Multiple decomposition pathways for the oxenium ion precursor O-(4-(4′-methylphenyl)phenyl)-N-methanesulfonylhydroxylamine

Article abstract of DOI:10.1021/jo9014062

(Chemical Equation Presented) Although O-arylhydroxylamine derivatives have been claimed to be sources of oxeniumions in a large number of studies, it is not clear that the products of these reactions are due to oxenium ions. Previously, we had shown thro

Full text of DOI:10.1021/jo9014062

pubs.acs.org/joc
Multiple Decomposition Pathways for the Oxenium Ion Precursor
O-(4-(40-Methylphenyl)phenyl)-N-methanesulfonylhydroxylamine
Yue-Ting Wang and Michael Novak*
Department of Chemistry and Biochemistry, Miami University, Oxford, Ohio 45056
novakm@muohio.edu
Received July 2, 2009
Although O-arylhydroxylamine derivatives have been claimed to be sources of oxenium ions in a large
number of studies, it is not clear that the products of these reactions are due to oxenium ions.
Previously, we had shown through azide trapping studies that the quinol ester 2a and the title
compound 3a generate the oxenium ion 1a. The ester 2a exclusively generates 1a in water and is also
a photoprecursor of 1a in water. This is not true of 3a. The oxenium ion pathway accounts for a
significant fraction of the reaction of 3a under neutral and acidic pH conditions, but there are three
other pathways that account for hydrolysis of 3a. Both 3a and its conjugate base 3a- are present in
aqueous solution under mild pH conditions. In addition to the oxenium ion product 4a, two other
significant products are generated: the phenol 6a and the rearrangement product 8a. Both 4a and 8a are
generated exclusively from 3a, whereas 6a is generated from both 3a and 3a-. Azide trapping studies
show that 6a and 8a are not generated from the oxenium ion. The phenol 6a is generated by two paths,
one involving an apparent radical intermediate 7a and the other through a stepwise R-elimination
pathway through 3a-. The rearrangement product 8a is generated either through a concerted
rearrangement or via an ion-pair rearrangement. Photolysis of 3a does not generate 1a. The only
products of photolysis of 3a in water are 6a (major) and 8a (minor). The weak O-N bond of 3a is
susceptible to homolysis under photolysis conditions, and the radical 7a is observed after laser flash
photolysis of 3a. The cation 1a that is observed during laser flash photolysis experiments on 2a cannot
be detected during similar experiments on 3a. These results suggest that the previous attribution of
oxenium ions as the source of the decomposition products of other O-arylhydroxylamine derivatives in
aromatic solvents via thermolysis or acid-catalyzed decomposition may not be correct.
Introduction
aromatic solvents via thermolysis or acid-catalyzed decom-
position.1-10 Interpretation of results was difficult because
reaction products generated from similar precursors were
significantly different from each other.1-4,6-9 Furthermore,
the isolated products were considerably different from those
Early mechanistic investigations into the potential gen-
eration of aryloxenium ions, 1, via heterolytic bond disso-
ciation processes concentrated on the utilization of O-aryl-
hydroxylamine derivatives as precursors to these ions in
(1) Abramovitch, R. A.; Inbasekaran, M.; Kato, S. J. Am. Chem. Soc.
1973, 95, 5428–5430.
(3) Abramovitch, R. A.; Inbasekaran, M. N. J. Chem. Soc., Chem.
Commun. 1978, 149–150.
(2) Abramovitch, R .A.; Alvernhe, G.; Inbasekaran, M. N. Tetrahedron
Lett. 1977, 1113–1116.
(4) Abramovitch, R. A.; Alvernhe, G.; Bartnik, R.; Dassanayake, N. L.;
Inbasekaran, M. N.; Kato, S. J. Am. Chem. Soc. 1981, 103, 4558–4565.
DOI: 10.1021/jo9014062
r
Published on Web 09/23/2009
J. Org. Chem. 2009, 74, 7697–7706 7697
2009 American Chemical Society

Wang and Novak
JOCArticle
SCHEME 1. Products Attributed to Aryloxenium Ions
Generated under Different Reaction Conditions
SCHEME 2. Previously Demonstrated Reaction Pathways for
2a and 3a17,22,23
obtained via oxidation reactions of phenols that were also
thought to generate oxenium ions.11-16 Specifically, the O-
arylhydroxylamine derivatives led to predominant attack by
nucleophiles at the ortho position of the purported cations,
whereas the apparent cations generated by phenol oxidation
led to predominant nucleophilic attack at the para position
(Scheme 1). The decomposition of the O-arylhydroxylamine
derivatives in aromatic solvents also often led to products of
apparent nucleophilic attack on the oxygen of the alleged
cation and to phenols that were thought to arise from H-
abstraction by a triplet oxenium ion.1-4,9 The triplet path-
way did not appear to be viable because low level ab initio
calculations (MP2/3-21G//STO-3G) suggested that the sing-
let is the ground state for unsubstituted (Y = H) 1 by ca. 30
kcal/mol.5 It has been argued that the products of apparent
nucleophilic attack on oxygen of the cation are due, instead,
to a concerted SN2-like process, but no convincing experi-
mental evidence for this pathway has been presented.9
More recently we have utilized the 4-(p-tolyl) quinol ester
2a and O-(4-(40-methylphenyl)phenyl)-N-methanesulfonyl-
hydroxylamine 3a to generate the oxenium ion 1a in an
aqueous environment (Scheme 2).17 Under hydrolysis con-
ditions from pH 1 to 8, 2a generates 4a as the only detectable
product. In the presence of N3- 4a is replaced by the azide-
adduct 5a with no increase in the rate constant for decom-
position of 2a.17 Phenol 6a was not detected under these
conditions.17 This result is consistent with recent calculations
suggesting that the triplet oxenium ion is ca. 20 kcal/mol less
stable than the singlet species.18,19 Under hydrolysis condi-
tions all available evidence is consistent with the conclusion
that 2a leads exclusively to the singlet ion 1a. Similar con-
clusions were reached for the 4-phenyl quinol ester 2b and the
4-(p-bromophenyl) quinol ester 2c.17,20,21 Under laser flash
photolysis conditions both 1a and the aryloxy radical 7a are
generated from photolysis of 2a in aqueous solution.22,23
Since their UV spectra do not overlap significantly, these
species can be separately characterized. The oxenium ion
-
(λmax 460 nm) decays in a first-order manner, reacts with N3
at a diffusion controlled rate, and is insensitive to O2,
whereas the radical (λmax 360 nm) undergoes reversible
dimerization resulting in biphasic decomposition kinetics,
is insensitive to N3-, is somewhat sensitive to O2, and does
yield 6a as a major reaction product.22,23 The hydroxylamine
derivative 3a was shown, in part, to generate the oxenium ion
1a in aqueous solution through heterolytic O-N bond
dissociation because the same hydrolysis product 4a
was observed, and that product was trapped out by N3- to
yield 5a with a selectivity ratio, kaz/ks, 1.2 ꢀ 103 M-1 that was
experimentally equivalent to that observed for 2a
(Scheme 2).17
(5) Li, Y.; Abramovitch, R. A.; Houk, K. N. J. Org. Chem. 1989, 54,
2911–2914.
(6) Endo, Y.; Shudo, K.; Okamoto, T. J. Am. Chem. Soc. 1977, 99, 7721–
7723.
(7) Shudo, K.; Orihara, Y.; Ohta, T.; Okamoto, T. J. Am. Chem. Soc.
1981, 103, 943–944.
(8) Endo, Y.; Shudo, K.; Okamoto, T. J. Am. Chem. Soc. 1982, 104, 6393–
6397.
(9) Iijima, H.; Endo, Y.; Shudo, K.; Okamoto, T. Tetrahedron 1984, 40,
4981–4985.
(10) Uto, K.; Miyazawa, E.; Ito, K.; Sakamoto, T.; Kikugawa, Y.
Heterocycles 1998, 48, 2593–2600.
(11) Swenton, J. S.; Carpenter, K.; Chen, Y.; Kerns, M. L.; Morrow, G.
W. J. Org. Chem. 1993, 58, 3308–3316.
The hydrolysis chemistry of 2a is particularly simple since
it generates only 1a at all pH examined.17 Although azide
(12) Kerns, M. L.; Conroy, S. M.; Swenton, J. S. Tetrahedron Lett. 1994,
35, 7529–7532.
(13) Swenton, J. S.; Callinan, A.; Chen, Y.; Rohde, J. L.; Kerns, M. L.;
Morrow, G. W. J. Org. Chem. 1996, 61, 1267–1274.
(14) Rieker, A.; Beisswenger, R.; Regier, K. Tetrahedron 1991, 47, 645–
654.
(18) Glover, S. A.; Novak, M. Can. J. Chem. 2005, 83, 1372–1381.
(19) Wang, Y.-T.; Jin, K. J.; Myers, L. R.; Glover, S. A.; Novak, M. J.
Org. Chem. 2009, 74, 4463–4471.
(20) Novak, M.; Glover, S. A. J. Am. Chem. Soc. 2004, 126, 7748–7749.
(21) Novak, M.; Glover, S. A. J. Am. Chem. Soc. 2005, 127, 8090–8097.
(22) Wang, Y.-T.; Wang, J.; Platz, M. S.; Novak, M. J. Am. Chem. Soc.
2007, 129, 14566–14567.
(15) Rieker, A.; Speiser, B.; Straub, H. DECHEMA Monogr. 1992, 125,
777–782.
(16) Pelter, A.; Ward, R. S. Tetrahedron 2001, 57, 273–282.
(17) Novak, M.; Poturalski, M. J.; Johnson, W. L.; Jones, M. P.; Wang,
Y.-T.; Glover, S. A. J. Org. Chem. 2006, 71, 3778–3785.
(23) Wang, Y.-T.; Jin, K. J.; Leopold, S. H.; Wang, J.; Peng, H.-L.; Platz,
M. S.; Xue, J.; Phillips, D. L.; Glover, S. A.; Novak, M. J. Am. Chem. Soc.
2008, 130, 16021–16030.
7698 J. Org. Chem. Vol. 74, No. 20, 2009

Wang and Novak
JOCArticle
TABLE 1. Derived Rate Parameters for Decomposition of 3aa
T (°C)
ko (s-1
(1.62 ( 0.03) ꢀ 10-4
)
k- (s-1
(2.17 ( 0.07) ꢀ 10-5
)
pKa
30
7.77 ( 0.04
aDetermined by a nonlinear least-squares fit of the kinetic data to eq 1.
The pH dependence of the decomposition of 3a is sum-
marized in Figure 1. The kobs for decomposition of 3a at pH
1-13 were fit to the rate law of eq 1. The derived rate
constants are presented in Table 1.
kobs ¼ koð½Hþꢁ=ðKaþ½HþꢁÞÞþk-ðKa=ðKaþ½HþꢁÞÞ
ð1Þ
FIGURE 1. pH dependence of the decomposition of 3a at 30 °C.
Red circles correspond to kobs for decomposition of 3a obtained by
UV spectroscopy. Blue triangles correspond to the rate constants
obtained by HPLC. Data were fit to eq 1 as described in the text.
The first term of eq 1 corresponds to a pH-independent
decomposition of neutral 3a. The second term of eq 1
corresponds to the uncatalyzed decomposition of its con-
jugate base 3a-. The dominant process at pH g 9 is the
uncatalyzed decomposition of 3a-. The rate constant for
decomposition of 3a-, k-, is ca. 7.5-fold smaller than the rate
constant for decomposition of its conjugate acid, ko. Over
the entire range of pH examined the rate of decomposition of
3a varies by less than an order of magnitude. The kinetic pKa
of 3a, 7.77 ( 0.04, corresponds to the previously observed
spectrophotometric pKa of 7.8 ( 0.1.17
Three major reaction products were detected by HPLC
throughout the pH range of the study. These are the pre-
viously identified quinol 4a,17 the phenol 6a, and the rear-
ranged product 8a. At pH e 8, 4a and 8a are the major
products. At pH g 10, 6a is the only significant product
detected. The time course of the decomposition of 3a and the
formation of these products can be followed by HPLC.
Figure 2 shows the kinetics of reactant decay and product
formation for reactions carried out in pH 8.7 and 9.2 Tris
buffers and in pH 11.5 NaOH solution.
trapping clearly shows that 3a also generates 1a, there were
indications in our earlier study that the chemistry of 3a is
more complicated than that of 2a. Ionization of 3a to form its
conjugate base 3a- near neutral pH was demonstrated by
spectrophotometric titration (pKa = 7.8 ( 0.1). The rate of
decomposition of 3a slowed considerably near its pKa, but
there were indications that its conjugate base was also
reactive in aqueous solution.17 Preliminary results showed
that the yield of 4a decreased at pH > 8, and products other
than 4a were detected by HPLC at all pH examined. Because
of the previous ambiguity of interpretation of the results for
the decomposition of other O-arylhydroxylamine derivatives
we decided to examine the chemistry of 3a in more detail. The
results presented herein show that 3a has three distinct
decomposition paths in aqueous solution: a heterolysis to
generate the cation 1a, a rearrangement pathway, and an
apparent homolysis to generate 7a, and ultimately 6a. The
conjugate base 3a- has another unique decomposition me-
chanism involving an R-elimination to form 6a and a nitrene.
Photolysis of 3a in aqueous solution is not a viable pathway
to the oxenium ion, 1a, but does yield 7a as a readily
detectable transient intermediate.
Results and Discussion
Hydrolysis Reactions. Within the pH range examined
previously, the decomposition of 3a remains pH-indepen-
dent from pH 1 to 7 and exhibits a decrease in the reaction
rate from pH 7 to 9 that appears to be associated with
ionization of 3a.17 For the current study, additional reaction
kinetics were measured at 30 °C in pH 5-8 NaH2PO4/
Data for 3a and the two products 4a and 6a fit a standard
first-order rate equation at all three pH. A double exponen-
tial rate equation was used to fit the data for 8a. The rate
constants obtained for the reaction of 3a based on HPLC
analysis are shown in Table 2. They are comparable to those
determined by UV spectroscopy under the same conditions.
The rate constants for the disappearance of 3a from the
HPLC experiments are shown with the rate constants ob-
tained from UV experiments in Figure 1.
The yield of the quinol 4a decreases in Tris buffers at pH >
8.0 compared to less basic buffers, but control experiments
and the kinetic results of Figure 2A and B shows that
authentic 4a is stable in Tris buffers. A fast decomposition
of authentic 4a in pH 12.8 NaOH solution was observed. No
HPLC peak corresponding to 4a can be detected within the
detection limit after 1 h. Instead, a new peak that has an
HPLC retention time similar to 4a was found. It is likely that
4a undergoes a base-catalyzed acyloin rearrangement.24,25
Na2HPO4 buffers, pH 7-9 TrisHþ/Tris buffers, pH 9.8
2-
HCO3-/CO3
buffer, and pH 11-13 NaOH solutions
(5 vol % CH3CN/H2O, μ = 0.5 (NaClO4)) by monitoring
changes in UV absorbance as a function of time. Because of
the low solubility of 3a in water, all reactions were performed
with an initial concentration of 3a of ca. 1 ꢀ 10-5 M to
prevent precipitation. Rate constants taken at two wave-
lengths were averaged at each pH to obtain kobs. All kobs for
3a are provided in Supporting Information. Rate constants
measured in 0.01-0.04 M phosphate and Tris buffers are
independent of buffer concentration (Supporting
Information). Rate constants reported in Supporting Infor-
mation were taken in 0.02 M buffer and were not extrapo-
lated to 0 M buffer. Kinetic data were also gathered by
monitoring changes in HPLC peak area versus time for 3a
and its decomposition products.
(24) Uno, H.; Yayama, A.; Suzuki, H. Chem. Lett. 1991, 7, 1165–1168.
(25) Nishinaga, A.; Itahara, T.; Matsuura, T.; Berger, S.; Henes, G.;
Rieker, A. Chem. Ber. 1976, 109, 1530–1548.
J. Org. Chem. Vol. 74, No. 20, 2009 7699

Wang and Novak
JOCArticle
FIGURE 3. Yields of decomposition products of 3a vs pH. Data
for 4a (blue triangles) and 8a (green diamonds) were obtained at
227 nm, and data for 6a (magenta inverted triangles) were obtained
at 260 nm. Initial concentration of 3a is ca. 1 ꢀ 10-5 M. Data were fit
as described in the text.
basic NaOH solution, peaks corresponding to 8a or its
decomposition product were not detected. Phenol 6a is stable
under all reaction conditions according to HPLC data.
The three major products observed by HPLC for the
decomposition of 3a in the absence of strong nonsolvent
nucleophiles in aqueous solution were generated in pH and
buffer independent yields in the pH range 4-7. These
products were isolated from a large scale hydrolysis reaction
of 3a in pH 4.6 acetate buffer at 30 °C. Both 4a and 6a were
identified by comparison to authentic samples. After isola-
tion and purification from reaction mixtures, the identity of
8a was confirmed by analysis of spectroscopic data. The
individual yields of 4a, 6a, and 8a determined by HPLC
quantification after 10 half-lives (ca. 12 h) of the hydrolysis
of 3a are (37.4 ( 1.1)%, (9.1 ( 1.1)%, and (34.3 ( 1.5)%,
respectively, in pH 4.6 acetate buffer and (34.5 ( 3.5)%,
(12.1 ( 1.1)%, and (28.6 ( 0.5)%, respectively, in pH 7.1
phosphate buffer. In basic solution at pH > 8, the yield of
phenol 6a increases while the yields of 4a and 8a decrease
with increasing pH. Control experiments confirmed that
product yields were not buffer-dependent in Tris buffers.
At pH g 10, 6a becomes the only significant product while 4a
and 8a are no longer observed. On the basis of HPLC
quantification 6a accounted for (70.6 ( 2.8)% of the decom-
position products of 3a in pH 11.5 NaOH solution. Yields of
the three major decomposition products of 3a in aqueous
solution at pH 4-13 obtained by HPLC quantification are
summarized in Figure 3. Since 8a decomposes in basic Tris
buffers, its yield under those conditions was determined by
extrapolation of its HPLC decay curve to zero time. Data for
4a and 8a were fit to eq 2 and eq 3, respectively, in which [4a]o
or [8a]o are the maximum concentrations of 4a or 8a formed
under acidic conditions. A titration curve (eq 4) was used to
fit the concentration vs pH data for 6a. [6a]o and [6a]-
represent the limiting concentration of 6a produced from
the decomposition of 3a under acidic and basic conditions,
respectively.
FIGURE 2. HPLC peak area as a function of time for the decom-
position of 3a (red circles) and the formation of 4a (blue triangles),
6a (magenta inverted triangles), and 8a (green diamonds) at 30 °C in
pH 8.7 Tris buffer (A), pH 9.2 Tris buffer (B), and pH 11.5 NaOH
solution. In A and B, peak area data were obtained at 260 nm except
for 4a, which was monitored at 227 nm. In C all data were obtained
at 269 nm. The data for 3a, 4a, and 6a were fit to a first-order rate
equation. The data for 8a were fit to a double exponential rate
equation.
TABLE 2. Rate Constants Obtained by HPLC Analysis of Decom-
position of 3a at pH 8.7, 9.2, and 11.5
105kobs (s-1
)
from 3a at from 4a at from 6a at from 8a at from 8a at
260/269 nm 260 nma
pH 260/269 nm
227 nm
260 nmb
8.7 4.2 ( 0.7
9.2 3.9 ( 0.1
11.5 1.7 ( 0.1
3.5 ( 0.3
3.7 ( 0.2
4.1 ( 0.6
3.8 ( 0.2
1.7 ( 0.1
3.8 ( 1.9
4.9 ( 1.3
2.7 ( 1.4
2.7 ( 0.7
aRate constant for the generation of 8a as 3a decomposes. bRate
constant for the disappearance of 8a.
Neither 4a nor its decomposition product were observed
during the decomposition of 3a monitored by HPLC in pH
11.5 NaOH solution (Figure 2C). Neither 4a nor its decom-
position product were detected in NaOH solutions or in
½4aꢁ ¼ ½4aꢁoð½Hþꢁ=ðKa0 þ½HþꢁÞÞ
½5aꢁ ¼ ½5aꢁoð½Hþꢁ=ðKa0 þ½HþꢁÞÞ
ð2Þ
ð3Þ
2-
HCO3-/CO3 buffer at the completion of the decomposi-
½6aꢁ ¼ ½6aꢁoð½Hþꢁ=ðKa0 þ½HþꢁÞÞþ½6aꢁ-Ka0 =ðKa0 þ½HþꢁÞÞ ð4Þ
tion of 3a. The results indicate that 4a is not generated at
pH g 10.
Control experiments confirm that the rearrangement pro-
duct 8a is stable in phosphate buffers but does decompose in
Tris buffers with rate constants that are somewhat smaller
than the rate constants for its formation (Table 2). In more
0
0
The average apparent pKa, pKa , is 8.84 ( 0.19. This pKa
differs from the thermodynamic pKa of 3a (7.77 ( 0.04) since
the yield of each product depends on the pKa of 3a, and the
7700 J. Org. Chem. Vol. 74, No. 20, 2009

Wang and Novak
JOCArticle
The dependence of decomposition kinetics and product
yields on pH and the lack of effect of N3- on the yield of 6a
and 8a at neutral pH indicate that there are several compet-
ing pathways for the decomposition of 3a over the pH range
0
examined. At pH < pKa the uncatalyzed decomposition of
neutral 3a via ko dominates the product distribution. The
quinol 4a is clearly generated through an oxenium ion path-
way, but 6a and 8a cannot be generated by that same path-
0
way. At pH > pKa the product distribution is dominated by
the uncatalyzed decomposition of 3a- via k-. Under these
conditions the only reaction product detected by HPLC is
the phenol 6a. Because 6a derived under basic pH conditions
is demonstrably generated from 3a- and 3a- cannot account
for the small but readily detectable yield of 6a under acidic
and neutral pH conditions, there must be two independent
pathways leading to 6a. A mechanistic scheme consistent
with the experimental data is presented in Scheme 3.
The scheme proposes that the three major decomposition
products are generated by four different pathways. Paths I
through III are proposed to account for the products gener-
ated from 3a via ko, whereas path IV is proposed to account
for the generation of 6a from 3a- via k-.
FIGURE 4. Yields of 4a (blue triangles), 5a (brown circles), 6a
(magenta inverted triangles), and 8a (green diamonds) at pH 7.0 in
0.02 M phosphate buffer as a function of [N3-]. Data were obtained
from HPLC analysis at 227 nm (4a) and 260 nm (5a, 6a, 8a). Yields
of 4a and 5a were fit to the standard “azide clock” equations.26,27
.
a
0
TABLE 3. pKa Obtained from Yield vs pH Data for 4a, 6a, and 8a
0
4a
6a
8a
av pKa
0
pKa
8.76 ( 0.07
9.15 ( 0.15
8.71 ( 0.09
8.84 ( 0.19
aFrom nonlinear least-squares fits of the yield data to eqs 2-4.
relative magnitudes of ko and k-. Since ko is 7.5-fold larger
than k-, the products generated from 3a have a kinetic
advantage over those from 3a-. The expected relationship
Since the quinol 4a is the most abundant product at pH <
0
pKa , path I is a major pathway for the decomposition of 3a
under these conditions. This pathway was previously de-
scribed.17 Evidence for the existence of 1a includes the results
of azide trapping experiments on both 2a and 3a and the
direct detection and characterization of 1a after laser flash
photolysis of 2a.17,22,23 The results shown in Figure 4 de-
monstrate that the generation of 6a and 8a cannot be
attributed to 1a.
0
between pKa , pKa, and ko/k- is shown in eq 5. The calculated
0
pKa (8.64 ( 0.05), obtained from eq 5 and the kinetic
parameters of Table 1, is in good agreement with the average
0
pKa derived from product yield versus pH data. These
results indicate that 4a and 8a are formed exclusively from
decomposition of 3a, whereas 6a is generated from decom-
position of both 3a and 3a- via independent pathways.
The rearrangement product 8a is reminiscent of similar
rearrangement products generated during the decomposi-
tion of N-arylhydroxylamine derivatives in aqueous solu-
pKa0 ¼ pKaþlogðko=k-Þ
ð5Þ
-
tion.28,29 The insensitivity of the yield of 8a toward [N3
]
Previously we showed that the quinol 4a was efficiently
trapped out of the reaction mixtures by N3- and replaced by
the azide adduct 5a with no acceleration of the rate of
decomposition of 3a.17 The selectivity ratio for 3a, kaz/ks,
(see Scheme 2) measured previously was (1.2 ( 0.1) ꢀ 103
could be used to support a mechanism involving a concerted
1,3-rearrangement accompanying O-N bond cleavage30 or,
alternatively, an ion-pair rearrangement proceeding through
a very short-lived tight ion pair.28,29,31 In aqueous solution
free ion 1a is a relatively stable cation (lifetime = 170 ns)
with a high azide/solvent selectivity ratio, kaz/ks, of 1.0 ꢀ
M-1 17
. This ratio was experimentally indistinguishable from
kaz/ks of (1.0 ( 0.2) ꢀ 103 M-1 measured from azide trapping
experiments performed on 2a.17 The equivalence of the
measured selectivity ratio and of the two reaction products
4a and 5a generated from the two different precursors 2a and
3a indicated that they both generated the same cation, 1a.17
In the current study, the effect of [N3-] on the yields of the
two other products was examined in pH 7.0 phosphate buffer
under reaction conditions identical to those used previously.
Results of that study are presented in Figure 4. Data for 4a
and the azide adduct 5a were fit by least-squares procedures
to the standard “azide-clock” equation.26,27 The observed
value of kaz/ks of (1.3 ( 0.2) ꢀ 103 M-1 is experimentally
equivalent to the value previously reported.17 Although the
data are somewhat scattered, the yields of neither 6a nor 8a
103 M-1 17,22
A systematic study of the effect of arylnitre-
.
nium ion stability on the involvement of ion pairs in the
hydrolysis of the corresponding esters of N-arylhydroxyla-
mines found that if the nitrenium ion is highly selective with
kaz/ks in the range of that observed for 1a, the yield of the
rearrangement product arising from the ion pair is typically
very low (less than 5%).29 This is due to the fact that the ion
pair has a very short lifetime in water (ca. 10-11 to10-10 s), and
if the ion is selective, the ion pair does not react significantly
before diffusional separation occurs.29,32 In the current case,
the yield of rearrangement product 5 is almost equal to that of
quinol 4a, and the cation 1a is highly selective. In order for the
rearrangement product 8a to be generated through the ion
-
responds significantly to changes in N3 concentration
under the reaction conditions. This indicates that the path-
(28) Fishbein, J. C.; McClelland, R. A. J. Chem. Soc., Perkin Trans. 2
1996, 663–671.
ways leading to 6a and 8a do not involve the cationic
-
(29) Novak, M.; Kahley, M. J.; Lin, J.; Kennedy, S. A.; James, T. G.
J. Org. Chem. 1995, 60, 8294–8304.
intermediate 1a that is trapped by N3
.
(30) Oae, S.; Sakurai, T. Tetrahedron 1976, 32, 2289–2294.
(31) Gutschke, D.; Heesing, A. Chem. Ber 1973, 106, 2379–2394. Gessner,
U.; Heesing, A.; Keller, L.; Homann, W. K. Chem. Ber 1982, 115, 2865–2871.
(32) Hand, E. S; Jencks, W. P. J. Am. Chem. Soc. 1975, 97, 6221–6230.
Eigen, M. Angew. Chem., Int. Ed. Engl. 1964, 3, 1–19.
(26) Jencks, W. P. Acc. Chem. Res. 1980, 13, 161–169.
(27) Richard, J. P.; Jencks, W. P. J. Am. Chem. Soc. 1982, 104,
4689-4691; 1982, 104, 4691-4692; 1984, 106, 1383-1396.
J. Org. Chem. Vol. 74, No. 20, 2009 7701

Wang and Novak
JOCArticle
SCHEME 3. Proposed Mechanistic Pathway for the Decomposition of 3a
0
SCHEME 4. Possible Ion Pair Mediated Pathway
explain the increasing generation of 6a at pH > pKa . There
can be little doubt that the rate-limiting step is the O-N
bond dissociation, because 3a- is detected spectrophotome-
trically at basic pH. This is consistent with the modest
leaving group ability of 6a-.
This mechanism requires that the nitrene 9 is formed along
with 6a-. Sulfonyl nitrenes, RSO2N, are known reactive
intermediates that can be generated from sulfonyl azides and
other precursors by thermolysis in aprotic solvents or photo-
lysis in protic solvents.33,34 Methanesulfonyl nitrene, 9, has
a triplet ground state based on a triplet EPR signal observed
at very low temperature.35,36 Nitreme 9 can react as a
diradical (triplet) undergoing H-abstraction or as an electro-
phile (singlet) inserting into aromatic or aliphatic hydrocar-
bons.33,34 It was reported that 9, generated from photolysis
of methanesulfonyl azide in EtOH underwent both H-ab-
straction and O-H insertion leading to methanesulfonamide
10 (yield 43%) and N-ethoxymethanesulfonamide (yield
48%).37 In an attempt to trap 9, MeOH solutions containing
1 and 10 mM NaOMe were employed as reaction media for
the decomposition of 3a at 30 °C in the dark to replace the
aqueous NaOH solutions. Authentic N-methoxymethane-
sulfonamide 12 was synthesized for comparison purposes.38
pair, kr (Scheme 4) must be nearly equivalent to k-d, the
diffusion-limited rate constant for the ion-pair separation,
based on the relative yields of 4a and 8a (ca. 1.1:1). However,
a kr of 1010 s-1 is ca. 100-fold larger than that observed in
similar reactions of selective arylnitrenium ion pairs .29 The
anionic component of the ion pair in those cases was typically a
carboxylate ion. The methylsulfonamide anion of the ion
pair in Scheme 4 is expected to be a stronger nucleophile, so
the ion-pair pathway cannot be ruled out.
T0he detection of a significant yield of 6a (ca. 10%) at pH <
pKa suggests that a third pathway for the reaction of 3a
(path III) occurs under acidic conditions. The triplet oxe-
nium ion is an unlikely source of 6a because recent calcula-
tions of the singlet-triplet gap for a variety of oxenium
ions including 1a indicate that the triplet state is about
20 kcal/mol less stable than the singlet state ion.18,19 Also,
if the triplet ion were in thermal equilibrium with the singlet
ion 1a, the yield of 6a should be sensitive to [N3-]. The
aryloxy radical 7a, derived from a homolytic cleavage of the
weak O-N bond, is a potential source of 6a. It is known that
7a derived from photolysis of 2a in aqueous solution and
CH3CN gives rise to 6a.23 Photolysis results for 3a, described
below, also show that this substrate readily generates 7a
under conditions in which 2a yields both 1a and 7a.
The reaction process was followed by HPLC and
GC-MS. The decomposition rate of 3a is dependent on
[MeO-]; 3a decayed faster in MeOH solution with lower base
concentration. Compared to the reaction of 3a in basic
aqueous solution, the decomposition in basic MeOH at the
(33) Abramovitch, R. A.; Sutherland, R. G. Fortschr. Chem. Forsch.
1970, 16, 1–33.
ꢀ
(34) L’abbe, G. Chem. Rev. 1969, 69, 345–363.
Under basic conditions the yield of phenol 6a increases
dramatically up to ca. 70-80%, whereas 4a and 8a are
simultaneously suppressed and finally not detectable in
NaOH solutions. The dominant form of 3a in basic solution
is the ionized species, 3a-; paths I and II, and III are no
longer kinetically viable at pH > pKa . A stepwise base-
catalyzed R-elimination (path IV) proceeding via 3a- can
(35) Smolinsky, G.; Wasserman, E.; Yager, W. A. J. Am. Chem. Soc.
1962, 84, 3220–3221. Wasserman, E.; Smolinsky, G.; Yager, W. A. J. Am.
Chem. Soc. 1964, 86, 3166–3167.
(36) Moriarty, R. M.; Rahman, M.; King, G. J. J. Am. Chem. Soc. 1966,
88, 842–843.
(37) Shingaki, T.; Inagaki, M.; Torimoto, N.; Takebayashi, M. Chem.
Lett. 1972, 1181–1184.
(38) Conway, T. T.; DeMaster, E. G.; Lee, M. J. C.; Nagasawa, H. T. J.
Med. Chem. 1998, 41, 2903–2909.
0
7702 J. Org. Chem. Vol. 74, No. 20, 2009

Wang and Novak
JOCArticle
same temperature was much slower. It required ca. 1 month
for 3a with an initial concentration of 0.01 M to disappear in
10 mM NaOMe/MeOH solution incubated at 30 °C. HPLC
analysis of the reaction mixture shows that 6a is the most
significant decomposition product of 3a in basic MeOH
solution. The sulfonamide 10 was detected by GC-MS with
an increasing yield during the course of the reaction. This is
the expected product arising from triplet 9 through a H-
abstraction mechanism. However, the expected O-H bond
insertion product of singlet 9, 12, was not detected. This
result is consistent with that recently reported by Desikan
et al.39 These workers reported a nearly quantitative yield of
10 for direct photolysis of a precursor of 9, N-mesyl diben-
zothiphene sulfilimine, in MeOH but no observation of 12.39
It is likely that in MeOH the singlet-triplet intersystem
crossing is dominant over the reactions of singlet nitrene.
Photolysis Reactions. Previously we directly detected the
generation of two transients, the cation 1a and aryloxy
radical 7a, from nanosecond laser flash photolysis (ns-LFP)
of quinol ester 2a in aqueous solution and exclusive genera-
tion of 7a from LFP of 2a in CH3CN.22,23 Photolysis of 3a
was carried out at initial concentrations of ca. 1 ꢀ 10-5 M in
N2-saturated 5 vol % CH3CN/H2O buffered with pH 7.1
0.02 M phosphate buffer, and in N2 or O2-saturated CH3CN
with an initial concentration of 3a of 5 ꢀ 10-5 M with UVC
lamps. The aqueous reaction solutions were examined by
HPLC immediately after short periods (15-45 s) of irradia-
tion. Since the initial concentration of 3a for the reactions in
CH3CN was higher and 3a is unreactive in CH3CN in the
dark, photolysis of 3a in CH3CN was monitored by HPLC as
the photolysis proceeded in short, punctuated (ca. 30 s)
irradiation pulses for total irradiation times up to 360 s.
HPLC analyses showed that there was little difference in
photolysis products between the reactions in aqueous buffers
and in CH3CN, or between the reactions carried out in N2-
saturated and O2-saturated reaction media. In all cases the
phenol 6a and the rearrangement product 8a were formed as
a result of photolysis. The initial relative yield of 6a and 8a is
ca. 3:1 in aqueous solution and in CH3CN. Both 6a and 8a
decompose with long irradiation time but 8a was more
photosensitive than 6a, so relative yields of the two photo-
products were determine after short UV pulses (ca. 15-30 s).
Quinol 4a was not identified as a photoproduct of 3a in any
experiment.
FIGURE 5. Transient absorbance spectra obtained after 266 nm
excitation of 3a in O2-saturated 50 vol % CH3CN/H2O at pH 7.1.
Key: red, 20 ns after the flash; green, 100 ns after the flash.
1a is not detected during photolysis of 3a in an aqueous
environment. The feature at 300 nm was not observed during
photolysis of 2a.22,23 In CH3CN the transient spectra are very
similar except the 360 nm absorbance band has shifted to
350 nm. This same solvent-dependent shift was observed in
the band attributed to 7a during LFP experiments on 2a.22,23
The bands decay over a period of ca. 500 μs. Kinetics of the
decay of A-300 and A-360 monitored at 300 and 360 nm,
respectively, at ambient temperature are summarized in
Figure 6. The transient associated with A-360 decays in a
biphasic manner. Data in Figure 6A were fit by a double
exponential rate equation. Two rate constants derived from
this fit are similar in the magnitude to those for the biphasic
decay of 7a previously observed in 5 vol % CH3CN/H2O
(Table 4).23
On the basis of the evidence presented above, the transient
species with λmax 360 nm generated from LFP of 3a is
identified as aryloxy radical 7a. Cation 1a is not generated
by photolysis of 3a in aqueous solution. Its characteristic UV
absorbance at 460 nm was not observed after LFP of 3a. This
indicates that the homolytic dissociation of 3a to generate 7a
is the dominant process during the photolysis of 3a. This
conclusion is supported by the product study results of
steady-state photolysis of 3a: quinol 4a, produced through
the intermediacy of 1a, was not observed under photolysis
conditions. Instead, phenol 6a was the most abundant
photolysis product.
Modeling of the decay kinetics of the other transient
species associated with A-300 is difficult because of the small
absorbance change and the scattered data. Fitting the data
by a triple exponential rate equation (Figure 6B) shows that
an additional process with a large rate constant of (3.22 (
0.46) ꢀ 105 s-1 occurs in the first few microseconds of the
reaction. This rate constant is an order of magnitude larger
than any of those observed for the decay of 7a. The other two
rate constants derived from the same fit, shown in Figure 6
appear to be comparable with the decay rate constants
attributed to 7a (Table 4). This indicates that the absorption
at 300 nm is due, in part, to 7a and to another shorter lived
species. The absorbance band associated with 7a that is
generated during LFP of 2a does have residual absorbance
at 300 nm, so it is reasonable that the kinetics of decay of
A-300 do include contributions from the decay of 7a.23 The
identity of the short-lived species is unclear. It is not likely to
be either singlet or triplet nitrene 9, since 9 cannot be detected
by ns-TRIR that has a lower detection limit of about 50 ns.39
Compound 3a was also subjected to ns-LFP (266 nm laser)
in Ar- or O2-saturated 50 vol % CH3CN/H2O buffered by
pH 7.1 phosphate buffer, and in CH3CN. The larger propor-
tion of CH3CN in the LFP experiments performed in the
aqueous environment was necessitated by the need to in-
crease the concentration of 3a to obtain sufficient absor-
bance at 266 nm. The UV spectra for the reaction in O2-
saturated aqueous solution taken within 100 ns of the laser
flash are shown in Figure 5. Two transient absorbance bands
at λmax 300 nm (A-300) and λmax 360 nm (A-360) were
observed. No significant absorbance band was observed at
longer wavelength. An absorbance band at 360 nm that was
attributed to the radical 7a was previously observed during
ns-LFP experiments on 2a.22,23 The band at 460 nm observed
during the photolysis of 2a that was attributed to the cation
(39) Desikan, V.; Liu, Y.; Toscano, J. P.; Jenks, W. S. J. Org. Chem. 2008,
73, 4398–4414.
J. Org. Chem. Vol. 74, No. 20, 2009 7703

Wang and Novak
JOCArticle
Azide trapping that occurs without overall rate acceleration
demonstrates that 3a, similar to the quinol ester 2a, does
generate 1a under solvolysis conditions. Unlike the hydro-
lysis of 2a, however, ionization that leads to the cation 1a is
not the only pathway for the decomposition of 3a in aqueous
solution since three major products including the quinol 4a,
phenol 6a, and rearrangement product 8a were identified.
The formation of these products is pH-dependent. The
insensitivity of the yields of 6a and 8a to N3- indicates that
they result from processes that do not involve 1a. The
rearrangement product 8a is likely to be produced directly
from 3a through concerted intramolecular rearrangement or
through rearrangement of an ion pair. Formation of phenol
6a under acidic conditions may require homolysis of the
O-N bond in 3a. Formation of 6a under basic conditions, on
the other hand, occurs through a base-catalyzed R-elimina-
tion of 3a via its conjugate base 3a-. The byproduct of the R-
elimination is the nitrene 9. Decomposition of 3a in basic
MeOH leads to formation of methanesulfonamide 10, the
expected product of H-abstraction by triplet nitrene 9. The
expected O-H insertion product, methoxymethanesulfona-
mide 12, was not observed. This result is consistent with a
recent observation that 9 generated photolytically in MeOH
yields only 10.39
It has been demonstrated that LFP of 2a in aqueous
solution generates two transient species, the oxenium ion
1a and the aryloxy radical 7a, that are responsible for the
formation of a variety of photolysis products in aqueous
solution. LFP of 3a in aqueous solution does not generate 1a,
but the phenoxy radical 7a is detected. These results are
consistent with the observation that 6a is a major photolyisis
product of 3a, but 4a is not detected in the photolysis
reaction mixtures of 3a. In addition to the biphasic decay
of 7a observed at 360 nm, another faster decay process is
observed at 300 nm. Since both phenol 6a and rearrangement
product 8a were generated from steady-state photolysis of
3a, it is possible that the fast decay process observed at
300 nm is due to an intermediate leading to 8a.
Although the quinol ester 2a exclusively generates oxe-
nium ion 1a under hydrolysis conditions and is a reliable
photoprecursor of 1a, the generation of 1a from the
O-arylhydroxylamine derivative 3a is less favorable. The
oxenium ion pathway does account for a significant fraction
of the reaction of 3a under neutral and acidic conditions, but
there are three other pathways that account for decomposi-
tion of 3a under hydrolysis conditions. The phenol 6a is
generated by two different paths, one involving an apparent
radical intermediate 7a and the other proceeding via a
stepwise R-elimination pathway through the conjugate base
of 3a, 3a-. The rearrangement product 8a is also produced
by a pathway that does not involve 1a. Photolysis of 3a does
not generate 1a. The weak O-N bond of 3a is susceptible to
homolysis under photolysis conditions, and the radical 7a is
readily observed after laser flash photolysis of 3a. Photo-
induced rearrangement to generate 8a also is a minor path-
way under photolysis conditions.
FIGURE 6. Decay of UV absorbance bands A-360 (A) and A-300
(B) in O2-saturated pH 7.1 phosphate buffer (50 vol % CH3CN/
H2O, μ = 0.5 (NaClO4)). Data were fit to a double exponential rate
equation (A) and a triple exponential equation (B) (blue curves).
TABLE 4. Decay Rate Constants for 7a in O2-saturated 5 vol %
CH3CN/H2O and for Two Absorbance Bands A-360 and A-300 Observed
after LFP of 3a in O2-Saturated 50 vol % CH3CN/H2O
transient/
conditions
k1 (s-1
)
k2 (s-1
)
decay of 7a in
5 vol % CH3CN/
H2O, O2 sat.a
(8.06 ( 0.08) ꢀ 104
(3.44 ( 0.48) ꢀ 104
(1.45 ( 0.13) ꢀ 104
(1.29 ( 0.02) ꢀ 104
(0.68 ( 0.09) ꢀ 104
(0.24 ( 0.23) ꢀ 104
decay of A-360 in
1/1 CH3CN/
H2O, O2 sat.b,d
decay of A-300 in
1/1 CH3CN/
H2O, O2 sat.c,d
aData from ref 23. bRate constants and error limits are averages and
standard deviations of four independent measurements. cDerived from a
single measurement. Three rate constants were obtained from the data
fit; only the two smaller rate constants are shown. The larger rate
constant is described in the text dAll absorbance data except the data
collected within the first 1 μs after the flash were used in data fitting.
Since the rearrangement product 8a was also generated
photochemically, this rapidly decaying species might be an
intermediate leading to 8a. Alternatively, it could be the
radical 11, the expected byproduct of the formation of 7a.
This species does not appear to have been described in the
literature.
Because O-arylhydroxylamine derivatives have alterna-
tive reaction pathways and aromatic solvents are not effec-
tive at solvating cations, the previous attribution of oxenium
ions as the source of the decomposition products of other
O-arylhydroxylamine derivatives in aromatic solvents via
thermolysis or acid-catalyzed decomposition (Scheme 1)
Conclusion
Originally methanesulfonylhydroxylamine 3a was consi-
dered to be a potential precursor of the oxenium ion 1a.
7704 J. Org. Chem. Vol. 74, No. 20, 2009

Wang and Novak
JOCArticle
should be reconsidered. The previously observed phenol
products1-4,9 are likely generated by homolysis or R-elimi-
nation mechanisms. The predominance of ortho-substituted
products1-4,6-9 suggests an intramolecular pathway related
to the process that generates 8a in this study.
buffer incubated at 40 °C. The steady-state concentration of 3a
in the reaction medium was maintained as ca. 1 ꢀ 10-5 M to
avoid precipitation of 3a during the reaction. After completion
of the addition the mixture was incubated in the dark at 40 °C
for another 10 half-lives (ca. 160 min; at 40 °C t1/2 is ca. 16 min).
The reaction was quenched by neutralization with saturated
aqueous NaHCO3 solution and was extracted with CH2Cl2 (3 ꢀ
250 mL) until HPLC examination indicated no products re-
mained in the aqueous layer. The combined extract was dried
over anhydrous Na2SO4, filtered, and evaporated to dryness
under vacuum. The residue was subjected to separation and
purification by multiple application of radial chromatography
on silica gel using 20/1 CH2Cl2/EtOAc as eluent. Two products
(4a and 6a) are known compounds identified by comparison to
authentic samples. The third product 8a was characterized by
NMR, IR, and LC-MS.
N-(4-Hydroxy-40-methylbiphenyl-3-yl)methanesulfonamide
(8a). IR 3377, 3255, 3020, 2919, 1614, 1502, 1385, 1298, 1135,
1111, 983 cm-1; 1H NMR (300 MHz, DMSO-d6) δ 2.32 (3H, s),
2.97 (3H, s), 6.96 (1H, d, J = 9.0 Hz), 7.23 (2H, d, J = 9.0 Hz),
7.33 (1H, dd, J = 10.4 Hz, 2.1 Hz), 7.44 (3H, m), 8.80 (1H, s(br)),
9.98 (1H, s(br)); 13C NMR (75.5 MHz, DMSO-d6) δ 20.6, 116.2,
124.3, 124.6, 124.7, 125.8, 129.4, 131.4, 135.8, 136.8, 150.3; LC-
MS (ESI, positive) m/e 300 (M þ Na)þ, 199 (M - SO2Me þ H)þ;
(ESI, negative) m/e 276 (M - H)-, 183 (M - NHSO2Me)-;
high-resolution MS (ES, positive) C14H16NO3S (M þ H) calcd
278.0845, found 278.0853; C14H15NO3SNa (M þ Na) calcd
300.0664, found 300.0668.
Experimental Section
Synthesis. The synthesis and characterization of 3a, 4a, and
6a have been described previously.17 The isolation and char-
acterization of 8a is described below. A sample of N-(methoxy)-
methanesulfonamide 12 was obtained from a previously pub-
lished procedure.38
Kinetic Studies. Kinetic studies of the decomposition reac-
tions of 3a were performed in 5 vol % CH3CN/H2O, μ = 0.5
(NaClO4) over a wide range of pH at 30 °C in the dark. The pH
was maintained with HClO4 solutions (pH < 3.0), NaOH
solutions (pH > 10.0), or with HCO2H/NaHCO2, AcOH/
AcONa, NaH2PO4/Na2HPO4, TrisHþ/Tris base, and HCO3
/
-
CO32- buffers. Owing to the low solubility of 3a in water, a 0.002
M stock solution of 3a in CH3CN was prepared and injected
(15 μL) into 3 mL of reaction medium to obtain a low initial
concentration of 1 ꢀ 10-5 M. Reactions were monitored by
measuring the changes in UV absorbance as a function of time at
all examined pH and by measuring the changes in the HPLC
peak area as a function of time for 3a and its decomposition
products in pH 8.7, 9.2 Tris buffers, and pH 11.5 NaOH solution
with dual wavelength monitoring. (HPLC condition: 20 μL
injections on a 4.7 mm ꢀ 250 mm C-8 reverse phase column,
65/35 MeOH/H2O elution solvent, flow rate 1.0 mL/min). The
wavelength chosen for monitoring the decomposition process
varied slightly for different reaction media due to a slight shift of
the maximum absorbance changes observed during a repetitive
scan of the reaction solution: 227 and 260 nm used for reactions
in all buffers, 232 and 269 nm for reactions in aqueous NaOH
solutions. For the reactions monitored by UV spectroscopy, the
observed rate constants were obtained as an average of two rate
constants taken at each wavelength. For the reactions analyzed
by HPLC, rate constants were obtained for each compound at
the same wavelengths used for the UV study. The buffer effect
on both reaction kinetics and product formation of 3a was
investigated in 0.01-0.04 M 1/9 NaH2PO4/Na2HPO4 buffers
and in 0.01-0.04 M 3/1 TrisHþ/Tris base buffers, respectively.
Control experiments were performed that measured the stability
of rearranged product 8a in pH 7.1 phosphate buffer and in pH
9.1 Tris buffer, and the stability of quinol 4a in pH 12.8 NaOH
solution at 30 °C. Reactions were initiated by injection of 15 μL
of ca. 2 mM of stock solution of 4a or 8a in CH3CN into 3 mL of
the reaction solution incubated at 30 °C. The control mixtures
were subjected to HPLC examination periodically. Most rate
constants were evaluated by fits to a standard first-order rate
equation. Formation and decay kinetics of 8a in Tris buffer were
determined by fitting the peak area versus time data to a double
exponential rate equation.
Azide Trapping Studies. Azide trapping experiments were
performed on the hydrolysis reaction of 3a in pH 7.0, 0.02 M
phosphate buffer (5 vol % CH3CN/H2O, μ = 0.5 (NaClO4))
-
containing different amount of N3 (0.0005-0.03 M). HPLC
examinations were performed on the reaction solutions after 10
half-lives (ca. 12 h). HPLC peak area for 6a and 8a, the
decomposition products of 3a, and for the azide adduct 5a were
plotted as a function of [N3-]. The production and character-
ization of azide adduct 5a have been published elsewhere.17
HPLC conditions were the same as those used for product
studies but only the data obtained at 260 nm were used.
Attempted Trapping of Nitrene, 9. The decomposition of 3a in
0.01 M MeONa/MeOH solution was monitored periodically by
GC-MS and by HPLC until the HPLC peak for 3a was no
longer observable. 3a (13.8 mg, 0.05 mmol) was directly added
into 5 mL of basic MeOH solution incubated at 30 °C to gain an
initial concentration of 3a of 0.01 M. Simultaneously 0.01 M of
12 in the same basic MeOH solution was incubated at 30 °C as a
control experiment. The samples used for HPLC analysis were
prepared by diluting 100 μL of the reaction mixture with 1 mL of
HPLC eluent. Another 100 μL of reaction mixture or the control
experiment mixture was dried by a stream of N2, redissolved in
50 μL of EtOAc, and then subjected to GC-MS examination.
HPLC conditions were 20 μL injections on a 4.7 mm ꢀ 250 mm
C-8 reverse phase column, 65/35 MeOH/H2O elution solvent,
flow rate 1.0 mL/min, monitoring wavelengths: 227, 262 nm.
GC-MS conditions were 30 m ꢀ 0.25 mm ꢀ 0.25 μm UF-35 ms
column, initial column temperature 50 °C for 3 min, a 10 °C/min
ramp to 250 °C, remaining at that temperature for 12 min,
column pressure 0.1 psi. Identities of the major products gener-
ated from 3a under this reaction condition were determined by
comparison to the authentic samples using both HPLC and
GC-MS.
Product Studies. The same reaction solutions of 3a used for
kinetic studies were subjected to HPLC analyses after the
completion of the hydrolysis reaction. Individual yields of three
major products were determined by HPLC quantification and
studied as a function of pH over the pH range 4-13. HPLC
conditions were 20 μL injections on a 4.7 mm ꢀ 250 mm C-8
reverse phase column, 65/35 MeOH/H2O elution solvent, flow
rate 1.0 mL/min, monitoring wavelengths: 227, 260 nm for
reactions in buffers, 232, 269 nm for reactions in NaOH solu-
tions. Three major products were isolated from a large-scale
hydrolysis reaction of 3a in HOAc/NaOAc buffer at pH 4.6. A
general procedure is as follows: 3a (0.104 g, 0.376 mmol) was
dissolved in 5 mL of CH3CN, and the resulting solution was
slowly added by syringe pump to 1 L of pH 4.6 0.02 M acetate
Steady-State Photolysis Experiments. Steady-state photolyses
of 3a in 0.02 M pH 7.1 phosphate buffer (5 vol % CH3CN/H2O,
μ = 0.5(NaClO4)) and in CH3CN were performed in a Rayonet
photochemical reactor in a jacketed quartz vessel kept at 30 °C.
UVC lamps that have emission (ca. 90% of total lamp energy) in
the range from 235 to 280 nm with a sharp maximum at 254 nm
were used as the UV sources. A brief description of the
J. Org. Chem. Vol. 74, No. 20, 2009 7705

Wang and Novak
JOCArticle
apparatus for steady-state photolysis has been published.22 For
reactions proceeding in the aqueous buffer, an initial concentra-
tion of ca. 1 ꢀ 10-5 M in 3a was obtained by adding 0.5 mL of
0.002 M stock solution of 3a in CH3CN to 100 mL of N2-
saturated phosphate buffer. Reaction mixture was subjected to
either 15 or 45 s irradiation. For reactions in CH3CN, 5 mL of
0.01 M stock solution of 3a in CH3CN was injected into 100 mL
of N2- or O2-saturated CH3CN to obtain an initial concentra-
tion of 3a of 5 ꢀ 10-5 M. Reactions performed in aqueous
solutions were examined by HPLC immediately after the irra-
diation (C-8 reverse phase column, 65/35 MeOH/H2O elution
solvent, 1 mL/min, monitored by UV absorbance at 227 and 260
nm). In the reactions carried out in CH3CN, 3a was exposed to
the UV irradiation for short intervals (ca. 30 s) for a total
irradiation time of 270 s in N2-saturated CH3CN and 360 s in
O2-saturated CH3CN. The reaction mixture was examined by
HPLC after each irradiation in order to monitor the progress of
the photoreaction.
CH3CN/H2O, μ = 0.5(NaClO4)) by injecting 15 μL of ca.
0.02 M stock solution of 3a into 3 mL of the buffer, so that
the initial concentration of 3a was ca. 2.5 ꢀ 10-4 M. Since 3a
undergoes slow hydrolysis under these conditions, all solutions
were used promptly after mixing. Transient absorbance spectra
were monitored in the range 280-540 nm. Kinetics of the decay
of transients was monitored at 300 and 360 nm.
Acknowledgment. We thank the Donors of the American
Chemical Society Petroleum Research Fund (Grant no.
43176-AC4) for support of this work. Y.-T.W. thanks the
Department of Chemistry and Biochemistry at Miami Uni-
versity for a Dissertation Fellowship. The laser flash photo-
lysis experiments were performed with the assistance of
Jin Wang in the laboratory of Prof. Matthew S. Platz at
The Ohio State University.
Laser Flash Photolysis Experiments. A standard procedure
provided elsewhere was employed for LFP of 3a.22,23 Laser flash
photolysis was carried out using a Nd:YAG laser (266 nm, ∼5 ns
pulse) with 1 cm light path. All reactions were performed at
ambient temperature (22 °C). Solutions of 3a were made in O2 or
Ar saturated pH 7.1, 0.02 M phosphate buffer (50 vol %
Supporting Information Available: Table S1, containing rate
constants at various pH and buffer concentrations; Figure
S1, showing the dependence of product yields on buffer con-
1
centrations; and H and 13C NMR spectra of 8a. This mate-
rial is available free of charge via the Internet at http://
pubs.acs.org.
7706 J. Org. Chem. Vol. 74, No. 20, 2009