Probing the compound I-like reactivity of a bare high-valent oxo iron porphyrin complex: The oxidation of tertiary amines

Article abstract of DOI:10.1021/ja077286t

The mechanisms of oxidative N-dealkylation of amines by heme enzymes including peroxidases and cytochromes P450 and by functional models for the active Compound I species have long been studied. A debated issue has concerned in particular the character of the primary step initiating the oxidation sequence, either a hydrogen atom transfer (HAT) or an electron transfer (ET) event, facing problems such as the possible contribution of multiple oxidants and complex environmental effects. In the present study, an oxo iron(IV) porphyrin radical cation intermediate 1, [(TPFPP)^?+Fe ^IV=O]⁺ (TPFPP = meso-tetrakis (pentafluorophenyl) porphinato dianion), functional model of Compound I, has been produced as a bare species. The gas-phase reaction with amines (A) studied by ESI-FT-ICR mass spectrometry has revealed for the first time the elementary steps and the ionic intermediates involved in the oxidative activation. Ionic products are formed involving ET (A^?+, the amine radical cation), formal hydride transfer (HT) from the amine ([A(-H)]⁺, an iminium ion), and oxygen atom transfer (OAT) to the amine (A(O), likely a carbinolamine product), whereas an ionic product involving a net initial HAT event is never observed. The reaction appears to be initiated by an ET event for the majority of the tested amines which included tertiary aliphatic and aromatic amines as well as a cyclic and a secondary amine. For a series of N,N-dimethylanilines the reaction efficiency for the ET activated pathways was found to correlate with the ionization energy of the amine. A stepwise pathway accounts for the C-H bond activation resulting in the formal HT product, namely a primary ET process forming A^?+, which is deprotonated at the α-C-H bond forming an N-methyl-N-arylaminomethyl radical, A(-H)^?, readily oxidized to the iminium ion, [A(-H)]⁺. The kinetic isotope effect (KIE) for proton transfer (PT) increases as the acidity of the amine radical cation increases and the PT reaction to the base, the ferryl group of (TPFPP)Fe^IV=O, approaches thermoneutrality. The ET reaction displayed by 1 with gaseous N,N-dimethylaniline finds a counterpart in the ET reactivity of FeO⁺, reportedly a potent oxidant in the gas phase, and with the barrierless ET process for a model (P)^?+Fe^IV=O species (where P is the porphine dianion) as found by theoretical calculations. Finally, the remarkable OAT reactivity of 1 with C₆F ₅N(CH₃)₂ may hint to a mechanism along a route of diverse spin multiplicity.

Full text of DOI:10.1021/ja077286t

Published on Web 02/16/2008
Probing the Compound I-like Reactivity of a Bare High-Valent
Oxo Iron Porphyrin Complex: The Oxidation of
Tertiary Amines
Barbara Chiavarino,† Romano Cipollini,† Maria Elisa Crestoni,†
Simonetta Fornarini,*,† Francesco Lanucara,† and Andrea Lapi‡
Dipartimento di Chimica e Tecnologia delle Sostanze Biologicamente AttiVe, UniVersita` di Roma
“La Sapienza” P.le A. Moro 5, I-00185, Roma, Italy, and Dipartimento di Chimica and Istituto
CNR di Metodologie Chimiche-IMC, Sezione Meccanismi di Reazione, UniVersita` di Roma
“La Sapienza” P.le A. Moro 5, I-00185, Roma, Italy
Received September 20, 2007; E-mail: simonetta.fornarini@uniroma1.it
Abstract: The mechanisms of oxidative N-dealkylation of amines by heme enzymes including peroxidases
and cytochromes P450 and by functional models for the active Compound I species have long been studied.
A debated issue has concerned in particular the character of the primary step initiating the oxidation
sequence, either a hydrogen atom transfer (HAT) or an electron transfer (ET) event, facing problems such
as the possible contribution of multiple oxidants and complex environmental effects. In the present study,
an oxo iron(IV) porphyrin radical cation intermediate 1, [(TPFPP)•+FeIVdO]+ (TPFPP ) meso-tetrakis
(pentafluorophenyl)porphinato dianion), functional model of Compound I, has been produced as a bare
species. The gas-phase reaction with amines (A) studied by ESI-FT-ICR mass spectrometry has revealed
for the first time the elementary steps and the ionic intermediates involved in the oxidative activation. Ionic
products are formed involving ET (A•+, the amine radical cation), formal hydride transfer (HT) from the
amine ([A(sH)]+, an iminium ion), and oxygen atom transfer (OAT) to the amine (A(O), likely a carbinolamine
product), whereas an ionic product involving a net initial HAT event is never observed. The reaction appears
to be initiated by an ET event for the majority of the tested amines which included tertiary aliphatic and
aromatic amines as well as a cyclic and a secondary amine. For a series of N,N-dimethylanilines the reaction
efficiency for the ET activated pathways was found to correlate with the ionization energy of the amine. A
stepwise pathway accounts for the CsH bond activation resulting in the formal HT product, namely a primary
ET process forming A•+, which is deprotonated at the R-CsH bond forming an N-methyl-N-arylaminomethyl
radical, A(sH)•, readily oxidized to the iminium ion, [A(sH)]+. The kinetic isotope effect (KIE) for proton
transfer (PT) increases as the acidity of the amine radical cation increases and the PT reaction to the
base, the ferryl group of (TPFPP)FeIVdO, approaches thermoneutrality. The ET reaction displayed by 1
with gaseous N,N-dimethylaniline finds a counterpart in the ET reactivity of FeO+, reportedly a potent oxidant
in the gas phase, and with the barrierless ET process for a model (P)•+FeIVdO species (where P is the
porphine dianion) as found by theoretical calculations. Finally, the remarkable OAT reactivity of 1 with
C6F5N(CH3)2 may hint to a mechanism along a route of diverse spin multiplicity.
Introduction
by an unstable carbinolamine intermediate that evolves by
nonenzymatic cleavage between the hydroxylated carbon and
The oxidative N-dealkylation of tertiary amines is an impor-
tant process, well documented in chemistry and biochemistry.
In particular, it constitutes a major pathway for the disposition
of xenobiotic amines including many drugs. In the biological
environment the process is catalyzed by heme enzymes, notably
by the peroxidase and cytochrome P450 families of enzymes,1
and results from the formal hydroxylation of a carbon adjacent
to nitrogen effected by a short-lived oxidant named Compound
I and abbreviated as (Porp)•+FeIVdO, where Porp is the
protoporphyrin IX dianion. The reaction conceivably proceeds
the nitrogen atom. For the enzymatic N-dealkylation reaction
two mechanisms have been formulated, differing for the primary
reactive step (Scheme 1).1-3 In the first route, a single electron
oxidation (ET) of the heteroatom gives a radical cation which
undergoes subsequent deprotonation (PT) at a CsH bond on
the carbon adjacent to the heteroatom, forming an R-carbon
centered neutral radical. Alternatively, a hydrogen atom transfer
(HAT) mechanism has been found to operate for example in
the cytochrome P450-catalyzed oxidation of 4-substituted N,N-
dimethylanilines and of N-cyclopropyl-N-alkyl-anilines.2,3 The
varied pattern of mechanistic pathways activated by cytochrome
P450 and peroxidase enzymes and their model compounds is
recognized to depend on the chemical and electronic structures
† Dipartimento di Chimica e Tecnologia delle Sostanze Biologicamente
Attive.
‡ Dipartimento di Chimica and Istituto CNR di Metodologie Chimiche-
IMC.
9
3208
J. AM. CHEM. SOC. 2008, 130, 3208-3217
10.1021/ja077286t CCC: $40.75 © 2008 American Chemical Society

The Oxidation of Tertiary Amines
A R T I C L E S
Scheme 1. Proposed Mechanisms in Oxidative N-Demethylation of Tertiary Amines Catalyzed by (Porp)•+FeIVdO, Initiated by Either
H-Atom Abstraction (HAT) or Single Electron Transfer (ET)
of substrates and active catalytic species that may affect the
partition between competitive routes.1,4,5 However, in spite of
intense interest in the past two decades, not all ambiguities have
been solved and there is a clear need for further work.1,6 In
particular, the direct observation of the significant steps in the
oxidation process is crucial for the elucidation of the detailed
mechanism. This is the aim of the present work, providing a
close, direct view of the charged intermediates involved in the
oxidation of gaseous amines by a synthetic oxo iron(IV)
porphyrin π-cation radical used as chemical model of the
Compound I intermediate.
In the study of complex reaction mechanisms that involve
the formation and decay of transient ionic intermediates, the
tools of gas-phase ion chemistry have proven valuable. In a
highly dilute environment reactive intermediates are isolated
and may be characterized about their structures, spectroscopic
properties, reactivities, and energetics.7 In this way a vast amount
of information on gaseous metal ion complexes has been
accessed and detailed insight into patterns of their catalytic
activity has been obtained.8 The advent of electrospray ionization
(ESI)9 coupled to mass spectrometry (MS) has added a new
dimension to the potential information that can be gathered from
the gas-phase chemistry of metal ion complexes.10 Ionic species
in dilute solution can be transferred intact directly to the gas
phase. One can take advantage of ESI-MS to isolate postulated
intermediates and investigate their reactivity patterns in the gas
phase, also relying on the remarkable similarity that is frequently
observed between ion molecule reactions in the gas phase and
the corresponding solution-phase reactions.10
In a recent report, a model of the high-valent oxo iron(IV)
species that is invoked as the key intermediate in enzymatic
oxygenation reactions has been produced and studied as a naked
ion in the gas phase.11 The oxo iron(IV) porphyrin radical cation
intermediate 1, [(TPFPP)•+FeIVdO]+ (where TPFPP is a 5,10,-
15,20-tetrakis(pentafluorophenyl)porphinato dianion and the
whole species within square brackets owns a unit positive
charge), was prepared by the reaction of the iron(III) porphyrin
chloride, (TPFPP)FeIIICl, with H2O2 in methanol and transferred
to the gas phase under mild ESI conditions. The elementary
steps of its gas-phase reaction with simple molecules of
biological significance (L) have been studied by FT-ICR mass
spectrometry providing data on intrinsic reactivity features and
on the formation of transient intermediates. The bare [(TPFPP)•+-
FeIVdO]+ ion (1) is found to react by oxygen atom transfer to
L (eq 1a), releasing [(TPFPP)FeIII]+, and by addition (eq 1b),
yielding [(TPFPP)Fe(L)O]+. While it behaves as a sluggish
(1) (a) Ortiz de Montellano, P. R. Cytochrome P450: Structure, Mechanism
and Biochemistry, 3rd ed.; Kluwer Academic/Plenum Publishers: New
York, 2005. (b) Guengerich, F. P. Chem. Res. Toxicol. 2001, 14, 611. (c)
Groves, J. T. J. Inorg. Biochem. 2006, 100, 434. (d) Groves, J. T. Proc.
Natl. Acad. Sci. U.S.A. 2003, 100, 3569. (e) Sono, M.; Roach, M. P.;
Coulter, E. D.; Dawson, J. H. Chem. ReV. 1996, 96, 2841. (f) Loew, G.
H.; Harris, D. L. Chem. ReV. 2000, 100, 407. (g) Meunier, B.; de Visser,
S. P.; Shaik, S. Chem. ReV. 2004, 104, 3947. (h) Shaik, S.; Kumar, D.; de
Visser, S. P.; Altun, A.; Thiel, W. Chem. ReV. 2005, 105, 2279. (i) Denisov,
I. G.; Makris, T. M.; Sligar, S. G.; Schlichting, I. Chem. ReV. 2005, 105,
2253. (j) Ortiz de Montellano, P. R.; De Voss, J. J. Nat. Prod. Rep. 2002,
19, 477. (k) Kim, D.; Guengerich, F. P. Ann. ReV. Pharmacol. Toxicol.
2005, 45, 27.
(2) (a) Bhakta, M. N.; Hollenberg, P. F.; Wimalasena, K. J. Am. Chem. Soc.
2005, 127, 1376. (b) Bhakta, M. N.; Hollenberg, P. F.; Wimalasena, K.
Chem. Commun. 2005, 265. (c) Bhakta, M. N.; Wimalasena, K. J. Am.
Chem. Soc. 2002, 124, 1844. (d) Bhakta, M. N.; Wimalasena, K. Eur. J.
Org. Chem. 2005, 4801.
(3) (a) Karki, S. B.; Dinnocenzo, J. P.; Jones, J. P.; Korzekwa, K. R. J. Am.
Chem. Soc. 1995, 117, 3657. (b) Dinnocenzo, J. P.; Karki, S. B.; Jones, J.
P. J. Am. Chem. Soc. 1993, 115, 7111. (c) Shaffer, C. L.; Morton, M. D.;
Hanzlik, R. P. J. Am. Chem. Soc. 2001, 123, 8502. (d) Shaffer, C. L.;
Harriman, S.; Koen, Y. M.; Hanzlik, R. P. J. Am. Chem. Soc. 2002, 124,
8268. (e) Cerny, M. A.; Hanzlik, R. P. J. Am. Chem. Soc. 2006, 128, 3346.
(f) Manchester, J. I.; Dinnocenzo, J. P.; Higgins, L.; Jones, J. P. J. Am.
Chem. Soc. 1997, 119, 5069.
(4) (a) Watanabe, Y. In The Porphyrin Handbook; Kadish, K. M., Smith, K.
M., Guilard, R., Eds.; Academic Press: San Diego, 2000; Vol. 4, Chapter
30, p 97. (b) Meunier, B.; Robert, A.; Pratviel, G.; Bernardou, J. In The
Porphyrin Handbook; Kadish, K. M., Smith, K. M., Guilard, R., Eds.;
Academic Press: San Diego, 2000; Vol. 4, Chapter 31, p 119. (c) Montanari,
F.; Casella, L. Metalloporphyrins Catalyzed Oxidations; Kluwer Academic
Publishers: Dordrecht, The Netherlands, 1994. (d) Fujii, H. Coord. Chem.
ReV. 2002, 226, 51.
(7) (a) Nibbering, N. M. M. Int. J. Mass Spectrom. 2000, 200, 27. (b) DePuy,
C. H. J. Org. Chem. 2002, 67, 2393. (c) Kass, S. R.; Broadus, K. M. J.
Phys. Org. Chem. 2002, 15, 461. (d) Olmstead, W. N.; Brauman, J. I. J.
Am. Chem. Soc. 1977, 99, 4219. (e) Caldwell, G.; Magnera, T. F.; Kebarle,
P. J. Am. Chem. Soc. 1984, 106, 959. (f) Morton, T. H. Tetrahedron 1982,
38, 3195.
(8) (a) Rodgers, M. T.; Armentrout, P. B. Acc. Chem. Res. 2004, 37, 989. (b)
Armentrout, P. B.; Beauchamp, J. L. Acc. Chem. Res. 1989, 22, 315. (c)
Organometallic Ion Chemistry; Freiser, B. S., Ed.; Kluwer: Dordrecht,
1996. (d) Bohme, D. K.; Schwarz, H. Angew. Chem., Int. Ed. 2005, 44,
2336. (e) Schroeder, D.; Schwarz, H. Angew. Chem., Int. Ed. Engl.
1995, 34, 1973. (f) Schroeder, D.; Schwarz, H. Top. Organomet. Chem.
2007, 22, 1. (g) Operti, L.; Rabezzana, R. Mass Spectrom. ReV. 2006, 25,
483.
(9) (a) Whitehouse, C. M.; Dreyer, R. N.; Yamashita, M.; Fenn, J. B. Anal.
Chem. 1985, 57, 675. (b) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S.
F.; Whitehouse, C. M. Science 1989, 246, 64.
(5) (a) Baciocchi, E.; Lanzalunga, O.; Lapi, A.; Manduchi, L. J. Am. Chem.
Soc. 1998, 120, 5783. (b) Baciocchi, E.; Lapi, A. Tetrahedron Lett. 1999,
40, 5425. (c) Baciocchi, E.; Gerini, M. F.; Lanzalunga, O.; Lapi, A.; Lo
Piparo, M. G.; Mancinelli, S. Eur. J. Org. Chem. 2001, 2305. (d) Goto,
Y.; Watanabe, Y.; Fukuzumi, S.; Jones, J. P.; Dinnocenzo, J. P. J. Am.
Chem. Soc. 1998, 120, 10762. (e) Lindsay Smith, J. R.; Mortimer, D. N.
J. Chem. Soc., Perkin Trans. 2 1986, 1743. (f) Baciocchi, E.; Gerini, M.
F.; Lanzalunga, O.; Lapi, A.; Mancinelli, S.; Mencarelli, P. Chem. Commun.
2000, 393. (g) Baciocchi, E.; Bietti, M.; Gerini, M. F.; Lanzalunga, O. J.
Org. Chem. 2005, 70, 5144.
(10) (a) Chen, P. Angew. Chem., Int. Ed. 2003, 42, 2832. (b) Gerdes, G.; Chen,
P. Organometallics 2003, 22, 2217. (c) Traeger, J. C. Int. J. Mass Spectrom.
2000, 200, 387. (d) Sabino, A. A.; Machado, A. H. L.; Correia, C. R. D.;
Eberlin, M. N. Angew. Chem., Int. Ed. 2004, 43, 2514. (e) Bossio, R. E.;
Hoffman, N. W.; Cundari, T. R.; Marshall, A. Organometallics 2004, 23,
144. (f) Schroeder, D.; Roithova, J.; Schwarz, H. Int. J. Mass Spectrom.
2006, 254, 197. (g) Trage, C.; Diefenbach, M.; Schroeder, D.; Schwarz,
H. Chem.sEur. J. 2006, 12, 2454. (h) Feng, W. Y.; Gronert, S.; Lebrilla,
C. J. Phys. Chem. A 2003, 107, 405. (i) Turecek, F. Mass Spectrom. ReV.
2007, 26, 563.
(6) (a) Nehru, K.; Seo, M. S.; Kim, J.; Nam, W. Inorg. Chem. 2007, 46, 293.
(b) Nam, W. Acc. Chem. Res. 2007, 40, 522.
(11) Crestoni, M. E.; Fornarini, S. Inorg. Chem. 2005, 44, 5379.
9
J. AM. CHEM. SOC. VOL. 130, NO. 10, 2008 3209

A R T I C L E S
Chiavarino et al.
paraformaldehyde and sodium borohydride as described in the
literature.3a,5c These compounds were purified by preparative GLC using
a 3 m column filled with Chromosorb 80/100 W-AW coated with a
base deactivated polyethyleneglycol stationary phase, mounted on a
Carlo Erba FRACTOVAP Mod. ATC/f series 410 gas chromatograph.
The identity and the purity of the products, besides their deuterium
content, were checked by GLC-MS on a Hewlett-Packard 5890 gas
chromatograph connected to a model 5989B quadrupole mass spec-
trometer, performing the analysis on a 50 m long, 0.2 mm i.d. fused-
silica capillary column, coated with cross-linked methylsilicone film.
Instrumental. All experiments were performed with a Bruker
BioApex Fourier Transform ion cyclotron resonance (FT-ICR) mass
spectrometer with an external Analytica of Branford Inc. ESI source.
The instrument is equipped with a cylindrical infinity cell and a 4.7 T
superconducting magnet. The vacuum is maintained by rotary vacuum
pumps and turbomolecular pumps. Analyte solutions were infused into
a 50 µm i.d. fused-silica capillary at a flow rate of 200 µL h-1 by a
syringe pump, and ions were accumulated in an rf-only hexapole ion
guide for 0.6 s. The ion population was pulsed into the ICR cell at
room temperature, and the ion of interest was isolated by broad-band
radio frequency pulses and exposed to a neutral reagent leaked by a
needle valve at a constant pressure value, comprised in the range of 1
× 10-8-2 × 10-7 mbar. Pressure readings were obtained from a cold
cathode gauge, calibrated by using the rate constant k ) 1.1 × 10-9
cm3 s-1 for the reference reaction CH4•+ + CH4 f CH5+ + CH3• and
corrected for different response factors.14,15 In the mass spectra,
corrections for 13C isotopic contributions were effected whenever the
ion of interest appeared as an isotopic cluster contaminated by the
presence of adjacent or overlapping peaks from another ionic species.
In particular, the correction was applied in evaluating kinetic isotope
effects by the ratio of ion abundances for species involving either H or
D atom loss.
Pseudo-first-order-rate constants were obtained from the slope of
the semilog decrease of the reactant ion signal vs time and divided by
the known pressure of the neutral to derive bimolecular rate constants
(kexp). The rate constants, obtained as average values from at least three
determinations at different neutral pressures at a room temperature of
300 K, were also expressed as percentages (φ, namely the reaction
efficiency) of the collision rate constant (kcoll) calculated by the
parametrized trajectory theory.16 The reproducibility of kexp values was
within 10%, while the error of the absolute rate constants is estimated
to be (30%. Branching ratios for parallel reaction channels were
obtained by extrapolation of product ion intensities at initial times, in
order to minimize any interference due to possible consecutive
processes. Collision induced dissociation (CID) experiments were made
in the FT-ICR cell, and argon was admitted by a pulsed valve and
used as a collision gas. Isolated ions were translationally excited by an
on-resonance radio frequency pulse. The excited ions were allowed to
collide with argon and dissociate for as long as 5 s before the product
ion radio frequency sweep.
oxidant toward olefins, the reactivity with phosphites, pyridines,
and sulfides is pronounced and goes in parallel with the
oxophilic character of the active site of L. An insight into the
dynamics of the oxygen transfer reaction has been gained
probing the so-formed [(TPFPP)Fe(L)O]+ adduct ions by their
ion-molecule reactions with a second neutral (M). Ligand
addition and ligand exchange products are observed (eq 2a and
2b, respectively). In all instances, the ligand exchange reaction
involves the displacement of an L(O) unit, never occurring by
displacement of L (eq 2c). This result points to the formation
of an oxidized species, L(O), bound to an axial site in the iron
coordination sphere, while the second axial coordination site
remains vacant and thus susceptible to undergo ligand addition.
The latter process would not be directly open to an isomeric
[(TPFPP)FeO(L)]+ species whereby an axial ligand L were
bound in a trans relationship to the oxo iron moiety, as more
explicitly indicated by the [(TPFPP)(L)FedO]+ notation. In-
terestingly, the same reactivity is displayed by [(TPFPP)Fe-
(L)O]+ ions extracted by ESI from a solution containing L
besides the precursors of 1. This finding suggests a common,
distinct intermediate along the pathway to oxygen transfer
products and provides evidence for the remarkable similarities
that emerge when comparing reactivity patterns and transient
intermediates occurring either in the gas phase or in solution,
based on the assay of ESI formed ions. The reactivity of 1
toward neutrals such as NO and NO2 is notably high, in line,
though, with the recognized role of prosthetic heme groups in
governing the biological chemistry of nitrogen oxides and in
particular with a recently suggested protective action of nitric
oxide toward highly oxidizing forms of heme proteins.12
Aiming to gain further information on the Compound I-like
reactivity of high-valent oxo iron intermediates in a solvent free
environment, the gas-phase reactivity of 1 with amines including
N,N-dimethylanilines has been examined and is presently
reported.
Experimental Section
Computational Details. DFT calculations were performed with the
Spartan’04 program package (Wavefunction, Inc., Irvine, CA) at the
B3LYP level (unrestricted formalism for open-shell systems) using the
6-311+G** basis set.
The reported energies include the zero-point vibrational energy,
calculated at the same level. The frequency calculations also verified
that the structures represented true minima. The bond dissociation
energies (BDEs) were calculated using the standard definition of the
BDE, that is, the difference in energy between the substrate and the
sum of the radical and hydrogen atom, using optimized geometries:
BDE ) E(radical) + E(H) - E(substrate). The ionization energy (IE)
of a substrate (S) and the proton affinity (PA) of a base (B) at 0 K
Materials. All chemicals used in the experiments and precursors
needed for the synthesis of useful substrates, including (5,10,15,20-
tetrakis(pentafluorophenyl)porphinato)iron(III) chloride ((TPFPP)FeIII-
Cl), iodosobenzenediacetate, [D2]paraformaldehyde, aniline, pentaflu-
oroaniline, 4-fluoroaniline, 4-bromoaniline, 4-(trifluoromethyl)aniline,
4-methylaniline, and sodium borohydride, were research grade products
obtained from commercial sources and used as received. Iodosylbenzene
(C6H5IO) was prepared according to a published procedure13 and stored
at -20 °C. The ring substituted N,N-bis(dideuteriomethyl)anilines were
prepared by the reaction of the corresponding anilines with [D2]-
(12) (a) Scheidt, W. R.; Ellison, M. K. Acc. Chem. Res. 1999, 32, 350. (b)
Gorbunov, N. V.; Osipov, A. N.; Day, B. W.; Zayas-Rivera, B.; Kagan,
V. E.; Elsayed, N. M. Biochemistry 1995, 34, 6689. (c) Herold, S.;
Rehmann, F.-J. K. Free Radical Biol. Med. 2003, 34, 269.
(13) Saltzman, H.; Sharefkin, J. G. Organic Syntheses; Wiley: New York, 1973;
Collect. Vol. 5, p 658.
(14) Meot-Ner, M. In Gas Phase Ion Chemistry; Bowers, M. T., Ed.; Academic
Press: New York, 1979; Vol. 1.
(15) Bartmess, J. E.; Georgiadis, R. M. Vacuum 1983, 33, 149.
(16) Su, T.; Chesnavich, W. J. J. Chem. Phys. 1982, 76, 5183.
9
3210 J. AM. CHEM. SOC. VOL. 130, NO. 10, 2008

The Oxidation of Tertiary Amines
A R T I C L E S
were similarly obtained (IE ) E(S•+) - E(S) and PA ) E(B) - E(BH+),
where PA is defined as -∆H° of the process: B + H+ f BH+).
Preparation and Sampling of [(TPFPP)•+FeIVdO]+ Ions. Prepara-
tion of the complex 1 and transfer into the gas phase have been
described previously in our earlier mechanistic work on the cytochrome
P450-like reactivity of high valent oxo iron intermediates in the
gas phase.11 For the present study iodosylbenzene was used as the
oxidant and the pure solid was added to a 5 × 10-5 M solution of the
parent (TPFPP)FeIIICl complex (making a 12:1 molar ratio) in CH3-
OH/CH2Cl2 (1:1) cooled at -40 °C. These experimental conditions
result in a good compromise between a fast and efficient formation of
the high-valent iron intermediate and an acceptable stability of the
solution. Fresh solutions kept at -40 °C are found to be stable for
about 1 h. Electrospray ionization of these solutions allows 1 to be
transferred to the gas phase and finally sampled in the FT-ICR cell.
Ions 1 appear in the FT-ICR mass spectrum as a prominent cluster
centered at m/z 1044 (henceforth the m/z value of the dominant peak
in the isotopic cluster is given). However, the overall ion signal is not
totally accounted for by [(TPFPP)•+FeIVdO]+ species because the
synthetic procedure leads to a fraction of isomeric species oxidized on
the porphyrin frame that cannot be removed. In order to evaluate this
fraction, one can exploit the ion chemistry specific of this species. In
fact, the oxidation of the porphyrin ligand, delivering an oxygen atom
presumably on the porphyrin periphery or on a pyrrolic nitrogen, does
not alter appreciably the reactivity of the iron center toward ligand
addition. This porphyrin oxidized species reacts in a closely similar
way as [(TPFPP)FeIII]+ ions.11 For example, both species react with
NO yielding a ligand addition product, whereas 1 releases an oxygen
atom to NO.11 Due to the easy disposal of this gaseous reagent from
the FT-ICR cell, NO has thus been used to quantify the relative
abundance of metal-oxidized and porphyrin-oxidized isomers, titrating17
in this way the m/z 1044 ion population. Differences in chemical
reactivity have long been recognized as a useful means for resolving
isomeric ion mixtures.18
Figure 1. ESI-FT-ICR mass spectrum from a CH3OH/CH2Cl2 (1:1) solution
of (TPFPP)FeIIICl (5 × 10-5 M) and iodosylbenzene (6 × 10-4 M). The
inset shows an enlargement of the isotopic cluster at m/z 1044.
atom transfer process to 1, the presence of [(TPFPP)FeIV(OH)]+
in substantial amounts already in the reagent ion mixture could
possibly obscure an important reaction pathway. The use of
iodosylbenzene for forming high valent oxo metal complexes
to be further characterized by ESI-MS is not unprecedented.20
Exploiting this oxygen atom donor21 is not completely void of
inconveniences though. Under the adopted reaction conditions,
optimizing the yield of the oxo iron complex and its lifetime in
solution, the ions that appear as an isotopic peak cluster centered
at m/z 1044, denote the neat addition of an oxygen atom to the
reagent [(TPFPP)FeIII]+ ions at m/z 1028. In the ESI-FT-ICR
mass spectrum (Figure 1) the two isotopic clusters of the reagent
ion and the oxidized species are characterized by similar ion
abundance ratios, as expected. However, when the ion cluster
at m/z 1044 is isolated in the FT-ICR cell and sampled by its
ion-molecule reactivity toward NO, it becomes clear that two
isomeric species are present. While a major fraction of ions
displays O-atom transfer reactivity, as already reported,11
releasing [(TPFPP)FeIII]+ ions and NO2 as the likely neutral
product, a second, substantial fraction appears less reactive,
yielding ultimately an addition product at m/z 1074. The NO
ligand addition reactivity is typical of tetracoordinate [(TPFPP)-
FeIII]+ ions11 and [(Porp)FeIII]+ ions19 in the gas phase. On this
basis it appears that those ions at m/z 1044 undergoing NO
addition are species oxidized on the porphyrin ligand, presum-
ably due to formation of a N-oxide on a pyrrole unit.22 Figure
2 displays the time dependence of ion abundances that follow
the isolation of the cluster at m/z 1044 which is then allowed
to react with a stationary concentration of NO. The decay of
the parent ion abundance is intrinsically biphasic, though it
hardly appears so because of the close values of the time
constants for the reaction of the two isomeric components. The
Results and Discussion
SynthesisandCharacterizationofGaseous[(TPFPP)•+FeIVd
O]+ Ions. Obtaining a positively charged high-valent iron(IV)
oxo porphyrin cation radical by a gas-phase reaction is by no
means a trivial task.11,19 Potential oxidants such as O2, N2O,
NO2, and oxirane were tested for reaction with [(TPFPP)FeIII]+
and [(Porp)FeIII]+; however they were found either to be
unreactive (O2, N2O) or to lead to products other than the desired
high valent oxo iron species. The synthesis of 1 was then
performed in solution by controlled oxidation of (TPFPP)FeIII-
Cl. The reagent used in a previous report, H2O2,11 was avoided
in the present work in favor of iodosylbenzene. The reason lies
in the dual reactivity of H2O2 which undergoes not only
heterolytic cleavage, yielding 1, but also a homolytic cleavage
of the peroxidic bond leading to [(TPFPP)FeIV(OH)]+.4b,11 The
gas-phase reactivity of the latter species is known to be markedly
different compared with that of 1, so that it may be expected to
cause little interference with the reactivity pattern of the high
valent oxo iron species under study.11 However, because
[(TPFPP)FeIV(OH)]+ product ions could result from a hydrogen
(17) (a) Bueker, H.-H.; Gruetzmacher, H.-F.; Crestoni, M. E.; Ricci, A. Int. J.
Mass Spectrom. Ion Processes 1997, 160, 167. (b) Mormann, M.; Salpin,
J.-Y.; Kuck, D. Int. J. Mass Spectrom. 2006, 249/250, 340.
(18) (a) McEwan, M. J. In AdVances in Gas Phase Ion Chemistry; Adams, N.,
Babcock, L. M., Eds.; JAI Press: Greenwich, 1992; Vol. 1. (b) Ausloos,
P.; Lias, S. G. J. Am. Chem. Soc. 1981, 103, 6505.
(19) (a) Angelelli, F.; Chiavarino, B.; Crestoni, M. E.; Fornarini, S. J. Am. Soc.
Mass Spectrom. 2005, 16, 589. (b) Irikura, K. K.; Beauchamp, J. L. J. Am.
Chem. Soc. 1991, 113, 2767. (c) Chen, O.; Groh, S.; Liechty, A.; Ridge,
D. P. J. Am. Chem. Soc. 1999, 121, 11910. (d) Chen, O.; Hagan, S. E.;
Groh, S.; Ridge, D. P. J. Am. Chem. Soc. 1991, 113, 9669.
(20) Feichtinger, D.; Plattner, D. A. J. Chem. Soc., Perkin Trans. 2 2000, 1023.
(21) Groves, J. T.; Nemo, T. E.; Myers, R. S. J. Am. Chem. Soc. 1979, 101,
1032.
(22) (a) Nam, W.; Lim, M. H.; Oh, S.-Y. Inorg. Chem. 2000, 39, 5572. (b)
Groves, J. T.; Watanabe, Y. J. Am. Chem. Soc. 1988, 110, 8443. (c) Machii,
K.; Watanabe, Y.; Morishima, I. J. Am. Chem. Soc. 1995, 117, 6691.
9
J. AM. CHEM. SOC. VOL. 130, NO. 10, 2008 3211

A R T I C L E S
Chiavarino et al.
concentration of tertiary amines (A), three ionic products may
be observed (eq 4a-c). The first one is the product of electron
transfer (ET), namely the radical cation of the amine (A•+),
implying the concomitant formation of neutral (TPFPP)FeIVd
O. A second product derives from a formal hydride transfer
(HT) process producing [A(sH)]+ ions (iminium ions, the [A(s
H)]+ notation indicates the species obtained from A by hydride
loss) and neutral (TPFPP)FeIII(OH). This reaction was found
to proceed by a multistep sequence to be discussed in a
following section. The third process releases [(TPFPP)FeIII]+
ions, implying an oxygen atom transfer to the amine (OAT). A
formal ligand addition product is observed as well. Upon close
inquiry, however, the origin of this product can be traced to
the reaction of the isobaric [(TPFPP+O)FeIII]+ species (eq 5)
which can obviously not be differentiated by the mass selection
routines of FT-ICR mass spectrometry.
Figure 2. Time dependence of relative ion intensities (I%) following the
selection of ions at m/z 1044 (a mixture of [(TPFPP)•+FeIVdO]+ and
[(TPFPP+O)FeIII]+) in the FT-ICR cell in the presence of NO at 3.8 ×
10-8 mbar.
product of the oxygen atom transfer reaction, [(TPFPP)FeIII]+,
subsequently adds NO giving [(TPFPP)FeIII(NO)]+ at m/z 1058
(eq 3a). In the slower process, the isobaric porphyrin oxidized
species, henceforth depicted as [(TPFPP+O)FeIII]+, yields the
NO adduct at m/z 1074 (eq 3b).
This conclusion is supported by several lines of evidence. The
ligand addition reactivity is well documented for tetracoordinate
ferric porphyrin ions such as [(TPFPP)FeIII]+ ions11 (as found
characteristic of the NO reaction with the porphyrin oxidized
[(TPFPP+O)FeIII]+ isomer). Notably, the extent of amine
addition is quantitatively matched by the fraction of nitrosyl
adduct formed from the same sampled ion population. This
outcome is expected from an ion mixture where the relative
amount of [(TPFPP+O)FeIII]+ isomer remains constant during
the time of the experimental assay, displaying the characteristic
ligand addition reactivity toward both NO and amines. Further
evidence pointing to coordination of the amine to a vacant axial
site on the metal comes from CID experiments activating the
[(TPFPP+O)FeIII(A)]+ adduct toward dissociation. The exclu-
sive loss of the intact amine is observed. No evidence is found
for the possible formation of an oxidized amine product within
the complex. This behavior is clearly different from the reactivity
displayed by the active oxidant 1 with pyridine, reported
previously.11 In this case the pyridine adduct showed clear
evidence of pyridine-O bond formation, releasing a formal
[pyridine(O)] unit both when activated to CID or when allowed
to undergo a ligand substitution process. The collective evidence
gathered on the reactivity toward amines, including trialky-
lamines, N-phenylpyrrolidine, piperidine, and N,N-dimethy-
lanilines, showed a uniform behavior whereby 1 ions yield ET,
HT, and OAT products and [(TPFPP+O)FeIII]+ ions yield the
corresponding adducts. The latter species are not of primary
interest. Their contribution to the overall pattern causes little
interference and can be subtracted with no consequence
whatsoever. Thus their reactions will not be further discussed.
In no instance is an HAT product, retaining the charge on
the iron porphyrin complex, ever observed. If formed, this
species, corresponding to [(TPFPP)FeIV(OH)]+ ions with a
formal protonated ferryl unit, would appear at m/z 1045 (eq
As a consequence of the two independent processes character-
ized by different rates, the ratio of the sum of ion abundances
for the species at m/z 1028 and m/z 1058 relative to the sum of
ion abundances for m/z 1044 and m/z 1074 ions reaches a
constant value at a sufficiently long reaction time. In the
experiment depicted in Figure 2 the ratio approaches the value
85:15. This constant ratio finally corresponds to the end ion
abundances of [(TPFPP)FeIII(NO)]+ and [(TPFPP+O)FeIII-
(NO)]+, which do not react further with NO. The reaction with
NO may then be used to probe the relative amount of the oxo
iron species in each new sample solution, providing a means to
titrate the relative amount of the isomer of interest in the ion
population produced by ESI.17 The fraction of 1 with respect
to the total ions at m/z 1044 was found to depend on the reaction
conditions adopted for the synthesis. At typically low temper-
atures and a constant relative amount of iodosylbenzene, as
described in the Experimental Section, the amount of the oxo
iron species was never below 60%. Furthermore, with all the
amines assayed in the present study, the isomeric [(TPFPP)•+-
FeIVdO]+ and [(TPFPP+O)FeIII]+ complexes proved to be
readily discriminated by their distinct reactivities. The careful
scrutiny of the reactivity patterns with both the NO probe and
the sampled amine provided in all cases an overall consistent
picture of their ion chemistry.
Oxidation of Gaseous Amines by [(TPFPP)•+FeIVdO]+
Ions. When ions 1 obtained by ESI are conveyed into the FT-
ICR cell, mass selected, and allowed to react with a stationary
9
3212 J. AM. CHEM. SOC. VOL. 130, NO. 10, 2008

The Oxidation of Tertiary Amines
A R T I C L E S
Table 1. Kinetic Data for the Gas-Phase Reaction of
[(TPFPP)•+FeIVdO]+ with Tertiary Amines (A)a
product distributione (%)
c
A (IE)b
kexp
φd
ET
HT
OAT
(CH3)3N (7.85)
0.44
1.4
1.1
0.3
1.2
0.8
0.7
0.7
0.3
1.0
4.5
14
8.9
2.7
11
6.0
5.3
5.5
2.1
7.0
100
85
70
100
40
67
45
27
48
(C2H5)3N (7.53)
15
30
c-C4H8N-C6H5f (6.8)
c-C5H10NHg (8.03)
p-CH3-C6H4N(CH3)2 (6.93)
C6H5N(CH3)2 (7.12)
p-F-C6H4N(CH3)2 (7.5)
p-Br-C6H4N(CH3)2 (7.3)
p-CF3-C6H4N(CH3)2 (n.a.)
C6F5N(CH3)2 (8.5, vertical)
44
33
55
55
10
16
18
42
100
a Reactions were run at least in triplicate at the temperature of the FT-
ICR cell of 300 K, and averaged results are presented. Piperidine, a
secondary amine, is also included. bIonization energies (IE)23 in eV are
given in parentheses (n.a. stands for not available). cSecond-order rate
constants in units of 10-10 cm3 molecule-1 s-1. The estimated error is (30%,
while the internal consistency of the data is within (10%. dReaction
efficiency, φ ) kexp/kcoll × 100. Collision rate constants (kcoll) evaluated
with the parametrized trajectory theory.16 eProduct branching resulting from
electron transfer (ET), hydride loss (HT), and oxygen atom transfer (OAT)
Figure 3. Time dependence of relative ion intensities (I%) following the
selection of 1 in the FT-ICR cell and allowing it to react with p-CF3-
C6H4N(CH3)2 (A) at 2.3 × 10-8 mbar.
grounds. An alternative possibility that the amine radical cation,
formed by ET and released free in the gas phase, may be a
precursor of [A(sH)]+ ions (paths a f b, in eq 6) is in fact
hardly compatible with an appearance energy of the iminium
ion that is substantially higher than the IE value of the parent
amine.23,24 This conclusion obviously holds also in those cases,
as in the trimethylamine reaction, where the release of an A•+
product is not accessed and the formation of the iminium ion
(path 6c) may still be thermochemically allowed within the ion
neutral complex 2 (eq 6), driven by the concomitant formation
of neutral (TPFPP)FeIII(OH). Therefore, the point that [A(s
H)]+ product ions do not proceed by the formation of free A•+
does not exclude a stepwise mechanism, whereby the same
collision complex (2) accounts for the formation of both A•+
and [A(sH)]+, as further discussed in a next section. However,
f
g
channels. N-Phenylpyrrolidine. Piperidine.
4d). Thus, with reference to Scheme 1, from a first inspection
of this gas-phase reaction, an ET process is clearly discernible
whereas an HAT path is not emerging.
The product distribution and the kinetic data for the reaction
of 1 with amines are summarized in Table 1. The second-order
rate constants for the overall reaction of 1 are reported together
with the reaction efficiencies (φ ) kexp/kcoll × 100), measuring
the fraction of reactive collisions. In other words, φ yields the
percent number of collision events leading to products and is
obtained from the ratio of the experimental rate constant, kexp
,
and the rate for ion-neutral collision, kcoll, calculated by
parametrized trajectory theory.16 The progress of the reaction
is recorded by means of ion abundance versus time plots such
as the one depicted in Figure 3. For the sake of clarity, the
contribution of [(TPFPP+O)FeIII]+ and the ensuing ligand
addition product are omitted. The distinct pathways of eqs 4a-c
appear to be parallel and independent, leading to ET, HT, and
OAT products in ion abundance ratios that remain constant with
elapsed time. The [(TPFPP)FeIII]+ product ion of the OAT
channel, in the presence of neutral A, undergoes an addition
process, conforming to the usual reactivity behavior of porphyrin
iron(III) ions with gaseous amines.
while a high IE prevents the formation of the free amine radical
cation both for aliphatic and aromatic amines, their oxidation
products are different. HT product ions are obtained from the
former and an OAT path is prevailing for the latter.
A View into the Mechanism of Oxidation of N,N-Di-
methylanilines by [(TPFPP)•+FeIVdO]+. The N,N-dimethyl-
anilines represent a useful series of structurally similar com-
pounds whose electronic properties at the reaction center may
be modulated by varying the substituents on the aromatic ring.
Accordingly, such a series have been studied in solution with
regard to their oxidative dealkylation by both monooxygenase
enzymes and model compounds.2,3,5,6a,25 Among the paths
depicted in eq 4, the formation of the aniline based ions, namely
Typically, two or more channels are active for the amines
listed in Table 1. Notable exceptions are trimethylamine and
piperidine (a secondary amine), yielding only an HT product,
and C6F5N(CH3)2 which reacts only by the OAT pathway.
Interestingly, these three amines are the only ones in the table
that do not show any radical cation formation (ET). This finding
appears related to the comparatively higher ionization energies
of these neutrals. The experimental IE values from the NIST
database (also listed in Table 1) seem to indicate an onset for
the formation of the ET product. The amine radical cation
appears in fact only for IE values e7.53 eV, suggesting that
the ET process may otherwise be endothermic. From this
evidence, an approximate lower limit of 7.5 eV for the IE value
of [(TPFPP)FeIVdO] in the gas phase is inferred. The formation
of an HT product by a direct path starting from the reagents, 1,
and A (as described in eq 6, path c) is not only consistent with
the observed kinetics but also conceivable on thermochemical
(23) Hunter, E. P.; Lias, S. G. NIST Chemistry Webbook, NIST Standard
Reference Database Number 69; Linstrom, P. J., Mallard, W. G., Eds.;
National Institute of Standards and Technology: Gaithersburg, MD, 2005;
http://webbook.nist.gov.
(24) Traeger, J. C. J. Phys. Chem. A 2007, 111, 4643.
9
J. AM. CHEM. SOC. VOL. 130, NO. 10, 2008 3213

A R T I C L E S
Chiavarino et al.
Table 2. Kinetic and Computed Thermodynamic Data for the
Reaction of [(TPFPP)•+FeIVdO]+ with N,N-Dimethylanilines
(ArN(CH3)2)
a φET+φHT is the sum of the efficiencies for the formation of the ET and
b
HT products, obtained from data reported in Table 1, at 300 K. BDE (in
Figure 4. Plot of log(φET + φHT) (where φET + φHT is the sum of the
reaction efficiencies for the ET and HT channels) as a function of the IE
for the reactant p-X-substituted N,N-dimethylaniline (X is given for each
datum on the plot).
kJ mol-1) for the R C-H bond in the neutral N,N-dimethylanilines and IE
(eV) of the neutral aniline were determined at B3LYP/6-311+G** level,
including zero point vibrational energy corrections. cExperimental values26
are reported in parentheses.
C6F5N(CH3)2. Indeed, in recent, thorough theoretical studies27
of the R C-H activation of N,N-dimethylanilines with a
Compound I model endowed with an SH axial ligand, it is found
that the hydrogen atom abstraction is the rate-limiting step of
the hydroxylation reaction leading to a carbinolaniline inter-
mediate. These studies, however, provide an interpretation of
the several facets of the enzymatic oxidation of anilines in terms
of a two-state reactivity model.28 In particular, in a series of
para-substituted N,N-dimethylanilines, as the substituent be-
comes more electron withdrawing the reactivity pattern moves
from spin-selective to two-state reactivity. In the presence of a
para-NO2 group, a substantial fraction of the reaction proceeds
via the high spin quartet besides the lower energy low spin
doublet route. While the NO2-substituted aniline could not be
sampled due to its scant volatility, it is possible that the gas-
phase reactivity behavior of C6F5N(CH3)2 toward 1 may
similarly reflect a switch to two-state reactivity. The formation
of a carbinolaniline as the primary hydroxylation product formed
within the collision complex would also be in agreement with
the small binding energy calculated for the association of this
species to the heme in both the doublet and the quartet energy
profile.27 In contrast, the formation of an amine N-oxide is
expected to yield a stable adduct, based on the evidence from
the pyridine reaction likely forming this species, as mentioned
previously. Thus, the formation of an aniline N-oxide as an OAT
product does not seem to be occurring under the present
conditions.1e,29 In this context, recent theoretical studies, aimed
at modeling the generation of active species of cytochrome P450
from different precursors and the ensuing oxidative N-de-
methylation of N,N-dimethylanilines, have investigated the
detailed pathway leading to the high valent oxo iron intermedi-
ate, using N,N-dimethylaniline-N-oxide as an oxygen atom
donor, along the possible different spin state surfaces.30 From
the amine radical cation (4a) and the formal hydride abstraction
product or iminium ion (4b), is more easily accounted for. The
finding that the two paths appear in parallel, with no evidence
of intervening free intermediates, may be explained by a
mechanism initiated by an ET step leading to a complex (2 in
eq 6) that may either dissociate to give the aniline radical cation
or proceed by further reaction to form the iminium ion. Overall
these pathways depend on an ET event and are thus expected
to depend on the IEs of the neutral anilines. Because experi-
mental IEs are available only in part for the tested N,N-
dimethylanilines, computational data have been acquired and
are listed in Table 2 together with the computed values of the
C-H bond dissociation energy (BDE) for the methyl group R
to the amino nitrogen. The first column in Table 2 reports the
sum of the efficiencies for the formation of the ET and HT
products (φET+φHT) which decreases along a trend of increasing
IEs. The relationship between φET+φHT and the computed IE
values is illustrated in the plot of Figure 4 where the data are
fitted by a straight line. As shown in Table 1, the partition into
the two products is not a simple function of the electronic
properties of the anilines, at least as described by their IE values,
although the least oxidizable p-trifluoromethyl compound
(among the ones still reacting by ET/HT) shows the smallest
release of the radical cation product.
As the formation of the amine radical cation becomes
endothermic, which is likely occurring with C6F5N(CH3)2, the
OAT path takes over, making the reaction of this amine a
relatively efficient process (φ ) 7.0, Table 1). If this reaction
were conforming to an ET initiated process, C6F5N(CH3)2,
owning the highest IE, should behave as the least reactive among
the probed N,N-dimethylanilines. However, also invoking an
HAT initiated process, proceeding by a “rebound” step, one
should not expect an enhanced reactivity in view of the similar
BDEs for the R C-H bond in the aniline series,26 as also shown
by the computed data reported in Table 2. So, also this possible
mechanism does not explain the relatively higher reactivity of
(27) (a) Li, C.; Wu, W.; Kumar, D.; Shaik, S. J. Am. Chem. Soc. 2006, 128,
394. (b) Wang, Y.; Kumar, D.; Yang, C.; Han, K.; Shaik, S. J. Phys. Chem.
B 2007, 111, 7700.
(28) (a) Shaik, S.; Hirao, H.; Kumar, D. Acc. Chem. Res. 2007, 40, 532. (b)
Shaik, S.; de Visser, S. P.; Ogliaro, F.; Schwarz, H.; Schro¨der, D. Curr.
Opin. Chem. Biol. 2002, 6, 556. (c) Schro¨der, D.; Shaik, S.; Schwarz, H.
Acc. Chem. Res. 2000, 33, 139. (d) Shaik, S.; Filatov, M.; Schroder, D.;
Schwarz, H. Chem.sEur. J. 1998, 4, 193.
(29) Guengerich, F. P.; Muller-Enoch, D.; Blair, I. A. Mol. Pharmacol. 1986,
30, 287.
(30) Cho, K.-B.; Moreau, Y.; Kumar, D.; Rock, D. A.; Jones, J. P.; Shaik, S.
Chem.sEur. J. 2007, 13, 4103.
(25) (a) Dowers, T. S.; Rock, D. A.; Rock, D. A.; Jones, J. P. J. Am. Chem.
Soc. 2004, 126, 8868. (b) Bhakta, M. N.; Wimalasena, K. Eur. J. Org.
Chem. 2005, 4801.
(26) Dombrowski, G. W.; Dinnocenzo, J. P.; Farid, S.; Goodman, J. L.; Gould,
I. R. J. Org. Chem. 1999, 64, 427-431.
9
3214 J. AM. CHEM. SOC. VOL. 130, NO. 10, 2008

The Oxidation of Tertiary Amines
A R T I C L E S
Table 3. Kinetic Isotope Effects for the HT Path in the Reaction of
[(TPFPP)•+FeIVdO]+ with N,N-Bis(dideuteriomethyl)anilines
(ArN(CHD2)2) and Computed Thermodynamic Data
these data, still regarding the Compound I model endowed with
an SH axial ligand, it may be observed that the OAT reaction
between the Compound I model and N,N-dimethylaniline
leading to N,N-dimethylaniline-N-oxide is exothermic by ca.
24 kJ mol-1. However, the reaction pathway proceeds by a
transition state lying comparatively high in energy, implying
activation energy barriers greater than 50 kJ mol-1 30
. Altogether,
the computational results concur in suggesting that the formation
of an N-oxide as the product of the OAT process is highly
unlikely.
Kinetic Isotope Effects in the Oxidation of N,N-Di-
methylanilines by [(TPFPP)•+FeIVdO]+ Ions. In the several
studies addressing the oxidative N-dealkylation of N,N-di-
methylaniline catalyzed by cytochromes P450 and iron porphy-
rin complexes, the hydrogen/deuterium kinetic isotope effect
(KIE) has largely been exploited as a mechanistic tool.3,5,6,25,27
The same probe is used in the present study to gain an insight
into the gas-phase reaction pathway. In fact, while the observed
ionic products, the amine radical cation and the iminium ions,
testify an ET initiated reaction, the path leading to the iminium
ion (overall, a formal HT product) from the intermediate ion-
neutral complex (2 in eq 7) may be envisioned as either a
concerted hydrogen atom transfer (HAT) or a stepwise proton
transfer (PT) and electron transfer (ET) route. The occurrence
a KIE values obtained according to eq 8; n.a. stands for not available.
bBDE (in kJ mol-1) for the R C-H bond in the radical cation of N,N-
dimethylanilines, PA (in kJ mol-1) for the N-methyl-N-arylaminomethyl
radical (namely the conjugate base of the N,N-dimethylaniline radical cation),
and IE (eV) for the N-methyl-N-arylaminomethyl radical were determined
for the unlabeled compounds at the B3LYP/6-311+G** level, including
zero-point vibrational energy corrections.
groups was preferred, relative to N-methyl N-(trideuteriomethyl)-
anilines, where a masking effect may arise if the rotation of
CH3 and CD3 groups around the C(aromatic)-N bond is slower
than hydrogen or proton transfer, an event found to affect the
reaction occurring within an enzyme pocket.3b,5f When N,N-
bis(dideuteriomethyl)anilines are allowed to react with 1, the
HT path yields two products implying formal H- and D-
abstraction from the amine methyl groups.32 The ion intensity
ratio of the so-formed species, [A(-H)]+ and [A(-D)]+, is time
independent and, corrected by a statistical factor (eq 8), yields
the KIE values reported in Table 3.
I[A(-H)]+
I[A(-D)]+
KIE ) 2 ×
(8)
of a PT event is conceivable in view of the pronounced acidity
of N,N-dimethylaniline radical cations5c and of the basic
properties of the iron(IV)-oxo group.5b,c,31 In particular, the
(TPFPP)FeIVdO ferryl species is estimated to be more basic
than pyridine.5b,c Assuming the same relative basicity to hold
in the gas phase, a PA value g 930 kJ mol-1 is derived.23 The
PA values of the radicals representing the conjugate bases of
the N,N-dimethylaniline radical cations, as obtained by calcula-
tions at the B3LYP/6-311+G** level, provide a homogeneous
set of data listed in Table 3. The N-methyl-N-arylaminomethyl
radicals so obtained by deprotonation of the aniline radical
cations are expected to be easily oxidized to the iminium ion,
in view of their low IE values, also listed in Table 3.
Finally, Table 3 provides also the computed values of the
C-H bond dissociation energy for the methyl group R to the
amino nitrogen in the N,N-dimethylaniline radical cations. The
computed thermochemical data provide a basis to survey the
KIE affecting the formation of the iminium ion. To this end
the reaction of N,N-bis(dideuteriomethyl)anilines was investi-
gated. The use of substrates with partial labeling in both methyl
The KIE values increase as the para substituent changes from
an electron donor (CH3) to an electron withdrawing group (CF3).
The trend is paralleled by the decrease in the PA of the
N-methyl-N-arylaminomethyl radical. Both the magnitude of the
KIE values, indicative of a primary KIE, and their variation
with the PA of the conjugate base of the aniline radical cation
(computed values are listed in Table 3) are consistent with the
operation of a KIE affecting the PT step of eq 7. In fact, based
on the relative PA values, the KIE increases as the PA of the
N-methyl-N-arylaminomethyl radical approaches the estimated
PA of the iron(IV)-oxo group (PA g 930 kJ mol-1) and the PT
reaction approaches thermoneutrality, in agreement with the
Melander-Westheimer KIE model.33 In contrast, in the event
of a concerted HAT reaction one may expect the magnitude of
the KIE values to scale with the BDE(C-H) for the methyl
group R to the amino nitrogen in the aniline radical cation.
However, a clear trend should not result for the measured KIEs
in this case, based on the somewhat uniform values of the
relative BDE(C-H) (Table 3). The KIE data appear then to
(31) (a) Green, M. T.; Dawson, J. H.; Gray, H. B. Science 2004, 304, 1653. (b)
Behan, R. K.; Hoffart, L. M.; Stone, K. L.; Krebs, C.; Green, M. T. J. Am.
Chem. Soc. 2006, 128, 11471. (c) Behan, R. K.; Green, M. T. J. Inorg.
Biochem. 2006, 100, 448. (d) Horner, O.; Mouesca, J.-M.; Solari, P. L.;
Orio, M.; Oddou, J.-L.; Bonville, P.; Jouve, H. M. J. Biol. Inorg. Chem.
2007, 12, 509. (e) Alfonso-Prieto, M.; Borovik, A.; Carpena, X.; Murshu-
dov, G.; Melik-Adamyan, W.; Fita, I.; Rovira, C.; Loewen, P. C. J. Am.
Chem. Soc. 2007, 129, 4193. (f) Rovira, C. ChemPhysChem 2005, 6, 1820.
(g) Silaghi-Dumitrescu, R. J. Biol. Inorg. Chem. 2004, 9, 471.
(32) The possibility of a ring hydrogen being removed from the radical cation
of an aromatic compound, owning alternative acidic sites on a ring
substituent, has been discarded in previous studies using appropriately
labelled compounds. (a) Chiavarino, B.; Crestoni, M. E.; Fornarini, S. Chem.
Phys. Lett. 2003, 372, 183. (b) Baciocchi, E.; Bietti, M.; Chiavarino, B.;
Crestoni, M. E.; Fornarini, S. Chem.sEur. J. 2002, 8, 532.
(33) (a) Westheimer, F. H. Chem. ReV. 1961, 61, 265. (b) Melander, L.; Saunders,
W. H. Reaction Rates of Isotopic Molecules; Krieger Publishing Com-
pany: Malabar, 1987.
9
J. AM. CHEM. SOC. VOL. 130, NO. 10, 2008 3215

A R T I C L E S
Chiavarino et al.
Conclusions
favor a stepwise pathway to the iminium ion, comprising a PT
event followed by the fast oxidation of the so-formed N-methyl-
N-arylaminomethyl radical. In a related context, it may be
relevant to note, however, that HAT reactions are a topic of
intensive study. The reaction of an exemplary hydrogen abstrac-
tor, the tert-butoxyl radical, was found to display rates that are
rather entropy- than enthalpy-controlled for substrates with
relatively low C-H bond dissociation energy.34 Regarding the
Cytochrome P450 catalyzed N-demethylation, the KIE associ-
ated to HAT from neutral N,N-dimethylanilines has been
accounted for in terms of electronic substituent effects on the
reaction barrier and of two-state reactivity.3f,27b In the theoretical
study on a compound I model of Cytochrome P450 it is in fact
shown that both low- and high-spin states may contribute to
the overall reactivity when strongly electron withdrawing
substitutents such as cyano or nitro groups are present on the
aromatic ring.27b
A high-valent iron(IV) oxo porphyrin cation radical complex,
1, has been prepared and characterized in the gas phase,
providing a model for the high valent oxo iron intermediates
commonly referred to as Compound I species in biological
contexts. The gas phase may grant an extreme simplification
relative to the complex environment of the heme group within
the enzyme pocket, which is known to markedly affect the
reactivity of the catalytically active species.37 The mass
spectrometric study of the reactivity of naked 1 with gaseous
amines has thus unveiled elementary pathways that could only
be inferred in condensed phase studies where mechanistic
information typically relied on the analysis of the neutral end
products. The major relevant information thus gained may be
summarized in the following points. The reaction is initiated
by an ET event for the majority of the tested amines which
included tertiary aliphatic and aromatic amines as well as a
cyclic and a secondary amine. For a series of N,N-dimethyl-
anilines the reaction efficiency for the ET activated pathways
was found to correlate with the IE of the base. It is then implied
that the ET process is allowed to occur, at least within the ion
neutral complex formed upon the encounter of 1 with the amine,
in line with the pronounced oxidizing power of 1, which is due
in part to the electron deficient character of the porphyrin
ligand.6b,38 An important factor is also the lack of an axial ligand
in the pentacoordinate complex 1. Indeed, theoretical calcula-
tions have investigated a Compound I model where the heme
is represented by an iron-porphine complex and an SH group
replaced the axial cysteine ligand which is present in P450,
showing that the reaction with N,N-dimethylaniline proceeds
by an HAT mechanism.27 However, the same authors find that
on removing the SH ligand the reaction pathway turns to initial
spontaneous ET from the aniline, followed by proton abstrac-
tion.39 The same stepwise pathway is presently inferred from
the kinetic study of the reaction of naked 1 with a series of
N,N-dimethylanilines. The primary ET event, yielding complex
2, may either release the amine radical cation (A•+) or proceed
by proton transfer from the acidic methyl groups of A•+ to the
iron(IV)-oxo group. Accordingly, the PT process is character-
ized by a KIE that increases as the PA of the conjugate base,
the N-methyl-N-arylaminomethyl radical (A(-H)•), decreases
and the reaction approaches thermoneutrality. The so-formed
radical is finally oxidized to the iminium ion. The reaction
of 1 with C6F5N(CH3)2, the amine owning the highest IE in
the N,N-dimethylaniline series, presents unique reactivity fea-
tures, that may only in part be accounted for by the fact that
the ET route has become thermodynamically unfavorable.
The observed relatively high efficiency of the OAT pathway,
A Parallel with the Gas-Phase Oxidation of Aromatic
Amines by Bare FeO+. In the quest for simplified models for
the high valent oxo iron species in the active sites of heme
enzymes, the most essential model system is the bare FeO+
cation, which has been generated and extensively studied in the
gas phase.35,10f The dissociation of metastable FeO+ complexes
formed with N-methylaniline upon chemical ionization revealed
a dehydration process leading to an imine/Fe+ complex,36
potentially an intermediate to carbinolamine formation in an
aqueous environment. The bimolecular reactivity of FeO+ with
N,N-dimethylaniline was also studied by FT-ICR displaying
remarkably similar facets with respect to the reaction of 1 that
is presently reported.36 The major product ions are in fact the
amine radical cation (70%) and an iminium ion by a formal
hydride transfer reaction (30%). The KIE affecting the latter
process was obtained by the reaction with N-methyl-N-(trideu-
teriomethyl)aniline. The KIE proved to be sensitive to the
presence and electronic properties of a ligand, increasing from
2.0 for bare FeO+ to 2.6 and 3.0 for (L)FeO+ with L ) C6H6
and C6H5CN, respectively, thus coming to a close, though
probably fortuitous, agreement with the KIE displayed by 1,
where the oxo iron unit is embedded in the porphyrin coordina-
tion environment.
However, a unifying feature for the reaction of N,N-
dimethylaniline with both FeO+ and 1 is the ET step as the
initial elementary step in the gas-phase oxidation. The ET step
is permitted to FeO+ by the notably high IE(FeO) of 8.9 eV
and appears to be allowed to 1 as well. Comparably important
is the C-H bond activation leading to the formation of an
iminium ion, a possible central intermediate in the N-dealky-
lation process in the condensed phase.
(34) (a) Tanko, J. M.; Friedline, R.; Suleman, N. K.; Castagnoli, N. J. Am. Chem.
Soc. 2001, 123, 5808. (b) Finn, M.; Friedline, R.; Suleman, N. K.; Wohl,
C. J.; Tanko, J. M. J. Am. Chem. Soc. 2004, 126, 7578.
(37) Watanabe, Y.; Nakajima, H.; Ueno, T. Acc. Chem. Res. 2007, 40, 554.
(38) (a) Goh, Y. Mee; Nam, W. Inorg. Chem. 1999, 38, 914. (b) Dolphin, D.;
Traylor, T. G.; Xie, L. Y. Acc. Chem. Res. 1997, 30, 251.
(35) (a) Jackson, T. C.; Jacobson, D. B.; Freiser, B. S. J. Am. Chem. Soc. 1984,
106, 1252. (b) Schroeder, D.; Schwarz, H. Angew. Chem., Int. Ed. Engl.
1990, 29, 1433. (c) Schroeder, D.; Schwarz, H. Angew. Chem., Int. Ed.
Engl. 1990, 29, 1431. (d) Fiedler, A.; Schroeder, D.; Shaik, S.; Schwarz,
H. J. Am. Chem. Soc. 1994, 116, 10734. (e) Clemmer, D. E.; Chen, Y.-M.;
Khan, F. A.; Armentrout, P. B. J. Phys. Chem. 1994, 98, 6522. (f) Baranov,
V.; Javahery, G.; Hopkinson, A. C.; Bohme, D. K. J. Am. Chem. Soc. 1995,
117, 12801. (g) Schroeder, D.; Schwarz, H.; Clemmer, D. E.; Chen, Y.;
Armentrout, P. B.; Baranov, V.; Bohme, D. K. Int. J. Mass Spectrom. Ion
Processes 1997, 161, 175. (h) Broenstrup, M.; Schroeder, D.; Schwarz, H.
Chem.sEur. J. 1999, 5, 1176.
(39) Supplementary material in ref 30.
(40) de Visser, S. P.; Shaik, S.; Sharma, P. K.; Kumar, D.; Thiel, W. J. Am.
Chem. Soc. 2003, 125, 15779.
(41) (a) Newcomb, M.; Toy, P. H. Acc. Chem. Res. 2000, 33, 449. (b) Pan, Z.;
Zhang, R.; Newcomb, M. J. Inorg. Biochem. 2006, 100, 524. (c)
Harischandra, D. N.; Zhang, R.; Newcomb, M. J. Am. Chem. Soc. 2005,
127, 13776. (d) Dey, A.; Ghosh, A. J. Am. Chem. Soc. 2002, 124, 3206.
(e) Koppenol, W. H. J. Am. Chem. Soc. 2007, 129, 9686. (f) Collman, J.
P.; Chien, A. S.; Eberspacher, T. A.; Brauman, J. I. J. Am. Chem. Soc.
2000, 122, 11098. (g) Song, W. J.; Ryu, Y. O.; Song, R.; Nam, W. J. Biol.
Inorg. Chem. 2005, 10, 294.
(36) Broenstrup, M.; Schroeder, D.; Schwarz, H. Can. J. Chem. 1999, 77, 774.
9
3216 J. AM. CHEM. SOC. VOL. 130, NO. 10, 2008

The Oxidation of Tertiary Amines
A R T I C L E S
forming a carbinolamine product, may rather suggest the
incursion of a mechanism along a route of different spin
multiplicity.
It is worth noting, however, that the mechanistic pathway
displayed by 1 in the activation of gaseous amines may
constitute a representation of what happens in enzymes like
horseradish and lignin peroxidase,2d,5c,d where the Compound I
species is a much better electron acceptor than the corresponding
intermediate of P450.40
their temporal evolution, may provide an ultimate archetype
against which the intriguingly complex behavior of Compound
I intermediates and model complexes in condensed phases1-6,41
may be fruitfully compared.8
Acknowledgment. This project was supported by the Uni-
versita` di Roma “La Sapienza” and by the Italian Ministero
dell’Universita` e della Ricerca.
In conclusion, the oxidation of amines by 1 studied in the
gas phase, by viewing directly the charged intermediates and
JA077286T
9
J. AM. CHEM. SOC. VOL. 130, NO. 10, 2008 3217