.6 × 10 s , compared to 2.0 × 10 s-1 without Ce(IV)
-
2
-1
-9
18O exchange of the methylphosphonic acid product under
our hydrolysis conditions.
3
2
at pH 7.6, 30 °C.
Despite this significant progress in the rate enhancement
of phosphodiester hydrolysis, relatively little is known about
the site of metal cation mediated hydrolysis. To propose any
detailed mechanism for these metal cation mediated phos-
phodiester cleavages, it is absolutely essential to know
whether the P-O or C-O bond is cleaved.
For DMP, it was necessary to first determine the P-O/
C-O selectivity of MP (5) hydrolysis because the Ce(IV)-
mediated cleavage of MP is slightly faster than the DMP f
MP reaction; MP does not sufficiently accumulate, and the
site selectivity of the DMP cleavage must be extracted from
the results of the overall (DMP f phosphate + 2MeOH)
conversion.
DMP is hydrolyzed with 99.5% O-Me cleavage in neutral
1
8
or basic water, and with 78% O-Me cleavage at pH 1.24,
but the cleavage site is unknown for Ce(IV) or Co(III)-
MP (5 mM) was heated at pH 2.2, 60 °C for 5 h (>95%
1
8
reaction) with 25 mM Ce(IV) in 47% O-enriched water.
10
31
cyclen mediated hydrolyses. A very recent report indicates
The product phosphate singlet at δ 1.648 (59.3%) for P-
5
16
that bis(η -cyclopentadienyl)molybdenum(IV) effects P-O
O was accompanied by a 0.019 ppm upfield singlet for
scission of DMP.11 Additionally, Komiyama has demon-
strated P-O cleavage in the heterogeneous Ce(IV) hydrolysis
of thymidylyl(3′,5′)thymidine.12a
31
18
18
P- O (40.7%). Taking account of the O enrichment, this
corresponds to 86.6 ((1.7)% P-O cleavage of the MP. A
18
control experiment demonstrated no O exchange of phos-
Here we report that the (acidic) Ce(IV) hydrolyses of DMP
and MMP as well as methyl phosphate (5, MP) involve
extensive (but not exclusive) P-O cleavage. These site-
selectivity results help refine our mechanistic conception of
these enormously accelerated Ce(IV)-mediated hydrolyses.
phate or MP under the reaction conditions.
Next, 5 mM of DMP was cleaved by 25 mM Ce(IV) in
1
8
47% O-enriched water at pH 2.2, 60 °C over 5 h. The
-
4
-1
reaction proceeded with k ) 1.8 × 10
s
(t1/2 ) 1.07 h).
Hydrolysis occurs more slowly here, with a cation/substrate
13
13
concentration ratio of Ce(IV)/DMP ) 5, than it does with
C-Labeled MMP (2, O- CH
3
) was prepared from methyl
OH (Et N, Et O, 0 °C, 3 h;
O/acetone, 12h). 13C-2 (7.5 mM) and 25 mM
Ce(IV) (as ceric ammonium nitrate) were heated to 60 °C at
-
4
-1 7
13
Ce(IV)/DMP ) 10, where k ) 5.3 × 10 s . After
phosphonic dichloride and CH
then 1:5 H
3
3
2
31
chelation of the Ce(IV), the P NMR spectrum showed 87%
2
of phosphate at δ 1.648, 6% of MP (δ 2.85, with a significant
3
1
18
18
P- O upfield satellite), and 7% of residual DMP (δ 4.01,
pH 2.2 for 2 h in 1 mL of 43.8% enriched O-water. The
1
8
-
4
-1
no O exchange).
substrate cleaved with khydrol ) 4.8 × 10
s
(t1/2 ) 24
The δ 1.648 P- O singlet was accompanied by 31P-
O and P- O
3
1
16
1
min) as followed by the formation of MeOH in the H NMR
1
8
31
18
spectrum.13
2
upfield satellites at 1.627 and 1.608 ppm.
18
The distributions are shown in eq 1, where O is represented
After 2 h, the Ce(IV) was chelated with 25 mg of sodium
3
1
EDTA, and the P NMR spectrum (161.9 MHz) revealed
the product, methylphosphonic acid, at δ 25.94 (vs external
3
1
16
8
5% H
3 4
PO ). The singlet resonance for P- O was
31
18
accompanied by a second singlet ( P- O) 0.026 ppm
upfield. Deconvolution analysis (with Varian VNMR
software, version 5.2) gave the O/ O P integral ratio as
1.3/38.7, which, corrected for 43.8% O enrichment,
1
4
16
18
31
18
6
18
indicates 88.3 ((1.8)% O incorporation as P-O cleavage
by a “dark” atom. The problem is to determine the 16O/18O
product ratio of the DMP f MP cleavage, eq 2.
15
during the Ce(IV) hydrolysis of MMP.
The 13C spectrum (100.57 MHz) showed the product 13
-
CH
3
OH singlet accompanied by a small (4.6%) satellite for
13
18
14
18
3
CH OH 0.019 ppm upfield. After correction for the O
content of the water, this corresponds to 10.5 ((0.2)% water
attack at the methyl group. The directly measured P-O
(88%) and C-O (10.5%) cleavages of MMP are in good
experimental balance. A control experiment revealed <1%
We assume that the 59.3/40.7 16O/ O partition observed
in the MP f phosphate hydrolysis (above) holds for the
overall DMP f phosphate conversion. The source of
unlabeled phosphate from DMP in eq 1 must be unlabeled
MP (x in eq 2), while the source of doubly labeled phosphate
18
(
9) Moss, R. A.; Ragunathan, K. G. Langmuir 1999, 15, 107. The Ce-
IV) concentration was 1 mM, with [3] ) 0.05 mM, in the presence of 2
mM Brij surfactant.
(
(10) Kim J. H.; Chin, J. J. Am. Chem. Soc. 1992, 114, 9792.
(11) Kuo, L. Y.; Barnes, L. A. Inorg. Chem. 1999, 38, 814.
(12) (a) Komiyama, M.; Takeda, N.; Takahashi, Y.; Uchida, H.; Shiiba,
T.; Kodama, T.; Yashiro, M. J. Chem. Soc., Perkin Trans. 2 1995, 269. (b)
Komiyama, M.; Takeda, N.; Shigekawa, H. Chem. Commun. 1999, 1443.
18
in eq 1 must be O-MP (y in eq 2). Therefore, the 31.7% of
16
(
13) With a higher Ce(IV)/MMP ratio (25/2.5, pD 2.2, 60 °C), khydrol )
unlabeled phosphate in eq 1 must arise from x% of O-MP,
-
3
-1
1
.2 × 10
s
.
18
hydrolyzed without O incorporation (59.3%, see above),
(
14) (a) Parente, J. E.; Risley, J. M.; Van Etten, R. L. J. Am. Chem. Soc.
determining x as 53.5%. Similarly, the 17.5% of doubly
1
984, 106, 8156. (b) Williams, N. H.; Cheung, W.; Chin, J. J. Am. Chem.
18
Soc. 1998, 120, 8079.
15) Error analysis indicates that the deconvolution procedure is the
labeled phosphate, eq 1, reflects 40.7% of O-incorporating
(
1
8
cleavage of O-MP (y in eq 2), fixing y as 43%. The
deviation of the sum of x and y (96.5%) from 100% is
principal source of error; an error of 2% can be assigned to the deconvoluted
integral values, which carries over to the 18O enrichment results.
1792
Org. Lett., Vol. 1, No. 11, 1999