Hydrolysis of cyclic phosphoramides. Evidence for syn lone pair catalysis

Article abstract of DOI:10.1039/b300916e

Hydrolysis between 1.5 < pH < 4 of five and six membered cyclic phosphoramides has been followed by UV and ₃₁PNMR spectroscopy. The observed rates fit the equation: k_obs = k_H2O[H⁺]/([H⁺] + Ka) + k′

Full text of DOI:10.1039/b300916e

View Article Online / Journal Homepage / Table of Contents for this issue
Hydrolysis of cyclic phosphoramides. Evidence for syn lone pair
catalysis†
Andrés Núñez, Dyanna Berroterán and Oswaldo Núñez*
Laboratorio de Fisicoquímica Orgánica y Química Ambiental, Departamento de Procesos y
Sistemas, Universidad Simón Bolívar, Apartado Postal 89000, Caracas, Venezuela
Received 28th January 2003, Accepted 14th May 2003
First published as an Advance Article on the web 30th May 2003
Hydrolysis between 1.5 < pH < 4 of five and six membered cyclic phosphoramides has been followed by UV
and 31PNMR spectroscopy. The observed rates fit the equation: kobs = kH O [Hϩ]/([Hϩ] ϩ Ka) ϩ kЈH O, where kH O
2
2
and kЈH O are the pseudo first-order rate constants of water attack on the2protonated phosphoramide and its
2
unprotonated form, respectively, and Ka is the phosphoramide acidity equilibrium constant. Although, faster
hydrolysis rates on the five membered ring are expected due to the energy released in going from a strained cyclic
to a “strained free” trigonal-bipyramidal-pentacoordinated intermediate, with one of the cyclic nitrogens occupying
the apical position, these compounds react slightly faster (kH O values) but slower regarding the kЈH O values than
2
2
the six membered analogs. The balance in reactivity is attributed to the additional stability obtained in the six
membered cyclic compounds by a syn orientation of the two lone pairs of the cyclic nitrogen to the water attack.
This stabilization does not exist in the five membered phospholidines since the water attack is perpendicular to
the electron pairs of the cyclic nitrogen. In agreement with the incoming water orientation, the product ratios
from the hydrolysis show that in the five membered rings the main product is the one produced by endocyclic
cleavage; meanwhile, in the six membered cyclic phospholines the kinetic product is the one produced by exocyclic
cleavage. The syn orientation of two electron pairs on nitrogen stabilizes the transition state of water approach
to the phosphoramides by ca. 3 kcal molϪ1 when compared to the orthogonal attack.
important point when considering phosphorus as compared
Introduction
with carbon. Due to the involvement of the 3d orbitals, the
attack or departure from phosphorus could be oriented
orthogonal to the plane formed by the n(X) electrons and the
P–X bond. Therefore, a syn or anti-periplanar orientation may
be compared to an orthogonal one, in the latter there is no
possibility for overlap between the n orbital and the leaving or
incoming group σ* orbital. This provides a good model to
compare and quantify the SCT effect. The last point is quite
important since antiperiplanar stabilization has been found
to be relatively small in cyclic orthoesters,9 thioesters10 and in
cyclic amidines5 since the competing path is a syn orientation
that is also recognized11 as a stabilizing one. In fact, it has been
found9,10 that the anti orientation is equally or slightly more
stabilizing than the syn one (see Scheme 1).
The stereoelectronic control theory (SCT)
The stereoelectronic control theory1 (SCT) claims that there is
preferential cleavage when two electron pairs are antiperiplanar
to the leaving group, and SCT has received intense investigation
on several systems. At the aldehyde oxidation level interesting
models based on restricted-conformation flexibility have been
used2 as evidence for SCT. However, against SCT, the principle
of least nuclear motion has emerged3 to explain the relative
reactivity in pyranoside systems. Acyl oxidation hydrolysis of
cyclic-orthoesters1,4 and amidines,5 are some of the classical
studies that have been carried out. The reactivity or kinetic-
product ratios have been presented as evidence in favour
of SCT. The conclusion that one can reach after a general
revision6 is that the SCT effect where it does exist, is weak. A
similar conclusion is reached when attempts are made to extra-
polate the hypothesis to systems where phosphorus is involved.
Although some evidence in favor of the hypothesis has been
presented,7 there is also doubt of its application.6 The latter
conclusion is also supported by theoretical calculations.8
In general, the hypothesis is based on a stabilization intro-
duced by the n–σ* orbital interaction at the transition state.
Qualitatively, its application to phosphorus looks more prom-
ising than to carbon based on two facts: phosphorus is slightly
less electronegative than carbon and the apical bonds (from
which heteroatoms leave or arrive) in the pentacoordinated
intermediate, are elongated relative to the equatorial bonds or
its equivalent carbon bonds. These two properties make the σ*
orbital involved more stable and therefore improves its inter-
action with the n orbital. If the electron donor capacity of the n
orbital is also improved (for instance n(N) instead of n(O)) then
we may observe an effective SCT effect. There is also a third
Scheme 1 Orientation of an incoming group L to yield anti or syn
products. Since the carbon is sp2 hybridized, the axial electron pair on
X, at the transition state is either syn or anti oriented to L. Both
interactions are stabilizing and there is no significant difference between
† Electronic supplementary information (ESI) available: Kobs values at
various pH values. See http://www.rsc.org/suppdata/ob/b3/b300916e/
kanti and ksyn.
T h i s j o u r n a l i s © T h e R o y a l S o c i e t y o f C h e m i s t r y 2 0 0 3
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 2 8 3 – 2 2 8 9
2283

View Article Online
Scheme 2 Water orientation at the phosphorus attack in the five (n = 5) and six (n = 6) membered phosphoramides. For n = 5 the attack is orthogonal
oriented to the electron pairs on the cyclic nitrogen meanwhile, it is syn oriented for n = 6. In agreement with the water orientation, for n = 5 only endo
cleavage is detected in the products and mainly exo cleavage kinetic product is observed for n = 6.
Orientation of the water attack
The mixture was filtered again to separate the triethylamine
hydrochloride and the solvent was evaporated under reduced
pressure. The residual was purified using column chromato-
graphy packed with Al2O3. Chloroform was used as eluent.
A colorless liquid that decomposes when heating was finally
found. The phospholidines (five membered ring phosphor-
amides) or phospholines (six membered ring phosphoramides)
were characterized using GCMS (Shimadzu GC-17A with a
QP-5000 detector), 1H NMR and 31P NMR (Bruker 300 MHz).
The yield and the corresponding signals were as follows:
We propose that in the phosphoramides studied in this work,
the acid catalyzed water attack occurs on the conjugated acid
form of the reactant. Preliminary12 molecular mechanics
optimizations MMϩ and semiempirical PM3 methods, per-
formed on the phosphoramides acid forms, show that the
α conformer is more stable than the corresponding β one. These
calculations also show that the five membered phosphoramides
are less stable than the six membered conformers. Five mem-
bered destabilization is quite well documented13 for the case of
cyclic phosphate esters and is based on the five membered ring
strain. This strain is released when one of the phospholidine
nitrogens is oriented apical (longer bond) in the trigonal bi-
pyramidal pentacoordinated intermediate (TBP). Therefore, in
the five membered ring phosphoramides the water attack must
occur orthogonal to the ring nitrogen electron pairs in order
to accomplish ring release. In the case of the six membered
ring phosphoramides, there is no strain to release but an extra
stabilization to gain (via SCT) at the transition state, if the
water attack occurs syn to the cyclic nitrogen electron pairs
(Scheme 2). A strong piece of evidence for this water attack
orientation, is the fact that in the case of the five membered
phosphoramides only endocyclic cleavage hydrolysis product
is obtained. Meanwhile, in the case of the six membered ring
phosphoramides only exocyclic cleavage kinetic hydrolysis
product is observed.
1,3-Dimethyl-2-morpholinyl-1,3,2-diazaphospholidine-2-oxide
(M5). Yield: 80–90%. Mass Spectrum: m/z (rel. intensity %):
219[M ϩ](8), 133(100), 86(49), 44(91), 42(63). 1H NMR(CDCl3):
ؒ
δ: 2.55 (6H, d, J = 9.60 Hz), 3.10 (8H, m), 3.57 (4H, t, J = 3.96
Hz). 31P NMR: (CDCl3) δ 22.71.
1,3-Dimethyl-2-morpholinyl-1,3,2-diazaphospholine-2-oxide
(M6). Yield: 31%. Mass Spectrum: m/z (rel. intensity %): 233
1
[M ϩ](11), 147(100), 86(23), 44(63), 42(83). H NMR(CDCl3):
ؒ
δ: 1.75 (2H, m), 2.49 (6H, d, J = 10.55 Hz), 3.03 (8H, m), 3.56
(4H, m). 31P NMR: δ 19.93.
1,3-Dimethyl-2-piperidinyl-1,3,2-diazaphospholidine-2-oxide
(P5). Yield: 80–90%. Mass Spectrum: m/z (rel. intensity %):
217[M ϩ](5), 133(20), 84(100), 44(37), 42(29). 1H NMR(CDCl3) :
ؒ
δ: 1.45 (6H, m), 2.48 (6H, d, J = 9.80 Hz), 3.02 (8H, m). 31P
NMR (CDCl3): δ 23.46.
Experimental
1,3-Dimethyl-2-piperidinyl-1,3,2-diazaphospholine-2-oxide
Materials
(P6). Yield: 59%. Mass Spectrum: m/z (rel. intensity %):
231[M ϩ](18), 147(40), 84(100), 44(58), 42(73). 1H NMR-
ؒ
The following compounds were used without further purifi-
cation: phosphorus oxychloride (Aldrich), ethylphosphonic
dichloride (Aldrich), phenylphosphonic dichloride (Aldrich),
N,NЈ-dimethylethylenediamine (Aldrich), N,NЈ-dimethyl-1,3
propanediamine, morpholine (Aldrich), piperidine (Merck),
diethylamine (Merck), benzene (Merck), triethylamine (Riedel-
de Haen) and chloroacetic acid (Merck).
(CDCl3): δ:1.50 (6H, m), 1.82 (2H, m), 2.63 (10H, m), 2.92(4H,
m). 31P NMR (CDCl3): δ 20.63.
1,3-Dimethyl-2-diethylamino-1,3,2-diazaphospholidine-2-oxide
(D5). Yield 80–90%. Mass Spectrum: m/z (rel. intensity %):
205[M ϩ](8), 190(7), 133(98), 72(100), 44(63), 42(55). 1H
ؒ
NMR(CDCl3): δ: 1.06 (6H, t, J = 7.20 Hz), 2.51 (6H, d, J = 9.80
Hz), 3.05 (8H, m). 31P NMR(CDCl3): δ: 28.13.
Synthesis
The general procedure14 used to synthesize the phosphoramides
studied in this work was as follows:
1,3-Dimethyl-2-diethylamino-1,3,2-diazaphospholine-2-oxide
(D6). Yield: 32%. Mass Spectrum: m/z (rel. intensity %):
219[M ϩ](15), 147(100), 72(63), 44(68), 42(82). 1H
ؒ
A solution of 10 mmol of the corresponding secondary
amine and 10 mmol of triethylamine in 25 ml of dry benzene
were added while stirring to a solution of 10 mmol of phos-
phorus oxychloride in 25 ml of dry benzene. The mixture was
kept at room temperature while stirring during two hours and
filtered to separate the triethylamine hydrochloride. To the
resultant liquid, a solution of 10 mmol of N,NЈ-dimethylethyl-
enediamine (or N,NЈ-dimethyl-1,3 propanediamine) and
20 mmol of triethylamine in 25 ml of dry benzene were added.
NMR(CDCl3): δ: 1.14 (6H, m), 1.46 (2H, m), 2.59 (6H, m), 2.97
(8H, m). 31P NMR(CDCl3): δ: 22.14.
1,3-Dimethyl-2-hydroxy-1,3,2-diazaphospholidine-2-oxide
potassium salt (O5). A solution of 0.88 g (1.06 mL, 10 mmol) of
N,NЈ-dimethylethylenediamine and 2.02 g (2.78 mL, 20 mmol)
of triethylamine in 25 ml of dry benzene were added while
stirring over a solution of 1.53 g (0.93 mL, 10 mmol) of
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 2 8 3 – 2 2 8 9
2284

View Article Online
phosphorus oxychloride in 25 ml of dry benzene. The reaction
mixture was kept at room temperature during two hours while
stirring. The liquid was filtered and 2 g (35 mmol) of KOH was
added to the liquid while stirring vigorously. After two hours
the mixture was extracted with water. The water phase was
evaporated under reduced pressure and the solid was suspended
in methanol and filtered. The obtained yield is 70%. 1H
NMR(D2O): δ: 2.31 (6H, d, J = 10.50 Hz), 2.81 (4H, d, J = 11.10
Hz). 31P NMR(D2O/H2O): δ: 27.86.
A typical kinetic is shown in Fig. 1 for compound P5. The rates
constants were directly obtained from a plot of ln(It Ϫ Iinf.) vs.
time; where I is the ratio (phosphoramide/H2P04Ϫ) of 31P NMR
signal integrals. Alternatively, kinetics were also followed by
UV, using diode array HP instrument following the dis-
appearance of the absorption band at 230–240 nm. Com-
parison between the rate constants obtained by NMR and UV
show a good agreement (ca. 5% deviation) between both
methods. To test for general acid catalysis the pH was fixed at
3.2 and the [buffer] was varied in an interval from 1.6 to 2.4 M.
General acid catalysis was not detected. For the proton inven-
tory,16 rates of hydrolysis of M5 were measured at pH = 3.2,
using the 31P NMR technique described above. The D2O frac-
tions used were: 0, 0.1 and 1. The rates as a function of the D2O
fraction (n) show a flat plot in agreement with the equation:
kn = kH(1 Ϫ n ϩ nꢀOH)(1 Ϫ n ϩ nꢀOH) / (1 Ϫ n ϩ nꢀOH),2 where
ꢀOH = fractionation factor of OH = 1, in agreement with the
mechanism shown in Scheme 4.
1,3-Dimethyl-2-ethyl-1,3,2-diazaphospholidine-2-oxide (E5)
and 1,3-dimethyl-2-ethyl-1,3,2-diazaphospholine-2-oxide (E6).
According to the general reported15 procedure, a solution of
1.47 g (1.07 ml, 10 mmol) of ethylphosphonic dichloride in
25 ml of dry benzene were added while stirring to a solution of
10 mmol of N,NЈ-dimethylethylenediamine (for E5) or N,NЈ-
dimethyl-1,3-propanediamine (for E6) and 2.02 g (0.78 ml, 20
mmol) of triethylamine in 25 ml of dry benzene. The reaction
mixture was maintained at room temperature during two hours
while stirring. The solid was filtered and the solvent evaporated
under reduced pressure. The residual was purified using a
column packed with Al2O3 and chloroform as eluent. E5: Mass
Spectrum: m/z (rel intensity %): 162[M ϩ](22), 134(25), 133(76),
ؒ
90(98), 44(100), 42(55). 1H NMR(CDCl3) δ: 0.84 (3H, dt, JHH
7.53 Hz, JPH = 19.87 Hz), 1.76 (2H, dq, JHH = 7.26 Hz, JPH
=
=
16.17 Hz), 2.45 (6H, d, J = 9.90 Hz), 2.00 (4H, m). 31P
NMR(CDCl3): δ: 41.22. E6: Yield 83%. m/z (rel intensity %):
176[M ϩ](15), 147(100), 70(30), 44(69), 42(70). 1H NMR-
ؒ
(CDCl3): δ: 0.96 (3H, m), 1.66 (2H, m), 1.70 (2H, m), 2.58 (6H,
d), 2.95 (4H, m). 31P NMR(CDCl3): δ: 36.65.
1,3-Dimethyl-2-phenyl-1,3,2-diazaphospholidine-2-oxide
(Ph5) and 1,3-dimethyl-2-phenyl-1,3,2-diazaphospholine-2-oxide
(Ph6). 5 mmol of phenylphosphonic dichloride, were dissolved
in 50 ml of dry benzene. This solution was slowly added over a
5 mmol solution of N,NЈ-dimethyl-1,3-propanediamine (Ph6)
or 5 mmol of N,NЈ-dimethylethanediamine (Ph5) in 10 mmol
of triethylamine. After the addition, the reaction was kept
stirring for 45 min. The reaction mixture was filtered to separate
the triethyl ammonium chloride. To the filtered solution, 20 g of
alumina were added and the mixture was kept on stirring for
5 h. The last mixture was filtered and the solution was evapor-
ated on reduced pressure to eliminate benzene. A final yellow
liquid was obtained. This liquid was purified using a column
packed with alumina and dry chloroform was used as eluent.
Compounds were characterized by means of mass spectrometry
Fig. 1 31P NMR spectra for the hydrolysis (pH = 3.0) of P5 at different
times. The signals at 30, 8.0, 7.9 and 0.0 ppm correspond to: P5, I1, I3
and H3PO4, respectively (see Scheme 3). For the kinetics, the ratio of the
signal integrations at 30 ppm and 0 ppm, were used to make the plot:
ln(It Ϫ Iinf) vs. t.
31P NMR signals assignments
Ϫ
In all cases H2PO4 was used as chemical shift reference at
δ = 0.0 ppm.
1,3-Dimethyl-2-hydroxy-1,3,2-diazaphospholidine-2-oxide
potassium salt (O5). As expected the hydrolysis of compound
O5 produces signals at 6.8 ppm and 0 ppm, corresponding to
aminoamide-phosphoric acid and to phosphoric acid, respect-
ively. The reaction at pH < 4 is very fast and the signal of O5
cannot be detected after 1 min when the first 31P spectrum is
taken. However, the signal at 6.8 ppm can be followed. This
signal also disappears rapidly with a t1/2 ca. 3 min at pH = 3. It is
important to recognize that the signal at 6.8 ppm is important
for evaluating products ratio in the hydrolysis of the phos-
phoramides of this work since its intensity measures the
amount of exocyclic cleavage (see Scheme 3, exo path).
1
and NMR (31P and H). Ph5: Yield 87%. m/z (rel intensity %):
210[M ϩ](20), 133(40), 77(23), 42(100). 1H NMR(CDCl3):
ؒ
δ: 2.56 (6H, m), 4.27 (4H, m), 7.46 (5H, m). 31P NMR(CDCl3):
δ: 26.80. Ph6: Yield 90%. m/z (rel intensity %): 224[M ϩ](8),
ؒ
147(19), 77(12), 44(50), 42(100). 1H NMR(CDCl3): δ: 1.98 (2H,
m), 2.42 (6H, d, J = 14.3 Hz), 3.08 (4H, m), 7.36 (3H, m), 7.65
(2H, m). 31P NMR(CDCl3): δ: 32.55.
Kinetics
The hydrolysis kinetics at different pHs were run using chloro-
acetic–chloroacetate as a buffer. Use of phosphate buffer at pH
ca. 3, resulted in formation of byproducts that incorporate the
buffer to phosphorus.
M5, P5 and D5. The 31P NMR signals corresponding to the
hydrolysis of the phospholidines M5, P5 and D5 are summar-
ized in Scheme 3. Only for compound M5 exocyclic cleavage is
observed (signal at 6.8 ppm). The exocyclic product is com-
pound O5. However, its hydrolysis is fast and its 31P NMR
signal is not directly observed. What is observed is the signal at
6.8 ppm corresponding to the intermediate I2 in Scheme 3, that
is, the aminoamide-phosphoric acid. As shown in the scheme,
the latter compound can also be produced from the endocyclic
hydrolysis product intermediate I1, the aminoamide phos-
phoamide. However we have discarded this possibility based
on the fact17 that the endocyclic first intermediate I1 cleaves
further to I3 faster than its cleavage to I2. Therefore, if I2
Each pH was adjusted using a total chloroacetic concen-
tration of 2.0 M. Kinetics were monitored at 24 0.2 ЊC using
31P NMR (300 MHz Bruker). In a typical run, 125 µL of a
solution of 0.25 M phosphoramide or phosphonamide in water
was transferred to a NMR tube of 5 mm. To the tube, 500 µl of
buffer solution were added and spectra were taken in an interval
of 8–60 min (depending on the phosphoramide) using 16 scan/
spectrum. A small amount of NaH2PO4 was added to the
reaction mixture as internal reference. Kinetics were followed
by observing the disappearance of the phosphoramide signal.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 2 8 3 – 2 2 8 9
2285

View Article Online
Scheme 3 Hydrolysis products paths and 31P NMR chemical shift for the five membered ring phospholidines.
Scheme 4 Energy diagram and hydrolysis mechanism for the six membered (—) and five membered (---) phosphoramides.
is observed, it is formed exclusively via exocyclic cleavage. How-
ever, according to the experimental results, its formation is also
limited compared with the endocyclic cleavage, even though the
leaving ability of the morpholine (pKa = 8.618) is favored com-
pared to the endocyclic amine departure (pKa = 9.9)18 corre-
sponding to the second pKa of N,N-dimethylethylenediamine.
(equivalent to I2 in Scheme 3) appears at 7.2 ppm. Meanwhile,
the ones corresponding to the equivalent to I1 intermediate
(endocyclic cleavage in Scheme 3) appear at 8.67 ppm (M6) and
9.64 ppm (P6).
Results and discussion
Product ratio
M6, P6. The hydrolysis products for these two compounds
were also followed with time. The signal (31P NMR) corre-
sponding to the intermediate obtained from exocyclic cleavage
The acid hydrolysis of O5, is quite fast (t ca. 1 min) and it
½
produces only two signals at 6.8 ppm and 0 ppm in the 31P
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 2 8 3 – 2 2 8 9
2286

View Article Online
NMR. These signals correspond to the intermediate I2 of
Scheme 3 and phosphoric acid, respectively. Therefore, I2
corresponds to the exo cyclic product of cleavage. In the
hydrolysis of P5 and D5, the signal at ca. 6.8 ppm is not
detected and a signal at 8–8.2 ppm is observed. We conclude,
that for these two compounds only endocyclic cleavage occurs.
In the case of M5, the signal at 6.8 is observed, but a plot of
exo/endo intensity vs. time shows an intercept of 0.03. Accord-
ing to Scheme 3, compound I2 (with 31P NMR signal at 6.8
ppm) may be produced from I1, however we have discarded17
this possibility. Therefore, the kinetic product for M5 is also
the endo product (endo/exo ca. 97%). With these results we
can conclude that the hydrolysis of the phospholidines (five
membered phosphoramides) produces exclusively endocyclic
cleavage kinetic products.
The hydrolysis of the six membered compounds (phos-
pholines) produce 31PNMR signals that resemble the five
membered ones shown in Scheme 3. However, for these
compounds the kinetic product is the exo one instead of the
endo. As shown in Fig. 2, a plot of endo/exo intensity vs. time
for compound P6 gives an intercept < 0. The results for M6 are
similar to the P6 ones. However, for M6 some kinetic endo
product is detected. This result is unexpected since the pKa
(8.618) of the exocyclic amine morpholine is quite low when
compared to pKa (10.7)18 the first pKa of the amine N,NЈ-
dimethyl-1,3 propanediamine that cleaves via endocyclic
rupture. Therefore, for M6, from consideration of the leaving
ability, more exocyclic product as compared to P6 (with a pKa =
11.1 for the exo group) is expected. Since the result is the
opposite, it is quite probable that the protonation of the amine
instead of its leaving ability is what makes the difference.
Protonation occurs on the more basic amine (endocyclic
nitrogen) and the equilibrium is shifted toward positioning
the endocyclic nitrogen in the apical position via phosphorus
pseudorotation (see Scheme 4) that is promoted based on the
apicophilicity19 that favors the protonated groups at apical
position.
Fig. 3 log kobs vs. pH plot for the phosphoramides M5 (right) and M6
(left). The solid lines correspond to the best fitting of the experimental
points (average kobs values with a uncertainty ≤ 5%) to equation 1. The
kЈH O (high pH leveling) value for M5 is less reliable than the one for
2
M6 since it was obtained from extrapolation of the best fitting value.
Fig. 4 log kobs vs. pH plot for the phosphonamides Ph5 (top) and Ph6
(bottom). The solid lines correspond to the best fit of the experimental
points (average kobs values with an uncertainty ≤ 5%) to equation 1.
Ϫ
(or phosphonamide) equilibrium acidity constant and kOH the
second order rate constant of hydroxy attack to the phos-
phoramide (or phosphonamide) conjugated acid. The second
term of the equation is kinetically equivalent to kЈH O, where
2
this pseudo-first order rate constant corresponds to the water
attack on the conjugated base of the phosphoramide (or
phosphonamide).
From the described best fit, rate constants and Ka values were
obtained. These values are shown in Table 1. In Table 1, the pKa
values titrimetrically obtained are also shown. As shown in
Table 1, the titrimetric and kinetic pKa are in good agreement.
It is worth noting that for the phospholines (six membered
phosphoramides) the pKa values are ca. 1–1.5 pKa lower than
the five membered ones (phospholidines). We attributed this
difference to the steric destabilization induced by the axial
cation water solvation that switches the acid equilibrium toward
its conjugated base in the six membered cycle. The ∆∆GЊ from
this extra destabilization (compared to the five membered
Fig. 2 Relative 31PNMR integration signals vs. time for: exo product
(squares), endo product (triangles) and [endo]/[exo] (circles joined with
straight line) for compound P6 hydrolysis.
From the product ratio measurements vs. time it is concluded
that in the five membered phospholidines the hydrolysis kinetic
product is endo and for the six membered phospholines it is exo.
This provides strong evidence for the water molecule orien-
tation proposal and syn orientation catalysis as the main factors
controlling the process.
analog acidity) is ca. 1.7 kcal molϪ1
.
Ϫ
The kH O and kЈH O ( kOH Kw/Ka) are shown in Table 1 and
2
2
#
#
in Table 2, the ∆G5 Ϫ ∆G6 are also shown. These values are
negative for water attack on the conjugated acid of the
phosphoramides studied (first row of values). However, for
the case of water attack on the conjugated base of the
kobs vs. pH profile
In Fig. 3, the log kobs vs. pH profiles for the hydrolysis of M5
and M6 (phosphoramides) are shown. In Fig. 4, the hydrolysis
profiles for Ph5 and Ph6 (phosphonamides) are also depicted.
The solid lines correspond to the best fit to the equation:
#
#
phosphoramides, the ∆G5 Ϫ ∆G6 values are positive (second
row of values), meaning that there is an extra stabilization in
the six membered compounds that more than balances the
destabilization induced by the ring strain of the five membered
cycle. At this point it is important to emphasize that the values
kobs = k [Hϩ]/([Hϩ] ϩ Ka) ϩ kOH Kw/Ka
(1)
Ϫ
H2O
#
#
(positive or negative) of ∆G5 Ϫ ∆G6 are small meaning that
there is a small difference in reactivity between the 5- and
6-membered ring systems. When the Table 2 third row values
are multiplied by the corresponding Ka and divided by Kw
Where, kH O, corresponds to the pseudo first order rate con-
2
stant for water attack on the conjugated acid of the phos-
phoramides (or phosphonamides), Ka is the phosphoramide
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 2 8 3 – 2 2 8 9
2287

View Article Online
Table 1 Rate and equilibrium constants obtained experimentally by best fit of the experimental points according to equation: kobs = kH O[Hϩ]/([Hϩ]
2
Ϫ
ϩ Ka) ϩ kOH Kw/Ka (or kЈH O). The experimental (titrimetrically obtained) pKa values for the phosphoramides and the corresponding values for the
2
exo cyclic amines, are also shown
n = 5 five membered
n = 6 six membered
D
P
M
Ph
E
05
kH O (minϪ1
)
10%
2
n = 5
n = 6
Ϫ
0.030
0.050
0.17
0.08
0.15
0.10
0.90
0.34
0.20
0.10
> 2.3
0.1
Ϫ1
kOH Kw/Ka (min
n = 5
)
5%
0.003
0.017
0.015
0.050
0.020
0.035
0.002
0.011
0.10
0.07
n = 6
kOH (minϪ1 MϪ1
)
Ϫ
n = 5
n = 6
pKa (kinetic)
n = 5
n = 6
3.8 × 108
1.1 × 1011
4.7 × 1010
2.5 × 1011
2.0 × 109
8.7 × 1010
2.0 × 108
2.0 × 109
2.0 × 1010
7.0 × 109
1 × 1011
2.0
2.9
1.2
(2.3)
2.5
1.3
(1.6)
3.0
1.6
(1.9)
3.0
2.7
2.7
3.0
ꢀꢀGЊ pKa
pKa (titrimetric)
n = 5
3.2
1.4
10.5
3.3
1.9
11.1
3.2
1.9
8.6
3.0
2.7
—
—
3.3
—
—
—
n = 6
pKa (exo group, as free amine)
Table 2 Kinetic energy barriers difference (∆G#5 Ϫ ∆G#6), thermodynamic internal energy difference (∆∆GЊ (5–6)) and syn stereoelectronic control
energy barrier (∆∆G#SCT) for the water attack on the phosphoramide protonated form ∆∆G# (kH O), for the water attack on the unprotonated
2
phoshoramide ∆∆G# (kЈH O) and for the OHϪ attack on the protonated phosphoramide ∆∆G# (kOH
)
Ϫ
2
kcal molϪ1
D
P
M
Ph
E
∆G#5 Ϫ ∆G#6 (kH O), 25 ЊC
0.30
1.03
3.34
Ϫ0.45
0.71
2.35
Ϫ0.24
0.33
2.24
Ϫ0.57
1.01
1.41
Ϫ0.41
Ϫ0.21
Ϫ0.62
2
∆G#5 Ϫ ∆G#6 (kЈH O), 25 ЊC
2
∆G#5 Ϫ ∆G#6 (kOH , 25 ЊC)
Ϫ
∆∆GЊ (5–6) MMϩ Ϫ PM3
Protonated
6.5
2.8
4.1
2.3
3.6
(2.0)
3.0
6.4
5.8
1.7
4.4
1.2
3.8
3.8
3.0
3.4
Non-protonated
∆∆G#SCT (kH O
)
6.8
5.5
2
(4.2)a
3.8
(3.6)a
2.0
∆∆G#SCT (kЈH O
)
2.2
5.8
2.8
3.2
2
∆∆G# (kOH
)
9.8
8.0
Ϫ
SCT
a Corrected for axial solvation destabilization (see text)
∆∆G# = ∆G#5 Ϫ ∆G#6 ϩ ∆∆GЊ
(Kw =1 × 10Ϫ14 M2), the second order rate constant for OHϪ
nucleophilic attack on the protonated phospholidine are
obtained.
diagram of Scheme 4, the SCT energy contribution can be
estimated from the relation: ∆∆G#SCT = ∆G5# Ϫ ∆G6# ϩ ∆∆GЊ.
In Table 2 the values of ∆∆G#
have been obtained from the
SCT
experimental ∆G5# Ϫ ∆G6# data and from ∆∆GЊ semiempirical
PM3 calculations performed on mechanical MMϩ optimized
conformations of the protonated and unprotonated five vs. six
membered phosphoramides. The first observation worth
Proposed mechanism and stereoelectronic control effect (SCT)
In Scheme 4, the energy diagram for the hydrolysis of the five
and six membered phosphoramides is shown. The proposed
mechanism is also depicted. For the case of the six membered
ring phosphoramides (solid line in the energy diagram) the
water attack occurs syn oriented to the in-ring electron pairs.
This syn orientation causes a stabilization of the transition state
that decreases significantly the reaction ∆G# values. Once the
pentacoordinated intermediate is formed exo cleavage proceeds
to yield exo product. The intermediate may also pseudorotate
to orient one of the phospholine nitrogens to an apical position
allowing endo cleavage to occur. If phosphorus pseudorotation
is slow compared to the TBP intermediate cleavage, only kinetic
exo cleavage product is expected. Only exo cleavage will be
observed (even if pseudorotation is fast) if for instance, the
intermediate 3 of Scheme 4 is too unstable when compared to
intermediate 2. For the five membered phospholidines, proton
transfer to one of the phospholidine nitrogens might occur
before the water attack (see also dash lines in the energy
diagram) that is oriented orthogonal to the in-ring electron
pairs to form the TBP intermediate that cleaves to yield endo
cleavage hydrolysis product. However, through pseudorotation
the intermediate may locate the exo group in apical position to
yield exo cleavage product. As deduced from the energy
mentioning, is that the ∆∆G#
values for the mechanism of
SCT
OHϪ attack on the protonated phosphoramides (last row on
Table 2) are the highest values obtained. However, lower
∆∆G#
values are expected for this diffusion controlled
SCT
reaction since it should be less selective. Therefore, this
inconsistency in ∆∆G#
allows us to discard this mechanism
SCT
as the reason for the second leveling observed in the kobs vs. pH
profile and to propose a kinetically equivalent one that consists
of water attack on the unprotonated phosphoramides.
The ∆∆G# values for the mechanism of water attack on the
protonated phospholidine are also high compared to the
corresponding attack of the unprotonated phosphoramides.
However, these values must be corrected for the destabilization
factor (Table 1, four row of data) observed in the reduction
of pKa for the six membered ring as compared to the five
membered ones that is due to axial solvation. This extra
destabilization of the six membered ring protonated phos-
phoramide is not taken into account in the gas phase calcu-
lations. This destabilization energy of the protonated six
membered ring is 1.5–1.9 kcal molϪ1 (see Table 1). When the
∆∆G#SCT data are corrected with these values the obtained SCT
stabilizations are close to the ones corresponding to the water
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 2 8 3 – 2 2 8 9
2288

View Article Online
2 A. J. Kirby, Acc. Chem. Res., 1984, 17, 305; A. J. Kirby and
attack on the unprotonated phosphoramides. As shown in
Table 2, these values are in the range of 2.0–3.8 kcal molϪ1. It
is also important to note that there is no difference between the
∆∆G#SCT obtained for the phosphoramides D,P and M and the
phosphonamides Ph and E. This means that the exocyclic
electron pair on nitrogen of the phosphoramides does not play
an important role on the SCT stabilization. As shown in
Scheme 4, in the intermediate 2 formed by water attack on the
six membered phospholidines there are two electron pairs from
nitrogen oriented syn to the water molecule and a third electron
pair on OH that may be also oriented syn to the water. In the
case of the intermediate 2 from the five membered ring phos-
pholidine, there is one electron pair on the exo amine that may
be oriented syn to the water and one electron pair on OH that
R. J. Martin, J. Chem. Soc., Perkin Trans. 2, 1983, 1627; A. J. Kirby
and R. J. Martin, J. Chem. Soc., Perkin Trans. 2, 1983, 1633; A. J.
Briggs, C. M. Evans, R. Glenn and A. J. Kirby, J. Chem. Soc., Perkin
Trans. 2, 1983, 1637.
3 M. L. Sinnott, Adv. Phys. Org. Chem., 1988, 24, 113.
4 A. J. Kirby, The Anomeric Effect and Related Stereoelectronic Effects
at Oxygen, Springer-Verlag, Berlin, 1983.
5 C. L. Perrin and G. M. L. Arrhenius, J. Am. Chem. Soc., 1982, 104,
2839; C. L. Perrin and O. Núñez, J. Am. Chem. Soc., 1986, 108,
5997; C. L. Perrin and O. Núñez, J. Am. Chem. Soc., 1987, 109, 522;
C. L. Perrin and J. D. Thoburn, J. Am. Chem. Soc., 1993, 115, 3140.
6 The Anomeric Effect and Associated Stereoelectronic Effects,
ed.G. R. J. Thatcher, ACS Symposium Series 539, Washington, DC,
1993.
7 D. G. Gorenstein, Chem. Rev., 1987, 87, 1047.
8 C. Lim and P. Tole, J. Phys. Chem., 1992, 96, 5217; P. Tole and
C. Lim, J. Am. Chem. Soc., 1994, 116, 3922; N. Chang and C. Lim,
J. Am. Chem. Soc., 1998, 120, 2156.
also may be oriented syn. Since there is no difference in ∆∆G#
SCT
between phosphoramides and phosphonamides, the exo amino
group in equatorial position of the five membered ring does not
contribute to the stabilization of its intermediate 2, because
substitution of the amino for ethyl or phenyl does not influence
the estimated ∆∆G#SCT. This is probably due to the entropic
destabilization induced by the loss of two of the three bond
rotation modes. Finally, we must conclude that the values
estimated in this work of ca. 3 kcal molϪ1 correspond to the
stabilization of two electron pairs on nitrogen syn oriented
to an incoming water vs. none (the OH electron pair effect is
cancelled since it is present in the five and six membered
phospholidines) in the formation of HO–P bond.
9 C. L. Perrin and R. E. Engler, J. Am. Chem. Soc., 1997, 119,
585.
10 M. C. Caserio, P. Shih and C. L. Fisher, J. Org. Chem., 1991, 56,
5517.
11 S. Li, A. J. Kirby and P. Deslongchamps, Tetrahedron Lett., 1993,
34, 7757; S. Li and P. Deslongchamps, Tetrahedron Lett., 1993, 34,
7759.
12 HyperChem for Windows, Molecular Mechanics MMϩ and
Semiempirical PM3 Method.
13 R. Kluger and D. T. Scott, J. Am. Chem. Soc., 1990, 112, 6669;
N. Chang and C. Lim, J. Am. Chem. Soc., 1998, 120, 2156.
14 H. Ulrich, B. Tucker and A. Sajigh, J. Org. Chem., 1967, 32, 1360.
15 C. R. Hall, T. D. Inch and N. E. Williams, J. Chem. Soc. Perkin
Trans. 1, 1982, 639.
16 R. L. Schowen, Prog. Phys. Org. Chem., 1972, 9, 275.
17 If I2 is produced from I1, parallel kinetics should be expected when
following I2 and I3 formation. However, a plot of I2/I3 vs. time is
not a straight line.
18 Titrimetrically determined in our laboratory using morpholine and
N,NЈ-dimethylethanediamine (second pKa) for the phospholidines
(five membered cycles) and piperidine and N,NЈ-dimethyl-1,3
propanediamine (second pKa) for phospholines (six membered
cycles).
Acknowledgements
This research was supported by the UGAUSB (Simon Bolívar
University Environmental Management Unit).
References
1 P. Deslongchamps and R. J. Tailleefer, Can. J. Chem., 1975, 53, 3029;
P. Deslongchamps, Stereoelectronic Effects in Organic Chemistry,
Pergamon, Oxford, 1983.
19 G. R. J. Thatcher and R. Kluger, Adv. Phys. Org. Chem., 1989, 25,
99.
O r g . B i o m o l . C h e m . , 2 0 0 3 , 1, 2 2 8 3 – 2 2 8 9
2289