl)ethylenediamine) oxidize C-H bonds of cyclohexane (BDE ≈
99.3 kcal mol-1) in MeCN: Kaizer, J.; Klinker, E. J.; Oh, N. Y.;
Rohde, J.-U.; Song, W. J.; Stubna, A.; Kim, J.; Mu¨nck, E.; Nam, W.;
Que, L., Jr. (h) tri-tert-butylphenoxyl radical, BDE ) 82.3 kcal mol-1
in DMSO: Bordwell, F. G.; Liu, W. Z. J. Am. Chem. Soc. 1996, 118,
10819.
V for cis-[Ru(hfac/acac)2(MeCN)2]+/0 15,46
The Lever param-
.
eters47 predict a change of E1/2 by 0.97 V between ruthenium
bis-acac and bis-hfac complexes, in very good agreement
with the observed ∆E1/2 for the bis-acetonitrile and imidazole
complexes, but somewhat overestimating the change for the
bpy and imidazoline species.
The acac complex 1 has a 17.6 ( 0.4 kcal mol-1 weaker
N-H bond than the hfac derivative 4.42 This is a dramatic
difference in BDFEs. For comparison, replacing CH3 for CF3
in substituted toluenes, p-CH3C6H4CH3 versus p-CF3C6H4-
CH3, shifts the benzylic C-H bond dissociation enthalpies
(BDE)48 only by 0.9 kcal mol-1 (for organic compounds,
(46) The ∆E1/2 (0.99 V) reported here is calculated from E1/2 values of
cis-[Ru(acac)2(MeCN)2]+/0 and cis-[Ru(hfac)2(MeCN)2]+/0 (-0.29 and
0.70 V vs Cp2Fe+/0) measured in this work, which agree very well
with those predicted by Lever parameters (-0.28 and 0.69 V)47 but
differ significantly from those reported in ref 15 (∆E1/2 ) 1.22 V).
(47) Lever, A. B. P. Inorg. Chem. 1990, 29, 1271.
(48) Pratt, D. A.; DiLabio, G. A.; Mulder, P.; Ingold, K. U. Acc. Chem.
Res. 2004, 37, 334.
(43) Gas-phase thermochemical data from NIST Chemistry Webbook, June
2005 release.
(44) This analysis follows that in Soper, J. D.; Rhile, I. J.; DiPasquale, A.
G.; Mayer, J. M. Polyhedron 2004, 23, 323 and references therein.
(49) (a) Raebiger, J. W.; Miedaner, A.; Curtis, C. J.; Miller, S. M.;
Anderson, O. P.; DuBois, D. L. J. Am. Chem. Soc. 2004, 126, 5502.
(b) Fraze, K.; Wilson, A. D.; Appel, A. M.; Rakowski, DuBois, M.;
DuBois, D. L. Organometallics 2007, 26, 3918-3924.
11198 Inorganic Chemistry, Vol. 46, No. 26, 2007

Ruthenium Bis(â-diketonato) Pyridine-Imidazole Complexes
1
Physical Techniques and Instrumentation. H (300 and 500
hfac complexes, F∆E1/2 is 5.6 times as large as 2.3RT∆pKa.
The disconnection between ∆E1/2 and ∆pKa is likely due to
the four CF3/CH3 groups being six bonds removed from the
N-H bond, causing little effect on the loss of the proton,
but only three bonds removed from the ruthenium center that
at least formally loses the electron.
MHz), 13C{1H} (75 and 126 MHz), and 19F (282 MHz) NMR and
1H 2D COSY spectra were recorded on Bruker Avance spectrom-
eters at room temperature, referenced to a residual solvent peak or
an external CF3C(O)OH standard (δ -78.50),24 and reported as:
δ (multiplicity, number of protons, assignment, coupling constant).
The error for NMR integration is estimated to be (10%. Electron
impact mass spectra (EI/MS) were obtained on a Kratos Profile
HV-3 direct probe instrument. Electrospray ionization mass spectra
(ESI/MS) were obtained on a Bruker Esquire-LC ion trap mass
spectrometer and reported as m/z for the most-abundant peak in a
ruthenium isotopic pattern. Samples were infused as MeCN
solutions and acquired in positive- or negative-ionization mode.
UV-vis spectra were acquired with a Hewlett-Packard 8453 diode
array spectrophotometer in anhydrous MeCN, and are reported as
Conclusions
A ruthenium acac pyridine-imidazole system has been
developed that is very well suited for the study of metal-
mediated hydrogen atom transfer. Both the RuII protonated
and RuIII deprotonated complexes, RuII(acac)2(py-imH) (1)
and RuIII(acac)2(py-im) (3), have been isolated and well
characterized, fulfilling our design criteria of suitable one-
electron reduction potential couples between protonated and
mono-deprotonated species for the RuII and RuIII states. The
reduction potential and pKa measurements indicate that the
removal of a H• from the imidazole N-H in 1 has a BDFE
of 62.0 kcal mol-1 in MeCN at 298 K, and 79.6 kcal mol-1
in RuII(hfac)2(py-imH) (4). The remarkable 17.6 kcal mol-1
difference in BDFEs is primarily due to an increase in E1/2
(0.93 V, 21.4 kcal mol-1) with small compensation from the
decrease of pKa (2.8 units, 3.8 kcal mol-1), when substituting
two acac for hfac ligands. Consistent with the BDFEs in 1
and 4, complex 1 is very rapidly oxidized by TEMPO• to
give 3 and TEMPO-H in a net HAT reaction, for which
∆G° ) -4.5 kcal mol-1. In contrast, no reaction was ob-
served between 4 and TEMPO•, consistent with a very uphill
∆G° ) +13.1 kcal mol-1, and a facile reaction occurs in
the opposite direction: RuIII(hfac)2(py-im) (6) + TEMPO-H
f 4 + TEMPO•. Detailed studies of HAT reactions with
these systems are underway, including HAT self-exchange,
kinetic isotope effects, and application of the Marcus cross
relation.
λ
max/nm (ꢀ/M-1 cm-1). CV measurements in 0.1 M (nBu4N)PF6/
MeCN were performed using a platinum disc working electrode, a
platinum wire auxiliary electrode, and a silver wire/AgNO3 refer-
ence electrode with Cp2Fe as an internal standard, and potentials
are reported versus Cp2Fe+/0 ((0.01 V). Elemental analyses were
performed by Atlantic Microlab (Norcross, GA).
RuII(acac)2(py-imH) (1). A solution of cis-RuII(acac)2(MeCN)2
(150 mg, 0.393 mmol) and py-imH (69 mg, 0.48 mmol) in C6H6
(15 mL) was stirred and heated in a 80 °C oil bath for 5 h under
N2. The solution was cooled to room temperature to yield a brown
precipitate, which was filtered by a swivel frit and dried in vacuo.
Yield: 136 mg (78%). 1H NMR (CD3CN): 1.51 (6H), 2.00 (3H),
2.05 (3H) (s, acac-CH3); 5.29, 5.32 (s, 1H each, acac-CH); 7.09
3
(t), 7.53 (t), 7.81 (d), 8.75 (d) (1H each, py-H, JHH ) 6-8 Hz);
3
7.14, 7.36 (d, 1H each, im-CH, JHH ) 2 Hz); 11.31 (s, 1H, im-
NH). An adequate 13C{1H} NMR spectrum has not been obtained
due to low solubility of 1. EI/MS: 444 [M - H]+, 401, 344 [M-
acacH]+, 300 [M - py-imH]+, 259, 247. UV-vis: 272 (27 000),
428 (6700), 568 (7000). CV: E1/2 ) -0.64 V (RuIII/II). Anal. Calcd
(Found) for C18H21N3O4Ru: C, 48.64 (48.84); H, 4.76 (4.71); N,
9.45 (9.18).
[RuIII(acac)2(py-imH)]OTf (2). A solution of cis-[RuIII(acac)2-
(MeCN)2]OTf (150 mg, 0.283 mmol) and py-imH (49 mg, 0.34
mmol) in C6H6 (15 mL) was stirred under N2 at 80 °C for 5 h. The
solution was cooled to room temperature to yield a brick-red
precipitate, which was filtered by a swivel frit. The solid was
reprecipitated with CH2Cl2/hexanes, filtered, and dried in vacuo at
78 °C. Yield: 125 mg (74%). 1H NMR (CD3CN): (all br s) -21.71
(6H), -18.88 (3H), -17.48 (3H) (acac-CH3); -64.83, -54.40,
-23.61, -8.07 (1H each, acac-CH or im-CH); -8.87, 0.07, 2.14,
3.91 (1H each, py-H); 5.71 (1H, im-NH). EI/MS: 444 [M-H]+,
401, 344 [M - acacH]+, 300 [M - py-imH]+, 259, 247.
ESI/MS+: 445 (M+); ESI/MS-: 149 (OTf-). UV-vis: 288
(20 000), 360sh (5000), 520 (2000). CV: E1/2 ) -0.64 V (RuIII/II).
Anal. Calcd (Found) for C19H21N3O7F3SRu: C, 38.45 (38.27); H,
3.57 (3.59); N, 7.08 (7.31).
RuIII(acac)2(py-im) (3). A solution of 1 (200 mg, 0.450 mmol)
and TEMPO• (84 mg, 0.54 mmol) in MeCN (30 mL) was stirred
under N2 for 10 min at room temperature. The solvent was removed
under vacuum, and the residue was sublimed for 16 h with a vacuum
cold finger apparatus to remove TEMPO-H. The product was
reprecipitated with CH2Cl2/hexanes to yield a dark-brown solid,
which was filtered and dried in vacuo at 78 °C. Yield: 130 mg
Experimental Section
Materials. All of the reagent grade solvents were purchased from
Fisher Scientific, EMD Chemicals, or Honeywell Burdick &
Jackson (for anhydrous MeCN). Various efforts to purify MeCN,
including treatments with CaH2/P2O5 and various oxidants, have
only decreased the stability of strongly oxidizing materials in MeCN
(perhaps due to amine impurities). Therefore, the high-purity
Burdick & Jackson MeCN was simply sparged with N2 and piped
from a steel keg directly into a glove box. Deuterated solvents were
obtained from Cambridge Isotope Laboratories. CD3CN was dried
over CaH2, vacuum transferred to P2O5, then over to CaH2, and
then to an empty glass vessel. DBU, 2,4-lutidine, TEMPO•, and
(nBu4N)PF6 were purchased from Aldrich, HOTf from Acros, and
Et3N from Fisher. Et3N was distilled from KOH and then dried
over CaH2.50 TEMPO• was sublimed onto a cold finger. (nBu4N)-
PF6 was recrystallized from EtOH before use. cis-RuII(acac)2-
(MeCN)2,14 cis-[RuIII(acac)2(MeCN)2]OTf,15 cis-RuII(hfac)2(MeCN)2,15
py-imH,12 py-imnH,35 TEMPO-H,32 and [N(tol)3]PF6 were pre-
33
pared according to literature procedures. All of the reactions were
performed in the absence of air, using glove box/vacuum line
techniques unless otherwise noted.
1
(65%). H NMR (CD3CN): (all br s) -17.58, -15.52, -11.00,
-5.09 (3H each, acac-CH3); -47.33, -39.08, -21.31, -19.45
(1H each, acac-CH or im-CH); -8.56, -4.46, -2.95, 8.75 (1H
each, py-H). EI/MS: 444 (M+), 401, 344 [M - acac]+, 300 [M -
py-im]+, 259, 247. ESI/MS+: 445 [M + H]+. UV-vis: 286
(50) Armarego, W. L. F.; Chai, C. L. L. Purification of Laboratory
Chemicals 5th Ed.; Butterworth-Heinemann: Amsterdam, The Neth-
erlands, 2003.
Inorganic Chemistry, Vol. 46, No. 26, 2007 11199

Wu et al.
(19 000), 331sh (13 000), 486 (1600). CV: E1/2 ) -1.00 V (RuIII/II).
Anal. Calcd (Found) for C18H20N3O4Ru‚0.2H2O: C, 48.36 (47.89);
by UV-vis titration (Figure S4 in the Supporting Information).
Inside a glove box, solutions of 5 (0.053 mM), [N(tol)3]PF6 (2.7
mM), and TEMPO-H (2.7 mM) in MeCN were prepared at room
temperature. An aliquot of 5 (2.5 mL) in a UV-vis cuvette was
titrated with 0.1 equiv (5 µL) increments of [N(tol)3]PF6 until 1
equiv, as 6 was generated. UV-vis of 6: 455 (4700), 508 (3400).
The solution was further titrated with increments of 0.1 equiv (5
µL) of TEMPO-H until 1 equiv, as 4 was produced. The yield for
4 was 100 ( 10% on the basis of starting 5. CV: E1/2 ) -0.07 V
(RuIII/II) for 6, generated from 5 (2.5 mM, 2.0 mL) + 1 equiv of
[N(tol)3]PF6 (62 mM, 80 µL) in MeCN.
1
H, 4.60 (4.51); N, 9.40 (9.35); H NMR spectra of 3 in CD3CN
typically show ∼0.2 equiv of H2O per ruthenium although an NMR
spectrum of the batch sent for elemental analysis was not obtained.
RuII(hfac)2(py-imH) (4). A solution of cis-RuII(hfac)2(MeCN)2
(1000 mg, 1.67 mmol) and py-imH (420 mg, 2.89 mmol) in C6H6
(50 mL) was refluxed for 16 h under air. The solvent was removed
on a rotary evaporator, and the residue was loaded onto a silica
gel column and eluted with 9:1 CH2Cl2/CH3OH. The first brown
fraction was unreacted cis-RuII(hfac)2(MeCN)2 (207 mg, 21%), and
4 was isolated as the second red-brown fraction, which was rotary
evaporated to dryness, reprecipitated with CH2Cl2/hexanes, filtered,
and dried in vacuo at 78 °C. Yield: 298 mg (27%). 1H NMR (CD3-
CN): 6.20 (s, 2H, hfac-H); 7.42 (t), 7.97 (t), 8.09 (d), 8.48 (d) (1H
each, py-H, 3JHH ) 6-8 Hz); 7.22, 7.49 (d, 1H each, im-CH, 3JHH
) 2 Hz); 11.82 (s, 1H, im-NH). 19F NMR (CD3CN): -75.06,
-75.04, -74.99, -74.94 (s, hfac-CF3). 13C{1H} NMR (CD3OD):
92.81, 93.00 (hfac-CH); 117.84, 117.86, 119.02, 119.09 (q, hfac-
CF3,1JCF ) 282 Hz); 120.83, 124.81, 137.99, 153.41 (py-CH);
121.53, 132.38 (im-CH); 149.67, 153.26 (py-N-C-C-N-im);
168.94, 169.10, 172.35, 172.53 (q, hfac-C(O), 2JCF ) 33 Hz). ESI/
MS+: 661 (M+). UV-vis: 291 (26 000), 481sh (9600), 519
(10 000). CV: E1/2 ) 0.29 V (RuIII/II). Anal. Calcd (Found) for
C18H9F12N3O4Ru: C, 32.74 (32.80); H, 1.37 (1.38); N, 6.36 (6.54).
[DBU-H][RuII(hfac)2(py-im)] (5). DBU (25 µL, 0.165 mmol)
was added to a red-brown solution of 4 (109 mg, 0.165 mmol) in
MeCN (10 mL) under air to immediately generate a dark-purple
solution, which was rotary evaporated to dryness. The residue was
reprecipitated with CH2Cl2/hexanes to yield a black-purple solid,
which was filtered and dried in vacuo at 78 °C. Yield: 102 mg
(76%). 1H NMR (CD3CN): 1.70 (m, 6H), 1.98 (quintet, 2H), 2.60
(m, 2H), 3.30 (t, 2H), 3.46 (t, 2H), 3.52 (m, 2H) (DBU-H+); 6.11,
6.12 (s, 1H each, hfac-H); 7.06 (t), 7.71 (t), 7.87 (d), 8.21 (d) (1H
each, py-H, 3JHH ) 6-8 Hz); 6.99, 7.19 (d, 1H each, im-CH, 3JHH
) 2 Hz). 19F NMR (CD3CN): -75.06, -74.91, -74.88, -74.74
(s, hfac-CF3). 13C{1H} NMR (CD3OD): 20.43, 24.94, 27.49, 29.96,
33.78, 39.42, 49 (overlapped with CD3OD), 55.36 (DBU-CH2);
93.38, 93.47 (hfac-CH); 118.23, 119.74, 119.58 (q, hfac-CF3,1JCF
) 282 Hz), the fourth quartet is obscured by overlapping with py-
CH; 118.56, 121.18, 137.01, 152.09 (py-CH); 129.80, 132.11 (im-
CH); 155.76, 167.45 (py-N-C-C-N-im); 158.29 (DBU-NdC-
RuII(acac)2(py-imnH) (7). Complex 7 was synthesized analo-
gous to 1 using cis-RuII(acac)2(MeCN)2 (200 mg, 0.52 mmol) and
py-imnH (93 mg, 0.63 mmol) and was isolated as a black-green
1
powder. Yield: 121 mg (52%). H NMR (CD3CN): 1.55 (3H),
1.60 (3H), 2.00 (6H) (s, acac-CH3); 3.6-4.0 (m, 4H, imn-CH),
5.27, 5.31 (s, 1H each, acac-CH); 6.12 (s, 1H, imn-NH); 7.12 (t),
3
7.49 (t), 7.61 (d), 8.74 (d) (1H each, py-H, JHH ) 6-8 Hz). An
adequate 13C{1H} NMR spectrum has not been obtained because
of low solubility of 7. EI/MS: 447 [M]+, 348 [M - acac]+, 300
[M - py-imnH]+, 282, 276, 260, 248. UV-vis: 274 (24 000), 428
(6900), 610 (7700). CV: E1/2 ) -0.68 V (RuIII/II). Anal. Calcd
(found) for C18H23N3O4Ru: C, 48.42 (48.13); H, 5.19 (5.26); N,
9.41 (9.38).
RuII(hfac)2(py-imnH) (8). Complex 8 was synthesized analogous
to 4 except using cis-RuII(hfac)2(MeCN)2 (200 mg, 0.33 mmol) and
py-imnH (99 mg, 0.67 mmol) and was isolated as a brown-purple
powder. Yield: 62 mg (28%). 1H NMR (CD3CN): 3.72 (1H), 3.85
(1H), 4.00 (2H) (m, imn-CH); 6.17, 6.20 (s, 1H each, hfac-H);
6.90 (s, 1H, imn-NH); 7.50 (t), 7.95 (t), 7.97 (d), 8.54 (d) (1H each,
3
py-H, JHH ) 6-8 Hz). 19F NMR (CD3CN): -75.14, -75.11,
-75.01, -74.83 (s, hfac-CF3). 13C{1H} NMR (CD3OD): 46.12,
55.58 (imn-CH); 92.86 (both hfac-CH); 117.89, 117.96, 119.04,
119.12 (q, hfac-CF3,1JCF ) 282 Hz); 125.02, 127.16, 137.45, 153.84
(py-CH); 152.92, 169.34 (py-N-C-C-N-imn); 168.22, 169.59,
172.25, 172.50 (q, hfac-C(O), 2JCF ) 33 Hz). ESI/MS+: 663 (M+).
UV-vis: 225 (6200), 269 (5300), 289 (5300), 484sh (4200), 524
(5300). CV: E1/2 ) 0.14 V (RuIII/II). Anal. Calcd (Found) for
C18H11F12N3O4Ru: C, 32.64 (32.82); H, 1.67 (1.67); N, 6.34 (6.30).
1H NMR Titration of 2 and 3. Stock solutions were prepared
for DBU (111 mM, 16.9 mg in 1 mL) and HOTf (111 mM, 16.7
mg in 1 mL) in CD3CN. A solution of 3 in an NMR tube (11 mM,
2.5 mg in 0.5 mL CD3CN) was titrated to 2 by adding 1 equiv of
HOTf in 0.1 equiv (5 µL) increments. 1H NMR spectra were
recorded initially and after each addition of HOTf. Each peak in
the spectra was a weighted average of the corresponding peaks for
2 and 3, indicating fast proton exchange on the NMR time scale.
The reverse titration, adding 1 equiv of DBU in 0.1 equiv (5 µL)
2
N); 165.70, 166.68, 170.13, 170.83 (q, hfac-C(O), JCF ) 33 Hz).
ESI/MS+: 661 [M + H]+, 153 (DBU-H+); ESI/MS-: 660 (M-).
UV-vis: 292 (20 000), 472 (8300), 564 (11 000). CV: E1/2
)
-0.07 V (RuIII/II). Anal. Calcd (Found) for C27H25F12N5O4Ru: C,
39.91 (39.92); H, 3.10 (3.11); N, 8.62 (8.74).
In Situ Generation of RuIII(hfac)2(py-im) (6) and Reaction
of 6 + TEMPO-H. In a N2 glove box, solutions of 5 (2.5 mM, 4.0
mg in 2.0 mL), [N(tol)3]PF6 (61.5 mM, 26.6 mg in 1 mL), and
TEMPO-H (123 mM, 38.7 mg in 2.0 mL) in CD3CN were prepared
at room temperature. A trace amount of (Me3Si)2O was added to
the solution of 5 as internal standard. Each of three J-Young NMR
tubes were filled with 0.5 mL of the solution of 5. To tubes 2 and
3 was added 1 equiv of [N(tol)3]PF6 (20 µL), with immediate color
changes from purple-red to pale-brown 6. After mixing tube 3 well,
1 equiv of TEMPO-H (10 µL) was added, giving an immediate
1
increments, was also monitored by H NMR.
UV-Vis Titration of 2 and pKa Determination. Stock solutions
were prepared for 2 (0.11 mM), DBU (6.5 mM), and HOTf
(6.5 mM) in MeCN. An aliquot of 2 (3.0 mL, 0.11 mM) was
transferred to a UV-vis cuvette and was titrated with increments
of 0.1 equiv (5 µL) of DBU. UV-vis spectra were recorded for
the initial 2 and after each addition of DBU. A total of 1.3 equiv
of DBU was added, but the spectrum stopped changing after 1.0
equiv, showing a stoichiometric conversion to the deprotonated 3.
The titration was reversible, and protonated 2 was regenerated
stoichiometrically by 1 equiv of HOTf, by adding 0.1 equiv (5 µL)
increments.
1
color change to red-brown 4. The H NMR spectrum of tube 2
after ∼20 min showed resonances of DBU-H+, N(tol)3 [δ 2.27
(s, 9H, CH3); 6.87, 7.07 (d, 6H each, Ar-H, 3JHH ) 7 Hz)], and a
trace of 4; paramagnetic 6 was not observed. The 1H NMR spectrum
of tube 3 showed 100 ( 10% yield for 4, on the basis of the
integration of starting 5. The generation of 6 was also monitored
A stock solution of 2,4-lutidine (647 mM) in MeCN was
prepared, and was serial diluted twice to make two other solutions
(64.7 and 6.47 mM). An aliquot of 2 (3.0 mL, 0.11 mM) was
11200 Inorganic Chemistry, Vol. 46, No. 26, 2007

Ruthenium Bis(â-diketonato) Pyridine-Imidazole Complexes
transferred to a UV-vis cuvette and was titrated with increments
of 0.1 equiv (5 µL) of 2,4-lutidine (6.47 mM) until 2.0 equiv. The
titration was continued by adding 1 equiv (5 µL) of 64.7 mM base
until 20 equiv and then with 10 equiv (5 µL) of 647 mM base until
200 equiv. UV-vis spectra were recorded for the initial 2 and after
each addition of 2,4-lutidine. The UV-vis data were analyzed using
the absorbance at 340 nm, yielding [3]/[2] ) (A - A2)/(A3 - A),
where A2 and A3 are the absorbances for pure 2 and 3 at 340 nm:
[3] ) [2,4-lutidine-H+] ) {((A - A2)/(A3 - A2)) × [Ru]total} and
[2,4-lutidine] ) [2,4-lutidine]total - [2,4-lutidine-H+] ) [2,4-
lutidine]total - {((A - A2)/(A3 - A2)) × [Ru]total}. Plotting [3][2,4-
lutidine-H+]/[2] versus [2,4-lutidine] yielded a straight line (Figure
S3 in the Supporting Information), whose slope is Keq, ) 0.011 (
0.001 for 2 + 2,4-lutidine a 3 + (2,4-lutidine-H)OTf. The pKa of
X-ray Structural Determinations. Crystals of 1 were grown
from the slow evaporation of MeCN/C6H6 solutions inside a N2
glove box. Crystals of 2-5 and 8 were grown by vapor diffusion
of Et2O/hexanes to CH2Cl2 solutions of the complex under air. The
crystals were mounted onto glass capillaries with oil. The data were
collected on a Nonius Kappa CCD diffractometer. The data were
integrated and scaled using hkl-SCALEPACK.51 This program
applies multiplicative correction factor (S) to the observed intensities
(I) and has the following form: S ) exp(-2B(sin2θ)/λ2)/scale. S is
calculated from the scale, and B factor determined for each frame
and is then applied to I to give the corrected intensity (Icorr). Solution
by direct methods (SIR97) produced a complete heavy-atom phasing
model consistent with the proposed structure.52 All of the hydrogen
atoms were located using a riding model. All of the non-hydrogen
atoms were refined anisotropically by full-matrix least-squares
(SHELXL-97).53 Half of a solvent molecule per Ru is found in the
unit cells of 1 (0.5 C6H6) and 2 (0.5 CH2Cl2). In the structures of
4 and 8, each of the unit cells contains two independent ruthenium
complexes. The structure of 5 contains a disordered CF3 group,
with major F1, F2, and F3 and minor F1A, F2A, and F3A
components (Figure S5 in the Supporting Information); only the
major fluorine atoms are shown in part b of Figure 2. The major to
minor occupancy was modeled as 80 and 20%, and the thermal
ellipsoids for minor components F1A, F2A, and F3A were
restrained during refinement.
2 was calculated from pKa(2) ) pKa(2,4-lutidine-H+) - log Keq
)
16.0 ( 0.1 using the known pKa of 14.05 for 2,4-lutidine-H+.27
UV-Vis Titration of 4 and pKa Determination. Following the
procedure above, 3.0 mL of a 0.033 mM solution of 4 was titrated
with DBU (19.6 mM) and then with HOTf (19.7 mM) (all in
MeCN). One equiv of DBU completely converted 4 to 5, which
was converted back to 4 by 1 equiv of HOTf. Again, following the
procedure above, 4 (3.0 mL, 0.030 mM) was titrated with solutions
of Et3N, adding 0.1 equiv (5 µL) of Et3N (1.78 mM) until 2.0 equiv,
then adding 1 equiv (5 µL) of Et3N (17.8 mM) until 20 equiv. UV-
vis spectra were recorded for the initial 4 and after each addition
of Et3N, and the data were analyzed using the absorbance at 565
nm. The plot of [5][Et3NH+]/[4] versus [Et3N] yielded a straight
line with slope Keq ) 0.14 ( 0.01. The pKa of 4 is given by pKa(4)
) pKa(Et3NH+) - log Keq ) 19.3 ( 0.1 using pKa ) 18.46 for
Et3NH+.27
Acknowledgment. We would like to thank the U.S.
National Institutes of Health (GM50422) for financial support
of this work. J.M. also thanks Ithaca College Dana Internship
Program for a summer fellowship. We acknowledge Mr. Eric
Carter and Dr. Mira Kanzelberger for preliminary studies
of benzimidazole systems, Ms. Elizabeth Mader for valuable
discussions, and Mr. Loren Kruse for assistance with mass
spectrometry.
1H NMR Reactions with TEMPO•. Many reactions were
1
monitored by H NMR in sealable J-Young tubes. In a typical
procedure, solutions of 1 (1.8 mM, 1.6 mg in 2 mL) and TEMPO•
(90 mM, 28.0 mg in 2 mL) were prepared in CD3CN in a N2 glove
box. A trace of (Me3Si)2O was added to the solution of 1 as an
internal standard. Each of the two J-Young tubes was charged with
0.5 mL of the solution of 1. TEMPO• (1 equiv, 10 µL) was added
to one of the tubes, accompanied by an instant color change from
red-purple to pale-brown. 1H NMR spectra of 1 and 1 + TEMPO•
were recorded after ∼20 min, the latter showing the product yield
for 3 (86%) and TEMPO-H (98%) [TEMPO-H: δ 1.06 (s, 12H,
CH3), 1.45 (s, 6H, CH2), 5.34 (s, 1H, OH)].32 The 1H NMR
spectrum of 4 (3.0 mM, 0.5 mL) and 36 equiv of TEMPO• (8.5
mg) in CD3CN showed resonances only for 4 and TEMPO• at room
temperature after 1 d [TEMPO•: δ -29.74 (4H, 3,5-CH2), -16.51
(12H, CH3), 15.33 (2H, 4-CH2) (all br s)].
Supporting Information Available: Crystallographic data for
1-5 and 8 in CIF format and Figures S1-S5. This material is
available free of charge via the Internet at http://pubs.acs.org.
IC7015726
(51) Otwinowski, Z.; Minor, W. In Methods in Enzymology; Carter, C.
W., Jr., Sweet, R. M., Eds.; Academic Press: New York, 1997; Vol.
276, pp 307-326.
(52) Altomare, A.; Cascarano, G.; Giacovazzo, C.; Guagliardi, A.; Burla,
M. C.; Polidori, G.; Camalli, M. J. Appl. Crystallogr. 1994, 27, 435.
(53) Sheldrick, G. M. SHELXL-97: Program for the Refinement of Crystal
Structures; University of Go¨ttingen: Go¨ttingen, Germany, 1997.
Inorganic Chemistry, Vol. 46, No. 26, 2007 11201