Substrate oxidation by copper-dioxygen adducts: Mechanistic considerations

Article abstract of DOI:10.1021/ja045191a

A series of copper-dioxygen adducts [{Cu^II(MePY2) ^R}₂(O₂)](B(C₆F₅) ₄)₂ (1^R), systematically varying in their electronic properties via ligand pyridyl donor substituents (R = H, MeO, and Me₂N), oxidize a variety of substrates with varying C-H or O-H bond dissociation enthalpies. Detailed mechanistic studies have been carried out, including investigation of 1^R thermodynamic redox properties, 1 ^R tetrahydrofuran (THF) and N,N′-dimethylaniline (DMA) oxidation kinetics (including analyses of substrate dicopper binding equilibria), and application of mechanistic probes (N-cyclopropyl-N- methylaniline (CMA) and (p-methoxyphenyl)-2,2-dimethylpropanol (MDP)), which can distinguish if proton-coupled electron-transfer (PCET) processes proceed through concerted electron-transfer proton-transfer (ETPT) or consecutive electron-transfer proton-transfer (ET/PT) pathways. The results are consistent with those of previous complementary studies; at low thermodynamic driving force for substrate oxidation, an ET/PT is operable, but once ET (i.e., substrate one-electron oxidation) becomes prohibitively uphill, the ETPT pathway occurs. Possible differences in coordination structures about 1^Me2N/1 ^Meo compared to those of 1^H are also used to rationalize some of the observations.

Full text of DOI:10.1021/ja045191a

Published on Web 03/26/2005
Substrate Oxidation by Copper-Dioxygen Adducts:
Mechanistic Considerations
†
†
‡
Jason Shearer, Christiana Xin Zhang, Lev N. Zakharov,
‡,¶
,†
Arnold L. Rheingold, and Kenneth D. Karlin*
Contribution from the Department of Chemistry, Johns Hopkins UniVersity,
400 North Charles Street, Baltimore, Maryland 21218, and Department of Chemistry,
UniVersity of Delaware, Newark, Delaware 19716
3
Received August 9, 2004; E-mail: karlin@jhu.edu
II
R
R
Abstract: A series of copper-dioxygen adducts [{Cu (MePY2) }
in their electronic properties via ligand pyridyl donor substituents (R ) H, MeO, and Me
of substrates with varying C-H or O-H bond dissociation enthalpies. Detailed mechanistic studies have
2
(O
2
)](B(C
6
F
5
)
4
)
2
(1 ), systematically varying
2
N), oxidize a variety
R
R
been carried out, including investigation of 1 thermodynamic redox properties, 1 tetrahydrofuran (THF)
and N,N′-dimethylaniline (DMA) oxidation kinetics (including analyses of substrate dicopper binding
equilibria), and application of mechanistic probes (N-cyclopropyl-N-methylaniline (CMA) and (p-methoxy-
phenyl)-2,2-dimethylpropanol (MDP)), which can distinguish if proton-coupled electron-transfer (PCET)
processes proceed through concerted electron-transfer proton-transfer (ETPT) or consecutive electron-
transfer proton-transfer (ET/PT) pathways. The results are consistent with those of previous complementary
studies; at low thermodynamic driving force for substrate oxidation, an ET/PT is operable, but once ET
(
i.e., substrate one-electron oxidation) becomes prohibitively uphill, the ETPT pathway occurs. Possible
Me N MeO
H
differences in coordination structures about 1
2
/1
compared to those of 1 are also used to rationalize
some of the observations.
magnetically coupled dinuclear CuII active site (Scheme 1).2,5,11
Introduction
2
CO performs the catalytic conversion of ortho-catechols to
ortho-quinones, while Tyr will both ortho-hydroxylate phenols
and convert ortho-catechols to ortho-quinones (Scheme 1).7
Tyr facilitates phenol hydroxylations by first coordinating a
phenol (or phenolate) to one of the copper centers followed by
an electrophilic attack of the phenol ring by the peroxo
Copper-mediated oxidations are prominent in chemical and
1-9
2,10
catalytic processes,
while tyrosinase (Tyr), catechol oxi-
,16
2
,11
3,12
dase (CO),
dopamine-â-hydroxylase (DBH),
and pepti-
1
3-15
dylglycine R-hydroxylating monooxygenase (PHM)
are
widely studied copper-containing monooxygenases/oxidases. Tyr
and CO are structurally related, containing a peroxo-bridged
2
,10,17-19
moiety.
In CO substrate dehydrogenation, concerted
†
‡
¶
two-electron (two-proton) oxidation and radical-type proton-
coupled electron-transfer (PCET) reaction mechanisms have
Johns Hopkins University.
University of Delaware.
Current address: Chemistry Department, University of California, San
1
,11,20
been proposed.
DBH and PHM proceed through initial
Diego, La Jolla, CA 92093.
•
(
1) Kitajima, N.; Moro-oka, Y. Chem. ReV. 1994, 94, 737-757.
hydrogen atom (H ) abstraction reactions, which have been
(
2) Solomon, E. I.; Sundaram, U. M.; Machonkin, T. E. Chem. ReV. 1996, 96,
II
-•
proposed to occur from either a mononuclear Cu -O2 or a
2
563-2605.
II
12-15
(
(
3) Klinman, J. P. Chem. ReV. 1996, 96, 2541-2561.
Cu -OOH active-site oxidant.
4) Kopf, M.-A.; Karlin, K. D. In Biomimetic Oxidations; Meunier, B., Ed.;
Imperial College Press: London, 2000; Chapter 7, pp 309-362.
5) Solomon, E. I.; Chen, P.; Metz, M.; Lee, S.-K.; Palmer, A. E. Angew. Chem.,
Int. Ed. 2001, 40, 4570-4590.
To aid in understanding the fundamental properties of the
oxidative process in these metalloenzymes, as well as to
potentially guide the design of oxidation reagents or catalysts,
copper-dioxygen reactivity studies have received considerable
(
(
(
(
6) Que, L., Jr.; Tolman, W. B. Angew. Chem., Int. Ed. 2002, 41, 1114-1137.
7) Quant Hatcher, L.; Karlin, K. D. J. Biol. Inorg. Chem. 2004, 9, 669-683.
8) Mirica, L. M.; Ottenwaelder, X.; Stack, T. D. P. Chem. ReV. 2004, 104,
7
-9,21,22
attention.
The use of copper complexes and dioxygen
1
013-1045.
(
9) Lewis, E. A.; Tolman, W. B. Chem. ReV. 2004, 104, 1047-1076.
10) Decker, H.; Dillinger, R.; Tuczek, F. Angew. Chem., Int. Ed. 2000, 39,
22-27
in organic oxidations is well established,
but the emphasis
(
(
(
(
(
1
591-1595.
11) Gerdemann, C.; Eicken, C.; Krebs, B. Acc. Chem. Res. 2002, 35, 183-
91.
12) Evans, J. P.; Ahn, K.; Klinman, J. P. J. Biol. Chem. 2003, 278, 49691-
9698.
13) Chen, P.; Bell, J.; Eipper, B. A.; Solomon, E. I. Biochemistry 2004, 43,
(16) Sanchez-Ferrer, A.; Rodriguez-Lopez, J. N.; Garcia-Canovas, F.; Garcia-
Carmona, F. Biochim. Biophys. Acta 1995, 1247, 1-11.
(17) Pidcock, E.; Obias, H. V.; Zhang, C. X.; Karlin, K. D.; Solomon, E. I. J.
Am. Chem. Soc. 1998, 120, 7841-7847.
1
4
(18) Itoh, S.; Kumei, H.; Taki, M.; Nagatomo, S.; Kitagawa, T.; Fukuzumi, S.
J. Am. Chem. Soc. 2001, 123, 6708-6709.
5
735-5747.
14) (a) Chen, P.; Solomon, E. I. J. Am Chem. Soc. 2004, 126, 4991-5000. (b)
(19) Yamazaki, S.-i.; Itoh, S. J. Am Chem. Soc. 2003, 125, 13034-13035.
(20) Klabunde, T.; Eicken, C.; Sacchettini, J. C.; Krebs, B. Nat. Struct. Biol.
1998, 5, 1084-1090.
Chen, P.; Solomon, E. I. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 13105-
1
3110.
(
15) Prigge, S. T.; Eipper, B.; Mains, R.; Amzel, L. M. Science 2004, 304, 864-
(21) Stack, T. D. P. Dalton Trans. 2003, 1881-1889.
8
67.
(22) Karlin, K. D.; Gultneh, Y. Prog. Inorg. Chem. 1987, 35, 219-327.
10.1021/ja045191a CCC: $30.25 © 2005 American Chemical Society
J. AM. CHEM. SOC. 2005, 127, 5469-5483
9
5469

A R T I C L E S
Scheme 1
Shearer et al.
3
9
on detailed mechanistic studies employing discrete copper-
dimers) through ET/PT-type pathways. Furthermore, for both
dioxygen adducts is more recent.7 Oxidative N-dealkylation
,9
phenolate oxygenation by a Cu 2-(η :η -µ-peroxo) complex
II
2
2
9
40
reactions of copper ligands are common, with those involving
to give o-catechols and phenol oxidation promoted by a
trinuclear Cu-dioxygen species, it is suggested that substrate
coordination to the metal center is important for subsequent
III
41
intramolecular ligand oxidations of Cu 2-bis-µ-oxo complexes
by far being the best represented.2
8-31
From both theoretical
32,33
and experimental (kinetic) arguments,28-31 it has been suggested
that such reactions occur through an initial rate-limiting
concerted proton-coupled electron-transfer (PCET) reaction (an
ETPT reaction, often referred to as hydrogen atom abstraction
substrate oxidative transformations.
Recently, we reported the synthesis, properties, and reactivity
II
R
2- 2+
of a series of Cu2(O2) complexes, [{Cu (MePY2) }2(O2 )]
R
(R ) H, MeO, and Me2N, 1 ; Scheme 3), that contain copper
(
HAT), Scheme 2), as opposed to a consecutiVely occurring
Scheme 3
34,35
PCET (an ET/PT reaction, Scheme 2).
Scheme 2
Intermolecular exogenous substrate oxidations by Cu2-O2
species are less well documented.1
8,30,36-38
In studies by Itoh,
II
2
2
it has been suggested that both Cu 2-(η :η -µ-peroxo) and
III
Cu 2-bis-µ-oxo complexes oxidize phenols (to coupled phenol
(
23) Oxidation in Organic Chemistry; Trahanovsky, W. S., Ed.; Academic
Press: New York, 1973; Vol. Part B, Chapter 1.
(24) Sprecher, C. A.; Zuberb u¨ hler, A. D. Angew. Chem., Int. Ed. Engl. 1977,
1
6, 189.
(
25) Gampp, H.; Zuberb u¨ hler, A. D. Met. Ions Biol. Syst. 1981, 12, 133-189.
26) Zuberb u¨ hler, A. D. In Copper Coordination Chemistry: Biochemical and
Inorganic PerspectiVes; Karlin, K. D., Zubieta, J., Eds.; Adenine: Guild-
erland, NY, 1983; pp 237-258.
(
(
(
(
(
(
(
27) Reglier, M.; Amadei, E.; Alilou, E. H.; Eydoux, F.; Pierrot, M.; Waegell,
B. Bioinorg. Chem. Copper 1993, 348-362.
28) Mahapatra, S.; Young, V. G., Jr.; Kaderli, S.; Zuberb u¨ hler, A. D.; Tolman,
W. B. Angew. Chem., Int. Ed. Engl. 1997, 36, 130-133.
29) Mahadevan, V.; Hou, Z.; Cole, A. P.; Root, D. E.; Lal, T. K.; Solomon, E.
I.; Stack, T. D. P. J. Am. Chem. Soc. 1997, 119, 11996-11997.
30) Taki, M.; Itoh, S.; Fukuzumi, S. J. Am. Chem. Soc. 2001, 123, 6203-
6
204.
31) Mahapatra, S.; Halfen, J. A.; Tolman, W. B. J. Am. Chem. Soc. 1996, 118,
1575-11586.
32) Cramer, C. J.; Kinsinger, C. R.; Pak, Y. THEOCHEM 2003, 632, 111-
20.
1
1
(
(
(
33) Cramer, C. J.; Pak, Y. Theor. Chem. Acc. 2001, 105, 477-480.
34) Cukier, R. I. J. Phys. Chem. B 2002, 106, 10089-10100.
35) Since the specifics of these reactions are unknown from a theoretical
standpoint (i.e., we do not know if the electron and proton are going to the
same (a hydrogen atom transfer (HAT)) or different (a concerted PCET)
acceptor orbitals), we prefer to use Cukier’s more general nomenclature
ligated by para-substituted bispyridyl(ethylamine) ligands,
(ref 34) for thermodynamically coupled proton/electron-transfer reactions.
R 42-44
Here, “ETPT” means that the ET and PT occur in one kinetically
irresolvable step, while “ET/PT” designates that the ET and PT processes
can be broken down into two kinetically resolvable steps, with the ET event
occurring prior to proton transfer (PT).
(MePY2) .
These complexes are capable of oxidizing a
variety of substrates in good yields, making them excellent
(
(
(
36) Fukuzumi, S.; Ohtsu, H.; Ohkubo, K.; Itoh, S.; Imahori, H. Coord. Chem.
(39) Osako, T.; Ohkubo, K.; Taki, M.; Tachi, Y.; Fukuzumi, S.; Itoh, S. J. Am.
Chem. Soc. 2003, 125, 11027-11033.
ReV. 2002, 226, 71-80.
37) Mahadevan, V.; Henson, M. J.; Solomon, E. I.; Stack, T. D. P. J. Am.
Chem. Soc. 2000, 122, 10249-10250.
(40) Itoh, S.; Kumei, H.; Nagatomo, S.; Kitagawa, T.; Fukuzumi, S. J. Am.
Chem. Soc. 2001, 123, 2165-2175.
38) Mahadevan, V.; DuBois, J. L.; Hedman, B.; Hodgson, K. O.; Stack, T. D.
P. J. Am. Chem. Soc. 1999, 121, 5583-5584.
(41) Taki, M.; Teramae, S.; Nagatomo, S.; Tachi, Y.; Kitagawa, T.; Itoh, S.;
Fukuzumi, S. J. Am. Chem. Soc. 2002, 124, 6367-6377.
5470 J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005

Substrate Oxidation by Copper−Dioxygen Adducts
A R T I C L E S
fate,52 were all prepared through literature methods. NH
(B(C F ) ) and
4 6 5 4
candidates for detailed mechanistic studies of Cu2-O2-mediated
exogenous substrate oxidations. An initial study was published
concerning PCET reactions between 1R and para-substituted
42
NBu (B(C F ) ) were prepared by the method of Geiger; NR Br
4
6
5
4
4
underwent a metathesis reaction with Li(B(C
6
F
5
)
4
)‚Et
2
O (Boulder
53
Scientific, Boulder, CO) in EtOH/H
2
O. Ferrocene, dimethylferrocene,
N,N-dimethylanilines (R′-DMAs), which undergo N-dealkyla-
II
II
R
2+
R
ethylferrocene, decamethylferrocene, [Fe (phen) (phen-NO )](ClO ) ,
2 2 4 2
tions yielding [{Cu (MePY2) }2(OH)2] (2 ), para-substituted
methylanilines, and formaldehyde (Scheme 3).44 On the basis
of linear-free energy and deuterium kinetic isotope effect (KIE)
studies, we determined that both ET/PT and ETPT pathways
are possible for these oxidations, with ET being rate limiting
N-methylaniline (MA), para-methoxybenzaldehyde (BA), 9,10-dihy-
droanthracene (DHA), and N,N-dimethylaniline (DMA) were all
obtained through commercial sources and purified prior to use according
54
to standard procedures. Solvents were purified using a commercially
available double alumina column solvent purification system (Innovative
Technologies, Inc.) and were degassed prior to use. All other reagents
were used as received. Electronic absorption spectra were recorded on
a Hewlett-Packard 8453 UV-vis spectrometer in a quartz optical Dewar
using Schlenk cuvettes with a 1 cm path length. The methanol bath
inside the Dewar was maintained at the desired temperature (within
H
in the ET/PT case. For 1 , we suggested that all R′-DMA
oxidations occurred through an ET/PT pathway, while there was
MeO
a switch-over in the mechanism for R′-DMA oxidation by 1
Me N
and 1
2
. Since the R′-DMA substrate is made more difficult
to oxidize by one electron, the mechanism changes from ET/
PT to ETPT.
(0.1 °C) by cooling with a Neslab ULT-95 low-temperature circulating
cooler. NMR spectra were recorded on a Bruker AMX 300 FTNMR
spectrometer. EPR spectra were recorded in quartz EPR tubes on a
Bruker EMX CWEPR spectrometer. The cavity was maintained at 65
K with the use of a He cryostat. Stopped-flow kinetic traces were
recorded on a Hi-Tech SF-40 variable-temperature stopped-flow unit
(1 × 0.2 cm quartz cell) equipped with a J&M-Tidas spectrometer.
In the present study, we disclose a much more detailed
investigation on the mechanism of PCET reactions facilitated
by 1 . Given that both KIEs and linear free-energy correlations
R
can be difficult to interpret, we decided to further explore the
R
nature of PCET reactions by 1 (i.e., ETPT versus ET/PT) using
two mechanistic probes. Mechanistic probes are substrates that
give unambiguous and different products depending on reaction
pathways. The probes chosen for this study are N-cyclopropyl-
N,N-Dimethylaniline and Tetrahydrofuran (DMA and THF)
II
R
Reaction Kinetic Analysis. Solutions of [{Cu (MePY2) }
2
(O
2
)]-
R
R
-5
(B(C F ) ) (1 ) in CH Cl ([1 ] ranged between 4.0 × 10 and 6.5
6
5
4
2
2
2
45
-5
I
R
N-methylaniline (CMA) and (p-methoxyphenyl)-2,2-dimeth-
× 10 M) were prepared by bubbling solutions of [Cu (MePY2) ]-
4
6
ylpropanol (MDP). Here, the differential stabilities of radical
B(C ) in CH Cl with dioxygen at low temperatures (-85 to -65
F
6 5
2
2
•
-
°C). UV-vis reaction monitoring was carried out using an apparatus
with a quartz cuvette at one end of a long (14 in.) glass tube and air-
free handling stopcocks and ground-glass joints at the other of which
(
H loss) versus radical cation (e loss) intermediates formed
following initial probe oxidation can allow us to differentiate
R
between reaction pathways. To delineate the scope of 1
1
0 in. is completely immersed in a cold methanol bath. After complete
R
reactivity, we also present the reaction patterns of 1 toward
R
Me N
2
formation of 1 , as noted by the Abs at 355 nm (or 360 nm for 1
the solutions were purged of excess dioxygen with argon. A solution
of DMA (500 µL) in CH Cl was then slowly layered over the top of
),
substrates possessing both similar (i.e., 2,5-dimethyltetrahydro-
furan (Me2THF), toluene, diethyl ether (Et2O)) and significantly
weaker bond dissociation enthalpies (BDEs) (i.e., 9,10-dihy-
droanthracene (DHA) and 2,2,6,6-tetramethylpiperidin-1-ol
2
2
the reaction mixture by adding it down the side of the tube so that no
solution mixing was obtained. After a brief period to allow for
temperature equilibration, the reaction was initiated by mixing the
solutions together with a gentle stream of argon, and then the reaction
(
TEMPO-H)) compared to those of DMAs. Electrochemical and
R
thermodynamic information for the 1 series is also presented
to further aid us in gaining mechanistic insight for these
oxidations. Furthermore, evidence for a pre-equilibrium involv-
ing substrate coordination to the metal center prior to oxidation
is provided. Overall, the present study provides not only new
insights but also a more detailed mechanistic picture concerning
the mechanism of substrate oxidations by copper-dioxygen
adducts.
was monitored by UV-vis spectroscopy. For the higher-temperature
Me N
reactions (-80 to -65 °C) involving DMA and 1
2
, stopped-flow
methods were used to follow the course of the reaction. Here, an O
saturated solution of CH Cl was injected into a CH Cl solution of
) and DMA, and the reaction was monitored.
2
-
2
2
2
2
I
Me N
2
[
Cu (MePY2)
6 5
]B(C F
First-order rate constants for all DMA oxidation reactions were then
extracted either manually using the change in Abs at 355 nm (360 nm
Me N
for 1
2
) or automatically using either SpecFit/32 (Spectrum Software
Associates, Marborough, MA)55 or an in-house series of procedures
written for Igor Pro 4.0 (Wavemetrics, Lake Oswego, OR). All three
methods gave identical first-order rate constants (kobs). Data were
collected at several different DMA concentrations (0.8-30 mM) over
the temperature range of -65 to -85 °C. Plots of kobs versus [DMA]
were then fit to a Lineweaver-Burk-type plot (eq 1):
Experimental Section
General Methods. All reactions were performed under an atmo-
sphere of argon using standard Schlenk techniques. Cu complexes,
I
I
H
I
MeO
I
[
(
Cu (MePY2) ]B(C
6
F
5
)
4
,
[Cu (MePY2) ]B(C
, were prepared according to previously published
Mechanistic probes, N-methyl-N-cyclopropylaniline
6
F
5
)
4
,
and [Cu -
-1
-1
Me N
MePY2)
procedures.
2
6 5 4
]B(C F )
4
2,44
47
48
(
CMA), (p-methoxyphenyl)-2,2-dimethylpropanol (MDP), and TEM-
PO-H, as well as cyclopropylaniline (CA), (p-methoxyphenyl)-2,2-
kobs- ) kox + (Keqkox)-1[DMA]-1
1
-1
(1)
4
9
50
5
1
+
dimethylpropanone (MDK), and N-methylquinolinium (MQ ) sul-
(
47) Shaffer, C. L.; Morton, M. D.; Hanzlik, R. P. J. Am. Chem. Soc. 2001,
(
42) (a) Zhang, C. X.; Liang, H.-C.; Kim, E.-i.; Shearer, J.; Helton, M. E.; Kim,
E.; Kaderli, S.; Incarvito, C. D.; Zuberb u¨ hler, A. D.; Rheingold, A. L.;
Karlin, K. D. J. Am Chem. Soc. 2003, 125, 634-635. (b) Liang, H.-C.;
Karlin, K. D.; Dyson, R.; Kaderli, S.; Jung, B.; Zuberb u¨ hler, A. D. Inorg.
Chem. 2000, 39, 5884-5894.
123, 349-350.
(48) Baciocchi, E.; Bietti, M.; Putignani, L.; Steenken, S. J. Am. Chem. Soc.
1996, 118, 5952-5960.
(49) Mader, E. A.; Larsen, A. S.; Mayer, J. M. J. Am. Chem. Soc. 2004, 126,
8066-8067.
(50) Cui, W.; Loeppky, R. N. Tetrahedron 2001, 57, 2953-2956.
(51) Smyth, T. P.; Corby, B. W. J. Org. Chem. 1998, 63, 8946-8951.
(52) Hammond, P. R. U.S. Patent Appl. 566,924, November 9, 1984.
(53) LeSuer, R. J.; Geiger, W. E. Angew. Chem., Int. Ed. 2000, 39, 248-250.
(54) Armarego, C. Purification of Laboratory Chemicals, 5th ed.; Butterworth
Heinemann: New York, 2003.
(
(
(
(
43) Henson, M. J.; Vance, M. A.; Zhang, C. X.; Liang, H.-C.; Karlin, K. D.;
Solomon, E. I. J. Am. Chem. Soc. 2003, 125, 5186-5192.
44) Shearer, J.; Zhang, C. X.; Hatcher, L. Q.; Karlin, K. D. J. Am. Chem. Soc.
2
003, 125, 12670-12671.
45) Shaffer, C. L.; Harriman, S.; Koen, Y. M.; Hanzlik, R. P. J. Am. Chem.
Soc. 2002, 124, 8268-8274.
46) Baciocchi, E.; Bietti, M.; Lanzalunga, O. Acc. Chem. Res. 2000, 33, 243-
(55) Gampp, H.; Maeder, M.; Meyer, C. J.; Zuberb u¨ hler, A. D. Talanta 1985,
32, 95-110.
2
51.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5471

A R T I C L E S
Shearer et al.
Table 1. Product Yields in the Oxidation of Mechanistic Probes
CMA and MDP (see Schemes 7, 8, and 10) by 1 in CH2Cl2 at
sample; therefore, we had to estimate yields of this product by
R
Me N
integration relative to N-methylaniline. The reaction between 1
2
and
-
80 °C as Assessed by GC/MS Analysis
+
CMA produced N-methylquinolinium (MQ ), which was isolated by
column chromatography of the reaction mixture (alumina). The resulting
solution was evaporated to dryness and weighed on an analytical
1H (%)
1MeO (%)
1Me2N (%)
Oxidation of CMA
PMAa
MA
20(5)
48(2)
6(1)
30(5)
55(1)
2(1)
<0.1
88
<0.1c
2(1)
2(1)
balance. The H NMR spectrum of the product was then compared to
1
1
+
the H NMR spectrum of an authentic sample of MQ and found to be
nearly identical (see Discussion and Supporting Information, Figure
S15). All results represent the average of three separate experiments.
Exogenous Substrate Oxidation. The oxidation of exogenous
CA
MQ
+
c
c
d
<0.1
74
50(20)
total yield of CMAb
50-100e
Oxidation of MDP
substrates was monitored in two ways. First, we monitored the decay
MDK
10(3)
31(2)
41
47(3)
< 0.1
47
63(8)
<0.1
63
R
c
c
of 1 by UV-vis (-80 °C) in 0.5 M CH
2
Cl
THF) by noting the loss of the
). The first-order oxidation rate was
2
solutions of substrate
MBA
total yield of MDPb
2 2
(TEMPO-H, toluene, DHA, Et O, Me
band at 355 nm (360 nm for 1
then extracted and compared to the decay of 1 in the absence of
substrate (k < 10
Me N
2
a
R
Product yields of PMA were calculated by comparison to the peak area
b
of MA (see Schemes 7 and 10, and the Experimental Section). Total yields
are based on the amount of product quantified by GC analysis. No product
was detected by GC. Based upon gravimetric analysis of the isolated
product. See text for details.
-6
-1). If the rate of oxidation in the presence of
s
c
substrate was comparable to that in the absence of substrate, it was
d
R
e
concluded that 1 did not react with that substrate. We also examined
the oxidation of substrate on a “synthetic” scale using a procedure
identical to that used for mechanistic probe oxidations. Products were
then analyzed by GC/MS. For the case of diethyl ether oxidation, we
noted oxidation was taking place; however, GC/MS revealed no
where kox is the oxidation rate constant, and Keq is the equilibrium
R 18
constant for DMA binding to 1 . Rate constants for THF oxidation
Me N
by 1
2
were determined in a similar manner using both stopped-
flow and benchtop UV-vis methods over the temperature range of
90 to -70 °C. Here, the THF concentration in CH Cl was varied
identifiable products. Therefore, we used Fehling’s test (CuSO
Rochell’s salt) to confirm the presence of an aldehyde.5
4
/
7
-
2
2
between 0.1 and 2.0 M. First-order rate constants (kobs) were extracted
as mentioned above and plotted versus THF concentration. The slope
Electrochemical Measurements and Spectroelectrochemical Ti-
trations. Cyclic voltammograms (CVs) were obtained at -78 °C in a
dry ice/acetone bath and recorded with a BAS-100B potentiostat
interfaced to a personal computer. To aid in data quality of the CV
waves and subsequent data analysis, CVs were recorded in 5 mV (as
opposed to 1 mV) increments. All measurements were recorded in a
of this fit was then taken as the second-order rate constant, kox
.
Activation and thermodynamic parameters were then obtained using
5
6
van’t Hoff or Eyring plots.
Mechanistic Probe (CMA and MDP, see Schemes 7 and 8)
Oxidation Reactions and Product Analysis (Table 1). In a typical
Faraday cage, using Bu
4 6 5 4
N(B(C F ) ) (0.1 M) as the supporting
I
R
reaction, 50 mg of [Cu (MePY2) ](B(C
6
F
5
)
4
) was dissolved in 2 mL
electrolyte in CH Cl . In a typical experiment, a 0.01 M solution of
2
2
I
R
of CH Cl and cooled in a dry ice/acetone bath to -78 °C. Dioxygen
2
2
the Cu complex was exposed to dioxygen (resulting in 0.005 M 1 )
R
was then slowly bubbled through the reaction mixture for ∼30 s and
and then purged with argon after full formation of 1 was noted. For
H
MeO
then through the reaction headspace for an additional 120 s. After full
1 and 1 , a standard three-electrode cell was used containing a glassy
II
R
R
formation of [{Cu (MePY2) }
0 min), the reaction mixtures were then purged with argon to remove
all residual dioxygen. Ten equivalents of substrate CMA or MDP and
equiv of decane were dissolved in 500 µL of argon-sparged CH Cl
2
(O
2
)](B(C
6
F
5
)
4
)
2
(1 ) was obtained (2-
carbon working electrode, a platinum wire counter electrode, and an
t
3
SCE reference electrode in contact with the cell through a CH
Bu N(B(C
2
Cl
2
-
4
6 5
F
4
) ) salt bridge (2.0 M). The glassy carbon electrode was
5
8
1
2
2
activated according to the methods of Thorp and co-workers. Scan
rates utilized for these two complexes ranged between 0.5 and 2.5 V
and then injected into the reaction mixture. The reaction mixture was
then placed in a low-temperature freezer (-80 °C) for 24 h to 7 days.
By this time, the reaction went to completion as noted by full formation
-1
Me N
2
s
. For 1
, a 10 µm diameter platinum ultramicroelectrode was used
as the working electrode. The concentrations of electrolyte and copper
complex were kept the same as those listed above.
Reduction potentials were also obtained by spectroeletrochemical
II
R
R
of [{Cu (MePY2) }
2
(OH)
](B(C F ) )
2 6 5 4 2
(2 ). The reaction mixture was
R
then passed through a short plug of alumina to remove all 2 . All
reaction mixtures were then analyzed by GC/MS on a Shimadzu
GCMS-QP5050A equipped with a DB-5M5 packed J&W Scientific
H
“titrations” using ferrocene derivatives (ferrocene for 1 , ethylferrocene
MeO
Me N
2
for 1 , and dimethylferrocene for 1
). In a typical experiment, a
-
5
R
(30 m/0.25 mm i.d.) column. Quantification of yields was then
solution of ∼5 × 10 M 1 in CH
2
Cl
2
was cooled to -78 °C. A
determined by standard calibration curves of decane versus the various
products. Due to the unstable nature of ortho-(propan-3-al)-N-methy-
laniline, a calibration curve could not be constructed using authentic
solution of the ferrocene derivative was then added to the copper
solution and mixed gently with argon for ∼5 s. Reduction of the Cu
complex was noted by the decrease in the charge-transfer band of 1
2
O
2
R
Table 2. Comparison of Geometric Parameters for Structures Discussed Here
distances
2MeO
2Me2N
2H
1H
a
and angles
Cu-O
Cu-O
1.945(2)
1.946(3)
2.277(2) (NR3)
2.011(3) (py)
2.000(2)
0.279
1.951(3)
1.944(3)
2.302(2) (NR3)
2.011(3) (py)
2.004(4)
0.250
1.958(3)
1.948(3)
2.234(4) (py)
2.075(4) (NR3)
2.005(4)
0.215
1.905(5)
1.922(6)
2.177(6) (py)
2.027(6) (NR3)
1.989(6)
Cu-Namine/PY (axial)
Cu-NPY/amine(eq)
Cu-NPY (eq)
b
Cu/N2O2 plane
Cu‚‚‚Cu
3.057(1)
93.4(1)
96.0(1)
95.1(1)
0.07
3.037(1)
90.6(1)
96.2(1)
94.1(1)
0.09
3.012(1)
96.7(1)
96.7(1)
94.3(2)
0.04
3.445(2)
97.9(3)
103.1(3)
98.2(3)
0.06
Neq/PY-Cu-Nax
Neq/PY(amine)-Cu-Nax
Neq/PY(amine)-Cu-Neq
c
τ
a
Intermediate structure as a mixture of peroxo and bis-µ-oxo isomers.67 Distance (Å) of Cu(II) out of the basal plane. c With (a - b)/60 (a ) largest
b
interatomic angle including Cu; b ) second largest such angle); square pyramid, τ ) 0.0; trigonal bipyramid, τ ) 1.0.79
472 J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005
5
9

Substrate Oxidation by Copper−Dioxygen Adducts
A R T I C L E S
centered around 355 nm (ꢀ ) 16 000-23 000 M-1 cm-1). The end
point was determined by noting no further change in the electronic
absorption spectrum at 355 nm upon further addition of ferrocene or
Scheme 4
derivative.
R
EPR Experiments. 1 in 0.25 mL of a 1:1 mixture of CH
2
Cl
2
/
-
4
toluene (2 × 10 M) was prepared in an EPR tube at -78 °C in a dry
ice/acetone bath. A cold solution of 0.05 mL of CH Cl solution of
2
2
substrate (0.1 equiv of decamethylferrocene of TEMPO-H) was then
injected into the EPR tube, and then the tube was quickly immersed in
liquid nitrogen. EPR spectra were then recorded at 65 K as frozen
•
glasses. In the case of TEMPO-H, we determined the yield of TEMPO
formed in the reaction by integrating the EPR spectra and comparing
-
4
•
this to a standard (2 × 10 M) solution of TEMPO .
MeO
Me N
. Both 2MeO and
X-ray Crystallographic Studies of 2
and 2
2
Me N
MeO
2
2
compounds were prepared by the oxidative decomposition of 1
Me N
and 1
2
2 2
in THF and then recrystallized from CH Cl /pentane. Suitable
greenish-blue crystals were mounted with epoxy cement to the tip of
a glass fiber. Intensity data were collected at 150(2) K with a Bruker
SMART APEX CCD diffractometer with graphite monochromated Mo
KR radiation (λ ) 0.71073 Å). An absorption correction was done using
the SADABS program (Sheldrick, G. M. SADABS (2.01), Bruker/
Siemens Area Detector Absorption Correction program, Bruker AXS,
Madison, WI, 1998). The structure was solved by direct methods,
completed by difference Fourier syntheses, and refined by full-matrix
“rebound”-type reaction analogous to that of cytochrome P450
chemistry.
4
2,44
In the present study, we explore other potential substrates
2
least-squares refinement procedures based on F . All non-hydrogen
R
that are oxidized by 1 (Scheme 4) and possess thermodynamic
atoms were refined with anisotropic thermal parameters and assigned
to idealized geometric positions using a rigid group model. All software
and sources of the scattering factors are contained in the SHELXTL
properties similar to those previously examined. Examining
thermodynamically similar (i.e., one-electron oxidation poten-
tials and H-X BDEs, X ) C or O) but structurally different
substrates will also allow us to gain insight into potential
(5.1) program library (G. Sheldrick, Siemens XRD, Madison, WI).
Relevant crystallographic information and the main details of the
diffraction experiment and refinement of the crystal structure are
provided in Table 2, while the full X-ray structural reports (in CIF
format) may be found in the Supporting Information.
substrate structural and/or electronic requirements necessary for
R
1
to facilitate substrate oxidation. Toluene oxidation (BDE ∼
-1 60 R
89 kcal mol ) was not facilitated by 1 despite possessing a
C-H bond strength similar to that known for THF (BDE ∼ 92
Results
-
1 61
kcal mol ). It was also observed that substrates which are
structurally similar to THF either showed no reactivity, such
as 2,5-dimethyltetrahydrofuran (Me2THF), or showed dimin-
This study is aimed at better understanding the fundamental
mechanism of PCET oxidation reactions mediated by peroxo
II
R
2- 2+
R
62
complexes, [{Cu (MePY2) }2(O2 )] (1 ). We first present
ished (i.e., slower) reactivity, such as diethyl ether. This is
our survey of the scope of PCET reactions that can or cannot
despite the fact that these ethers are thermodynamically similar
R
-1
be facilitated by 1 . Next, we briefly discuss relevant thermo-
to THF; all have R C-H BDEs of ∼90 kcal mol .
R
R
dynamic properties of 1 pertaining to such reactions, that is,
Next, we investigated reactions between 1 and substrates
one-electron reduction potentials, and how oxo/peroxo proto-
nation influences ET reactions. We then present kinetic and
reactivity studies aimed at better understanding the mechanism
with significantly weaker X-H bonds (where X ) C or O).
TEMPO-H (2,2,6,6-tetramethylpiperidin-1-ol), which has an
-
1 63
R
O-H BDE of ∼70 kcal mol , reacts with 1 exceptionally
R
R
of 1 substrate oxidation reactions.
fast and quantitatively at -80 °C, affording 2 and 2 equiv of
TEMPO as the only organic product; see further discussion
below. Both EPR spin integration (vide infra) and UV-vis
II
R
•
Oxidation Reaction Facilitated by [{Cu (MePY2) }2-
2
+
R
II
R
2+
64
(
(
O2)] (1 ). Cu-dioxygen adducts [{Cu (MePY2) }2(O2)]
R
1 ) are capable of performing several oxidation reactions,
titrations demonstrate that 2 equiv of TEMPO-H gives 2 equiv
5
9
•
R
including O atom transfers (PPh3 to OdPPh3) and formal
of TEMPO and 1 equiv of 2 .
•
R
hydrogen atom (H ) abstraction reactions (tetrahydrofuran (THF)
Previously, we had stated that 1 would react with 9,10-
dihydroanthracene (DHA, which has a C-H BDE of ∼78 kcal
to 2-hydroxytetrahydrofuran (THF-OH), alcohols to aldehydes
and ketones, and dimethylaniline (DMA) to methylaniline (MA)
and formaldehyde).4 In the latter case, the resulting copper-
-1
42,65
mol ) at low temperatures.
Reinvestigation of DHA oxida-
2,44
tion now reveals that, in fact, it will not undergo a redox reaction
II
R
containing products are bis-µ-hydroxo complexes, [{Cu -
with 1 under the conditions investigated (-80 °C in CH Cl ).
2
2
R
2+
R
(
MePY2) }2(OH)2] (2 ), and the organic substrate is either
•
(60) Berkowitz, J.; Ellison, G. B.; Gutman, D. J. Phys. Chem. 1994, 98, 2744-
765.
oxidized by (formally) one or two H or hydroxylated via a
2
(
61) Laarhoven, L. J. J.; Mulder, P. J. Phys. Chem. B 1997, 101, 73-77.
(
56) Wilkins, R. G. Kinetics and Mechanism of Reactions of Transition Metal
Complexes, 2nd ed.; VCH: New York, 1991.
(62) Due to the low yields and low molecular weights of products in diethyl
ether oxidations, we were only able to confirm that an aldehyde was
produced. This suggests an oxidative O-dealkylation through a rebound-
like process.
(
57) Furniss, B. S.; Hannaford, A. J.; Smith, P. W. G.; Tatchell, A. R. Vogel’s
Textbook of Practical Organic Chemistry, 5th ed.; Pearson Education
Ltd.: Essex, England, 1989.
(63) Bordwell, F. G.; Liu, W.-Z. J. Am. Chem. Soc. 1996, 118, 10819-10823.
(64) The reaction is too fast for us to monitor by stopped-flow methods (i.e.,
TEMPO-H oxidation is faster than dioxygen adduct formation), that is,
(58) Thorp, H. H.; Sarneski, J. E.; Brudvig, G. W.; Crabtree, R. H. J. Am. Chem.
Soc. 1989, 111, 9249-9250.
I
+
R
(
59) Obias, H. V.; Lin, Y.; Murthy, N. N.; Pidcock, E.; Solomon, E. I.; Ralle,
M.; Blackburn, N. J.; Neuhold, Y.-M.; Zuberb u¨ hler, A. D.; Karlin, K. D.
J. Am. Chem. Soc. 1998, 120, 12960-12961.
the overall reaction observed is [Cu (MePY2)] f 2 ).
(65) Bordwell, F. G.; Cheng, J.; Ji, G. Z.; Satish, A. V.; Zhang, X. J. Am. Chem.
Soc. 1991, 113, 9790-9795.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5473

A R T I C L E S
Scheme 5
Shearer et al.
In the earlier study, organic product reaction yields were
analyzed by GC after 24 h from the start of the reaction. Doing
so involves a workup where the reaction mixture is warmed
and then passed through a plug of alumina prior to GC/MS
analysis to remove all copper-containing products. It is during
this warmup that DHA oxidation appears to occur. Dichlo-
Figure 1. Cyclic voltammograms of 1H (black) and 1MeO (red) recorded
at -78 °C in CH2Cl2 with Bu4N(B(C6F5)4) as the supporting electrolyte.
R
romethane solutions of 1 kept at -80 °C for 7 days show no
reactivity toward DHA. Instead, UV-vis monitoring shows that
R
R
1
“autodecomposes” to 2 through an unidentified oxidation
H
of ferrocene (E1/2 ) 560 mV in CH2Cl2) to 1 resulted in the
pathway at a rate identical to that when no substrate is present
complete quenching of its LMCT band. However, excess
-6 -1
(
k < 10 s ). Furthermore, no anthracene product is observed,
MeO
Me N
2
ferrocene was incapable of readily reducing either 1
On the other hand, 1
or 1
.
as noted by both GC and UV-vis spectroscopies.
MeO
is readily reduced by excess ethylfer-
R
Thus, it appears that for 1 substrate oxidations, thermodynam-
Me2N
rocene (E1/2 ) 380 mV versus SCE in CH2Cl2), while 1
required the substrate, dimethylferrocene (E1/2 ) 230 mV versus
SCE in CH2Cl2), for its reduction. On the basis of these
experiments/evidence, it appears that the complex with the least
ic considerations (i.e., X-H BDEs) cannot adequately predict
which substrates will react. Scheme 4 shows a summary of the
R
reactivity of 1 that we investigated for the current and previous
studies, excluding the reactivity probe studies described below.
H
electron-donating 4-pyridyl substituent (1 ) is the most oxidiz-
R
Thermodynamic Properties of 1 . We next sought to
ing, while the complex with the most electron-donating sub-
II
R
2+
determine the oxidizing abilities of [{Cu (MePY2) }2(O2)]
stituent (1Me2N
) is the least powerful oxidant; not surprisingly,
to determine if proton-uncoupled reactions (i.e., single-electron
oxidation (SET)) by 1R are thermodynamically uphill by
obtaining approximate redox potentials. SET oxidations effected
the more electron-donating ligand can better support and
stabilize the higher oxidation state, making it less reactive.
R
Dichloromethane solutions of 1 are stable at -80 °C in the
R
by 1 might, in principle, produce a one-electron-reduced mixed-
II
presence of the weak reductant, [Fe (phen)2(NO2-phen)](ClO4)2
II
III
+
valence dicopper [Cu Cu (O)2] species (see Scheme 5).
(E1/2 ) 1.3 V versus SCE in CH2Cl2). However, the addition
II
R
2+
R
Cyclic voltammetry studies of [{Cu (MePY2) (O2)] (1 )
were carried out in CH2Cl2 at -78 °C. Complexes 1 are
of a slight excess of NH (B(C F ) ) (pK ) 16.5 in MeCN at
4
6
5 4
a
R
66
25 °C) as an acid/proton source leads to the rapid decay of
Me N
reasonably stable toward the supporting electrolyte, 0.1 M Bu4N-
the band at 355 nm (360 nm for 1
2
), demonstrating the
(B(C6F5)4). The latter has been shown to improve the quality
reduction of the Cu -dioxygen adducts. The resulting products
2
of electrochemical data recorded in solvent with low dielectric
constants at low temperatures.53 Reasonably “clean” reductions
that can best be described as quasi-reversible are observed for
are the corresponding two-electron/two-proton-reduced (relative
R
II
R
2+
to 1 ) bis-µ-hydroxo complexes, [{Cu (MePY2) } (OH) ]
2
2
R
III
3+
(2 ) and [Fe (phen) (NO -phen) ] (as assessed by UV-vis
2
2
2
H
MeO
both 1 and 1
(Figure 1). We can, therefore, estimate
spectroscopy). When acetic acid is instead used as a proton
H
66
reduction potentials (Ered), ∼550 mV for 1 and ∼400 mV for
source (pK ) 22.3 in MeCN at 25 °C) under otherwise
a
MeO
1
, based on the apparent midpoints of the reduction waves.
identical conditions, no immediate observable reaction takes
place. The results, discussed further below, suggest coupling
protonation with a weak reductant can, in fact, lead to PCET-
Me N
In contrast, 1
2
does not yield readily interpretable electro-
chemical data by CV, even with the use of a platinum
-1
R
ultramicroelectrode at a scan velocity of 51 V s . This is
type reduction of the peroxo complexes 1 .
Me N
presumably due to 1
2
s extreme instability toward reducing
Oxidation Reactions Probed by EPR. We next sought to
determine if a mixed-valence dicopper µ-oxo µ-hydroxo species,
conditions (see Supporting Information for a representative CV
Me N
2+
II
R
of 1
2
).
2
[Cu (O)(OH)] , could be readily produced from [{Cu (MePY2) -
R
2+
R
•
To put an upper limit on the SET oxidizing ability of 1 , we
(O )] (1 ) by using less than 1 equiv of an easily oxidized H
or e donor (see Scheme 5). Such species are likely to be
2
R
-
decided to expose solutions of 1 to ferrocene derivatives that
are progressively harder to oxidize. Here, we can monitor the
reduction of 1 , and hence the oxidation of the ferrocene
derivative, by observing the quenching of the peroxo f CuII
“initial” intermediates in hydrogen atom abstraction oxidation
R
31,44,67,68
reactions;
however, their instability may preclude
•
-
R
observation. Reactions between H or e donors and 1 were
performed in an EPR tube at -78 °C, and then the resulting
ligand-to-metal charge transfer (LMCT) at ∼355 nm (360 nm
Me N
for 1
2
) by UV-vis spectroscopy upon the addition of the
(
66) Izutsu, K. Acid-Base Dissociation Constants in Dipolar Aprotic SolVents;
Blackwell Scientific Publications: Oxford, U.K., 1990; Chemical Data
Series, No. 35.
appropriate ferrocene derivative (see Supporting Information
Figure S2). It was determined that addition of a slight excess
5474 J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005

Substrate Oxidation by Copper−Dioxygen Adducts
A R T I C L E S
solutions were quickly frozen in a liquid nitrogen bath. These
frozen glasses were subjected to analysis by EPR spectroscopy.
To test for single versus double ET to 1 , /10 of an equivalent
mined by both stopped-flow and benchtop UV-vis kinetic
techniques, each of which gave data consistent with one another.
No UV-vis detectable intermediates were observed in the
R
1
Me N
Me N
2
of decamethylferrocene was added per copper, which reacted
oxidation of THF by 1
2
; only the clean conversion of 1
is noted. Under pseudo-first-order conditions ([THF]
) 0.1-2.0 M), plots of kobs versus [THF] demonstrated second-
R
Me N
2
quantitatively with 1 at -78 °C (UV-vis monitoring). This
to 2
resulted in the formation of an EPR silent solution (at 65 K),
consistent with the conversion of all decamethylferrocene into
decamethylferrocenium ion, which has a ground spin state of
Me N
order behavior (first order in THF, first order in 1
2
), with
at -90 °C
at -70 °C. The second-order behavior
-
3
-1 -1
rate constants ranging from 8.9(1) × 10
M
s
3
-2
-1 -1
/
2 and would not be observable at this high temperature. Also,
to 2.9(1) × 10
M
s
II
III
this suggests that a Cu Cu mixed-valence complex was not
stable under these conditions, and that the copper complex
generated was fully reduced. If a mixed-valence compound was
formed, it should have an easy-to-observe distinctive EPR
spectrum, resulting from electron delocalization over the two
might seem to suggest a straightforward “outer-sphere”-type
oxidation mechanism in the conversion of THF to THF-OH
Me N 71
by 1
2
. However, an Eyring analysis demonstrates that THF
Me N
oxidation by 1
2
proceeds with a very low activation enthalpy
-
1
of 3.7(1) kcal mol , which strongly implicates the presence of
a rapid but left-lying pre-equilibrium which occurs prior to the
THF oxidation step, possibly that involving THF coordination.
3
1
copper (I ) /2) centers. Similarly, /10 of an equivalent of
TEMPO-H results in the complete formation of TEMPO (as
•
72
assessed by EPR spin integration), but no signal that could be
(Such a pre-equilibrium involving substrate binding is clearly
evident in the oxidation of DMA by 1 , vide infra.) Thus, the
THF oxidation mechanism is perhaps better described as “inner-
sphere oxidation” because of the likely substrate binding step.
II
III
R
attributed to a Cu /Cu dimer is seen. This also suggests that
II
R
2+
R
molecules of [{Cu (MePY2) (O2)] (1 ) are reduced by two
hydrogen atoms, independent of the amount of TEMPO-H
reductant added (i.e., less than stoichiometric amounts).
Reactions of 1R with N,N′-Dimethylaniline (DMA). As
II
III
44
To summarize, we could not generate a Cu /Cu binuclear
complex by the addition of a substoichiometric amount of a
reductant or a H-atom donor. Furthermore, these data seem to
suggest that 2 equiv of FeCp2R and TEMPO-H is consumed
previously described, -85 to -65 °C dichloromethane solu-
R
tions of 1 rapidly oxidize DMA to formaldehyde and methyl-
amine (MA) within several minutes (see Scheme 4). Here, our
further studies show that observed first-order rate constants (kobs)
plotted against substrate concentration ([DMA] ranging between
0.8 and 30 mM) exhibit Michaelis-Menten-type saturation
kinetics at higher DMA concentrations (Figure 2). Fitting these
data to eq 1 (vide supra) allowed us to extract both an
II
R
2+
per Cu-peroxo binuclear complex, [{Cu (MePY2) (O2)]
(
R
1 ), possibly through a disproportionation reaction (Scheme
5
; see Discussion).
Me N
Oxidation of THF by 1
2
. It was previously noted that
II
R
2+
R
eq
equilibrium constant (K ) for DMA binding and subsequent
THF solutions of [{Cu (MePY2) (O2)] (1 ) are unstable, with
R
II
R
2+
R
first-order oxidation rate constants (k ) for DMA oxidation
1
readily converted into [{Cu (MePY2) }2(OH)2] (2 ) over
ox
R
6
9,70
(Scheme 6). The 1 series exhibit fairly weak binding of DMA,
the course of several minutes to hours (depending on R).
A
as may be expected, with Keq varying between approximately
further study using dichloromethane solutions of THF demon-
-
1
R
850 and 10 M over the temperature range of -85 to -65 °C
see Supporting Information). A van’t Hoff analysis provides
estimates of the thermodynamic parameters; DMA binding is
strated that this substrate is, in fact, being oxidized by 1 to
(
THF-OH; the oxygen atom incorporated into the THF-OH
R 42
product is derived from the peroxo moiety of 1 . Here, we
-1
Me N
enthalpically favored (∆H° > 10 kcal mol ), while it is strongly
disfavored entropically, as would be expected (∆S° < -40 cal
present new data and insights for THF oxidation by 1
2
.
H
Difficulties were seen to arise for THF oxidation by 1 and
-
1
-1
73
MeO
mol K ; see Supporting Information). DMA binding to
1
, as kinetics of THF oxidation are complicated by the
MeO
-75 C
-1
1
(Keq
) 140(30) M ) appears to be stronger than that
presence of the MeCN ligand found in the starting material
H
-75 C
-1
I
R
+
42
observed for 1 (K
) 27(11) M ), while DMA binding
) 105(45) M ) is intermediate (Supporting
complexes, [Cu (MePY2) (MeCN)] (R ) H or MeO). As
eq
Me N
-75 C
-1
to 1
2
(Keq
seen previously, small variable amounts of MeCN complicate
O2 complex formation,4 and here, differing quantities of MeCN
2b
Information).
H
MeO
dramatically slow THF oxidation by 1 and 1 . However,
Oxidation rates at low temperatures (e-80 °C) appear to
the variations are not predictable, quite possibly due to variable
follow trends related to the electron-donating abilities of the
4-pyridyl ligand substituent. Concerning these DMA reactions,
I
R
+
I
R +
contamination of [Cu (MePY2) (MeCN)] by [Cu (MePY2) ] ,
I
Me2N
the Cu complex with a “lost” acetonitrile ligand. The MeCN
the more electron-rich metal center in 1
affords reaction
I
Me N + 42
-1 -1
ligand is not ever present in [Cu (MePY2)
2
] , which is a
rate constants (kox ) 3.4(1) × 10
s
at -85 °C) nearly an
order of magnitude larger than that for 1
MeO
pure three-coordinate complex. It should be noted that the
(kox ) 1.1(1) ×
I
-2 -1
additional MeCN derived from the Cu starting material does
10
s
at -85 °C), and nearly 2 orders of magnitude faster
H
MeO
H
-3 -1
not appear to influence the kinetics of 1 or 1
oxidation of
than that of 1 (kox ) 5.5(1) × 10
s
at -85 °C). These
dimethylaniline (DMA) (vide infra).
findings may seem to be unexpected, with the strongest oxidant,
Me N
Dichloromethane solutions of 1
2
readily and rapidly
(
71) An initial rate-limiting oxidation event combined with the fact that the
transferred oxygen atom is derived from the peroxo moiety suggests that
THF oxidation to 2-hydroxytetrahydrofuran occurs through a rebound-type
oxidize THF over the temperature range of -90 to -70 °C (see
Supporting Information). Oxidation rate constants were deter-
4
2
mechanism similar to that of cytochrome P450 chemistry.
(
72) (a) Karlin, K. D.; Kaderli, S.; Zuberb u¨ hler, A. D. Acc. Chem. Res. 1997,
30, 139-147. (b) Zhang, C. X.; Kaderli, S.; Costas, M.; Kim, E.-i.; Neuhold,
Y.-M.; Karlin, K. D.; Zuberb u¨ hler, A. D. Inorg. Chem. 2003, 42, 1807-
1824.
(
67) Pidcock, E.; DeBeer, S.; Obias, H. V.; Hedman, B.; Hodgson, K. O.; Karlin,
K. D.; Solomon, E. I. J. Am. Chem. Soc. 1999, 121, 1870-1878.
68) Chan, S. I.; Chen, K. H. C.; Yu, S. S. F.; Chen, C. L.; Kuo, S. S. J.
Biochemistry 2004, 43, 4421-4430.
(
(73) Weak binding at higher temperatures adds a great deal of uncertainty in
reported Keq values, which leads to uncertainty in the calculated ∆S° term.
See further details in the Discussion.
(
69) Zhang, C. X. Johns Hopkins University, 2001.
(
70) Obias, H. V. Johns Hopkins University, 1998.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5475

A R T I C L E S
Shearer et al.
Scheme 7
whether these complexes support concerted or consecutive
3
4
proton-coupled reactions (ETPT versus ET/PT). If CMA is
R
oxidized by 1 through an ETPT pathway, then N-methyl
dealkylation should be the prevalent oxidation pathway (produc-
ing N-cyclopropylaniline (CA)), with N-cyclopropyl dealkylation
being the minor oxidation pathway (producing MA).75 However,
under ET/PT conditions, CA is a minor (<5%) product;
furthermore, this pathway will also result in a cyclopropyl ring
Figure 2. (A) DMA oxidation reaction monitored by the decay of the λmax
355 nm absorption due to peroxo complex 1H in CH2Cl2 at -80 °C
)
{
[DMA] ) 6.0 mM; time ) 3000 s; time interval ) 30 s}. (B) ObserveHd
first-order rate constants (kobs) as a function of DMA concentration for 1
76
MeO
(red) and 1
(blue) oxidation of DMA at -80 °C in CH2Cl2. The
experimental data are shown as solid circles, and the best fit is displayed
as the solid line. The inset plots kobs for DMA oxidation by 1
Me N
2
(green)
opening, which can then produce a wide variety of observable
versus DMA concentration at -80 °C in CH2Cl2. Trend lines are based on
45,47,76
organic products.
This is summarized in Scheme 7.
fits to the data according to kobs ) kox[DMA]/(Keq + [DMA]).74
Actual reactions between 1R and 10 molar equiv of CMA
Scheme 6
are relatively fast, going to completion within a 24 h period.
MeO
H
The mixture produced when CMA is oxidized by 1
and 1
is rich in products, including a compound we tentatively assign
7
7
as ortho-(propan-3-al)-N-methylaniline (PMA) (20-30%)
based on GC/MS (m/z ) 163.1) analysis and mechanistic
considerations (see Discussion). MA (50-60%) and a small
amount of CA (<6%) (Scheme 7) are also observed in the
reaction mixture. These yields assume that one substrate CMA
R
is oxidized per dicopper 1 oxidant. There is at least one other
1H, affording the slowest oxidation rates, and this subject is
further discussed below. This trend appears to hold at higher
temperature. However, due to the weak binding of DMA at
higher temperatures (Supporting Information Table S2), extrac-
tion of accurate oxidation rate constants is difficult, and further
comment/conclusions reached concerning the rate data obtained
would be unwarranted. It should be pointed out that the presence
of additional “free” O2 in the reaction mixtures (i.e., not purging
the reactions with argon) did not significantly change the
H
MeO
product obtained in the oxidation of CMA by 1 and 1 , as
assessed by GC/MS, which has eluded identification (see
Supporting Information).
In contrast, there are relatively few oxidation products
Me2N
observed by GC/MS for CMA oxidation by 1
, with MA
(
74) Ambundo, E. A.; Friesner, R. A.; Lippard, S. J. J. Am. Chem. Soc. 2002,
124, 8770-8771.
(75) Shaffer, C. L.; Harriman, S.; Koen, Y. M.; Hanzlik, R. P. J. Am. Chem.
Soc. 2002, 124, 8268-8274.
(
76) Bhakta, M. N.; Wimalasena, K. J. Am. Chem. Soc. 2002, 124, 1844-1845.
R
calculated DMA rates or binding constants to 1 .
Reactions With Mechanistic Probes: N-Cyclopropyl-N-
methylaniline (CMA). CMA was used as a probe to investigate
(77) Attempts were made to obtain a pure standard of PMA. These attempts all
yielded impure material which quickly decomposed into a brown polymeric
substance, presumably due to inter- and intramolecular iminium formation
between the aldehyde and methylamine moiety.
5476 J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005

Substrate Oxidation by Copper−Dioxygen Adducts
A R T I C L E S
Scheme 8
and CA produced in relatively low yields (∼2% each). It should
be noted that the oxidation product seen with 1H and 1
MeO
,
which we assigned to PMA, is not produced in this reaction. A
nonvolatile compound was obtained following column chro-
matography of the reaction mixture, which was identified (by
1
H NMR spectroscopy; see Supporting Information) as the
+
Figure 3. Representation of the cationic portion of 2Me2N, generated from
the compound cif file using CrystalMaker 6.3 (CrystalMaker, Ltd., Yarnton,
Oxfordshire, U.K.). Hydrogen atoms (except those bound to the hydroxo
core) have been removed for clarity.
N-methylquinolinium cation (MQ ), presumed to be isolated
with the B(C6F5)4 counteranion since (i) the MQ salt is highly
-
+
soluble in organic solvents, (ii) readily moved down a silica
1
column, and (iii) showed no H-NMR signal arising from the
+
counteranion. Due to the cationic nature of MQ , we have been
unable to determine exact yields; however, this is the major
oxidation product present in ∼50% (or possibly 100%; see
Discussion) yield, as indicated by gravimetric analysis. To
overview, these results (summarized in Table 1, see Experi-
R
mental Section) strongly suggest that 1 oxidizes CMA through
an ET/PT pathway (also see Scheme 7 and Discussion below).
MDP Oxidation. The oxidation of (p-methoxyphenyl)-2,2-
R
dimethylpropanol (MDP) by 1 in CH2Cl2 at -80 °C was also
investigated. Under ETPT conditions, MDP loses two hydrogen
atoms (formally), resulting in the formation of (p-methoxy-
phenyl)-2,2-dimethylpropanone (MDK), while an ET/PT path-
way results in an unstable radical cation, which undergoes an
R-â C-C bond cleavage and leads to the formation of
p-methoxybenzaldehyde (MBA) (see Scheme 8).46
Figure 4. Representation of the cationic portion of 2MeO, generated from
the compound cif file using CrystalMaker 6.3 (CrystalMaker, Ltd., Yarnton,
Oxfordshire, U.K.). Hydrogen atoms (except those bound to the hydroxo
core) have been removed for clarity.
Reaction times were significantly longer for MDP than for
CMA oxidations, requiring several days for the reaction to go
R
to completion, as followed by the disappearance of 1 and the
R
R
appearance of 2 . Conversion of the peroxo complexes 1 to
R
the bis-µ-hydroxo complexes 2 is accelerated in the presence
of MDP compared to “autodecomposition” (i.e., with no
substrate present); however, the latter pathways must compete
since the overall yields of organic product were significantly
lower when compared to those of CMA oxidations; yields
ranged from 40 to 60% depending on the metal complex (Table
2
4
). Curiously, the complex with the most electron-donating
Me N
-pyridyl substituent (1
2
) not only afforded the highest yield
Figure 5. Representation (ChemDraw 3D) of the previously reported crystal
H 53
of oxidized product but also approached completion the fastest.
As with CMA oxidation (vide supra), this may be unexpected
considering it is the weakest (one-electron) oxidant of the 1
series. MDP oxidation by 1MeO and 1
structure of 2 . Note the different way in which the pyridylalkylamine
tridentate (MePY2)H ligand (see Introduction) is wrapped about the copper
MeO
Me N
2
ions, contrasting with that observed for 2
and 2
.
R
Me N
II
2
produced only MDK
provide representations of these Cu 2-bis-µ-hydroxo com-
as the oxidation product, suggesting an ETPT mechanistic
pathway (Scheme 8). For MDP oxidation by 1 , the major
product was MBA, with MDK observed as a minor product,
plexes, while Figures 5 and 6 show the previously described
structures of 2 and 1 , respectively (a mixture; see discussion
H
H
H
5
9,67,70
below).
Structural comparisons (see also Table 2) show
-
R
2
2
2-
suggesting that both ETPT and ET/PT pathways may be
that the µ-(OH )2 (all 2 ) and µ-η :η -peroxo or µ-(O )2
(ligand)-Cu groups (in 1 ) lie in equatorial positions. For 2
and 1 , the coordination of the tridentate pyridyl ligand is nearly
identical in terms of wrapping around the copper ion. In both
cases, one pyridyl arm is bound axially; the alkylamine nitrogen
is bound equatorially, and the other pyridyl arm is bound in a
pseudoequatorial position. By contrast, the crystal structures of
7
8
H
H
operable.
X-ray Crystallographic Studies of 2MeO and 2
Me N
H
2
. Refine-
ment parameters for [{Cu (MePY2)MeO}2(OH)2] (2MeO) and
II
2+
II
Me N
2+
Me N
2
[
{Cu (MePY2)
2
}2(OH)2] (2
) are given in the Sup-
porting Information (Table S1), and selected structural metric
parameters are displayed here in Table 2. Figures 3 and 4
MeO
Me N
2
2
and 2
reveal a completely different ligand coordination
(
78) No change in product distributions is noted if the reaction temperature is
increased to -60 °C, which is the highest temperature at which these
complexes will cleanly oxidize substrates.
geometry, which results in near C2h symmetry for the metal
complex. In these cases, both pyridyl arms are bound in
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5477

A R T I C L E S
Shearer et al.
be expected considering the electron-donating abilities of the
para-substituents, with the least electron-donating ligand
H
(
MePY2 ) supporting the most strongly oxidizing metal com-
plex. Two methods were utilized to estimate these trends in
reduction potentials. One was cyclic voltammetry, which was
plagued by problems, including poor signal-to-noise (due to the
low temperatures utilized) and the instability of 1R toward
reduction, even on fast time scales. The instability of these
complexes toward reducing conditions may explain the nonideal
behavior exhibited in the CVs (Figure 1). Nonetheless, we were
H
MeO
able to obtain observable reduction waves for 1 and 1
Figure 6. Representation (ChemDraw 3D) of the previously described
H
corresponding to 550 and 400 mV versus SCE, respectively.
crystal structure of 1 , the mixed species peroxo/bis-µ-oxo complex with
H
63,68
Me N
the (MePY2) ligand.
Note the similarities between the crystal structure
2
Complex 1
yielded CV data that were not interpretable,
of this bis(µ-OH-)dicopper(II) complex 2H (Figure 5) but the different way
in which the pyridylalkylamine tridentate ligand is wrapped about the metal
compared to that in 2
-1
even at scan velocities exceeding 50 V s . Therefore, to
supplement these CV studies, reduction reactions between
ferrocene derivatives and 1 were monitored by UV-vis
MeO
Me N
2
and 2
(Figures 3 and 4).
R
R
II
2+
R
Scheme 9. [{(MePY2) Cu }2(OH)2] (2 ) Complex Structures
spectroscopy in CH2Cl2 at -78 °C. Due to the irreversible nature
of the peroxo reduction, we cannot determine exact values for
reduction potentials through such methods. However, we can
determine the trend in reduction potentials for these complexes
and, with the help of the CV data, place an upper limit on the
R
H
reduction potential for 1 . The strongest oxidant, 1 , could
readily oxidize a slight excess of ferrocene (∼6 equiv added to
-
5
H
a 6.7 × 10 M solution of 1 ); however, it was incapable of
readily oxidizing substrates with higher one-electron oxidation
potentials than ferrocene through an SET pathway. Complexes
MeO
Me N
2
1
and 1
required stronger reductants to facilitate their
reductions, as no immediate reaction is observed when excess
MeO
ferrocene is added to these solutions. Complex 1
was readily
required dimethylfer-
rocene as a reductant. Similar to results obtained with ferrocene
equatorial positions, while the alkylamine-nitrogen is weakly
bound axially, Cu-N ) 2.27-2.30 Å (Table 2 and Scheme 9).
These results indicate that stronger donor substituted pyridines
Me N
reduced by ethylferrocene, while 1
2
Me N
Me N
2
and 1
2
, ethylferrocene is incapable of reducing 1
results strongly suggest that the trend in redox potentials follows
. These
(
i.e., MeO- or Me2N-) prefer to bind equatorially within the
MeO
Me N
2
square-based pyramidal µ-hydroxo complexes, 2
and 2
.
the order 1H > 1
MeO
> 1
Me N
, with an upper limit on the
being below 400 mV versus SCE
Thus, we suggest that in structurally similar8 µ-peroxodi-
copper(II) (or bis-µ-oxo dicopper(III) isomer forms; see Discus-
2
Me N
2
reduction potential of 1
MeO
MeO
Me N
(based on the CV data for 1 ). These results are in good
agreement with those reported by Stack and co-workers, whom
to our knowledge, are the only others to have reported estimates
of reduction potentials involving Cu2-O2 species. Similar
approaches were employed in those studies; approximate
reduction potentials of Cu2-dioxygen adducts centered about
sion) complexes, 1
and 1
2
, a similar situation may occur
and the coordination (i.e., ligand wrapping) may be different
H
for 1 . Such structural differences or variations may need to
be considered to explain certain aspects of the reactivity patterns
observed.
Discussion
5
00 mV versus SCE using decamethylferrocene as the reductant
2
9,38
in redox titrations.
The major goal of this study was to investigate the initial
steps of PCET reactions facilitated by copper-dioxygen adducts,
Our EPR spectroscopic inquires (see Results) indicate that
II
R
2- 2+
R
II
R
2- 2+
R
[{Cu (MePY2) }2(O2 )] (1 ). To meet these ends, we
the ferrocene reductions of [{Cu (MePY2) }2(O2 )] (1 )
involve 2 equiv of ferrocene per Cu2-dioxygen adduct. We
suggest that following consumption of 1 equiv of the ferrocene,
utilized several different approaches. First, we investigated the
R
fundamental redox properties of copper-dioxygen adducts 1 ,
II
III
+
attempting to determine if PCET versus SET reactions can be
an unstable and spectroscopically undetectable [Cu Cu (O)2]
R
thermodynamically supported by 1 in CH2Cl2 at -80 °C. We
species is formed (Scheme 5) that transforms through a
disproportionation reaction to give 1 equiv of a “fully reduced”
then investigated the oxidation of DMA in some detail to see if
R
substrate coordination likely occurs and then related this back
species and 1 equiv of 1 . This makes the overall process appear
R
to different PCET reactions that 1 can (and cannot) support.
as though it consumes 2 equiv of ferrocene derivative. There
are two possible processes that may produce the fully reduced
species, which could be the bis-µ-hydroxo dicopper(II) com-
R
Finally, we investigated the reaction of 1 with two different
mechanistic probes, CMA and MDP, to determine if these PCET
reactions likely involve concerted or consecutive proton-coupled
electron-transfer reactions.
R
plexes 2 with protons which we presume derived from solvent.
II
III
One is that the Cu Cu mixed-valence one-electron-reduced
product can readily and rapidly accept a second electron from
a second equivlent of decamethylferrocene. Although possible,
we think this situation to be unlikely as the formation of a
R
Thermodynamics. Copper-dioxygen adducts 1 all appear
to be mild one-electron oxidants with reduction potentials all
H
below ∼550 mV versus SCE (-78 °C), with 1 being the most
Me N
II
2- 0
2- 2+
oxidizing complex and 1
2
being the least. This trend may
[Cu 2-O2 ] species from a stable [Cu2-(O2 )] complex
5478 J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005

Substrate Oxidation by Copper−Dioxygen Adducts
A R T I C L E S
II
R
should be thermodynamically unfavorable; no reduced Cu 2-
bis-µ-oxo complex has ever been observed. The second pos-
sibility, as we suggested above as being more likely, is that the
mixed-valence species is produced and undergoes a rapid
duction pathways of 1 , mentioned above, are accelerated by
acid sources. Therefore, photometric acid/base titrations would
be exceptionally difficult. Another reason that pKa estimations
are impractical in the present system is that many of the weak
acids that could be used, such as thiophenol or phenol, will
R
disproportionation reaction. Furthermore, the CV studies on 1
indicate that they are unstable with respect to reduction by one
electron even if no protons are present. It should also be pointed
out that although a trimeric Cu Cu 2-(O)2 species has been
a mixed-valence Cu Cu -(O)2 complex has
never been observed, suggesting that such species are highly
R
•
also act as substrates to 1 , giving up a formal H to the copper
complexes. A third and most important reason that a full pKa
study cannot be adequately performed is due to the lack of
quantitative or meaningful data for pKa’s in CH2Cl2. Although
the trend in acidity between MeCN and CH2Cl2 is similar,84
exact values are not known. We can, therefore, only place a
lower limit on the peroxo/bis-µ-oxo pKa values for 1 in CH2-
Cl2 as being ∼15; however, the actual pKas may be much larger
III
II
4
1,80-82
II
III
isolated,
unstable.
R
It appears that oxidations of substrates by one electron are
aided by the thermodynamic driving force imparted by a proton
transfer (PT) event. For example, all three complexes are
incapable of oxidizing [Fe (phen)2(phen-NO2)](ClO4)2 (E1/2 )
.3 V versus SCE in CH2Cl2). No reaction is observed when a
solution of [Fe (phen)2(phen-NO2)](ClO4)2 is injected into a
solution of 1 at -80 °C. However, upon the addition of
ammonium perfluorotetraphenylborate as a proton source,
Fe (phen)2(phen-NO2)](ClO4)2 is readily oxidized by all three
in magnitude.
II
Similar to the observations noted by monitoring the reduction
of 1 by decamethylferrocene, no mixed-valent copper species
(i.e., EPR detectable copper species) are observed when
substoichiometric amounts of TEMPO-H are added to 1 , yet
R
1
II
R
R
•
the full formation of TEMPO is noted by spin integration
II
II
III
2+
[
1
(Scheme 4). Therefore, a Cu Cu (O)(OH) species, which may
possibly have formed in such reactions, may be so unstable with
respect to disproportionation that they are undetectable using
these methods. This is supported by UV-vis titrations involving
TEMPO-H, which afforded no detectable intermediate. Only
R
R
, resulting in 2 as the sole copper-containing product.
R
Overall, these results clearly indicate that 1 complexes are
capable of oxidizing substrates with much less favorable ET
driving forces through a PCET pathway. In other words, the
availability of protons and PT makes the ET event thermody-
namically favorable, driving the overall reaction to completion.
This helps explain why substrates with Eox much more positive
II
R
2+
R
II
the clean conversion of [{Cu (MePY2) } (O )] (1 ) to [{Cu -
2
2
R
2+
R
(MePY2) } (OH) ] (2 ) is observed. We should also note that
2
2
R
R
if 1 equiv of 2 is added to a solution of 1 , no observable
reaction occurs, suggesting the mixed-valence complex is
thermodynamically unstable with respect to the combination of
R
44
than for 1 (>1.5 V; e.g., THF, Et2O, and 4-CN-DMA) can
be readily oxidized by 1 . In these reactions, a PT event
R
II
R
2+
R
II
R
2+
following (or coupled to) the ET reaction can thermodynamically
drive the reaction to completion. Similar results have been
observed by Tolman and co-workers, who have seen that
ferrocene is unreactive toward some tacn-ligated Cu2-O2
complexes unless an acid source is added.31 This facilitated a
redox reaction producing ferrocenium and reduced copper
products.
If we now make the PT event less favorable by using a weaker
acid (i.e., acetic acid, which is weaker than ammonium by ∼7
pKa units in MeCN), then no reaction is observed between
Fe (phen)2(phen-NO2)](ClO4)2 and 1 . This most likely results
from a combination of the low driving force for ET ([Fe (phen)2-
phen-NO2)](ClO4)2 is a weak reductant) coupled to the low
driving force for protonation (acetic acid is a fairly weak acid
in CH2Cl2); the resulting Cu2-OOH adduct formed under these
conditions would have a significantly weaker O-H bond than
that which was formally “broken”; therefore, the overall reaction
is not thermodynamically viable and will not proceed. At-
tempts to estimate peroxo/bis-µ-oxo pKas, which is vital to better
estimate Cu2-OOH BDEs for 1 , were not fruitful for three
reasons. One is that 1 complexes are not very stable toward
acid; over a relatively short period of time (minutes to hours),
we observe the decomposition of 1R to 2 , even after the
addition of relatively weak acids. We believe that the autore-
2 2 2 2
[{Cu (MePY2) } (O )] (1 ) and [{Cu (MePY2) } (OH) ]
R
(2 ) (i.e., these two compounds do not comproportionate).
Substrate Coordination and Oxidation. The oxidation of
has been studied by us and was shown to display
second-order reaction kinetics behavior (first order in 1 , first
order in THF). The low activation enthalpy (∆H ) 3.7 kcal
Me N
THF by 1
2
Me N
2
q
-
1
q
mol ) and large negative activation entropy (∆S ) -56 cal
-
1
-1
mol K ) (see Supporting Information) for this reaction
suggest a pre-equilibrium THF binding step followed by
oxidation; however, saturation behavior was not observed.
Furthermore, reactions with 1 /1 and THF suggested that
as little as 1 equiv of MeCN (as a competitive inhibitor) could
dramatically influence the observed second-order rate constant,
which lends more evidence to a substrate coordination event.
In contrast to THF oxidation, variable temperature kinetic studies
II
R
MeO
H
[
II
(
R
of DMA oxidation by 1 does provide clear evidence that
oxidation takes place following DMA binding to the metal
complex. At high DMA concentration (>20 mM), the observed
rate constant begins to plateau, clearly demonstrating saturation
behavior (i.e., substrate coordination; see Figure 2). This is
strongly pronounced at low temperatures (i.e., <-70 °C).
The quality of the fits to the data and narrow temperature
range necessitated by the stability of 1 makes it difficult to
draw any detailed conclusions concerning the thermodynamic
and kinetic properties of the 1 series, yet generalizations can
be reached. One is that binding of DMA to the metal center
appears to be entropically controlled, as may be expected. The
8
3
R 65
R
R
R
R
(
(
(
(
(
79) Addison, A. W.; Rao, T. N.; Reedijk, J.; van Rijn, J.; Verschoor, G. C. J.
Chem. Soc., Dalton Trans. 1984, 1349-1356.
80) Machonkin, T. E.; Mukherjee, P.; Henson, M. J.; Stack, T. D. P.; Solomon,
E. I. Inorg. Chim. Acta 2002, 341, 39-44.
-1
calculated ∆S° values for DMA binding are all <-40 cal mol
K , which is in line with the need to use low temperatures
81) Cole, A. P.; Root, D. E.; Mukherjee, P.; Solomon, E. I.; Stack, T. D. P.
Science 1996, 273, 1848-1850.
-1
82) Root, D. E.; Henson, M. J.; Machonkin, T.; Mukherjee, P.; Stack, T. D.
P.; Solomon, E. I. J. Am. Chem. Soc. 1998, 120, 4982-4990.
83) (a) Mayer, J. M. Acc. Chem. Res. 1998, 31, 441-450. (b) Roth, J. P.; Yoder,
J. C.; Won, T.-J.; Mayer, J. M. Science 2001, 294, 2524-2526.
(84) Kim, H.-s.; Chung, T. D.; Kim, H. J. Electroanal. Chem. 2001, 498, 209-
215.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5479

A R T I C L E S
Shearer et al.
(
1
<-70 °C) to observe clear evidence of DMA coordination to
analysis of MDA oxidation suggests an apparent third-order
reaction (first order in MDA, second order in Cu2 -bis-µ-oxo),
the mechanistic implications of which are ambiguous at this
time.
R
III
. In contrast, the calculated ∆H° values appear to favor DMA
-1
binding by ∼10 kcal mol . Once binding occurs, the oxidation
q
reactions are driven by fairly low ∆H , which were all under
-
1
1
5 kcal mol . Although these activation enthalpies are lower
ETPT versus ET/PT using Mechanistic Probes. Oxidation
II
R
2+
R
than those typically observed for metal-mediated PCET
reactions, they are not without precedent. For example, intra-
molecular N-dealkylations by Cu 2-bis-µ-oxo complexes
reported by Itoh and co-workers proceed with ∆H well below
of CMA by [{Cu (MePY2) } (O )]
(1 ) results in the
2
2
83a
R
formation of bis-µ-dihydroxy-dicopper(II) complexes, 2
III
(Scheme 3), as the sole isolable copper-containing products.
q
MeO
H
CMA oxidations by 1
and 1 will be discussed first. Here,
-
1 9
1
0 kcal mol .
GC/MS analysis revealed a number of different organic products
in the reaction. The major product, in fact, is N-methylaniline
(MA), comprising the majority of the resulting oxidation product
mixture, strongly supporting the conclusion that an ET/PT
pathway occurs (Schemes 7 and 10).
It was noted above that substrate oxidation rates as a function
R
of R (for 1 ) appear to be the opposite of what one may expect.
The weakest SET oxidant affords the fastest reaction rates, while
the strongest SET oxidant affords the slowest oxidation rates!
This can be the result of many different factors, including strong-
If an ETPT event occurred, then N-methyl dealkylation is
preferred due to steric, statistical, and thermodynamic reasons.
Therefore, N-cyclopropylaniline (CA) should be the major
oxidation product under ETPT conditions. We do find that CA
is formed in both reactions; however, its contribution to the
overall oxidation product mixture is less than 6% in both cases,
which lends additional evidence that a rate-limiting ET/PT
pathway is occurring during the oxidation of CMA. A third
product is also observed, which we tentatively assign as the
ring-opening product, PMA, although we should point out that
definitive proof (i.e., comparison with an authentic standard)
Me N
er hydroxo O-H bond strengths in the product 2
2
compared
H
III
to those in 2 , differing amounts of Cu 2-bis-µ-oxo isomer in
the 1 series, or different ligand coordination environments
R
about the metal center (vide infra). All three could have a dra-
matic influence on the oxidation rates of these PCET reactions,
and unfortunately, we do not see ways of sorting out how these
(or other) factors may be influencing reactivity with this system.
Substrate coordination to the metal center prior to oxidation
reactions has been observed via kinetic studies by Itoh for
18
phenolate ortho hydroxylations. There, the presence of an
7
7
anionic phenolate aids in coordination to the metal center, and
as to its identity is lacking. Formation of PMA is supported
by mass spectrometry (m/z ) 163.1) and makes mechanistic
sense if we consider the following scenario. SET (the first step
in the ET/PT scheme) would lead to an initially formed
N-cyclopropyl radical cation, unstable with respect to very rapid
-1 18
relatively weak binding occurs, Keq ∼ 500-1000 M , which
86
compares well with our results. In a related trinuclear Cu3-
O2 system,40 there is evidence for phenol coordination prior to
oxidation and corresponding ortho C-C coupling. There,
-1
8
-1
87
substrate binding is much weaker (K ) 44.6(22) M ) even at
low temperatures (203 K).40 Furthermore, the majority of known
oxidation reactions involving discrete Cu2-O2 complexes
ring opening, k > 10 s at -80 °C. This radical center could
then couple to the phenyl ring ortho to the amine moiety. The
unstable radical cation will then rapidly lose a proton, forming
9
•
involve ligand oxidative N-dealkylations. In all of these cases,
intermediate MQ , which has a radical center R to the amine
the “substrate” is already coordinated to the metal center,
negating the binding step.
group. This can then undergo a “rebound” step, capturing an
•
II
III
HO from the Cu OHCu O intermediate. Hydroxyl radical
capture by a structurally related amine radical, which forms an
R-amino alcohol, has been previously observed in photolysis
All of these findings suggest why thermodynamic consider-
ations could not predict that DHA, Me2THF, and toluene (all
7
9
R
experiments. Hydroxyl group capture during a rebound step
with weaker X-H bonds) would not be oxidized by 1 . It
R 42
is also supported by DMA and THF oxidation by 1 . The
appears that a substrate coordination event must take place prior
•
R
R-amino alcohol formed from MQ can then undergo a rapid
to oxidation by 1 , but these substrates would be (mostly)
ring-opening reaction, forming aldehyde PMA (see Scheme 10),
which would be in equilibrium with different iminium salts
formed through both inter- and intramolecular Schiff base
condensations. Iminium salt formation from PMA was indicated
by our attempted synthesis of PMA, which demonstrated that
PMA rapidly polymerizes; authentic PMA rapidly turns into
incapable of coordinating to the metal complex. DHA and
toluene are hydrocarbons with no heteroatom Lewis base to
facilitate coordination to the copper complex oxidants, while
the ether oxygen atom in Me2THF is rather sterically inacces-
sible compared to its counterpart in the unsubstituted THF
substrate. It should be pointed out that DHA oxidation (and
lack thereof) by Cu2-O2 complexes is not well understood.
Work by Stack and co-workers is consistent with our studies
7
7
brown, sticky, insoluble polymers.
Me N
The oxidation of CMA by 1
2
yields a distinctly different
III
product distribution, yet still supports the ET/PT mechanism
of reaction. First, almost no MA is observed. In fact, virtually
no organic products can be observed in the GC trace. Instead,
what is isolated from the reaction is N-methylquinolinium
where some Cu2 -bis-µ-oxo systems do not appear to react
3
8
with DHA. This is despite the fact that these complexes will
oxidize substrates with much stronger X-H bonds, all of which
happen to contain a heteroatom that is sterically accessible to
+
III
(MQ ) (Scheme 10). This product has been previously identified
the metal center. By contrast, Itoh’s Cu2 -bis-µ-oxo species
in heme-oxygenase oxidations of CMA and is readily formed
is reported to react with DHA as well as the structurally related
3
0
compound, 10-methyl-9,10-dihydroacridine (MDA). Kinetic
(
87) The reported rate constant for ring opening of a N-methyl-N-(trans-2-
phenylcyclopropyl)aminyl radical (a structurally related compound to CMA)
1
1 -1
(
85) Itoh, S.; Kumei, H.; Taki, M.; Nagatomo, S.; Kitagawa, T.; Fukuzumi, S.
J. Am. Chem. Soc. 2001, 123, 6708-6709.
has been measured at 7.2 × 10
s : Musa, O. M.; Horner, J. H.; Shahin,
H.; Newcomb, M. J. Am. Chem. Soc. 1996, 118, 3862-3868. Our estimate
of the lower limit for ring opening of CMA+ comes from this value and
then extrapolating down to -80 °C using the Arrhenius equation: k )
(
86) The phenolate oxidations reported by Itoh were carried out in acetone. This
solvent may compete somewhat with phenolate coordination, which could
suppress phenolate binding constants in this system.
-Ea/RT
Ae
.
5480 J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005

Substrate Oxidation by Copper−Dioxygen Adducts
A R T I C L E S
Scheme 10
from MQ•.47,88 If the rebound step is sufficiently slow with
by the vast body of work concerning coupled versus uncoupled
metal-mediated PT and ET reactions.
•
respect to a second ET/PT reaction, then a second H loss is
•
observed instead of HO capture. As indicated above in the EPR
II
R
Benzylic alcohol MDP is also oxidized by [{Cu (MePY2) }2-
R
studies, the mixed-valence forms of 1 are very unstable with
2+
R
(
O2)] (1 ) (Scheme 8); however, these reactions are all
•
42
respect to abstracting a second H . It has been previously
R
considerably slower than those between 1 and CMA, taking
days to go to completion. This could be due to either a change
in the thermodynamic driving force of the reaction (MDP is
Me N
is particularly reactive compared to 1H
and, therefore, may abstract a second H• more quickly
demonstrated that 1
2
MeO
and 1
H
MeO
•
than 1 or 1 . Loss of H results in 2-dihydro-N-methylquino-
1
.0 V less reducing) or potential differences in rates between
line (MQ), which rapidly oxidizes upon exposure to O2 (or to
ETPT and ET/PT, which can both dramatically influence
reaction rates. Differences in oxidation rates could also be due
to the increased steric bulk of MDP compared to that of CMA,
which may limit its access to the metal complex. Considering
that substrate coordination to the metal center seems to be a
requirement for oxidation activity, this seems to be a reasonable
supposition (vide supra).
Me N
+
a second equivalent of 1
2
), forming MQ (see Scheme 9).
Due to the inherently unstable nature of MQ, its direct
observation has thus far been elusive, but can be inferred from
+
the formation of MQ . As a result of the cationic nature of
+
MQ and the small reaction scales necessitated for experimental
Me N
reasons, quantification of reaction yields between 1
2
and
CMA was extremely difficult. Therefore, we could estimate a
In the case of 1MeO and 1
Me N
2
oxidation of MDP, the sole
+
lower limit on the yield of MQ (MQ) based on a gravimetric
isolable organic product is ketone MDK (Scheme 8). This is
strong evidence in favor of an ETPT reaction, in contrast to
the ET/PT oxidation of CMA facilitated by these complexes.
In an ET/PT scheme, oxidation of MDP will form a charged
radical cation intermediate, which will rapidly lead to R-â C-C
bond cleavage and result in MBA. However, ETPT reactions
would result in a relatively stable neutral radical species, which
would simply lose an additional hydrogen atom to form MDK.
The change in mechanism between CMA (see above) and MDP
analysis of the recovered product at ∼50%, which assumes a
two-electron oxidation of CMA to MQ followed by the air
+
oxidation of MQ to MQ . It is also plausible that the oxidation
+
of CMA to MQ does not require O2 and is produced directly
-
in the solution (i.e., the 2 e oxidation of CMA to MQ by 1
Me N
equiv of 1
2
would be rapidly followed by another equivalent
Me N
+
of 1
2
oxidizing MQ to MQ ), which would make this
oxidation a near quantitative process. This would be in line with
the fact that no MQ is observed, even during anaerobic workup
MeO
Me N
2
oxidations by 1
and 1
can be largely explained by the
of the reaction. Such a scenario is also consistent with the
fact that MDP is roughly 1.0 V harder to oxidize by one-electron
Me N
unstable nature of 1
2
with respect to reducing conditions.
than is CMA, yet both CMA and MDK have similar C-H BDEs
II
R
Thus, in all three cases, it appears that [{Cu (MePY2) }2-
O2)]2 (1 ) oxidizes CMA via an ET/PT pathway (vide supra).
-1
(
∼85 kcal mol ). Therefore, the overall MDK oxidation would
+
R
(
still be thermodynamically viable, and the ETPT pathway may
be the more easily accessible pathway (i.e., compare to ET/
This is despite the fact that the one-electron oxidation of CMA
1
is thermodynamically unfavorable by approximately /2 V.
MeO
Me N
2
PT) for 1
and 1
oxidations, with substrates possessing
Coupling in the protonation step (PT) makes the overall
oxidation favorable, as indicated by the acid-base/redox
titrations mentioned above. This type of mechanism is supported
such inaccessible oxidation potentials and “weak” X-H bonds.
This is supported by our previous study investigating para-
R 44
substituted dimethylaniline (R-DMA) oxidations by 1 . There,
we found that as the R-DMA was made more difficult to oxidize,
the mechanism switched over from ET/PT to ETPT. Switching
(
88) Shaffer, C. L.; Morton, M. D.; Hanzlik, R. P. J. Am. Chem. Soc. 2001,
1
23, 8502-8508.
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5481

A R T I C L E S
Shearer et al.
mechanisms between ET/PT and ETPT as the substrate is made
more difficult to oxidize has been observed in a wide variety
of metal-oxo-mediated substrate oxidations. Most notable are
high-valent Fe-oxo porphyrinates,89,90 which will perform both
ETPT and ET/PT reactions depending on the oxidation potential
of the substrate.
In contrast, [{Cu (MePY2) }2(O2)] (1 ) can oxidize MDP
in a different manner, mostly through an ET/PT pathway, as
indicated by formation of MBA as the major oxidation product
Scheme 11
II
H
2+
H
(Scheme 8). This is also supported by our previous R-DMA
study, which demonstrated that only an ET/PT pathway is
H
44
operable for 1 oxidations. Surprisingly, we also found that
1
approximately /3 of the MDP can be oxidized via the ETPT
pathway, as indicated by the formation of ketone MDK as the
minor product. This was not hinted at in the previous 4-sub-
stituted 4-R-DMA studies,44 probably due to the ambiguity in
interpreting kinetic isotope effect data.
•
+
that an oxygen atom derived H or H acceptor would lie
proximate to the bound substrate. Also, differential substrate
binding patterns could hinder the substrate from approaching
H
MeO
Me2N
the metal center in 1 versus 1
and 1
, which can have
In previous work, Baciocchi46 and Mayer91 utilized MDP as
a probe for metal-oxo-mediated oxidation reactions. Mayer and
co-workers have shown that MDP oxidation by Mn(hfacac)3
proceeds through an ET/PT pathway, which may be expected
a dramatic influence on reaction rates and pathways (see Scheme
11).
H
Another reason for the differential reactivity between 1
may be the consequence of varying amounts
of peroxo/bis-µ-oxo isomer in solution. Complex 1 is almost
MeO Me N
versus 1 /1
2
R
given that Mn(hfacac)3 is significantly more oxidizing than 1 .
H
In this case, MDP SET would be thermodynamically downhill
and, therefore, accessible.91 Baciocchi has shown that Co -W
in acidic aqueous solutions also results in an ET/PT oxidation
Me N
MeO
all peroxo (∼90:10 peroxo:bis-µ-oxo), while 1
2
and 1
III
have substantially more bis-µ-oxo isomer present (∼75:25
43
peroxo:bis-µ-oxo). It is possible that the bis-µ-oxo isomer is
+
•
chemistry, while cytochrome P450 (which contains a [(por )-
responsible for the ETPT pathway, as is suggested by the many
examples of H abstraction chemistry by bis-µ-oxo dicopper-
III) complexes and from computational results.
IV
Fe dO] moiety in its active form) will oxidize MDP through
•
4
8
an ETPT pathway.
Differential Reactivity and Relation to Copper Catalysis.
9
6,7,93
(
These
molecular orbital (MO) calculations have suggested that the
peroxo tautomer is incapable of facilitating ETPT reactions,
however, would be capable of facilitating ET/PT (and hydride
transfer) reaction. In contrast, MO calculations seem to support
the idea that the bis-µ-oxo tautomer is capable of supporting
ETPT reactions; however, its ET ability has not been probed,
and in the end, it may prove to be a more capable SET oxidant
than the peroxo tautomer. Therefore, it seems reasonable that
the complex with the most bis-µ-oxo isomer present would be
more likely to undergo an ETPT reaction, while the peroxo
isomer may be responsible for the ET/PT reaction. This could
explain why both ETPT and ET/PT pathways are operable for
II
H
The difference in oxidative reactivity of [{Cu (MePY2) }2-
2
+
H
MeO Me N
2
(O2)] (1 ) versus 1 /1
toward MDP, DMA, and CMA
could be the result of several different factors. One could be a
difference in substrate approach to the metal center. We have
R
evidence from the crystal structures of 2 that the ligand adopts
different geometries about the metal center of these µ-hydroxo
complexes (see discussion above, and Scheme 9). A similar
R
trend may exist for Cu2O2 complexes 1 . Further support for
MeO Me N
differential ligand coordination environments in 1 /1
2
H
versus 1 comes from resonance Raman spectroscopic studies
R 43
and vibrational analyses of 1 . The Cu-Npy stretching
MeO
frequencies are measurably higher in peroxo complexes 1
H
MDP oxidation by 1 ; however, it alone cannot explain why
there is no ET/PT product observed for MDP oxidation by 1
Me N
-1
and 1
2
(295 and 305 cm , respectively) compared to that
MeO
R
-1 43
in 1 (275 cm ); as the N donor strength of pyridine is
increased, the copper ion moves more into plane and the tertiary
nitrogen and oxygen ligands become poorer donors. Additional
precedent for the current suppositions concerning 1 /2 com-
parisons comes from Tolman and co-workers,92 who have
published crystallographic evidence that a similar ligand
coordination environment (i.e., wrapping of the multidentate
Me N
2
and 1
.
It should be pointed out that at this point we cannot state
which tautomer (peroxo, bis-µ-oxo, or both) is responsible for
the observed reactivity. However, it is quite interesting that the
H
H
Me N
complex that contains the most bis-µ-oxo tautomer (1
2
) is
the “best” oxidant in terms of both the fastest reaction rates
and the highest yields of product. Whether this is merely a
coincidence or suggestive that the bis-µ-oxo tautomer is the
reactive species remains to be seen. Furthermore, [{Cu -
) is also the complex that is the
weakest SET oxidant, which means that other factors (possibly
III
2-
nitrogenous ligand about Cu) exists for both Cu 2-(O )2 and
Cu 2-(OH)2 complexes. If the coordination environment is
II
II
similar in the peroxo and bis-µ-oxo cases, then the substrate
Me N
2+
Me N
2
(MePY2)
2
}2(O2)] (1
H
MeO
may have differential binding geometries in 1 versus 1
/
Me N
1
1
2
, which may make the ETPT pathway less favorable in
due to geometric constraints and a diminishing likelihood
9
4
differing bis-µ-oxo complex acidities) must be playing a
significant role in these oxidations.
H
(
(
(
89) Sato, H.; Guengerich, F. P. J. Am. Chem. Soc. 2000, 122, 8099-8100.
90) Hall, L. R.; Hanzlik, R. P. J. Biol. Chem. 1990, 265, 12349-12355.
91) Bryant, J. R.; Taves, J. E.; Mayer, J. M. Inorg. Chem. 2002, 41, 2769-
(93) Henson, M. J.; Mukherjee, P.; Root, D. E.; Stack, T. D. P.; Solomon, E. I.
J. Am. Chem. Soc. 1999, 121, 10332-10345.
III
(MePY2)Me2N(O)
]2+, because
2
776.
(94) We suggest that bis-µ-oxo complex [{Cu
2
2
(92) Mahapatra, S.; Halfen, J. A.; Wilkinson, E. C.; Pan, G.; Wang, X.; Young,
of its Me
2
N ligand substituents and resulting coordination properties, may
form a stronger O-H bond upon acceptance of H or H from a substrate,
+
•
V. G., Jr.; Cramer, C. J.; Que, L., Jr.; Tolman, W. B. J. Am. Chem. Soc.
Me2N
1
996, 118, 11555-11574.
explaining the enhanced reactivity of 2
.
5
482 J. AM. CHEM. SOC. VOL. 127, NO. 15, 2005
9

Substrate Oxidation by Copper−Dioxygen Adducts
A R T I C L E S
More broadly, the results presented here seem to suggest that
oxidations by catechol oxidase (CO) probably occur through a
PCET pathway. If analogies are to be made between these
copper complexes and the active site of CO, it seems reasonable
to suggest that the Cu2-O2 center is a weak one-electron oxidant
and may need to make use of a PT event following ET to
thermodynamically drive the oxidation of catechols to ortho-
quinones to completion. A major facet of the research presented
above suggested that substrate coordination to the metal centers
is a major factor in substrate oxidations. Substrate coordination
Scheme 12
has been suggested to take place in both CO and Tyr, which
this study (and those from Itoh)1
8,40
fully supports. Therefore,
if Cu2-O2 species were to be used in synthetic and catalytic
systems to promote practical oxidation reactions, it may be
prudent to design ligands in such a way that the copper center
would be accessible to the substrate. Also, these studies suggest
that a strong SET reagent is not necessary to promote these
PCET oxidations if the overall thermodynamics of the PCET
reactions are favorable.
statements by pointing out that since we implicate DMA, THF,
or other substrate binding to the copper-dioxygen complexes,
reference to ET processes may be better described as “inner-
sphere substrate electron-transfer” oxidations; nevertheless, clear
mechanistic differentiation based on established ETPT versus
Summary and Concluding Remarks
In this study, we have attempted to provide a more in-depth
understanding of the initial steps of the oxidation of substrates
R
ET/PT mechanistic criteria is observed here in these 1 -mediated
substrate oxidations. Investigations on other systems are needed
to determine if the findings here are general and also to provide
a more comprehensive understanding of copper-dioxygen
substrate reactivity patterns and mechanism of action.
II
R
R
by [{Cu (MePY2) }2(O2)](B(C6F5)4)2 (1 ). We have provided
clear evidence for substrate coordination to the metal complex,
most likely at one of the two copper centers; this is vital for
subsequent oxidation reactions to take place. Furthermore, as a
new approach to studies in copper-dioxygen substrate oxidation
chemistry, we have used mechanistic probes to show that these
complexes are capable of oxidizing substrates through both
ETPT and ET/PT reactions. Complexes 1R appear to be
moderately to weakly oxidizing, and the SET oxidation reactions
of all of the substrates investigated are certainly thermodynami-
cally uphill processes. Therefore, 1R makes use of PCET
pathways to effect substrate oxidations/oxygenations. At mod-
erately unfavorable thermodynamic driving forces for electron
Acknowledgment. This work was supported by a grant from
the National Institutes of Health (K.D.K., GM28962), and a
Public Health Service Postdoctoral Fellowship (J.S.,
F32GM067447).
Supporting Information Available: Plots of the temperature
R
dependence of kobs versus [DMA] for DMA oxidation of 1 ,
temperature dependence of kobs versus [THF] for THF oxidation
Me N
by 1
2
, van’t Hoff and Eyring plots for DMA binding and
R
Me N
2
oxidation by 1 , CV data for 1
ferrocene “titrations” by 1 and 1
, representative UV-vis of
R
transfer from substrate to 1 , it appears that an ET/PT process
H
Me N
2
, representative GCs for
is dominant. At very unfavorable thermodynamic driving forces,
the ETPT process is foremost. These results are summarized in
Scheme 12. A switchover in mechanism has also been observed
R
MeO
1
oxidation of CMA, and cif files for X-ray structures of 2
Me N
and 2
2
. This material is available free of charge via the
Internet at http://pubs.acs.org.
in other metal-mediated oxidations as a function of substrate
oxidation potential.4
4,91
We must qualify these concluding
JA045191A
J. AM. CHEM. SOC.
9
VOL. 127, NO. 15, 2005 5483