You are on page 1of 12

Organic nitrates, thionitrates,

peroxynitrites, and nitric oxide: a


molecular orbital study of the
RXNO, 2 RXONO (X = 0, S)
rearrangement, a reaction of potential
biological significance

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Dale R. Cameron, Alison M.P. Borrajo, Brian M. Bennett,


and Gregory R.J. Thatcher

Abstract: The rearrangement of organic thionitrate to sulfenyl nitrite potentially mediates the release of nitric
oxide from organic nitrates, such as nitroglycerin, in the presence of thiol. The biological activity of these
nitrovasodilators is proposed to result from release of nitric oxide in vivo. The thionitrate rearrangement bears
analogy to the rearrangement of peroxynitrous acid to nitric acid, which has been proposed to mediate the
biological toxicity of nitric oxide and superoxide. In this paper, the two concerted rearrangement processes and
competing homolytic reactions are explored using molecular orbital calculations at levels up to MP4SDQl631G*lIMP216-31G*. Examination of structure and energy for all conformers and isomers of RSONO, (R = H,
Me), models for organic thionitrates and their isomers, demonstrates that structures corresponding to
thionitrates and sulfenyl nitrates are of similar energy. Free energies of reaction for homolysis of these
compounds are low (AGO< 19 kcallmol), whereas the bamer for concerted rearrangement is large (AGf (aq.) =
56 kcaVmo1). The larger bamer for concerted rearrangement of peroxynitrous acid to nitric acid (aGz(aq.) = 6 0
kcallmol) again compares unfavourably with homolysis (AGOc 11 kcallmol for homolysis to NO2 or 'NO). The
transition state structures, confirmed by normal mode and intrinsic reaction coordinate analysis, indicate that
considerable structural reorganization is required for concerted rearrangement of the ground state species.
These results suggest that concerted rearrangement is not likely to be a viable step in either biological process.
However, rearrangement via homolysis and radical recombination may provide an energetically accessible
pathway for peroxynitrous acid rearrangement to nitric acid and rearrangement of thionitrate to sulfenyl nitrite.
In this case, NO, will be a primary product of both reactions.
Key words:thionitrate, nitric oxide, peroxynitrite, nitrovasodilator, nitrate.

RCsumC : Le rearrangement des thionitrates organiques en nitrites de sulftnyle pourrait Cventuellement


permettre de degager, en prtsence de thiols, de l'oxyde nitrique B partir de nitrates organiques, comme la
nitroglycerine. I1 est suggtrt que l'activite biologique de ces nitrovasodilatateurs est le resultat d'une expulsion
in vivo d'oxyde nitrique. Le rearrangement thionitrate resemble au rkarrangement de l'acide peroxynitreux en
-. acide nitrique qui serait la cause de la toxicitt biologique de l'oxyde nitrique et du superoxyde. Dans ce travail,
utilisant des calculs d'orbitales molCculaires k des niveaux allant jusqu'k MP4SDQl6-3 lG*//MP2/6-31G*, on a
examint les deux processus concertts du rearrangement et les rkactions homolytiques qui leur font competition.
Un examen de la structure et de I'tnergie de chacun des conformkres et des isomkres des RSONO, (R = H, Me),
modkles des thionitrates organiques et de leurs isomkres, permet de demontrer que les structures correspondant
aux thionitrates et aux nitrites de sulfenyles sont d'energies semblables. Les knergies libres de reaction pour
I'homolyse de ces composts sont basses (AGOc 19 kcallrnol) alors que la barrikre pour le rtarrangement
concert6 est ClevCe (aG#(aq) = 56 kcallmol). La barrikre encore plus ClevCe observte pour le rearrangement
concert6 de l'acide peroxynitreux en acide nitrique (AGz(aq) = 60 kcallmol) donne B nouveau lieu a une
comparaison non favorable avec l'homolyse (AGOc 11 kcallmol pour l'homolyse en NO, ou en 'NO). Les
structures des Ctats de transition, confirmees par une analyse du mode normal et des coordonnees de la reaction

Received May 10. 1995,

I1 '

D.R. Cameron, A.M.P. Borrajo, and G.R.J. hatcher.' Department of Chemistry, Queen's University, Kingston, O N K7L 3N6, Canada.
B.M. Bennett. Department of Pharmacology and Toxicology, Queen's University, Kingston, ON K7L 3N6, Canada.
Author to whom correspondence may be addressed. Telephone: (613) 545-2640. Fax: (613) 545-6669.

Can. J. Chem. 73: 1627-1638 (1995). Printed in Canada / Imprim6 au Canada

Can. J . Chem. Vol. 73, 1995


intrinskque, indiquent qu'une reorganisation structure importante est requise pour un rearrangement concert6
des espkces dans 1'Ctat fondamental. Ces resultats suggkrent qu'il est peu probable que le rCarrangement
concert6 soit une Ctape viable dans l'un ou l'autre des processus biologiques. Toutefois, un rearrangement par le
biais d'une homolyse et d'une recombinaison de radicaux peut s'averer une voie Cnergetiquement accessible
pour le rearrangement de l'acide peroxynitreux en acide nitrique et pour le riarrangement du thionitrate en
nitrite de sulf6nyle. Dans ce cas, le NO, sera le produit primaire de chacune de ces reactions.
Mots elks : thionitrate, oxyde nitrique, peroxynitrite, nitrovasodilatateur, nitrate.

[Traduit par la redaction]

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Introduction
Nitric oxide is the subject of substantial, current interest, stimulated by recent evidence suggesting many varied biological
roles for NO, extending beyond vasodilation to immune function and neurotransmission (1). Furthermore, there is significant interest in the toxicology and pharmaceutical applications
of 'NO and 'NO-releasing agents (2). Interestingly, glyceryl
trinitrate (GTN, nitroglycerin), which has been used since
1879 in the treatment of angina pectoris (3), is now widely
believed to exert its therapeutic effect through release of nitric
oxide ('NO) in vivo (4). Several simple organic nitrates
(RONO,) in addition to GTN, including erythrityltetranitrate,
isosorbide dinitrate, and isoidide dinitrate, are effective and
clinically important vasodilators (5). A substantial body of
evidence supports the hypothesis that the vasodilatory activity
of these organic nitrates, and indeed of other potent vasodilators such as nitroprusside and molsidomine, is largely the
result of activation of guanylate cyclase (GCase) leading to an
elevated level of intracellular cGMP, which mediates vascular
smooth muscle relaxation. Consideration of pharmacological
data has led to the conclusion that activation of GCase by
nitrovasodilators is effected by binding of a common proximal
ligand to a heme site. There is some consensus in the literature
that this ligand is 'NO and that Endothelium Derived Relaxing
Factor (EDRF) is in fact 'NO or a nitrosothiol (RSNO) (6).
The exact biotransformation mechanism by which GTN is
converted to 'NO or nitrosothiol remains unresolved. A sulfhydryl-dependent biotransformation pathway requires mediation by either glutathione-S-transferase (GST) or nonenzymically by a free thiol, possibly cysteine (7). The early,
seminal hypothesis of Ignarro et al. (4a) proposed that organic
nitrates enter the smooth muscle cell where breakdown occurs
by an enzymic or non-enzymic process to yield inorganic
nitrite. The pathway rests on the intermediacy of inorganic
nitrite (NO2-), which must be converted to 'NO (eqs. [I]-[3]).
Although this pathway was accepted for many years, substantial evidence now exists against the intermediacy of NO,- in
the biotransformation of GTN (8).
[I]

RONO, + R'SH

[2] NO,-

R'SH

NO,- + R'SSR'

+ ROH

+ H++HONO a 'NO

[3] 'NO + R"SH

a R"SN0

The task then is to formulate a chemically reasonable mechanism for the conversion of GTN to 'NO mediated by thiol
(RSH). Such a mechanism may be simply drawn as involving

the initial formation of a thionitrate (RSNO,), which rearranges to a sulfenyl (RSONO) or sulfinyl nitrite (RS(0)NO)
(eqs. [4], [5]). Either of these nitrites may be drawn as forming
NO on homolysis. Although a similar mechanism appeared in
the pharmacological literature in recent years, no experimental
work or computational studies have been reported to support
this mechanism (4c, 9). Furthermore, it is of interest that this
mechanism is analogous to that recently proposed by Koppeno1 et al. for the rearrangement reaction between peroxynitrous and nitric acids (10). This peroxynitrite rearrangement,
itself, is of substantial biological significance, as it is proposed
to mediate the biological toxicity of 'NO and superoxide (eqs.
[61, [71).
[4] RONO,

+ R'SH + R'SNO, + NO, + H+

[5] R'SNO,

+ [R'SONO +RfS(0)NO]+ 'NO + R'SO'

[6] 'NO + '0,- + 00NO[7] 00NO- + H+ + HOONO +HONO,


The molecular orbital calculations reported within examine
the thermodynamics, reaction barriers, and reaction profiles
for the thionitrate and peroxynitrite rearrangements and the
thermodynamics of related competing homolytic processes.
The results allow comparison of the two rearrangement processes, examination of the viability of these and competing
processes as general reactions of biological significance, and
speculation as to modes of catalysis. Activation barriers for
the rearrangement processes are calculated at levels up to
MP4SDQl6-3 1G:bNMP2/6-31G:% with aqueous solvation
energies applied and structures along the reaction coordinate
examined by intrinsic reaction coordinate (IRC) and Natural
Bond Orbital (NBO) analysis.

Methodology
Isomers of methyl thionitrate and HSNO, were employed as
truncated models of organic thionitrates and their isomers.
The energy and structure for each isomer of MeSNO, was calculated employing full geometry optimization at the following
levels: AM1 (1 I), PM3 (12), HFl6-3 lG*//HF16-3 1GzSand
MP216-31G*//HF/6-3 IG* using GAUSSIAN 92 (13) on an IBM
RISC 60001355, and SPARTAN 3 . 0 ~on an SGI Iris Indigo,
employing the standard basis sets supplied (14). Energies and
structures for the isomers of HSNO, were calculated employing full geometry optimization at the following levels: PM3,
v.3.0. Wavefunction, Inc. 18401 Von Karman, no. 210,
Irvine, CA 92715, U.S.A.

SPARTAN

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Cameron et al
HFl6-3 1G*//HF/6-3 1G*, MP216-3 1G"llHFl6-3 1G*, MP2163 1G*//MP2/6-3 1G'$, and MP4SDQl6-7 1G*//MP2/6-3 1G*.
Energies and structures for the isomers of HONO, were calculated employing full geometry optimization at the following
levels: PM3, HFl6-3 1G"lMF16-3 1G*, MP216-3 1 1G**//HF/
6-3 1G:*, MP216-3 lG*//MP2/6-3 1G*, and MP4SDQl6-3 lG*//
MP216-3 IG*, with the exception of the rearrangement transition state, which was obtained at levels up to MP216-3 1lGq"+l
lHFl6-31G*. Energies and structures for free radical species
were calculated employing full geometry optimization at the
levels UMP216-3 1G*/NMP2/6-3 lG* and UMP4SDQl63 lG"lNMP216-31G*. All MP calculations use the frozencore approximation. For selected structures, pre-orthogonal
NBO overlap matrices were obtained and second-order perturbational analyses and full NBO analyses performed with GAUSSIAN 92 (15). Mulliken charges and charges derived from the
electrostatic potential (CHELP) were recorded for each structure (16). The continuum dielectric solvation models
employed for aqueous solvation energies were SM3-PM3 (17)
as im lemented in SPARTAN and the CDM model of Lim and
Chan. These models were applied using geometries obtained
at the ab initio level. The identity of stationary points was confirmed by calculation of analytical frequencies at the appropriate level. Thermochemical data were obtained by normal
mode analysis and calculation of analytical frequencies (using
SPARTAN 4.0 for RMP2 structures. Free energies, enthalpies of
reaction (AG, AH), and free energies of activation (AG') were
calculated from various reactants including a PdVterm for dissociative processes. Transition state structures for both rearrangement reactions were located by full optimization using
linear synchronous transit geometries (obtained using SPARTAN)as starting points. The character of these transition state
structures was confirmed by (i) analysis of the normal mode
associated with the single imaginary frequency, and (ii) IRC
analysis (at the RHFl6-3 lG* level, using GAUSSIAN 92) along
the reaction coordinate from transition state to ground states
(19). The term ground state is used throughout to indicate stationary points on the calculated potential energy surfaces possessing zero imaginary frequencies. The more correct terms,
reactant and product, lead to confusion, since the two rearrangement reactions discussed in this paper are considered in
forward and reverse directions. Frequencies and zero point
energies (EZp)are scaled by 0.9 (20). A discussion on the suitability of the 6-3 lG* basis set in calculations at the MP2 level
for methyl nitrate and nitric acid has appeared previously (21).

I:

Results
Thorough exploration of the conformational space for nitric
acid isomers has been reported by Hehre and co-workers
(21b), therefore only one isomer of nitric acid (1) and the cis
(2c) and trans (2t) isomers of peroxynitrous acid were studied
herein (Scheme 1, Table 1). However, in the absence of previous work, a thorough study of the isomers of the thionitrates,
HSNO, and CH3SN0,, was required. Several rotamers were
located of structures corresponding to sulfenyl nitrites
(RSONO (3H, R = H; 3M, R = Me)), in addition to conformers
of sulfinyl nitrites (RS(0)NO (4H, R = H; 4M, R = Me)) and
thionitrate (RSNO, (5H, R = H; 5M, R = Me)) (Scheme I,
C. Lim and S.K. Chan. Institute of Biomedical Sciences,
Academics Sinica, Taipei, Taiwan. See ref. 18.

Scheme 1.

Table 1, Table 2, Fig. 1). The identity of these structures as


energy minima and first- or second-order saddle points was
confirmed by calculation of analytical frequencies (Table 1,
Fig. 1). A greater range of conformers was located for the Me
series due to the additional S-C bond rotational barriers. Calculations at the HFl6-31G* level show sulfenyl nitrite (3) and
thionitrate (5) isomers to be of similar energy (Table 1, Fig. 1,
Fig. 2), in particular in comparison to their oxygen congeners,
nitric and peroxynitrous acid, where the latter is considerably
the less stable. Estimation of the influence of aqueous solvation through various solvation approximations demonstrates
little difference in energy between the thionitrate and sulfenyl
nitrite isomers and their conformers (Table I, Fig. 1). Thus
sulfenyl nitrites are predicted to be of similar stability to their
thionitrate isomers in the gas phase and solution. Comparison
of energies obtained from geometry optimization at the MP2
level of 3H, 4H, and 5 H does not significantly change this
conclusion, but at this level the structure of the sulfinyl nitrites
(4H) is lost upon geometry optimization to yield association
complexes of NO with HSO (Tables 1, 2, 5). Again at the
higher MP4llMP2 level sulfenyl nitrites and thionitrates are
seen to be of similar energy (Tables 1, 5). The experimentally
observed frequencies in tert-butyl thionitrate associated with
NO, a stretch, NO, s stretch, and NO, bend are 1301, 1265,
and 831 cm-I, compared with 1322, 1226, and 806 cm-' calculated for HSONO,. The N=O stretch in HSONO is calculated at ==I593 cm-' (3Hcs, 3Hca), suggesting the possibility
for experimental detection of transient sulfenyl nitrite species.
We previously calculated structures and energies for all
conformers of HSNO, and CH3SN0, by full geometry optimization using the semi-empirical AM 1 and PM3 Hamiltonians
and ab initio calculation at the lower RHFl3-21G'Vevel (22).
At these lower levels of calculation, geometrical parameters
for the isomeric thionitrates and sulfenyl nitrite structures are
reproduced reasonably consistently when compared to the
data at the higher levels of calculation. However, molecular
energy is highly dependent on method and basis set (22). This
is illustrated for the HSNO, series, in which molecular energies at the PM3 level show sulfinyl nitrites as the most stable
structures, whereas at the HF16-31G*//HF/6-31G": sulfenyl
nitrites are most stable (Fig. 2). Comparison of these data and
those at higher levels of calculation suggest that, for these and
related molecules, the minimal level of calculation must
include MP216-31G* energies, although a lower level of
geometry optimization may be accommodated.

Table 1. Energies, atomic charges, solvation energies, and thermochemical data used for calculation of reaction free energies and activation parameters for peroxynitrous acid isomers,
thionitrate isomers, and free radical species.
CHELP charges, e'

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Molecular energies, hartrees

E(MP2IHF)"

E(HF/HF)"

E(MP2MP2)'

E(MP~MF)"

E(MP4MP2)' SIO

01

Energies, kcallmol

03

1
2t
2c
3Hca
3Hcs
4H
5H

6
7
OH
OOH
HSO
HS
NO?
NO
Note: MP2 level calculations and data on radical species have been published previously.
"MP216-311G**//HF/6-31G" for 1-2, 6; MP216-31 G"llHF16-3 1 G" for 3-5, 7.
'HFI6-31 G"lkIF16-3 1 G*'.
'MP216-3 1 G"IlMP216-3 1G".
d ~ P 4 S D Q / 6 -13G*l/HF/6-31G".
"MP4SDQl6-3 1 G"//MP2/6-3 1 G".
'For MP2 geometry (3, 4, 5, 7); HF geometry (1, 2, 6).
xFor MP2 geometry, except 1-2, 6 HF geometry.
"using HFIIHF charges and geometries (1, 2, 6); MP2//MP2 charges and geometries (3, 4, 5, 7).
'MP2IlMP2 1-3, 5, 7 and free radical species; HFl/HF for 4, 6; HFIIHF data, for other compounds, used to obtain data in Fig. 4a is not shown.

PM31
SM3"

CDM
aq"

Hv'

EZP'
scaled

SI,,',
cal mol-' K-'

cn
W

Cameron et al

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Table 2. Bond lengths


(Me series).

(A) and bond angles for selected structures optimized

at MP216-31G* (H series) or HFl6-31G*

HSONO 3Hca
HSONO, 5 H
HS(0)NO 4H
HSONO 3Hcs
MeSONO 3Mcs
MeSONO 3Mca
MeSNO, 5 M
MeS(0)NO 4M

Fig. 1. Structures, structural labels, and relative molecular energies (kcallmol) for methyl thionitrate
isomers. First- and second-order saddle points (two imaginary frequencies) are indicated by
superscripts($). Energies are MP216-3 lG*//HFl6-31G" + Ezp with aqueous solvation energies
(SM3) indicated in parentheses. The stereochemical labels employed to indicate isomer geometry
are applied in an identical fashion to the H series (3H, 4H, 5H).

Fig. 2. Relative molecular energies for conformers of 3H, 4H,


5H calculated at semi-empirical and ab initio levels: PM3 + EZP;
HF16-31G"//HF/6-31G* + E,,

30

Calculations on NO, and other free radical species treated


in this present paper have been reported previously (23). The
purpose of the calculations reported within is to provide comparative thermodynamic parameters for the homolytic dissociations of the nitric acid and thionitrate isomers. Energies
obtained at the UMP4SDQl6-3 lG*/NMP216-3 1G* level
(Table 1) are used with zero point energy (EZP)corrections
and vibrational data to yield free energy, enthalpy, and
entropy of reaction for homolysis (Tables 3-5; Figs. 3,4,6,7).
For comparison with these calculations, the experimentally
estimated 0-0 bond energy in peroxynitrous acid is -20
kcallmol (24). Calculated values of (s2)
are in the range 0.760.78 (doublets) and reveal little spin contamination. Projected
PMP2 energies with annihilation of contaminating spin states
deviate from UMP2 energies by 0.8 kcaVmo1 for thiyl radical
to 2.7 kcallmol for nitric oxide. Extension of electron correlated effects from the MP2 to MP4SDQ level leads to differences in relative molecular energies for radical and other
species, modifying quantitative reaction energies but not altering qualitative conclusions made from data calculated at the
MP2 level (Tables l , 4 , 5 ) .
The free energy profile for rearrangement of isomers of per-

Can. J. Chem. Vol. 73, 1995

Table 3. Relative molecular energies and free energies (at 298 K) for HXNOO
species and radical products, calculated at MP4SDQl6-31G:V/MP2/6-3 IG* level,
kcallmol.
HONO, 1

HOONO 2c
AE
~(+EzP)
AG

-22.41
-22.80
-9.74

-49.86
-48.80
-35.41

HSON03Hcs

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

NO

+ HOO

0.00
0.00
0.00

HSNO? 5H

HSO

(a)

+ NO

AG

-s

10.3

H+

ONO;

= HOONO-$

NO + HO,

8
.,,,-,

+ HO

-4.54
-7.50
-6.97

Fig. 3. Thermochernical data for (a) peroxynitrous acid, nitric


acid, and related free radical species; (b) thionitrate, sulfenyl
nitrite, and related free radical species. Free energies of reaction
shown are at the MP4SDQl6-3 lG*//MP216-31G* level, in kcaV
mol (see Table 3).

NO + 0-,

NO,

3.4
HO + NO,

HS

+ NO,

Fig. 4. Reaction coordinate - free energy profiles for homolysis


and rearrangement. (a) Free energy profile for peroxynitrite
rearrangement. Relative free energies below line are at MP4SDQl
6-3 lG*//MP2/6-31G* and those above line (and AG') are
MP4SDQ16-31G*//HF/6-3 lG* + E(CDM-aq.sol), in kcal/mol
(see Tables 1, 4). (b) Free energy profile for thionitrate
rearrangement. Relative free energies below line are at MP4SDQI
6-3 1G*//MP2/6-3 IG* and those above line (and AGZ) are
MP4SDQ16-31G*l/MP2/6-3 1G* + E(CDM-aq.sol.), in kcal/mol
(see Tables 1,5). E(CDM-aq.so1.) is derived from CDM data.

's
's

's

's

%3

3
B

-259
. $'

HONO,

HS + NO,

(b)

HSONO

AG

oxynitrous acid with nitric acid was explored by varying several bond angles and bond lengths separately and in
combination, approximating the reaction coordinate, at variin each case, the "reaction
ous levels of calculation:
profile" proceeded to plateau approximately 65-100 kcal/mol
above peroxynitrous acid, whereupon "dissociation" of the
molecule occurred. A transition state for rearrangement was
instead located by geometry optimization from the linear synchronous transit structure generated from high-energy conformers of HONO, and HOONO (Fig. 5). Clearly, destruction
of the 0 - N - 0 - 0 plane is essential for transition state location
and indeed reaction. To confirm that this structure corresponded to the transition state for rearrangement, the imaginary frequency corresponding to the reaction coordinate and
structures along the intrinsic reaction coordinate were examined. One imaginary frequency was located corresponding to
the reaction coordinate (-2143 cm-', RHF frequency) and the
reaction profile generated from the IRC did indeed correspond
to rearrangement of peroxynitrous acid to nitric acid (Fig. 6).
Examination of structures along the IRC from trans-peroxyni-

HONO,

ow ever,

+ HSO

0.0
HSNO,

-1.4 3Hcs

HSONO

trous acid (2t) to nitric acid demonstrates clearly the initial


rotation about the 0-0 bond to place HO orthogonal to the
plane of the nitro moiety ( 0 - N - 0 ) (Fig. 6). Reaction proceeds
by lengthening of the 0-0 distance and migration of HO
towards nitrogen and thence relaxation to the planar nitric acid

Cameron et at.

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Fig. 5. Structural data for peroxynitrite (6) and thionitrate (7)


rearrangement transition states obtained by optimization at
HF16-3 lG* and MP216-3 lG*, respectively. See Figs. 6 and 7
for stereochemical representations.

structure. In addition, the proximity of the hydrogen to the


oxygen 0 3 along the entire reaction profile is clear. Although
the 03...H distance reaches 2.0 A and there is likely electrostatic stabilization between these two centres, NBO analysis
reveals no significant charge transfer interactions indicative of
stronger hydrogen bonding.
The reaction energy profile depicted uses the N- OH distance as the x-axis reaction coordinate (Fig. 6). The profile
generated using the mass-scaled internal coordinate (not
shown) is smooth and more symmetric; however, use of the
N-OH distance as the reaction coordinate provides a profile
that clearly demonstrates a curious and informative "discontinuity" at d(H0-N) .=: 2.2 A. This discontinuity lies at the
crossover between the two processes, torsional and scissile.
The first process is initial torsional rotation about the bond
between the oxygen (03) and nitrogen in peroxynitrous acid
(2t), to place HO orthogonal to the 0 - N - 0 plane. This structure is close to the transition state for rotation about this 0-N
bond. The second process, structural deformation of this "rotamer" through 0-0 bond elongation, migration, and cleavage, is associated with the large energy barrier.
The reaction energy profile for thionitrate rearrangement
was located in a similar manner to that for peroxynitrite. IRC
energies and structures were calculated at the HFl6-3 1G;*//HF/
6-3 lG* level and reveal a similar reaction pathway to the peroxynitrite rearrangement. However, the dislocation of HS
from the nitro plane ( 0 - N - 0 ) is much less marked than that of
HO in the related rearrangement. Reaction again proceeds
from the trans isomer, in this case sulfenyl (SHca), to the planar thionitrate (3H). Again the proximity of the hydrogen with
the bridging oxygen along the profile is noted and, again, NBO
analysis did not provide evidence of a hydrogen bond involving significant charge transfer. Since the energies of the 5Hca
and 5Hcs conformers are close and d e ~ e n d e non
t level of calculation, this rearrangement reasonably may proceed equally
from either conformer. The structures and energies of the
ground and transition state species were reoptimized at the
MP4SDQl6-3 lG*I/MP2/6-3 1G* level (Fig. 7) with normal
mode analysis (Table 1, Table 2, Fig. 5). The imaginary frequency corresponding to the reaction coordinate for 7 is at
-886 cm-' (MP2 frequency).
Combination of the thermodynamic data for homolysis with
activation energies for rearrangement allows construction of
free energy profiles for reactions of peroxynitrite (Fig. 4n) and
thionitrate (Fig. 4b). Relative molecular energies, enthalpies,
and free energies have been calculated at a number of different
u

levels with zero point energies (Ezp) and aqueous solvation


approximations (Tables 3, 4, and 5). The PM3-SM3, CDM,
and other solvation models applied do not yield consistent values for aqueous solvation energies, presumably owing to lack
of parameterization for molecules similar to those in this study
(Table 1). Cramer and Truhlar state that the PM3-SM3 inodel
may give anomalously positive solvation energy values for
nitroaliphatics partly because of unusual atomic charges at N
in PM3 level calculations; SM2 performs equally poorly (17).
Comparison of PM3 electrostatic charges at atoms with those
from MP2 calculations (Table 1) demonstrates the anticipated
large positive charges at N in 1 H and 5 H at the PM3 level:
+1.81 (HF geometry) and +1.44 (MP2 geometry), respectively. The CDM aqueous solvation model utilizes charges
from the ab initio calculations and does indicate PM3-SM3 to
yield a relatively positive solvation energy for l H , but not for
5H. In addition, the choice of charge type, electrostatic, Mulliken, or natural, has a significant effect on CDM solvation
energies. Clearly, use of the calculated aqueous solvation
energies is limited to indication of the magnitude of perturbation of the reaction free energies calculated in the gas phase,
on transfer to aqueous solvent (Table 1). For these small neutral molecules this perturbation is sinall relative to the large
activation free energies. At the highest level of calculation
employed and with (CDM) solvation energies applied, it can
be estimated that the free energy of activation in aqueous solution at 298 K for peroxynitrous acid rearrangement to nitric
acid is 60 kcallmol and that for rearrangement of thionitrate to
sulfenyl nitrite is 56 kcallmol (Fig. 4).

Discussion
GTN biotransformation
Two general pathways for organic nitrate biotransformation
may be described, characterized as sulfhydryl dependent and
heme dependent (4, 7, 25). There is substantial debate about
whether the sulfhydryl-dependent pathway requires glutathione-S-transferase (7c). However, it has been conclusively
demonstrated that GTN in the presence of specific, simple thi01s is capable of activation of guanylate cyclase ( 7 4 . Therefore, it has been postulated that in the presence of these specific
thiols 'NO is a minor product of the reaction of thiol with GTN.
The major product of reaction of GTN with all thiols at neutral
pH is inorganic nitrite (8). Early hypotheses required generation of 'NO or nitrosothiol from the initial nitrite product (4n).
In recent years, alternative postulates have appeared widely,
involving thionitrate as an intermediate, with further reaction
such as (a) subsequent reduction to nitrosothiol (for example,
ref. 26), or (b) concerted reai-mngeinent to sulfinyl and sulfenyl
nitrites (4c, 9). Indeed, a chemically reasonable mechanism
may be drawn, in which thionitrate partitions between thiolysis
yielding nitrite ion and concerted rearrangement to the sulfenyl
or sulfinyl nitrite (Scheme 2). Support for the subsequent
homolytic generation of 'NO may be drawn from (a) the
reported facile homolytic decomposition of the related sulfenyl and sulfinyl nitrates (27), and (b) the relative stability and
ease of generation of sulfinyl radicals (28). However, only one
paper has appeared, by Oae et al., directed at the reactivity of
thionitrates, and this work did not deal with aqueous solution
(29). The paper of Oae et al. yields only circumstantial evidence for rearrangement, in that sulfonyl species are observed

Can. J. Chem. Vol. 73, 1995

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Fig. 6. Reaction energy profile for rearrangement of peroxynitrous acid (2t, right-hand side) to
nitric acid (1, left-hand side), generated from IRC analysis on the HF16-3 lG*l/HF16-31G"
transition state. The N-OH distance is chosen as the reaction coordinate and x-axis. Use of the
mass-scaled internal coordinate as reaction coordinate yields a smooth profile without discontinuity.
The reaction coordinate shown is chosen to better illustrate the dynamics of rearrangement.
Structures along the reaction profile are shown with nitrogen and the terminal oxygen eclipsed as
shown by the appended Newman projection.

1.3

1.5

1.7

1.9

2.1

2.3

d (HO-N)

Fig. 7. Reaction energy profile for rearrangement of thionitrate HSNO, ( S H , left-hand side) to
sulfenyl nitrite HSONO ( 3 H c a , right-hand side), generated from IRC analysis on the HF16-3 1G*//
IIFl6-31G* transition state. The S-N distance is chosen as the reaction coordinate and x-axis.
Structures along the reaction profile are shown with nitrogen and the terminal oxygen eclipsed as
shown by the appended Newman projection.

Cameron et al.
Table 4. Relative molecular energies and free energies of peroxynitric acid isomers, free radical products, and the rearrangement
transition state, kcal/mol.

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

NO + HOO
NO, + HO
HONOz 1
HOON02t
HOON02c
HOON06

52.61
36.50
0.00
36.11
33.15
-

49.86
45.32
0.00
29.59
27.445
-

0.00
32.40
3 1.50
133.61

"MP2/6-31G*NMP2/6-3 1 G*.
'MP4SDQ16-3 1G"NMP2/6-3 lG*.
'HF/6-3 lG"/kIF/6-3 1G*.
"MP2/6-31 lG"*//HF/6-31G*.

Table 5. Relative molecular energies and enthalpies (at 298 K) of thionitrate isomers, free radical products, and the rearrangement
transition state, kcal/mol.
-

AE(MP2MP2)"
HSONO 7
HSNO, 5H
HSNO, 5H'
HSON03Haa
HSON03Has
HSON03Hsa
HSON03Hss
HSOBI3Hca
HSON03Hcs
HS(O)N04H
HS(0)N04H1
NO + HSO
NO, + HS

57.98
0.00
6.17
12.40
6.37
12.32
7.92
7.12
4.90
8.68
23.49
29.53
34.19

AE(HF/HF)"

AE(MP2IHF)'

AE(MP2MP2)" + E,'

AH(MP~/MP~)"AE(MP4Mp2)" + E,,/

70.96
2.8 1
10.40
4.81
1.17
5.5 1
3.35
0.49
0.00
21.64
24.03
-

58.19
0.00
6.43
13.28
6.87
13.26
8.62
8.80
6.5 1
23.32
27.23
-

58.14
0.00
7.80
12.89
7.52
12.75
8.63
7.12
5.87
10.07
24.64

55.85
0.80

56.42
0.80

0.6 1
0.00

0.76
0.00

14.04
27.88

15.62
29.20

"MP2/6-3 1G*//MP2/6-3 1G*.


'HF/6-3 1G*/kIF/6-3 1G".
'MP216-31G*/kIF/6-31G".
" M P ~ s D Q / ~ -13G*//MP2/6-3 lG*
'HF on HF geometries.
IMP2 on MP2 geometries.

Scheme 2.

..

.
II

R-S-N=O

RS-(-

11

+ O

R-S-0-N=O

0*

R-S-N+
RSSR +

fQ

B=O
RS=O +'NO

&'&

0 0

II

R-S-S-R

II

0
I1

R-S-S-

II
0

y o 1 -RSH

fl

R-S-0-k0

0
II

R-s-0II

as products of thionitrate decomposition, presumably 'from


sulfinyl radical recombination (30) (Scheme 2). However, we
have recently demonstrated 'NO generation from tert-butyl
thionitrate in aqueous solution at neutral pH and ambient ternperature using an 'NO selective Clarke-type electrode (ISONO, World Precision Instruments) (Thatcher and Lock, unpub-

lished results). The question to be answered is whether this


NO may derive from sulfenyl nitrite after rearrangement of
the initially formed thionitrate (Scheme 2).
'

Thionitrates, sulfenyl and sulfinyl nitrites


The few thionitrates studied to date are reactive species, only

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Can. J. Chem. Vol. 73, 1995


tertiary alkyl thionitrates having sufficient stability for storage.
Sulfinyl and sulfenyl nitrites have not been isolated and similarly can be expected to be high-energy species. Sulfenyl
nitrates have beenobserved to undergo rapid homolysis in aqueous solution to NO2 and sulfinyl radical, yielding a sulfonothioate product by radical recombination (27). Although the planar
thionitrate is the only stable conformer, there are several lowenergy sulfenyl nitrite conformers. Thecalculations on the isomers of MeSNO, and HSNO, demonstrate that structures corresponding to thionitrates and sulfenyl nitrites are of similar
energy at the highest level of calculation with electron correlation incorporated at the MP2 or MP4SDQlMP2 levels (Tables
3,5; Fig. 1). Structures corresponding to sulfinyl nitrites are not
stable at the highest level of geometry optimization. The lowest
energy sulfenyl nitrite conformers are 1-7 kcallmol higher in
energy than the thionitrate in the H and Me series ofcompounds,
except with inclusion of electron correlation at the MP4SDQ
level when the sulfenyl nitrites become marginally more stable
(Fig. 1, Table 5). Thus rearrangement of thionitrate to sulfenyl
nitriteappears thermodynamically accessible.

Thionitrate rearrangement
The thermodynamic viability of rearrangement of thionitrate to
sulfenyl nitrite, in light of competing homolytic fission to NO,
and thiyl radical, can be judged from the data shown in Figs. 3
and 4 (Tables 2 and 5). Indeed comparison with data for 'NO and
sulfinyl radical gives information on the homolysis of the sulfenyl nitrite itself. Clearly, rearrangement of thionitrate to sulfenyl nitrite is thermodynamically viable: AGo(aq) = 0.6 kcall
mol. Competitive homolysis to NO2 and sulfinyl radical reflects
a low S-N bond dissociation energy: AGO=16.7 kcallmol (Fig.
4). However, the lower free energy of reaction of sulfenyl nitrite
to yield 'NO and thiyl radical (AGO= 3.8 kcallmol, Fig. 3) suggests that, with a low barrier to rearrangement, this pathway
would be viable for .NO generation from the initially formed
thionitrate (Scheme 2). It rkmains to examine the energy barrier
of the thionitrate +sulfenyl nitrite rearrangement step.
Inspection of the HOMO and other close-lying occupied
orbitals for thionitrate reveals problems in terms of orbital symmetry for a simple concerted reaction via migration of HS along
the nitro (ONO) plane. Indeed, the transition state located for
the rearrangement reaction has the sulfur dislocated out of the
nitro plane, although S, 0 , N, and 0 are coplanar in both ground
states. If the rearrangement is viewed from the sulfenyl nitrite,
reaction is initiated by rotation about the SO-NO and HSON bonds followed by S-0 bond elongation and migration of
sulfur to nitrogen (Fig. 7). The free energy of activation for the
rearrangement of thionitrate (5H) is 56 kcallmol in the gas
phase at the highest level of calculation. The PM3-SM3 model
suggests a small relative stabilization of the transition state
from aqueous solvation (Table l), but the CDM model yields a
negligible reduction of the barrier, which remains at 56 kcall
mol at 25C. This activation barrier is clearly substantial, especially in light of the small bond dissociation energies associated
with the S-0 and S-N bonds of the ground states. The size of
this barrier provoked us to examine the similar peroxynitrite
rearrangement. This reaction had been proposedin the literature, supported by MO calculations.
Peroxynitrite rearrangement
The rearrangement mechanism proposed by Koppenol et al. is

founded upon comparison of molecular orbital calculations


with experimental data (10). A mechanism via homolytic dissociation of peroxynitrous acid and radical recombination of
hydroxyl radical and nitrogen dioxide was re'ected partly on
.
the basis of an estimated AS*of 1-5 cal mol- ! K- I ,>which
was
argued to be too low for a homolytic reaction. Instead, a concerted rearrangement mechanism was proposed in which peroxynitrous acid undergoes concerted migration of OH from 0
to N via a high-energy rotamer to yield nitric acid as product.
The free energy barrier to rearrangement was presented as 17
kcallmol (PM3 level). The authors reported location of a
"potential transition state" at the PM3 level with LOON = 80"
and d(0-0) = 1.8 A (cf. Fig. 5), although no confirmation by
normal mode analysis was provided. Subsequent calculations
at the RHFl6-3 1G4:*:level on peroxynitrous acid conformers
did not locate any transition states, but a speculative rearrangement mechanism was expounded (3 1). Previous to this
work, the conformational space of peroxynitrous acid had
been fully explored at the RHF and MP216-31Ge levels by
Hehre and co-workers (21b).
The formation of peroxynitrite species and isomerization in
the gas phase are also of interest in atmospheric chemistry.
Atkinson et al. proposed that alkyl nitrates formed from reaction of peroxy radicals with nitric oxide might result from
rearrangement of the peroxynitrite to the nitrate via a cyclic
transition state akin to the transition state located in the present
calculations (a flattened analogue of 6 was drawn) (32). However, these workers emphasized a rearrangement proceeding
from a vibrationally and chemically excited peroxynitrite, for
which a reaction energy profile was drawn. Burkholder et al.
suggested a similar rearrangement of peroxynitrous to nitric
acid, again via a high-energy form of peroxynitrous acid (24).
The rearrangement pathway mapped by IRC analysis from
location of the transition state structure (Figs. 5, 6, and 7)
shows similarity to that speculated upon by Koppenol and
Klasinc (31). The initial process is rotation about the 0-N
bond of peroxynitrous acid towards the rotamer with the HO
oxygen orthogonal to the nitro (ONO) plane, in simile with the
thionitrate rearrangement mechanism. However, in this reaction further dislocation of the oxygen from the nitro plane is
required to permit orbital overlap in the transition state (Fig.
6). The underestimation of the barrier for rearrangement in
previous work results from the assumption that the initial, simple torsional motion surmounts the barrier to rearrangement.
In fact, the very substantial barrier results from the bond
lengthening and bending processes associated with migration
of oxygen consequent to the initial rotational motion. The free
energy of activation for rearrangement of peroxynitrous acid
at 25C with aqueous solvation corrections applied is 60 kcall
mol. The substantial hypothetical barrier for rearrangement of
nitric acid to peroxynitrous acid (AG' (aq.) = 93 kcallmol)
allows comparison with the rearrangement of thionitrate to
sulfenyl nitrite ( a G i (aq.) = 56 kcallmol). The higher barrier is
compatible with the more extreme geometrical distortion
required to reach the transition state in the peroxynitrite rearrangement.

Rearrangement versus homolysis


It has been clearly demonstrated experimentally that peroxynitrite in aqueous solution isomerizes to nitrate via peroxynitrous acid (33). Equally, it can be simply demonstrated that

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Cameron et al.
alkyl thionitrate spontaneously yields 'NO in aqueous solution. However, the substantial barriers to reaction, calculated
for the concerted rearrangement processes involved, strongly
argue against such concerted rearrangements mediating the
observed experimental phenomena. The free energies of reaction associated with the competitive homolytic processes are
calculated to be low: =17 kcallmol for dissociation of thionitrate to NO, and thiyl radical; and =3 kcallrnol for dissociation of peroxynitrous acid to NO, and hydroxyl radical.
Rearrangement via homolysis and radical recombination
appears a more likely pathway for both reactions. Of course,
the concerted rearrangement processes may be susceptible to
catalysis. Indeed, the limited range of thiols, which in the presence of GTN are able to activate GCase, all possess a p-carboxylate group; for example, thiosalicylic acid (7a, 4. The
ability of a P-carboxylate or carboxylic acid group to participate in intramolecular base, nucleophilic, and acid catalysis is
well documented. We are currently exploring potential catalytic pathways experimentally and through further MO calculations. The present calculations cannot rule out a
nonconcerted ionic or diradical rearrangement. However, the
low free energy for homolysis of the S-0 bond of thionitrate
coupled with that for dissociation of sulfenyl nitrite to 'NO
suggests that a homolysis-recombination pathway is more reasonable (eq. [a], Fig. 4). If separation and diffusion of the initial geminate pair competes with recombination, NO, may be
released before 'NO in both thionitrate and peroxynitrite rearrangements. The question then to be addressed is whether NO,
may mediate some of the biological activity of organic nitrates
and toxicity of 'NO.
[8] R'SNO,

+ {R'S' 'NO,) + R'SONO + R'SO' + 'NO

Conclusions
I . Structures representing thionitrates and conformers of sulfenyl nitrites are of similar energy at the higher levels of calculation employed, including electron correlation at the
MP4SDQIMP2 and MP2 levels. Structures corresponding to
sulfinyl nitrites do not represent stable equilibrium geometries
at the MP2 level of calculation.
2. Free energies of reaction for homolysis of thionitrates,
sulfenyl nitrites, and peroxynitrous acid, at 298 K, all lie below
17 kcallmol. The free energy for homolysis of nitric acid is 29
kcallmol.
3. Structural reorganization to reach the transition states for
concerted rearrangement of thionitrate to sulfenyl nitrite and
peroxynitrous acid to nitric acid requires destruction of the
ONOX plane of the ground states (X = S, 0 ) .
4. The energy barriers for concerted rearrangement of peroxynitrous acid and thionitrate are substantial: AG?(aq.) = 60
kcallrnol and 56 kcallrnol, respectively.
5. The small energies of homolysis and the large activation
barriers to concerted rearrangement suggest that the most reasonable mechanism for rearrangement involves homolysis followed by radical recombination. In this case, NO, is in initial
product of reaction.

Acknowledgments
This work was supported by the Heart and Stroke Foundation

of Ontario (Grant no. A2259) and Queen's University. Felicity


Wowk, Jen Barbour, Jodi Lock, and Christine Robb are
thanked for preliminary calculations.

References
1. J.S. Stamler, D.J. Singel, and J. Loscalzo. Science, 258, 1898
(1992); H-J Galla. Angew. Chem. Int. Ed. Engl. 23, 378 (1993);
T.G. Traylor and V.S. Sharma. Biochemistry, 31, 2847 (1992).
2. D.A. Wink, R.W. Nims, J.F. Darbyshire, D. Christodoulou, I.
Hanbauer, G.W. Cox, F. Laval, J. Laval, J.A. Cook, M.C.
Krishna, W.G. DeGraff, and J.B. Mitchell. Chem. Res. Toxicol.
7, 5 19 (1994); J.A. Hrabie, J.R. Klose, D.A. Wink, and L.K.
Keefer. J. Org. Chem. 58, 1472 (1993); S.A. Lipton, Y-B Chol,
Z-H Pan, S.Z. Lei, H-S. V. Chen, N.J. Sucher, J. Loscalzo, D.J.
Singel, and J.S. Stamler. Nature, 364, 626 (1993); S. Tamir,
R.S. Lewis, TdeR. Walker, W.M. Deen, J.S. Wishnok, and S.R.
Tannenbaum. Chem. Res. Toxicol. 6,895 (1993).
3. J. Abrams. Am. J. Cardiol. pp. IB-3B (1992), and following
articles.
4. (a) L.J. Ignarro, H. Lippton, J.C. Edwards, W.H. Baricos, A.L.
Hyman, P.J. Kadowitz, and C.A. Grueter. J. Pharmacol. Exp.
Ther. 218, 739 (1981); ( 0 ) G.S. Marks. J. Physiol. Pharmacol.
65, 1111 (1987); (c) R.A. Yeates. Arzneim-Forsch. 42, 1314
(1992).
5. B.M. Bennett. Trends Pharmacol. Sci. 15, 245 (1994); W.R.
Kukovetz and S. Holzman. Eur. J. Clin. Pharmacol. 38, S9
(1990).
6. M. Feelisch, M. te Poel, R. Zamora, A. Deussen, and S.
Moncada. Nature, 368, 62 (1994); S. Moncada, R.M.J. Palmer,
and E.A. Higgs. Pharmacol. Rev. 43, 109 (1 99 1 ).
7. (a) M. Feelisch and E. Noack. Eur. J. Pharmacol. 142, 465
(1987); (6) R.A. Yeates, M. Schmid, and M. Leitold. Biochem.
Pharmacol. 38, 1749 (1989); (c) M.A. Kurz, T.D. Boyer, R.
Whalen, T.E. Peterson, and D.G. Harrison. Biochem. J. 292,
545 (1993); (d) S. Chong and H-L. Fung. Biochem. Pharmacol.
42, 1433 (1991).
8. C. Romanin and W.R. Kukovetz. J. Mol. Cell Cardiol. 20, 389
(1988).
9. M. Feelisch. J. Cardiovasc. Pharmacol. 17 (S3), 25 ( 1991).
10. W.H. Koppenol, J.J. Moreno, W.A. Pryor, H. Ischiropoulos, and
J.S. Beckman. Chem. Res. Toxicol. 5, 834 (1992).
11. M.J.S. Dewar, E.G. Zoebisch, E.F. Healy, and J.J.P. Stewart. J.
Am. Chem. Soc. 107, 3902 (1985).
12. J.J.P. Stcwart. J. Comput. Chem. 10, 209 (1989); 10, 221
(1 989).
13. M.J. Frisch, G.W. Trucks, M. Head-Gordon, P.M.W. Gill, M.W.
Wong, J.B. Foresman, B.G. Johnson, H.B. Schlegel, M.A.
Robb, E.S. Replogle, R. Gomperts, J.L. Andes, K.
Raghavachari, J.S. Binkley, C. Gonzalez, R.L. Martin, and
D.J. GAUSSIAN 92, Revision C, Gaussian Inc., Pittsburgh, Pa.
1992.
14. M.M. Francl, W.J. Pietro, W.J. Hehre, J.S. Binkley, M.S. Gordon, D.J. DeFrees, and J.A. Pople. J. Am. Chem. Soc. 104,
5039 (1982); R. Krishnan, M.J. Frisch, and J.A. Pople. J. Chem.
Phys. 72, 4244 (1980); J.S. Binkley and J.A. Pople. Int. J.
Quantum. Chem. 9, 229 (1975).
15. J.P. Foster and F. Weinhold. J. Am. Chem. Soc. 102, 721 1
(1980); A.E. Reed and F. Weinhold. J. Chem. Phys. 78, 4066
(1983); A.E. Reed, R.B. Weinstock, and F. Weinhold. J. Chem.
Phys. 83, 735 (1985); J.E. Carpenter and F. Weinhold. J. Mol.
Struct. (Theochem), 169,41 ( 1988).
16. L.E. Chirlian and M.M. Francl. J. Comput. Chem. 8, 894
(1987).
17. C.J. Cramer and D.G. Truhlar. J. Comput. Chem. 13, 1089,
(1992); J. Cornput.-Aided Mol. Des. 6, 629 (1992).
18. S.L. Chan and C. Lim. J. Phys. Chem. 98,692 (1994).

Can. J. Chem. Downloaded from www.nrcresearchpress.com by 93.138.101.25 on 08/04/16


For personal use only.

Can. J. Chem. Vol. 73, 1995

19. C. Gonzalez and H.B. Schlegel. J. Phys. Chem. 90,2154 (1989).


20. W.J. Hehre, L. Radom, P.v.R. Schleyer, and J.A. Pople. Ab initio
molecular orbital theory. John Wiley & Sons, Toronto. 1986.
21. ( a )T.J. Lee and J.E. Rice. J. Am. Chem. Soc. 114, 8247 (1992);
(b) M.P. McGrath, M.M. Francl, F.S. Rowland, and W.J. Hehre.
J. Phys. Chem. 92,5353 (1988).
22. D.R. Cameron. Ph.D. Thesis, Queen's University, 1994.
23. D. Yu, A. Rauk, and D.A. Armstrong. J. Phys. Chem. 96, 6031
(1992).
24. I.B. Burkholder, P.D. Hammer, and C.J. Howard. J. Phys.
Chem. 91, 2136 (1987); D.J. Benton and P. Moore. J. Chem.
Soc. A, 3179 (1970).
25. B.M. Bennett, B.T. McDonald, and M.J. St James. J. Pharmacol.
Exp. Ther. 261, 716 (1992); H. Schroder. J. Pharmacol. Exp.
Ther. 262, 298 (1992).
26. E. Noack and M. Feelisch. Basic Res. Cardiol. 86 (S2), 37
(1991).

27. Y.J. Park, H.H. Shin, and Y.H. Kim. Chem. Lett. 1483 (1992);
R.M. Toppin and N. Kharasch. J. Org. Chem. 27, 4353 (1962);
S. Oae and K. Ikura. J. Chem. Soc. Jpn. 38,58 (1965).
28. C. Chatgilialoglu. In The chemistry of sulfones and sulfoxides.
Edited by S. Patai and C.J.M. Rapaport. Wiley, New York.
1988. Chap. 24.
29. S. Oae, K. Shinhama, K. Fujimori, and H.K. Yong. Bull. Chem.
Soc. Jpn. 53,775 (1980).
30. P.L. Folkins and D.N. Harpp. J. Am. Chem. SOC.113, 8998
(199 l), and references therein.
31. W.H. Koppenol and L. Klasinc. Int. J. Quantum Chem. S20, 1
(1993).
32. R. Atkinson, W.P.L. Carter, and A.M. Winer. J. Phys. Chem. 87,
2012 (1983).
33. M. von Gretzel, A. Henglein, and S. Taniguchi. Ber. BunsenGes. Phys. Chem. 74,292 (1970).

You might also like