Oxidation of dihydrogen by iridium complexes of redox-active ligands

Article abstract of DOI:10.1021/om9010593

Unsaturated organoiridium complexes were prepared with amidophenolate ligands derived from 2-(2-trifluoromethyl)amino-4,6-di-tert-butylphenol (H ₂^tBA^F) and 2-tert-butylamino-4,6-di-tert- butylphenol (H₂^tBA^tBu). The following 16e complexes were characterized: Cp*M(^tBA^R) with M = Ir (1^F and 1^t-Bu), Rh (2^F), and (cymene)Ru( ^tBA^F) (3^F). These complexes undergo two 1e oxidations at potentials of about 0 and -0.25 V vs Cp₂Fe ^0/+. The magnitude of δE_1/2 is sensitive to the counteranions, and the reversibility is strongly affected by the presence of Lewis bases, which stabilize the oxidized derivatives. Crystallographic measurements indicate that upon oxidation the amidophenolate ligands adopt quinoid character, as indicated by increased alternation of the C-C bond lengths in the phenylene ring backbone and shortened C-N and C-O bonds. Unlike the charge-neutral precursors, the cationic [Cp*M(^tBA ^R)]⁺ are Lewis acidic and form well-characterized adducts with PR₃ (R = Me, Ph), CN^-, MeCN (reversibly), and CO. In the absence of competing ligands, the cations oxidize H₂. Coulommetry measurements indicate that H₂ is oxidized by the monocations [Cp*M(^tBA^R)]⁺, not the corresponding dications. Oxidation of H₂ is catalytic in the presence of a noncoordinating base at potentials required for the generation of [Cp*M(^tBA^R)]⁺. The rate decreases in the order [Cp*M(^tBA^F)]BAr^F₄ > [Cp*M(^tBA^F)]PF₆ > [Cp*M( ^tBA^t-Bu)]PF₆. The reduction of ferrocenium by H₂ is catalyzed by Cp*M(^tBA^R).

Full text of DOI:10.1021/om9010593

1956 Organometallics 2010, 29, 1956–1965
DOI: 10.1021/om9010593
Oxidation of Dihydrogen by Iridium Complexes of Redox-Active Ligands
Mark R. Ringenberg, Mark J. Nilges,‡ Thomas B. Rauchfuss,* and Scott R. Wilson
School of Chemical Sciences, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801.
‡ The Illinois EPR Research Center, School of Molecular & Cell Biology
Received December 10, 2009
Unsaturated organoiridium complexes were prepared with amidophenolate ligands derived from
2-(2-trifluoromethyl)amino-4,6-di-tert-butylphenol (H2 BAF) and 2-tert-butylamino-4,6-di-tert-bu-
t
tylphenol (H2 BAtBu). The following 16e complexes were characterized: Cp*M(tBAR) with M = Ir
t
(1F and 1t-Bu), Rh (2F), and (cymene)Ru(tBAF) (3F). These complexes undergo two 1e oxidations at
potentials of about 0 and -0.25 V vs Cp2Fe0/þ. The magnitude of ΔE1/2 is sensitive to the
counteranions, and the reversibility is strongly affected by the presence of Lewis bases, which
stabilize the oxidized derivatives. Crystallographic measurements indicate that upon oxidation the
amidophenolate ligands adopt quinoid character, as indicated by increased alternation of the C-C
bond lengths in the phenylene ring backbone and shortened C-N and C-O bonds. Unlike the
charge-neutral precursors, the cationic [Cp*M(tBAR)]þ are Lewis acidic and form well-characterized
adducts with PR3 (R = Me, Ph), CN-, MeCN (reversibly), and CO. In the absence of competing
ligands, the cations oxidize H2. Coulommetry measurements indicate that H2 is oxidized by the
monocations [Cp*M(tBAR)]þ, not the corresponding dications. Oxidation of H2 is catalytic in
the presence of a noncoordinating base at potentials required for the generation of [Cp*M(tBAR)]þ.
The rate decreases in the order [Cp*M(tBAF)]BArF4 > [Cp*M(tBAF)]PF6 > [Cp*M(tBAt-Bu)]PF6.
The reduction of ferrocenium by H2 is catalyzed by Cp*M(tBAR).
Introduction
especially those associated with O2 processing.10,11
The advantages of redox-active ligands for the catalytic
transformations of small molecules have been demonst-
rated.12-14
Many redox-active noninnocent ligands (NILs) are
based on cis-disubstituted-1,2-alkene and 1,2-phenylene, i.e.,
X-C(R)dC(R)-X. These complexes include the dithiolenes
(X = S), diimines (X = NR), and catecholates (X = O).15-21
In this work, we are interested in converting M-NIL ensem-
bles into stable cationic Lewis acids. One problem we faced is
that conventional NILs typically oxidize to cations at very
high potentials (most redox interconversions involve anions).
Furthermore, complexes of many NILs tend to dimerize
Complexes of redox-active ligands have long attracted the
attention of inorganic chemists.1-6 Whereas metal-centered
redox dramatically alters the bonding of the metal-ligand
ensemble, changes in the redox state of the ligands have
subtler affects on the metal center.7-9 Redox-active ligands
may lead to faster catalysts for the reaction of small inor-
ganic molecules (e.g., H2, N2, O2) because reorganizational
barriers for redox-active organic substituents are typically
lower than for inorganic centers. Furthermore, redox-active
ligands supplement the electrons transferred from the metal,
which can facilitate reactions that require multielectron
transfers. Redox-active ligands are pervasive in enzymes,
(11) Jazdzewski, B. A.; Tolman, W. B. Coord. Chem. Rev. 2000,
200-202, 633–685.
(12) Knijnenburg, Q.; Gambarotta, S.; Budzelaar, P. H. M. Dalton
Trans. 2006, 5442–5448.
(13) Ben-Ari, E.; Leitus, G.; Shimon, L. J. W.; Milstein, D. J. Am.
*Corresponding author. E-mail: rauchfuz@uiuc.edu.
(1) Allgeier, A. M.; Mirkin, C. A. Angew. Chem., Int. Ed. 1998, 37,
895–908.
(2) Butin, K. P.; Beloglazkina, E. K.; Zyk, N. V. Russ. Chem. Rev.
Chem. Soc. 2006, 128, 15390–15391.
2005, 74, 531–553.
(14) Frech, C. M.; Ben-David, Y.; Weiner, L.; Milstein, D. J. Am.
Chem. Soc. 2006, 128, 7128–7129.
(15) Fourmigue, M. Coord. Chem. Rev. 1998, 178-180, 823–864.
(16) Stiefel, E. I. Dithiolene Chemistry, Synthesis, Properties, and
Applications; Prog. Inorg. Chem.; John Wiley: Hoboken, NJ, 2004.
(17) Hendrickson, D. N.; Pierpont, C. G. Valence Tautomeric Tran-
sition Metal Complexes. In Spin Crossover in Transition Metal Com-
pounds II; Springer: Berlin/Heidelberg, Germany, 2004; pp 786-786.
(18) Evangelio, E.; Ruiz-Molina, D. Eur. J. Inorg. Chem. 2005, 2957–
2971.
(3) De Bruin, B.; Hetterscheid, D. G. H.; Koekkoek, A. J. J.;
Gruetzmacher, H. Prog. Inorg. Chem. 2007, 55, 247–354.
(4) Kaim, W.; Schwederski, B. Pure Appl. Chem. 2004, 76, 351–364.
(5) Chakraborty, S.; Laye, R. H.; Paul, R. L.; Gonnade, R. G.;
Puranik, V. G.; Ward, M. D.; Lahiri, G. K. J. Chem. Soc., Dalton Trans.
2002, 1172–1179.
(6) Fenske, D. Ber. 1979, 112, 363–375.
(7) Klein, R. A.; Elsevier, C.; Hartl, F. Organometallics 1997, 16,
1284–1291.
ꢀ
(8) Hartl, F.; Stufkens, D. J.; Vlcek, A. Inorg. Chem. 1992, 31, 1687–
1695.
(9) Lorkovic, I. M.; Wrighton, M. S.; Davis, W. M. J. Am. Chem. Soc.
1994, 116, 6220–6228.
(10) Stubbe, J.; van der Donk, W. A. Chem. Rev. 1998, 98, 705–762.
(19) Poddel’sky, A. I.; Cherkasov, V. K.; Abakumov, G. A. Coord.
Chem. Rev. 2009, 253, 291–324.
(20) Pierpont, C. G.; Lange, C. W. Prog. Inorg. Chem. 1993, 41,
331-442.
(21) Moussa, J.;Amouri, H. Angew. Chem., Int. Ed. 2008, 47, 1372–1380.
r
pubs.acs.org/Organometallics
Published on Web 03/24/2010
2010 American Chemical Society

Article
Scheme 1. Proton-Induced Lewis Acidity versus Oxidation-
Induced Lewis Acidity of a 16e Metal Complex
Organometallics, Vol. 29, No. 8, 2010 1957
Scheme 2. Oxidation of H2 by a Metal Complex via a Hetero-
lytic Process
upon oxidation to neutral or cationic states, quenching any
coordinative unsaturation.22 The N-substituted 2-amido-
phenolates popularized by Wieghardt et al. alleviate the
problems of harsh redox potentials and dimerization.23-26
Such complexes undergo redox at potentials around -0.3 V
vs Cp2Fe0/þ. Furthermore the amidophenolates oxidize
to give complexes that remain monomeric by virtue of
their bulky substituents. The amidophenolates are easily
prepared and derivatized. For example, catechol is effici-
ently converted to a 3,5-di-tert-butyl derivative,28 and the
resulting disubstituted diol then reacts with primary
amines to give a range of aminophenols.29 We have demon-
strated how ligand-centered oxidation of a Pt-amido-
phenolate-olefin complex induces nucleophilic attack at the
olefin.27
Given their attractive properties, the amidophenolates
are promising candidates for reactions wherein ligand-
centered oxidation enhances the Lewis acidity of an un-
saturated metal center. Metal-centered oxidation is known
to cause increases in coordination number at the metal. The
subtle effects of ligand-centered oxidation are, however,
more applicable to catalytic processes where dramatic
changes at the metal can lead to large kinetic barriers.
The concept of oxidation-induced Lewis acidity (OI-LA)
is related to our previously studied protonation-induced
Lewis acidity (PI-LA, see Scheme 1), which was demonstrated
in the case of 16e- diamido iridium(III) complexes.30,31
Like protonation-induced Lewis acidity, oxidation-induced
Lewis acidity is expected to be operative when the start-
ing complex is formally coordinatively unsaturated.32 Since
the π-donor orbitals on the amidophenolates are also
the redox-active orbitals (HOMO),19,24,33 ligand-centered
oxidation would be expected to enhance the Lewis acidity at
the metal by simultaneously raising the formal positive
charge on the complex and weakening the metal-ligand
π-bonds.
In this project we apply oxidation-induced Lewis acidity to
the oxidation of dihydrogen. The oxidation of dihydrogen by
soluble metal complexes has long been studied,34,35 and in
recent years detailed mechanistic information has become
available.36,37 Recognizing that H2 oxidation begins with the
formation of dihydrogen complexes,38 we sought complexes
that would give H2 adducts susceptible to deprotonation.
Dihydrogen complexes generally resist oxidation, but the
corresponding hydrides are redox-active.39-41 Redox-active
complexes of H2/H- complexes are operative in the [NiFe]-
and [FeFe]-hydrogenases, and the mechanism in Scheme 2
shows the general pattern of their reactions.42-44
We have previously communicated that the reduction of
Agþ by H2 is catalyzed by amidophenolates of Cp*Ir(III).45
In this paper, we demonstrate the scope of oxidation-induced
Lewis acidity to other ligands, and we further probe the
proposed pathway for H2 activation.
Results
Preparation of Ir(III), Rh(III), and Ru(III) Amidopheno-
lates. The complexes Cp*Ir(tBAt-Bu) (1t-Bu), Cp*Ir(tBAF)
(1F), Cp*Rh(tBAF) (2F), Cp#Ir(tBAF) (Cp# = tetra-
methylcyclopentadienyl), and (cymene)Ru(tBAF) (3F) were
prepared from the corresponding dichloride dimers and
(22) Guyon, F.; Lucas, D.; Jourdain, I. V.; Fourmigue, M.; Mugnier,
Y.; Cattey, H. Organometallics 2001, 20, 2421–2424.
(23) Ray, K.; Petrenko, T.; Wieghardt, K.; Neese, F. Dalton Trans.
2007, 1552–1566.
ꢀ
(33) Hartl, F.; Rosa, P.; Ricard, L.; Le Floch, P.; Zalis, S. Coord.
Chem. Rev. 2007, 251, 557–576.
(34) Chalk, A. J.; Halpern, J.; Harkness, A. C. J. Am. Chem. Soc.
(24) Kokatam, S.-L.; Chaudhuri, P.; Weyhermueller, T.; Wieghardt,
1959, 81, 854–857.
K. Dalton Trans. 2007, 373–378.
(35) Harrod, J. F.; Halpern, J. Can. J. Chem. 1959, 37, 1933–1935.
(36) Ogo, S. Chem. Commun. 2009, 3317–3325.
(37) Rakowski DuBois, M.; DuBois, D. L. Chem. Soc. Rev. 2009,
38, 62.
(25) Bill, E.; Bothe, E.; Chaudhuri, P.; Chlopek, K.; Herebian, D.;
Kokatam, S.; Ray, K.; Weyhermueller, T.; Neese, F.; Wieghardt, K.
Chem.;Eur. J. 2005, 11, 204–224.
(26) Kokatam, S.; Weyhermueller, T.; Bothe, E.; Chaudhuri, P.;
Wieghardt, K. Inorg. Chem. 2005, 44, 3709–3717.
(27) Boyer, J. L.; Cundari, T. R.; DeYonker, N. J.; Rauchfuss, T. B.;
Wilson, S. R. Inorg. Chem. 2009, 48, 638–645.
(28) Khomenko, T.; Salomatina, O.; Kurbakova, S.; Il’ina, I.; Volcho,
K.; Komarova, N.; Korchagina, D.; Salakhutdinov, N.; Tolstikov, A. Russ.
J. Org. Chem. 2006, 42, 1653–1661.
(29) Blackmore, K. J.; Ziller, J. W.; Heyduk, A. F. Inorg. Chem. 2005,
44, 5559–5561.
(30) Heiden, Z. M.; Gorecki, B. J.; Rauchfuss, T. B. Organometallics
(38) Kubas, G. J. Adv. Inorg. Chem. 2004, 56, 127–177.
(39) Rocchini, E.; Rigo, P.; Mezzetti, A.; Stephan, T.; Morris, R. H.;
Lough, A. J.; Forde, C. E.; Fong, T. P.; Drouin, S. D. J. Chem. Soc.,
Dalton Trans. 2000, 3591–3602.
(40) Cheng, T.-Y.; Szalda, D. J.; Zhang, J.; Bullock, R. M. Inorg.
Chem. 2006, 45, 4712–4720.
(41) Bullock, R. M. Chem.;Eur. J. 2004, 10, 2366–2374.
(42) Artero, V.; Fontecave, M. Coord. Chem. Rev. 2005, 249, 1518–
1535.
(43) Fontecilla-Camps, J. C.; Volbeda, A.; Cavazza, C.; Nicolet, Y.
2008, 27, 1542–1549.
Chem. Rev. 2007, 107, 4273–4303.
(31) Heiden, Z. M.; Rauchfuss, T. B. J. Am. Chem. Soc. 2009, 131,
3593–3600.
(32) Nomura, M.; Fujii, M.; Fukuda, K.; Sugiyama, T.; Yokoyama,
(44) Gloaguen, F.; Rauchfuss, T. B. Chem. Soc. Rev. 2009, 38, 100–
108.
(45) Ringenberg, M. R.; Kokatam, S. L.; Heiden, Z. M.; Rauchfuss,
Y.; Kajitani, M. J. Organomet. Chem. 2005, 690, 1627–1637.
T. B. J. Am. Chem. Soc. 2008, 130, 788–789.

1958 Organometallics, Vol. 29, No. 8, 2010
Ringenberg et al.
Figure 1. Structures of 2F (left) and 3F (right). Thermal ellip-
soids are shown at the 50% probability level. Hydrogen atoms
have been omitted for clarity. Selected bond lengths are shown
in Table 1.
Figure 2. Cyclic (solid line) and differential pulse (dashed line)
voltammograms of 1t-Bu vs Cp2Fe0/þ. Conditions: CH2Cl2, 10-3
M 1t-Bu, 0.1 M Bu4NPF6, and scan rate = 0.100 V/s.
Scheme 3. Rotamers of (cymene)Ru(tBAF) (3F)
For CH2Cl2 solutions, the two one-electron oxidations of
1F are separated by 0.17 and 0.317 V when the electrolyte is
Bu4NPF6 and Bu4NBArF4, respectively. As observed by
Geiger and co-workers, the diminished value of ΔE for
Bu4NPF6 reflects the stabilization of the higher oxidation
state by ion pairing.46
aminophenols, 3,5-(t-Bu)2C6H2-6-OH-1-(NRH) where R =
C6H4-2-CF3 and t-Bu (eq 1).
From ΔEox (= Eox(1) - Eox(2)), one can calculate the
disproportionation constant (Kdisp) for the monocations.47
For example, in the system [1F]nþ/PF6-, ΔEox = 0.172 V
corresponds to Kdisp = 3.57 ꢀ 10-2 M. Thus, a solution of
nominally 0.50 mM [1F]PF6 in CH2Cl2 will contain about
20% each (0.069 mM) of both 1F and [1F](PF6)2. For the
[1F]nþ/BArF - system, Kdisp is 50 times smaller at 6.99 ꢀ 10-4
4
M. Similarly, a nominally 0.50 mM solution of [1F]BArF4 will
contain only ∼3% each (0.013 mM) of 1F and [1F](BArF4)2.
The (arene)Ru complex 3F exhibits a third quasi-reversible
oxidation at 0.570 V. We attribute this process to the RuII/III
couple. As shown in this and related work,48 a single amido-
phenolate ligand supports only two ligand-based oxidations.
[Cp*Ir(amidophenolate)]þ and Related Adducts. Oxida-
tions of 1F and 1t-Bu were conducted on a preparative scale
using AgPF6 and FcPF6, respectively. The BArF - salts were
4
The new complexes were obtained as deeply colored, air-
stable solids that are readily soluble in a range of organic
solvents.
obtained by extracting the cations into Et2O by salt ex-
change. The salts of [1F]PF6 and [1t-Bu]PF6 were obtained
as analytically pure brown solids. The salt [1F]BArF was
4
The 1H NMR spectra of 1t-Bu, 1F, and 2F are straightfor-
ward; however, spectra for the corresponding Ru(II) com-
pounds are more complex. For 3F, we observed two methyl
doublets for the cymene ligand, indicating that the isopropyl
methyl groups are diastereotopic. Upon warming the sample,
these peaks broaden, but complete coalescence was not
observed even at 85 ꢀC. This pattern indicates that rotation
of the cymene ligand is hindered by the bulky 2-(trifluoro-
methyl)phenyl substituent on the amine (Scheme 3).
Electrochemical Studies. The Rh and Ir amidophenolato
complexes undergo two reversible one-electron oxidations
(Figure 2), as analyzed by cyclic voltammetry in CH2Cl2
solutions. The first oxidation of the Ir complexes occurs
about 0.090 V more positive than for the Rh analogues. The
substituents on the nitrogen affect the redox potentials in the
expected way. For 1t-Bu and 1F, E1/2(0/þ) = -0.015 and
0.05 V and E1/2(þ/2þ) = 0.160 and 0.355 V, vs Cp2Fe0/þ
(Table 2).
further characterized crystallographically. The Ir-N and
˚
Ir-O bonds (2.010(3) and 2.044(7) A, respectively) are
elongated relative to Ir-N and Ir-O bonds (1.963(4) and
F 45
˚
1.996(3) A, respectively) in 1 . The carbon-carbon dis-
tances in the phenylene backbone of the amidophenolate
F
differ from those in the neutral 1 , which average 1.40 A
˚
(Table 1). Thus, the C11-C12 and C14-C15 bonds are
˚
elongated by 0.05 A, and the C13-C14 and C15-C16 bonds
are shortened by 0.04 A. The pattern of bond length alter-
˚
nation in the phenylene backbone is similar to that of other
semiquinone complexes.49
(46) LeSuer, R. J.; Geiger, W. E. Angew. Chem., Int. Ed. 2000, 39,
248–250.
(47) Bard, A. J.; Faulkner, L. R. Electrochemical Methods: Funda-
mentals and Applications, 2nd ed.; John Wiley & Sons Inc.: New York,
2001; p 856.
(48) Fujihara, T.; Okamura, R.; Tanaka, K. Chem. Lett. 2005, 34,
1562–1563.
(49) Pierpont, C. G.; Buchanan, R. M. Coord. Chem. Rev. 1981, 38,
45–87.
The potential and ΔE for the two redox steps depend
strongly on the supporting electrolyte and solvent conditions.

Article
Organometallics, Vol. 29, No. 8, 2010 1959
Table 1. Selected Bond Lengths (A) of Amidophenolato Complexes
˚
1F (M = Ir)45
2F (M = Rh)
3F (M = Ru)
[1F]BArF 45
[1F(MeCN)](OTf)2
4
M1-O1
1.963(4)
1.996(3)
1.342(4)
1.393(5)
1.416(5)
1.396(6)
1.396(6)
1.388(5)
1.395(6)
1.407(5)
1.958(2)
2.0032(18)
1.332(3)
1.390(3)
1.419(4)
1.390(4)
1.403(4)
1.378(4)
1.398(4)
1.405(4)
1.9600(18)
2.0032(15)
1.338(3)
1.396(3)
1.415(3)
1.391(3)
1.414(3)
1.379(3)
1.399(3)
1.403(3)
2.010(3)
2.044(7)
1.287(5)
1.342(7)
1.451(9)
1.398(1)
1.363(5)
1.453(6)
1.370(8)
1.427(8)
2.103(5)
2.095(6)
1.250(8)
1.314(9)
1.445(10)
1.344(10)
1.457(10)
1.356(10)
1.434(10)
1.480(10)
M1-N1
O1-C11
N1-C16
C11-C12
C12-C13
C13-C14
C14-C15
C15-C16
C16-C11
[Cp*IrMe2(PMe3)]þ, which exhibits a rhombic spectrum
centered at g = 2.183 (A(31P) was not resolved at -100 ꢀC
in frozen solution).50
Table 2. Redox Couples and Disproportionation Constants (M)
for Selected Complexes
E1/20/þ (V)a E1/2þ/2þ (V)a ΔEo
Kdisp
The magnetic moments of 1t-Bu(CN), [1t-Bu]BArF
,
4
1F, Bu4NPF6
2F, Bu4NPF6
3F, Bu4NPF6
1F, Bu4NBArF
0.073
-0.040
-0.015
0.005
-0.015
-0.262
0.245
0.227
0.110
0.380
0.160
0.172 3.57 ꢀ 10-2
0.267 5.67 ꢀ 10-3
0.125 8.88 ꢀ 10-2
0.375 6.99 ꢀ 10-4
0.175 3.36 ꢀ 10-2
0.202 2.02 ꢀ 10-2
[1F]PF6,45 and [1t-Bu(PPh3)]PF6 were found to be 1.78, 1.59,
1.75, and 1.55 μB, respectively. These values, which were
measured on solutions, are reasonable for typical Ir(IV)
complexes51 or organic radicals. The magnetic susceptibility
of [1t-Bu(PPh3)]PF6 obeys the Curie-Weiss dependence over
the temperature range -15 to 40 ꢀC.
4
1t-Bu, Bu4NPF6
1t-Bu, Bu4NBArF
-0.060
4
a Potentials vs Cp2Fe0/þ; conditions: CH2Cl2 solution, 0.1 ꢀ 10-3
M
The cyclic voltammograms of the amidophenolate com-
plexes are strongly affected by donor solvents. Thus, the first
oxidation of 1F in MeCN solution is quasi-reversible even at
scan rates greater than 2 V/s. The ic for the [1F]þ/0 couple shifts
cathodically and becomes smaller as the scan rate is increased
to 5 V/s. The cathodic shift and lower ic are consistent with an
EC process, whereby oxidation of 1F is followed by rapid
coordination of solvent to form [1F(MeCN)]þ. The second
oxidation is reversible, consistent with a couple where the
coordination number at iridium remains unchanged. The
voltammograms for the oxidation of 1F and 1t-Bu in MeCN
were digitally simulated, being sensitive to the parameters,
E1/2 forthe[1F]þ/0 and [1F(NCMe)]þ/2þ couples, therateof the
reaction MeCNþ [1F]þ, andthe bindingconstant ofMeCNto
[1F]þ (Supporting Information).
The voltammogram of 1F and 1t-Bu with the noncoordi-
nating solvent and electrolyte CH2Cl2/Bu4NBArF was
shown to be very informative in developing the concept of
OI-LA. The redox properties of 1F are strongly affected by
the addition of PPh3, consistent with the formation of the
adduct, [1F(PPh3)]þ. The subsequent addition of PPh3 causes
the voltammogram shown in Figure 3; the presence of a
common intersection of the i-V traces for several concen-
trations of PPh3 (isoalpha points52) indicates that the reac-
tion [1F]þ þ PPh3 gives a single product. The CV measure-
ments with variable amounts of PPh3 indicate that PPh3
in complex, 0.1 M in Bu4NPF6.
Although the neutral amidophenolate complexes show no
tendency to bind ligands, the cations are Lewis acidic. Thus
oxidation of 1F in MeCN solution gives the adduct [1F-
(NCMe)]PF6. Solid samples of the adduct [1F(NCMe)]PF6
were found to liberate MeCN, reverting to the naked salt
[1F]PF6 upon washing with ether and heating under vacuum.
Stable adducts arise from treatment of cations [1t-Bu]þ and
[1F]þ with CO, phosphines, and cyanide. A deep brown
solution of [1t-Bu]PF6 in CH2Cl2 rapidly forms a pale orange
adduct with CO, [1t-Bu(CO)]PF6 (νCO = 2049 cm-1). In cont-
rast to the MeCN adduct, the CO complex showed no tend-
ency to dissociate under vacuum. The salts [1t-Bu]þ and [1F]þ
were found to react with the phosphines PMe3 and PPh3 to
give the adducts [1R(PR3)]PF6, which were characterized by
EPR spectroscopy and electrochemically (see below). The
cherry-red charge-neutral cyanide 1t-Bu(CN) (νCN(KBr) =
2119 cm-1) was obtained by treating [1t-Bu]PF6 with NaCN.
The complex was also prepared by treating 1t-Bu with AgCN.
Isotropic liquid EPR spectra for the naked complex [1t-Bu]-
PF6 are characterized by giso = 1.9718, close to the g value
for an organic radical (2.0023). As a frozen glass in 1:3
CH2Cl2/toluene, [1t-Bu]PF6 exhibits an axial spectrum g|| =
2.001 and g^ = 1.948 with 14N coupling A|| = 23.5 and A^ =
63.6 MHz. Spectra for the adducts, [1t-Bu(PMe3)]PF6 and
1
t-BuCN, suggest greater delocalization of the electron onto
the metal: 1t-Bu(CN) giso = 1.9741, A(14N) = 31.6, 30.3, and
(50) Diversi, P.; Iacoponi, S.; Ingrosso, G.; Laschi, F.; Lucherini, A.;
Zanello, P. J. Chem. Soc., Dalton Trans. 1993, 351–352.
(51) Cotton, S. A. Chemistry of Precious Metals; Chapman & Hall:
London, 1997.
(52) Sanecki, P. T.; Skital, P. M. Electrochim. Acta 2008, 53, 7711–
7719.
A(1H) = 19.1 MHz. The isotropic spectrum of [1t-Bu
-
(PMe )]PF6 has giso = 1.972 and A(14N) = 53.9, A(1H) =
24.0, and A(31P) = 52.8 MHz. Examples of Ir(IV) phosphine
complexes have been characterized previously including
3

1960 Organometallics, Vol. 29, No. 8, 2010
Ringenberg et al.
Figure 3. Cyclic voltammogram of 1F þ 0, 0.5, and 1 equiv of
PPh3 vs Cp2Fe0/þ. Conditions: CH2Cl2, 0.12 ꢀ 10-3 M 1F, 0.2 M
Figure 5. Structure of the cation in [1F(MeCN)](OTf)2 with
thermal ellipsoids at 50% probability and hydrogen atoms
omitted. Selected bond lengths are shown in Table 1.
Bu4NBArF4, and scan rate = 0.1 V s-1
.
correspond to double bonds, indicative of the description of
this ligand as an iminoquinone.24
Recrystallization of [1F(MeCN)](OTf)2 from CH2Cl2/
ether as well as heating under vacuum did not liberate the
MeCN, in contrast to the behavior of the monocationic
derivatives [1F(MeCN)]þ. A MeCN solution of [1F(MeCN)]-
(OTf)2 was found to convert to [Cp*Ir(MeCN)3]2þ over the
course of hours at room temperature, indicating weak bind-
ing properties of the doubly oxidized ligand.
In CH2Cl2 solution, 1t-Bu(CN) was found to reversibly
oxidize at -0.28 V, which is 0.42 V more negative than the
[1t-Bu þ/2þ couple. Oxidation of 1t-Bu(CN) with [Cp2Fe]PF6
]
gave diamagnetic [1t-Bu(CN)]PF6.
Figure 4. Cyclic voltammogram of a solution of 1F and PPh3
Catalytic Oxidation of H2. Solutions of the neutral com-
plexes (1F, 1t-Bu, 2F, 3F) in CH2Cl2 solution are unreactive
toward 1 atm of H2 even for extended reaction times. A
solution of [1t-Bu(MeCN)]PF6 in CH2Cl2 was very slow to
react with H2. Under strictly analogous conditions, [1t-Bu]-
PF6 reacts with H2 over the course of several minutes at room
temperature, as indicated by a color change from deep red to
pale yellow. The reaction affords [Cp*Ir(μ-H)3IrCp*]PF6
and free aminophenol, consistent with the stoichiometry
in eq 2.
(black) vs Cp2Fe0/þ and simulation (dashed blue). Conditions:
CH2Cl2, 0.12 ꢀ 10-3 M 1F, 0.24 mM PPh3, 0.2 M Bu4NBArF
,
4
and scan rate = 0.1 V s-1
.
forms a one-to-one adduct (Supporting Information). The
height of the current response (ip) for observed redox couples
has a linear dependence with respect to (scan rate)1/2 in the
presence of 2 equiv of PPh3 below 1 V s-1. At scan rates above
1 V s-1 this dependence deviates from linearity, indicative of
the electron transfer rates (i.e., [1F]0/þ) being faster than the
binding of Lewis bases.53 The voltammogram of 1F with PPh3
was digitally simulated (Figure 4). The binding of PPh3 by
54
þ
ꢂ
ꢂ þ
2½1t-Buꢁ þ 4H2 f ½Cp Irðμ-HÞ3IrCp ꢁ
[1F]þ occurs at ca. 105 s-1 M-1 with Kequlib = 2 ꢀ 108 M-1
.
Conversely, the simulation indicated that dissociation of
þ 2HOðt-BuðHÞNÞC6H2ðt-BuÞ2 þ Hþ
ð2Þ
PPh3 from 1F(PPh3) proceeds at a rate of 103 s-1 with K =
The reaction was monitored by UV-vis spectroscopy (λmax
=
6.2 ꢀ 106 M. Cyclic voltammetry experiments on [1t-Bu 0/þ
The couples [1F(MeCN)]þ/2þ and [1F(PPh3)]þ/2þ differ by
only 0.091 V. The insensitivity of the couple to the nature of
the Lewis base suggests that the second oxidation is highly
localized on the amidophenolate ligand.
]
in
460 nm for [1t-Bu]PF6), and the rate was found to be first
order in both [1t-Bu]PF6 and PH2 with d[1t-Bu]/dt = 1.61 ꢀ
10-4 ((0.12 ꢀ 10-4) s-1 at 22 ꢀC with PH2 of 1 atm.
The oxidation of H2 also proceeded well in the presence of
a noncoordinating base; however hydrogenolysis of the
complex was suppressed. Thus, reduction of [1t-Bu]PF6
in CH2Cl2 containing 5 equiv of 2,6-di-tert-butylpyridine
(t-Bu2py) and 1 atm of H2 resulted in the clean formation of
neutral 1t-Bu (eq 3).
the presence of PPh3 gave similar findings.
Doubly Oxidized Species [Cp*Ir(tBAR)L]2þ. Oxidation of
1F with 2 equiv of AgOTf in MeCN solution gave the deep
purple diamagnetic salt [1F(MeCN)](OTf)2, which was char-
acterized by 1H and 19F NMR spectra. Crystallographic
analysis of [1F(MeCN)](OTf)2 supports the assignment of
the ligand to the iminoquinone oxidation state.24 Thus, the
2½1t-Buꢁþ þ H2 þ 2t-Bu2py f 21t-Bu þ 2t-Bu2pyHþ ð3Þ
The rate of the hydrogenation reaction in the presence of
t-Bu2py was evaluated by monitoring the disappearance of
˚
˚
Ir1-N1 and Ir1-O1 distances of about 2.10 A are ∼0.11 A
longer than in 1F (Table 1). The carbon-oxygen (O1-C11,
1.25 A) and carbon-nitrogen (N1-C12, 1.31 A) distances
(53) Creutz, C. Prog. Inorg. Chem. 1983, 1–73.
˚
˚
(54) Fettinger, J. C.; Pleune, B.; Poli, R. J. Am. Chem. Soc. 1996, 118,
4906–4907.

Article
Organometallics, Vol. 29, No. 8, 2010 1961
Table 3. Rate Constants for Reduction of
[(C5Me5-nHn)M(O(NR)C6H2(t-Bu)2)]þ by H2 (5 equiv t-Bu2Py,
CH2Cl2 soln, 22 ꢀC unless otherwise noted)
catalysta
rate, at 1 atm H2, s-1
[1t-Bu]PF6
[1F]PF6
1.01 ꢀ 10-3 ((2.66 ꢀ 10-5
1.85 ꢀ 10-3 ((1.06 ꢀ 10-4
4.66 ꢀ 10-2 ((2.28 ꢀ 10-3
1.13 ꢀ 10-3 ((8.69 ꢀ 10-4
4.74 ꢀ 10-3 ((2.18 ꢀ 10-4
no reaction
)
)
)
)
)
[1F]BArF
4
[Cp#Ir(tBAF)]PF6
[2F]BArF
4
[3F]PF6
[Ni(PCy2NBz2)2](BF4)2
AgOAc in pyridine (70 ꢀC)
10 (ref 55)
0.49 (ref 34)
a Cp# = C5Me4H-. PCy2NBz2 = [C6H11PCH2N(CH2Ph)CH2]2.
[1t-Bu]PF6. The rate was unaffected by the concentration of
base and exhibited a first-order dependence on both [1t-Bu]-
PF6 and pH2 (eq 4). The rate of reduction of [1t-Bu]PF6 by 1
atm of D2 in the presence of t-Bu2py gave kH/kD = 2.01.
d½1t-Buꢁ=dt ¼ ð1:01ꢀ10-3ÞðPH2Þ½ð1t-BuÞþꢁ ð22 ꢀCÞ ð4Þ
Similarly, the corresponding reduction of [1F]PF6 by H2 with
Figure 6. Solution of 0.1 M [Cp2Fe]PF6, 0.4 M t-Bu2py, and
1t-Bu (catalyst) in 10 mL of CH2Cl2 before (left) and 18 h after
pressurization with 3 atm of H2 (right).
t-Bu2py is nearly twice as fast as [1t-Bu]PF6 with a kH/kD
1.2. Changing from Cp* to C5Me4H only slightly affected the
rate. The rate of reduction was strongly affected by the
=
Although the “hydrogenophilicity” of metal centers normally
increases with charge,56 our data indicate that the singly
oxidized species are better catalysts than the dication. The
role of the singly oxidized species is supported by the finding
counteranion: reduction of [1F]BArF by H2 is about 30
4
times faster than [1F]PF6. Different metal cations were also
evaluated: the analogous [2F]BArF4 was reduced by H2 but
was 10 times slower than the iridium complex, and the
cymene ruthenium complex, [3F]PF6, does not react with
H2. A summary of the observed first-order rate constants at
1.0 atm of H2 is shown in Table 3.
F
-
that [1F]BAr4 is nearly 30 times faster than the PF6 salt.
Being subject to a far higher comproportionation constant,
solutions of the BArF - salt contains less of the dication (and
4
neutral) components. Coulommetry experiments are also
consistent with the activation of H2 by the monocations,
since the rate (current) of H2 oxidation is not enhanced at
potentials above the second oxidation. The inactivity of the
dications is attributed in part to their high Lewis acidity,
which may result in their binding counteranions, which
compete with the binding of H2. Matching a complex’s Lewis
acidity and hydrogenophilicity is a key consideration in the
design of catalysts for oxidation of H2.37
With 1t-Bu as a catalyst, [Cp2Fe]þ can be reduced with H2
(Figure 6). Thus, a blue-green H2-saturated solution consist-
ing of equal parts [Cp2Fe]PF6 and t-Bu2py became pale
yellow several hours after the addition of 2.5 mol % of
1
t-Bu. The rate (d[Cp2Fe]/dt) was shown to be independent of
the initial concentration of [Cp2Fe]þ. The overall rate con-
stant for the reduction of Fcþ is 9.00 ꢀ 10-4 s-1, which is
comparable to 1.01 ꢀ 10-3 s-1 for the reduction of [1t-Bu]PF6
in the absence of the [Cp2Fe]þ.
In terms of mechanism, the proposed catalytic cycle
involves the intact complex, consistent with the fact that
the reduction of [1t-Bu]þ by H2 occurs far faster than the H2-
induced hydrogenolysis to give free aminophenol. Substitu-
ents on the amine affect the rate: the trifluoromethylphenyl
derivative [1F]þ is reduced by H2 nearly twice as fast as the
tert-butyl derivative [1t-Bu]þ. Substituents on the cyclopenta-
dienyl ligands (C5Me5 vs C5Me4H) and the identity of the
metals (Rh vs Ir) also affect the rates. On the basis of the
above analysis, we propose that H2 oxidation begins with
formation of the complex [Cp*Ir(tBAR)(H2)]þ. This assump-
tion is consistent with the inhibition of H2 oxidation by
MeCN. In the next step, the H2 adduct undergoes deproto-
nation, consistent with the high Brønsted acidity of many
cationic dihydrogen complexes.56,57 Deprotonation of this
dihydrogen complex would give the hydride Cp*Ir(tBAR)H,
which would be readily oxidized by either [Cp*Ir(tBAR)]þ or
Fcþ. In terms of their redox potential, the (unobserved)
hydride can be modeled to some extent by the cyanide
Discussion
The 2-amidophenolate ligands are strong π-donors and as
such stabilize metal complexes that are formally electroni-
cally unsaturated. The oxidation-induced Lewis acidity in
such complexes has not been widely exploited. In this work
we show that oxidation of 16e amidophenolate derivatives of
Cp*M2þ (M = Rh, Ir) enhances their electrophilicity suffi-
ciently to cause an interaction with dihydrogen. The spectro-
scopic, crystallographic, and electrochemical measurements
indicate that the oxidation is significantly localized on the
amidophenolate ligand.
Since the complexes undergo two oxidations, we con-
ducted experiments aimed at determining which of the two
cationic states is responsible for the oxidation of H2. The
bimolecular kinetics are compatible with either (or both) the
monocation and a dication being the active oxidant.
1
t-BuCN, although the hydride is expected to be the better
electron donor of the two.58 By comparing the couples for
(55) Wilson, A. D.; Newell, R. H.; McNevin, M. J.; Muckerman,
J. T.; Rakowski DuBois, M.; DuBois, D. L. J. Am. Chem. Soc. 2006, 128,
358–366.
(56) Kubas, G. Metal Dihydrogen and σ-Bond Complexes; Kluwer
Academic/Plenum: New York, 2001.
(57) Zhang, K.; Gonzalez, A. A.; Hoff, C. D. J. Am. Chem. Soc. 1989,
111, 3627–3632.
(58) Lever, A. B. P. Inorg. Chem. 2002, 29, 1271–1285.

1962 Organometallics, Vol. 29, No. 8, 2010
Ringenberg et al.
Scheme 4. Proposed Pathway for the Oxidation of H2 by Iridium
Amidophenolate Complexes
Instrumentation. NMR spectra were recorded on Varian
UNITY INOVA 500NB or UNITY 500 spectrometers. FT-IR
spectra were recorded on a Mattson Infinity Gold FTIR spec-
trometer. UV-vis spectra were recorded on a Cary 50 UV/vis
spectrometer.
Electrochemical experiments were performed using a BAS
CV-50W voltammetric analyzer. For a typical cyclic voltamme-
try (CV) experiment the cell consisted of a 10-3 M analyte
solution containing 0.1 M of supporting electrolyte. The work-
ing electrode was a 1 mM glassy carbon tipped electrode, which
was polished between experiments using 1.0 μM alumina polish
(Buehler) in deionized water. The reference electrode was Ag/
AgCl in saturated aqueous KCl with a Vycor tip, and a Pt wire
was used as the counter electrode. Simulations employed Digi-
Sim 3.05 software (BAS, West Lafayette, IN).
Magnetic moments were recorded by the Evans method by
comparing the shifts of the residual 1H benzene signals (δ 7.37)
for solutions of C6D6 or CDCl3 with and without the para-
magnetic sample.63 Using the same method, we found that the
moment for [Cp2Fe]PF6 was 2.49 μB, vs the literature value of
2.62 μB.64
X-band EPR spectra were collected on a Varian E-122
spectrometer. Variable-temperature spectra were recorded
using an E-257 variable-temperature accessory using liquid
nitrogen as a coolant. The magnetic fields were calibrated with
a Varian NMR Gauss meter, and the microwave frequency was
measured with an EIP frequency meter. EPR spectra were
simulated with the program SIMPOW6.65 The sample was flame-
sealed under vacuum in a 3 ꢀ 4 mm quartz tube at liquid nitro-
gen temperatures.
[1t-Bu þ/2þ
]
and [1t-BuCN]0/þ, we see that cyanide stabilizes the
doubly oxidized state by 0.44 V. In fact the [1t-BuCN]0/þ
couple is 0.260 V more negative than the [1t-Bu 0/þ
couple;
]
that is, 1t-BuCN would be readily oxidized by [1t-Bu]þ. Thus, it
is apparent that [1t-BuH] would readily convert to [1t-BuH]þ in
the presence of [1t-Bu þ
] (eq 5).
[Cp*Ir(tBAF)(NCMe)]PF6, [1F(NCMe)]PF6. This salt was
prepared analogously to [1F]PF6 but in MeCN solution and
can be isolated by crystallization from Et2O. The MeCN was
removed by double recrystallization from CH2Cl2 by the addi-
tion of ether. The solid was heated to 40 ꢀC at reduced pressure.
Anal. Calcd for C31H39F9IrNOP(CH3CN) (found): C, 45.20
(44.60); H, 4.82 (5.04); N, 3.19 (3.14). After washing Anal. Calcd
for C31H39F9IrNOP (found): C, 44.54 (44.77); H, 4.67 (4.30); N,
1.67 (1.69).
½1t-BuXꢁ þ ½1t-Buꢁþ f ½1t-BuXꢁþ þ ½1t-Bu
ꢁ
ð5Þ
ðX- ¼ H-, CN-; for X ¼ CN, E ¼ 0:260 VÞ
Oxidation of metal hydrides is known to enhance their
acidities by as much as 10-20 pKa units.59 Although in the
present case the oxidation is ligand-centered, the acidity of
the resulting [1t-BuH]þ is apparently readily deprotonated by
t-Bu2Py, which has a pKa in MeCN of 11.5.60 The overall
proposed mechanism is shown in Scheme 4.
Cp*Ir(tBAt-Bu), 1t-Bu. A solution of 1 g (1.256 mmol) of
[Cp*IrCl2]2 in 40 mL of CH2Cl2 was combined with a solution
t
of 700 mg (2.512 mmol) of H2 BAt-Bu in 20 mL of CH2Cl2. After
5 min of stirring, 1 g (7.24 mmol) of K2CO3 and 10 mg (60.3
μmol) of Et4NCl were added. After about 1 h the solution color
had turned from orange to red. The reaction solution was
filtered through a course frit to remove K2CO3, and the solvent
was removed under reduced pressure. The solid was extracted
from the residue into 10 mL of pentane, and the pentane
solution was passed through a plug of Celite to remove the
Et4NCl. The resulting solution was cooled to -30 ꢀC for 12 h,
and the resulting crystals were collected and washed with cold
Experimental Section
Unless otherwise indicated, reactions were conducted using
Schlenk techniques, and all reagents were purchased from
Sigma-Aldrich. Reagents and routine solvents were obtained
commercially and were purified by standard methods.61 Unless
otherwise indicated, all solvents were HPLC grade and purified
using an alumina filtration system (Glass Contour, Irvine, CA).
Reagents. [Cp2Fe]PF6 was recrystallized by extraction into
acetone followed by precipitation with ethanol.62 Electrolyte,
Bu4NPF6, was crystallized prior to use by extracting into a
minimal amount of acetone followed by addition of an equal
part of ethanol and cooling to -30 ꢀC. CD2Cl2 and CD3CN
(Cambridge Isotope Laboratories) were dried with CaH2 and
1
pentane. Yield: 1.127 g (75%). H NMR (CD2Cl2): δ 1.30 (s,
9H), 1.48 (s, 9H), 1.83 (s, 9H), 1.89 (s, 15H), 6.71 (d, 1H), 7.33 (d,
1H). UV-vis (CH2Cl2) λmax, nm (ε): 456 (9436). Anal. Calcd for
C28H44IrNO (found): C, 55.78 (55.74); H, 7.36 (7.71); N, 2.32
(2.46).
[Cp*Ir(tBAt-Bu)]PF6, [1t-Bu]PF6. A solution of 63 mg (0.249
mmol) of AgPF6 in 5 mL of CH2Cl2 was added to a solution of
150 mg (0.249 mmol) of 1t-Bu in 40 mL of CH2Cl2. Upon
distilled under vacuum onto NMR samples. 3,5-Di-tert-butyl-2-
hydroxy-1-(2-tert-butlylamine)benzene (H2 BAt-Bu
) was pre-
t
pared according to Blackmore et al. (Supporting Information).29
Celite 545 was dried at 120 ꢀC prior to use. We previously
described the syntheses of 1F and [1F]PF6.45
(63) Girolami, G. S.; Rauchfuss, T. B.; Angelici, R. J. Synthesis and
Technique in Inorganic Chemistry, 3rd ed.; University of Science Books:
Sausalito, CA, 1999; pp 47-53.
(64) Gray, H. B.; Hendrickson, D. N.; Sohn, Y. S. Inorg. Chem. 2002,
10, 1559–1563.
(65) EPR spectra were simulated with the automated fitting program,
SIMPOW6, a program developed at the University of Illinois based on
POW (Nilges, M. J. Ph.D. Thesis, University of Illinois, Urbana-
Champaign, 1979) and MPOW (Chang, H.-R.; Diril, H.; Nilges,
M. J.; Zhang, X.; Potenza, J. A.; Schugar, H. J.; Hendrickson, D. N.;
Isied, S. S. J. Am. Chem. Soc. 1988, 110, 625-627).
(59) Ryan, O. B.; Tilset, M.; Parker, V. D. J. Am. Chem. Soc. 1990,
112, 2618–2626.
(60) Izutsu, K. Acid-Base Dissociation Constants in Dipolar Aprotic
Solvents; Blackwell Scientific Publications: Oxford, 1990; Vol. 35, p 166.
(61) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals, 3rd ed.; Pergamon Press: New York, 1988.
(62) Connelly, N. G.; Geiger, W. E. Chem. Rev. 1996, 96, 877–910.

Article
Organometallics, Vol. 29, No. 8, 2010 1963
addition of AgPF6, the reaction solution color changed from
bright orange to dark brick red. The reaction solution was
allowed to stir for 5 h. The reaction mixture was filtered through
Celite to remove precipitated silver, rinsing with about 20 mL of
CH2Cl2 until the effluent ran colorless. The solvent was removed
under reduced pressure, and the resulting brick-red solid was
crystallized from 5 mL of CH2Cl2 overlayered with 15 mL of
hexane. To remove any excess 1t-Bu, the resulting crystals were
washed with 10 mL portions of ether until the effluent ran
colorless. The solid was dried at 40 ꢀC at 0.5 mmHg and stored in
an N2-filled glovebox. Yield: 37 mg (20%). UV-vis (CH2Cl2,
nm (ε)) λmax: 460 nm (4211). Anal. Calcd for C28H44F6Ir-
(KBr) νmax: 2119 cm-1 (CN). Anal. Calcd for C29H44Ir-
N2O CH2Cl2: C, 50.48 (49.62); H, 6.49 (6.50); N, 3.92 (4.01).
3
Magnetic moment (291 K): The magnetic moment of 1t-BuCN
was determined to be 1.78 μB, by the Evans method in C6D6
solution.
[Cp*Ir(tBAt-Bu)CN]PF6, [1t-Bu(CN)]PF6. A solution of 24 mg
(0.078 mmol) of [Cp2Fe]PF6 was added to a solution of 50 mg
(0.079 mmol) of 1t-Bu(CN) in 10 mL of MeCN. The solution color
turned from cherry red to brown. The reaction solution was stirred
for 1 h, after which 30 mL of Et2O was added to precipitate a
brown solid. The product was crystallized from 2 mL of CH2Cl2
and 10 mL of Et2O, followed by washing with 10 mL of hexanes.
Yield: 47 mg (76%). 1H NMR (CD2Cl2): δ 1.01 (s, 9H), 1.35 (s,
9H), 1.48 (s, 9H), 1.58 (s, 15H), 6.15 (d, 1H), 7.35 (d, 1H). Anal.
NOP CH2Cl2 (found): C, 41.82 (41.20); H, 5.46 (5.38); N,
3
1.68 (2.29).
[Cp*Ir(tBAR)]BArF4, [1R]BArF4. The salt, [1R]BArF4, was
made by ion exchange in which the [1R]PF6 was overlayered
with ether containing 0.9 equiv of KBArF4. The reaction solu-
tion was vigorously stirred for ∼1 h, at which point the ether
solution color turned dark brown. The residual PF6- salts were
removed via filtration, and the filtrate was collected. The solvent
was removed under reduced pressure to yield a brown powder.
Calcd for C29H46IrN2OPF6 2(CH2Cl2): C, 39.45 (38.00); H, 5.12
(5.11); N, 2.96 (2.90).
3
t
Cp*Rh(tBAF), 2F. A solution of 240 mg of H2 BAF (0.65 mmol)
in 10 mL of MeCN was added to a slurry of 200 mg (0.32 mmol) of
[Cp*RhCl2]2 in 15 mL of MeCN followed by 0.2 mL (14 mmol) of
Et3N. The reaction solution was stirred for 4 h, and the solvent was
removed under vacuum. The air-stable purple microcrystals were
extracted into ca. 50 mL of n-hexane, and this extract was filtered to
remove Et3NHCl. Evaporation of filtrate afforded a purple pow-
der. Yield: 0.18 g (92%). Slow evaporation of a MeCN solution of
2F gave purple X-ray quality crystals. 1HNMR(CD2Cl2):δ1.09 (s,
9H), 1.53 (s, 9H), 1.67 (s, 15H), 6.16 (d, 1H), 6.86 (d, 1H), 7.24 (d,
1H), 7.38 (t, 1H), 7.65 (t, 1H), 7.78 (d, 1H). 19F NMR (CD2Cl2): δ
-60.5 (s). FD-MSþ: m/z = 601. UV-vis (CH2Cl2) λmax, nm (ε):
579 (3110), 353 (410), 288 (1300). Anal. Calcd for C31H39NF3ORh
(found): C, 61.95 (61.45); H, 6.54 (6.73); N, 2.33 (2.48).
The magnetic moment of [1t-Bu]BArF was determined to be
4
1.59 μB, by the Evans method in C6D6 solution.
[Cp*Ir(tBAt-Bu)(PPh3)]PF6, [1t-Bu(PPh3)]PF6. A solution of
110 mg (0.335 mmol) of Cp2FePF6 in 10 mL of CH2Cl2 was
added to a solution of 205 mg (0.340 mmol) of 1t-Bu in 10 mL of
CH2Cl2. Upon addition of the Cp2FePF6, the solution color
turned from orange to brown. The reaction solution was stirred
for 2 h, at which point 89 mg (0.340 mmol) of PPh3 was added,
and the reaction solution was stirred for an additional 3 h. The
solvent was removed under reduced pressure, and the resulting
brown solid was washed with 2 ꢀ 10 mL of ether and 10 mL
portions of n-hexanes until the effluent ran colorless. The
resulting brown powder was crystallized from 5 mL of CH2Cl2
overlayered with 10 mL of ether. The solid was isolated via
filtration and dried 20 ꢀC at reduced pressure. Yield: 140 mg
[Cp*Rh(tBAF)]BArF4, [2F]BArF4. A solution of 105 mg
(0.17 mmol) of 2 in 10 mL of CH2Cl2 was added a solution of
170 mg (0.169 mmol) of [Cp2Fe]BArF4 in 10 mL of CH2Cl2. An
immediate color change from purple to green was observed.
After 5 h, the solvent was removed under vacuum. The oily green
residue was crystallized by extracting into 7 mL of CH2Cl2 and
layered with 30 mL of hexane. The resulting green solid was
washed with 10 mL of hexane and dried at 0.5 mmHg. A MeCN
(40%). Anal. Calcd for C46H59F6IrNOP2 CH2Cl2 (found): C,
3
51.55 (50.89); H, 5.61 (5.65); N, 1.27 (1.47). Magnetic mom-
ent (291 K): The magnetic moment of [1t-Bu(PPh3)]PF6 was
determined to be 1.55 μB, by the Evans method in CDCl3
solution.
solution of the [Cp*Rh(tBAF)]BArF upon cooling to -30 ꢀC
4
afforded brown X-ray quality crystals of the adduct
[Cp*Rh(tBAF)(NCMe)]BArF4. Yield: 0.196 g (80%). ESI-
MSþ, m/z: 601 ([Cp*Rh(tBAF)]þ). UV-vis (CH2Cl2) λmax, nm
(ε): 653 (4000), 437 (7400). UV-vis (MeCN) λmax, nm (ε): 793
(800), 575 (1300), 447 (3700), 333 (sh). Anal. Calcd for
C63H51NBF27ORh (found): C, 53.44 (53.90); H, 3.63 (3.25);
N, 1.00 (1.38).
[Cp*Ir(tBAt-Bu)(PMe3)]PF6, [1t-Bu(PMe3)]PF6. A Schlenk
tube was charged with 23 mg (30.7 μmol) of [1t-Bu]PF6 and 5
mL of CH2Cl2. To a frozen solution excess PMe3 was trans-
ferred via vacuum distillation. The solution was thawed and
shaken periodically over the course of 40 min. The color did not
change substantially. Solvent and excess PMe3 were removed
under reduced pressure, resulting in a brick-red, oily solid. The
residue was triturated with 10 mL portions of n-hexanes until the
effluent ran colorless. The resulting powder was isolated via
filtration. UV-vis (CH2Cl2, nm) λmax: 460.
(cymene)Ru(tBAF), 3F. To a solution of 0.24 g (0.65 mmol)
t
of H2 BAF in 10 mL of MeCN was added a solution of 0.2 g
(0.33 mmol) of [(cymene)RuCl2]2 in 15 mL of MeCN followed
by 0.19 mL (13 mmol) of Et3N. The reaction solution was stirred
for 4 h, and the solvent was evaporated under reduced pressure.
The red residue was extracted into 100 mL of hexanes, and the
solid Et3NHCl was removed via filtration. The solvent volume
was reduced to about 20 mL and by using a slow stream of
argon, and the remaining solvent was evaporated to yield X-ray
[Cp*Ir(tBAt-Bu)(CO)]PF6, [1t-Bu(CO)]PF6. A solution of 22
mg (29 μmol) of [1t-Bu]PF6 in 5 mL of CH2Cl2 was sparged with
CO for 30 s. The color of the solution lightened from brick red to
yellow. The solvent was removed under reduced pressure,
and the resulting solid was rinsed with 10 mL portions of
hexanes until the extract was colorless. The resulting powder
was isolated via filtration and dried under vacuum. IR (KBr):
1
quality crystals. Yield: 170 mg (87%). H NMR (CD2Cl2): δ
1.08 (9H, s), 1.36 (6H, dd), 1.53 (9H, s), 1.73 (3H, s), 2.76 (1H,
m), 4.98 (1H, d), 5.13 (1H, d), 5.38 (1H, d), 5.45 (1H, d), 6.20
(1H, d), 6.50 (1H, d), 6.89 (2H, s), 7.03 (1H, s), 7.29 (2H, s), 7.34
ν
CO = 2038 cm-1
Cp*Ir(tBAt-Bu)(CN),1t-Bu(CN).A solution of 70 mg (0.116 mmol)
.
(1H, d), 7.44 (1H, d), 7.54 (1H, t), 7.64 (1H, t), 7.78 (1H, d). 19
F
of 1t-Bu in 10 mL of CH2Cl2 was added to a slurry of 15.6 mg
(0.116 mmol) of AgCN in 10 mL of MeCN. No immediate color
change was observed. After stirring for 12 h, the reaction solution
had turned from orange to cherry red. The mixture was filtered
through Celite to remove precipitated silver. The solvent volume was
concentrated to about 5 mL, and 20 mL of hexanes was added. The
resulting mixture was cooled to -30 ꢀC to yield a cherry-red solid.
The air-stable product was isolated via filtration. Yield: 44 mg
(60%). UV-vis (CH2Cl2) λmax, nm (ε): 480 (3887). MS-ESIþ
(relative intensity): 630.2 m/z (100%, [Cp*Ir(tBAt-Bu)(CN)]þ). IR
NMR (CD2Cl2): δ -59.36 (s). ESIþ: m/z = 599 ([C31H38-
RuNO]þ). Anal. Calcd for C31H38NORu (found): C, 68.73
(67.82); H, 7.07 (7.12); N, 2.58 (2.44).
[(cymene)Ru(tBAF)]PF6, [3F]PF6. A solution of 120 mg (0.36
mmol) of [Cp2Fe]PF6 in 15 mL of CH2Cl2 was added to 195 mg
(0.33 mmol) of 3F in 15 mL of CH2Cl2. The solution color
changed from red to purple. The reaction solution was stirred
for 3 h, and the solvent was evaporated under reduced pressure.
The resulting solid was dissolved in 2 mL of CH2Cl2 and layered
with 10 mL of hexanes to afford purple crystals. The solid was

1964 Organometallics, Vol. 29, No. 8, 2010
Ringenberg et al.
Scheme 5. Parameters Used in Simulation of Voltammograms of 1F and 1t-Bu in the Presence of PPh3 and in MeCNa
a Operating conditions for 1R þ MeCN (R = F, t-Bu): MeCN, 0.2 mM 1R, 0.20 M Bu4NPF6, and scan rate = 0.1 V/s. Operating conditions for 1F þ
PPh3: CH2Cl2, 0.12 mM 1F, 0.2 M Bu4NBArF4, and scan rate = 0.1 V/s.
washed with 2 ꢀ 10 mL of hexanes. Yield: 170 mg (70%). ESI-
MS: m/z = 743 ([(cymene)Ru(tBAF)]þ). Anal. Calcd for
C31H38F6NOPRu (found): C, 50.01 (49.88); H, 5.15 (5.21); N,
1.88 (2.16). Anal. Calcd for C31H39F9IrNOP(Ph3P) (found): C,
51.20 (48.67); H, 4.85 (4.51); N, 1.27 (1.35).
simulation. The fit was quite sensitive to the rate of dissociation
of PPh3 from [1F(PPh3)]þ and the associated equilibrium constant
(Scheme 5). The heterogeneous electron transfer rate and symmetry
parameter (R) were left at the default value of 1 ꢀ 105 cm s-1
and 0.5. For [1F]þ/0 and [1F]þ/2þ, the default values for the
symmetry parameter and electron transfer rates are consistent
with the observed ic/ia ≈ 1 for scan rates 0.025 to 5 V/s.67 The
diffusion coefficients for 1R and [1RL]þ were adjusted to be
proportional to the slopes of the (linear) plots of ip vs ν1/2, and
electrochemically coupled species (e.g., 1F and [1F]þ) were assigned
[Cp*Ir(tBAF)(MeCN)](OTf)2, [1F(MeCN)](OTf)2. To a solu-
tion of 50 mg (0.072 mmol) of 1F in 10 mL of CH2Cl2 with
0.1 mL of MeCN was added 37 mg (0.144 mmol) of AgOTf in
5 mL of CH2Cl2. The solution color turned from orange to dark
red upon addition of the AgOTf solution. The reaction solution
was stirred at room temperature for 5 h. The solvent volume was
reduced to about 1 mL and overlayered with 10 mL of Et2O to
yield a brick-red solid. The solid was isolated via filtration and
crystallized from 1 mL of CH2Cl2 and 7 mL of Et2O at -30 ꢀC to
yield X-ray quality crystals. Yield: 31 mg (42%) 1H NMR
(CD2Cl2): δ 1.147 (9H, s), 1.427 (9H, s), 1.588 (15H, s), 2.743
(3H, s), 6.034 (1H, s), 7.405 (1H, s), 7.472 (1H, d), 7.870 (1H, d),
8.022 (2H, t). 19F NMR (CD2Cl2): δ -59.5. The decomposition
of [1F(MeCN)](OTf)2 was monitored by adding 0.02 mL of
CD3CN to the CD2Cl2 solution, and the 19F NMR signal at δ
-59.5 was observed to disappear after about 5 h. The 1H NMR
spectrum was complicated.
the same diffusion coefficients of 4.533 ꢀ 10-5 cm2 s-1 68
.
The
diffusion coefficient of Cp2Fe in MeCN is 2.37 ꢀ 10-5 cm2 s-1 69
.
Reduction of [1t-Bu]PF6 with H2. A 50 mL stock solution of
[1t-Bu]PF6 (8.03 ꢀ 10-5 M) in CH2Cl2 was prepared. To an oven-
dried UV-vis cell that was fitted with an airtight Teflon stop-
cock was added 4 mL of the stock solution under argon. An
initial UV-vis spectrum was recorded. Taking care not to
introduce air, the solution was sparged with H2 for 20 s, and
the cell was sealed under an active purge of H2 to establish
approximately 1 atm of H2. The solution was shaken and placed
in the spectrometer, and the initial spectrum was obtained 48 s
after H2 was introduced. Spectra were recorded at 10 s intervals
for 40 min (296 K). The solution was not stirred for the duration
of the experiment. The rate was monitored by following the peak
at λmax = 460 nm. Data from t = 0 to 360 s were analyzed. The
observed first-order rate constant, an average of four experi-
Oxidation of 1F and 1t-Bu in the Presence of MeCN and PPh3.
Solution of 8.5 mg (0.0123 mmol) of 1F in 10 mL of CH2Cl2/
Bu4NPF6, 0.12 mM 1F (0.24 mM 1t-Bu) in 0.1 M Bu4NPF6/
MeCN, and 18.1 mM PPh3 in CH2Cl2/Bu4NPF6 were prepared
and sparged with solvent-saturated N2.
ments, was 1.01 ꢀ 10-3 ( 2.66 ꢀ 10-5 s-1
.
Before each electrochemical experiment, the working elec-
trode was polished using alumina,66 and the reaction solution
was briefly sparged with N2. The electrochemical experiment was
performed by the addition of varying equivalents of the PPh3
solution at a scan rate of 0.1 V/s. The final scan was performed after
the addition of ∼10 mg of Cp2Fe as an internal standard.
The voltammogram of 1F in MeCN was simulated according to
the mechanism and using the parameters in Scheme 5. With very
similar parameters, we also simulated the CV for 1t-Bu in MeCN.
The parameters E1MeCN-F and E1MeCN-Bu matched those observed
Reduction of [1t-Bu]PF6 with H2 in the Presence of Base. A
10 mL portion of stock solution of [1t-Bu](PF6) (1.01 mM) and
t-Bu2py (2.26 mM) in CH2Cl2 was prepared. To the UV-vis
reaction cell was added 1 mL of the stock solution followed by
10 mL of CH2Cl2. The initial UV-vis spectrum was recorded,
and the sample was frozen. The frozen solution was evacuated
and then pressurized with 1.02 atm of H2. The sample was
thawed by shaking the flask under running tepid tap water; the
initial spectrum was recorded at time = 124 s. Time = 0 was
assigned to the moment the sample was completely thawed. The
solution color was pale brown upon insertion into the spectro-
meter. Spectra were recorded at 20 s intervals for 40 min (296 K).
The sample was not stirred for the duration of the experiment.
Rates were determined by monitoring the disappearance of the
peak at 460 nm. The data used was from time = 124 to 900 s. The
directly for the couple [1F]0/þ and [1t-Bu 0/þ
indicate that affinities of both [1t-Bu þ and [1F]þ for MeCN are
]
. The simulations
]
low, consistent with the ease with which they can be desolvated.
The t-Bu derivative’s affinity is 10 times higher of the two, which
was unexpected. For these simulations involving MeCN, it was
unnecessary to include E3 and the fit was insensitive to Keq 2
.
average of four runs has a rate of reduction of 1.702 ꢀ 10-4
(
The simulation for the oxidation of 1F in the presence of PPh3
indicated a very high affinity (108-1) of [1F]þ for the phosphine.
2.04 ꢀ 10-6 s-1. When a solution of [1t-Bu]PF6 in CH2Cl2 was
treated with t-Bu2py, the optical spectrum remained unchanged.
Because of this high affinity, the couple [1F(PPh3)]þ/0 was
PP3-F
observable (E3
= -0.984 V) and incorporated into the
(67) Wang, R. L.; Tam, K. Y.; Compton, R. G. J. Electroanal. Chem.
1997, 434, 105–114.
(68) Taylor, A. W.; Qiu, F.; Hu, J.; Licence, P.; Walsh, D. A. J. Phys.
Chem. B 2008, 112, 13292–13299.
(69) Jacob, S. R.; Hong, Q.; Coles, B. A.; Compton, R. G. J. Phys.
Chem. B 1999, 103, 2963–2969.
(66) Tilset, M. Organometallic Electrochemistry: Thermodynamics
of Metal-Ligand Bonding. In Comprehensive Organometallic Chemistry
III; Crabtree, R. H., Mingos, D. M. P., Eds.; Elsevier: Oxford, 2007;
pp 279-305.

Article
Organometallics, Vol. 29, No. 8, 2010 1965
NMR analysis indicated the formation of 1t-Bu, [Cp*Ir(μ-
The compound [2F(MeCN)]BArF was crystallized from a
4
t
H)3IrCp*]þ, t-Bu2pyHþ, and H2 BAt-Bu. The reductions of
saturated MeCN solution at -30 ꢀC. The structure was solved
by direct methods. The crystal contained two MeCN solvates
per formula unit; one of these was disordered about an inversion
center. Atoms of the disordered solvent were refined as a rigid,
idealized group, and their displacement parameters were re-
strained to be isotropic with an effective standard deviation of
0.01. The Cp* ligand was also found to have a 2-fold disorder;
the atoms modeling the disorder were restrained to have the
same anisotropic displacement parameters with an effective
[1F]PF6 and [2F]BArF4 were performed in the same manner.
Reduction of Cp2Feþ Salts. A stock solution of 115 mg
(0.347 mmol) of [Cp2Fe]PF6, 5.2 mg (8.67 μmol) 1t-Bu, and 76 μL
(0.347 mmol) of t-Bu2py was prepared in 100 mL of CH2Cl2. To
the UV-vis reaction cell was added 10 mL of the stock solution.
An initial UV-vis spectrum was recorded, and the sample was
frozen. The frozen solution was evacuated and then pressurized
with 1.02 atm of H2. The sample was thawed by shaking the flask
under running tepid tap water; the initial spectrum was recorded at
time = 100 s. Time = 0 was assigned to the moment when the
sample was completely thawed. The sample was blue upon inser-
tion into the spectrometer. Spectra were recorded at 60 s intervals
for 500 min (296 K). The sample was not stirred for the duration of
the experiment. Rates were determined by monitoring the disap-
pearance of the peak at 625 nm. In a control experiment, a solution
equimolar in [Cp2Fe]PF6 and t-Bu2py under H2 was shown to be
stable for days at room temperature.
standard deviation of 0.01. For BArF -, all CF3 substituents
4
were found to be disordered over two sites. These F3C-C(aryl)
groups were restrained to similar, ideal geometry (esd 0.01 A for
˚
bond lengths and 0.02ꢀ for angles). H atoms were included as
riding idealized contributors. Methyl H atom U’s were assigned
as 1.5Ueq of the carrier atom; remaining H atom U’s were
assigned as 1.2Ueq.
The complex 2F was crystallized from a saturated MeCN
solution at -30 ꢀC. The structure of 2F was solved by direct
methods. Methyl H atom positions, R-CH3, were optimized by
rotation about R-C bonds with idealized C-H, R-H, and
H-H distances. Remaining H atoms were included as riding
idealized contributors. U’s for -CH3 atoms were assigned as
1.5Ueq of the carrier atom. Remaining H atom U’s were assigned
as 1.2ꢀ carrier Ueq.
Reaction of 2F, 3F, and of [2F]þ, [3F]þ with H2. The following
ligand hydrogenolysis experiments were conducted in CD2Cl2
solutions under ∼1 atm of H2 and analyzed by 19F NMR
spectroscopy (H2 BAF -62 ppm). The yields for the reaction
t
of [2F]0/þ with H2 were calculated from the ratio of anion 19F
NMR signal (-59 ppm, BArF4-) to free ligand signal. Treat-
ment of [3F]0/þ with H2 did not yield significant amounts of free
The compound 3F was crystallized from a saturated MeCN
solution at -30 ꢀC. The structure of 3F was solved by direct
methods. Methyl H atom positions, R-CH3, were optimized by
rotation about R-C bonds with idealized C-H, R-H, and
H-H distances. Remaining H atoms were included as riding
idealized contributors. U’s for -CH3 atoms were assigned as
1.5Ueq of the carrier atom. Remaining H atom U’s were assigned
as 1.2ꢀ carrier Ueq.
ligand. Yield of H2 BAF (time): 2F 20% (7 days), [2F]BArF
t
4
100% (∼2 h), 3F none (7 days), [3F]PF6 trace (7 days).
Crystallography. The compound [1F(NCMe)](OTf)2 was crys-
tallized from CH2Cl2 overlayered with hexanes. The structure
was solved by direct methods. The proposed model includes 1.5
idealized hexane solvate molecules per unit cell disordered about
the inversion center, one idealized CH2Cl2 solvate per unit cell
sharing space with the disordered hexane, two triflate anions, one
ordered and one disordered over two sites, and one host cation.
Similar geometry was imposed on the two disordered anion sites
Acknowledgment. This research was sponsored by the
U.S. Department of Energy. We thank Dr. Swarna Latha
Kokatam for preliminary studies on the Ru complexes.
˚
using a standard deviation of 0.01 A. Similar, rigid-bond re-
straints (esd 0.01) were imposed on displacement parameters for
overlapping disordered sites. Ordered methyl H atom positions,
R-CH3, were optimized by rotation about R-C bonds with
idealized C-H, R-H, and H-H distances. Remaining H atoms
were included as riding idealized contributors. Methyl H atom
U’s were assigned as 1.5Ueq of the carrier atom; remaining H
atom U’s were assigned as 1.2Ueq.
Supporting Information Available: Kinetic plots, spectra,
rate calculations. Crystallographic information files (cif) for
[1F(MeCN)](OTf)2, 2F, [2F(MeCN)]BArF4, and 3F. This mate-
rial is available free of charge via the Internet at http://
pubs.acs.org.