Formation of gas-phase dianions and distonic ions as a general method for the synthesis of protected reactive intermediates. Energetics of 2,3- and 2,6-dehydronaphthalene

Article abstract of DOI:10.1021/ja002351j

New methods for the regioselective formation of radical anions are described. Dicarboxylates are generated by electrospray ionization mass spectrometry and lose carbon dioxide and an electron upon collision-induced dissociation to afford distonic ions. These radical anions also can be synthesized via laser desorption, chemical ionization, and electron ionization of dibenzyl esters. Subsequent fragmentation of the remaining CO₂ affords new radical anions corresponding to neutral reactive intermediates. By measuring the proton affinities and electron binding energies of these species, quantitative energetic information on carbenes, biradicals, and other transient molecules can be obtained. This approach is demonstrated by measuring the heats of formation of 2,3- and 2,6-dehydronaphthalene, ancillary thermochemical data also are derived, and the results are compared to either o- or p-benzyne.

Full text of DOI:10.1021/ja002351j

J. Am. Chem. Soc. 2000, 122, 10689-10696
10689
Formation of Gas-Phase Dianions and Distonic Ions as a General
Method for the Synthesis of Protected Reactive Intermediates.
Energetics of 2,3- and 2,6-Dehydronaphthalene
†
Dana R. Reed, Michael Hare, and Steven R. Kass*
Contribution from the Department of Chemistry, UniVersity of Minnesota, Minneapolis, Minnesota 55455
ReceiVed June 29, 2000
Abstract: New methods for the regioselective formation of radical anions are described. Dicarboxylates are
generated by electrospray ionization mass spectrometry and lose carbon dioxide and an electron upon collision-
induced dissociation to afford distonic ions. These radical anions also can be synthesized via laser desorption,
chemical ionization, and electron ionization of dibenzyl esters. Subsequent fragmentation of the remaining
CO2 affords new radical anions corresponding to neutral reactive intermediates. By measuring the proton affinities
and electron binding energies of these species, quantitative energetic information on carbenes, biradicals, and
other transient molecules can be obtained. This approach is demonstrated by measuring the heats of formation
of 2,3- and 2,6-dehydronaphthalene, ancillary thermochemical data also are derived, and the results are compared
to either o- or p-benzyne.
Introduction
The key to broadly applying a thermodynamic cycle as shown
in eq 1 is the preparation of the required starting ions.
Unfortunately, there are few systematic and predictable methods
for synthesizing radical anions, even though they can serve as
precursors to many different classes of compounds. Of the
One of the major successes of gas-phase ion chemistry has
been the determination of a wide variety of thermochemical
1
data. Anions are useful in this regard for probing neutral
molecules including reactive intermediates because they often
are easier to generate and less prone toward rearrangements and/
or fragmentations than their uncharged counterparts. In other
words, an electron can serVe as a protecting group! The
energetics of free radicals can be obtained from carbanions while
bond energies and heats of formation of carbenes, alkenes,
alkynes, biradicals, and many other species can be derived from
radical anions. One way to accomplish this is by measuring an
ion’s proton affinity (PA) and electron binding energy (EBE)
and applying these data in a thermodynamic cycle (eq 1).
•
-
available procedures, the atomic oxygen ion (O ) has been
employed in a number of instances but its reactivity is capricious
9
and its selectivity cannot be controlled. The only other broadly
applicable approach for producing radical anions was developed
by Squires and co-workers and involves the reaction of a
trimethylsilyl-containing carbanion with molecular fluorine.1
For example, the radical anion of p-benzyne can be generated
first by reacting p-bis(trimethylsilyl)benzene with fluoride ion
and then allowing the resulting p-(trimethylsilyl)phenyl anion
to interact with F2 (eq 2). This methodology is extremely
0-12
+
-
-
HA f H + A
∆H° ) PA(A )
(1a)
(1b)
(1c)
•
-
•
-
-
A f A + e
∆H° ) EBE(A )
+
-
•
•
H + e f H
∆H° ) -IP(H )
•
•
-
-
HA f H + A
∆H° ) PA(A ) + EBE(A ) - IP(H )
powerful and has been elegantly applied in a number of
(1d)
13-19
instances,
but it does have several significant drawbacks:
This general approach is applicable to all anions, in principle,
and has been used to measure the energetics of a number of
transient species (e.g., o-benzyne, bicyclo[1.1.0]but-1(3)-ene,
cubene, cubyl radical, dichlorocarbene, phenyl radical, and vinyl
(5) Hare, M.; Emrick, T.; Eaton, P. E.; Kass, S. R. J. Am. Chem. Soc.
997, 119, 237-238.
1
(
6) Grabowski, J. J. In AdVances in Gas-Phase Ion Chemistry; Adams,
N. G., Babcock, L. M., Ed.; JAI Press: New York, 1992; Vol. 1, p 43.
(7) Davico, G. E.; Bierbaum, V. M.; Depuy, C. H.; Ellison, G. B.; Squires,
R. R. J. Am. Chem. Soc. 1995, 117, 2590-2599.
2
-8
radical).
(8) Ervin, K. M.; Gronert, S.; Barlow, S. E.; Gilles, M. K.; Harrison, A.
†
Current address: Department of Chemistry, Ohio University, Athens,
G.; Bierbaum, V. M.; DePuy, C. H.; Lineberger, W. C.; Ellison, G. B. J.
Am. Chem. Soc. 1990, 112, 5750-5759.
OH 45701.
1) NIST Chemistry WebBook, NIST Standard Reference Database
(
(9) Lee, J.; Grabowski, J. J. Chem. ReV. 1992, 92, 1611-1647.
(10) Wenthold, P. G.; Hu, J.; Squires, R. R. J. Am. Chem. Soc. 1994,
116, 6961-6962.
Number 69; Mallard, W. G., Linstrom, P. J., Eds.; National Institute of
Standards and Technology, Gaithersburg MD, 20899 (http://webbook.nist-
.
gov), February 2000.
(11) Hu, J.; Squires, R. R. J. Am. Chem. Soc. 1996, 118, 5816-5817.
(12) Wenthold, P. G.; Hu, J.; Hill, B. T.; Squires, R. R. Int. J. Mass
Spectrom. 1998, 180, 173-183.
(2) Guo, Y.; Grabowski, J. J. J. Am. Chem. Soc. 1991, 113, 5923-5931.
(3) Chou, P. K.; Kass, S. R. J. Am. Chem. Soc. 1991, 113, 697-698.
(4) Staneke, P. O.; Ingemann, S.; Eaton, P.; Nibbering, N. M. M.; Kass,
(13) Wenthold, P. G.; Squires, R. R. J. Am. Chem. Soc. 1994, 116,
11890-11897.
S. R. J. Am. Chem. Soc. 1994, 116, 6445-6446.
1
0.1021/ja002351j CCC: $19.00 © 2000 American Chemical Society
Published on Web 10/14/2000

10690 J. Am. Chem. Soc., Vol. 122, No. 43, 2000
Reed et al.
(
1) Molecular fluorine must be used! (F2 is highly toxic and
For ESI, an API custom-built source from Analytica was interfaced
to an Ultrasource electrostatic ion guide. This external source is
equipped with an adjustable-amplitude fixed-frequency (5.2 MHz)
hexapole ion guide for external trapping of ions prior to injection into
the magnetic field. We have found that lowering the rf amplitude from
the standard 600 Vp-p to between 50 and 250 Vp-p substantially
improves performance for low-mass (<200 Da) ions.
Aryne Radical Anion Generation By ESI. Lithium, sodium,
potassium, or cesium salts of o-, m-, and p-benzenedicarboxylic acids
and 2,3- and 2,6-naphthalenedicarboxylic acids were dissolved in 3:1
corrosive, and is harsh on vacuum systems.) (2) The reaction
is rarely clean and leads to many side products, in part, because
the product ions are unstable with respect to F2. (3) F is not
reactive enough to afford basic anions (PA > ∼400 kcal/mol)
at room temperature. (4) The requisite bis(trimethylsilyl) precur-
sors must be synthesized. In this paper, we describe a versatile
and general alternative for the selective formation of radical
anions. Electrospray ionization (ESI) of dicarboxylic acids
affords dianions that can be collisionally fragmented in a
synthetically useful manner to afford odd electron species in
violation of the “even-electron rule”. Likewise, laser desorption
ionization (LDI), electron impact (EI), and chemical ionization
(
lates which also can be broken apart to radical anions corre-
sponding to reactive intermediates. To demonstrate the utility
of this approach, the heats of formation of 2,3- and 2,6-
dehydronaphthalene were measured, ancillary thermochemical
data are derived, and the results are compared to either o- or
p-benzyne.
-
(v/v) H O/MeOH to a concentration of 1 mM. These solutions were
2
investigated, and diluted further, to find optimal conditions for the
production of the dicarboxylate dianion signals. In general, 750 µM
solutions injected at a rate of 3 µL/min were used since they gave the
most intense signals of the dianions. In these experiments, the ions
were focused through the analyzer cell using the analyzer trap plate
and conductance limit as lenses into the source cell where they were
trapped. This was advantageous because the amount of background
water from the external source is less in the source cell than in the
analyzer cell. Isolation of the dianions was carried out using a broad-
20
CI) of bisesters leads to the formation of monoester carboxy-
22
band chirp excitation, and then they were subjected to collision-
activated dissociation. In all cases, low-energy (0.5-1.0 eV lab)
sustained off-resonance irradiation (SORI)23 was found to be the best
method for cleaving the first molecule of carbon dioxide. The second
Experimental Section
2
CO also was removed by SORI except in the case of 2,3-naphtha-
Synthesis. Dibenzyl esters were prepared using a slightly modified
lenedicarboxylate, where on-resonance collision-induced dissociation
2
1
literature procedure. The dicarboxylic acid of interest was dissolved
in water containing 2 equiv of sodium hydroxide. The resulting solution
was pumped dry under vacuum at 0.1 Torr and 60 °C. Ether was added
to the remaining salt along with 0.2 equiv of tetra-n-butylammonium
bromide and 2.2 equiv of benzyl bromide. After the heterogeneous
mixture was stirred for 24 h at room temperature, water was added
and the organic layer concentrated via rotary evaporation. The crude
product was purified by recrystallization from methanol or a hexanes/
ethyl acetate mixture and, in some cases, a subsequent vacuum
sublimation (150 °C at ∼0.01 Torr).
(
CID) (∼10 eV lab) was used. In this instance, on-resonance CID was
used to suppress an undesired fragmentation pathway involving the
loss of carbon monoxide. After the aryne radical anions were formed,
-4
they were cooled with an argon pulse (g1 × 10 Torr) and carefully
isolated by a combination of a broad-band chirp excitation and SWIFT
24
waveform.
Aryne Radical Anion Generation By EI, CI, or LDI. Large
amounts of hydroxide were generated by pulsed valve addition of a
-5
∼
1:1 N
2
O/CH
4
mixture up to a pressure of 2 × 10 Torr followed by
electron ionization at 3 eV for 100-150 ms. These ions rapidly reacted
with the dibenzyl esters, which were introduced into the FTMS via a
solids probe inlet heated to 170 °C. The resulting monoester mono-
carboxylate ions also could be prepared by EI or LDI of the dibenzyl
esters. In the former (CI) case, they were transferred to the second
Gas-Phase Experiments. All experiments were carried out with a
Finnigan model 2001 Fourier transform mass spectrometer (FTMS)
equipped with a Sun workstation running the Odyssey Software 4.2
package. This instrument is outfitted with a 3.0-T superconducting
magnet and a differentially pumped dual cell. A base pressure of less
(
analyzer) cell and fragmented by an on-resonance excitation (∼20 eV
lab) in the presence of a high pressure (2 × 10 Torr) of argon; these
conditions were found to be optimal for the loss of one CO and one
-9
than 2 × 10 Torr was maintained in each cell by two Edwards model
-5
160M Diffstak diffusion pumps, each of which is backed by an Alcatel
2
model 2015 mechanical pump. The two cubic 2-in. cells are separated
by a common wall (or conductance limit) with a 1-mm opening so
that ions can be passed from one cell to the other at designated times
and a differential pressure of ∼2 orders of magnitude can be maintained
between the cells. Trapping potentials of -1 to -2.5 V were employed
except when ions were being transferred from one cell to the other
and the conductance limit was grounded for 50-150 µs, depending
upon the mass-to-charge ratio of the ions being transferred. Neutral
compounds were introduced into the instrument through a variety of
inlets including a heated solids probe, a batch inlet system, a Granville-
Phillips series 203 variable leak valve, or any of three independent
General Valve dual-solenoid pulsed valves. The FTMS also is equipped
benzyl radical. The resulting distonic ion (a dehydrocarboxylate) was
isolated using a broad-band chirp excitation and SWIFT waveform and
decarboxylated using SORI under high-pressure conditions (g10 Torr
-5
Ar). The resulting benzyne or dehydronaphthalene radical anion was
-4
collisionally cooled with a high-pressure pulse of argon (>10 Torr)
followed by a 700-ms pump out delay and careful isolation to avoid
any re-excitation. Subsequent reactions with a variety of probe reagents
were monitored as a function of time.
2
,6-Dehydronaphthalene Radical Anion (8) Derivatization. For
the conversion of 2,6-dehydronaphthalene radical anion (8) to 2-naph-
-8
thoxide (13), a static pressure of 3 × 10 Torr sulfur dioxide was
admitted into the FTMS cell and an argon pulse for the SORI generation
of 8 was carefully optimized. If the pressure is too low, not enough
fragmentation takes place, and if the pressure is too high, the subsequent
oxygen atom abstraction is largely replaced by adduct formation.
Reaction of the resulting dehydronaphthoxide with a pulse of t-BuSH
leads to hydrogen atom abstraction and the formation of 2-naphthoxide.
The identity of this product was confirmed by comparison with authentic
with a N
2
laser which emits 337-nm light with a 3-ns pulse width and
a power of 6 mW for laser desorption/ionization (LDI) experiments.
(
14) Wenthold, P. G.; Hu, J.; Squires, R. R. J. Am. Chem. Soc. 1996,
18, 11865-11871.
15) Wenthold, P. G.; Hu, J.; Squires, R. R.; Lineberger, W. C. J. Am.
Chem. Soc. 1996, 118, 8, 475-476.
16) Wenthold, P. G.; Hu, J.; Squires, R. R. J. Mass Spectrom. 1998,
3, 796-802.
17) Wenthold, P. G.; Squires, R. R. Int. J. Mass Spectrom 1998, 175,
15-224.
18) Wenthold, P. G.; Squires, R. R.; Lineberger, W. C. J. Am. Chem.
Soc. 1998, 120, 5279-5290.
19) Wenthold, P. G.; Hu, J.; Squires, R. R.; Lineberger, W. C. J. Am.
Soc. Mass Spectrom. 1999, 10, 800-809.
20) McLafferty, F. W.; Turecek, F. In Interpretation of Mass Spectra,
ed.; University Science Books: Mill Valley, CA, 1993; p 55.
21) Barry, J.; Bram, G.; Decodts, G.; Loupy, A.; Orange, C.; Petit, A.;
Sansoulet, J. Synthesis 1985, 40-45.
1
(
(
1- and 2-naphthoxide via bracketing and collision-induced dissociation
3
studies.
(
Computations. All calculations were performed using Gaussian 94
2
or 982 installed on IBM and SGI workstations. Geometry optimiza-
5,26
(
tions were carried out with the B3LYP functional and the 6-31+G(d)
(
(22) Marshall, A. G.; Roe, D. C. J. Chem. Phys. 1980, 73, 1581-1590.
(23) Gauthier, J. W.; Trautman, T. R.; Jacobson, D. B. Anal. Chim. Acta
1991, 246, 211-225.
(24) Wang, T. C. L.; Ricca, T. L.; Marshall, A. G. Anal. Chem. 1986,
58, 2935-2938.
(
4
(

Formation of Gas-Phase Dianions and Distonic Ions
J. Am. Chem. Soc., Vol. 122, No. 43, 2000 10691
Scheme 1
would result holds on to an electron less tightly than a
•
carboxylate (EA(RCO2 ) ∼3.3 eV), and the diminished distance
between the charges leads to an increase in the Coulomb
3
8
repulsion. The resulting distonic ion is interesting in its own
right, but another collision-induced decarboxylation should lead
to a new radical anion which can serve as a precursor to a variety
of reactive intermediates.
o-, m-, and p-Benzyne Radical Anions. To test this
approach, o-, m-, and p-benzenedicarboxylic acids were dis-
solved in a 3:1 (v/v) water/methanol solution containing cesium
hydroxide and were electrosprayed into the gas phase using an
Analytica source interfaced to a Finnigan FTMS. In all three
cases, the ESI mass spectra showed abundant signals for the
basis set. Unrestricted wave functions were used for open shell species.
Vibrational frequencies were computed to ensure that each stationary
point corresponds to an energy minimum and to provide zero-point
energy and temperature corrections to 298 K. Computed proton affinities
and reaction energies are given at 298 K while electron binding energies
are reported at 0 K; temperature corrections typically amount to e 0.2
kcal/mol in the case of electron binding energies.
2
8
dianions as previously observed by Siu et al. Mass selection
of the dianions 1 and collision-induced dissociation with argon
2
3
via SORI leads to 2, which can be further fragmented to the
known o- and m- and p-benzyne radical anions (3, eq 3).2
,10,14,16
Results and Discussion
Multiply charged anions such as dicarboxylates can be
produced in the gas phase via electrospray ionization.27-34 These
ions experience large Coulombic repulsions, especially when
the two charges are close to each other. If one uses the dielectric
constant for a vacuum (ꢀ ) 1), Coulomb’s law indicates that
the interaction energy between two point charges is 332/d kcal/
mol, where d is the distance between the charge centers in
angstroms. One can exploit this situation for the synthesis of
radical anions since fragmentation of a dicarboxylate should
lead to the regiospecific loss of carbon dioxide and an electron
As an alternative to electrospray, we generated 3m and 3p, but
not 3o, via EI, CI, and LDI of their corresponding dibenzyl
esters (Figure 1, eq 4). In particular, 337-nm irradiation of the
(
Scheme 1). Decarboxylation upon collision-induced dissociation
35-37
is a well-known process,
and the accompanying loss of an
electron in violation of the “even-electron rule” also has been
28,29
observed in the analytical literature.
The formation of an
odd-electron species occurs because the new charge center that
bisesters 4, direct electron ionization, or reaction with hydroxide
ion affords the monocarboxylates 5. Sequential fragmentation
of these ions via on-resonance and then off-resonance irradiation
first affords 2 and then 3. This two-step sequence was employed
rather than fragmenting 5 directly to 3 because the former route
is more efficient, and the intermediate dehydrocarboxylates 2
are of interest and can be isolated even in the presence of a
wide variety of reactive neutral reagents. The ortho isomers of
(25) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Gill, P. M. W.;
Johnson, B. G.; Robb, M. A.; Cheeseman, J. R.; Keith, T.; Peterson, G. A.;
Montgomery, J. A.; Raghavachari, K.; Al- Laham, M. A.; Zakrewski, V.
G.; Ortiz, J. V.; Foresman, J. B.; Cioslowski, J.; Stefanov, B. B.;
Nanayakkara, A.; Challacombe, M.; Peng, C. Y.; Ayala, R. Y.; Chen, W.;
Wong, M. W.; Andres, J. L.; Replogle, E. S.; Gomperts, R.; Martin, R. L.;
Fox, D. J.; Binkley, J. S.; Defrees, D. J.; Baker, J.; Stewart, J. P.; Head-
Gordon, M.; Gonzalez, C.; Pople, J. A. Gaussian 94 ReVisions A, B, C;
Gaussian, Inc.: Pittsburgh, PA, 1995.
(26) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
2
and 3 could not be generated in this fashion because
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A.; 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.; Ortiz, J.
V.; Stefanov, B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.;
Gomperts, R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng,
C. Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B. G.; Chen, W.; Wong, M. W.; Andres, J. L.; Head-Gordon, M.;
Replogle, E. S.; Pople, J. A. Gaussian 98; Gaussian, Inc.: Pittsburgh, PA,
fragmentation of 5o leads to the loss of carbon monoxide rather
than carbon dioxide and an alkyl radical. The effect of the ester’s
alkyl group on aryne formation also was briefly examined but
was not optimized. Using EI or CI on the bismethyl esters leads
to 5 where R ) Me, but fragmentation of this species to 2 is
less efficient. Ethyl and trimethylsilyl esters do not work at all
because they donate a hydrogen atom upon dissociation to give
benzoate anion.
1
998.
27) Maas, W. P. M.; Nibbering, N. M. M. Int. J. Mass Spectrom. Ion
Processes 1989, 88, 257-266.
28) Siu, K. W. M.; Gardner, G. J.; Berman, S. S. Org. Mass Spectrom.
989, 24, 931-942.
29) van Berkel, G. J.; Glish, G.; McLuckey, S. A. Anal. Chem. 1990,
2, 1284-1295.
30) Scheller, M. K.; Compton, R. N.; Cederbaum, L. S. Science 1995,
The reactivity of the m- and p-benzyne radical anions has
(
10,14,16
been characterized by Squires and co-workers
in a seminal
series of investigations and the ortho isomer has been thoroughly
(
2
explored by Guo and Grabowski in an important flowing
1
6
2
(
afterglow study. This enabled us to distinguish and confirm the
identity of the three structures of 3 and thereby demonstrate
that their preparation occurs without isomerization. For example,
in accord with the literature, 3o reacts with carbon disulfide
and sulfur dioxide via electron transfer while 3m and 3p undergo
(
70, 1160-1166.
(31) Schroder, D.; Schwarz, H. J. Phys. Chem. 1999, 103, 7385-7394.
32) Alpin, R. T.; Moloney, M. G.; Newby, R.; Wright, E. J. Mass
(
Spectrom. 1999, 34, 60-61.
39
sulfur atom abstraction with the former reagent and somewhat
different behavior from each other with the latter compound.
3m affords electron and oxygen atom transfer products (∼15:
(
(
33) Wang, L. S.; Wang, X. B. J. Phys. Chem A 2000, 104, 1978-1990.
34) Skurski, P.; Simons, J.; Wang, X. B.; Wang, L. S. J. Am. Chem.
Soc. 2000, 122, 4499-4507.
35) Froelicher, S. W.; Freiser, B. S.; Squires, R. R. J. Am. Chem. Soc.
986, 108, 2853-2862.
36) Graul, S. T.; Squires, R. R. J. Am. Chem. Soc. 1986, 112, 2506-
516.
37) Bowie, J. H. Mass Spectrom. ReV. 1990, 9, 349-379.
(
1
(38) Bartmess, J. E. Negative Ion Energetics Data. In Secondary NegatiVe
Ion Energetics Data; Mallard, W. G., Linstrom, P. J., Eds.; National Institute
of Standards and Technology, Gaithersburg MD 20899 (http://webbook.nist-
.gov), February 2000.
(
2
(

10692 J. Am. Chem. Soc., Vol. 122, No. 43, 2000
Reed et al.
by generating the 2,6- and 2,3-dehydronaphthalene radical
anions (8 and 9, respectively), measuring their electron binding
energies and proton affinities, and combining these data in a
thermodynamic cycle as described below.
2,6-Dehydronaphthalene Radical Anion. Generation. Elec-
trospray ionization of the lithium, sodium, potassium, or cesium
salts of 2,6-naphthalenedicarboxylic acid (10) from water/
methanol or water/acetonitrile solutions affords the dianion 11
as the base peak in the mass spectrum (Figure 2). Significant
+
amounts of the monodeprotonated diacid (10 - H , m/z 215)
and the corresponding alkali metal salt of the dianion (11 +
+
M , where M ) Li, Na, K, or Cs; m/z 253 when M ) K) also
are observed. After isolation of the dianion, collision-induced
dissociation with argon via SORI cleanly affords 6-dehydro-2-
naphthalenecarboxylate (12, eq 5). This distonic ion corresponds
to the loss of carbon dioxide and an electron from 11. Its
subsequent fragmentation yields 2,6-dehydronaphthalene radical
anion (8) in abundant quantities. This same ion also can be
prepared starting from the dibenzyl ester of 10 via LDI, EI, or
-
CI with OH as described for 5m and 5p. In addition, a labeling
1
8
-
experiment using OH was carried out. Oxygen-18 is incor-
porated into the carboxylate corresponding to 5, and subsequent
dissociation of this ion leads to the complete loss of the label
-
Figure 1. (a) Chemical Ionization (OH ) mass spectrum of p-dibenzyl
(eq 6). This indicates that only the carboxylate is lost and that
-
benzenedicarboxylate from m/z 18 to 400 (5p, m/z 255; (4p-H) , m/z
3
45). (b) Collision-induced dissociation of 5p with argon plotted from
1
3
-
2
m/z 18 to 400 (5p- C, m/z 256; (5p-CO ) , m/z 211; 2p, m/z 120; 3p,
m/z 76). Inset corresponds to the isolation of 2p. (c) Sustained off-
resonance irradiation of 2p with argon shown from m/z 18 to 400 (3p,
m/z 76). Inset corresponds to the isolation of 3p.
1), and the latter slowly (krel ) 1) undergoes a second oxygen
atom abstraction, while 3p reacts to a larger extent (∼30%) via
the CO2 in the ester group is retained; the same result was found
oxygen atom transfer, the second atom abstraction proceeds
1
3
using 5m with a C-labeled carboxylate and R ) Me and
relatively quickly (krel ) 40), and a significant amount (∼25%)
4
8
several other substrates, we therefore believe that this is a
general finding.
of adduct is observed.4
0,41
In addition, the m- and p-benzyne
radical anions can be isomerized to the more stable ortho isomer
To confirm the structure of 8, we examined its reactivity with
several neutral probe reagents. Carbon disulfide and sulfur
dioxide both react with 8 in a fashion analogous to 3p. With
CS2, two sulfur atom abstractions take place in a sequential
manner while the reaction with SO2 leads to two sequential
oxygen atom abstractions and adduct formation. The latter
reagent also provided a means to derivatize 8 to a known
species. In particular, the oxygen atom transfer product abstracts
a hydrogen atom from 2-methyl-2-propanethiol (t-BuSH) to
afford 2-naphthoxide (13, m/z 143, eq 7). Bracketing studies
1
6
upon reaction with deuterium oxide, as previously described.
2
,6- and 2,3-Dehydronaphthalene Radical Anions. To
demonstrate the utility of our methodology, we decided to
measure the heats of formation of 2,6- and 2,3-dehydronaph-
thalene (6 and 7, respectively) as these species are of interest
as reactive intermediates in synthetic transformations, combus-
42-47
tion processes, and drug design.
This was accomplished
(39) We were able to observe a second sulfur atom abstraction with 3p,
as described in ref 16 and carried out in Q2 of a triple quadrupole mass
analyzer, but not the analogous reaction with 3m.
(40) The reactivity of SO2 with 3p (and 8), but not 3m, is pressure
dependent. The low-pressure reactivity is cited in the text, but at high
-
5
pressures (>10 Torr), the oxygen atom abstraction product is replaced
by adduct formation. This pressure-dependent reactivity is also observed
with phenyl anion and agrees well with reactions carried out under flowing
afterglow conditions as described in ref 14.
show that 13 has the same proton affinity (343.8 ( 2.1 kcal/
mol) as authentic 2-naphthoxide (i.e., proton transfer is not
observed with t-BuSH and acetic acid (∆H°acid ) 352.5 ( 2.2
(41) The relative rates for the second oxygen atom abstraction are in
good agreement with experiments carried out on independently generated
m- and p-dehydrophenoxide. Reed, D. R.; Kass, S. R., unpublished results.
(
42) Bunnett, J. F.; Brotherton, T. K. J. Am. Chem. Soc. 1956, 78, 155-
58.
43) Roberts, J. D.; Semenow, D. A.; Simmons, H. E. Jr.; Carlsmith, L.
A. J. Am. Chem. Soc. 1956, 78, 601-611.
1
(44) Weimer, H. A.; McFarland, B. J.; Li, S.; Weltner, W., Jr. J. Phys.
Chem. 1995, 99, 1824-1825.
(45) Synder, J. P. J. Am. Chem. Soc. 1989, 111, 7630-7632.
(

Formation of Gas-Phase Dianions and Distonic Ions
J. Am. Chem. Soc., Vol. 122, No. 43, 2000 10693
Figure 2. (a) Electrospray ionization mass spectrum of the potassium salt of 2,6-naphthalenedicarboxylic acid (10) in a 3:1 water/methanol solution
-
-
from m/z 20 to 400 (11, m/z 107; (10-H) , m/z 215; (10-2H+K) , m/z 253). (b) Mass selection of 2,6-naphthalenedicarboxylate plotted from m/z
13
2
0 to 400 (11, m/z 107). (c) Sustained off-resonance irradiation of 11 with argon shown from m/z 20 to 200 (12, m/z 170; 11- C, m/z 107.5). (d)
Sustained off-resonance irradiation of 12 with argon plotted from m/z 20 to 200 (8, m/z 126). Inset (m/z 20 to 200) corresponds to the isolation of
8
.
and 348.1 ( 2.2 kcal/mol, respectively) but does take place
with benzoic acid (∆H°acid ) 340.1 ( 2.2 kcal/mol)), but this
result is unlikely to differentiate between the 1- and 2-iso-
mers.4 Therefore, collision-induced dissociation studies were
carried out. The m/z 143 ion produced from 8 and authentic
390.7 ( 0.1 kcal/mol) unless kinetic energy is provided. Upon
reaction with deuterium oxide (∆H°acid ) 392.9 ( 0.1 kcal/
mol), a maximum of 3 H/D exchanges is observed and an acid-
catalyzed isomerization to a more stable and less basic 1,2- and/
or 2,3-dehydronaphthalene radical anion takes place.52 The latter
pathway is indicated by the lack of reactivity of the d3 ion with
additional D2O and the failure to observe back-exchange with
H2O. Nevertheless, on the basis of these observations, we can
9,50
2
-naphthoxide (generated by proton abstraction of 2-naphthol
-
using F ) give identical spectra in which CO is lost over a wide
range of energies (eq 8). In contrast, 1-naphthoxide (generated
5
3,54
reasonably assign PA(8) ) 390 ( 4 kcal/mol.
in good accord with the predicted B3LYP/6-31+G(d) result of
92.5 kcal/mol and is consistent with this hybrid functional’s
ability to reproduce the acidity (∆H°acid) of phenyl radical (381.2
This value is
3
1
8
2
(
(
(
calcd), 377.4 ( 3.4, and 379 ( 5 (expt) kcal/mol), benzene
399.9 (calcd), 401.7 ( 0.5 (expt) kcal/mol), and naphthalene
394.4 (calcd for the 2-position), 395.5 ( 1.3 (expt) kcal/mol).55
The electron binding energy of 8 also was examined by the
bracketing method. Unfortunately, most of the reference com-
pounds in the desired range are substituted aromatic compounds
with acidic protons. These compounds (m-fluoromethylnitroben-
zene, o-trifluoromethylnitrobenzene, and m-trifluoromethylni-
trobenzene) react with 8 by proton transfer as well as isomer-
-
by proton abstraction of 1-naphthol using F ) exhibits an
additional fragmentation pathway to afford ketene enolate, which
becomes dominant with increasing collision energy.
5
1
Thermochemistry. The proton affinity of 8 was examined
by reacting it with a series of standard reference acids. 2,6-
Dehydronaphthalene radical anion was found to deprotonate
fluorobenzene and methanol-OD (∆H°acid ) 387.2 ( 2.5 and
(51) Since no 1-naphthoxide is observed upon derivatization of 8 and
CID of its m/z 143 ion, 1,n-dehydronaphthalene radical anion structures
for 8 are highly improbable. The only other possible structure that could
not give 1-naphthoxide is 2,7-dehydronaphthalene radical anion, but an
improbable hydrogen atom shift is required to form this species. Moreover,
3
83.5 ( 0.7 kcal/mol, respectively) but not water (∆H°acid )
8
reacts with SO2 to give two sequential oxygen atom abstractions at roughly
(
46) Thoen, K. K.; Thoen, J. C.; Uckun, F. M. Tetrahedron Lett. 2000,
1, 4019-4024.
47) Cioslowski, J.; Piskorz, P.; Moncrieff, D. J. Am. Chem. Soc. 1998,
equal rates. This behavior parallels that of 3p, but not 3m, and is consistent
with the fact that p-quinone and 2,6-naphthoquinone are stable closed-shell
molecules whereas m-quinone and 2,7-naphthoquinone are not; they are
non-Kekule molecules.
(52) B3LYP calculations indicate that the two isomers differ in energy
by only 0.2 kcal/mol and that the 1,2-species is more stable.
(53) “[A] dual cell FTMS can be used to bracket lower acidity sites in
multiple-acidic-site- molecules.” See: Kurinovich, M. A.; Lee, J. K. J. Am.
Chem. Soc. 2000, 122, 6258-6262.
4
1
(
20, 1695-1700.
(
48) Broadus, K. M.; Kass, S. R., unpublished results.
(49) All thermochemical data, unless otherwise noted, comes from ref
3
8.
(
50) The proton affinity of 1-naphthoxide is unknown, but most likely
is very similar to its 2-isomer.

10694 J. Am. Chem. Soc., Vol. 122, No. 43, 2000
Reed et al.
Table 1. Summary of Energetic Dataa
cmpd
∆H°f,298 (g)
PA
EBE
BDE
2
2
2
-naphthyl anion (15)
65.9 ( 1.3
120 ( 5
395.5 ( 1.3b
390 ( 4
377 ( 3
30.0 ( 2.0b
31.0 ( 2.0
13.8 ( 2.5
93 ( 3 (C3-H), 106 ( 4 (C6-H)
,6-dehydronaphthalene radical anion (8)
,3-dehydronaphthalene radical anion (9)
107 ( 4
c
111.9 ( 2.4 (C2-H)b
77 ( 4 (C3-H), 107 ( 4 (C6-H)
naphthalene
36.05 ( 0.25
95.9 ( 2.4
151 ( 5
2
2
2
-naphthyl radical (14)
,6-dehydronaphthalene (6)
,3-dehydronaphthalene (7)
121 ( 5
a
All values in kcal/mol. b References 55 and 56. c Reference 60.
izing it to more stable isomers. We did find, however, that
electron transfer occurs upon reaction with pentafluoronitroben-
zene (EBE ) 1.45 ( 0.11 eV) but not sulfur dioxide (EBE )
(EBE ) 31.0 ( 2.0 kcal/mol) and proton affinity (PA ) 390
( 4 kcal/mol) of 8 can be combined with the same quantities
1
.107 ( 0.008 eV), suggesting an EBE for 8 of 1.28 ( 0.19 or
9 ( 4 kcal/mol. The EBE of 8 also was determined via a
2
kinetic method which previously has been used for 1- and
1
4,55
2
-naphthyl anions and m- and p-benzyne radical anions.
Results from these experiments, which involve measuring the
•
-
SO2 to 8 ratio upon collision-induced dissociation of the (8
+
SO2)•- adduct ion and comparing it to a calibration line, give
a value of 31.0 ( 2.0 kcal/mol.
56
Electron binding energies of aryl anions such as phenyl and
- and 2-naphthyl anions are well-reproduced using B3LYP/6-
1+G(d) calculations.55 This method, however, is unsuitable
1
3
for anions whose corresponding neutral is a ground-state singlet
biradical or a singlet species with considerable biradical
character as is the case for 2,6- and 2,3-dehydronaphthalene.
This failing is a consequence of the fact that a singlet biradical
requires at least two configurations to adequately describe its
wave function, whereas the Hartree-Fock part of the B3LYP
functional is a single-determinant method. This difficulty can
be overcome by calculating the energy difference between the
anion of interest and its corresponding neutral in the triplet state,
as the latter species can be adequately described with a single
configuration. The resulting EBE must be corrected for the
singlet-triplet (S-T) gap, which can be done using experi-
mental data or some reliable (not B3LYP) computational
approach.57 Application of this procedure in conjunction with
the experimental S-T gap for o-benzyne (37.5 ( 0.3 kcal/mol)
leads to a computed electron binding energy for 3o of 12.9 kcal/
mol, which is in excellent agreement with the measured value
of 13.01 ( 0.16 kcal/mol.18 Likewise, if we use the S-T gap
for 2,6-dehydronaphthalene computed by Squires and Cramer
that we recently measured for 2-naphthyl anion (EBE ) 30.0
55,56
(
2.0 kcal/mol and PA ) 395.5 ( 1.3 kcal/mol)
along with
the ionization potential of hydrogen atom (IP ) 313.6 kcal/
mol) and the bond dissociation energy of molecular hydrogen
(
BDE ) 104.2 kcal/mol) in a thermodynamic cycle to yield a
value of 115 ( 5 kcal/mol for the heat of hydrogenation of
2
,6-dehydronaphthalene (eq 9). This result is in excellent accord
59
with the predicted B3LYP/6-31+G(d) value of 117.4 kcal/mol
and leads to an experimental heat of formation for 6 of 151 (
5
kcal/mol given the well-known heat of formation of naph-
6
0
thalene (36.05 ( 0.25 kcal/mol). One can use this quantity to
derive the C6-H bond dissociation energy for 2-naphthyl radical
(
BDE2 ) 107 ( 5 kcal/mol), which is 5 ( 5 kcal/mol less
5
8
at the CASPT2/cc-pVDZ level (1.4 kcal/mol at 0 K), then
the calculated B3LYP/6-31+G(d) EBE of 8 is 30.9 kcal/mol,
which is within the experimental uncertainty of our measure-
ment.
The thermochemistry of 2,6-dehydronaphthalene radical anion
can now be used to derive the heat of formation of 2,6-
dehydronaphthalene (Table 1). The electron binding energy
than the C2-H BDE for naphthalene (BDE1 ) 111.9 (
5
5,56
2
.4).
results for p-benzyne (BDE1 - BDE2 ) 3.8 ( 2.9 kcal/mol),
and the 298 K predicted S-T gap for 2,6-dehydronaphthalene
This small energy difference is in accord with the
18
5
8
of 1.6 kcal/mol. It also is in keeping with the small through-
5
8,61
bond coupling expected in such a system.
The absolute heat of formation of 2,6-dehydronaphthalene
radical anion and the C6-H homolytic bond energy of 2-naph-
thyl anion (106 ( 4 kcal/mol, Table 1) also can be readily
derived from our data. The latter result indicates that a radical
center, a lone pair of electrons, or a carbon-hydrogen bond at
C2 has little impact on the C6-H bond strength, a finding that
was previously noted in the benzene/p-benzyne system.18
2,3-Dehydronaphthalene Radical Anion. Generation. Elec-
trospray ionization of the cesium salt of 2,3-naphthalenedicar-
(54) The H/D exchange results with D2O indicate that 8 does not directly
isomerize to 1,2- or 2,3-dehydronaphthalene radical anion in a single process,
at least much of the time, and, therefore, strongly suggests PA(8) e 393
kcal/mol. In addition, if the PA was much greater than 393, then one would
-
expect to observe OH upon reaction with H2O. The observation that 8
deprotonates fluorobenzene suggests that PA(8) g 387 kcal/mol, and thus
we assign the proton affinity a value of 390 kcal/mol. We also adopt a
conservative estimate of (4 kcal/mol for the uncertainty.
(
55) Reed, D. R.; Kass, S. R. J. Mass Spectrom. 2000, 35, 534-539.
(56) The EBE was determined using the calibration line given in ref 55
•-
and the observed SO2 to 8 ratio of 11.0 ( 1.0; a value of 31.0 ( 0.5
kcal/mol results. This uncertainty, however, is apt to be too small, so a
more conservative estimate of (2 kcal/mol obtained by averaging the
uncertainties from both methods ((4 and (0.5) and rounding to the nearest
whole number was adopted. For consistency sake, the same (larger) error
also was used for the electron binding energy of 15.
(58) Squires, R. R.; Cramer, C. J. J. Phys. Chem. 1998, 102, 9072-
9081.
(59) Kolos, W.; Wolniewicz, L. Phys. ReV. Lett. 1968, 20, 243-244.
(60) Cox, J. D.; Pilcher, G. Thermochemistry of Organic and Organo-
metallic Compounds; Academic Press: New York, 1970.
(61) Hoffmann, R.; Imamura, A.; Hehre, W. J. J. Am. Chem. Soc. 1968,
90, 1499-1509.
(57) This approach for computing EBE was suggested to us by Prof. W.
T. Borden.

Formation of Gas-Phase Dianions and Distonic Ions
J. Am. Chem. Soc., Vol. 122, No. 43, 2000 10695
Figure 3. (a) Electrospray ionization mass spectrum of the cesium salt of 2,3-naphthalene dicarboxylic acid (16) in a 3:1 water/methanol solution
-
from m/z 20 to 400 (17, m/z 107; (16-H) , m/z 215). (b) Mass selection of 2,3-naphthalenedicarboxylate and its sustained off-resonance irradiation
1
3
with argon plotted from m/z 20 to 400 (18, m/z 170; 17- C, m/z 107.5). (c) Sustained off-resonance irradiation of 18 with argon shown from m/z
2
2
0 to 200 (18-CO, m/z 142; 9, m/z 126). (d) Collision-induced dissociation of 18 with argon plotted from m/z 20 to 200 (9, m/z 126). Inset (m/z
0 to 200) corresponds to the isolation of 9.
boxylic acid (16) from a 3:1 (v/v) water/methanol solution
affords the dianion 17 as the base peak in the mass spectrum
31+G(d) value of 375.5 kcal/mol. The electron binding energy
of 9 also was examined by the bracketing method. Electron
transfer takes place with pentafluoropyridine (EBE ) 0.68 (
0.11 eV) and sulfur dioxide, but not weaker reagents such as
carbon disulfide and carbonyl sulfide (EBE ) 0.51 ( 0.10 and
(Figure 3). After isolation of 17, two molecules of carbon
dioxide and an electron can be detached via sequential SORI
and CID fragmentations (eq 10). The reactivity of 9 is distinctly
0
.46 ( 0.20 eV, respectively). This enables us to assign EBE
(9) ) 0.60 ( 0.11 eV or 13.8 ( 2.5 kcal/mol, which is within
the experimental uncertainty of the computed (as described
above) B3LYP/6-31+G(d) value of 15.7 kcal/mol; the S-T gap
computed by Squires and Cramer for 2,3-dehydronaphthalene
at the CASPT2/cc-pVDZ level (39.6 kcal/mol) was used.58
The thermochemistry of 2,3-dehydronaphthalene radical anion
can be used to derive the heat of formation of 2,3-dehydronaph-
thalene. By combining the electron binding energy (EBE ) 13.8
(
2.5 kcal/mol) and proton affinity (PA ) 377 ( 3 kcal/mol)
different from 8 but analogous to o-benzyne radical anion (3o).
For example, 9 reacts with sulfur dioxide via electron transfer
rather than oxygen atom abstraction and does not undergo sulfur
atom transfer with carbon disulfide. It also undergoes rapid
of 9 as in eq 9, one obtains a value of 85 ( 5 kcal/mol for the
heat of hydrogenation of 2,3-dehydronaphthalene (Table 1). This
result is in essentially perfect accord with the predicted B3LYP/
6
-31+G(d) value of 85.4 kcal/mol and is in good agreement
2
signal loss with some alcohols (e.g., MeOD), as does 3o, and
with a BLYP/6-311G(d,p) energy of 84.0 kcal/mol previously
•
-
behaves similarly to the M - 2 ion(s) formed by reacting O
47
reported by Cioslowski et al. This quantity also leads to
6
2
with naphthalene.
∆
Hf°(7) ) 121 ( 5 kcal/mol, which is in good accord with a
Thermochemistry. The proton affinity of 9 was measured
by bracketing it with a series of standard reference acids.
Deprotonation of t-BuOD, CD3CN, FCH2CH2OH, and phenyl-
acetylene (∆H°acid ) 374.6 ( 2.1, 372.9 ( 2.1, 371.2 ( 2.9,
and 370.7 ( 2.3 kcal/mol, respectively) was observed but not
63
previous estimate of 126 ( 6 kcal/mol by Linnert and Riveros
but in poor agreement with an older estimate of 91 kcal/mol
64-66
based upon ionization potentials and appearance energies.
(
63) Linnert, H. V.; Riveros, J. M. Int. J. Mass Spectrom. Ion Processes
994, 140, 163-176.
64) Grutzmacher, H. F.; Lohmann, J. Liebigs Ann. Chem. 1970, 733,
88-100.
(65) Grutzmacher, H. F.; Lehmann, W. R. Liebigs Ann. Chem. 1975,
023-2032.
1
1
-pentyne (∆H°acid ) 379.8 ( 2.5 kcal/mol), MeOD, and weaker
(
acids. Consequently, we assign PA(9) ) 377 ( 3 kcal/mol,
which is in excellent agreement with the computed B3LYP/6-
2
(62) The product ion, presumably, consists largely of a mixture of 1,2-
(66) In ref 64, ∆H°f(1,2- - 2,3-dehydronaphthalene) ) 26 kcal/mol,
and 2,3-dehydronaphthalene radical anions. A small amount of (1,n) and/
or (2,n) radical anions also is formed as indicated by a small amount of
deuterium incorporation upon reaction with D2O.
which is unreasonably large and is much greater than the computed
difference of -2.7 (B3LYP/6-31G(d)) and -2.8 (BLYP/6-311G(d, p) (ref
47).

10696 J. Am. Chem. Soc., Vol. 122, No. 43, 2000
Reed et al.
One can use our experimental heat of formation to derive the
via the collision-induced dissociation of dianions formed by
electrospray ionization mass spectrometry, laser desorption,
chemical ionization, or electron ionization has been described.
Ion structures were authenticated by reaction studies and/or
classical derivatization experiments. Subsequent measurements
on 2,3- and 2,6-dehydronaphthalene radical anions provided their
proton affinities and electron binding energies, which were
combined in thermodynamic cycles to afford the heat of
formation of 2,3- and 2,6-dehydronaphthalene. From these data,
second bond dissociation energies of naphthalene were derived,
and the results were contrasted to the analogous energetic
quantities for benzene. Further application of this methodology
for obtaining valuable thermodynamic information on transient
neutral molecules will be reported.
C3-H bond dissociation energy for 2-naphthyl radical (BDE2′
)
77 ( 4 kcal/mol), which is 35 ( 5 kcal/mol less than BDE1.
This energy difference can be equated to the π bond strength
of 2,3-dehydronaphthalene and is the same as in o-benzyne (i.e.,
BDE1 - BDE2 ) 36.0 ( 3.1 kcal/mol).14
The absolute heat of formation of 2,3-dehydronaphthalene
radical anion and the C3-H homolytic bond energy of 2-naph-
thyl anion (93 ( 3 kcal/mol) also can be derived from our
results. In this case, as in the benzene/o-benzyne system, the
presence of a radical center, a lone-pair of electrons, or a
carbon-hydrogen bond at C2 has a large impact on the adjacent
C3-H bond strength. Not surprisingly, the BDE for the anion
is between that of the radical and the parent hydrocarbon since
the two-center three-electron interaction in the radical anion is
better than in the 2-naphthyl radical but not as effective as π
bond formation (a two-center two-electron interaction) in 2,3-
dehydronaphthalene.
Acknowledgment. Support from the National Science Foun-
dation and the donors of the Petroleum Research Foundation,
as administered by the American Chemical Society, and the
Minnesota Supercomputer Institute are gratefully acknowledged.
Conclusions
A new method for the selective formation of radical anions
JA002351J