Remarkable kinetic stability of α-thiocarbamoyl substituted 4-methoxybenzyl cations

Article abstract of DOI:10.1139/v98-225

4-Methoxybenzyl cations bearing α-(N,N-dimethylcarbamoyl) and α-(N,N-dimethylthiocarbamoyl) substituents have been generated photochemically upon irradiation of precursors with pentafluorobenzoate or 4-methoxybenzoate leaving groups. The ions have been observed with flash photolysis in 40:60 acetonitrile:water and in 50:50 methanol:water, and rate constants were measured for their decay in solvent alone and for their capture by azide ion. The cations so studied and their lifetimes in 40% acetonitrile are 6, ArC⁺H-CONMe₂, 0.6 μs; 2, ArC⁺H-CSNMe₂, 7 ms; and 4, ArC⁺(CH₃)-CSMe₂ 6 ms, where Ar = 4-MeOC₆H₄. The cation 4 reacts with solvent by elimination of a proton from the α-methyl group, and the rate constant for solvent addition must be less than 1 s^-1. The CSNMe₂ substituted cations are 10⁵-10⁷-fold longer lived than analogs where the thioamide group has been replaced with an α-methyl. The UV-visible absorption spectra of these two cations also show significant differences from those of typical 4-methoxybenzyl cations. Thus, both the lifetimes and spectra point to a strong interaction of the benzylic centre with the thioamide group.

Full text of DOI:10.1139/v98-225

1
910
Re m a rka ble kine tic s ta bility of ␣-thioc a rba m oyl
s ubs titute d 4 -m e thoxybe nzyl c a tions
Robe rt A. Mc Cle lla n d, Vic toria E. Lic e nc e , J ohn P. Ric h a rd,
Ka thle e n B. Willia m s , a nd Shrong-Sh i Lin
Abstract: 4-Methoxybenzyl cations bearing α-(N,N-dimethylcarbamoyl) and α-(N,N-dimethylthiocarbamoyl) substituents
have been generated photochemically upon irradiation of precursors with pentafluorobenzoate or 4-methoxybenzoate
leaving groups. The ions have been observed with flash photolysis in 40:60 acetonitrile:water and in 50:50
methanol:water, and rate constants were measured for their decay in solvent alone and for their capture by azide ion.
+
+
The cations so studied and their lifetimes in 40% acetonitrile are 6, ArC H-CONMe , 0.6 µs; 2, ArC H-CSNMe ,
2
2
+
7
ms; and 4, ArC (CH )-CSMe 6 ms, where Ar = 4-MeOC H . The cation 4 reacts with solvent by elimination of a
proton from the α-methyl group, and the rate constant for solvent addition must be less than 1 s . The CSNMe
substituted cations are 10 –10 -fold longer lived than analogs where the thioamide group has been replaced with an α-
3 2, 6 4
–
1
2
5
7
methyl. The UV–visible absorption spectra of these two cations also show significant differences from those of typical
4
-methoxybenzyl cations. Thus, both the lifetimes and spectra point to a strong interaction of the benzylic centre with
the thioamide group.
Key words: flash photolysis, thiocarbamoyl stabilized carbocation, photosolvolysis.
Résumé : L’irradiation de précurseurs appropriés portant des groupes partant pentafluorobenzoate ou 4-
méthoxybenzoate permet de générer photochimiquement des cationns 4-méthoxybenzyle portant des substituants α-
(
N,N-diméthylcarbamoyle) et α-(N,N-diméthylthiocarbamoyle). On a observé les ions par photolyse éclair dans des
mélanges 40 : 60, acétonitrile : eau et de 50 : 50, méthanol : eau et on a mesuré les constantes de décroissance dans
les solvants seuls ainsi que pour leur capture par l’ion azoture. Les cations ainsi étudiés et leurs temps de vie dans
+
+
+
l’acétonitrile à 40% sont : 6, ArC H-CONMe , 0,6 µs; 2, ArC H-CSNMe , 7 ms et 4, ArC CH -CSNMe , 6 ms dans
2
2
3
2
lesquels Ar = 4-MeOC H . Le cation 4 réagit avec le solvant pour donner une élimination de proton provenant du
6
4
–
1
groupe méthyle en α et la constante de vitesse pour l’addition du solvant doit être inférieure à 1 s . Les temps de vie
des cations substitués par CSNMe sont de 10 –10 fois plus grandes que celles de leurs analogues dans lesquels le
5
7
2
groupe thioamide a été remplacé par un α-méthyle. Les spectres d’absorption UV–visible de ces deux cations
présentent des différences importantes par rapport à ceux des cations 4-méthoxybenzyles typiques. Les temps de vie et
les spectres suggèrent donc tous les deux qu’il existe une forte interaction entre le centre benzylique et le groupe
thioamide.
Mots clés : photolyse éclair, carbocation stabilisé par un groupe thiocarbamoyle, photosolvolyse.
[
Traduit par la Rédaction] McClelland et al. 1915
There have been several recent studies of solvolysis reac-
tions involving carbocations substituted at the C center with
slowly, in some cases somewhat more quickly, and in some
cases considerably more quickly. On the other hand, an α-
CONMe2 substituted cation generally forms considerably
more slowly than its α-methyl analog.
Less is known about the rate constants for solvent capture
of the fully formed cations. The 9-thioamidofluorenyl cation
1 has been claimed as a transient intermediate in a laser
flash photolysis experiment in 2,2,2-trifluoroethanol (TFE)
+
a thioamide functional group such as CSNMe (1–9). There
2
is good evidence that the reactions do occur by a stepwise
D + A (S 1) mechanism proceeding via a carbocation in-
N
N
N
termediate (2, 4, 7). Rate constants for the formation of the
α-thioamido substituted cation relative to an analog with α-
methyl vary. In some cases the former forms slightly more
(
6). There are, however, now serious concerns over the
structural assignment in this study, since the chloro precur-
sor that was supposedly employed decomposes instantly
when dissolved in TFE (9).
Received September 10, 1998.
1
R.A. McClelland and V.E. Licence. Department of
Chemistry, University of Toronto, Toronto, ON M5S 3H6,
Canada.
J.P. Richard, K.B. Williams, and S.-S. Lin. Department of
Chemistry, State University of New York at Buffalo (SUNY),
Buffalo, NY 14260-3000, U.S.A.
1Author to whom correspondence may be addressed.
Telephone: (416) 978-3592. Fax: (416) 978-3592. E-mail:
rmcclell@alchemy.chem.utoronto.ca
Can. J. Chem. 76: 1910–1915 (1998)
© 1998 NRC Canada

McClelland et al.
Table 1. Rate constants for solvent decay and azide quenching of 4-methoxybenzyl cations
1911
+
4
-MeOC H C R ,R generated by flash photolysis of 7–9. Experiments at 20°C in 40:60
6 4 1 2
acetonitrile:water at ionic strength zero and in 50:50 methanol:water at ionic strength 0.5 with NaClO4.
7
→ 6
8 → 2
9 → 4
Constant
CONMe , H
CSNMe , H
CSNMe , Me
2
2
2
k (40% AN), s–1
(1.6±0.1) × 106
(1.5±0.2) × 102
(1.8±0.2) × 102
Not measured
s
kaz (40% AN), M–1 s–1
(5.4±0.3) × 10
9
(6.3±0.2) × 10
7
–
1
(8.6±0.4) × 106
(2.4±0.1) × 102
2
ks (50% MeOH), s
(2.1±0.1) × 10
(4.9±0.2) × 10
(2.3±0.2) × 10
kaz (50% MeOH), M–1
k :k (50% MeOH), M
s
–1
(2.1±0.3) × 10
9
(1.03±0.03) × 10
7
4
–
1
(2.4±0.4) × 102
2.1 × 102
(4.3±0.2) × 104
7 × 104
2
az
s
a
–1
2 b
k :k (50% MeOH-sol), M
1.8 × 10
az
s
a
Ratio obtained from analysis of products of ground state solvolysis (7).
Solvent reaction is elimination. Azide reaction is sum of elimination and addition.
b
In a second study, competition kinetics in 50:50 metha-
experiments needed to be carried out immediately after
preparation of the solution. It in fact proved possible to pre-
pare the solution and irradiate within 20 s.
nol:water revealed that the benzylic derivative 2 has an
4
–1
azide:solvent selectivity k :k of 7 × 10 M (7). This is
az
s
4
–1
1
0 -fold greater than the k :k of 7.2 M for the 4-methoxy-
az s
phenethyl cation 3 in the same solvent (7). With the assump-
tion that the azide combination is a diffusion-limited process
9
–1 –1
(
10), so that k is the same for both cations (~5 × 10 M s ),
az
the lifetime (1͞k ) of the thioamido-substituted cation is
s
1
0 000-fold longer. Unexpected behavior is also seen with
α-thioamido, α-methyl substituted cations (4, 7). The cation
, for example, reacts with solvent to give exclusively the
4
alkene product derived from elimination of one of the α-pro-
tons (7). Elimination even occurs with azide although, with
+
this ion, addition to the C center also occurs. Such behavior
contrasts with that of the cumyl cation analog 5, where only
nucleophilic addition is observed (e.g., less than 1% elimina-
tion in the reaction with solvent in 50:50 methanol:water)
Flash photolysis spectra obtained with 7–9 in 40% ace-
tonitrile are shown in Fig. 1. The benzoate 7 produces a
transient species with λmax around 350 nm, which decays
over 5–10 µs to essentially zero OD. The decay is first order,
with the same rate constant being obtained across the entire
spectrum consistent with the presence of a single transient
species. Absorbing species are also observed with 8 and 9.
However, these show no decay up to times as long as
(
7).
5
00 µs, the longest time that can be reliably employed with
the laser flash photolysis apparatus. Decay is seen, however,
using a lamp photolysis apparatus. Like 7, these decays are
first order with the same rate constant throughout the spec-
trum. The λmax of the transients with 8 and 9 lie in the range
3
10–325 nm, being slightly higher with 9. There is, how-
As noted in the original paper (7), the large k :k ratio for
az
s
ever, some distortion of the spectra, since the thioamido pre-
cursors absorb below 310 nm. (Note that these experiments
measure a ∆OD, the OD after the laser pulse minus the OD
before.) With 9, there is a second weaker absorption with
λmax around 500 nm. This is due to the same transient as the
absorbance at lower wavelength, since the decay here has
the same rate constant. It can be noted that with 8, aged so-
lutions that had undergone solvolysis to the alcohol failed to
produce any transient.
2
leads to concerns about the applicability of the azide clock
assumption. The competing elimination͞addition with 4
raises this question even more with this system. In conse-
quence, we have performed flash photolysis experiments so
as to obtain the absolute rate constants for these cations. The
α-amido substituted 6 was included for comparison pur-
poses. It transpires that these cations can be generated from
the same precursors employed in the studies of the ground
solvolysis. Moreover, the α-thioamido cations are remark-
ably long-lived.
Azide ion accelerates the decay of all three transients,
with linear plots of the observed rate constant versus azide
concentration. Values of k , the first-order rate constant for
s
decay in solvent alone, and k , the second-order rate con-
az
The substituted benzoates 7–9 were employed as photo-
chemical precursors of the cations 6, 2, and 4, respectively.
These esters undergo ground state solvolysis with half-lives
of 100 h (7), 130 s (8), and 5.2 h (9) in 50:50 methanol:wa-
ter at 25°C (7). With 8 this meant that the flash photolysis
stant for the quenching by azide ion, are given in Table 1.
Two solvents were employed, 40:60 acetonitrile:water, and
50:50 methanol:water. The former was chosen to provide
comparisons with rate constants measured by flash pho-
tolysis for other benzylic cations. The latter was the solvent
©
1998 NRC Canada

1
912
Can. J. Chem. Vol. 76, 1998
employed in the analysis of the products of the ground state
solvolysis of these esters.
Identification of transients as benzylic carbocations
While there is no spectrum of an authentic cation available
for comparison, there is good evidence that the transients
observed with the three esters 7–9 are cations forming by
photoheterolytic cleavage of the α-carbon–oxygen bond.
Firstly, the stable products of irradiation are those expected
of a photosolvolysis. In the case of the thioamides 8 and 9 in
fact, there is a very efficient conversion to exactly the same
products obtained in the ground state solvolysis reaction.
Secondly, the transients are quenched by azide ion, a typical
carbocation trap. More importantly, the absolute rate con-
The final products of the irradiation were examined in
0% aqueous acetonitrile. These experiments were carried
4
out by preparing a solution of the precursor benzoate of con-
centration ~100 µM, which was then split in two. One half
was irradiated at 254 nm, while the other was kept in the
dark under conditions sufficient for complete ground state
solvolysis. The HPLC chromatograms of the two were then
compared, along with the chromatograms of the authentic
products of the ground state hydrolysis.
With 8, which undergoes a relatively rapid ground state
solvolysis, the solvent was cooled to 0°C before addition of
the substrate. After substrate addition, half was irradiated at
this temperature for 30 s, a time sufficient to remove 95% of
the precursor. The other half was warmed to room tempera-
ture and allowed to stand for a total time of 1 h. (The half-
life at 20°C is 4 min.) The products observed in both the ex-
cited state and ground state reactions were the alcohol 10
and pentafluorobenzoic acid (eq. [1]). Based on the relative
areas of the peaks (with a small correction for remaining
precursor), the yield of the the alcohol in the irradiated solu-
tion was 102% of that in ground state. Thus, within experi-
mental error, there was quantitative formation of the same
product of the ground state solvolysis.
stants kaz and k obtained by flash photolysis give selec-
s
tivities k :k that are essentially the same as those obtained
az
s
in the ground state solvolysis (7). This comparison was
made in 50:50 methanol:water and is shown in Table 1. In
the case of the cations 4 and 6, the flash photolysis ratio is
slightly larger but the difference is not outside experimental
error. With 2, the flash photolysis ratio is about two thirds of
that obtained in the competition experiment. The latter was,
however, obtained with azide ion concentrations in the range
0.5–1.0 mM, where >95% of the products arise from azide
trapping (7). This means that the amount of alcohol product
being measured is quite small, creating some uncertainty in
the final ratio.
The same conclusion was reached with 9. Again, 30 s ir-
radiation was sufficient to photolyze 95% of the precursor.
Four hours at 50°C was required for the ground state reac-
tion. The observed products from both were the alkene 12
and 4-methoxybenzoic acid (eq. [2]). The yield of alkene in
the irradiated solution was 97% of that of the solvolyzed
one.
With 9, the situation was more complicated. The pho-
tolysis was considerably less efficient, 3 min of irradiation
for example only reacting 45% of the starting material. A
consequence was that the products were not stable, unlike
the situation with the two thioamides, but underwent further
photochemistry. The solvolysis reaction required overnight
at 50°C for completion, and gave the expected alcohol 11
and pentafluorobenzoic acid (eq. [1]). These were also the
major products in the irradiated solution, but their yields ex-
pressed as a percentage of reacted 7 decreased with
increased irradiation time. Based on a three-point extrapola-
tion to zero irradiation, the yield of alcohol is 60–70%. The
HPLC chromatogram of the irradiated solution did have other
peaks, but these were not identified.
Rate constants for addition and elimination in cation 4
The reaction of the cation 4 gives only alkene product in
50:50 methanol:water (7). Thus the decay of this ion ob-
served in the flash photolysis experiments represents the re-
–
1
action with solvent acting as a base, and the k of 210 s is
s
equal to k (B) of eq. [3]. There is less than 1% of the ether
s
and alcohol products that would arise form nucleophilic ad-
–
1
dition of solvent. Thus k (Nu) < 2 s .
s
With azide ion, both elimination and addition occur. The
azide-catalyzed elimination is seen in the observation that
the alkene:RN ratio does not approach zero at high concen-
3
trations of azide, but plateaus, in fact, with a greater amount
of alkene (7). The limiting [alkene]:[RN ] is 1.69; this is
3
equal to the ratio of rate constants for azide acting as a base
and a nucleophile, k (B):k (Nu). The quenching that is ob-
az
az
served with azide ion in the flash photolysis studies repre-
sents the sum of these two processes, i.e., k (flash
az
photolysis) = k (B) + k (Nu). Since the ratio is known from
az
az
the product analysis, the absolute values of the two can be
4
–1 –1
calculated. Thus k (B) = 3.1 × 10 M
s
and k (Nu) =
az
az
©
1998 NRC Canada

McClelland et al.
1913
+
Table 2. Lifetimes of 4-methoxybenzyl cations 4-MeOC H C R R measured by flash
6
4
1 2
photolysis.
R1
R2
(1͞k ) in TFE
(1͞k ) in W:ANa
Ref.
s
s
H
H
CH3
H
H
H
CH3
CH3
COOMe
CONMe2
CSNMe2
CSNMe2
0.2 µs
2.5 µs
60 µs
6 µs
12
12
12
13
25 ns
50 ns
0.6 µs
7 ms
This work
This work
This work
H
CH3
6 ms (elim.)
>
1 s (add.)
a
The lifetimes of carbocations in aqueous acetonitrile solutions (0–60% acetonitrile) do not change
by more than 10–20% with changing acetonitrile content (14).
.8 × 10 M s–1. The latter is a remarkably low number for
4
–1
As was observed previously with an α-COOMe derivative
1
(
13), the α-CONMe cation 6 examined in this work is lon-
the addition of azide ion to a carbocation. Even the tris(4-
2
7
–1
ger lived than its α-methyl analog. The differences, how-
ever, are not large. The amide is an order of magnitude
longer lived than the ester, which in turn is probably about
another order of magnitude longer lived in water than the α-
methyl derivative (7, 10). These kinetic stabilizations occur
in spite of the fact that the α-CO cations are much less ther-
modynamically stable. Such observations are also made with
cations bearing α-CF3 substituents. Explanations for this be-
haviour have been proposed in the literature and need not be
repeated here (7, 13, 15, 16).
methoxyphenyl)methyl cation reacts with k around 10 M
az
–
1
s
(11).
It is also worth noting that for stabilized triarylmethyl cat-
ions k :k is approximately constant with changing reactiv-
az
s
ity (11), although the ratio starts to sharply decrease when
the azide reaction approaches the diffusion limit. For the cat-
ion 2, where both azide and solvent react as nucleophiles,
4
k :k = 4.5 × 10 . If the same ratio were to apply to the
az
s
nucleophilic reactions of 4, then k (Nu) for this cation would
s
–
1
be 0.4 s , only 0.2% of the k (B) value.
s
The cations bearing α-CSNMe on the other hand are ex-
2
traordinarily long-lived. The secondary ion 2, for example,
4
6
7
Lifetimes of 4-methoxybenzyl cations
is 10 longer lived than its CONMe analog and 10 –10
2
As suspected in the earlier study (7), the assumption of
diffusion control for the reactions of azide ion with the cat-
ions 2 and 4 is incorrect. The reactions with azide are con-
siderably slower than diffusion, and the lifetimes of these
two cations are substantial. A comparison with other 4-meth-
oxybenzylic cations whose rate constants have been directly
determined by flash photolysis is presented in Table 2. In the
case of the parent 4-methoxybenzyl and the derivative with
one α-methyl group, the lifetimes are too short in aqueous
solutions for the cations to be studied by ns laser flash
photolysis. These cations can, however, be observed in the
less nucleophilic TFE.
longer lived than the 4-methoxyphenethyl cation. The ter-
5
tiary 4 is 10 longer lived than the 4-methoxycumyl cation.
+
If this comparison involves only solvent addition to the α-C
center, this factor rises to 10 .
7
There are two possible interactions that could account for
the stabilization. One involves a resonance interaction as
shown in structure 13b where the positive charge is del-
ocalized to the sulfur atom. A large fraction of this interac-
tion would be lost in the transition state for solvent capture,
leading to an unusually low reactivity (15, 16). The second
possibility is the formation of a bridged ion, shown as the
resonance structures 13c and 13d.
©
1998 NRC Canada

1
914
Can. J. Chem. Vol. 76, 1998
Fig. 1. Flash photolysis spectra obtained on irradiation of 7 (a),
(b), and 9 (c) in 40:60 acetonitrile:water. Figure 1(a) was
Ab initio calculations have found that an α-thioamido cat-
ion where R = H and R = H is stabilized by 35.7 kcal͞mol
8
1
2
obtained with 248 nm laser flash photolysis; Figs. 1(b) and 1(c)
were obtained with lamp flash photolysis. In each figure the
uppermost spectrum is the ∆OD measured immediately after the
laser or lamp flash, the middle spectrum is the ∆OD at a time
corresponding to one half-life of the decay, and the bottom
spectrum is the ∆OD at the completion of the decay (6–8 half-
lives). The points drawn as full squares in Fig. 1(b) were
obtained with the laser flash photolysis apparatus, and represent
the ∆OD immediately after the laser flash, normalized to the
same ∆OD’s obtained in the lamp flash photolysis experiment.
by the first interaction, but a further 47.2 kcal͞mol stabiliza-
tion accompanies conversion to the cyclic structure (17).
The cation 13 with R = H and R = 3,5-bis(trifluorometh-
1
2
yl)phenyl reacts predominantly as a closed ion, giving 86%
of a product that has arisen from reaction of solvent (acetic
acid) at the thioamide carbon (2). The cation 2 of this study,
however, reacts to give products of addition at the α-carbon
(7). This cation of course has the strongly electron-donating
4-methoxy group, which would favour the open structure,
and the contrasting behaviour with the bis-CF cation im-
3
plies it must be reacting through this form. There is the pos-
sibility that the open cation is a minor species in an
equilibrium mixture with an unreactive closed form (7). This,
however, cannot account for all of the kinetic stabilization.
This conclusion is based on the measured kaz for 2 in aque-
ous acetonitrile, which is about 1% of the kaz of a diffusion-
controlled combination (e.g., the kaz for 6). Providing azide
ion reacts with the open cation (as the azide product implies)
there must be at least 1% of the open cation in the mixture
with the closed one. A percentage less than 1% requires kaz
to be greater than the encounter limit. This in turn means
4
–1
that kw for the open cation must still be less than 10 s .
This is still two orders of magnitude lower than the k for 6
w
4
and 10 lower than the value for the α-methyl cation.
The UV–visible spectra observed for the three cations in
this study (Fig. 1) are also interesting in the context of the
structural issue. The spectrum for the α-amido cation 6 is
typical of 4-methoxybenzyl cations. The 4-methoxyphene-
thyl and 4-methoxycumyl ions, for example, have similarly
shaped absorption bands (somewhat narrower) with λmax at
3
40 and 360 nm, respectively (12), in the same region as 6.
With this cation, therefore, there is no indication of anything
highly unusual that could be attributed to a strong interac-
tion with the α-carbonyl group.
This, however, cannot be said for the α-CSNMe cations.
2
With 2 the λmax is shifted to a lower wavelength, perhaps not
highly unusual for a benzyl cation, but there is a long tailing
absorbance, and cation decay is even observed at 550 nm.
The methyl cation 4 has a similar λmax at the lower wave-
length, and the absorbance at higher λ has now become so
pronounced that there is a second band centered around
5
00 nm. These spectra are clearly different from the spec-
trum for 6, and also from those of other benzylic derivatives.
It is tempting to ascribe the high-wavelength absorption ob-
served with 4 to increased delocalization, in other words to a
strong resonance interaction of the type 13a ↔ 13b in an
open cation. The same absorption is seen with 2, but it is
much weaker. This suggests the possibility that there is less
open cation in this system, although clearly there is still
some. Such conclusions are highly speculative, since there
are no suitable models that allow unambiguous comparisons
to be made. What can certainly be concluded is that the be-
haviours of the α-CSNMe cations are unusual, and this is
2
paralleled in both the absorption spectra and the kinetic sta-
bilities.
The photochemical precursors 7–9 and the solvolysis
©
1998 NRC Canada

McClelland et al.
1915
products 10–12 were from a previous study (7). Laser flash
photolysis experiments involved ca. 20 ns pulses at 248 nm
(
60–120 mJ ͞ pulse) from a Lumonics excimer laser (KrF)
1
. F.J. Ablenas, B.E. Georges, M. Maleki, R. Jain, A.C.
Hopkinson, and E. Lee-Ruff. Can. J. Chem. 65, 1800 (1987).
emission. A pulsed Xenon lamp provided monitoring light.
After passing through a monochromator, the signal from the
photomultiplier tube was digitized and sent to a computer
for analysis. Conventional flash photolysis experiments were
performed with equipment previously described (18), with
the sample being irradiated with a broad-band flash lamp of
ca. 100 ms duration.
2. X. Creary and T. Aldridge. J. Org. Chem. 53, 3888 (1988).
3. X. Creary and T. Aldridge. J. Org. Chem. 56, 4280 (1991).
4. X. Creary, H.N. Hatoum, A. Barton, and T. Aldridge. J. Org.
Chem. 57, 1887 (1992).
5. X. Creary and C. Zhu. J. Am. Chem. Soc. 117, 5859 (1995).
6. C.S.Q. Lew, D.F. Wong, L.J. Johnston, M. Bertone, A.C.
Hopkinson, and E. Lee-Ruff. J. Org. Chem. 61, 6805 (1996).
7
8
9
The products of irradiation were determined by preparing
mL of a solution of 7–9 in 40% aqueous acetonitrile of
. J.P. Richard, S.-S. Lin, J.M. Buccigross, and T.L. Amyes. J.
Am. Chem. Soc. 118, 12603 (1996).
. J.P. Richard, S.-S. Lin, and K.B. Williams. J. Org. Chem. 61,
5
concentration 100 µM. With 8, the solvent had been cooled
to 0°C in an ice bath before the substrate was added. One
half (2.5 mL) of these solutions were placed in a UV
cuvette, the other 2.5 mL in a capped vial wrapped in tin
foil. The former were immediately irradiated at 254 nm in a
spinning merry-go-round in a Rayonet reactor, then immedi-
ately analyzed by HPLC. The latter solutions were left under
such conditions that no precursor remained (see Results), as
shown by HPLC. Identical volumes of the irradiated and
solvolyzed solutions were injected into a Waters HPLC,
eluting with 60:40 water:acetonitrile through a C18 column.
Products were identified by comparison with known stan-
dards. The yields of the products in the irradiated solutions
were calculated from the relative peak areas of the irradiated
and solvolyzed solutions.
9
033 (1996).
. X. Creary and J. Tricker. J. Org. Chem. 63, 4907 (1998).
1
0. J.P. Richard, M.E. Rothenberg, and W.P. Jencks. J. Am. Chem.
Soc. 106, 1361 (1984).
1
1. R.A. McClelland, V.M. Kanagasabapathy, N. Banait, and S.
Steenken. J. Am. Chem. Soc. 113, 1009, (1991).
1
1
1
2. R.A. McClelland, C. Chan, F. Cozens, A. Modro, and S.
Steenken. Angew. Chem. Int. Ed. Engl. 30, 1337 (1991).
3. N.P. Schepp and J. Wirz. J. Am. Chem. Soc. 116, 11749
(
1994).
4. R.A. McClelland, V.M. Kanagasabapathy, N. Banait, and S.
Steenken. J. Am. Chem. Soc. 111, 3966 (1989).
15. T.L. Amyes, I.W. Stevens, and J.P. Richard. J. Org. Chem. 58,
6057 (1993).
1
1
6. J.P. Richard. Tetrahedron, 51, 7981 (1994).
7. M.H. Lien and A.C. Hopkinson. J. Am. Chem. Soc. 110, 3788
(
1988).
The continued financial support of the Natural Sciences
and Engineering Research Council of Canada is gratefully
acknowledged.
18. A.A. Allen, A.J. Kresge, N.P. Schepp, and T.T. Tidwell. Can.
J. Chem. 65, 1719 (1987).
©
1998 NRC Canada