α,2-, α,3-, and α,4-dehydrophenol radical anions: Formation, reactivity, and energetics leading to the heats of formation of α,2-, α,3-, and α,4-oxocyclohexadienylidene

Article abstract of DOI:10.1021/ja029571c

We have regiospecifically generated the α2-, α3-, and α4-dehydrophenoxide anions by collisional activation of o-, m-, and p-nitrobenzoate. The α,2 and α,4 isomers also were synthesized by reacting o-benzyne radical anion with carbon dioxide and electron i

Full text of DOI:10.1021/ja029571c

Published on Web 03/21/2003
r,2-, r,3-, and r,4-Dehydrophenol Radical Anions: Formation,
Reactivity, and Energetics Leading to the Heats of Formation
of r,2-, r,3-, and r,4-Oxocyclohexadienylidene
Dana R. Reed,† Michael C. Hare,†, Alireza Fattahi,† Gyusung Chung,*,‡
Mark S. Gordon,*,§ and Steven R. Kass*,†
Contribution from the Department of Chemistry, UniVersity of Minnesota,
Minneapolis, Minnesota 55455, Department of Chemistry, Konyang UniVersity, Chungnam,
320-711, Korea, and Department of Chemistry, Iowa State UniVersity, Ames, Iowa 50011-2030
Received December 3, 2002; E-mail: kass@chem.umn.edu
Abstract: We have regiospecifically generated the R,2-, R,3-, and R,4-dehydrophenoxide anions by
collisional activation of o-, m-, and p-nitrobenzoate. The R,2 and R,4 isomers also were synthesized by
reacting o-benzyne radical anion with carbon dioxide and electron ionization of p-diazophenol. All three
dehydrophenol radical anions were differentiated from each other and identified by probing their chemical
reactivity with several reagents. Each isomer was converted to phenoxide and its corresponding quinone
as well. Thermochemical measurements were carried out on all three radical anions and their hydrogen-
atom affinities, proton affinities, and electron binding energies are reported. These measured quantities
are combined in thermodynamic cycles to derive the heats of formation of each of the radical anions and
their corresponding carbenes (i.e., R,2-, R,3-, and R,4-dehydrophenol). These results are compared to
MCQDPT2, G3, G2+(MP2), and B3LYP calculations and experimental data for appropriate reference
compounds.
Introduction
thermally and photochemically from 4-diazo-2,5-cyclohexadi-
enone.3 It was found to be a ground-state triplet on the basis of
Carbenes are formed in many useful chemical transformations
and have been the subject of extensive experimental investiga-
tions.1 They also provide a demanding testing ground for
computational methods and, consequently, have been the focus
of a large number of theoretical studies. Diazo precursors of
p-oxocyclohexadienylidene (1p) and its derivatives (i.e., quino-
nediazides) are particularly important compounds as they are
used in technological processes such as photolithography, the
stabilization of rubbers, polyolefins, diesel fuels, and lubricating
oils, and the preservation of foods and fats.2 The parent carbene,
R,4-dehydrophenol (1p), has been generated in solution both
the observed low-temperature ESR signal.3b From an analysis
of the spectrum at 4 K the spin density at the para carbon (C4)
was found to be 0.42-0.45, which is similar to the value for
phenoxyl radical (0.4) at the same site. This led to the suggestion
that 1p is a σ1π1 diradical, which has the π-system of a phenoxyl
radical and the σ-system of phenyl radical. In accord with this
view, Sander and co-workers found that the carbon-oxygen
bond stretch in the matrix-isolated IR spectrum of 1p is at 1496
cm-1 3d
This frequency is about halfway between the C-O
.
single and double bond stretches of phenol (1250 cm-1) and
p-quinone (1682 cm-1) and leads to a bond order of ∼1.5.
* To whom correspondence should be addressed.
† University of Minnesota.
‡ Konyang University.
§ Iowa State University.
Current address: Department of Chemistry, Ohio University, Athens,
OH 45701.
(1) (a) Kirmse, W. In Carbene Chemistry, 2nd ed.; Academic: New York,
1971; p 622. (b) Diradicals; Borden, W. T., Ed.; Wiley: New York, 1982;
p 343. (c) Kinetics and Spectroscopy of Carbenes and Biradicals; Platz,
M. S. Ed.; Plenum: New York, 1990; p 372. (d) AdVances in Carbene
Chemistry; Brinker, U. H., Ed.; JAI Press: Greenwich, 2001; Vol. 3, p
322. (e) Carbenes; Moss, R. A., Jones, M., Jr., Eds.; Wiley: New York,
1975; Vol. 2, p 373. (f) Platz, M., Ed. Tetrahedron 1985, 41, 1423-2036.
(g) Carbene Chemistry: From Fleeting Intermediate to Powerful Reagents;
Bertrand, G., Ed.; Marcel Dekker: New York, 2002; p 301. (h) Houben-
Weyl Methods of Organic Chemistry, Regitz, M.; Giese, B. Eds.; Verlag
Georg Thieme: Stuttgart, 1989; Vol. E 19, Low-Valence Carbon Com-
pounds, Part B, Carbenes (Carbenoids); p 2276. (i) Nefedov, O. M. In
Chemistry of Carbenes and Small-Sized Cyclic Compounds; Mir: Moscow,
1989; p 312.
Recently, multiconfiguration self-consistent field (MCSCF)
calculations have been carried out on R,2-, R,3-, and R,4-
(3) (a) Dewar, M. J. S.; Narayanaswami, K. J. Am. Chem. Soc. 1964, 86, 2422-
2427. (b) Wasserman, E.; Murray, R. W. J. Am. Chem. Soc. 1964, 86,
4203-4204. (c) Sander, W. W. J. Org. Chem. 1988, 53, 2091-2093. (d)
Sander, W.; Muller, W.; Sustmann, R. Angew. Chem., Int. Ed. Engl. 1988,
27, 572-574. (e) Sander, W.; Bucher, G.; Reichel, F.; Cremer, D. J. Am.
Chem. Soc. 1991, 113, 5311-5322. (f) Bucher, G.; Sander, W. J. Am. Chem.
Soc. 1992, 57, 1346-1351. (g) Sander, W.; Bucher, G.; Wandel, H.; Kraka,
E.; Cremer, D.; Sheldrick, W. S. J. Am. Chem. Soc. 1997, 119, 10 660-
10 672. (h) Sander, W.; Hubert, R.; Kraka, E.; Grafenstein, J.; Cremer, D.
Chem Eur. J. 2000, 6, 4567-4579.
(2) Ershov, V. V.; Nikiforov, G. A.; de Jonge, C. R. H. I. Quinonediazides;
Elsevier: Amsterdam, 1981; p 301.
9
10.1021/ja029571c CCC: $25.00 © 2003 American Chemical Society
J. AM. CHEM. SOC. 2003, 125, 4643-4651
4643

A R T I C L E S
Reed et al.
oxocyclohexadienylidene (1o, 1m, and 1p).4 The ground
electronic states of the R,2 and R,4 isomers were found to be
open-shell triplets with singlet-triplet (S-T) gaps of 8.7 and
10.6 kcal mol-1, respectively, using multiconfiguration quasi-
degenerate second-order perturbation theory (MQDPT2/6-
311+G(d)//MCSCF(8,8)/6-31+G(d)). In contrast, the R,3 or
meta isomer 1m is predicted to be a ground-state singlet with
a small S-T gap of 0.6 kcal mol-1. All three isomers are
structurally similar and have short carbon-oxygen bonds
(1.216-1.230 Å) and distinct carbon-carbon bond lengths
spanning a range of 0.089-0.137 Å. These geometries suggest
that 1o, 1m, and 1p are best described as quinoid-type biradicals.
From an energetic point of view, the relative ground-state
energies of these carbenes are predicted to be nearly the same
(1.1 (1p), 0.0 (1m), and 2.9 (1o) kcal mol-1).5 Based upon an
estimate for the heat of formation of 1H-bicyclo[3.1.0]hexa-
3,5-dien-2-one (1b) by Sander et al. (57 kcal mol-1),3e the
computed energy difference between this compound and 1p
reported by Sole et al. (19 kcal mol-1),4a and the relative
energies of 1o, 1m, and 1p, one can obtain 40, 37, and 38 kcal
mol-1 for the respective heats of formation of these carbenes.
These estimates are markedly different from the value obtained
by assuming bond additivity (75 (1 kcal mol-1) and indicate
that something is amiss. No experimental data, however, are
available to determine where the error(s) lies.
An electron can serve as a protecting group, and as such,
gas-phase negative ion chemistry can provide thermodynamic
information on neutral, uncharged, species.6 This approach is
particularly valuable for obtaining the energetics of transient
molecules because it is difficult to do this by other methods.
Radical anions are needed to obtain information on carbenes
and biradicals, and we recently described a general method for
their preparation.6a In the process of developing this methodol-
ogy, we rediscovered that the collision-induced fragmentation
of o-, m-, and p-nitrobenzoate (2) leads to the sequential loss
of carbon dioxide and nitric oxide.7 This sequence violates the
“even-electron rule” and affords radical anions with mass-to-
charge ratios of 92. We now report on the formation and
structural authentication of R,2-, R,3-, and R,4-dehydrophenol
radical anions (3o, 3m, and 3p) along with energetic measure-
ments that lead to a determination of the heats of formation of
the corresponding biradicals. G3, G2+(MP2), B3LYP, and
MCQDPT2 calculations also were carried out to compare with
the experimental results.
Figure 1. (a) Deprotonation of p-nitrobenzoic acid with F-; p-O2NC6H4-
CO2 (2p, m/z 166). (b) Formation of p-nitrophenide (m/z 122) via 5-6
eV sustained off-resonance irradiation of 2p; the ion at m/z 167 is the
residual 13C isotopomer of p-nitrobenzoate. (c) Immediate application of a
second 2-3 eV SORI excitation causing the loss of nitric oxide and the
formation of p-dehydrophenoxide radical anion (3p, m/z 92). (d) Isolation
of 3p for further study via ion-molecule reactions.
-
been described.6a Briefly, our instrument is equipped with a dual cell
inside the bore of a 3-T superconducting magnet and is outfitted with
an Ultrasource electrostatic ion guide interfaced to an Analytica
electrospray ionization (ESI) source. Ions were initially generated by
electron ionization or ESI and the desired species were isolated by using
stored-waveform inverse Fourier transform (SWIFT) and/or chirp
excitations.8,9 Translational and vibrational cooling of the ions was
carried out by collisions with argon, which was pulsed into the
appropriate cell up to pressures of ∼1 × 10-4 Torr. Neutral reagents
were added via pulsed or slow leak valves, and all reactions were
monitored as a function of time.
Generation of r,2-, r,3-, and r,4-Dehydrophenoxide Radical
Anions by Collision-Induced Dissociation of o-, m-, and p-Nitroben-
zoate. o-, m-, and p-Nitrobenzoic acid were purchased from Aldrich
Chemical Co. and used without further purification. In separate
experiments each of these acids was admitted into the source cell of
the FTMS via the solids inlet probe. Upon heating to 100-120 °C
each acid had sufficient vapor pressure to react with fluoride ion to
afford an abundant signal of its conjugate base (Figure 1). The resulting
carboxylate anions were transferred to the analyzer cell and fragmented
in the presence of argon (∼10-5 Torr) by sustained off-resonance
irradiation (SORI)10 at excitation energies between 5 and 6 eV. Each
newly produced nitrophenide ion was subjected to a second SORI pulse
at 2-3 eV to bring about the loss of nitric oxide. The energy for the
latter fragmentation was carefully optimized in each experiment because
insufficient fragmentation takes place if the energy is too low and nitrite
Experimental Section
Gas-phase experiments were carried out with a Finnigan model 2001
Fourier transform mass spectrometer (FTMS), which previously has
(4) (a) Sole, A.; Olivella, S.; Bofill, J. M.; Anglada, J. M. J. Phys. Chem.
1995, 99, 5934-5944. (b) Chung, C.; Pak, M. V.; Reed, D. R.; Kass, S.
R.; Gordon, M. S. J. Phys. Chem. A 2000, 104, 11 822-11 828.
(5) These energies were computed at the MCQDPT/6-11+G(d) level as in
ref 4b but include a zero-point energy and temperature correction to 298
K using unscaled MCSCF/6-31G(d) vibrational frequencies. Such adjust-
ments were not made for the previously reported energetics.
(6) (a) Reed, D. R.; Hare, M. C.; Kass, S. R. J. Am. Chem. Soc. 2000, 122,
10 689-10 696. (b) Broadus, K. M.; Kass, S. R. J. Am. Chem. Soc. 2000,
122, 10 697-10 703. (c) Broadus, K. M.; Kass, S. R.; Osswald, T.;
Prinzbach, H. J. Am. Chem. Soc. 2000, 122, 10 964-10 968. (d) Broadus,
K. M.; Kass, S. R. J. Am. Chem. Soc. 2001, 123, 4189-4196. (e) Reed, D.
R.; Kass, S. R.; Mondanaro, K. R.; Dailey, W. P. J. Am. Chem. Soc. 2002,
124, 2790-2795. (f) Glasovac, Z.; Eckert-Maksic, M.; Dacres, J. E.; Kass,
S. R. J. Chem. Soc., Perkin Trans 2, 2002, 410-415. (g) Broadus, K. M.;
Kass, S. R. J. Phys. Org. Chem. 2002, 15, 461-468.
(8) Marshall, A. G.; Roe, D. C. J. Chem. Phys. 1980, 73, 1581-1590.
(9) Wang, T. C. L.; Ricca, T. L.; Marshall, A. G. Anal. Chem. 1986, 58, 2935-
2938.
(7) Hunt, D. F.; Shabanowitz, J.; Giordani, A. B. Anal. Chem. 1980, 52, 386-
390.
(10) Gauthier, J. W.; Trautamn, T. R.; Jacobson, D. B. Anal. Chim. Acta 1991,
246, 211-225.
9
4644 J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003

R,2-, R,3-, and R,4-Dehydrophenol Radical Anions
A R T I C L E S
anion (NO2-) is formed at higher energies. Isolation of the resulting
C6H4O•- (m/z 92) ions were carried out by a combination of SWIFT
and chirp excitation pulses,8,9 and these ions were subsequently cooled
with argon pulsed to a pressure of ∼10-4 Torr. In some experiments,
the carboxylate anions were prepared by ESI of the conjugate acid from
a 1:1 methanol-water solution (v/v) containing 0.1% ammonia.
Scheme 1
Generation of r,4-Dehydrophenoxide Radical Anion by Dis-
sociative Electron Capture of 4-Diazo-2,5-cyclohexadienone. 4-Diazo-
2,5-cyclohexadienone was prepared by literature methods and stored
at -40 °C in the dark.11 The wet diazide was dried in the sample lock
of the solids probe at ∼10-2 Torr and then was introduced into the
high vacuum system. After briefly heating the sample to 40 °C to drive
off residual water, it was ionized with 6 eV electrons to afford a
C6H4O•- anion at m/z 92. This ion was subsequently transferred to the
analyzer cell where it was cooled with a pulse of argon (∼10-4 Torr)
and isolated for further investigation.
Generation of r,2-Dehydrophenoxide Radical Anion by Oxygen
Atom Transfer to o-Benzyne Radical Anion. Atomic oxygen radical
anion was generated by dissociative electron impact on nitrous oxide
and then was reacted with benzene to afford o-benzyne radical anion.12
This ion was allowed to react with a static pressure of carbon dioxide
(∼10-7 Torr) in the source cell and at these low pressures oxygen atom
transfer takes place in preference to adduct formation. The resulting
R,2-dehydrophenoxide radical anion was transferred to the analyzer
cell where it was cooled with argon (∼10-4 Torr) and isolated for further
investigation.
include ZPE corrections and are at 0 K (electron affinities) or 298 K
(proton affinities and bond dissociation energies).
In an attempt to obtain reliable energetic data and to assess several
popular computational approaches, B3LYP/6-311+G(3df,2p)//MP2/6-
31+G(d), G2+(MP2),16 and G317 calculations were carried out. The
second of these methods effectively corresponds to a QCISD(T)/6-
311+G(3df,2p) energy but was obtained on a MP2(FC)/6-31+G(d)
geometry as follows:
Structure Authentication. Each of the C6H4O•- isomers was
converted into phenoxide ion by reaction with tert-butyl mercaptan or
diethylhydroxylamine. The resulting hydrogen atom transfer product
was bracketed with standard reference acids and further characterized
by collision-induced dissociation (CID). These results were compared
to control experiments performed on authentic phenoxide ion generated
by fluoride anion deprotonation of phenol. The R,2- and R,4-isomers
also were derivatized to the corresponding quinone radical anions by
reaction with sulfur dioxide. These latter species were characterized
by bracketing their electron binding energies and comparing the results
to the known electron affinities of o- and p-quinone.
Computations. All calculations were carried out using GAMESS13
and Gaussian 9814 or earlier versions of the program on SGI and IBM
workstations. Geometries were optimized at the HF and MP2 levels
with the 6-31G(d) and 6-31+G(d) basis sets. Closed-shell systems were
treated by using restricted wave functions and open-shell species were
handled by using the unrestricted formalism. In both cases, analytical
vibrational frequencies were computed at both theoretical levels for
each stationary point. Zero-point energies and temperature corrections
to 298 K were computed by scaling the vibrational frequencies using
the following factors: 0.9135 (HF, ZPE), 0.8929 (HF, ν’s), 0.9646
(MP2, ZPE), and 0.9427 (MP2, ν’s), where ZPE refers to the vibrational
zero-point energy correction.15 All of the cited energies in this work
E[G2+(MP2)] ) E[QCISD(T)/6-311+G(d,p)] + E[MP2/6-311+
G(3df,2p)] - E[MP2/ 6-311+G(d,p)] + hlc + ZPE + TC (1)
where hlc is the high level correction term for G2 theory and TC is a
temperature correction to 298 K. Hartree-Fock ZPEs and vibrational
frequencies for the temperature corrections were used in each case,
but in some instances there are significant differences (g1.0 kcal mol-1
)
with the MP2 corrections. In these situations both sets of results are
given.
As the species of interest in this work are inherently multiconfigu-
rational, MCSCF geometries and Hessians also were computed with
the 6-31G(d) and 6-31+G(d) basis sets.18 To account for some of the
dynamic correlation in these species, single-point calculations were
carried out at the MCQDPT2 level with the 6-311+G(d) basis set. All
of the resulting energies were ZPE corrected and adjusted to 298 K,
except for the electron affinities which are at 0 K, by using unscaled
MCSCF vibrational frequencies.
Results and Discussion
Deprotonation of o-, m-, and p-nitrobenzoic acid by fluoride
ion in the gas phase affords the corresponding carboxylate anions
(2), which lose carbon dioxide upon 5-6 eV SORI-CID.
Further fragmentation of the phenide isomers at 2-3 eV yields
the desired R,2-, R,3-, and R,4-dehydrophenoxide radical anions
(Scheme 1 and Figure 1). These cleavages previously have been
reported in the analytical chemistry literature,7 but the structures
and reactivities of the resulting C6H4O•- ions were not explored.
To establish the structures of 3o, 3m, and 3p, their reactivities
were examined. It was anticipated that these radical anions
should have the σ-system of a phenyl radical and the π-system
of a phenoxide ion and would display radical reactivity similar
to related distonic radical cations containing an aryl radical
moiety. In accord with these expectations, all three isomers
(11) Puza, M.; Doetschman, D. Synthesis 1971, 481-488.
(12) Guo, Y.; Grabowski, J. J. J. Am. Chem. Soc. 1991, 113, 5923-5931.
(13) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M.
A.; Cheeseman, J. R.; Zakrewski, V. G.; Montgomery, J. A. Jr.; Stratmann,
R. E.; Burant, J. C.; Dapprich, S.; Millam, J. M.; Daniels, A. D.; Kudin,
K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone, V.; Cossi, M.; Cammi,
R.; Mennucci, B.; Pomelli, C.; Adamo, C.; Clifford, S.; Ochterski, J.;
Petersson, G. A.; Ayala, P. Y.; Cui, Q.; Morokuma, K.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Cioslowski, J.; Oritz,
J. V.; Baboul, A. G.; Stefanov, B. B.; Lui, G.; Liashenko, A.; Piskorz, P.;
Komaromi, I.; Gomperts, R.; Martin, R. L.; Fox, D. J.; Ketih, T.; Al-Laham,
M. A.; Peng, C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.;
Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.;
Gonzalez, C.; Head-Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian
98, ReVisions A.6-9, Gaussian, Inc., Pittsburgh, PA, 1998.
(14) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.; Gordon, M.
S.; Jensen, J. H.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.; Su, S. J.;
Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem. 1993,
14, 1347-1363.
(15) Pople, J. A.; Scott, A. P.; Wong, M. W.; Radom, L. Isr. J. Chem. 1993,
33, 345-350.
(16) (a) Curtiss, L. A.; Raghavachari, K.; Pople, J. A. J. Chem. Phys. 1993, 98,
1293-1298. (b) Gronert, S. J. Am. Chem. Soc. 1993, 115, 10 258-10 266.
(17) Curtiss, L. A.; Raghavachari, K.; Redfern, P. C.; Rassolov, V.; Pople, J.
A. J. Chem. Phys. 1998, 109, 7764-7776.
(18) For more details, see ref 4b.
9
J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003 4645

A R T I C L E S
Reed et al.
Table 1. Reaction Products of 3o, 3m, and 3p with Probe Reagentsa
observed products
reagent
MeSSMe
3o
3m
3p
o-MeSC6H4O- (139)
o-ClC6H4O- (127, 66%),
CCl3- (117, 33%)
m-MeSC6H4O- (139)
p-MeSC6H4O- (139)
p-ClC6H4O- (127, 66%),
CCl3- (117, 33%)
CCl4
m-ClC6H4O- (127, 60%),
CCl3- (117, 40%)
NO
o-ONC6H4O- (122)
m-ONC6H4O- (122, 70%),
m-OC6H4O•- (108, 3%), CN-
(26, 6%), NO2- (46, 15%),
C4H4N- (66, 7%)
p-ONC6H4O- (122)
SO2
o-OC6H4O•- (108)
m-OC6H4O•- (108, ∼33%),
m-OC6H4SO2•- (156, ∼66%)b
NO2- (46, 40%),
p-OC6H4O•- (108)
NO2
NO2- (46, 58%),
NO2- (46, 48%),
o-OC6H4O•- (108, 38%),
C6H4NO2- (122, 4%)
o-SC6H4O•- (124)
m-OC6H4O•- (108, 50%),
C5H4O- (80, 10%)
p-OC6H4O•- (108, 48%),
C6H4NO2- (122, 4%)
p-SC6H4O•- (124, ∼66%),
p-OC6H4COS•- (152, ∼33%)
t-BuS- (89, 20%),
COS
m-SC6H4O•- (124, ∼66%),
m-OC6H4COS•- (152, ∼33%)
t-BuS- (89, 35%), C6H5O-
(93, 60%), m-HSC6H4O- (125, 5%)
t-BuSH
t-BuS- (89, 20%),
C6H5O- (93, 70%),
o-SC6H4OH- (125, 10%)
C8H6O•- (118, 50%),
C4HO3- (97, 20%),
C9H5O2- (145, 20%),
C9H4O2•- (146, 10%)
C6H5O- (93, 75%),
p-HSC6H4O- (125, 5%)
C8H6O•- (118, 45%),
C4HO3- (97, 25%),
maleic anhydride
C9H5O2- (145, 80%),
C9H4O2•- (146, 20%)
C9H5O2- (145, 20%),
C9H4O2•- (146, 10%)
p-Me3SiC6H4O- (165)
C12H8O3•- (200, 50%),
C11H8O2•- (172, 50%)
Me3SiSiMe3
p-benzoquinone
C12H6O2•- (182, 33%),
C12H7O2- (183, 66%)
C12H8O3•- (200, 50%),
C11H8O2•- (172, 50%)
a Values in parentheses correspond to the mass-to-charge ratio of the product ion followed by its percent abundance. b The product ratio was obtained at
1.1 × 10-7 Torr of SO2 and is sensitive to this pressure.
Scheme 2
undergo a number of atom and group transfers with a variety
of reagents (Table 1). This type of behavior has been observed
with radicals in condensed media and with other radical anions
and cations in the gas phase.6a,6d,19
Structural Verification. All three dehydrophenol radical
anions can be differentiated on the basis of their qualitative
reactivity with the probe reagents listed in Table 1. The R,3
isomer is the easiest one to differentiate in the group. For
example, it reacts ∼50 times slower with sulfur dioxide than
the R,2 and R,4 radical anions and gives two product ions as
opposed to one (i.e., oxygen-atom transfer and an adduct vs
only the former reaction channel). The reactivity of 3m with
maleic anhydride also is distinct in that no (adduct - CO -
CO2)•- (m/z 118) is observed from the R,3 isomer, whereas this
is the major product ion from 3o and 3p. Reaction of
350.4 ( 0.6 kcal mol-1 for phenoxide anion.20 Likewise, all of
the CID spectra show that C5H5 is the dominant fragment at
p-benzoquinone with the three dehydrophenol radical anions
-
serves to differentiate 3o from 3m and 3p. In particular, 3o gives
(adduct-HO)- and (adduct-H2O)•- ions (m/z 183 and 182,
respectively) whereas the other two isomers do not.
-
low energies (∼50%) and that smaller amounts of C3H3
(∼35%), HCtCO- (<10%), and HCtC- (<10%) are formed.
At higher energies, the cyclopentadienide fragment decreases
in intensity relative to ketene enolate. These results preclude
ring-opened structures for 3o, 3m, and 3p and, given their
unique reactivity, the three isomers cannot be interconverting
between each other.
To further characterize the structures of 3o, 3m, and 3p, they
were converted to phenoxide ion (m/z 92) upon reaction with
tert-butyl mercaptan or diethylhydroxylamine (Scheme 2). In
each case the identity of the phenoxide ion was verified by
bracketing its proton affinity and fragmenting it by CID. All of
the derivatized dehydrophenoxide radical anions and indepen-
dently prepared phenoxide, generated by deprotonating phenol
with F-, displayed identical behavior. In particular, these ions
As a final structural confirmation to prove the structures of
3o, 3m, and 3p beyond a reasonable doubt, they were converted
to their corresponding benzoquinones upon reaction with sulfur
dioxide or nitrogen dioxide (Scheme 2). The resulting benzo-
quinone radical anions (4o, 4m, and 4p) were characterized by
deprotonate acetic acid (∆H°acid ) 348.1 ( 2.2 kcal mol-1
)
but not 2-methyl-2-propanethiol (∆H°acid ) 352.5 ( 2.2 kcal
mol-1), which is in accord with the literature proton affinity of
(20) All thermodynamic data unless otherwise noted comes from: Bartmess, J.
E. NIST Chemistry WebBook, NIST Standard Reference Database Number
69, W. G. Mallard and P. J. Linstrom (Eds.). In Secondary NIST Chemistry
WebBook, NIST Standard Reference Database Number 69, W. G. Mallard
and P. J. Linstrom (Eds.), National Institute of Standards and Technology,
Gaithersburg, MD 20899 (http://webbook.nist.gov).
(19) (a) Heidbrink, J. L.; Ramirez-Arizmendi, L.; Thoen, K. K.; Guler, L.;
Kenttamaa, H. I. J. Phys. Chem. A. 2001, 105, 7875-7884. (b) Kochi, J.
In Free Radicals; Wiley: New York, 1973.
9
4646 J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003

R,2-, R,3-, and R,4-Dehydrophenol Radical Anions
A R T I C L E S
Table 2. Electron Affinity Bracketing Results for 1o, 1m, and 1p
found that the reactions with 1,4-dinitrobenzene are inefficient
(k ∼ 10-11 cm3 molecule-1 s-1) and should be considered a
“no” in terms of bracketing the electron affinities. Electron trans-
fer also was observed with ozone, but the reaction rates were
not determined because of the difficulty in making these meas-
urements. Consequently, the fast reaction (∼10-9 cm3 molecule-1
s-1) with nitrogen dioxide (EA ) 2.273 ( 0.005 eV) and the
slow process with 1,4-dintrobenzene (2.00 ( 0.10 eV) are used
to assign EA(1o) ) EA(1m) ) EA(1p) ) 2.14 ( 0.17 eV.
To verify our measured electron affinities, a kinetic deter-
mination using an approach developed by Squires and co-
workers was carried out.21 The collision-induced dissociation
spectra of the conjugate base of 5-nitro-1,3-cyclopentadiene22
and the radical anions of 2,5-difluoro-1-nitrobenzene and
pentafluoronitrobenzene were recorded as a function of pressure
(∼1 × 10-8-6 × 10-8 Torr) and energy.23 The resulting (M -
NO2-)/NO2- ratios (R) were extrapolated to zero pressure (single
collision conditions) and energy (threshold values) and were
found to be 0.039 ( 0.018, 0.040 ( 0.005, and 12.3 ( 1.7,
respectively. These data were used to construct a calibration
line by plotting the known electron binding energies of the M
- NO2 anions versus the natural logarithm of R,20 and the
following equation was obtained: EA (eV) ) 0.223(ln R) +
2.62, r2 ) 0.96. In the exact same way R was obtained from
o-, m-, and p-nitrophenoxide (R ) 0.189 ( 0.014, 0.170 (
0.028, and 0.190 ( 0.018, respectively) to afford EA(1o) )
EA(1p) ) 2.25 ( 0.15 eV and EA(1m) ) 2.23 ( 0.15 eV.
These values are in good accord with the bracketing results,
are the same as the electron affinity of phenoxyl radical (EA )
2.253 ( 0.006 eV)24a and, as expected, are much larger than
for phenyl radical (EA ) 1.096 ( 0.006 eV).24b There also is
good accord with the MCQDPT2 electron affinities which span
a narrow range of 0.1 eV and, after correcting for the error in
the computed value for phenoxyl radical (10.0 kcal mol-1), give
predicted electron affinities of 2.39 eV (1o), 2.40 eV (1m), and
2.28 eV (1p) (Table 3). G3, G2+(MP2), and B3LYP calcula-
tions were carried out as well on 1o, 1m, and 1p and several
well-established reference compounds (Tables 3 and 4) to test
these more commonly employed single-configuration ap-
proaches. The G3 and B3LYP predictions reproduce the EAs
of methoxyl, phenoxyl, and phenyl radicals to within 2.3 kcal
mol-1, but the latter results are systematically too small by 1-2
kcal mol-1. G2+(MP2) fares less well, missing the EA's of
electron transfer
reagent
EA, eV
3o
3m
3p
SO2
1.107 ( 0.008
no
no
no
no
no
no
no
no
no
no
no
no
3-trifluoromethylnitrobenzene 1.41 ( 0.10
maleic anhydride
3,5-bis(trifluoromethyl)-
nitrobenzene
1.44 ( 0.09
1.79 ( 0.10
1,4-benzoquinone
1.91 ( 0.10,
1.990 ( 0.048
2.00 ( 0.10
2.103 ( 0.004
2.273 ( 0.005
no
no
no
1,4-dinitrobenzene
O3
NO2
noa
yes/nob yes/nob yes/nob
yes yes yes
noa
noa
a Electron transfer is observed, but the reaction is slow (i.e., k ∼ 10-11
cm3 molecule-1 s-1). b Electron transfer is observed, but the reaction rate
is unknown.
their electron binding energies (EBEs); the EBE of an anion is
equivalent to the electron affinity (EA) of its corresponding
neutral species. 3,5-Bis(trifluoromethyl)nitrobenzene (EA ) 1.79
( 0.10 eV) reacts with o-benzoquinone radical anion (4o) via
electron transfer but not with the para isomer, as expected given
the known electron affinities of o- and p-benzoquinone (EA )
1.620 ( 0.048 and 1.91 ( 0.10 or 1.990 ( 0.048 eV,
respectively).20 All three species (3o, 3m, and 3p) undergo
electron transfer with nitrogen dioxide (EA ) 2.2730 ( 0.0050
eV), but while this is the sole ionic product with the ortho and
para isomers, 3m also gives oxygen atom transfer (m-OC6H4O•-,
m/z 108) and an adduct (m-O2NC6H4O•-, m/z 138).
Two additional methods for generating 3o and 3p were used.
Dissociative electron impact on 4-diazo-2,5-cyclohexadienone
(6 eV) leads to the loss of molecular nitrogen and formation of
an ion that behaves identically to 3p prepared by CID of
p-nitrobenzoate (eq 2). We tried to use the same precursor for
the formation of 3o, but the o-quinonediazide is difficult to work
with and the desired ion can be produced by oxygen-atom
transfer to o-benzyne radical anion upon reaction with low
pressures of carbon dioxide (eq 3). As expected, the reactivity
of this ion was found to be identical to 3o produced from
o-nitrobenzoate.
phenoxyl and phenyl radicals by 2.9 and 5.2 kcal mol-1
,
respectively. When applied to 1o, 1m, and 1p, the small spread
is reasonably reproduced by G3 theory, and all three methods
do an adequate job of reproducing the absolute energies, with
errors in the ranges of 1-4.5 kcal mol-1 (G3), 3.7-6.1 kcal
mol-1 (G2+(MP2)), and 1.1-6.1 kcal mol-1 (B3LYP); the cited
values have been corrected for the small computed errors
(0.9-2.9 kcal mol-1) in the electron affinity of phenoxyl radical.
Proton Affinities. Proton affinities of 3o, 3m, and 3p, or
equivalently the acidity of phenoxyl radical at the 2-, 3-, and
Thermochemistry. Electron Affinities. The electron binding
energies of 3o, 3m, and 3p or equivalently the electron affinities
of 1o, 1m, and 1p were determined by reacting these ions with
standard reference compounds and observing the occurrence or
nonoccurrence of electron transfer. All three radical anions
undergo electron transfer upon reaction with 1,4-dinitrobenzene,
ozone, and nitrogen dioxide (Table 2). This process does not
occur with 1,4-benzoquinone, 3,5-bis(trifluoromethyl)nitroben-
zene, and additional compounds with lower electron affinities.
These results suggest that the R,2-, R,3-, and R,4-dehydrophenol
radical anions all have the same electron binding energies, and
we initially assigned EA(1o) ) EA(1m) ) EA(1p) ) 1.96 (
0.10 eV.4b However, upon closer examination of the data it was
(21) (a) Wenthold, P. G.; Hu, J.; Squires, R. R. J. Am. Chem. Soc. 1996, 118,
11 865-11 871. (b) Seburg, R. A.; Squires, R. R. Int. J. Mass Spectrom.
Ion Processes 1997, 167, 541-557.
(22) Methyl nitrate was used instead of ethyl nitrate as this facilitated the
proceedure and gave better yields (∼40% vs 20%). For further details,
see: Kerber, R. C.; Chick, M. J. J. Org. Chem. 1967, 32, 1329-1333.
(23) These experiments were carried out on a Finnigan FTMS which has been
outfitted with an IonSpec OMEGA-2001 data station and vacuum computer.
(24) (a) Gunion, R. F.; Gilles, M. K.; Polak, M. L.; Lineberger, W. C. Int. J.
Mass Spectrom. 1992, 117, 601-620. (b) Raymond, T. M.; Davico, G. E.;
Schwartz, R. L.; Lineberger, W. C. J. Chem. Phys. 2000, 112, 1158-1169.
9
J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003 4647

A R T I C L E S
Reed et al.
Table 3. Comparison of Experimental and Directly Computed Proton Affinities, Electron Affinities, and Bond Dissociation Energies for o-,
m-, and p-Dehydrophenol Radical Anion (3o, 3m, and 3p) and Related Speciesa
B3LYP
G2+(MP2)
G3
MCQDPT2
exptl
EA(X)
1o
51.6 (53.8)
55.3 (57.5)
48.5 (50.7)
60.9 (58.0)
58.4 (55.5)
51.1 (48.2)
53.8 (52.9)
56.8 (55.9)
56.6 (55.7)
45.0 (55.0)
45.4 (55.4)
42.6 (52.6)
51.9 ( 3.5
51.4 ( 3.5
51.9 ( 3.5
1mb
1p
PA(X-)
3o
(C)
(O)
3m
(C)
374.9
346.5
366.7
339.1
374.0
369.4
376.1
368.5
337.4
369.4 ( 1.7
369.4 ( 1.7
369.4 ( 1.7
369.9
342.8 [341.2]
367.8
342.0 [340.3]
365.1
334.1
(O)
3p
(C)
(O)
376.5
347.4 [346.1]
374.8 [374.8]
346.6 [345.3]
369.2
337.4
BDE (HX)
C6H5O• (C-H)
ortho
112.9 (114.5)
111.5 (113.1)
111.4 (113.0)
114.0 (110.6)
112.5 (109.1)
112.2 (108.8)
114.1 (112.5)
113.1 (111.5)
112.3 (110.7)
99.8 (114.4)
96.9 (111.5)
98.0 (112.6)
107.7 ( 3.9
107.3 ( 3.9
107.7 ( 3.9
metab
para
C6H5O- (C-H)
ortho
111.1 (112.8)
[113.2]
106.1 (107.7)
[107.4]
112.6 (114.3)
[115.8]
108.1 (104.7)
[110.2]
109.1 (105.7)
[110.4]
116.1 (112.7)
[119.2]
113.3 (111.7)
108.7 (107.1)
115.4 (113.8)
96.8 (111.4)
93.3 (107.9)
97.4 (112.0)
107.7 ( 1.0
107.7 ( 1.0
107.7 ( 1.0
meta
para
C6H4OH• (O-H)
ortho
84.5 (86.1)
[83.2]
84.4 (86.0)
[82.8]
82.4 (84.0)
[81.1]
86.3 (82.9)
[85.1]
86.7 (83.3)
[85.0]
84.0 (80.6)
[82.8]
68.7 (83.3)
65.9 (80.5)
66.3 (80.9)
metab
para
a All values in kcal mol-1. B3LYP and MCQDPT2 energies are with the 6-311+G(3df,2p) and 6-311+G(d) basis sets, respectively. All energies are ZPE
corrected using HF/ 6-31+G(d), MP2/6-31+G(d) (values in brackets), and MCSCF/6-31G(d) vibrational frequencies. Proton affinities and BDEs also include
a temperature adjustment to 298 K. Parenthetical values have been corrected for the error in the appropriate reference compound (see Table 4). b G2+(MP2),
G3, and B3LYP calculations were carried out using the triplet of R,3-dehydrophenol (1m). Since this is not the ground state at the MCQDPT2 level, these
energies were corrected by the S-T gap (0.6 kcal mol-1).
Table 4. Comparison of Experimental and Computed Proton Affinities, Electron Affinities, and Bond Dissociation Energies for Reference
Compoundsa
B3LYP
G2+(MP2)
G3
MCQDPT2
exptl
EA(X•)
CH3O•
34.8
36.9
35.8
52.9
27.6
36.3 ( 0.1
52.0 ( 0.1
25.3 ( 0.1
C6H5•O•
49.8 [52.9]
23.8 [28.1]
1.7(-2.2)
54.9 [58.1]
30.5 [34.8]
2.9(5.2)
42.0
14.9
10.2(10.4)
C6H5
absolute mean error (max error)
1.2(2.3)
PA(X-)
CH3O-
380.9
348.1
401.7
1.3(2.3)
383.0
348.6
400.0
1.0(1.8)
383.6
349.5
401.3
382.4 ( 0.5
350.4 ( 0.6
401.7 ( 0.5
C6H5-O-
340.5
397.8
6.9(9.9)
C6H5
absolute mean error (max error)
0.8(1.2)
BDE(HX)
CH3OH
C6H5OH
C6H5H
101.9
106.3
106.0
88.9
115.1
105.2 ( 0.7
88.8 ( 0.6
113.5 ( 0.5
84.2 [86.9]
111.9 [116.0]
3.2(4.6)
89.9 [92.6]
116.9 [120.9]
1.9(3.4)
68.7b
98.9b
absolute mean error (max error)
0.8(1.6)
17.4(20.1)
a All values in kcal mol-1 and include a ZPE correction using HF/6-31+G(d), MP2/ 6-31+G(d) (values in brackets) or MCSCF vibrational frequencies.
Proton affinities and BDEs are at 298 K. B3LYP and MCQDPT2 energies are with the 6-311+G(3df,2p) and 6-311+G(d) basis sets, respectively. b The
C6H5O-H and C6H5-H bond energies increase by 8.5 and 8.9 kcal mol-1, respectively, if the ZPEs are omitted.
4-positions, also were examined by the bracketing method.
Proton transfer was observed between all three radical anions
and the N,N-dimethylamide of propiolic acid ((CH3)2NCOCt
CH) and stronger acids, but not with p-fluoroaniline, acetone,
N,N-diethylhydroxylamine, phenylacetylene, and weaker species
(Table 5). Hydrogen atom abstraction takes place as well with
many of the reference acids (HX) that were employed, and this
competing pathway could result in the failure to observe X-
when proton transfer is an exothermic process. This complication
arises because phenoxyl radical has a large electron affinity,
which in most cases is greater than the value for X•. As a result,
the [C6H5O• X-] complex formed upon proton transfer probably
is better described as [C6H5O- X•] and preferentially dissociates
to phenoxide ion and X• rather than phenoxyl radical and X-
(Scheme 3). This potential problem can be addressed by using
reference acids whose conjugate bases have electron affinities
9
4648 J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003

R,2-, R,3-, and R,4-Dehydrophenol Radical Anions
A R T I C L E S
Table 5. Proton Affinity Bracketing Results for 3o, 3m, and 3pa
ortho, meta, and para C-H bond dissociation energies of
phenoxide ion are greater than the O-H bond strengths of
methanol (105.2 ( 0.7 kcal mol-1), ethanol (105.4 ( 1.4 kcal
mol-1), tert-butyl alcohol (106.8 ( 1.0 kcal/mol), and 2,2,2-
trifluoroethanol (107.0 ( 2.5 kcal mol-1).20,24b Ammonia,
benzene, and deuterium oxide (BDEs ) 108.5 ( 0.3, 113.5 (
0.5, and 121.4 ( 0.2 kcal/mol, respectively),25 on the other hand,
do not undergo hydrogen or deuterium atom abstraction with
3o, 3m, and 3p, so we assign HA(3o, 3m, and 3p) ) 107.7 (
1.7 kcal/mol. These values are in reasonable accord with
MCQDPT2 predictions of 111.4 (3o), 107.9 (3m), and 112.0
(3p) kcal mol-1 after correcting for the 14.6 kcal mol-1 error
in the C-H BDE for benzene (Table 3). Alternatively, the
directly computed non-ZPE corrected 0 K bond energies are
105.2, 101.9, and 105.8 kcal mol-1, respectively. This latter
approach is consistent with Wiberg’s finding that hydrocarbon
BDEs are well reproduced at the MP2 level if ZPE corrections
are omitted.26 As for the G3, G2+(MP2), and B3LYP calcula-
tions, they do a better job reproducing the C-H bond strength
for benzene and are in good accord with our experimental results
for 3o and 3m; the computed bond energy for 3p is consistently
somewhat too strong.
3o
3m
3p
reagent
∆H°acid
BDE
PT HT PT HT PT HT
NH3
C6D6
D2O
CH3OH
C6H5CH3
CH3CH2OH
(CH3)3COH
CH3CN
404.3 ( 0.3 108.5 ( 0.3 no no no no no no
401.7 ( 0.5 113.5 ( 0.5 no no no no no no
393.0 ( 0.2 121.4 ( 0.2 no no no no no no
382.4 ( 0.5 105.2 ( 0.7 no yes no yes no yes
382.3 ( 0.5 89.8 ( 0.6 no no no no no no
379.1 ( 1.2 105.4 ( 1.4 no yes no yes no yes
376.4 ( 0.7 106.8 ( 1.0 no yes no yes no yes
372.9 ( 2.1 94.8 ( 2.1 no yes no yes no yes
371.6 ( 0.2 136.4 ( 0.2 no no no no no no
b
HF
C6H5CtCH
(CH3CH2)2NOH
CH3COCH3
p-FC6H4NH2
(CH3)2NCOCtCH 362.5 ( 2.1 132 ( 5
CF3CH2OH
370.6 ( 2.3 133 ( 5
no no no no no no
370.6 ( 2.1 69.6 ( 1.9 no yes no yes no yes
369.1 ( 2.1 96.0 ( 2.2 no yes no yes no yes
364.3 ( 2.1 92.9 ( 2.1 no yes no yes noc yes
yes no yes no yes no
361.7 ( 2.5 107.0 ( 2.5 yes yes yes yes yes yes
358.7 ( 2.1 132 ( 5 yes no yes no yes no
352.5 ( 2.2 86.6 ( 2.2 yes yes yes yes yes yes
HCtCCO2CH3
(CH3)3CSH
a All values in kcal mol-1. PT ) proton transfer and HT ) hydrogen
atom transfer. b Values for the protio compound are given. c A trace of
proton-transfer product was observed, but this result was not reproducible.
Scheme 3
By combining the measured hydrogen atom affinities with
the known electron affinity of phenoxyl radical (51.95 ( 0.14
kcal mol-1) and the ionization potential of hydrogen atom (313.6
kcal mol-1, eq 5), we obtain PA(3o, 3m, and 3p) ) 369.4 (
1.7 kcal mol-1. These values are in accord with the bracketing
results for these anions (g367 ( 5 kcal mol-1) and are in good
agreement with MCQDP2 predictions of 368.5 (3o), 365.1 (3m),
and 369.2 (3p) kcal mol-1. The G2+(MP2) results also are in
good accord with experiment for 3o (366.7 kcal mol-1) and
3m (367.8 kcal mol-1) but not 3p (374.8 kcal mol-1). As for
the G3 and B3LYP proton affinities, they reproduce the value
for the R,3-isomer but give poor values for the R,2- and R,4-
species (Table 3).
Heats of Formation. Our thermochemical data on the R,2-,
R,3-, and R,4-dehydrophenol radical anions can now be used
to derive their heats of formation and those for the corresponding
oxocyclohexadienylidenes 1o, 1m, and 1p. Taking the former
quantities first, one can combine the known heat of formation
of phenoxide ion (-38.3 ( 0.6 kcal mol-1, Table 6) with the
measured hydrogen atom affinities of 3o, 3m, and 3p to derive
∆H°f(3o, 3m, and 3p) ) 17.3 ( 1.8 kcal mol-1. To put this
quantity in perspective, we consider the isodesmic reaction
illustrated in eq 6, which we term the distonic ion separation
in excess of 2.3 eV (e.g., HF and acetylene derivatives).
Consequently, we used (CH3)2NCOCtCH (∆H°acid ) 362.5
( 2.1 kcal/mol) and PhCtCH (∆H°acid ) 370.7 ( 2.3 kcal/
mol) as our brackets and assign PA(3o, 3m, and 3p) g 367 (
5 kcal/mol. These values are assigned as lower numerical limits
because proton transfer to 3o, 3m, and 3p is apt to have a kinetic
barrier; C-protonation is thermodynamically favored, but the
bulk of the charge is on the more electronegative (and less basic)
oxygen atom.
To derive the heats of formation of the three oxocyclohexa-
dienylidenes (1o, 1m, and 1p) from our data, we need to use
the proton affinities of 3o, 3m, and 3p. Fortunately, these
quantities could be determined more precisely by bracketing
the hydrogen-atom affinity (HA) of 3o, 3m, and 3p (eq 4) and
making use of the thermodynamic cycle shown in eq 5. All
three dehydrophenol radical anions undergo hydrogen atom
abstraction with the majority of reagents that were examined,
and all of the alcohols (MeOH, EtOH, tert-BuOH, and CF3-
CH2OH) listed in Table 5. To determine the abstraction site in
energy (DISE). This proposal is directly modeled on the bi-
radical separation energy (BSE, see below) originally described
by Wenthold et al.27 The DISE provides a correction to bond
(25) (a) Davico, G. E.; Bierbaum, V. M.; Depuy, C. H.; Ellison, G. B.; Squires,
R. R. J. Am. Chem. Soc. 1995, 117, 2590-2599. (b) Wickham-Jones, C.
T.; Ervin, K. M.; Ellison, G. B.; Lineberger, W. C. J. Chem. Phys. 1989,
91, 2762-2763. (c) Schulz, P. A.; Mead, R. D.; Jones, P. L.; Lineberger,
W. C. J. Chem. Phys. 1982, 77, 1153-1165. (d) Mihalick, J. E.; Gatev, G.
G.; Brauman, J. I. J. Am. Chem. Soc. 1996, 118, 12 424-12 431.
(26) Wiberg, K. B.; Hadad, C. M.; Sieber, S.; Schleyer, P. v. R. J. Am. Chem.
Soc. 1992, 114, 5820-5828.
the latter compounds (C vs O), 3o, 3m, and 3p were reacted
with deuterated (ROD) alcohols. In each case, hydrogen and
deuterium atom transfer was observed in approximately a
statistical (nonselective) fashion. These results indicate that the
(27) Wenthold, P. G.; Wierschke, S. G.; Nash, J. J.; Squires, R. R. J. Am. Chem.
Soc. 1994, 116, 7378-7392.
9
J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003 4649

A R T I C L E S
Reed et al.
Table 6. Ancillary Thermochemical Data Used and Generated in This Worka
∆H°f(C6H6) ) 19.82 ( 0.12
BDE(C6H5-H) ) 113.5 ( 0.5
∆H°f(C6H5-) ) 55.8 ( 0.5
BDE(C6H5O-H) ) 88.8 ( 0.6
∆H°f(C6H5O-) ) -38.3 ( 0.6
∆H°acid(C6H6) ) 401.7 ( 0.5
•
•
EA(C6H5 ) ) 25.3 ( 0.1
∆H°f(C6H5 ) ) 81.2 ( 0.6
∆H°f(C6H5OH) ) -23.03 ( 0.13
∆H°acid(C6H5OH) ) 350.4 ( 0.6
EA(C6H5O•) ) 51.95 ( 0.14
∆H°f(C6H5O•) ) 13.7 ( 0.6
a All values come from refs 20, 24, and 25.
Table 7. Distonic Ion Separation Energies and Heats of Formation of the R,2-, R,3-, and R,4-Dehydrophenol Radical Anions
DISE, kcal mol-1
∆H°f, kcal mol-1
method
R,2
R,3
R,4
R,2
R,3
R,4
B3LYP
G2+(MP2)
G3
MCSCF
MCQDPT2
exptl
0.8 [2.8]
8.8 [10.7]
1.8
0.2
2.1
5.8 [8.6]
7.8 [10.5]
6.4
3.1
5.6
-0.7 [0.2]
0.8 [1.7]
-0.3
22.3 [20.3]
14.3 [12.4]
21.3
23.0
21.1
17.3 [14.5]
15.3 [12.3]
16.7
20.1
17.6
23.8 [22.9]
22.3 [21.4]
23.4
23.9
21.7
-0.7
1.5
5.8 ( 1.8
5.8 ( 1.8
5.8 ( 1.8
17.3 ( 1.8
17.3 ( 1.8
17.3 ( 1.8
Table 8. Biradical Separation Energies and Heats of Formation of R,2-, R,3-, and R,4-Dehydrophenol
BSE, kcal mol-1
∆H°f , kcal mol-1
R,3
method
R,2
R,3
R,4
R,2
R,4
B3LYP
G2+(MP2)
G3
-1.0 [3.1]
2.9 [6.9]
1.0
0.4 [4.5]
4.4 [8.4]
2.0
0.5 [4.6]
4.7 [8.7]
2.8
76.1 [72.0]
72.2 [68.2]
74.1
74.7 [70.6]
70.7 [66.7]
73.1
74.6 [70.5]
70.4 [66.4]
72.3
MCSCF
MCQDPT2
expt.
0.7
-0.9
5.8 ( 4.0
3.6
2.0
5.4 ( 4.0
2.9
0.9
5.8 ( 4.0
74.5
76.1
69.3 ( 3.9
71.6
73.2
68.9 ( 3.9
72.3
74.3
69.3 ( 3.9
additivity by providing a measure of the stabilizing or desta-
bilizing interaction of a charged center with a radical center in
a given molecule. If there is mutual stabilization the DISE will
be a positive number, and if the two centers destabilize each
other then the DISE will be a negative number. In the case
illustrated in eq 6, the DISE is equivalent to BDE (C6H5-H)
+ IP(H•) - EA(C6H5O•) - PA(p-C6H4O•-) or 375.2 ( 0.5 kcal
mol-1 - PA(p-C6H4O•-). The same relationship holds for the
R,2- and R,3-isomers, and thus the DISEs for all three R,n-
dehydrophenol radical anions are 5.8 ( 1.8 kcal mol-1. This
means that there is a slight stabilizing interaction between the
charged and radical centers, and consequently the carbon-
hydrogen BDE of benzene is somewhat larger than any of the
C-H bonds in phenoxide ion. This result is in keeping with
Wenthold et al., who found that the σ-π exchange energy is
worth about 4-5 kcal mol-1 in these type of systems.27 All of
the computational procedures that were examined reproduce the
DISE for the R,3-isomer but are somewhat too small for the
other two isomers (Table 7). In any case, the calculated DISEs
can be combined with the known heats of formation of benzene,
phenyl radical, and phenoxide to predict the heats of formation
of 3o, 3m, and 3p. The resulting MCQDPT2 energies only span
a 4 kcal mol-1 range (17.6 (3m), 21.1 (3o), and 21.7 (3p) kcal
mol-1) and are in reasonable accord with our experimental
findings. Similar results also are obtained at the G3, G2+(MP2),
and B3LYP levels.
illustrated for R,4-dehydrophenol (eq 7, all values in kcal mol-1).
These quantities (92.3 ( 3.9 kcal mol-1 for 1o and 1p and 91.9
( 3.9 kcal mol-1 for 1m) when combined with the known heat
of formation of phenol (-23.03 ( 0.13 kcal/mol) lead to ∆H°f-
(1o and 1p) ) 69.3 ( 3.9 kcal mol-1 and ∆H°f(1m) ) 68.9 (
3.9 kcal mol-1 (Table 8). These values are the same within the
experimental uncertainties and are ∼6 kcal mol-1 less than
predicted by additivity. In other words, the biradical separation
energy is 5-6 kcal mol-1 for 1o, 1m, and 1p (eq 8).
The heats of hydrogenation of oxocyclohexadienylidenes
1o, 1m, and 1p can be derived from our data by combining
the electron affinities of 1o, 1m, and 1p, the proton affinities
of 3o, 3m, and 3p, the ionization potential of hydrogen atom,
the bond dissociation energy (BDE) of molecular hydrogen,
and the O-H BDE of phenol in a thermodynamic cycle as
9
4650 J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003

R,2-, R,3-, and R,4-Dehydrophenol Radical Anions
A R T I C L E S
dehydrophenoxide anions (3o, 3m, and 3p). The R,2- and R,4-
isomers were independently prepared from o-benzyne radical
anion and p-diazophenol, respectively, and the structures of all
three isomers were rigorously established (i.e., 3o, 3m, and 3p
have different reactivities, they all were converted to phenoxide
and their corresponding quinones, and these derivatives were
found to be identical to the authentic species). Electron affinities,
proton affinities, and hydrogen atom affinities were measured
by bracketing, and in the first case, a kinetic technique also
was used. The results are in reasonable accord with MCQDPT2,
G3, G2+(MP2), and B3LYP calculations and were combined
in thermodynamic cycles to afford the heats of formation of
3o, 3m, and 3p and their corresponding R,n-dehydrophenols.
In both cases, the energies of the three isomers are the same,
and they are more stable than predicted by bond additivity by
5-6 kcal mol-1. These results are analogous to Wenthold and
Squire’s earlier work on R,n-dehydrotoluenes and can be
attributed to the σ-π exchange energy. Our data also suggests
that 1H-bicyclo[3.1.0]hexa-3,5-dien-2-one (1b) is ∼30 kcal
mol-1 less stable than previously suggested.
Our findings are analogous to those reported by Wenthold et
al. for the R,n-dehydrotoluenes (i.e., they have the same heats
of formation and they are 5 kcal mol-1 less than predicted by
bond additivity) and are in reasonable accord with theory.28 It
is also worth noting that Wenthold et al. measured the heats of
formation of 1o and 1p by a totally independent technique
(energy-resolved collision-induced dissociation) and obtained
69-70 kcal mol-1 for 1o and 1p, which is in very good
agreement with our results; the CID method failed for 1m,
presumably because a rearrangement occurred.29,30
Based upon our data, the C-H bond energies of phenoxyl
radical can be derived (Table 3). In addition, by using the
calculated energy difference between 1H-bicyclo[3.1.0]hexa-
3,5-dien-2-one (1b) and 1p reported by Sole et al.,4a we estimate
that ∆H°f(1b) ) 88 kcal mol-1. This quantity is ∼30 kcal mol-1
larger than a previously reported value (57 kcal mol-1) based
upon MP2 calculations.3e The smaller prediction must be
incorrect since the measured heat of formation of 1p (69 kcal
mol-1) is larger and 5 is known to thermally isomerize to 1p at
35 K. Presumably, MP2 fails in this case because 5 requires
more than a one configuration wave function to be adequately
described.
Acknowledgment. We thank Prof. Paul Wenthold for kindly
sharing unpublished data on the R,n-dehydrophenols and Dr.
Lev Lis for providing a sample of 5-nitro-1,3-cyclopentadiene.
Support from the National Science Foundation, the donors of
the Petroleum Research Foundation, as administered by the
American Chemical Society, the Air Force Office of Scientific
Research, and the Minnesota Supercomputer Institute are
gratefully acknowledged. G.C. thanks the Korea Science and
Engineering Foundation for financial support (Grant No. R05-
2000-000-00060-0 (2002)).
Conclusions
Sequential fragmentation of carbon dioxide and nitric oxide
from o-, m-, and p-nitrobenzoate affords the R,2-, R,3-, and R,4-
(28) For a detailed discussion relating to the computational shortcomings of
σ,π biradicals and BSEs, see ref 27.
(29) Wenthold, P. G. Ph.D. Thesis, Purdue University, 1994.
(30) Wenthold, P. G.; Nash, J. J.; Hill, B. T.; Squires, R. R., unpublished work.
JA029571C
9
J. AM. CHEM. SOC. VOL. 125, NO. 15, 2003 4651