Endocyclic restriction test: Applications to transfers of oxygen from nitrogen and from sulfur to phosphorus(III)

Article abstract of DOI:10.1021/ja9527986

The geometries allowed for formal transfers of oxygen to the phosphorus(III) of a phosphine from a nitrone, an O-acetylhydroxylamine, and a sulfoxide have been evaluated by the endocyclic restriction test. Investigations of the conversions of 1 to 2, 16 to 17, and 28 to 29, by isotopic labeling, substituent effect, and kinetic and spectroscopic experiments, reveal the operation of different mechanisms for each of these transfers. For 1, oxygen addition to phosphorus is the preferred mechanism. In the case of 16, the mechanism involves nucleophilic displacement of oxygen from nitrogen by phosphorus to give 26 followed by oxygen addition to phosphorus. In acetic acid, the oxygen is added to 26 from water in the workup whereas in toluene the oxygen is provided by the acetate produced by the displacement. For 28, either addition by oxygen of the sulfoxide to activated phosphorus or addition by phosphorus to sulfur of the sulfoxide precedes oxygen transfer.

Full text of DOI:10.1021/ja9527986

3426
J. Am. Chem. Soc. 1996, 118, 3426-3434
Endocyclic Restriction Test: Applications to Transfers of
Oxygen from Nitrogen and from Sulfur to Phosphorus(III)
Mitchell L. Kurtzweil and Peter Beak*
Contribution from the Department of Chemistry, UniVersity of Illinois at Urbana-Champaign,
Urbana, Illinois 61801
ReceiVed August 14, 1995. ReVised Manuscript ReceiVed February 7, 1996X
Abstract: The geometries allowed for formal transfers of oxygen to the phosphorus(III) of a phosphine from a
nitrone, an O-acetylhydroxylamine, and a sulfoxide have been evaluated by the endocyclic restriction test.
Investigations of the conversions of 1 to 2, 16 to 17, and 28 to 29, by isotopic labeling, substituent effect, and kinetic
and spectroscopic experiments, reveal the operation of different mechanisms for each of these transfers. For 1,
oxygen addition to phosphorus is the preferred mechanism. In the case of 16, the mechanism involves nucleophilic
displacement of oxygen from nitrogen by phosphorus to give 26 followed by oxygen addition to phosphorus. In
acetic acid, the oxygen is added to 26 from water in the workup whereas in toluene the oxygen is provided by the
acetate produced by the displacement. For 28, either addition by oxygen of the sulfoxide to activated phosphorus
or addition by phosphorus to sulfur of the sulfoxide precedes oxygen transfer.
Reactions involving transfers of oxygen from heteroatoms
radical reaction, and (5) initial bonding between oxygen and
phosphorus in an addition-elimination sequence. Mechanisms
3-5 are also applicable to an oxygen formally multiply bonded
to a heteroatom. For multiply bonded systems two other
possibilities, (6) addition to the heteroatom or (7) addition to
the adjacent atom, are plausible mechanisms.
to phosphorus(III) giving the deoxygenated heteroatom and the
corresponding phosphine oxides are well-known processes.1 A
number of different mechanisms have been suggested for these
oxidations. The present work evaluates the geometries allowed
for transfers of oxygen from a nitrone, an O-acetylhydroxy-
lamine, and a sulfoxide to phosphorus(III) by use of the
endocyclic restriction test in order to distinguish between the
alternative possible mechanisms.2
The possible mechanisms for reactions of heteroatom-oxygen
single bonds with the phosphorus(III) of a phosphine are shown
in generalized form in eqs 1-5.4-13 The mechanisms are (1)
classic SN2 substitution at oxygen,3 (2) classic SN2 substitution
at the heteroatom followed by addition of oxygen to phosphorus,
(3) biphilic insertion into the oxygen heteroatom bond, (4) a
X Abstract published in AdVance ACS Abstracts, March 15, 1996.
(1) (a) Kirby, A. J.; Warren, S. G. The Organic Chemistry of Phosphorus;
Elsevier: New York, 1967. (b) Emsley, J.; Hall, D. The Chemistry of
Phosphorus; John Wiley and Sons: New York, 1976. (c) Smith, D. J. H.
In ComprehensiVe Organic Chemistry; Barton, D. H. R., Ollis, W. D., Eds.;
Pergamon Press: Oxford, England, 1979; pp 1127-1188. (d) Rowley, A.
G. In Organophosphorus Reagents in Synthesis; Cadogan, J. I. G., Ed.;
Academic: New York, 1979; pp 295-350.
(2) Beak, P. Acc. Chem. Res. 1992, 25, 215. We note that endocyclic
restriction provides only a permissive test of reaction geometry (vide infra).
(3) While a literal definition of an SN2 reaction denotes a second-order
nucleophilic substitution at carbon, we believe the experimental and
theoretical precedents are sufficient to use the term classic SN2 to denote
nucleophilic substitution which proceeds at any atom through a trigonal
bipyramidal transition state with the ending and leaving groups in the apical
positions.
(4) For the first report of deoxygenation of a nitrone by a phosphine,
see: Horner, L.; Hoffmann, H. Angew. Chem. 1956, 68, 473.
(5) (a) Hashimoto, S.; Furukawa, I.; Fujimoto, S. Nippon Kagaku Kaishi
1972, 391; Chem. Abstr. 1972, 76, 112403. (b) Hashimoto, S.; Furukawa,
I.; Katami, T. Nippon Kagaku Kaishi 1974, 511; Chem. Abstr. 1974, 80,
145023.
Mechanistic investigations of the reactions of phosphorus-
(III) with nitrones and with O-acylhydroxylamines have been
reported.4 Hasimoto et al. have proposed deoxygenation of
nitrones by triethyl phosphite occurs by initial addition of
phosphorus to nitrogen followed by formation of an oxygen
phosphorus bond and cleavage of the nitrogen phosphorus bond
to produce trimethyl phosphate (eq 6).5 Milliet and Lusinchi
have proposed initial addition of trimethyl phosphite to the
iminyl carbon of the nitrone followed by loss of a methyl group
from a phosphite oxygen, bonding of the oxygen to phosphorus,
and elimination of dimethyl phosphate to give the imine (eq
7).6 The reaction of triphenyl phosphine and N,O-dibenzoyl-
hydroxylamine has been shown by Wasserman and Koch to
(6) Milliet, P.; Lusinchi, X. Tetrahedron 1979, 35, 43.
(7) Wasserman, H. H.; Koch, R. C. Chem. Ind. 1956, 1014.
(8) Stec, W.; Okruszek, A.; Michalski, J. Bull. Acad. Pol. Sci., Ser. Sci.
Chim. 1973, 21, 445.
(9) Emerson, T. R.; Rees, C. W. J. Chem. Soc. 1964, 2319. Ramirez, F.;
Aguiar, A. M. Abstracts of Papers, 134th Meeting of the American Chemical
Society; American Chemical Society: Washington, DC, 1958; 42N.
(10) Amonoo-Neizer, A. H.; Ray, S. K.; Shaw, R. A.; Smith, B. C. J.
Chem. Soc. 1965, 4296.
(11) Luckenbach, R.; Herweg, G. Liebigs Ann. Chem. 1976, 2305.
(12) Szmant, H. H.; Cox, O. J. Org. Chem. 1966, 31, 1595.
(13) Olah, G. A.; Gupta, B. G. B.; Narang, S. C. Synthesis 1978, 137.
0002-7863/96/1518-3426$12.00/0 © 1996 American Chemical Society

Endocyclic Restriction Test
J. Am. Chem. Soc., Vol. 118, No. 14, 1996 3427
proceed by displacement at nitrogen by phosphorus to give a
stable phosphorus ylide (eq 2).7 In the transfer of oxygen from
an amine oxide to phorphorus(III), Stec observed a high degree
of retention at phosphorus and proposed an addition-elimination
sequence (eq 5).8 Similar pathways have been suggested by
Rees, Ramirez, and Milliet.6,9
Three different mechanisms involving valence expansion at
phosphorus followed by loss of the phosphine oxide have been
proposed for the transfer of oxygen from the sulfur of a sulfoxide
to phosphorus(III) by Amonoo-Neizer et al.10 These mecha-
nisms include initial nucleophilic attack of oxygen at phosphorus
to give a valence-expanded intermediate (eq 5), initial nucleo-
philic attack of phosphorus at sulfur to give a thiophosphonium
salt which undergoes ring closure to form a three-membered
ring intermediate (eq 6), and direct formation of the three-
membered ring intermediate through biphilic insertion of
phosphorus (eq 3).10 In separate studies Luckenbach, Szmant
and Olah have shown that oxygen transfers from sulfoxides to
phosphorus(III) require a Lewis acid and have proposed
mechanisms with electrophilic or nucleophilic attack by phos-
phorus at oxygen (eq 5).11-13 The geometrical dependence of
oxygen transfer in these reactions has not been determined.
We have previously used the endocyclic restriction test to
provide evidence that the transfer of oxygen from a hydroxy-
lamine to the phosphorus(III) of an arylphosphine does not
follow a classic SN2 pathway at either oxygen or nitrogen (eqs
1 and 2) and is not a biphilic insertion (eq 3) nor a radical
reaction (eq 4). We proposed that the mechanism involves
oxygen addition to phosphorus via a pentavalent phosphorane
(eq 5).14
Figure 1. Crystallographic structure of 1.
sion to the Grignard reagent and treatment with the appropriate
diarylphosphinous chloride gave the corresponding phosphine
which was deprotected and condensed with N-tert-butylhy-
droxylamine to provide the nitrones 1, 5, and 6. An X-ray
crystallographic structure of 1 established that the oxygen and
the aryl ring bearing the triarylphosphine are syn disposed
(Figure 1). Synthesis of doubly labeled 1-13C,18O for the
endocyclic restriction test (vide infra) was carried out by
lithiation of (2-bromophenyl)diphenylphosphine with t-BuLi
followed by reaction with methyl formate-13C to give labeled
aldehyde. Reaction of the aldehyde with N-tert-butylhydroxy-
lamine-18O afforded 1-13C,18O.17,18
We now report application of the endocyclic restriction test
to formal transfers of oxygen from the nitrone, an acetylhy-
droxylamine, and sulfoxide groups to phosphorus(III). The
geometrical dependence of the transfer is evaluated by the
reaction shown in general form in eq 8. If direct transfer can
occur at an oblique phosphorus-oxygen-heteroatom angle, the
reaction should be intramolecular when the tether is short and
n is a small number. If a large phosphorus-oxygen-hetero-
atom angle is required for direct transfer, the reaction will be
intermolecular for short tethers and intramolecular for long
tethers.2,15 The present work illustrates the information that can
be gained for three different reactions in which n ) 1.
Results and Discussion
Transfer of Oxygen from a Nitrone to Phosphorus(III).
The substrates 1, 5, and 6 have been used to investigate the
transfer of oxygen from a nitrone to phosphorus(III). The
syntheses of 1, 5, and 6 were carried out by modification of the
synthesis of 2-(diphenylphosphino)benzaldehyde reported by
Rauchfuss.16 Protection of 2-bromobenzaldehyde with ethylene
glycol afforded the corresponding 2-bromodioxolane. Conver-
The transfer of oxygen from nitrogen to phosphorus(III)
occurs on heating 1 at 110 °C for 57 h in toluene to give
phosphine oxide 2. This product was not isolable in pure form
but was identified by its spectral properties. Crude 2 was
converted by reduction to 3 and by hydrolysis on chromatog-
raphy to 4. An authentic sample of 3 was prepared by reaction
of 4 with tert-butyl amine followed by reduction. On heating
in toluene, compounds 5 and 6 were converted to the imine-
phosphine oxides 7 and 8, respectively, and these imines were
hydrolyzed to the corresponding benzaldehydes 9 and 10.
(14) Kurtzweil, M. L.; Loo, D.; Beak, P. J. Am. Chem. Soc. 1993, 115,
421.
(15) In addition to the seminal work of Eschenmoser and Hogg cited
for the endocyclic restriction test, we want to include a number of other
precedents which have been brought to our attention:2 Cruickshank, P.;
Fishman, M. J. Org. Chem. 1969, 34, 6061. Wudl, F.; Lee, T. B. K. J. Am.
Chem. Soc. 1973, 95, 6349. Danishefsky, S.; Dynak, J.; Hatch, W. E.;
Yamamoto, M. J. Am. Chem. Soc. 1974, 96, 1256. Lok, R.; Howard, J. K.
Bioorg. Chem. 1976, 5, 169.
(16) Hoots, J. E.; Rauchfuss, T. B.; Wrobleski, D. A. Inorg. Synth. 1982,
21, 175.
(17) Woods, K. W.; Beak, P. J. Am. Chem. Soc. 1991, 113, 6281.
(18) There is 4% 1-13C,16O in the doubly-labeled reactant.

3428 J. Am. Chem. Soc., Vol. 118, No. 14, 1996
Kurtzweil and Beak
Scheme 1
The molecularity of the conversion of 1 to 2 was determined
by double labeling experiments carried out with 1 and 1-13C,18O
as outlined in Scheme 1. If the reaction is intramolecular, the
label integrity of the reactants will be preserved to give 2 and
2-13C,18O in the same isomeric ratio as the reactants 1 and
1-13C,18O, and provide 4 and 4-13C,18O. If the reaction is
intermolecular, then four products, 2, 2-13C, 2-18O, and 2-13C,18O,
will be obtained with a statistical distribution of the isotopic
labels, and lead to 4, 4-13C,18O, 4-13C, and 4-18O.
A 59:36 mixture of 1 and 1-13C,18O in toluene at a combined
concentration of 0.075 M was heated at 100 °C for 18 hours.15
Purification of the product mixture by chromatography on silica
gave 4 as the hydrolysis product of 2 along with unreacted 1.
Table 1 shows the results expected for intramolecular and
intermolecular oxygen transfers and the experimental isotopic
distributions for 4 as determined by FI/MS. A similar distribu-
tion was obtained in an additional double labeling experiment
carried out for 20 h.
The results in Table 1 establish that oxygen transfer is an
intramolecular reaction in the conversion of 1 to 2. Therefore,
the geometry of a classic SN2 attack by phosphorus at oxygen
(eq 1) does not have to be available for reaction because the
constraints of the six-membered endocyclic ring would not allow
a linear P-O-N geometry.2,3 A classic SN2 reaction at the
nitrogen (eq 2) would require a front side displacement at an
sp2 atom which seems unlikely. The operations of a biphilic
insertion or in-cage radical recombination (eqs 3 and 4) are not
ruled out by the data but seem less likely than the alternatives
(vide infra). This labeling result does rule out the possibility
of an intermolecular oxygen transfer by a radical chain mech-
anism initiated by an adventitious radical source or by N-O
bond cleavage.
Table 1. Isotopic Distribution of 1 and 4 for the Double Labeling
Experiment with 1 in Toluene at 100 °Ca
reactant
theoretical
intramolecular intermolecular
theoretical
experiment
label
1
4
12C16O
13C16O
12C18O
13C18O
59
4
1
59
4
1
38
25
22
15
56
7
1
36
36
36
a Error is estimated to be (5% for the FIMS analysis.
Scheme 2
Three mechanisms which are consistent with intramolecular
transfer of oxygen in the conversion of 1 to 2 are shown in
Scheme 2. In mechanism A, which is a specific case of eq 6,
phosphorus attack at nitrogen to form intermediate 11 is
followed by oxygen-phosphorus bond formation to give
intermediate 12. Cleavage of the P-N and N-O bond of 12
gives 2. Mechanism B, which derives from eq 7, involves initial
addition by phosphorus to the iminyl carbon, leading to 13.6
Subsequent addition of oxygen to phosphorus provides 14 which
undergoes bond cleavage to 2. In mechanism C, which is an
example of eq 5, initial attack by oxygen at phosphorus gives
15, which upon subsequent N-O bond cleavage provides 2.8
If it is assumed that the first step of these mechanisms is rate
determining, substituent effects may be used to distinguish
between the possibilities.19 In mechanisms A and B, there is
electron donation by phosphorus in the first step, whereas in
mechanism C, phosphorus accepts electrons. Consequently,
electron-donating groups on the aryl rings which are bonded
only to phosphorus should accelerate oxygen transfer and
electron-withdrawing groups should decelerate the transfer by
mechanisms A and B. The opposite order of substituent effects
would be expected for mechanism C.
We have carried out a limited substituent effect study with
substrates 1, 5, and 6. The CF3 and CH3O groups were chosen
because they would be expected to have large effects on the
relative rates of oxygen transfer. The conversions of 1 to 2 , 5
to 7, and 6 to 8 were followed by NMR at 100 °C. First-order
kinetics were observed for each reactant with values of kobs
(19) Hammett studies have been used to distinguish between alternative
mechanisms for reactions at phosphorus with absolute values of F as low
as 0.35: (a) Sundberg, R. J.; Lang, C. J. Org. Chem. 1971, 36, 300. (b)
Shulman, J. I. J. Org. Chem. 1977, 42, 3970. (c) Baumstark, A. L.; Vasquez,
P. C. J. Org. Chem. 1984, 49, 793. (d) Lloyd, J. R.; Lowther, N.; Hall, C.
D. J. Chem. Soc., Perkin Trans. 2 1985, 245. (e) Jarvis, B. B.; Marien, B.
A. J. Org. Chem. 1976, 41, 2182.

Endocyclic Restriction Test
J. Am. Chem. Soc., Vol. 118, No. 14, 1996 3429
(×10-2 h-1) of 1.99 ((0.15), 7.2 ((1.3), and 1.59 ((0.07) for
1, 5, and 6, respectively. The relative rate order of 5 > 1 > 6
is consistent with pathway C and inconsistent with pathways A
and B.20
which the products were isolated gave 18, 20, and the hydrolysis
product 21 in 9%, 32%, and 29% yields, respectively.23
These results provide the first determination of the geometry
of transfer of a formally negative oxygen to phosphorus. It
should be noted that the endocyclic restriction test is permissive
only; it establishes that the oxygen of a nitrone can be transferred
to phosphorus at an oblique angle. Recent calculations which
favor a large P-O-N bond angle for the transition state
structure for this transfer suggest further studies to determine
the preferred pathway are warranted.21 It also is interesting to
note that the X-ray structure of 1 (Figure 1) which has a close
approach of the phosphorus to the nitrogen and a large separation
of the phorphorus and oxygen might have been taken to predict
reaction via 13. In this case the structure-reaction correlation
model would not have been useful.22
Transfer of Oxygen from an O-Acetylhydroxylamine to
Phosphorus(III). The transfer of oxygen from an acetoxy
group on nitrogen to phosphorus(III) has been investigated with
16. The products obtained by heating 16 under different
conditions are 17-21.
The kinetics of the conversion of 16 to 17 were measured in
deuterioacetic acid at 40 °C. The reaction is first-order with a
rate constant of (9.76 ( 0.54) × 10-2 min-1 (r2 ) 0.992). The
potential intermediates for a bimolecular reaction of 16 with
itself were tested for kinetic competence. When an acetic acid
solution of 19 and 22, 0.075 M in each, was heated at 100 °C
for 1.75 h, only a minor amount of 17, along with major amounts
The O-acetylhydroxylamine 16 was synthesized from 21 by
reaction with N-isopropylhydroxylamine, reduction, and acetyl-
ation. Independent syntheses of 17 and 18 were accomplished
by treatment of 4 with N-isopropylamine followed by reduction
to give 17 and by acetylation of 17 to give 18. Compound 19
was synthesized by treatment of 4 with isopropylhydroxylamine
followed by reduction and subsequent acetylation. The synthesis
of 22 (vide infra) was accomplished by treatment of 21 with
isopropylamine followed by reduction. Both 16a-18O and 16b-
18O were prepared for labeling studies (vide infra). Treatment
of 4 with N-isopropylhydroxylamine-18O followed by reduction
and acetylation gave 16a-18O with 74% enrichment. Labeled
16b-18O was prepared with 24% incorporation by treatment of
25 with acetic acid-18O in the presence of DCC.
1
of 4, 23, and unreacted 22, was detected by H NMR. Thus,
the kinetics, as well as the product profile on exposure of
possible intermediates to the reaction conditions, rule out a
classic bimolecular SN2 reaction.24 These results also rule out
a radical chain reaction which would be greater than first order
in 16 if initiated by 16.
The reaction mechanism could involve 24 formed via a classic
intramolecular SN2 reaction by phosphorus at nitrogen followed
by addition of acetate to phosphorus or by direct biphilic
insertion.19,25 The intermediate 24 could then lose an acetyl
group and undergo ring opening to 17. This possibility was
investigated initially by 18O-labeling experiments.
Oxygen transfer appears to occur in the conversion of 16 to
17. When 16 is heated in acetic acid at 100 °C, 17 is produced.
Heating 16 in toluene at 100 °C affords a more complex product
mixture containing 18 (20%), 19 (20%), and 20 (45%) on the
basis of NMR analysis. Another experiment in toluene from
When 16a-18O was heated in acetic acid at 100 °C for 1.75
h, the product 17 obtained after aqueous workup was found by
(20) The possibility that the steps 11 to 12 to 2 or 13 to 14 to 2 are the
rate-determining steps or that 12 and 14 are formed directly cannot be ruled
out definitively. However, since those mechanisms are more complex and
not required by the data, mechanism C is preferred.
(23) The isolation experiment was carried out by D. K. Loo: Loo, D.
K. Ph.D. Thesis, University of Illinois at Champaign-Urbana, 1987 (available
from Dissertation Abstracts, Ann Arbor, MI). The differences between the
NMR and isolation experiments are attributed to hydrolysis during the latter
and the possible presence of 22 in the former which was not detected due
to overlapping signals.
(24) The SN2 mechanism proposed for reaction of the O-O linkage in
diacyl peroxides with phosphorus has been tested kinetically: Greenbaum,
M. A.; Denney, D. B.; Hoffman, A. K. J. Am. Chem. Soc. 1956, 78, 2563.
Horner, L.; Jurgeleit, W. Liebigs Ann. Chem. 1955, 591, 138.
(21) Bachrach, S. M. J. Org. Chem. 1990, 55, 1016.
(22) Rosenfield, R. E., Jr.; Parhasarathy, R.; Dunitz, J. D. J. Am. Chem.
Soc. 1977, 99, 4860. Burgi, H. B.; Dunitz, J. D. Acc. Chem. Res. 1983, 16,
153. Burgi, H. D.; Dunitz, J. D.; Shefter, E. J. Am. Chem. Soc. 1973, 95,
5065. Cases have been discussed for which the X-ray structure does not
indicate a probable reaction trajectory: Venvogoplan, P.; Venkatesan, K.;
Klausen, J.; Novotny-Bregger, E.; Leuman, C.; Eschemoser, A.; Dunitz, J.
D. HelV. Chem. Acta 1991, 74, 662.

3430 J. Am. Chem. Soc., Vol. 118, No. 14, 1996
Kurtzweil and Beak
mass spectral analysis to have no excess 18O over natural
abundance. Reactions of 16a and 16b were carried out at 56
°C to allow for the recovery of reactant along with the formation
of product. Both reactants were found to have full label
integrity, and 17 from either labeled reactant 16a or 16b was
found to contain no detectable level of 18O isotopic enrichment.
A control experiment showed that 18O label was not lost from
the product under the reaction or workup conditions. To test
the possibility that the loss of label occurred in an intermediate,
which exchanges its label with the acetic acid, a reaction was
carried out in which 16 was heated in acetic acid-18O containing
60% 16O2, 34% 16O18O, and 6% 18O2 label at 100 oC for 5 min.25
Analysis of the product after aqueous workup showed that there
was no excess 18O incorporation in the 17 produced from this
reaction.
Scheme 3
neutralize the residual AcOH. Mass spectral analysis showed
the product from this sequence 17 to have 83% 18O enrich-
ment.28
Direct evidence for 26 as a stable intermediate was obtained
by 31P NMR. When 16 is heated in acetic acid-d4, a new 31P
NMR signal is observed at +50.7 ppm. The signal remains
unchanged after 2 h at 100 °C. After several days at room
temperature, a minor 31P NMR signal which corresponds to that
of 17 appears at +40.5 ppm. Thus, 26 is assigned to the 31P
NMR signal at 50.7 ppm. The salt 26 was isolated by removal
of acetic acid in vacuo and shown to have 31P chemical shifts
of +50.2 ppm in methanol-d and +49.7 ppm in C6D6. This
chemical shift assignment is consistent with the 31P NMR
chemical shift of +47.6 ppm (CDCl3) for the isolable salt
(dimethylamino)triphenylphosphonium chloride and also with
assignments made by Evans and by Krannich for similar
structures.29-31
The pathway for the conversion of 16 to 17 in toluene was
also investigated. On heating 16 in toluene at 100 °C, the 31P
NMR of 16 at -15.1 ppm disappeared after 3 h. Signals for
20 appear at -12.1 ppm, along with four smaller signals in the
region between +29 and +32 ppm which correspond to the
phosphine oxides 18 and 19, and an unidentified product. Two
signals are attributed to 18 because of rotational isomerism about
the imide bond. The 31P NMR signal which is observed at
+49.4 ppm is similar to the value of +49.7 in C6D6 for 26
(vide supra), and the same assignment is made. When toluene
was added to 26, prepared by removal of acetic acid (vide supra),
and the reaction mixture heated at 100 °C for 3 h, 18 was
obtained in 70% yield. Heating 16a-18O (74% 18O) in toluene
provided 18-18O with 71% 18O enrichment. These results
establish that 18 can be produced from 26 in toluene and that
the oxygens on phosphorus and in the acetyl group come from
the acetate.
Since neither the acetyl group of 16 nor the acetic acid solvent
provides the oxygen which appears in 17 from the conversion
of 16 to 17 in acetic acid, other sources for the oxygen were
investigated. Air oxidation, as well as more complicated
possibilities involving dioxygen, is possible.26 However, when
16 was heated in AcOH saturated with 97% 18O2 under an 18O2
atmosphere at 100 °C for 10 min, 17 was obtained with no 18
O
enrichment.
A mechanism consistent with the loss of label is displacement
at nitrogen to give an intermediate, which is a stable species
and does not react further in acetic acid. Thus, a displacement
by phosphorus at the nitrogen of 16 could afford 26 as a stable
species which subsequently reacts with water only on workup
to give 27 that leads to 17. Analogies for the stability of 26
may be found in the reaction of triphenylphosphine with an
O-benzoylhydroxylamine to give a phosphorus-nitrogen ylide
and in nucleophilic substitutions by triphenylphosphine and
O-acyl-, O-phosphinyl-, or O-sulfonylhydroxylamines to give
aminophosphonium salts.7,27
When a mixture of doubly-labeled 16-13C,18O and unlabeled
16 was heated in toluene, a statistical mixture of 18, 18-13C,
18-18O, and 18-13C,18O was obtained as shown in Scheme 3
and Table 2. The recovered reactant was found to have intact
isotopic content, and a control study in acetic acid showed that
the 18O label was retained in the product (vide supra). These
results can be taken to show label scrambling had not occurred
In order to test this mechanism, a reaction was carried out in
which 16 was heated in AcOH at 100 °C for 5 min, followed
by removal of the solvent. The residue was treated with an
excess of H218O with 98% 18O followed by NaHCO3 to
(28) Dilution of the 18O label from 98% to 89% would be excepted for
statistical redistribution of the 18O into the NaHCO3.
(29) Evans, S. A., Jr.; Pautard, A. M. J. Org. Chem. 1988, 53, 2300.
(30) Krannich, L. K.; Kajolia, R. K.; Watkins, C. L. Org. Magn. Reson.
1987, 25, 320.
(31) The possibility that 26 is in equilibrium with appreciable amounts
of 24 is unlikely since phosphoranes containing a structure similar to 24
would be expected to have 31P NMR chemical shifts from -10 to -70
ppm. For typical acyclic examples, see: Chang, L. L.; Denney, D. B.;
Denney, D. Z.; Kazior, R. J. J. Am. Chem. Soc. 1977, 99, 2293. For typical
cyclic and spirobicyclic examples, see: Dennis, L. W.; Bartuska, V. J.;
Maciel, G. E. J. Am. Chem. Soc. 1982, 104, 230.
(25) Biphilic reactions have been proposed: Lloyd, J. R.; Lowther, N.;
Hall, C. D. J. Chem. Soc., Perkin Trans. 2 1985, 245. Clennan, E. L.; Heah,
P. C. J. Org. Chem. 1984, 49, 2284. Denney, D. B.; Denney, D. Z.; Hall,
C. D.; Marsi, K. L. J. Am. Chem. Soc. 1972, 94, 245.
(26) Watts, G. B.; Ingold, K. V. J. Am. Chem. Soc. 1972, 94, 2528.
(27) Tamura, Y.; Minamikawa, J.; Sumoto, K.; Fujii, S.; Ikeda, M. J.
Org. Chem. 1973, 38, 1239. Harger, M. P. J. Chem. Soc., Perkin Trans. 1
1981, 3284. Krueger, J. H.; Sudbury, B. A.; Blanchet, P. F. J. Am. Chem.
Soc. 1974, 96, 5733.

Endocyclic Restriction Test
J. Am. Chem. Soc., Vol. 118, No. 14, 1996 3431
Table 2. Isotopic Distribution of 16 and 18 for the Double
Labeling Experiment with 16 in Toluene at 100 °C
tolylphosphine with enantiomerically enriched menthyl benzene-
sulfinate.34,35
experimenta
When 28 is heated in acetic acid at 50 °C, 29 is produced in
85% yield. Although 28 is stable to heating in DMF at 80 °C
for 4.5 h, when 28 is heated in DMF at 75 °C in the presence
of 1 equiv of CCl4 or I2, 29 is produced along with a small
amount of 30. Oxidative promotion of this type of oxygen
transfer reaction has been reported by Vedejs et al.33
reactant
16
theoretical theoretical
intramolecular intermolecular
label
18
12C16O
13C16O
12C18O
13C18O
49
17
7
49
17
7
36
29
19
15
40
29
18
12
28
28
a Error is estimated to be (5% for the FIMS analysis.
prior to or after oxygen transfer. The data in Table 2 show
that the conversion of 16 to 18 in toluene involves intermolecular
transfer of the oxygen from the acetoxy group to phosphorus.
Thus, the first step of the mechanisms for the conversion of
16 to 17 in acetic acid and to 18 in toluene is formation of the
aminophosphonium salt 26. In acetic acid, 26 provides 17 upon
addition of water. In toluene, the phosphonium salt could add
acetate to give 24. Rearrangement of 24 with acyl transfer to
nitrogen and ring opening could provide 18. The question of
whether the rearrangement of 24 to 18 is intramolecular or
intermolecular is unresolved. In order for 24 to come directly
from 16, there would need to be a step which allows the
observed isotopic scrambling. The product 19 is considered to
be formed by oxidation by adventitious oxygen. The failure
of 19 and 22 to provide products in toluene rules out these
compounds as intermediates.32
Benzyl tolyl sulfoxide has been shown to undergo thermal
racemization at sulfur at 135-155 °C by carbon-sulfur bond
cleavage.36 Although the reactions of 28 occur at 100 °C, a
similar process with 28 would compromise any conclusion about
transition structure geometry since the bond-breaking and bond-
forming steps would be disassociated. The possibility of
carbon-sulfur bond cleavage prior to oxygen transfer was
addressed for 28 by exposure of enantioenriched 28 to the
reaction conditions using Mislow’s approach.35 If the enan-
tioenriched reactant is unchanged, then a mechanism involving
carbon-sulfur bond cleavage and recombination prior to oxygen
transfer is unlikely. If the reactant is racemized, then reversible
formation of intermediate 32 is indicated.
Treatment of (R)-28 (64% ee) in AcOH at 50 °C for 2 h and
separately at 77 °C in DMF in the presence of CCl4 for 2.3 h
afforded 17% of unreacted 28 (66% ee) and 50% of unreacted
(R)-28 (70% ee), respectively. These results show that 28 is
stable to racemization under the reaction conditions and the
reversible formation of 32 can be discounted.
The conversion of 28 to 29 was found to follow first-order
kinetics with a rate constant of (9.10 ( 0.54) × 10-3 min-1 (r2
) 0.992) for the reaction in deuterioacetic acid at 50 °C. This
result rules out a classic SN2 substitution at oxygen and a radical
chain process.
In an attempted double-labeling experiment reaction of 28
and 28-13C,18O in acetic acid, no 18O label was observed in the
product 29.37 To test for the possibility of a reaction intermedi-
ate which undergoes exchange under the reaction conditions,
28 was heated in acetic acid containing 60% 16O2, 34% 16O18O,
and 6% 18O2 label. The 29 isolated from this experiment had
20% 18O incorporation as determined by FI/MS analysis. This
Transfer of Oxygen from a Sulfoxide to Phosphorus(III).
The transfer of oxygen from sulfur to phosphorus(III) has been
investigated for the conversion of 28 to 29 in acetic acid and in
DMF, the latter with oxidative catalysis.11-13,33 The synthesis
of 28 was carried out by lithiation of diphenyl-o-tolylphosphine
with n-BuLi/TMEDA followed by reaction with methyl ben-
zenesulfinate. For 28-13C,18O, diphenyl-o-tolylphosphine-13C
was synthesized from diphenyl(o-bromophenyl)phosphine by
halogen-metal exchange with n-BuLi and reaction with 13CH3I.
The labeled compound was lithiated and allowed to react with
methyl benzenesulfinate-18O to provide 28-13C,18O. The methyl
benzenesulfinate-18O was prepared from the sodium salt of
benzenesulfinic acid by treatment with thionyl chloride, H218O,
and diazomethane sequentially. An authentic sample of 29 was
prepared in two steps by lithiation of diphenyl-o-tolylphosphine
with n-BuLi/TMEDA followed by addition of diphenyl disulfide
to give 31. Oxidation of 31 with 30% H2O2 led to 29. The
synthesis of enantiomerically enriched (R)-28 (64% ee) (vide
infra) was accomplished by reaction of lithiated diphenyl-o-
(34) Herbrandson, H. F.; Dickerson, R. T., Jr. J. Am. Chem. Soc. 1959,
81, 4102.
(35) Pirkle, W. H.; Beare, S. D.; Muntz, R. L. Tetrahedron Lett. 1974,
2295. Pirkle, W. H.; Beare, S. D. J. Am. Chem. Soc. 1968, 90, 6250.
(36) Miller, E. G.; Rayner, D. R.; Thomas, H. T.; Mislow, K. J. Am.
Chem. Soc. 1968, 90, 4861.
(32) The fact that the origin of the second oxygen in 19 is probably
from adventitious oxygen dissolved in the solvent is supported by the
reaction of 18a-18O with 78% 18O to give 19-18O which has 79% 18O. If
both oxygens of 19 had come from the acetate, a detectable amount of
19-18O2 would have been expected.
(33) Vedejs, E.; Mastalerz, H.; Meier, G. P.; Powell, D. W. J. Org. Chem.
1981, 46, 5253.
(37) A control experiment showed that labeled 29 was recovered with
intact label after heating in acetic acid at 40 °C for 1.5 h. Reactant 28
recovered from the attempted double labeling experiment also appeared to
show a retention of 18O, suggesting exchange had not occurred prior to
reaction. However, FAB/MS evaluation of the isotopic composition of 28
had errors due to some apparent exchange in the analysis which prevents
an unambiguous conclusion.

3432 J. Am. Chem. Soc., Vol. 118, No. 14, 1996
Kurtzweil and Beak
Table 4. Reactants and Transition Structure For Oxygen Transfers
to Phosphorus(III)
Scheme 4
Table 3. Isotopic Distributions of 28 and 29 for the Double
Labeling Experiments with 28 in DMF at 77 °Ca
reactant theoreticalb
theoreticalb experiment
conditions
1 equiv of
label
12C16O 52
28 intramolecular intermolecular
29
perhaps to continue a chain, gives 29. This mechanism is
analogous to previous proposals for similar reactions such as
generalized eq 5.11,33,38
52
17
0
30
51
18
0
36
32
16
14
35
34
16
15
53
17
0
30
56
16
0
added CCl4 13C16O 17
12C18O
0
13C18O 30
12C16O 48
13C16O 18
1 equiv of
added I2
12C18O
7
13C18O 26
31
28
a Experimental error is estimated to be (5% for isotopic distributions
in 29 determined by FAB/MS. b For the I2 case the theoretical intra-
and intermolecular distributions were calculated on the basis of weighed
amounts of unlabeled and doubly-labeled starting materials because
FAB/MS of 28 did not give consistent results. The comparison of the
columns for this reaction shows the deviation.
result is within experimental error of the 23% 18O label expected
on the basis of the statistical availability of 18O label in the
labeled acid. Analysis of recovered 28 showed no 18O
incorporation.37 On the basis of these results, we suggest the
formation of 33 which adds acetic acid to give 34 which leads
to 29 by loss of acetic acid. This is consistent with the
generalized case of eq 7.
Summary. The mechanisms of oxygen transfer from a
nitrone nitrogen, an O-acetylhydroxylamine nitrogen, and a
sulfoxide sulfur to a phosphine phosphorus(III) have been
investigated using the endocyclic restriction test, isotopic
labeling, reaction kinetics, and NMR spectroscopy. The
structures of the reactive intermediates which are consistent with
the data are shown in Table 4. The transfer of oxygen from
the nitrone to phosphorus involves valence expansion at
phosphorus via intermediate 15. This pathway is similar to
oxygen transfer from the corresponding hydroxylamine and
corresponds to eq 5.2 A different pathway is followed for
oxygen transfer from an O-acetylhydroxylamine, in which
displacement at nitrogen occurs to give the intermediate 26
which is a precursor to 17 or 18 by reaction with water or
acetate. The oxygen is indirectly transferred in these reactions.
This pathway may result from the decreased ability of the
oxygen of the acetate to bond to phosphorus and the increased
leaving group ability of the acetate. This reaction is an example
of eq 2. For the transfer of oxygen to phosphorus from sulfur,
mechanisms involving phosphorus bonding to sulfur and oxygen
through 33 and 36 are supported. These mechanisms represent
the mechanisms of eqs 5 and 6. Clearly there is not a single
general mechanism for transfer of oxygen from a heteroatom
to phosphorus(III) and the pathways represented by eqs 1-7
provide a useful framework for understanding these reactions.
Experimental Section
The conversion of 28 to 29 in DMF in the presence of CCl4
or I2 has been investigated by double-labeling experiments as
shown in Scheme 4 and Table 3. The isotopic distribution in
the products is the same as that in the reactants within
experimental error, consistent with intramolecular oxygen
transfer.
The conversion of 28 to 29 in DMF with CCl4 or I2 is
consistent with initial halogenation at phosphorus to give 35
which undergoes ring closure to 36. Loss of a positive halogen,
General Methods. The 31P NMR spectra were recorded on a GN-
300NB spectrometer (121 MHz) in CDCl3 using 85% H3PO4 as an
external standard unless otherwise specified. Mass spectra were
performed on Finnigan-MAT 311A and 731 or a Varian MAT CH-5
spectrometer with ionization energies of 10 or 70 eV by the University
of Illinois Mass Spectrometry Laboratory. Data are reported in the
(38) Preux, M.; Lerous, Y.; Savignac, P. Synthesis 1974, 506. Castrillon,
J. P. A.; Szmant, H. H. J. Org. Chem. 1965, 30, 1338.

Endocyclic Restriction Test
J. Am. Chem. Soc., Vol. 118, No. 14, 1996 3433
form m/z (area). Elemental analyses were performed by the University
of Illinois Microanalytical Laboratory. Flash and medium-pressure
(MP) liquid chromatographies were performed using Merck 50-200
mm and 32-63 mm silica gel, respectively. Preparative thin layer
chromatography (Prep TLC) was performed on Kieselgel 60 F254 plates
from Merck containing indicator with visualization by UV light.
Analytical thin layer chromatography was performed on Merck silica
plates with F-254 indicator. Visualization was accomplished by UV
light or iodine. Analytical gas chromatography was performed on a
HP 5790S or a HP 5890A chromatograph using a Hewlett-Packard
fused silica capillary column cross-linked with 5% phenyl methyl
silicone (column i.d. 0.20 mm; column length 25 m) equipped with a
flame ionization detector. The injector temperature was 250 °C, and
the detector temperature was 300 °C. When microanalysis data were
not available, the purity of the title compound was judged to be >90%
oil weighing 21 mg. 1H NMR of this material gave a spectrum which
was consistent with that of 3. The product from the original organic
layer was redissolved in CH2Cl2, washed with saturated NaCO3, dried
over Na2SO4, and concentrated to give a light brown oil weighing 196
mg which slowly began to crystallize. Preparative TLC (10% MeOH-
EtOAc) of the combined organic fractions gave 78 mg (44%) of 3 as
a light brown oil which eventually solidified. Trituration (EtOAc-
hexanes) gave a fine solid from which the supernatant was removed.
Concentration and trituration (10% EtOAc/hexanes) gave 3 as a waxy,
1
light yellow solid: mp 86-90 °C; H NMR δ 6.98-7.71 (m, 14 H),
3.81 (s, 3 H), 2.10 (br s, 1 H), 1.02 (s, 9 H); 13C NMR δ 126.15-
133.50, 45.58 (d, J ) 3.7 Hz); 31P NMR δ 32.8. Anal. Calcd for
C23H26NOP: C, 76.01; H, 7.21; N, 3.86; P, 8.52. Found: C, 76.35;
H, 7.42; N, 3.49; P, 8.18.
r-tert-Butyl-2-[bis(4-trifluorophenyl)phosphino]benzene-
1
by H NMR spectral determinations unless otherwise noted.
methanimine N-Oxide (5). A mixture of the benzaldehyde (251 mg,
All solvents and reagents were obtained from commercial sources
and used without further purification, except where noted. Tetrahy-
drofuran (THF) and diethyl ether (Et2O) were distilled from sodium/
benzophenone under an atmosphere of nitrogen. The ethyl acetate and
hexane solvents used for MPLC and flash chromatography were distilled
from bulk solvent over sodium carbonate and molecular sieves,
respectively. Acetic acid was purified by treatment with acetic
anhydride and distillation from CrO3 as described by Perrin and
Armarego39 followed by distillation from Ph3P under nitrogen. Di-
methylformamide (DMF) was distilled from CaH under reduced
pressure. 18O-labeled water (10, 50, and 98+% 18O) was purchased
from Cambridge Isotope Labs or Isotec, Inc. Methyl formate-13C (99%
13C) was purchased from Isotec, Inc. Methyl-13C iodide (99% 13C)
was purchased from Cambridge Isotope Labs. 18O2 (98.9% 18O) was
purchased from Isotec.
-
0.59 mmol), t-BuNHOH-HCl (81.7 mg, 0.65 mmol), and CH3CO2
-
Na+-3H2O (97 mg, 0.71 mmol) in 15 mL of 5:1 EtOH-H2O was
stirred at ambient temperature for 10 h. An additional 81 mg of the
hydroxylamine was added along with 96 mg of CH3CO2-Na+. After
11.5 h the reaction was diluted with CH2Cl2 and washed with 10%
NH4Cl (50 mL) followed by H2O (1×). The organic layer was dried
over Na2SO4 and concentrated to give a yellow semisolid. Chroma-
tography (20% EtOAc-hexanes) gave 98.3 mg of recovered aldehyde
1
and 108.5 mg (37%) of 5 as an off-white solid: mp 152-154 °C; H
NMR δ 9.23 (m, 1 H), 8.24 (d, J ) 6.4 Hz, 1 H), 7.26-7.65 (m, 10
H), 6.86 (m, 1 H), 1.40 (s, 9 H); 13C NMR δ 121.16-140.10, 71.8,
28.1; 31P NMR δ -12.1. Anal. Calcd for C25H22F6NOP: C, 60.36;
H, 4.46; N, 2.82; P, 6.23. Found: C, 60.45; H, 4.45; N, 2.66; P, 6.30.
r-tert-Butyl-2-[bis(4-methoxyphenyl)phosphino]benzene-
methanimine N-Oxide (6). To a solution of the benzaldehyde (102
mg, 0.29 mmol) in 5:1 EtOH-H2O (6 mL) was added CH3CO2-Na+-
3H2O (60.5 mg, 0.44 mmol) followed by t-BuNHOH-HCl (48.0 mg,
0.38 mmol). After stirring at room temperature (RT) for 17 h, a light
yellow precipitate had formed. The reaction mixture was filtered to
give 49 mg (40%) of 6 as a light yellow solid, mp 162.5-163.5 °C.
The filtrate was concentrated, redissolved in Et2O, and dried over Na2-
SO4. Chromatography (30% EtOAc-hexanes) gave an additional 25
All organometallic reagents used were obtained commercially and
titrated according to the procedure of Tischler,40 Shapiro,41 or Suffert42
prior to use. All reactions involving air- or water-sensitive reagents
were carried out under nitrogen or argon in oven-dried or flame-dried
glassware which was cooled under a nitrogen atmosphere.
Isotope ratios were detrermined either by FI/MS on a Finnigan MAT-
731 instrument or by FAB/MS on a VG ZAB-SE instrument.43
r-tert-Butyl-2-(diphenylphosphino)benzenemethanimine N-Oxide
(1). A mixture of 2-(diphenylphosphenyl)benzaldehyde (0.50 g, 1.7
mmol), t-BuNHOH-HCl (0.20 g, 1.6 mmol), and CH3CO2-Na+ (0.16
g, 2.0 mmol) in 83% EtOH-H2O (30 mL) was stirred at room
temperature for 24 h. The reaction was filtered to give 142 mg of
recovered aldehyde as a yellow solid, mp 115.5-118.5 °C (lit.17 mp
118-119 °C). The filtrate was diluted with CH2Cl2, and H2O was
added. The aqueous layer was back-extracted with CH2Cl2 (2×), and
the combined organic layers were dried over Na2SO4. Concentration
followed by chromatography (20% EtOAc-hexanes) gave 255 mg
(44%) of 1 as a white solid after crystallization from 10% EtOAc-
hexanes: mp 150.5-151.5 °C; 1H NMR δ 9.25 (m, 1 H), 8.18 (d, J )
6.4 Hz, 1 H), 6.80-7.38 (m, 13 H), 1.36 (s, 9 H); 13C NMR δ 128.19-
137.33, 71.83, 28.44; 31P NMR δ -10.3. Anal. Calcd for C23H24-
NOP: C, 76.43; H, 6.69; N, 3.88; P, 8.57. Found: C, 76.51; H, 6.80;
N, 3.77; P, 8.43.
N-[2-(Diphenylphosphinyl)benzyl]-N-tert-butylamine (3). A solu-
tion of 4 (151.4 mg, 0.49 mmol) and t-BuNH2 (0.10 mL, 0.95 mmol)
in EtOH (15 mL) was stirred at room temperature for 3 days. The
reaction mixture was cooled to 0 °C, and BH3-pyridine (0.25 mL, 2.5
mmol) was added via syringe. After warming slowly to room
temperature over 20 h, 3 M HCl (20 mL) was added and stirring was
continued overnight. The aqueous layer was extracted with CH2Cl2.
The resulting organic layer was dried over Na2SO4 and concentrated
to give a brown oil. The aqueous layer was treated with saturated Na2-
CO3 until basic and was then extracted with CH2Cl2 (3×). This organic
layer was dried over Na2SO4 and concentrated to give a yellow tacky
1
mg (60%) of 6 as an off-white solid: mp 158-160 °C; H NMR δ
9.23 (m, 1 H), 8.24 (d, J ) 6.4 Hz, 1 H), 6.87-7.45 (m, 11 H), 3.80
(s, 6 H), 1.38 (s, 9 H); 13C NMR δ 160.5, 128.40-135.8, 114.4, 55.2,
28.0; 31P NMR δ -13.9. Anal. Calcd for C23H28NO3P: C, 71.24; H,
6.70; N, 3.32; P, 7.35. Found: C, 71.05; H, 6.70; N, 3.21; P, 7.31.
N-[2-(Diphenylphosphinyl)benzyl]-N-isopropylamine (17) was
prepared as previously described.14 A solution of 0.35 g (1.0 mol) of
N-[2-(diphenylphosphino)benzyl]-N-isopropylhydroxylamine in 10 mL
of toluene was heated at reflux for 1 h under a nitrogen atmosphere.
The solution was cooled to ambient temperature and evaporated in
vacuo to provide a yellow oil which was purified on the Chromatotron
with 5% methanol in ethyl acetate as an eluent to give 0.32 g (90%) of
17 as a pale yellow solid: mp 93-95 °C; 1H NMR δ 6.8-7.8 (m, 14
H), 4.8 (br s, 1 H), 3.85 (s, 2 H), 2.62 (hept, 1 H, J ) 6.3 Hz), 0.9 (d,
6 H), J ) 6.3 Hz); 13C NMR δ 145.6 (d, JPC ) 8.2 Hz), 133.9, 133.5,
133.3, 132.3, 132.2, 132.1, 131.9, 131.8, 131.7, 131.5, 130.0, 128.6,
128.5, 128.4, 126.3, 126.1, 50.1 (d, JPC ) 5.3 Hz), 48.0, 22.6. Anal.
Calcd for H22H24PNO: C, 75.62; H, 6.92; N, 4.01; P, 8.87. Found:
C, 75.28; H, 7.09; N, 3.92; P, 8.70.
A solution of 4 (0.75 g, 2.45 mmol) in EtOH (25 mL) containing
isopropylamine (1.0 mL, 11.7 mmol) was stirred for 36 h at room
temperature. The reaction was cooled with an ice bath, and BH3-
pyridine (1.7 mL, 16.8 mmol) was added under a nitrogen atmosphere.
After 1 h the cooling bath was removed, and stirring was continued at
ambient temperature for another 20 h. The mixture was cooled to 0
°C, 10% hydrochloric acid was added dropwise until hydrogen evolution
ceased, and stirring was continued at ambient temperature overnight.
The mixture was basified with Na2CO3 to pH ≈ 12 and extracted with
CH2Cl2 (3 × 25 mL), and the combined organic layers were dried over
anhydrous Na2CO3. Concentration of the organic layer in vacuo gave
an orange oil, which was purified by column chromatography (50%
EtOAc-hexanes followed by 10% MeOH-EtOAc) to give a light
yellow solid. Recrystallization from EtOAc-hexanes gave 394 mg
of 17 as an off-white solid, mp 95-98 °C. A second crop gave an
(39) Perrin, D. D.; Armarego, W. L. F. Purification of Laboratory
Chemicals; Pergamon: New York, 1988; p 67.
(40) Tischler, A. N.; Tischler, M. H. Aldrichim. Acta 1987, 11, 20.
(41) Lipton, M. F.; Sorensen, C. M.; Sadlet, A. C.; Shapiro, R. H. J.
Organomet. Chem. 1980, 186, 155.
(42) Suffert, J. J. Org. Chem. 1989, 54, 509.
(43) Lambert, J. B.; Shurvell, H. F.; Lightner, D.; Crooks, R. G.
Introduction to Organic Spectroscopy; Macmillan: New York, 1987; pp
351-353.

3434 J. Am. Chem. Soc., Vol. 118, No. 14, 1996
Kurtzweil and Beak
1
additional 91.0 mg of 17. The 31P NMR (CDCl3) δ 32.3, H, NMR,
mixture was cooled to -78 °C and a solution of methyl benzenesulfinate
(1.15 g, 7.4 mmol) in Et2O (5 mL) was added. The reaction was stirred
for 15 min, and then the cooling bath was replaced with a 0 °C bath
and allowed to warm slowly to ambient temperature over 10.5 h before
H2O and Et2O were added. The resulting aqueous layer was back-
extracted with Et2O, and the combined organic layers were dried over
MgSO4. Concentration in vacuo gave 2.0 g of a peach-colored oil
which was redissolved in Et2O and dry loaded into silica gel. Elution
with a gradient of hexanes, 20% EtOAc-hexanes and 50% EtOAc-
hexanes, gave a partially purified product. Additional purification of
the Chromatotron (31% EtOAc-hexanes) gave 0.577 g (31%) of 28
as a colorless tacky oil which eventually solidified. Recrystallization
from EtOAc-hexanes gave an analytical sample: mp 114-116 °C;
1H NMR (CDCl3) δ 6.98-7.67 (m, 19 H), 4.48 (dd, J ) 13.0 Hz, 1.6,
1 H), 4.27 (d, J ) 13.0 Hz, 1 H); 13C NMR (CDCl3) δ 124.2-143.7,
63.3 (d, J ) 21.1 Hz); 31P (CDCl3) δ -15.9; FI/MS m/z 310 (M+).
Anal. Calcd for C25H21OPS: C, 74.98; H, 5.29; P, 7.73; S, 8.01.
Found: C, 74.83; H, 5.35; P, 7.64; S, 7.92.
and 13C NMR correspond to those above.
N-[2-(Diphenylphosphino)benzyl]-N-isopropyl-O-acetylhydroxy-
lamine (16). To a solution of 25 (100 mg, 0.29 mmol) in CH2Cl2 (5
mL) containing pyridine (0.04 mL, 4.94 mmol) and a catalytic amount
of DMAP (<10 mg) was added acetyl chloride (30.5 mL, 4.29 mmol).
After stirring for 20 min at room temperature, the reaction was treated
with 10% NH4Cl (5 mL). The aqueous layer was extracted once more
with CH2Cl2. The combined organic layers were dried over MgSO4
and conentrated to give 116 mg of a light yellow oil. Flash
chromatography (CH2Cl2) gave 94 mg (84%) of 16 as a light yellow
oil which slowly crystallized on standing in the freezer: mp 50.5-
52.5 °C; 1H NMR (CDCl3, 200 MHz) δ 6.83-7.75 (m, 14 H), 4.24 (d,
J ) 2.5 Hz, 2 H), 3.17 (m, J ) 6.3 Hz, 1 H), 1.76 (s, 3 H), 1.08 (d, J
) 6.3 Hz, 6 H); 13C NMR (CDCl3) δ 127.2-141.4, 57.3, 56.6 (d, JPC
) 25.3 Hz), 19.7, 18.5; 31P NMR (CDCl3) δ -15.3. Anal. Calcd for
C24H26NO2P: C, 73.64; H, 6.70; N, 3.58; P, 7.91. Found: C, 73.77;
H, 6.76; N, 3.51; P, 7.83.
To a solution of 25 (250 mg, 0.72 mmol) in CH2Cl2 (10 mL) at 0
°C was added AcOH (0.12 mL, 2.10 mmol) followed by DCC (440
mg, 2.10 mmol). After 0.5 h the cooling bath was removed, and stirring
was continued for 0.5 h at RT. The reaction was then treated with
CH2Cl2 and the combined organic layers were dried over MgSO4.
Concentration followed by chromatography (90% CH2Cl2-hexanes)
gave 229 mg (81%) of 16 as a light yellow oil which was identical by
1H NMR with 16 prepared by other routes.
2-(Diphenylphosphinyl)benzyl Phenyl Sulfide (29). To a mixture
of 31 (151.2 mg, 0.29 mmol) in MeoH (25 mL) at 0 °C was added
30% H2O2 (31 mL, 0.35 mmol). After 0.5 h the cooling bath was
removed, and a solution of the reaction was attained within 0.5 h at
room temperature. Column chromatography (30% EtOAc-hexanes)
gave 128 mg (82%) of 29 as a colorless oil. Trituration in hexanes
gave a white solid, mp 108-111 °C. The 1H NMR spectrum was
identical with that of a sample obtained by oxygen transfer from 28.
An analytical sample of 29 was prepared by recrystallization from
Conversion of 16 to 17. A solution of 16 (22.3 mg, 0.057 mmol)
in AcOH (0.76 mL) was heated at 100 °C for 1.75 h. The solvent was
removed in vacuo, and the product was dissolved in CH2Cl2. The
organic layer was washed with 10% NaHCO3 and dried over Na2SO4.
Evaporation of solvent gave 14 mg (70%) of an oil which slowly
1
EtOAc-hexanes to give a white solid: mp 114-115 °C; H NMR
(CDCl3) δ 7.10-7.72 (m, 19 H), 4.58 (s, 2 H); 13C NMR (CDCl3) δ
125.79-142.47, 36.49 (d, JPC ) 6.0 Hz); 31P NMR (CDCl3) δ 32.6.
Anal. Calcd for C25H21OPS: C, 74.98; H, 5.29; P, 7.73; S, 8.01.
Found: C, 75.01; H, 5.30; P, 7.65; S, 7.94.
1
crystallized. The H NMR spectrum of the product was identical to
that of an authentic sample of 17.
Conversion of 28 to 29 Using AcOH. A solution of 28 (28 mg,
0.07 mmol) in AcOH (0.70 mL) was heated in an NMR tube at 31 °C
for 13 h. The reaction was diluted with CH2Cl2, and the acid was
neutralized with saturated NaHCO3. The organic layer was dried over
MgSO4 and concentrated to give 24 mg (85%) of 29 as an oil which
N-[2-(Diphenylphosphinyl)benzyl]-N-isopropylacetamide (18). To
a solution of 17 (0.100 g, 0.29 mmol) in CH2Cl2 (5 mL) at 0 °C was
added AcOH (49 mL, 0.78 mmol) followed by DCC (0.18 g, 0.87
mmol). After 0.5 h the cooling bath was removed, and the reaction
was stirred for 4 h and allowed to stand for 20 h. Removal of solvent
in vacuo gave a white foam which after column chromatography
(EtOAc followed by 10% MeOH-EtOAc) gave 72.0 mg (64%) of 18
as a colorless tacky oil which eventually crystallized; mp 136.5-140.5
°C. Recrystallization from EtOAc-hexanes gave a white solid: mp
135.5-137.5 °C; 1H NMR (CDCl3, 200 MHz, observed as an
approximate 2:1 ratio of rotational isomers) δ 7.00-7.77 (m, 14 H),
4.80 (s, 0.66 H), 4.78 (m, 0.66 H), 4.69 (s, 1.33 H), 4.10 (m, 0.33 H),
2.22 (s, 1 H), 1.81 (s, 2 H), 1.05 (d, J ) 6.6 Hz, 2 H), 0.80 (d, J ) 6.6
Hz, 4 H); 13C NMR (CDCl3, 75 MHz) δ 171.6, 143.7, 126.2-133.6,
49.9, 49.5, 44.8-45.4, 19.5-22.3; 31P NMR (CDCl3, 121 MHz,
observed as a pair of singlets in an approximate 1:2 ratio) δ 32.26,
32.97; HRES/MS for C24H26NO2P, m/z calcd 391.1701, found 391.1702.
Anal. Calcd for C24H26NO2P: C, 73.64; H, 6.70; N, 3.58. Found: C,
73.12; H, 6.66; N, 3.56.
1
was identical to an authentic sample as judged by H NMR.
Acknowledgment. We are grateful to the National Science
Foundation and the National Institutes of Health-Institute of
General Medical Science for support of this work and to Dr.
Dekai Loo for the preparation of 17 and the isolations of 18,
22, and 21 from the reaction of 16.
Supporting Information Available: Text describing the
syntheses of 21, (R)-28, and labeled compounds and control
experiments and tables listing crystallographic data, bond lengths
and angles, positional and thermal parameters, and kinetic data
(49 pages). This material is contained in libraries on microfiche,
immediately follows this article in the microfilm version of the
journal, can be ordered from the ACS, and can be downloaded
from the Internet; see any current masthead page for ordering
information and Internet access instructions.
2-(Diphenylphosphino)benzyl Phenyl Sulfoxide (28). To a solu-
tion of n-BuLi in hexanes (2.4 mL, 5.4 mmol) and TMEDA (0.82 mL,
5.4 mmol) in hexanes (10 mL) was added a solution of (2-methylphen-
yl)diphenylphosphine (1.0 g, 3.62 mmol) dissolved in hexanes (10 mL).
After stirring for 1 h at ambient temperature, the orange heterogeneous
JA9527986