The kinetics of thiyl radical-induced reactions of monounsaturated fatty acid esters

Article abstract of DOI:10.1021/ja027428d

The time-dependent isomerizations and thiol additions of several Z- and E-monounsaturated fatty acid methyl esters catalyzed by alkanethiyl radicals during γ-radiolysis of tert-butyl alcohol solutions are analyzed on the basis of the radiation chemical yield of radicals and established rate data. This provides room-temperature rate constants for the reversible thiyl addition. Within experimental errors, they do not depend on the double bond position in the alkyl chains. Particularly noteworthy is the very fast β-elimination of thiyl radicals from alkyl radicals which carry a second β-substituent. It is supported by additional evidence obtained with a radical clock methodology, and the large preference of fragmentation to the E-isomers is attributed to different barriers for the formation of the E- and Z-transition states from the equilibrium radical structure.

Full text of DOI:10.1021/ja027428d

Published on Web 10/08/2002
The Kinetics of Thiyl Radical-Induced Reactions of
Monounsaturated Fatty Acid Esters
Chryssostomos Chatgilialoglu,* Alessio Altieri, and Hanns Fischer*,†
Contribution from the ISOF, Consiglio Nazionale delle Ricerche,
Via P. Gobetti 101, 40129 Bologna, Italy
Received June 24, 2002
Abstract: The time-dependent isomerizations and thiol additions of several Z- and E-monounsaturated
fatty acid methyl esters catalyzed by alkanethiyl radicals during γ-radiolysis of tert-butyl alcohol solutions
are analyzed on the basis of the radiation chemical yield of radicals and established rate data. This provides
room-temperature rate constants for the reversible thiyl addition. Within experimental errors, they do not
depend on the double bond position in the alkyl chains. Particularly noteworthy is the very fast â-elimination
of thiyl radicals from alkyl radicals which carry a second â-substituent. It is supported by additional evidence
obtained with a radical clock methodology, and the large preference of fragmentation to the E-isomers is
attributed to different barriers for the formation of the E- and Z-transition states from the equilibrium radical
structure.
Scheme 1
Introduction
The radical-catalyzed addition of thiols to unsaturated CC-
bonds is a well-established chain reaction and has found many
synthetic applications.1,2 Thiol additions to Z- and E-alkenes
are accompanied by isomerizations, and this has led to the
general mechanism shown in Scheme 1 for a case including
isomerization.
Sivertz et al.3 and Walling et al.4 found early on that the rate
constants for the addition of thiyl radicals to the Z- and
E-isomers differ from one another as do the fragmentation rate
constants of the intermediate carbon-centered thiyl adduct
radicals to the Z- and E-compounds, respectively. At small thiol
concentrations, the isomerization is much faster than the thiol
adduct formation. Hence, thiyl radicals efficiently catalyze the
conversion of the naturally occurring Z-forms to the equilibrium
of Z- and E-configurations of mono- or polyunsaturated fatty
acid esters in which the thermodynamically more stable E-forms
dominate.5
Recently, the thiyl radical-catalyzed isomerization has re-
ceived additional attention because it may contribute to the Z-
to E-conversion of natural unsaturated phospholipids in cell
membranes and cause detrimental structural changes.6-8 In fact,
the isomerization was observed for mono- and polyunsaturated
species under biomimetic conditions, that is, using organized
systems such as lipid vesicles and biologically relevant thiols
such as glutathione.6 This process, which does not cause lipid
degradation but a permanent modification of these membrane
constituents, can ultimately influence the barrier properties and
functions of biological membranes.
So far, there is little information on the rate constants of the
individual reactions of Scheme 1 which could help to model
(6) Ferreri, C.; Costantino, C.; Perrotta, L.; Landi, L.; Mulazzani, Q. G.;
Chatgilialoglu, C. J. Am. Chem. Soc. 2001, 123, 4459. Chatgilialoglu, C.;
Ferreri, C.; Ballestri, M.; Mulazzani, Q. G.; Landi, L. J. Am. Chem. Soc.
2000, 122, 4593. Ferreri, C.; Costantino, C.; Landi, L.; Mulazzani, Q. G.;
Chatgilialoglu, C. Chem. Commun. 1999, 5, 407. Chatgilialoglu, C.;
Ballestri, M.; Ferreri, C.; Vecchi, D. J. Org. Chem. 1995, 60, 3826.
(7) Adhikari, S.; Sprinz, H.; Brede, O. Res. Chem. Intermed. 2001, 4, 5, 549.
Sprinz, H.; Schwinn, J.; Naumov, S.; Brede, O. Biochim. Biophys. Acta
2000, 1483, 91. Schwinn, J.; Sprinz, H.; Droessler, K.; Leistner, S.; Brede,
O. Int. J. Radiat. Biol. 1998, 74, 359.
(8) (a) The effectiveness of Z-E isomerization by thiyl radicals in the presence
of the most common antioxidants vitamins has also been addressed, see:
Chatgilialoglu, C.; Zambonin, L.; Altieri, A.; Ferreri, C.; Mulazzani, Q.
G.; Landi, L. Free Radical Biol. Med. 2002, in press. (b) For ‚NO2 radical
induces Z-E isomerization of arachidonic acid within cellular phospho-
lidids, see: Jiang, H.; Kruger, N.; Lahiri, D. R.; Wang, D.; Vate`le, J.-M.;
Balazy, M. J. Biol. Chem. 1999, 274, 16235.
* To whom correspondence should be addressed. E-mail: chrys@
area.bo.cnr.it.
† Permanent address: Physikalisch-Chemisches Institut, University of
Zuerich, Winterthurer Strasse 190, 8057 Zuerich, Switzerland.
(1) Walling, C. Free Radicals in Solution; Wiley: New York, 1957.
(2) Chatgilialoglu, C.; Bertrand, M. P.; Ferreri, C. In S-Centered Radicals;
Alfassi, Z. B., Ed.; Wiley: New York, 1999. Chatgilialoglu, C.; Guerra,
M. In Supplement S: The Chemistry of Sulphur-Containing Functional
Groups; Patai, S., Rappoport, Z., Eds.; Wiley: New York, 1993.
(3) Pallen, R. H.; Sivertz, C. Can. J. Chem. 1957, 35, 723. Graham, D. M.;
Mieville, R. L.; Sivertz, C. Can. J. Chem. 1964, 42, 2239. Graham, D. M.;
Mieville, R. L.; Pallen, R. H.; Sivertz, C. Can. J. Chem. 1964, 42, 2250.
(4) Walling, C.; Helmreich, W. J. Am. Chem. Soc. 1959, 81, 1144.
(5) Sgoutas, D. S.; Kummerow, F. A. Lipids 1969, 4, 283.
9
12816
J. AM. CHEM. SOC. 2002, 124, 12816-12823
10.1021/ja027428d CCC: $22.00 © 2002 American Chemical Society

Reactions of Monounsaturated Fatty Acid Esters
A R T I C L E S
Scheme 2
The yield of the disulfide amounts to one-half of the thiyl radical
yield, and two experiments with thiol consumptions of at most
13% led to G(RS‚) ) (0.98 ( 0.04) × 10-6 mol/J, which is
significantly larger than G ) 0.65 × 10-6 mol/J assumed
before,6 and we expect the same value for â-mercaptoethanol.
The thiyl radical yield provides the radical generation rate Ri
from the dose rate D˙ in Gy/min and the solvent density F )
0.775 kg/dm3 as Ri ) D˙ GF/60. In the studies described below,
the total doses did not exceed 4 kGy and led to a maximum
thiol consumption by the thiyl self-termination of 3 × 10-3 M.
This is small as compared to the minimum initial thiol
concentration of 4.3 × 10-2 M. Also, the thiol consumption by
thiol adduct formation was much smaller than the initial thiol
concentration; that is, the thiol consumption was always
negligible, and Ri was constant.
From the radical generation rate Ri and the known thiyl self-
termination rate constant10 2kt ) 3 × 109 M-1 s-1, one obtains
the steady-state concentration of thiyl radicals [RS‚] ) (Ri/2kt)1/2.
In the course of this work, the dose rate decreased from 23.0 to
16.8 Gy/min. Therefore, Ri decreased from about 3.0 × 10-7
to 2.2 × 10-7 M/s, and the steady-state thiyl radical concentra-
tion decreased from 10 × 10-9 to 8.6 × 10-9 M. This was taken
into account in the analysis.
biological processes. Furthermore, the literature is in part
controversial. The product distribution study of Walling et al.4
suggests quite large rate constants kf . 107 s-1 for the
â-elimination of the CH3S‚ radical from its adducts to 2-butenes,
whereas other authors7,9 attribute at least 2 orders of magnitude
smaller values to kf for the adducts of several alkanethiyl species
to mono- and polyunsaturated fatty acid esters. Moreover, in
contrast to the earlier views,1-4 equal rate constants were
attributed to the different addition and fragmentation reactions.7,9
To obtain more kinetic information, we have extended our
previous studies of thiyl radical-induced isomerizations.6 This
work presents the kinetics and the product yields of isomeriza-
tion and thiol adduct formation for a variety of Z- and
E-monounsaturated fatty acid esters in tert-butyl alcohol solution
at room temperature and for a wide range of thiol concentrations.
As thiol, â-mercaptoethanol was used in most cases. 1-Butane-
thiol was also employed and provided practically identical
results. The choice of these nonbiologically relevant thiols was
obligatory to quantify the corresponding adducts and the
disulfides (Scheme 1). The reactions were initiated by continu-
ous 60Co γ-radiation of deoxygenated and N2O saturated
solutions. This generates thiyl radicals according to Scheme 2,
and their radiation chemical yield was measured. Using this and
the established rate constants for the self-termination of al-
kanethiyl radicals and for the hydrogen abstraction from
alkanethiols, the data provide the rate constants of addition and
fragmentation. The latter are rather large but are supported by
the fragmentation rates of independently produced â-alkanethio
substituted alkyl radicals as determined from the product yields
of competing reactions.
As mentioned above, the rate constant for the hydrogen atom
transfer from the thiol to the primary alkyl radical ‚CH2(CH3)2-
COH must be rather large. In fact, several authors11,12 have
determined rate constants very close to 1.0 × 107 M-1 s-1 for
the reaction between alkanethiols and primary, secondary, and
tertiary alkyl radicals and biradicals at room temperature. The
Arrhenius parameters for the reaction of primary alkyl radicals
with t-BuSH are log(A/M-1 s-1) ) 8.15(0.18) and Ea ) 1.86-
(0.23) kcal/mol.12 Hence, the rate of hydrogen abstraction from
the thiol by ‚CH2(CH3)2COH is about kSH[RSH] ) 7.5 × 105
s-1 for [RSH] ) 0.075 M, and this is much larger than the
overall termination rate (2ktRi)1/2 ≈ 30 s-1
.
Isomerization Mechanism and Kinetics. The product
distribution obtained by GC analysis after radiolysis of tert-
butyl alcohol solutions of monounsaturated fatty acid esters (0.15
M) containing 0.043 to 1.3 M â-mercaptoethanol agrees very
well with the mechanism of Scheme 1. The products were the
isomer of the starting ester and the thiol adduct, and the yield
of the latter increased with increasing thiol concentration. The
disulfide formed by coupling of thiyl radicals appeared as minor
product but in the concentrations expected from the thiyl radical
radiation chemical yield. Traces of unidentified species amounted
to at most 1% of the total products.
Results and Discussion
Radiation Chemical Yield and Thiyl Radical Concentra-
tions. The analysis of the isomerization kinetics required the
knowledge of the thiyl radical concentrations during the
continuous reactions. It was calculated from their radiation
chemical yield. To determine this, deoxygenated and N2O
saturated 1-butanethiol solutions in tert-butyl alcohol were
exposed to total doses of 6.5 and 18.1 kGy. Butyl disulfide
RSSR was the only product found by GC analysis. It is formed
by the coupling of 1-butanethiyl radicals, and blank experiments
ensured its formation only in the presence of radiation. Appar-
ently, the primary species formed by the radiolysis abstract
hydrogen atoms from the thiol much faster than they terminate.
Figure 1 displays the development of the isomer concentra-
tions as functions of the radiation dose during irradiations of
four Z-compounds methyl oleate, methyl palmitoleate, methyl
Z-vaccenate, and oleic acid 0.15 M in tert-butyl alcohol and in
the presence of â-mercaptoethanol in low concentration (0.075
M). Within the experimental errors, the four compounds
approach the same equilibrium of isomers within the same time;
that is, the equilibrium and the rate constants are practically
(10) Self-termination constants of aliphatic thiyl radicals in liquid solution are
close to the diffusion controlled limit, for example, t-BuS‚, 2kt ) (2-4) ×
109 M-1 s-1 at room temperature in n-hexane. Ito, O.; Matsuda, M. Bull.
Chem. Soc. Jpn. 1978, 51, 427.
(9) Schoeneich, C.; Asmus, K.-D.; Dillinger, U.; von Bruchhausen, F. Biochem.
Biophys. Res. Commun. 1989, 161, 113. Schoeneich, C.; Bonifacic, M.;
Dillinger, U.; Asmus, K.-D. In Sulfur Centered ReactiVe Intermediates in
Chemistry and Biology; Chatgilialoglu, C., Asmus, K.-D., Eds.; Plenum
Press: New York, 1990; p 367. Schoeneich, C.; Dillinger, U.; von
Bruchhausen, F.; Asmus, K.-D. Arch. Biochem. Biophys. 1992, 292, 456.
(11) Tronche, C.; Martinez, F. N.; Horner, J. H.; Newcomb, M.; Senn, M.; Giese,
B. Tetrahedron Lett. 1996, 37, 5845. Enzinas, M. V.; Wagner, P. J.; Scaiano,
J. C. J. Am. Chem. Soc. 1980, 102, 1357.
(12) Newcomb, M.; Choi, S.-Y.; Horner, J. H. J. Org. Chem. 1999, 64, 1225.
9
J. AM. CHEM. SOC. VOL. 124, NO. 43, 2002 12817

A R T I C L E S
Chatgilialoglu et al.
Figure 3. Isomerization of E- and Z-vaccenate (0.15 M) by â-mercapto-
ethanol derived thiyl radicals in tert-butyl alcohol at a low thiol concentration
[RSH] ) 0.075 M. For solid lines, see text.
Figure 1. Isomerization of different monounsaturated fatty acid methyl
esters (0.15 M) by â-mercaptoethanol derived thiyl radicals in tert-butyl
alcohol starting from the Z-isomers at a low thiol concentration [RSH] )
0.075 M. 2, oleate; 9, palmitoleate; [, Z-vaccenate; b, oleic acid. For
solid lines, see text.
Figure 4. Isomerization of methyl oleate and methyl elaidate (0.15 M) by
â-mercaptoethanol derived thiyl radicals in tert-butyl alcohol for [RSH] )
0.38 M. For solid lines, see text.
Figure 2. Isomerization of methyl oleate (0.15 M) by â-mercaptoethanol
derived thiyl radicals in tert-butyl alcohol and formation of the thiol adduct
for [RSH] ) 0.516 M. For solid lines, see text.
equal. At the low thiol concentration, the total concentration of
the unsaturated compounds remains approximately constant.
This means that the fragmentation of the intermediate radical
A‚ is faster than its reaction with the thiol to the adduct AH.
For larger thiol concentrations, there is loss of unsaturation.
It is shown in Figure 2 for the reaction of methyl oleate in the
presence of 0.516 M â-mercaptoethanol. As it was found also
in many other cases, the loss of unsaturated esters is exactly
balanced by the formation of the thiol adduct. Figures 3 and 4
show that the same equilibria are reached in practically equal
times if one starts either from the Z- or from the E-isomers
both for small and for appreciable thiol concentrations. Finally,
Figure 5 compares the isomerizations of methyl oleate at two
thiol concentrations as another example for the increasing loss
of unsaturation at high thiol concentrations.
All of this supports the mechanism of Scheme 1, but side
reactions are also at work. They give the same products but
change the rates. Thus, for the rather high thiol concentrations
of 0.30 and 0.93 M, blank experiments on nonirradiated samples
showed some isomerization. This indicates an unknown ad-
ditional thermal generation of thiyl radicals. Furthermore, several
runs exhibited a small induction period such as those seen in
Figures 3 and 5. It is probably due to thiyl radical removal by
unknown and minor scavenging impurities. Because of their
Figure 5. Isomerization of methyl oleate (0.15 M) by â-mercaptoethanol
derived thiyl radicals in tert-butyl alcohol at thiol concentrations [RSH] )
0.2 M and [RSH] ) 1.37 M. For solid lines, see text.
spurious appearance, the side reactions are ignored in the
following analysis of the isomerization kinetics but lead to
appreciable errors of the resulting rate constants.
According to the product distribution and supported by the
large rate constant for the hydrogen abstraction, the only relevant
radical termination is the bimolecular coupling of the thiyl
radicals RS‚. Hence, the concentrations of the primary 2-hy-
9
12818 J. AM. CHEM. SOC. VOL. 124, NO. 43, 2002

Reactions of Monounsaturated Fatty Acid Esters
A R T I C L E S
droxypropyl substituted methyl radicals X‚ and of the carbon-
centered thiyl radical adducts A‚ are much smaller than the
concentration of RS‚.
These were used by Walling and Helmreich4 to obtain kSH/kf
for the isomerizations of Z- and E-2-butene by methyl thiyl
radicals. However, because of the small induction periods,
extrapolations of reaction rates to zero time were not practical
here, and a different analysis was used.
The concentration of A‚ obeys
d[A‚]
dt
For reactions starting from the pure Z-isomer, the integration
of eqs 4a and 4b provides
) (kZa[Z] + kEa [E])[RS‚] - (kfZ + kEf )[A‚] -
kSH[RSH][A‚] (1)
and in the steady state, one has
k11 - λ
k11 - λ
λ1 - λ2
[Z] )
2 exp(-λ1t) -
1 exp(-λ2t) ‚[Z]0 )
[
]
λ1 - λ2
[b exp(-λ1t) + (1 - b) exp(-λ2t)]‚[Z]0 (6a)
kZa[Z] + kEa [E]
[A‚] )
[RS‚]
(2)
k21
kZf + kEf + kSH[RSH]
[E] )
[exp(-λ2t) - exp(-λ1t)]‚[Z]0 )
λ1 - λ2
The concentrations of the unsaturated fatty acid esters obey the
rate equations
a[exp(-λ2t) - exp(-λ1t)]‚[Z]0 (6b)
and for initially pure E-isomer
d[Z]
dt
) -kZa[Z][RS‚] + kZf [A‚]
) -kEa [E][RS‚] + kEf [A‚]
(3a)
(3b)
k12
[Z] )
[exp(-λ2t) - exp(-λ1t)]‚[E]0 )
λ1 - λ2
d[E]
dt
c[exp(-λ2t) - exp(-λ1t)]‚[E]0 (7a)
k22 - λ
λ1 - λ2
k22 - λ
λ1 - λ2
2 exp(-λ1t) -
1 exp(-λ2t) ‚[E]0 )
or with eq 2
[E] )
[
]
kZa(kEf + kSH[RSH])[RS‚]
kZf + kEf + kSH[RSH]
[d exp(-λ1t) + (1 - d) exp(-λ2t)]‚[E]0 (7b)
d[Z]
dt
) -
[Z] +
where
kEa kZf [RS‚]
kZf + kEf + kSH[RSH]
λ1,2 ) 1/2(k11 + k22) ( (1/4(k11 - k22)2 + k12k21)1/2 (8)
[E] ) -k11[Z] + k12[E] (4a)
Hence, the time dependencies of the isomer concentrations
are sums of two exponential functions. Such functions were
constructed to reproduce the measured concentrations [Z] and
[E] as functions of the dose, allowing for the induction when
necessary. Examples are shown in Figures 1-5. This yielded
the dimensionless parameters a-d and λ1 and λ2 per dose units.
The latter were converted to reciprocal time units via 1 kGy ≡
6 × 104D˙ t with the dose rate in Gy/min and t in s.
kZakEf [RS‚]
kZf + kEf + kSH[RSH]
kEa (kZf + kSH[RSH])[RS‚]
kZf + kEf + kSH[RSH]
d[E]
dt
)
[Z] -
[E] ) k21[Z] - k22[E] (4b)
The total concentration of the isomers decreases according to
For small thiol concentrations [RSH] < 0.1 M, the total
isomer concentration [Z] + [E] is practically constant (Figures
1, 3), and the fragmentation of the thiyl radical adduct occurs
much faster than the hydrogen abstraction from the thiol. Hence,
the hydrogen abstraction can be neglected. One then has λ2 ≈
0, and
(kZa[Z] + kEa [E])[RS‚]
d([Z] + [E])
) -
kSH[RSH] (4c)
kZf + kEf + kSH[RSH]
dt
and it is exactly matched by the thiol adduct formation. When
a decay of the total unsaturation was observed, that is, for thiol
concentrations larger than about 0.1 M (Figures 2 and 5), it
vastly exceeded the maximum formation of the disulfide of 3
mM, and, therefore, the loss of unsaturation also equals the net
loss of thiol.
[E]∞ kZa kEf
K Z
K )
)
)
(9a)
(9b)
kZf kEa K E
[Z]∞
kZakEf + kEa kZf
kEf + kZf
λ1 )
[S‚] ) kis[S‚]
Equations 4 lead to expressions for initial rate ratios and
kSH[RSH]
d[RSH]
d([Z] + [E])
d[E]
Equations 6 and 7 become monoexponential, and fits to these
provide the equilibrium constant K and an effective isomeriza-
tion rate constant kis ) λ1/[RS‚]. The latter is the average of
the addition constants weighted by the fragmentation ratios.
However, K and kis alone do not lead to the individual addition
and fragmentation rate constants.
) -
) -
) -
(
)
(
(
)
kfE
d[E]
tf0
tf0
tf0
tf0
for [E]0 ) 0 (5a)
kSH[RSH]
d[RSH]
d([Z] + [E])
d[Z]
) -
(
)
)
kfZ
d[Z]
For [RSH] > 0.1 M, the hydrogen abstraction reaction
competes with fragmentation. For this case, one derives from
for [Z]0 ) 0 (5b)
9
J. AM. CHEM. SOC. VOL. 124, NO. 43, 2002 12819

A R T I C L E S
Chatgilialoglu et al.
M), palmitoleate, Z-vaccenate, oleic acid, and â-mercaptoethanol
([RSH] ) 0.075 M) and starting from E-isomers methyl elaidate,
E-vaccenate, and â-mercaptoethanol ([RSH] ) 0.38 and 0.075
M, respectively) were analyzed. Essentially the same rate
constants were obtained for the different starting compounds.
The reproducibilities of the equilibrium constant K and of
the parameters a, b, c, and d were excellent, but λ1 and λ2
showed a considerable scatter. Especially for low and for high
thiol concentrations did outliers from averages occur, but, as
was mentioned above, in these cases side reactions are likely.
Hence, we present in Table 1 the average rate constants obtained
from a set of 10 experiments with Z-isomers without the
apparent outliers and from three runs with E-isomers. The same
average rate constants resulted from the two sets.
â-Elimination of Thiyl Radicals. Support for the results is
obtained from the fragmentation constants kf of â-alkanethio
substituted alkyl radicals measured relative to a rate constant
other than the hydrogen abstraction from the thiol with a free
radical clock methodology.13 For that purpose, the reactions of
Scheme 3 were used.6
For sufficiently high concentrations of (TMS)3SiH, the radical
2 partitions between the hydrogen abstraction kH and the
â-elimination kf. The hydrogen transfer from (TMS)3SiH to the
thiyl radical is efficient. Hence, the ratio of the concentrations
of the products alkane and thiol reflects the ratio of rate
constants.
Figure 6. Competition between hydrogen abstraction and radical frag-
mentation during the isomerization of methyl oleate (0.15 M) versus the
thiol concentration. The linear increase is expected from eq 10a.
eqs 4 and 6-8 the general relations
λ2 + (b - a)(λ1 - λ2) kSH[RSH]
)
for [E]0 ) 0 (10a)
for [Z]0 ) 0 (10b)
kfE
a(λ1 - λ2)
λ2 + (d - c)(λ1 - λ2) kSH[RSH]
)
kfZ
c(λ1 - λ2)
k
[alkane]
[RSH]
)
H[(TMS)3SiH]
kf
(12)
which are formally identical to eqs 5. The rate constant kSH
)
1.0 × 107 M-1 s-1 for the hydrogen transfer from the thiol to
the secondary thiyl adduct radical is known,11,12 and the
application of eqs 10 to the data then yields the fragmentation
constants kEf and kfZ.
As before,6 the yields of alkane and thiol were determined
by GC analysis, following a photochemical radical generation
and using an internal standard.
Furthermore, one notices from Figures 1-5 that the isomer-
ization occurs well before the loss of unsaturation, and this held
in all of the cases studied. It means λ2 , λ1, and, under this
condition, λ1 is still well approximated by eq 9b and can be
used to obtain kis. The isomer ratio at long times provides the
total equilibrium constant K ) a/(1 - b) ) (1 - d)/c (eq 9a),
and λ2 is given by
The radical 2a derived from compound 1a carries methyl
groups in the R- and â-positions and closely resembles the
intermediate A‚ of the fatty acid ester isomerizations. From nine
experiments at 203 K with silane concentrations ranging from
0.2 to 1 M, a linear dependence of the product ratio on the
(TMS)3SiH concentration with negligible intercept was found.
Equation 12 led to kH/kf ) (0.15 ( 0.02) M-1, where kf ) kfE
+ kZf . In the earlier work,6 we found kH/kf ) (0.37 ( 0.09) M-1
from six data at the same temperature, but we consider this value
less reliable now. At temperatures above 203 K, kH/kf could
not be measured because the product ratio increased too little
with increasing (TMS)3SiH concentration. This means that kf
increases more than kH with increasing temperature.
kSH[RSH]
K
1 + K
λ2 )
λ2 )
kEa [RS‚]
kZa[RS‚]
for [E]0 ) 0
for [Z]0 ) 0
(11a)
(11b)
kfE
kSH[RSH]
1
kfZ
1 + K
The Arrhenius relation for the hydrogen abstraction from
(TMS)3SiH by secondary alkyl radicals is log(kH/M-1 s-1) )
8.5-4.3/Θ, where Θ ) 2.303RT in kcal/mol.14 Hence, one has
kH ) 4700 M-1 s-1 at 203 K and obtains kf ) 3.1 × 104 s-1
for 2a from kH/kf. The data given in Table 1 for room
temperature are much larger, but this is in accord with a strong
temperature dependence of the â-elimination.
This temperature dependence was addressed conveniently by
using the same methodology with compound 1b which lacks
the R- and â-alkyl groups. Five series of experiments with silane
concentrations between 0.2 and 1 M led to larger values of kH/
In detail, for reactions starting with the Z-isomer ([E]0 ) 0),
the equilibrium constant K was obtained from the isomer ratios
at long times, and kis followed from λ1. For sufficiently large
thiol concentrations, the fragmentation constant kEf was de-
duced via eq 10a (Figure 6), and the addition constant kEa was
determined using eq 11a. The remaining parameters kEf and kaE
were then extracted from K and kis. Hence, all of the desired
rate constants were obtained alone from experiments employing
the Z-isomers as starting compounds, but the same parameters
also followed from the analogous analysis of reactions starting
with the E-isomers.
(13) Newcomb, M. Tetrahedron 1993, 49, 1151. Griller, D.; Ingold, K. U. Acc.
Chem. Res. 1980, 13, 317.
(14) Chatgilialoglu, C.; Dickhaut, J.; Giese, B. J. Org. Chem. 1991, 56, 6399.
In total, isomerizations starting from Z-isomers methyl oleate
and â-mercaptoethanol or 1-butanethiol ([RSH] ) 0.049-1.37
9
12820 J. AM. CHEM. SOC. VOL. 124, NO. 43, 2002

Reactions of Monounsaturated Fatty Acid Esters
A R T I C L E S
Table 1. Rate Constants for the Isomerization of Monounsaturated Fatty Acid Methyl Esters Catalyzed by Thiyl Radicals in tert-Butyl
Alcohol at Room Temperaturea
runs
K
k /s-1
kE/s-1
kZ/s-1
kEa/M-1 s-1
kZa/M-1 s-1
is
f
f
Z f E
E f Z
10
3
5.1(0.1)
5.1(0.1)
1.4(0.3) × 105
1.6(0.3) × 108
1.4(0.4) × 108
2.2(0.9) × 107
2.2(0.6) × 107
2.0(0.8) × 105
1.5(0.3) × 105
1.1(0.4) × 105
1.4(0.5) × 105
1.1(0.4) × 105
a Errors in brackets represent one standard deviation.
Scheme 3
Scheme 4
phosphatidyl choline in tert-butyl alcohol, K decreases from 4.9
at 22 °C to 4.0 at 54 °C and 3.6 at 71 °C.6 With K ) exp(-
∆G/RT), this provides ∆H ) -5.4 kJ/mol and ∆S ) -5.5
J/molK for the Z- to E-conversion, similar to the differences of
the thermodynamic parameters of the 2-butenes.
â-Elimination of Thiyl Radicals. In the early classic study,
Walling and Helmreich4 related the rate of the azo-initiated
isomerization of 2-butenes by methanethiyl in a 2-butene/butane
mixture at 60 °C to the thiol consumption via eqs 5 and deduced
Table 2. Kinetic Data for the Reaction of Bromide 1 with
(TMS)3SiH in Toluene at Various Temperaturesa
b
-1
T, K
kH/k, M
intercept
f
253
273
294
294c
313
0.82 ( 0.03
0.48 ( 0.06
0.15 ( 0.01
0.18 ( 0.01
0.08 ( 0.01
0.02 ( 0.02
-0.02 ( 0.04
0.02 ( 0.01
0.04 ( 0.01
0.00 ( 0.01
kE/kSH ) 85 and kZ/kSH ) 19.5. With kSH ) 1.0 × 107 M-1 s-1
f
at 25 °C and Ea )f1.86(0.23) kcal/mol,12 one derives kSH ) 1.4
× 107 M-1 s-1 at 60 °C, and, hence, kEf ) 11.9 × 108 s-1 and
kZf ) 2.7 × 108 s-1. As expected, these rate constants are larger
than those of the â-eliminations of the monounsaturated fatty
acid esters at room temperature (Table 1) but support the latter.
Further, the selectivity of the fragmentation for the 2-butenes
kEf /kfZ ) 4.4 at 60 °C4 is smaller than kfE/kZf ) 7.2 observed at
room temperature for the esters.17
a Errors represent one standard deviation. Range of silane concentrations
employed 0.2-1.0 M in at least five experiments. b Photochemical initiation.
c Et3B/O2 initiation.
kf which decrease with increasing temperature. They are given
in Table 2 and provide the Arrhenius expression log((kH/kf) M-1
)
kEf . kZf indicates that the barrier for the fragmentation to the
more stable E-isomers is smaller than that to the less stable
Z-alkenes. The molar enthalpy difference of the isomers of ∆H
) hf(E) - hf(Z) ) -(4-5) kJ/mol is small. It may contribute
to the barrier difference, but this may also be explained by the
different energies required for the formation of the transition
states from the equilibrium structure of the intermediate radical
A‚. It carries one R- and two â-substituents, and such species
adopt average conformations in which space demanding sub-
stituents avoid each other (Scheme 4). In the transition state
for the â-elimination, the S-atom of the thiyl moiety eclipses
the pZ-axis of the unpaired electron orbital, and two structures
are possible. The less energy demanding leads to the E- and
the more costly one to the Z-configuration of the product, and
consequently the formation of the E-isomer is preferred.
The steric effect also suggests â-alkanethio substituted alkyl
radicals such as radicals A‚ and 2a with a second â-substituent
fragment considerably faster than species lacking this group such
as 2b, in accord with the experimental findings. R-Substituents
) (-5.4 ( 0.4) + (6.2 ( 0.6)/Θ. Combination with log(kH/
M-1 s-1) ) (8.9 ( 0.4) - (4.5 ( 0.5)/Θ for the hydrogen
abstraction from the silane by primary alkyl radicals14 provides
log(kf/s-1) ) (14.2 ( 0.8) - (10.6 ( 1.1)/Θ. The frequency
factor is in the expected range for a unimolecular decomposition,
and the rather larger activation energy supports the strong
temperature dependence of the fragmentation indicated above.
For the â-elimination from the primary alkyl radical 2b, one
extrapolates kf ) 560 s-1 at 203 K. For the secondary alkyl
radical 2a, we found kf ) 3.1 × 104 s-1 at the same temperature;
that is, its fragmentation is 56 times faster. This is probably
due to steric strain effects of the additional alkyl groups (see
below). If the same rate ratio holds at room temperature, the
rate constant kf,298 ) 2.5 × 106 s-1 for the primary species 2b
yields kf,298 ) 1.5 × 108 s-1 for the secondary 2a. From the
isomerization kinetics, we deduced kEf + kfZ ) (1.8 ( 0.3) ×
108 s-1 for such species at room temperature (Table 1), in
excellent agreement with the extrapolated value for 2a.
Discussion of the Isomerization Rate Data. Equilibrium
Constant. K ) 5.1 ( 0.1 at room temperature agrees well with
our earlier findings,6 and K ) 4.3 was reported for the
equilibrium of methyl oleate and methyl elaidate in methanol
by other authors.7 For the structurally related 2-butenes, one
calculates K ) 3.4 at room temperature from the gas-phase
molar enthalpy and entropy differences of the Z- and E-isomers
∆H ) -4.4 kJ/mol and ∆S ) -4.3 J/mol K.3,15,16 Obviously,
the equilibrium constant of the monounsaturated fatty acid esters
is slightly larger than that of the 2-butenes presumably because
of the longer alkyl chain residues. Furthermore, for dioleoyl
(15) Landolt-Boernstein: Zahlenwerte und Funktionen aus Physik, Chemie,
Astronomie, Geophysik und Technik, 6th Ed.; Schaefer, K., Lax, E., Eds.;
Springer: Berlin, 1961; Vol 2, Part 4, p 263. ∆H ) -4.4 kJ/mol: Lias, S.
G.; Bartmess, J. E.; Liebman, J. F.; Holmes, J. L.; Levin, R. D.; Mallard,
W. G. J. Phys. Chem. Ref. Data 1988, 17, Supplement 1.
(16) Graham, D. M.; Soltys, J. F. Can. J. Chem. 1969, 47, 2529. Graham, D.
M.; Soltys, J. F. Can. J. Chem. 1969, 47, 2719. Graham, D. M.; Soltys, J.
F. Can. J. Chem. 1970, 48, 2137.
(17) Boothe, T. E.; Greene, J. L., Jr.; Shevlin, P. B. J. Am. Chem. Soc. 1976,
98, 951 reports kEf /kfZ ) 2.3 at 80 °C and kfE/kZf ) 5.8 at -67 °C for the
elimination of benzenethiyl from its 2-butene adduct. kEf > kfZ is also
evident from the early gas phase work of Sivertz and Graham et al.,3,16
and it is known for the fragmentation of the benzenethiyl radical from its
â-methyl styrene adduct.18
9
J. AM. CHEM. SOC. VOL. 124, NO. 43, 2002 12821

A R T I C L E S
Chatgilialoglu et al.
alone do not lead to rate enhancements but seem to cause a
slower fragmentation which may be due to radical stabilization.
Thus, kf,298 ) 2.7 × 105 s-1 found by Wagner et al.19 for the
elimination of the 1-butanethiyl radical from a 1,4-diradical
carrying an R-CH2R and a â-CH2SR group is smaller than kf,298
) 2.5 × 106 s-1 for 2b which has no R-substituent.
assumed to be irreversible in the observation times of about 50
µs. A competition analysis neglected the differences of the
addition rate constants to the E- and Z-configurations as well
as of the fragmentations, although this would lead to an
unreasonable K ) 1. It provided ka ) 4.5 × 106 M-1 s-1 and
kf ) 3 × 105 s-1 for individual double bonds.7,9 In comparison
to the present results and the literature, ka is by 1 order of
magnitude larger and kf is by nearly 3 orders smaller. Therefore,
the observed inhibition of the formation of the bisallylic radical7
may need a different explanation and experiments to clarify this
point are in progress.
Addition of Thiyl Radicals to Alkenes. There are only very
few literature data with which our addition rate constants can
be compared. Nelson et al.20 report an upper limit kZa < 3.4 ×
106 M-1 s-1 for the gas-phase addition of the methanethiyl
radical to Z-2-butene at room temperature. It agrees with our
data, but the fast fragmentation was not taken into account. Ito
et al.18 followed the addition of arenethiyl radicals to a variety
of alkenes. For the addition of benzenethiyl to monosubstituted
nonconjugated monomers, the rate constants range from 1.7 ×
104 (vinyl acetate) and 4.6 × 105 (acrylonitrile) to 2 × 107 M-1
s-1 (styrene) and exhibit enthalpic plus nucleophilic polar
substituent effects. Alkanethiyls should add faster than the
resonance stabilized benzenethiyl, and this agrees with our data.
It is also noteworthy that the addition of benzenethiyl to the
E-disubstituted alkenes is faster than to the Z-isomers (â-methyl
styrene at room temperature kEa ) 6.8 × 106 M-1 s-1, kaZ ) 2.0
× 106 M-1 s-1, kaE/kZa ) 3.4). We find the same tendency for
the alkanethiyl radical adding to the monounsaturated esters
(kEa /kZa ) 1.3), and a faster addition to E- than to Z-isomers is
also known for C-centered radicals.21
Experimental Section
Materials. The methyl esters of oleic, elaidic, palmitoleic, palm-
itelaidic, Z-vaccenic, E-vaccenic, and stearic acids, â-mercaptoethanol,
1-butanethiol, 2-hydroxyethyl disulfide, butyl disulfide, and tert-butyl
alcohol were commercially available from Aldrich, Fluka, or Sigma
and were used without further purification. 2-Bromoethylthiooctane (1b)
was obtained by reaction of n-octyl thiolate with 2-bromoethanol
followed by bromination,24 whereas threo-2-bromo-3-butylthiobutane
(1a) was obtained by reaction of cis-2-butene with in situ prepared
BuSBr.25 Ethylthiooctane (3b) and 2-butylthiobutane (3a) were obtained
by reaction of bromoethane and 2-bromobutane with the corresponding
thiolate, respectively.24
General Methods. GC analyses for the determination of the isomeric
ratio of the fatty acid methyl esters were performed by using a Varian
CP-3800 equipped with a flame ionization detector. As a stationary
phase, a Rtx-2330 column (60 m × 0.25 mm of 10% cyanopropylphenyl
and 90% biscyanopropyl polysiloxane) was used with helium as carrier
gas (2 mL/min). Column heating to 156 °C for 40 min was followed
by an increase of 10 °C/min up to 250 °C. The methyl esters were
identified by comparison with the retention times of authentic samples.
GC analyses for the determination of the disulfide and/or methy-
loleate-thiol adduct yields were performed by using a Carlo Erba HRGC
5300 equipped with a flame ionization detector. As a stationary phase,
a HP 5 column (30 m × 0.25 mm cross-linked 5% phenylsilicone)
was used with helium as carrier gas (2 mL/min). Column heating to
50 °C for 5 min was followed by an increase of 15 °C/min up to 200
°C.
Finally, Griller et al.22 determined the addition rate constant
of t-BuS‚ to 1-octene at 298 K in isooctane as ka ) 1.9 × 106
M-1 s-1 and confirmed an earlier value of Davies and Roberts.23
Because of the absence of a substituent at the attacked site, the
addition to 1-octene is expected to be faster than that to the
monounsaturated fatty acid methyl esters (Table 1) as it is
common for carbon-centered species.21 Griller et al.22 also
measured kf/kSH ) 0.034 for the 1-octene adduct. If we take
kSH ) 1.0 × 107 M-1 s-1 as before, the fragmentation constant
becomes kf ) 3.4 × 105 s-1, similar to Wagner’s value for the
related elimination from a biradical19 and smaller than the data
of Table 1 because of the lacking â-substituent.
The above literature rate constants for the reactions involved
in the thiyl radical-induced reactions of alkenes support the
results of this work. In particular, the â-elimination of thiyl
radicals from â-alkanethio substituted alkyl radicals is very fast
if these carry a second â-substituent. However, recently quite
different values have been suggested for the related isomeriza-
tions of polyunsaturated fatty acid residues (PUFA). Several
authors7,9 found that thiyl radicals abstract hydrogen atoms at
the bisallylic positions and that this abstraction is increasingly
inhibited by increasing PUFA or other alkene concentrations.7
The latter findings were interpreted by a loss of thiyl radicals
due to the formation of the adduct A‚, and this was implicitly
GC analyses for the kinetic experiments were performed by using a
HP 5890 Series II equipped with a flame ionization detector. As a
stationary phase, a HP 5 column (30 m × 0.25 mm cross-linked 5%
phenylsilicone) was used with helium as carrier gas (2 mL/min).
Column heating to 70 °C was followed by an increase of 15 °C/min
up to 280 °C.
Continuous radiolyses were performed at room temperature (22 (
2) °C on 100 µL samples using a 60Co-Gammacell at different dose
rates. The exact absorbed radiation dose was determined with the Fricke
chemical dosimeter, by taking G(Fe3+) ) 1.61 µmol J-1 26
.
Preparation of Methyl Oleate/HOCH2CH2SH Adduct. A N2O-
saturated tert-butyl alcohol solution (7.2 mL) containing methyl oleate
(1.7 mmol) and HOCH2CH2SH (8.16 mmol) was γ-irradiated at 22 °C
(dose rate 19 Gy/min) overnight. The reaction mixture was evaporated
under vacuum. Silica gel chromatography, using eluent cyclohexane
with increasing amounts of acetone (up to 7.5%), provided 332 mg of
pure adduct (52% yield) oil. 1H NMR (CDCl3): δ 0.88 (t, 3H, J ) 7.6
Hz, CH3), 1.26 (m, 18H, CH2), 1.40 (m, 4H, CH2 γ to the sulfur atom),
1.52 (m, 4H, CH2 â to the sulfur atom), 1.61 (m, 2H, CH2 â to the
carbonyl group), 2.30 (t, 2H, J ) 7.6 Hz, CH2 R to the carbonyl group),
2.58 (quintet, 1H, J ) 6.4 Hz, H R to the sulfur atom), 2.71 (t, 2H, J
) 6 Hz, CH2S), 3.66 (s, 3H, OCH3), 3.70 (t, 2H, J ) 6.4 Hz, CH2O).
(18) Ito, O. In S-Centered Radicals; Alfassi, Z. B., Ed.; Wiley: New York,
1999. Ito, O. In General Aspects of the Chemistry of Free Radicals; Afassi,
Z. B., Ed.; Wiley: New York, 1999. Ito, O.; Matsuda, M. Bull. Chem.
Soc. Jpn. 1978, 51, 427. Ito, O.; Matsuda, M. J. Am. Chem. Soc. 1979,
101, 1815, 5732.
(19) Wagner, P. J.; Sedon, J. H.; Lindstrom, M. J. J. Am. Chem. Soc. 1978,
100, 2579.
(20) Balla, R. J.; Weiner, B. R.; Nelson, H. H. J. Am. Chem. Soc. 1987, 109,
4808.
(21) Fischer, H.; Radom, L. Angew. Chem., Int. Ed. 2001, 40, 1340.
(22) McPhee, D. J.; Campredon, M.; Lesage, M.; Griller, D. J. Am. Chem. Soc.
1989, 111, 7563. Griller, D.; Martinho Simoes, J. A. In Sulfur Centered
ReactiVe Intermediates in Chemistry and Biology; Chatgilialoglu, C., Asmus,
K.-D., Eds.; Plenum Press: New York, 1990; p 327.
(24) Salerni, O. L.; Clark, R. N.; Smart, B. E. J. Chem. Soc. C 1966, 645.
(25) Trost, B. M.; Ziman, S. D. J. Org. Chem. 1973, 38, 932.
(26) Spinks, J. W. T.; Woods, R. J. An Introduction to Radiation Chemistry,
3rd ed.; Wiley: New York, 1990; p 100.
(23) Davies, A. G.; Roberts, B. P. J. Chem. Soc. B 1972, 1830.
9
12822 J. AM. CHEM. SOC. VOL. 124, NO. 43, 2002

Reactions of Monounsaturated Fatty Acid Esters
A R T I C L E S
13C NMR (CDCl3): δ 14.4 (CH3), 22.9, 25.12, 25.14, 26.8, 26.9, 27.0,
29.2, 29.3, 29.7, 29.8, 29.9, 30.0, 32.1, 34.1, 34.3, 35.1 (each CH2),
45.9 (CH), 51.7 (CH3), 60.8 (CH2), 174.6 (CdO). GC/MS m/z, 329
(M+ - CH2CH2OH), 297, 264, 243, 229, 187, 169, 155, 11, 87, 69,
55.
Isomerization of Fatty Acid Methyl Esters in tert-Butyl Alcohol.
A N2O-saturated tert-butyl alcohol solution of monounsaturated fatty
acid methyl ester (0.15 M) containing thiol from 0.075 to 1.5 M was
γ-irradiated at 22 °C (dose rate 17-23 Gy/min). Methyl esters of both
cis and trans isomers and the thiol adduct were examined by GC
analysis in comparison with the retention times of authentic samples.
The saturated methyl stearate was used as the internal standard for
quantitative studies.
calculated by assuming a solution density of 0.775 g/mL (density of
pure tert-BuOH) and quantitative formation of disulfide. G ) (0.97 (
0.04) µmol/J is obtained.
Kinetic Experiments for the â-Elimination of the RS‚ Radical.
Toluene containing a small amount of hexadecane as an internal GC
standard was used as solvent. Bromide 1a or 1b (0.01 M) and
(TMS)3SiH (0.2-1.0 M) were added, and the resulting solutions were
degassed and photolyzed with a 1000 W high-pressure mercury lamp
at the appropriate temperature for a few minutes. Alternatively, the
radical initiator (Et3B/O2) was added via a syringe pump at 21 °C. The
products of interest, that is, compounds 3a and 3b and RSH, were
identified by comparison of their retention times with those of authentic
materials.
Determination of G(RS‚) Value. A N2O-saturated tert-butyl alcohol
solution of 75 mM BuSH was γ-irradiated at 22 °C (dose rate ca. 17
Gy/min). BuSSBu was the only product, and the calibration was
performed by analyzing four standard solutions containing different
ratios of BuSH and BuSSBu. Two experiments were carried out with
a dose of 18.1 and 6.5 kGy, respectively, as determined by Fricke
dosimetry with an augmentation of 2.5% to account for the different
electronic densities of tert-butyl alcohol and water.26 The G-value was
Acknowledgment. H.F. thanks the ISOF, CNR Bologna, for
the hospitality during academic visits in 2001 and 2002. We
thank C. Ferreri and Q. G. Mulazzani for helpful discussions
and M. Ballestri for technical assistence.
JA027428D
9
J. AM. CHEM. SOC. VOL. 124, NO. 43, 2002 12823