Hydrocarbon oxidation by Bis-mu-oxo manganese dimers: electron transfer, hydride transfer, and hydrogen atom transfer mechanisms.

Article abstract of DOI:10.1021/ja020204a

Described here are oxidations of alkylaromatic compounds by dimanganese mu-oxo and mu-hydroxo dimers [(phen)(2)Mn(IV)(mu-O)(2)Mn(IV)(phen)(2)](4+) ([Mn(2)(O)(2)](4+)), [(phen)(2)Mn(IV)(mu-O)(2)Mn(III)(phen)(2)](3+) ([Mn(2)(O)(2)](3+)), and [(phen)(2)Mn(III)(mu-O)(mu-OH)Mn(III)(phen)(2)](3+) ([Mn(2)(O)(OH)](3+)). Dihydroanthracene, xanthene, and fluorene are oxidized by [Mn(2)(O)(2)](3+) to give anthracene, bixanthenyl, and bifluorenyl, respectively. The manganese product is the bis(hydroxide) dimer, [(phen)(2)Mn(III)(mu-OH)(2)Mn(II)(phen)(2)](3+) ([Mn(2)(OH)(2)](3+)). Global analysis of the UV/vis spectral kinetic data shows a consecutive reaction with buildup and decay of [Mn(2)(O)(OH)](3+) as an intermediate. The kinetics and products indicate a mechanism of hydrogen atom transfers from the substrates to oxo groups of [Mn(2)(O)(2)](3+) and [Mn(2)(O)(OH)](3+). [Mn(2)(O)(2)](4+) is a much stronger oxidant, converting toluene to tolyl-phenylmethanes and naphthalene to binaphthyl. Kinetic and mechanistic data indicate a mechanism of initial preequilibrium electron transfer for p-methoxytoluene and naphthalenes because, for instance, the reactions are inhibited by addition of [Mn(2)(O)(2)](3+). The oxidation of toluene by [Mn(2)(O)(2)](4+), however, is not inhibited by [Mn(2)(O)(2)](3+). Oxidation of a mixture of C(6)H(5)CH(3) and C(6)H(5)CD(3) shows a kinetic isotope effect of 4.3 +/- 0.8, consistent with C-H bond cleavage in the rate-determining step. The data indicate a mechanism of initial hydride transfer from toluene to [Mn(2)(O)(2)](4+). Thus, oxidations by manganese oxo dimers occur by three different mechanisms: hydrogen atom transfer, electron transfer, and hydride transfer. The thermodynamics of e(-), H(*), and H(-) transfers have been determined from redox potential and pK(a) measurements. For a particular oxidant and a particular substrate, the choice of mechanism is influenced both by the thermochemistry and by the intrinsic barriers. Rate constants for hydrogen atom abstraction by [Mn(2)(O)(2)](3+) and [Mn(2)(O)(OH)](3+) are consistent with their 79 and 75 kcal mol(-)(1) affinities for H(*). In the oxidation of p-methoxytoluene by [Mn(2)(O)(2)](4+), hydride transfer is thermochemically 24 kcal mol(-)(1) more facile than electron transfer; yet the latter mechanism is preferred. Thus, electron transfer has a substantially smaller intrinsic barrier than does hydride transfer in this system.

Full text of DOI:10.1021/ja020204a

Published on Web 08/01/2002
Hydrocarbon Oxidation by Bis-µ-oxo Manganese Dimers:
Electron Transfer, Hydride Transfer, and Hydrogen Atom
Transfer Mechanisms
Anna S. Larsen, Kun Wang,† Mark A. Lockwood,‡ Gordon L. Rice, Tae-Jin Won,§
Scott Lovell,# Martin Sad´ılek, Frantisˇek Turecˇek, and James M. Mayer*
Contribution from the Department of Chemistry, UniVersity of Washington, Campus Box 351700,
Seattle, Washington 98195-1700
Received February 11, 2002
Abstract: Described here are oxidations of alkylaromatic compounds by dimanganese µ-oxo and µ-hydroxo
dimers [(phen)2MnIV(µ-O)2MnIV(phen)2]4+ ([Mn2(O)2]4+), [(phen)2MnIV(µ-O)2MnIII(phen)2]3+ ([Mn2(O)2]3+), and
[(phen)2MnIII(µ-O)(µ-OH)MnIII(phen)2]3+ ([Mn2(O)(OH)]3+). Dihydroanthracene, xanthene, and fluorene are
oxidized by [Mn2(O)2]3+ to give anthracene, bixanthenyl, and bifluorenyl, respectively. The manganese
product is the bis(hydroxide) dimer, [(phen)2MnIII(µ-OH)2MnII(phen)2]3+ ([Mn2(OH)2]3+). Global analysis of
the UV/vis spectral kinetic data shows a consecutive reaction with buildup and decay of [Mn2(O)(OH)]3+
as an intermediate. The kinetics and products indicate a mechanism of hydrogen atom transfers from the
substrates to oxo groups of [Mn2(O)2]3+ and [Mn2(O)(OH)]3+. [Mn2(O)2]4+ is a much stronger oxidant,
converting toluene to tolyl-phenylmethanes and naphthalene to binaphthyl. Kinetic and mechanistic data
indicate a mechanism of initial preequilibrium electron transfer for p-methoxytoluene and naphthalenes
because, for instance, the reactions are inhibited by addition of [Mn2(O)2]3+. The oxidation of toluene by
[Mn2(O)2]4+, however, is not inhibited by [Mn2(O)2]3+. Oxidation of a mixture of C6H5CH3 and C6H5CD3
shows a kinetic isotope effect of 4.3 ( 0.8, consistent with C-H bond cleavage in the rate-determining
step. The data indicate a mechanism of initial hydride transfer from toluene to [Mn2(O)2]4+. Thus, oxidations
by manganese oxo dimers occur by three different mechanisms: hydrogen atom transfer, electron transfer,
and hydride transfer. The thermodynamics of e-, H•, and H- transfers have been determined from redox
potential and pKa measurements. For a particular oxidant and a particular substrate, the choice of mechanism
is influenced both by the thermochemistry and by the intrinsic barriers. Rate constants for hydrogen atom
abstraction by [Mn2(O)2]3+ and [Mn2(O)(OH)]3+ are consistent with their 79 and 75 kcal mol-1 affinities for
H•. In the oxidation of p-methoxytoluene by [Mn2(O)2]4+, hydride transfer is thermochemically 24 kcal mol-1
more facile than electron transfer; yet the latter mechanism is preferred. Thus, electron transfer has a
substantially smaller intrinsic barrier than does hydride transfer in this system.
Introduction
pound. Many of the same issues arise in biological and
biomimetic oxidations.2 Progress in these areas will be facilitated
The selective oxidation of hydrocarbons by homogeneous,
heterogeneous, and enzymatic agents is of much fundamental
and technological interest.1 The use of hydrocarbons as feed-
stocks for commodity and fine chemicals typically requires an
oxidation step, usually catalyzed by a transition metal com-
by a better mechanistic understanding of how an oxidizing metal
site interacts with hydrocarbons. For C-H bond oxidations,
there are well-documented reactions that involve initial electron
transfer, hydrogen atom transfer, oxidative addition, and elec-
trophilic attack at a C-H bond.3 Described in this report are
oxidations of alkylaromatic compounds by µ-oxo manganese
dimers which occur by electron transfer, hydride transfer, or
hydrogen atom transfer.4
* To whom correspondence should be addressed. E-mail: mayer@
chem.washington.edu.
† Current address: ExxonMobil Research & Engineering Co., Annandale,
NJ 08801.
‡ Current address: Somatocor Pharmaceuticals, Atlanta, GA.
Current address: Neah Power Systems Inc., 22118 20th Ave. SE, Suite
142, Bothell, WA 98021.
Manganese 1,10-phenanthroline (phen) dimers are a good
testing ground for oxidation mechanisms because a number of
§ At UW 1999-2000 on sabbatical leave from Changwon National
University, Changwon, South Korea.
(2) Biomimetic Oxidations Catalyzed by Transition Metal Complexes; Meunier,
B., Ed.; Imperial College Press: London, 2000.
# Current address: Department of Biochemistry, University of Wisconsin,
Madison, WI.
(3) (a) Eberson, L. Electron Transfer Reactions in Organic Chemistry; Springer-
Verlag: Berlin, 1987. (b) Baciocchi, E.; Bietti, Massimo, M.; Lanzalunga,
O. Acc. Chem. Res. 2000, 33, 243-251. (c) Mayer, J. M. Acc. Chem. Res.
1998, 31, 441-450. (d) Crabtree, R. H. Chem. ReV. 1995, 95, 987-1007.
(e) Shilov, A. E.; Shulpin, G. B. Chem. ReV. 1997, 97, 2879-2832. (f)
Stahl, S. S.; Labinger, J. A.; Bercaw, J. E. Angew. Chem., Int. Ed. 1998,
37, 2181-2192.
(1) (a) Sheldon, R. A.; Kochi, J. K. Metal-Catalyzed Oxidation of Organic
Compounds; Academic Press: New York, 1981. (b) Olah, G. A.; Molna´r,
A. Hydrocarbon Chemistry; Wiley: New York, 1995. (c) Special issue on
Bioinorganic Enzymology. Chem. ReV. 1996, 96, 2237-3042. (d) Weis-
sermel, K.; Arpe, H.-J. Industrial Organic Chemistry; VCH: New York,
1997.
9
10112
J. AM. CHEM. SOC. 2002, 124, 10112-10123
10.1021/ja020204a CCC: $22.00 © 2002 American Chemical Society

Hydrocarbon Oxidation by Manganese Dimers
A R T I C L E S
compounds are available, with different oxidation states and
different levels of protonation.5-8 These and related manganese
complexes have long been studied as model systems for the
manganese cluster active site of the oxygen evolving complex
in Photosystem II.9-11 There has also been much parallel activity
in oxidations by di-µ-oxo complexes of iron, copper, and nickel
as models for various metalloenzymes and biomimetic oxidation
systems.12-20
have shown that a Marcus-type treatment using self-exchange
rates can be successful for hydrogen atom transfer.25 This
follows similar demonstrations for organic hydride transfer
reactions by Kreevoy et al.26 and for proton transfer reactions
of metal hydrides by Kristja´nsdo´ttir and Norton.27 Studies of
hydride transfer have been facilitated by Parker’s electrochemi-
cal determination of hydride affinities,28 which has been used
in understanding reactions of metal hydrides.29
The development of selective oxidants will be aided by
understanding what mechanisms are possible and how each
mechanism is affected by the properties of the oxidant and
substrate. We argue here that whether reactions occur by electron
transfer, hydride transfer, or hydrogen atom transfer depends
both on the energetics and on the intrinsic barriers of the various
pathways. The separation of driving force and intrinsic barriers
is a characteristic of Marcus theory, which was developed for
electron transfer reactions.21 A similar separation has been
observed for organic hydrogen atom transfer reactions, where
rate constants within a given class correlate with enthalpic
driving force (∆H, from bond strengths) via the Polanyi
relation.22 Studies in our laboratories and elsewhere have found
that hydrogen atom abstraction from alkanes and arylalkanes
by Cr, Mn, Fe, and Cu complexes can also be understood on
the basis of the bond strengths involved.23,24 Most recently, we
Results
I. Synthesis and Characterization of Manganese Dimers.
[(phen)2MnIV(µ-O)2MnIV(phen)2](ClO4)4 and [(phen)2MnIV(µ-
O)2MnIII(phen)2](PF6)3 (abbreviated [Mn2(O)2]4+ and [Mn2-
(O)2]3+) have been previously prepared and structurally
characterized.5-8 The bridging hydroxide dimers [(phen)2MnIII-
(µ-O)(µ-OH)MnIII(phen)2](PF6)3 ([Mn2(O)(OH)]3+) and [(phen)2-
MnIII(µ-OH)2MnII(phen)2](PF6)3 ([Mn2(OH)2]3+) were each
generated from [Mn2(O)2]3+and stoichiometric amounts of
hydroquinone, following the chemistry of the bipyridine ana-
logue.30 The compounds were characterized by elemental
analysis, IR and optical spectra, mass spectrometry, and cyclic
voltammetry. Solution IR spectra of [Mn2(O)2]3+ show a band
at 686 cm-1 6 which is characteristic of M2(µ-O)2 cores.31
,
Exchange with H218O is rapid (as is typical of Mn2(µ-O)2
compounds10,11) and shifts this band to 665 cm-1. Identification
of the hydroxide products in reaction mixtures can be prob-
lematic because their optical and IR spectra lack distinct bands.
Important support for the assignment of these materials was
obtained by analysis of their redox state by iodometric titration32
and/or by addition of hydroquinone and monitoring of its
(4) (a) Wang, K.; Mayer, J. M. J. Am. Chem. Soc. 1997, 119, 1470-1471. (b)
Lockwood, M. A.; Wang, K.; Mayer, J. M. J. Am. Chem. Soc. 1999, 121,
11894-11895.
(5) (a) Goodwin, H. A.; Sylva, R. N. Aust. J. Chem. 1965, 18, 1743-1749.
(b) Goodwin, H. A.; Sylva, R. N. Aust. J. Chem. 1967, 20, 629-637.
(6) Cooper, S. R.; Calvin, M. J. Am. Chem. Soc. 1977, 99, 6623-6630.
(7) Manchanda, R.; Brudvig, G. W.; de Gala, S.; Crabtree, R. H. Inorg. Chem.
1994, 33, 5157-5160.
(8) Stebler, M.; Ludi, A.; Bu¨rgi, H.-B. Inorg. Chem. 1986, 25, 4743-4750.
(9) (a) Tommos, C.; Babcock, G. T. Acc. Chem. Res. 1998, 31, 18-25. (b)
Wieghardt, K. Angew. Chem., Int. Ed. Engl. 1989, 28, 1153-1172. (c)
Pecoraro, V. L.; Hsieh, W.-Y. In Metals in Biological Systems; Sigel, A.,
Sigel, H., Eds.; Marcel-Dekker: Basel, Switzerland, 2000; Vol. 37, pp 429-
504.
1
quantitative conversion to benzoquinone by H NMR.
Cyclic voltammetry of [Mn2(O)2]3+ in MeCN shows quasi-
reversible waves at 0.90 and -0.01 V versus Cp2Fe+/° (Figure
1A), consistent with previous studies.6,8 These are assigned to
the redox couples [Mn2(O)2]4+/[Mn2(O)2]3+ and [Mn2(O)2]3+
/
(10) Limburg, J.; Vrettos, J. S.; Chen, H.; de Paula, J. C.; Crabtree, R. H.;
Brudvig, G. W. J. Am. Chem. Soc. 2001, 123, 423-430.
[Mn2(O)2]2+. [Mn2(O)(OH)]3+ and [Mn2(OH)2]3+ show quasi-
reversible reductions at -0.03 and -0.74 V, respectively.
Addition of acid to CV solutions causes protonation of the
reduced forms, as indicated by the reduction in anodic current
(Figure 1B). One equivalent of [PhNH3]ClO4 (pKa ) 10.5633)
(11) Baldwin, M.; Law, N. A.; Stemmler, T. L.; Kampf, J. W.; Penner-Hahn, J.
E.; Pecoraro, V. L. Inorg. Chem. 1999, 38, 4801-4809.
(12) Que, L., Jr. J. Chem. Soc., Dalton Trans. 1997, 3933-3940. (b) Chen, K.;
Que, L., Jr. J. Am. Chem. Soc. 2001, 123, 6327-6337.
(13) Lee, D.; Lippard, S. J. J. Am. Chem. Soc. 2001, 123, 4611-4612.
(14) Me´nage, S.; Galey, J. B.; Dumats, J.; Hussler, G.; Seite´, M.; Luneau, I.
G.; Chottard, G.; Fontecave, M. J. Am. Chem. Soc. 1998, 120, 13370-
13382.
(15) Tolman, W. B. Acc. Chem. Res. 1997, 30, 227-237.
(25) Roth, J. P.; Yoder, J. C.; Won, T.-J.; Mayer, J. M. Science 2001, 294,
2524-2526.
(16) Obias, H. V.; Lin, Y.; Murthy, N. N.; Pidcock, E.; Solomon, E. I.; Ralle,
M.; Blackburn, N. J.; Neuhold, Y.-M.; Zuberbu¨hler, A. D.; Karlin, K. D.
J. Am. Chem. Soc. 1998, 120, 12960-12961.
(26) (a) Kreevoy, M. M.; Oh, S. J. Am. Chem. Soc. 1973, 95, 4805. (b) Lee,
I.-S. H.; Jeoung, E. H.; Kreevoy, M. M. J. Am. Chem. Soc. 2001, 123,
7492.
(17) Mahadevan, V.; Henson, M. J.; Solomon, E. I.; Stack, T. D. P. J. Am.
Chem. Soc. 2000, 122, 10249-10250.
(27) Kristja´nsdo´ttir, S. S.; Norton, J. R. J. Am. Chem. Soc. 1991, 113, 4366-
(18) Elliott, S. J.; Zhu, M.; Tso, L.; Nguyen, H.-H. T.; Yip, J. H.-K.; Chan, S.
I. J. Am. Chem. Soc. 1997, 119, 9949-9955.
4367.
(28) (a) Cheng, J.-P.; Handoo, K. L.; Parker, V. D. J. Am. Chem. Soc. 1993,
115, 2655-2660. (b) Handoo, K. L.; Cheng, J.-P.; Parker, V. D. J. Am.
Chem. Soc. 1993, 115, 5067-5072. (c) Parker uses a conversion from
measured potentials versus Cp2Fe+/° to values versus NHE of +0.528 V
(refs 28a,b, Parker, V. D., personal communication, 1999), which differs
from the value for Cp2Fe+/° in acetonitrile quoted by others (cf., Connelly,
N. G.; Geiger, W. E. Chem. ReV. 1996, 96, 877-910).
(29) (a) Cheng, T. Y.; Brunschwig, B. S.; Bullock, R. M. J. Am. Chem. Soc.
1998, 120, 13121-13127; 1999, 121, 3150-3155. (b) Sarker, N.; Bruno,
J. W. J. Am. Chem. Soc. 1998, 120, 2174-2180. (c) Berning, D. E.; Noll,
B. C.; DuBois, D. L. J. Am. Chem. Soc. 1998, 120, 11432-11447.
(30) Ghosh, M. C.; Reed, J. W.; Bose, R. N.; Gould, E. S. Inorg. Chem. 1994,
33, 73-78.
(19) Itoh, S.; Bandoh, H.; Nakagawa, M.; Nagatomo, S.; Kitagawa, T.; Karlin,
K. D.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 11168-11178.
(20) Shiren, K.; Ogo, S.; Fujinami, S.; Hayashi, H.; Suzuki, M.; Uehara, A.;
Watanabe, Y.; Moro-oka, Y. J. Am. Chem. Soc. 2000, 122, 254-262.
(21) Marcus, R. A.; Sutin, N. Biochim. Biophys. Acta 1985, 811, 265. Sutin, N.
Prog. Inorg. Chem. 1983, 30, 441.
(22) (a) Kochi, J. K., Ed. Free Radicals; Wiley: New York, 1973; Vol. 1, pp
275-331. (b) Tedder, J. M. Angew. Chem., Int. Ed. Engl. 1982, 21, 401-
410.
(23) (a) Cook, G. K.; Mayer, J. M. J. Am. Chem. Soc. 1994, 116, 1855-1868
(correction, J. Am. Chem. Soc. 1994, 116, 8859). (b) Cook, G. K.; Mayer,
J. M. J. Am. Chem. Soc. 1995, 117, 7139-7156. (c) Gardner, K. A.; Mayer,
J. M. Science 1995, 269, 1849-1851. (d) Gardner, K. A.; Kuehnert, L. L.;
Mayer, J. M. Inorg. Chem. 1997, 36, 2069-2078. (e) Roth, J. P.; Mayer,
J. M. Inorg. Chem. 1999, 38, 2760-2761. (f) Lockwood, M. A.; Blubaugh,
T. J.; Collier, A. M.; Lovell, S.; Mayer, J. M. Angew. Chem., Int. Ed. 1999,
38, 225-227. (g) Reference 4a.
(24) (a) Jonas, R. T.; Stack, T. D. P. J. Am. Chem. Soc. 1997, 119, 8566-8567.
Goldsmith, C. R.; Jonas, R. T.; Stack, T. D. P. J. Am. Chem. Soc. 2002,
124, 83-96. (b) Bakac, A. J. Am. Chem. Soc. 2000, 122, 1092-1097. (c)
Strassner, T.; Houk, K. N. J. Am. Chem. Soc. 2000, 122, 7821-7822. (d)
Reitz, J. B.; Solomon, E. I. J. Am. Chem. Soc. 1998, 120, 11467-11478.
(31) Costas, M.; Rohde, J. U.; Stubna, A.; Ho, R. Y. N.; Quaroni, L.; Mu¨nck,
E.; Que, L., Jr. J. Am. Chem. Soc. 2001, 123, 12931-12932 and references
therein.
(32) (a) Jeffery, G. H.; Bassett, J.; Mendham, J.; Denney, R. C. Vogel’s Textbook
of QuantitatiVe Chemical Analysis, 5th ed.; Wiley: New York, 1989; pp
384-391. (b) Lee, D. G.; Perez-Benito, J. F. J. Org. Chem. 1988, 53, 5725-
5728.
(33) Izutsu, K. Acid-Base Dissociation Constants in Dipolar Aprotic SolVents;
IUPAC Chemical Data Series No. 35; Blackwell Scientific Publications:
Boston, MA, 1990.
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10113

A R T I C L E S
Larsen et al.
solutions, however, decompose rapidly in the presence of excess
water or nitrogen bases (pyridine, 2,6-lutidine, 2,6-di-tert-
butylpyridine, 1,10-phenanthroline), giving predominantly
[Mn2(O)2]3+ (by UV/vis). Comproportionation of [Mn2(O)2]4+
and [Mn2(O)(OH)]3+ occurs rapidly to form [Mn2(O)2]3+
,
although the fate of the proton is uncertain. Adding protons to
[Mn2(O)2]3+ (excess triflic acid) generates [Mn2(O)2]4+ in ca.
50% yield, apparently by disproportionation. Surprisingly,
comproportionation of [Mn2(O)2]3+and [Mn2(OH)2]3+ is very
slow; the optical spectrum of a mixture of 1 mM solutions of
these materials is identical to the sum of the components’
spectra. Spectra of the mixture change slowly over 3 days,
consistent with partial formation of [Mn2(O)(OH)]3+, but
[Mn2(O)2]3+ is still observed.
II. Oxidations by [Mn2(O)2]3+: Hydrogen Atom Transfer
Reactions. (A) Reaction with 9,10-Dihydroanthracene (DHA).
[Mn2(O)2]3+ reacts with DHA over 12 h at 65 °C in MeCN,
with a change in color from olive green to light brown (eq 1).
GC/MS analysis of the products revealed the presence of
anthracene and traces of anthrone and anthraquinone. Iodometric
titration of the isolated manganese product gave an average
oxidation state of 2.37 ((0.04), consistent with the predominant
formation of the MnIIMnIII dimer, [Mn2(OH)2]3+. The observed
products account almost quantitatively for the manganese
oxidizing equivalents consumed, so eq 1 is close to a balanced
reaction. When reaction 1 is run under oxygen, substantially
more anthrone and anthraquinone are formed (quantitation is
difficult because of the lesser but competing autoxidation of
anthracene under these conditions). The influence of O2 is
consistent with the involvement of carbon radicals. Oxidation
of a mixture of DHA and DHA-d12 (15 µmol each) by [Mn2-
(O)2]3+ (15 µmol) at 55 °C gave an anthracene to anthracene-
Figure 1. Cyclic voltammograms of 2.5 mM [Mn2(O)2]3+ in MeCN (100
mV s-1, nBu4NPF6, vs Cp2Fe+/°) (A); with added 2.0 mM lutidine H+ (B).
Scheme 1. Redox Potentials (vs Cp2Fe+/°), pKa Values, H• and
H- Affinities in MeCN
d10 ratio of 4.2 ( 0.3 (55 °C) (by GC/MS), indicating a primary
isotope effect of ∼4.
The oxidation of excess DHA by [Mn2(O)2]3+, as monitored
by UV/vis spectroscopy, shows no induction period (Figure 2).
Absorbance versus time traces indicate a consecutive reaction
causes complete loss of the anodic current for [Mn2(O)2]3+
,
while 1 equiv of NH4PF6 (pKa ) 16.4633) has no effect. With
2,4-lutidinium perchlorate (pKa ) 14.0533), three independent
measurements indicated a pKa of 14.6 ( 0.5 (following the
procedure of Baldwin and Pecoraro34). Similarly, titrations of
solutions of [Mn2(O)(OH)]3+ with PhNH3ClO4 yielded a pKa
for [Mn2(OH)2]3+ of 11.5 ( 0.5. The conclusion that the
reduced forms are basic is consistent with the reported aqueous
electrochemistry.35 The redox potentials and pKa values are
summarized in Scheme 1 (the values for H• and H- addition
will be discussed below).
pattern, with the formation and decay of [Mn2(O)(OH)]3+
,
because the absorbance at 526 nm increases and then decreases
as the reaction proceeds. An isosbestic point at 610 nm is
observed in the first phase of reaction, [Mn2(O)2]3+
f
[Mn2(O)(OH)]3+, which goes away when the intermediate is
converted to [Mn2(OH)2]3+. The 400-800 nm spectral-kinetic
data were processed by SpecFit software36 utilizing a kinetic
model in which both [Mn2(O)2]3+and [Mn2(O)(OH)]3+ oxidize
DHA to the 9-hydroanthracenyl radical (HA•), which is rapidly
oxidized to anthracene (A) by both oxidants (Scheme 2).
Trapping of HA• (kA3, kA4) is expected to be much faster than
reaction with DHA, as the C-H bond strength is substantially
weaker in the radical. SpecFit refinements started from ap-
All of the dimers are fairly stable in acetonitrile solution (the
solvent for all of the reactions reported here). [Mn2(O)2]4+
(34) Baldwin, M. J.; Pecoraro, V. L. J. Am. Chem. Soc. 1996, 118, 11325-
11326.
(35) Manchanda, R.; Thorp, H. H.; Brudvig, G. W.; Crabtree, R. H. Inorg. Chem.
1992, 31, 4040-4041. Attempts in our hands to repeat this aqueous
electrochemistry gave evidence (e.g., color changes) that [Mn2(O)2]3+ was
not stable under our conditions.
(36) Binstead, R. A.; Jung, B.; Zuberbu¨hler, A. D. Specfit/32 Global Analysis
System; Spectrum Software Associates, Malborough, MA; specsoft@
compuserve.com.
9
10114 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002

Hydrocarbon Oxidation by Manganese Dimers
A R T I C L E S
Table 1. Rate Constants for the Oxidation of Hydrocarbons by [Mn2O2]3+ in Acetonitrile
substrate (D[C−H]a)
temp, °C
k1, M-1 s-1 b
k2, M-1 s-1 b
k3, M-1 s-1
k4, M-1 s-1
DHAc (78)
32
40
47
54
64
20
30
40
50
55
65
1.2 × 10-3
2.4 × 10-3
4.3 × 10-3
7.3 × 10-3
1.4 × 10-2
6.7 × 10-3
2.2 × 10-2
3.1 × 10-2
6.9 × 10-2
6.2 × 10-4
1.3 × 10-3
3.0 × 10-4
6.7 × 10-4
1.2 × 10-3
1.9 × 10-3
4.4 × 10-3
6.4 × 10-3
1.9 × 10-2
2.7 × 10-2
6.7 × 10-2
4.3 × 10-4
1.0 × 10-3
g1 × 105 d
g1 × 104 d
xanthenee (76)
fluorenee (80)
2 × 105 f
3 × 105 f
5 × 105 f
6 × 105 f
2 × 104 f
3 × 104 f
2 × 104 f
3 × 104 f
5 × 104 f
6 × 104 f
2 × 103 f
3 × 103 f
d
a Substrate C-H bond dissociation energies, in kcal mol-1, from Table 3. b Estimated errors (20%. c Kinetic model in Scheme 2. kA3/kA4 = 10, and kA3
g 1 × 105. e Kinetic model in Scheme 3. f Estimated errors (50%.
Figure 3. Visible spectra of [Mn2(O)2]3+
, , and
[Mn2(O)(OH)]3+
[Mn2(OH)2]3+ measured on independent samples (broken lines) and
determined from the global kinetic fit of the data in Figure 2 according to
Scheme 2 (solid lines).
Simultaneous refinement of the four rate constants by SpecFit
was unsuccessful, so kA3 and kA4 were entered as fixed
parameters and varied manually. Good fits could only be
obtained with kA3 = 10kA4 and when both were >104 M-1 s-1
.
The kA3/kA4 ratio in large part determines how much [Mn2(O)-
(OH)]3+ builds up as an intermediate. The globally optimized
fits give the rate constants shown in Table 1.
Various checks provide good support for the kinetic model
in Scheme 2. Figure 2 shows the close agreement between
experimental and modeled absorbance versus time changes. The
spectra predicted for the colored species by the global fit closely
match the observed spectra for [Mn2(O)2]3+, [Mn2(O)(OH)]3+
,
Figure 2. Spectral changes observed for the reaction of [Mn2O2]3+ and
DHA in acetonitrile solution at 32 °C. (A) Overlay of the spectra, with the
first spectrum marked by a heavier line. Arrows indicate the sequence and
direction of absorbance changes at 526 and 700 nm. (B) and (C):
Absorbance changes at 526 and 700 nm, respectively; open circles are
experimental data; solid line is from the global kinetic fit (Scheme 2).
and [Mn2(OH)2]3+ (Figure 3). Independent experiments showed
that [Mn2(O)(OH)]3+ oxidizes DHA to yield anthracene and
trace amounts of anthraquinone and anthrone, and the rate
constant at 50 °C was in good agreement with the kinetic
simulation result for kA2
.
Scheme 2. Kinetic Model A, for Oxidation of DHA by [Mn2(O)2]3+
Rate constants determined over the temperature range of 32-
64 °C (Table 1) give the activation parameters ∆HA1 ) 15.2
q
(10) kcal mol-1, ∆SA1q ) -22 (3) eu; ∆HA2q ) 16.2 (10) kcal
mol-1, ∆SA2q ) -22 (3) eu. The negative entropies of activation
are consistent with bimolecular rate-limiting steps and are similar
to what we have observed in other hydrogen atom transfer
processes.3c,23 The rate constants reported here are somewhat
different from those given in our preliminary report,4a in part
because of more sophisticated kinetic modeling and in part
because the reactions proceed more slowly with more purified
materials.
proximate values of 2kA1 and 2kA2 that were obtained from
pseudo-first-order absorbance/time plots at 700 and 526 nm (the
factor of 2 is needed to account for the rapid trapping of HA•).
Initial spectra were used to obtain kA1, while kA2 was derived
from spectral changes after the loss of the isosbestic point.
To address whether the cleavage of the dimers occurs during
DHA oxidation, a ligand scrambling experiment was carried
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10115

A R T I C L E S
Larsen et al.
Scheme 3. Kinetic Model B, for Oxidation of Xanthene (XnH) by
[Mn2(O)2]3+
out. A 1:1 mixture of [Mn2(O)2]3+ and an analogue with 4,7-
dimethylphenanthroline (Me2phen) ligands was reacted with
DHA. (The Me2phen analogue oxidizes DHA at a rate compa-
rable to that of the phen complex.) After a little more than 1
half-life, analysis of the manganese complexes present by FAB-
MS showed (phen)4 and (Me2phen)4 complexes with only a very
small amount of mixed-ligand dimers. A mixture of dimers
3+
[Mn2(O)2(phen)x(Me2phen)(4-x)
]
was independently synthe-
sized from a mixture of phen and Me2phen and would have
been observable by FAB-MS. Thus, the dimers remain intact
through the reaction.
(B) Reactions with Fluorene and Xanthene. Fluorene (10
mM) is oxidized by [Mn2(O)2]3+ (1 mM) at 50 °C to a roughly
equal mixture of fluorenone (FldO, 0.28 mM) and 9,9-
bifluorenyl (Fl2, 0.20 mM). Xanthene is similarly oxidized by
a roughly 10:1 mixture of xanthone (XndO) and 9,9′-bixan-
thenyl (Xn2). Xn2 and Fl2 are most likely formed by coupling
of their respective radicals. The presence of fluorenyl radicals
is further supported by reaction in the presence of the radical
trap CBrCl3, which yields 9-bromofluorene (6%), fluorenone
(45%), and traces of bifluorenyl and 9-trichloromethylfluorene
(∼0.5% each).
Figure 4. Spectral changes observed for the reaction of [Mn2(O)2]3+ and
xanthene in acetonitrile solution at 40 °C (in both cases, only a few traces
are shown for clarity). (A) Spectra obtained by stopped-flow over the first
600 s of a reaction of 1.0 mM [Mn2(O)2]3+ and 13 mM xanthene (first
trace at 10 s; every 60th spectrum shown). (B) Spectra obtained using a
diode array spectrophotometer over 7000 s of a reaction of 1.0 mM
[Mn2(O)2]3+ and 12 mM xanthene (first trace at 10 m, spectra taken every
2 min, only a few spectra shown).
Xanthene oxidation is about an order of magnitude faster than
that of DHA; fluorene is about an order of magnitude slower
(Table 1). Global fitting of the UV/vis spectral data with SpecFit
used the kinetic model in Scheme 3. While similar to the scheme
used for DHA reactions, in this case, radical trapping by
[Mn2(O)2]3+ and [Mn2(O)(OH)]3+ (kB3, kB4, see below) is
assumed to lead to ketone products because neither Xn• nor Fl•
has an abstractable hydrogen. The partitioning between ketone
and dimer products is due to competition between trapping by
manganese and radical coupling (kB5), which occurs at 109 M-1
s-1 37
kB1 was determined by global analysis of just the first
.
Figure 5. Concentration profiles of all species for the reaction between
[Mn2(O)2]3+ and xanthene derived from global kinetic analysis of the data
phase of the reaction (which had to be monitored on a stopped-
flow instrument for xanthene). With these values fixed, the
spectral changes were analyzed by SpecFit over the whole
reaction varying only kB2 (Figure 4) to give the values in Table
1. Over the range of 20-50 °C, the activation parameters for
in Figure 4 (according to Scheme 3). [O, [Mn2(O)2]3+; b, [Mn2(O)(OH)]3+
9, [Mn2(OH)2]3+; 0, xanthyl radical; 4, xanthone; 2, bixanthyl.]
;
fits to the spectral data and to match the observed final ratios
of XndO to Xn2 (or FldO to Fl2) determined by GC-MS
analysis. As found for the DHA reaction, the buildup of
[Mn2(O)(OH)]3+ requires kB3 = 10kB4; their magnitude is set
by competition with radical coupling (kB5). kB3 for xanthene
varies from 2 × 105 to 6 × 105 M-1 s-1 in the 20-50 °C
temperature range (Table 1). The concentration profiles from
the kinetic fit are shown in Figure 5; it can be seen that xanthone
is predominantly produced in the early stages of reaction when
the concentrations of [Mn2(O)2]3+ and [Mn2(O)(OH)]3+ are
high. The uncertainties for kB3 and kB4 are estimated to be
(50%, because the fitting procedure is not very sensitive to
their changes, and the product ratios are uncertain by ∼15%.
q
q
xanthene are ∆HB1 ) 13.3 (1) kcal mol-1, ∆SB1 ) -23 (3)
eu; ∆HB2 ) 13.4 (1) kcal mol-1, ∆SB2 ) -23 (3) eu.
Equations B3 and B4 encompass multiple reaction steps and
likely involve multiple manganese compounds and trace water.
They are written in this nonbalanced form because no more
information is available. It is possible that other manganese
products may be formed in addition to the dimers shown,
although introduction of additional species in the kinetic model
does not improve the overall fit and results in meaningless
predicted spectra. kB3 and kB4 were varied both to give good
q
q
(37) Arends, I. W. C. E.; Mulder, P.; Clark, K. B.; Wayner, D. D. M. J. Phys.
Chem. 1995, 99, 8182-8189.
9
10116 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002

Hydrocarbon Oxidation by Manganese Dimers
A R T I C L E S
(o:m:p ) 42:7:40 by GC/MS; eq 3). These products, whose
identities were confirmed by independent synthesis,39 are often
formed when benzyl cation (PhCH2+) is an intermediate and it
alkylates the excess toluene present.40 No bibenzyl was ob-
served, but small amounts of methyl-benzophenones, benz-
aldehyde, and N-benzylacetamide are formed. The yields of
benzaldehyde and N-benzylacetamide are larger when undried,
HPLC grade MeCN is used. Independent oxidation of diphe-
nylmethane by [Mn2(O)2]4+ gave exclusively benzophenone;
competitive oxidation of toluene and diphenylmethane indicates
the latter to be ∼180 times more reactive per benzylic hydrogen.
Spectrophotometric titration of the final manganese containing
product gave an average oxidation state of 3.52 (( 0.05),32b
consistent with predominant formation of [Mn2(O)2]3+, which
is also indicated by the UV/vis spectra of completed reactions.
The observed products in the toluene oxidation account for
∼90% of the manganese oxidizing equivalents consumed.
Oxidation of an equimolar mixture of C6H5CH3 and C6H5-
CD3 gives d 0, d 2, d 3, and d 5 isotopomers of each of the three
isomeric tolyl-phenylmethane products. The d 2 products, for
Figure 6. Molecular structure of Mn(η1-OClO3)2(phen)2.
A confirmation of this kinetic model is its ability to predict
the changes of the XndO to Xn2 ratio depending on the initial
conditions. Increasing the initial xanthene concentration 4 times
is predicted to decrease XndO/Xn2 by a factor of 2.6, in
reasonable agreement with the observed ratio of 3.8. Quadru-
pling the initial concentration of [Mn2(O)2]3+ was predicted to
increase XndO/Xn2 2.1 times, the same as the experimental
value. As a further check on the model, the predicted molar
absorptivities of the absorbing species are close to the experi-
mentally determined spectra.
Addition of water to the oxidation of fluorene shifts the
product ratio substantially toward fluorenone over bifluorenyl
(g10:1). Addition of water at the end of the reaction had no
effect. In the presence of excess 20% enriched H218O, the
isolated fluorenone is 25 ( 5% enriched, consistent with the
rapid exchange of the oxo ligands with H218O mentioned above.
II. Oxidations by [Mn2(O)2]4+: Electron Transfer and
Hydride Transfer Reactions. DHA is oxidized by [Mn2(O)2]4+
over hours at ambient temperatures to give anthracene, anthrone,
and anthraquinone (eq 2). Faster oxidation of DHA by
[Mn2(O)2]4+ is consistent with it being a more powerful oxidant
than [Mn2(O)2]3+ (E1/2 ) +0.90 vs -0.01 V vs Cp2Fe+/° in
MeCN).
+
instance, result from attack of C6H5CD2 on C6H5CH3. The
ratios of products indicate a kinetic isotope effect of 4.3 ( 0.8
on the formation of benzyl cation, consistent with C-H bond
cleavage in the rate-determining step. (The apparent isotope
effects vary among the isomers and with reaction conditions,
consistent with previous studies of related reactions.41)
p-Xylene is oxidized by [Mn2(O)2]4+ to give 2,4′,5-trimeth-
yldiphenylmethane as the primary organic product (by GC-MS,
1H and 13C NMR; eq 4), as expected for an attack of the benzylic
cation on the hydrocarbon. As in most of the oxidations of
methylbenzenes by [Mn2(O)2]4+, small amounts of analogously
substituted benzophenone and acetamide products are also
typically observed. m-Xylene reacts similarly (eq 5). As a
(38) (a) McCann, M.; Casey, M. T.; Devereux, M.; Curran, M.; McKee, V.
Polyhedron 1997, 16, 2741-2748. (b) Geraghty, M.; McCann, M.; Casey,
M. T.; Curran, M.; Devereux, M.; McKee, V.; McCrea, J. Inorg. Chim.
Acta 1998, 277, 257-262.
(39) Friedel-Crafts addition of benzyl chloride to toluene: Hayashi, E.;
Takahashi, Y.; Itoh, H.; Yoneda, N. Bull. Chem. Soc. Jpn. 1993, 66, 3520-
3521.
The average oxidation of the final manganese product in eq
2 was 2.0 (( 0.05), and the organic products account for 85%
of manganese oxidative equivalents consumed. Layering the
reaction solution with ether precipitates a manganese(II) product,
Mn(OClO3)2(phen)2. Its X-ray structure shows monomeric
pseudo-octahedral molecules, with each perchlorate ion bound
η1 to Mn (Tables S1 and S2; Figure 6). The structure is similar
to that of [Mn(H2O)2(phen)2]2+ with somewhat longer Mn-O
bond lengths, as expected for perchlorate versus water.38
Toluene is oxidized by [Mn2(O)2]4+ in acetonitrile over 12
h at 65 °C, yielding primarily tolyl-phenylmethane products
(40) (a) Uemura, S.; Ikeda, T.; Tanaka, S.; Okano, M. J. Chem. Soc., Perkin
Trans. 1 1978, 2574-2576. (b) Darbeau, R. W.; White, E. H.; Song, F.;
Darbeau, N. R.; Chou, J. J. Org. Chem. 1999, 64, 5966-5978. (c) Andrulis,
P. J., Jr.; Dewar, M. J. S.; Dietz, R.; Hunt, R. L. J. Am. Chem. Soc. 1966,
88, 5473-5478. (d) Uemura, S.; Tanaka, S.; Okano, M. J. Chem. Soc.,
Perkin Trans. 1 1976, 1966-1969. (e) Uemura, S.; Ikeda, T.; Tanaka, S.;
Okano, M. J. Chem. Soc., Perkin Trans. 1 1979, 2574-2576. (f) McKillop,
A.; Turrell, A. G.; Young, D. W.; Taylor, E. C. J. Am. Chem. Soc. 1980,
102, 6504-6512. (g) Necsoiu, I.; Ghenciulescu, A.; Rentea, M.; Rentea,
C. N.; Nenitzescu, C. D. ReV. Roum. Chim. 1967, 12, 1503-1510. (h) King,
S. T. J. Catal. 1991, 131, 215-225. (i) Bryant, J. R.; Taves, J. E.; Mayer,
J. M. Inorg. Chem. 2002, 41, 2769-2776. (j) Reference 4b.
(41) Effenberger, F.; Maier, A. H. J. Am. Chem. Soc. 2001, 123, 3429-3433
and references therein.
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10117

A R T I C L E S
Larsen et al.
Table 2. Rate Constants for the Oxidation of Hydrocarbons by [Mn2O2]4+ in Acetonitrile
substrate
k (M-1 s-1 a
)
substrate
k (M-1 s-1 a
)
IE (eV)b
c
C6H5CH3
m-C6H4(CH3)2
2.9 × 10-5
2.9 × 10-4
2 × 10-3
1.0 × 101
naphthalene
2-methylnaphthalene
1-methylnaphthalene
4.6
8.14
d
5.1 × 102
6.2 × 102
1.8 × 103
3.4 × 102
7.96(3)
7.91(6)
7.78(10)
e
p-C6H4(CH3)2
p-MeOC6H4CH3
1.1 × 102
2,6-dimethylnaphthalene
p-MeOC6H4CH3 + [Mn2O2]3+ f
2,6-dimethylnaphthalene + [Mn2O2]3+ g
a Rate constants ((e20%) for the reaction of [Mn2O2]4+ with the given substrate at 298 K in MeCN. Experimental details are given in the Experimental
Section; products are given in the text. For the first three entries, these are calculated from ∆Hq and ∆Sq measured at higher temperatures. b Ionization
energies in eV.50 c From rate constants measured at 308-358 K: ∆Hq ) 16.8 ( 1.0 kcal mol-1 and ∆Sq ) -23 ( 3 cal K-1 mol-1 d From rate constants
.
measured at 298-328 K: ∆Hq ) 20 ( 2 kcal mol-1 and ∆Sq ) -6 ( 4 cal K-1 mol-1 e From rate constants measured at 308-338 K: ∆Hq ) 16 ( 2
.
kcal mol-1 and ∆Sq ) -15 ( 4 cal K-1 mol-1
× 10-4 M; initial [Mn2O2]4+ 9.5 × 10-5 M).
.
f With added [Mn2O2]3+ (3.7 × 10-4 M; initial [Mn2O2]4+ 8.0 × 10-5 M). g With added [Mn2O2]3+ (2.3
mechanistic probe, the oxidation of isodurene (1,2,3,5-tetra-
methylbenzene) was explored; commercial material containing
some durene (1,2,4,5-C6H2Me4) was used, and recrystallization
does not significantly change the purity. Two predominant
products formed (eq 6), along with a number of other materials
in trace quantities. The major products result from attack of
the 2,4,6-trimethylbenzyl cation on durene and isodurene, as
confirmed by observing the same products from Friedel-Crafts
alkylation of the isodurene/durene mixture by ZnCl2/2,4,6-
trimethylbenzyl chloride.39
occur on the stopped-flow time scale. Spectral overlays show
an isosbestic point at 668 nm and are consistent with a simple
[Mn2(O)2]4+ f [Mn2(O)2]3+ reaction pattern. Both single
wavelength and global analyses of the spectra showed first-
order kinetics to approximately 3 half-lives. After this time,
significant retardation of the reaction rate is often observed,
apparently due to the buildup of the reduced species, [Mn2(O)2]3+.
Addition of 4.6 equiv of [Mn2(O)2]3+ to an independent oxida-
tion of p-methoxytoluene by [Mn2(O)2]4+ led to a reduction in
rate of about an order of magnitude. Similarly, addition of 2.4
equiv [Mn2(O)2]3+ to an oxidation of 2,6-dimethylnaphthalene
slowed the rate by roughly a factor of 5. The pseudo-first-order
rate constants varied linearly with substrate concentration with
a zero intercept within the estimated errors, except for toluene
where decomposition of [Mn2(O)2]4+ was slightly competitive
under some conditions (<10%). Selected second-order rate
constants and activation parameters are given in Table 2.
Discussion
para-Methoxytoluene reacts much more quickly with [Mn2-
(O)2]4+ than do the alkyl-substituted toluenes, while p-nitro-
toluene is inert to oxidation. para-Methoxytoluene gives
primarily biaryl species, with only ∼2% of the substituted
diarylmethanes (eq 7). Oxidation of CD3C6H4OMe gives only
the biaryl products, indicating that there is a larger isotope effect
(kH/kD) for formation of the diarylmethanes.
I. Overview of the Reactions. The oxo-bridged manganese
dimers are versatile, multielectron oxidants capable of oxidizing
alkylaromatic compounds via different pathways to different
products. Because of the basicity of the reduced forms, these
reagents often accept both electrons and protons. Mechanisms
of alkylarene oxidations have received much attention,1-3,40,42
and the pathways relevant to the chemistry described here are
illustrated in Scheme 4, using toluene as an example. Initial
electron transfer forms an arene radical cation, in what is
typically an uphill preequilibrium step. Arene radical cations
form biaryls (as observed here for naphthalenes and p-
methoxytoluene) or can be deprotonated to benzylic radicals.
Radicals can also be formed directly from the arene by hydrogen
atom removal. Radicals can couple (as observed here for
xanthene and fluorene), can lose another hydrogen (as in the
conversion of DHA to anthracene), or can be oxidized to
carbocations. Benzylic carbocations can also be formed by direct
removal of a hydride ion. Under the nonnucleophilic conditions
used here, the benzylic cation predominantly attacks the excess
arene present in a Friedel-Crafts alkylation, forming diaryl-
methane products.
Naphthalene is oxidized by [Mn2(O)2]4+ to give binaphthyl
as the only identifiable organic product (GC-MS, NMR). The
average oxidation state of the isolated manganese product was
3.54 (( 0.05), consistent with the formation of the [Mn2(O)2]3+
.
The yield of binaphthyl is essentially quantitative. 1-Methyl-,
2-methyl-, and 2,6-dimethyl-naphthalene are oxidized faster than
naphthalene, and GC/MS analysis shows formation of binaph-
thyl products and (in the case of the 2-methyl derivative) some
naphthalene carboxaldehyde.
The kinetics of [Mn2(O)2]4+ oxidations were studied under
pseudo-first-order conditions. The toluene and xylene reactions
are slow enough to be studied with a diode-array spectropho-
tometer, but the p-methoxytoluene and naphthalene reactions
II. Rate-Limiting Steps. (A) Hydrogen Atom Abstraction
by [Mn2(O)2]3+. A wide range of evidence points to the
(42) (a) Eberson, L. J. Am. Chem. Soc. 1983, 105, 3192-3199. (b) Reference
40c. (c) Baciocchi, E.; Rol, C.; Mandolini, L. J. Am. Chem. Soc. 1980,
102, 7597-7598. (d) Baciocchi, E.; D’Acunzo, F.; Galli, C.; Lanzalunga,
O. J. Chem. Soc., Perkin Trans. 2 1996, 133-140. (e) Reed, R. A.; Murray,
R. W. J. Phys. Chem. 1986, 90, 3829-3835. (f) Schlesener, C. J.; Amatore,
C.; Kochi, J. K. J. Phys. Chem. 1986, 90, 3747-3756. (g) Del Giacco, T.;
Baciocchi, E.; Steenken, S. J. Phys. Chem. 1993, 97, 5451-5456.
9
10118 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002

Hydrocarbon Oxidation by Manganese Dimers
A R T I C L E S
Table 3. Thermochemical Properties of Oxidants and Substrates
+
BDE(X−H)a
D(X −H-)b
E
1/2 (V)c
pK d
IE (eV)e
a
[(phen)2Mn(O)2Mn(phen)2]3+
0.90
-0.01
[(phen)2Mn(O)(OH)Mn(phen)2]3+
79
75
90f
122
118
14.6
11.5
50.3
[(phen)2Mn(OH)2Mn(phen)2]3+
c
C6H5CH3
∼2.1
8.83
m-C6H4(CH3)2
8.55(2)
8.44(5)
7.90(5)
p-C6H4(CH3)2
112
107
53.0
54.4
p-MeOC6H4CH3
9,10-dihydroanthracene
xanthene
fluorene
naphthalene
1.29
78g
76g
74g
i
∼1.5h
90
30
7.65(2)
7.91(2)
8.14
1.55
2-methylnaphthalene
1-methylnaphthalene
2,6-dimethylnaphthalene
7.96(3)
7.91(6)
7.78(10)
a Bond dissociation energies, in kcal mol-1. (Mn)O-H bonds from application of the thermochemical cycle in ref 44. b Heterolytic bond dissociation
energies to remove H-, from ref 28 for the organic compounds. c Potential for oxidation, in MeCN versus Cp2Fe+/°. Values for organic compounds are from
ref 43. d In MeCN; values for organic compounds from ref 28. e Ionization energies; ref 50. f Reference 54. g See ref 23c for references. h Estimated as being
the same as durene.43 i The aromatic C-H bond strength in benzene is 113 kcal mol-1 50
.
The ability of a metal oxidant to abstract H• is related to its
thermodynamic affinity for H•, in other words, the strength of
the bond it can make to a hydrogen atom.3c,23,24 This bond
strength is in essence the sum of the affinity for H+ (the pKa)
and e- (the redox potential, E). The bond strengths shown in
Scheme 1, 79 kcal mol-1 for [Mn2(O)(O-H)]3+ and 75 kcal
mol-1 for [Mn2(OH)(O-H)]3+, were derived using the ther-
mochemical cycle of Tilset and Parker44 (following Bordwell45).
Scheme 4. Mechanisms for Oxidation of Alkylaromatic
Compounds
The C-H bond strength in DHA is 78 kcal mol-1 23d
so H-atom
,
transfer to the manganese dimers is roughly thermoneutral. The
bond strengths thus provide a qualitative understanding of why
these reactions occur, and they are also quantitative predictors
of reactivity. Rate constants for hydrogen atom abstraction from
DHA (per hydrogen, at 30 °C) roughly correlate with bond
strengths for abstraction by various metal complexes and oxygen
radicals, as shown in Figure 7. The line drawn in this figure is
defined by the values for tBuO• and sBuOO•, to show a perfect
correlation. Such a Bell-Evans-Polanyi correlation is typical
of hydrogen atom transfer reactions.22 The rates of H-atom
abstraction by [Mn2(O)2]3+ and [Mn2(O)(OH)]3+ also correlate
with the substrate C-H bond strengths (Table 1).
intermediacy of carbon radicals in the reactions of [Mn2(O)2]3+
.
The formation of bixanthyl and bifluorenyl implicates the
presence of xanthyl and fluorenyl radicals. This is further
supported by the trapping with CBrCl3 to produce 9-bromo-
fluorene, as well as the effect of added O2 on the reactions.
The primary isotope effect is consistent with initial H-atom
transfer. Optical monitoring of the reactions shows that the first
step is conversion of [Mn2(O)2]3+ to [Mn2(O)(OH)]3+ by net
addition of H•. [Mn2(O)(OH)]3+ is not a result of compropor-
tionation of [Mn2(O)2]3+ and [Mn2(OH)2]3+ because this
reaction is slow. The success of the kinetic models with multiple
checks supports the proposed mechanisms.
The carbon radicals are formed by a one-step hydrogen atom
transfer mechanism (which could alternatively be described as
a coupled proton and electron transfer). An alternative pathway
of initial electron transfer and subsequent deprotonation (Scheme
4) is ruled out by the redox potentials involved. Electron transfer
from DHA to [Mn2(O)2]3+ is uphill by ∼1.5 V or ∆G° = 34
kcal mol-1, much larger than the observed ∆Gq of 21 kcal mol-1
for DHA oxidation. The potential for DHA is conservatively
estimated as equal to that of durene (1,2,4,5-tetramethylben-
zene), and following the redox potentials assembled by Eber-
son.43 A mechanism of initial proton transfer is even less likely,
as [Mn2(O)2]3+ is a very weak base and DHA a very weak
acid.
We have recently shown that rates of hydrogen atom transfer
reactions depend on both bond strengths (driving force) and
intrinsic barriers, the latter being revealed in self-exchange
rates.25 Unfortunately, hydrogen-atom self-exchange rates are
not easily determined for [Mn2(O)2]3+/[Mn2(O)(OH)]3+ or for
[Mn2(O)(OH)]3+/[Mn2(OH)2]3+, all being NMR silent. The
Polanyi correlation in Figure 7 essentially assumes that the
manganese dimers’ intrinsic barriers are similar to those for the
oxygen radicals and for the iron tris(biimidazoline) system we
have examined.46 The fact that the manganese dimers lie below
the line in Figure 7 may reflect a larger intrinsic barrier,8 which
would be consistent with the slow comproportionation of
[Mn2(O)2]3+and [Mn2(OH)2]3+
.
(B) Electron Transfer to [Mn2(O)2]4+. The dehydrocoupling
of arenes to biaryl derivatives (the Scholl reaction47a) occurs
by an electron transfer mechanism (Scheme 4).3b,40f,42d,48 The
(43) (a) Reference 3a, p 44, based primarily on (b) Howell, J. O.; Goncalves, J.
M.; Amatore, C.; Klasnic, L.; Wightman, R. M.; Kochi, J. K. J. Am. Chem.
Soc. 1984, 106, 3968-3976 and ref 42f. (c) E° for durene and Cp2Fe given
in ref 3a as 2.07 and 0.53 V versus NHE in MeCN. (d) It should be noted
that a range of potentials have been reported in different sources for
alkylaromatic compounds, and no source gives a consistent list for all of
the substrates reported here.
(44) Parker, V. D.; Handoo, K. L.; Roness, F.; Tilset, M. J. Am. Chem. Soc.
1991, 113, 7493-7498.
(45) cf. (a) Bordwell, F. G.; Gheng, J.-P.; Satish, G.-Z.; Zhang, A. V. J. Am.
Chem. Soc. 1991, 113, 9790-9795. (b) Bordwell, F. G.; Liu, W.-Z. J. Am.
Chem. Soc. 1996, 118, 8777-8781. (c) Parker, V. D. J. Am. Chem. Soc.
1992, 114, 7458-7462; correction, J. Am. Chem. Soc. 1993, 115, 1201.
(46) Roth, J. P.; Lovell, S.; Mayer, J. M. J. Am. Chem. Soc. 2000, 122, 5486.
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10119

A R T I C L E S
Larsen et al.
indicate a partially cationic arene at the transition state. This
could occur either by hydride or by electron transfer. Similarly,
isodurene oxidation proceeding predominantly at the central
methyl group (eq 6) is consistent with initial hydride or electron
transfer but not hydrogen atom transfer.52
Electron transfer from toluene, however, is quite difficult
because of its very high redox potential, 2.61 V versus NHE in
MeCN (0.53 V higher than naphthalene).43a Electron transfer
from toluene to [Mn2(O)2]4+ is ∼1.2 V (27 kcal mol-1
)
.
endergonic, larger than the observed ∆Gq ) 23.6 kcal mol-1
Given that back electron transfer cannot be faster than 1010 M-1
s-1, the equilibrium constant of ∼10-20 means that the forward
rate must be slower than 10-10 M-1 s-1 at 298 K. Electron
transfer is therefore not kinetically competent to be a preequi-
librium step in toluene oxidation by [Mn2(O)2]4+ with k ) 2.9
(( 0.3) × 10-5 M-1 s-1 at 298 K.53 Although there is some
uncertainty associated with the redox potential for toluene, it
would have to be at least 0.25 V lower to be consistent with
rate-limiting electron transfer. In addition, the primary kinetic
isotope effect indicates that initial electron transfer would have
to be a preequilibrium, followed by rate-limiting deprotonation
of the toluene radical cation. Yet toluene oxidation is not
inhibited by addition of [Mn2(O)2]3+. The data are only
consistent with a mechanism of initial hydride transfer (Scheme
1). The [Mn2(O)2]3+ product is apparently formed by com-
proportionation of [Mn2(O)2]4+ with the initially formed
[Mn2(O)(OH)]3+. We have been unable to determine the fate
of the proton in either the oxidations or the comproportionation,
because of the reactivity of [Mn2(O)2]4+ with bases.
Figure 7. Plot of log k (the rate constant for hydrogen abstraction, per
hydrogen) versus the strength of the X-H bond formed by the oxidants.
The straight line is drawn through the two oxygen radical points.
oxidations of naphthalenes and p-methoxytoluene are additional
examples. A mechanism of preequilibrium electron transfer is
indicated by the inhibition of the oxidations of 2,6-dimethyl-
naphthalene and p-methoxytoluene by the reduced form of the
oxidant, [Mn2(O)2]3+ 40c,42a
.
Electron transfer is a reasonable pathway for [Mn2(O)2]4+
because this reagent, unlike [Mn2(O)2]3+, is a strong outersphere
oxidant (E1/2 ) +0.90 V vs Cp2Fe+/o). Electron transfer from
naphthalene to [Mn2(O)2]4+ is uphill by only ∼0.6 V (∆G° )
14 kcal mol-1, Keq = 10-10). The Keq indicates a maximum
electron transfer rate of about 1 M-1 s-1 if the reverse reaction
proceeds at 1010 M-1 s-1. This is reasonably consistent with
the room-temperature rate constant of 4.6 M-1 s-1 given the
uncertainty in the hydrocarbon redox potentials.43,49 The mea-
sured rates of oxidations of naphthalenes roughly parallel the
relative Keq’s for electron transfer, as estimated from the
naphthalene ionization energies (Table 2).50 This correlation,
however, ignores changes in the rates of the step(s) that follow.
(C) Hydride Transfer to [Mn2(O)2]4+. The formation of
substituted diarylmethanes from toluene and xylenes clearly
implicates the presence of a benzylic cation which then adds to
excess arene. The N-benzylacetamide is formed by the addition
of a benzylic cation to acetonitrile solvent, followed by
hydrolysis (the Ritter reaction47b). The large substituent effect
on the rates - for example, p-xylene oxidized 68 times faster
than toluene (a factor of 34 per methyl group) - corresponds
to a Hammett F+ of roughly -5. The large substituent effects
are inconsistent with a hydrogen atom transfer mechanism51 and
III. Understanding the Choice of the Rate-Determining
Step. The first step in understanding the preferred mechanism
for each oxidant/substrate combination is to determine the
energetics of each possible rate-limiting step. [Mn2(O)2]3+ and
[Mn2(O)(OH)]3+ are hydrogen atom abstracting agents because
they form fairly strong bonds to H• (79 and 75 kcal mol-1), a
result of their high affinity for both an electron and a proton
(H• ≡ H+ + e-). H-atom abstractions from DHA and xanthene
[D(C-H) ) 78, 76 kcal mol-1 23d
]
are close to thermoneutral,
is
while abstraction from toluene [D(C-H) ) 90 kcal mol-1 54
]
too far uphill. [Mn2(O)2]4+ has a much higher redox potential
than [Mn2(O)2]3+, but it does not act as a hydrogen atom
abstractor because its reduced form ([Mn2(O)2]3+) has es-
sentially no basicity.
[Mn2(O)2]4+ acts as an electron transfer oxidant because of
its high redox potential, while the 0.91 V lower potential for
[Mn2(O)2]3+ precludes it oxidizing hydrocarbons by electron
transfer. [Mn2(O)2]4+ reacts by electron transfer for the lower-
potential substrates, p-methoxytoluene and naphthalenes, but it
is apparently not a potent enough outersphere oxidant to remove
an electron from toluene. Instead, a hydride transfer mechanism
is adopted. Parker et al. have reported that removal of H- from
(47) March, J. AdVanced Organic Chemistry: Reactions, Mechanisms, and
Structure, 4th ed.; McGraw-Hill: New York, 1992; (a) p 539, (b) pp 879-
880.
toluene requires 118 kcal mol-1 28
, which is essentially the same
(48) Tanaka, M.; Nakashima, H.; Fujiwara, M.; Ando, H.; Souma, Y. J. Org.
Chem. 1996, 61, 788-792.
as the hydride affinity of [Mn2(O)2]4+, -122 kcal mol-1
(Scheme 1; these values were derived using the same thermo-
chemical cycle in the same solvent, so they are directly
(49) A reviewer has pointed out that one would normally expect ∆Gq > ∆G°
because of the intrinsic barrier λ, which would increase the discrepancy.
Yet when ∆G°′ ) λ, the Marcus equation predicts ∆Gq ) ∆G°′ (ignoring
work terms). In these highly endoergic reactions, the transition state closely
resembles the products.
(50) IEs from NIST Standard Reference Database Number 69 - July 2001
Release (http://webbook.nist.gov/chemistry/) except for 2,6-dimethylnaph-
thalene from Frey, J. E.; Andrews, A. M.; Combs, S. D.; Edens, S. P.;
Puckett, J. J.; Seagle, R. E.; Torreano, L. A. J. Org. Chem. 1990, 55, 606.
(b) IEs are good indicators of relative redox potentials.42f,43b
(51) Howard, J. A.; Chenier, J. H. B. J. Am. Chem. Soc. 1973, 95, 3054-3055.
(52) Baciocchi, E.; Rol, C.; Mandolini, L. J. Org. Chem. 1977, 42, 3682-3686.
(53) Similar arguments in favor of hydride transfer and ruling out electron
transfer can be found in Bunting, J. W. Bioorg. Chem. 1991, 19, 456-491
and references therein.
(54) Bierbaum, V.; DePuy, C.; Davico, G.; Ellison, B. Int. J. Mass Spectrom.
Ion Phys. 1996, 156, 109-131.
9
10120 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002

Hydrocarbon Oxidation by Manganese Dimers
A R T I C L E S
Conclusions
comparable). Therefore, the proposed hydride transfer rate-
limiting step for toluene oxidation has ∆G° = -4 kcal mol-1
(eq 8).
This paper compares and contrasts the reactivity of three
dimanganese complexes with various alkylaromatic substrates.
[(phen)2MnIV(µ-O)2MnIII(phen)2](PF6)3 ([Mn2(O)2]3+) and
[(phen)2MnIII(µ-O)(µ-OH)MnIII(phen)2](PF6)3 ([Mn2(O)(OH)]3+)
oxidize 9,10-dihydroanthracene, xanthene, and fluorene by initial
hydrogen atom abstraction. The radicals formed either transfer
a second hydrogen (yielding anthracene), dimerize (to bixan-
thenyl, bifluorenyl), or form C-O bonds on reactions with the
oxidant (to give xanthone, fluorenone). [(phen)2MnIV(µ-O)2-
MnIV(phen)2](ClO4)4 ([Mn2(O)2]4+) oxidizes naphthalenes and
para-methoxytoluene to binaphthyl and biphenyl derivatives,
by a mechanism of preequilibrium electron transfer. Oxidation
of toluene by [Mn2(O)2]4+, however, yields predominantly
ortho- and para-tolyl-phenylmethanes and appears to proceed
by initial hydride transfer.
The thermodynamic ability of the manganese dimers to accept
e-, H•, or H- has been determined from redox potentials and
pKa values, using established thermocycles (Scheme 1). These
values are critical to understanding what determines which
pathway is followed. [Mn2(O)2]3+ and [Mn2(O)(OH)]3+ react
by hydrogen atom abstraction because they can form strong
O-H bonds (79 and 75 kcal mol-1, respectively). They are not,
however, good outersphere oxidants, so electron transfer is not
favorable. In contrast, [Mn2(O)2]4+ does not react by hydrogen
atom abstraction because that requires accepting both an elec-
tron and a proton, and [Mn2(O)2]3+ has very low basicity.
[Mn2(O)2]4+ can react either by electron transfer - it is a
powerful one-electron oxidant - or it can accept a hydride ion
(two electrons and a proton) to form the stable dimer
[Mn2(O)(OH)]3+. There is a large thermodynamic bias for
[Mn2(O)2]4+ to accept a hydride rather than an electron from
alkylaromatic compounds. For toluene, this bias is 31 kcal mol-1
and is large enough that the reaction proceeds by hydride
transfer. However, [Mn2(O)2]4+ oxidizes para-methoxytoluene
by electron transfer despite hydride transfer being 24 kcal mol-1
more energetically favorable. Electron transfer must be intrinsi-
cally more facile than hydride transfer in this system; it has a
substantially smaller intrinsic barrier.
[Mn2(O)2]4+ + PhCH3 f [Mn2(O)(OH)]3+ + PhCH2
+
∆G° = -4 kcal mol-1 (8)
For [Mn2(O)2]4+ oxidizing toluene, there is a thermochemical
bias of ∼31 kcal mol-1 favoring hydride transfer over electron
transfer. This bias is large enough for hydride transfer to be the
mechanism used. p-Methoxytoluene, however, is oxidized by
electron transfer even though hydride transfer is 24 kcal mol-1
easier: ∆G°(hydride transfer) = -15 kcal mol-1 and
∆G°(electron transfer) = +9 kcal mol-1. Evidently, this 24 kcal
mol-1 bias is not sufficient to favor hydride transfer. Electron
transfer is evidently inherently more facile than hydride transfer;
in other words, electron transfer has a much smaller intrinsic
barrier than does hydride transfer.
IV. Radical Trapping by Manganese Dimers. The mech-
anisms for DHA, xanthene, and fluorene oxidation (Schemes
2, 3) involve trapping of intermediate radicals by [Mn2(O)2]3+
and [Mn2(O)(OH)]3+. This trapping could occur either by
electron transfer or by radical addition to a bridging oxo group.
Electron transfers from xanthyl and fluorenyl radicals (Xn•, Fl•)
to [Mn2(O)2]3+ have ∆G° ) -0.27 and +0.35 V, respectively.28
Uphill electron transfer from Fl• is difficult to rationalize with
the trapping rate constant of 2 × 104 M-1 s-1 at 50 °C. Most
telling, Fl• is 0.63 V harder to oxidize than Xn•, yet is trapped
only 40 times slower. It is much more likely that trapping occurs
by addition to a bridging oxo ligand and formation of a carbon-
oxygen bond. Subsequent oxidation of the alkoxide ligand to
the observed ketone (XndO, FldO) should be quite rapid. C-O
bond formation must be highly exothermic, almost like a radical
recombination, because formation of an O-H bond by H•
addition has ∆H° g 75 kcal mol-1. The observation of faster
trapping by [Mn2(O)2]3+ versus [Mn2(O)(OH)]3+ is consistent
with their H-atom affinities (although perhaps a smaller effect
would have been predicted given the early transition state for
these exothermic reactions).
It is interesting to compare the rates of radical trapping by
t
the hydrogen atom abstractors shown in Figure 7. BuO• and
Experimental Section
sBuOO• react with carbon radicals at essentially the diffusion
General Methods. All manipulations were done in dry solvents
under a N2 atmosphere unless otherwise noted. Toluene, p-xylene, and
m-xylene (Aldrich) were dried over Na/Ph2CO and vacuum transferred
prior to use. Dihydroanthracene (DHA, Aldrich) was recrystallized from
water/ethanol, and fluorene (Aldrich) was purified by sublimation.
DHA-d12 was synthesized from anthracene-d10;56 DHA-d4 and fluorene-
d2 were prepared by exchange using sodium dimsylate and DMSO-d6,
following ref 57, and C6H5CD358 and CH3OC6H5CD340c were prepared
as noted. Toluene-d8 (Cambridge Isotope Laboratories, Inc.) was dried
over sodium and vacuum transferred prior to use. Dry electrochemical
grade acetonitrile (Burdick & Jackson) was degassed and dispensed in
the glovebox. Mn(OAc)2 (Aldrich), 1,10-phenanthroline monohydrate
limit by radical recombination, in other words, by C-O bond
formation. MnO4 is also a diffusion-limited radical trap, for
instance, reacting with CH3 at 1.05 × 109 M-1 s-1 in H2O.55
-
•
The permanganate reaction also likely occurs by C-O bond
formation, as CH3• is not easy to oxidize. Because of the rapid
trapping, MnO4- oxidations do not form radical dimer products
such as bifluorenyl; they resemble the rebound mechanism
discussed for hydroxylations by cytochrome P450.2 In reactions
of [Fe(Hbim)(H2bim)2]2+, we have no evidence for trapping of
Xn• or Fl• by the oxidant.23e In sum, the rates of H-atom
abstraction by MnO4-, [Mn2(O)2]3+, and [Fe(Hbim)(H2bim)2]2+
vary by less than 104 (the H-atom affinities vary by 4 kcal
mol-1), but the rates of radical trapping appear to vary by g107.
Perhaps steric effects are more important in these radical
recombination-type processes than in the hydrogen atom
abstractions, or perhaps terminal oxo ligands are inherently more
reactive than bridging oxo ligands toward carbon radicals.
(Lancaster), perchloric acid (Aldrich, 70% solution in water), and H218
O
(Cambridge Isotopes) were used as received. nBu4NPF6 (98%, Aldrich)
was triply recrystallized from EtOH, melted under vacuum, and dried
under vacuum for 2 days. Other reagents were purchased from Aldrich
and used as received. Literature procedures were used to prepare 2-
(56) Bass, K. C. Org. Synth. 1958, 42, 48-49.
(57) Bailey, R. J.; Card, P. J.; Schechter, H. J. Am. Chem. Soc. 1983, 105, 6096-
6103.
(58) Johnstone, R. A. W.; Price, P. J. Tetrahedron 1985, 41, 2493-2501.
(55) Steenken, S.; Neta, P. J. Am. Chem. Soc. 1982, 104, 1244-1248.
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10121

A R T I C L E S
Larsen et al.
and 4-methyldiphenylmethanes, 2,4′,5-trimethyldiphenylmethane, 2,3′,6-
trimethyldiphenylmethane, and 2,4,4′-trimethyldiphenylmethane.59 Ring
coupling of 4-methylanisole was achieved with NOBF4,60 and the
benzylic products were prepared following ref 40a.
large amount of Et2O, and dried in vacuo (∼90%). Average Mn
oxidation state: 3.07 (3). Anal. Calcd for C48H33N8F18Mn2O2P3: C,
44.39; H, 2.56; N, 8.63. Found: C, 44.51; H, 2.63; N, 8.62. FAB-MS:
1153 [(M - PF6)+], 973 [(M - PF6 - phen)+]. IR (Nujol): 3380 (b),
1521 (s), 1457 (s), 1376 (s), 837 (s), 720 (s), 695 (m), 658 (w), 644
(w), 557 (s) cm-1. UV/vis (MeCN), no distinct features; a few ꢀ values
are listed, 526 (930), 556 (700), 700 (300).
[(phen)2Mn(µ-OH)2Mn(phen)2](PF6)3. This compound was pre-
pared following the procedure for [Mn2(O)(OH)]3+, using 1.0 equiv
of hydroquinone. Average Mn oxidation state: 2.66 (3). Anal. Calcd
for C48H34N8F18Mn2O2P3: C, 44.36; H, 2.64; N, 8.62. Found: C, 45.58;
H, 2.79; N, 8.44. FAB-MS: 1154 [(M - PF6)+], 974 [(M - PF6 -
phen)+]. IR (Nujol): 3377 (b), 1519 (s), 1460 (s), 1372 (s), 838 (s),
720 (s), 642 (w), 556 (s) cm-1. UV/vis (MeCN), no distinct features;
a few ꢀ values are listed, 526 (370), 556 (280), 700 (90).
Procedure for Reactions of [Mn2(O)2]3+. [Mn2(O)2]3+ (80 mg, 60
µmol) and DHA (21.6 mg, 120 µmol) in 3.0 mL of MeCN were heated
at 65 °C for 11 h. An aliquot of the solution was analyzed by GC. The
solvent was removed in vacuo, and the resulting solid was washed
thoroughly with benzene to remove the organic products, filtered, and
dried. The average Mn oxidation state was 2.37 (4). The oxidation of
fluorene (20 mM) by [Mn2(O)2]3+ (10 mM) in the presence of CBrCl3
(1 M) was left at room temperature for 15 h, and the solution of fluorene
in MeCN was run through a column of activated alumina before use.
Reaction of [Mn2(O)2]4+ with DHA; Preparation of (phen)2Mn-
(OClO3)2. A solution of [Mn2(O)2]4+ (90.6 mg, 69 µmol), excess DHA,
and MeCN was allowed to stand overnight at room temperature, during
which the solution color changed from red/brown to yellow. GC/MS
analysis showed anthracene (0.026 mmol), anthrone (0.009 mmol), and
anthraquinone (0.018 mmol). Retention times, fragmentation patterns,
and 1H NMR spectra were consistent with authentic samples. The
average oxidation of the final Mn product was 2.0 (( 0.05). The
remaining solution was layered with ether to give yellow crystalline
Mn(OClO3)2(phen)2.
Infrared spectra were recorded on a PE1720 FTIR with DTGS
detector. For both background and sample spectra, 32 scans were
recorded at a resolution of 2 cm-1. Solution IR spectra were recorded
in acetonitrile, with manganese dimer concentrations of 15-20 mM,
using a ZnSe cell with a 0.1 mm path length; acetonitrile peaks were
removed by spectral subtraction. Kinetic experiments and optical spectra
[reported as λ (nm) (ꢀ, M-1 cm-1)] were done by monitoring the
absorbance over the 350-820 nm range using a HP8452A diode array
spectrophotometer equipped with a custom built aluminum cell block
holder connected to a Peltier electronic control module ((0.1 °C) or a
HP8453 diode array spectrophotometer equipped with a multicell
transport and a constant-temperature bath. Fast reactions were monitored
using an OLIS rapid scanning monochromator equipped with an OLIS
USA stopped-flow apparatus and a constant-temperature bath. Global
analysis of the kinetic data was performed with OLIS or SpecFit
software (Spectrum Software Associates, NC).36 Product analyses were
done using an HP5890 series GC equipped with a SUPELCO Nukol
fused silica capillary column and a FID detector. Mass spectometric
analyses were performed on an HP5890/5971 GC-MS instrument using
a Supelco SPB-1 capillary column. FAB mass spectra were taken on
a VG 70 SEQ tandem hybrid instrument of EBqQ geometry.
[(phen)2Mn(µ-O)2Mn(phen)2](ClO4)4 ([Mn2(O)2]4+). This com-
pound was prepared from MnCl3(phen)‚H2O following literature
methods.5 Warning: Perchlorate salts are potentially explosiVe and
should be handled with care. UV/vis (CH3CN): 580 (1550), 494 (3440),
410 (5050). IR (KBr pellet): 3468 (br), 3049 (w), 1629 (s), 1607 (s),
1582 (s), 1520 (s), 1427 (s), 1341 (s), 1315 (s), 1226(s), 1095 (br),
850 (vs), 718 (vs), 686 (m), 656 (m), 626 (m) cm-1. The average Mn
oxidation state was 3.95 ((0.05) by adding hydroquinone to
1
[Mn2(O)2]4+ in CD3CN and monitoring by H NMR, integrating the
Reactions of [Mn2(O)2]4+ with Methylbenzenes. In a typical
procedure, a solution of [Mn2(O)2]4+ (30.0 mg, 23 mmol), MeCN (5.0
mL), and toluene (1.0 mL, 9.4 mmol) was heated at 65 °C for 12 h,
causing a color change to yellow-green. GC/MS analysis revealed the
presence of ortho, meta, and para-tolyl-phenylmethanes (10 µmol,
o:m:p ) 42:7:40; comparing retention times and fragmentation patterns
with authentic samples39), PhCHO (0.90 µmol), substituted benzophe-
nones (0.16 µmol), and N-benzylacetamide (trace). Removal of the
solvent in vacuo left a green/brown residue which was washed with
ether and dried; the average Mn oxidation state was 3.52 ((0.05).
p-Xylene gave 2,4′,5-trimethyldiphenylmethane (78%), a trace of the
analogous benzophenone, and p-methyl N-benzylacetamide (19%);
m-xylene gave 2,3′,4- and 2,3′,6-trimethyldiphenylmethanes (33% and
40%) as the major products. The average oxidation states of the final
manganese products were 3.43 (p), 3.50 (m) ( 0.05; the organic
products account for >95% (p), 80% (o) of the manganese oxidizing
equivalents consumed. Ph2CH2 (0.19 g, 1.1 mmol) and [Mn2(O)2]4+
(89 mg, 68 µmol) in MeCN (11.0 mL) for 16 h at 50 °C gave Ph2CO
(120 ( 20%); average Mn oxidation state in the product ) 2.8 ( 0.1.
Reactions of [Mn2(O)2]4+ with Naphthalenes. In a typical proce-
dure, a solution of [Mn2(O)2]4+ (32.5 mg, 25 µmol), naphthalene (20
mg, 160 µmol), and MeCN (5 mL) was stirred overnight at ambient
temperature, turning olive green. GC-MS analysis showed binaphthyl
as the sole organic product; the average oxidation of the final manganese
containing product was 3.54 ((0.05). Reactions with methylnaphtha-
lenes and 4-methylanisole were analogous, except that the color changes
were immediate upon mixing.
benzoquinone formed.
[(phen)2Mn(µ-O)2Mn(phen)2](PF6)3 ([Mn2(O)2]3+). This compound
was prepared as reported7 except product precipitation used NaPF6.
The precipitate was washed with copious amount of Et2O, redissolved
in MeCN, and filtered through Celite. The solvent was then removed
in vacuo to give an olive green solid (∼80%). FAB-MS: 1152 [(M -
PF6)+], 972 [(M - PF6 - Phen)+]. Anal. Calcd for [Mn2(O)2]3+‚2H2O,
C48H36N8F18Mn2O4P3: C, 43.23; H, 2.70; N, 8.40. Found: C, 43.32;
H, 2.37; N, 8.37. UV/vis (MeCN): 526 (580), 556 (460), 684 (554).
IR (Nujol): 3416 (b), 1704 (s), 1629 (s), 1607 (s), 1583 (s), 1521 (s),
1427 (s), 839 (vs), 720 (vs), 686 (m), 659 (m), 642 (m) cm-1. Average
Mn oxidation state: 3.45 (3). [Mn2(18O)2]3+ was prepared by stirring
[Mn2(O)2]3+ and 10 H218O (70% 18O) in MeCN for 30 min and
evacuating to dryness. IR (ν[18O]): 686 (665), 659 (642), 647 (636);
all bands are broad, so the smaller shifts are not well resolved.
[(Me2phen)2Mn(O)2Mn(Me2phen)2](PF6)3. This compound was
prepared analogously to [Mn2(O)2]3+, except that 4,7-dimethylphen-
anthroline was used, and NaPF6 was added to precipitate the product
(80%).7 IR (Nujol): 692 (m) cm-1 (νMn2O2). UV/vis (MeCN): 526
(620), 556 (490), 684 (610). FAB-MS: 1264 [(M - PF6)+]. Using a
50/50 mixture of phen and 4,7-Me2phen gives mixed ligand compounds
by FAB-MS: [Mn2(O)2(Me2phen)x(phen)4-x](PF6)2+, x ) 0-4.
[(phen)2Mn(µ-O)(µ-OH)Mn(phen)2](PF6)3 ([Mn2(O)(OH)]3+). Hy-
droquinone (5.1 mg, 0.045 mmol) was added to a solution of
[Mn2(O)2]3+ (120 mg, 0.09 mmol) in 20 mL of MeCN. The initial green
solution turned brown immediately. The mixture was stirred at room
temperature for 30 min, and 40 mL of Et2O was slowly added to cause
precipitation. The resulting brown solid was filtered, washed with a
Electrochemical Measurements. Cyclic voltammograms were
recorded at room temperature using a BAS B-100W potentiostat under
n
(59) Reference 40d and (a) Tanaka, S.; Uemura, S.; Okano, M. J. Chem. Soc.,
Perkin Trans. 1 1978, 431-434.
(60) Tanaka, M.; Nakasima, H.; Fujiwara, M.; Ando, H.; Souma, Y. J. Org.
Chem. 1996, 61, 788-792.
N2 in freshly distilled dry MeCN with 0.1 M Bu4NPF6 using Pt wire
and Pt disk working and auxiliary electrodes. The reference electrode
was a silver wire in 0.1 M AgNO3/CH3CN; Cp2Fe or (C5Me5)2Fe was
9
10122 J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002

Hydrocarbon Oxidation by Manganese Dimers
A R T I C L E S
used as an internal standard. Because of the quasi-reversible nature of
the couples, the error for E1/2 is estimated as (50 mV. Concentrations
of the compounds were in the range of 1-3 mM, and a scan rate of
100 mV/s was used unless otherwise noted. For [Mn2(O)2]3+, two quasi-
reversible waves were observed, at -0.01 and 0.90 V versus Cp2Fe+/°
((C5Me5)2Fe was used as the standard, and the E1/2 values were
converted as referenced to Cp2Fe). For [Mn2(O)(OH)]3+, one quasi-
reversible wave was observed at -0.03 V versus Cp2Fe+/°, and, for
[Mn2(OH)2]3+, one quasi-reversible wave was observed at -0.74 V
versus Cp2Fe+/° (in both cases, (C5Me5)2Fe as standard). Because of
the quasi-reversible nature of these couples, the error in E1/2 is estimated
to be (50 mV.
with varying concentrations of xanthene. The stopped-flow instrument
was used to collect initial data at 40 and 50 °C, with a spectrum taken
every 1 s. Upon completion, solutions were analyzed by GC/MS using
bifluorenyl as an internal standard (assuming the GC/MS response of
bixanthenyl to be equal to that of bifluorenyl, and using independent
calibration for xanthone). The reaction rates were sensitive to the purity
of the solvent and the concentrations of dissolved H2O and O2; all of
the organic substrate solutions were freshly prepared before each kinetic
run, otherwise much faster rates were observed. In a similar fashion,
toluene (0.25 mL, 2.35 mmol) was added by syringe to an equilibrated
solution of [Mn2(O)2]4+ (2.0 mg, 1.5 µmol) in MeCN (3.0 mL) in the
spectrophotometer. The final solutions were analyzed by GC/MS, and
the kinetic data were fit at 494 or 526 nm or modeled using SpecFit
with similar results. The kinetics of the methyl-naphthalene oxidations
were examined by stopped-flow, monitoring the decay of [Mn2(O)2]4+
at 500 nm; pseudo-first-order plots were linear to 3 half-lives.
Crystallographic Studies. Crystals of [(phen)2Mn(η1-OClO3)2] were
grown by slow diffusion of ether into an acetonitrile solution and
mounted on a glass fiber with epoxy. Data were collected at 161 K on
an Enraf-Nonius KappaCCD diffractometer using Mo KR radiation (λ
) 0.7107 Å). The data were scaled and averaged using HKL
SCALEPACK which applies a correction factor to the reflections
essentially correcting for the absorption anisotropy. All heavy atoms
were refined anisotropically by full matrix least-squares. Hydrogen
atoms were located by difference Fourier methods and refined with a
riding model.
pKa measurements were carried out using cyclic voltammetry
in MeCN following the procedure in ref 34. Protonation of
[Mn2(O)(OH)]3+ was monitored in CVs of [Mn2(O)2]3+ by loss of the
anodic current relative to the cathodic current in the [Mn2(O)2]3+
f
[Mn2(O)2]2+ couple upon titration of an acid with appropriate pKa into
the electrochemical solution. pKa values of the reference acids in MeCN
were from ref 33. Titration using 1 equiv of NH4PF6 (pKa ) 16.46)
had no effect on the CV, while 1 equiv of anilinium (PhNH3ClO4, pKa
) 10.56) causes complete loss of the anodic current. Using 2,4-
lutidinium (LuHClO4, pKa ) 14.05), we determined the pKa of
[Mn2(O)(OH)]3+ to be 14.6 ( 0.5 from three independent measure-
ments. The pKa of [Mn2(OH)2]3+ was measured similarly by monitoring
the loss of the anodic current in the [Mn2(O)(OH)]3+
f
[Mn2(O)(OH)]2+ couple upon titration. Using 1 equiv of either NH4-
PF6 or LuHClO4 has no obvious effect on the CV; by using PhNH3-
ClO4, the pKa was determined to be 11.5 ( 0.5.
Acknowledgment. We are grateful to the NIH (R01 GM50422-
05) for financial support of this work. J.M.M. and M.A.L. also
gratefully acknowledge insightful discussions with Dr. L.
Eberson.
Kinetics. All kinetic runs were done under pseudo-first-order
conditions; the following are typical procedures. A quartz cuvette
equipped with septum screw cap (Spectrocell) containing a solution of
[Mn2(O)2]3+ (4.0 mg, 3.0 mmol) in MeCN (2.5 mL) was equilibrated
at the desired temperature in the spectrophotometer. A 1.0 mL aliquot
of a solution of xanthene (18.1 mg) in MeCN (3 mL) was added to the
cuvette via syringe. Accumulation of spectra was started immediately
and every 180 s for 36 000 s. Three runs were done at each temperature
Supporting Information Available: Details of the X-ray
structure of [(phen)2Mn(η1-OClO3)2] (PDF). This material is
available free of charge via the Internet at http://pubs.acs.org.
JA020204A
9
J. AM. CHEM. SOC. VOL. 124, NO. 34, 2002 10123