Observation of a betaine lithium salt adduct during the course of a wittig reaction

Article abstract of DOI:10.1002/(sici)1099-0690(199806)1998:6<1085::aid-ejoc1085>3.0.co;2-g

A stable betaine lithium salt adduct is observed during the course of a Wittig reaction. The stabilization of this adduct has been achieved by complexation with lithium ions and by using the chelating effect of pyridyl ligands. Dynamic NMR spectra are obs

Full text of DOI:10.1002/(sici)1099-0690(199806)1998:6<1085::aid-ejoc1085>3.0.co;2-g

FULL PAPER
Observation of a Betaine Lithium Salt Adduct During the Course of a Wittig
Reaction
Robert A. Neumann and Stefan Berger*
Institut für Analytische Chemie,
Linn e´ straße 3, D-04103 Leipzig
Fax: (internat.) ϩ49(0)341/9711833 or 9736115
e-mail: stberger@rz.uni-leipzig.de
Received December 30, 1997
Keywords: Wittig reaction / Dynamic NMR / Oxaphosphetanes / Metal ion complexation
A stable betaine lithium salt adduct is observed during the
course of a Wittig reaction. The stabilization of this adduct spectra
has been achieved by complexation with lithium ions and by oxaphosphetanes complex with sodium or lithium ions.
using the chelating effect of pyridyl ligands. Dynamic NMR
are observed when the corresponding
[9]
The Wittig reaction between ylides (1) and aldehydes or of 6 with butyllithium under RI conditions did not reveal
3
1
ketones (2) to form olefins (5) is one of the fundamental
P NMR signals in the expected betaine chemical shift re-
[1]
transformations in organic chemistry and is constantly gion of δ ϭ ϩ10 to ϩ30 (see Scheme 2) but instead showed
P
used in natural product synthesis as well as industrial syn- a signal consistent with an oxaphosphetane.
[
2]
thesis (see Scheme 1).
Scheme 2
Scheme 1
In a recent communication it was shown by Ustynyuk[11]
,
that thiobetaines can be formed during the reaction of yl-
ides with thioketones. Subramanyam[ reported the failure
of a Wittig reaction when triphenylmethylenephosphorane
12]
The mechanism of the reaction, however, is still being (1) and 2,2Ј-dipyridyl ketone were allowed to react after
[
3][4]
[5]
debated.
Whereas in his first publication G. Wittig generation of the ylide using butyllithium as base. Substi-
proposed oxaphosphetanes (4) as intermediates, the sub- tuting lithium by potassium, however, resulted in the nor-
[
6]
sequent literature postulated betaines (3) even after in mal olefin formation. This behaviour was explained by the
[
7]
31
1
973 Vedejs was able to demonstrate unequivocally by
P
proposal of the formation of betaine (3a) complexed by a
NMR the occurence of only oxaphosphetanes. Indeed, lithium ion; however, no spectroscopic evidence has been
since then, no direct spectroscopic proof for the possibility given. The chelating effect of the pyridyl nitrogen atoms
of the existence of betaines has been published to our should stabilize the betaine form according to the author.
knowledge. Early postulations, mainly put forward by We have reported[ earlier on the complexation of oxaphos-
Schlosser and others, lacked convincing spectroscopic evi- phetanes by lithium ions and were thus prepared to investi-
9]
[
8]
dence. With the Rapid Injection NMR (RI) technique we gate the reaction of Subramanyam in detail. Here we wish
[9]
have not been able to observe these species, even not in to report the first direct spectroscopic evidence of the for-
the case, when the synthesis of a betaine was attempted mation of a betaine lithium salt adduct during the course
from a β-hydroxy phosphonium salt (6).[ Deprotonation of a Wittig reaction.
10]
Eur. J. Org. Chem. 1998, 1085Ϫ1087

WILEY-VCH Verlag GmbH, D-69451 Weinheim, 1998
1434Ϫ193X/98/0606Ϫ1085 $ 17.50ϩ.50/0
1085

R. A. Neumann, S. Berger
FULL PAPER
Scheme 3
Figure 1. Temperature-dependent 31P NMR spectra of 4a in the
presence of NaI in THF
Using sodium hexamethyldisilazide as base[13] to gener-
ate the ylide 1, followed by treatment with 2,2Ј-dipyridyl
ketone results in the formation of the corresponding oxa-
3
1
phosphetane 4a. The P NMR spectrum reveals only a sin-
glet at δP ϭ Ϫ63.2. On addition of lithium bromide, the
signal of 4a disappears and a weak signal at δ ϭ ϩ23.7
P
emerges from the spectrum. The recording of this signal,
however, is hindered by precipitation of solid material. We
ascribe the resonance signal to betaine 3a, because by ad-
dition of 12-crown-4 to scavenge the lithium ion one can
observe the reformation of oxaphosphetane 4a. The high ions to 4a leads immediately to the precipitation of betaine
insolubility of 3a prevented so far the growing of crystals 3a, as described above. In addition to the two main signals
suitable for an X-ray determination and the recording of which coalesce at about 223 K below 210 K a third weak
1
3
C NMR spectra; however we could purify the material to signal is observed which is probably due to double com-
3
1
[14]
the extent that a solid state CP/MAS P NMR spectrum
plexation. Neglecting this signal, the dynamic NMR spectra
could be recorded under high power proton decoupling at were analyzed by line shape analysis and the thermo-
a spinning speed of 4000 Hz; it reveals only one signal at dynamic parameters have been determined. We find an acti-
϶
δP ϭ ϩ16.8. If one tries to redissolve this material in vation energy ∆G (298) of 43 kJ/mol with an activation
϶
϶
DMSO or acetonitrile, the Wittig reaction sets in as shown enthalpy ∆H of 25 kJ/mol and an activation entropy ∆S
by an olefinic signal in the proton NMR spectrum. of Ϫ60 J/mol. The large negative entropy value clearly indi-
Replacing two phenyl groups in 1 by the 2,2Ј-biphenyl cates an intermolecular exchange process which corrobo-
ligand should stabilize the oxaphosphetane making ring rates our interpretation. On cooling of a solution of 4b in
opening and olefin formation less likely.[ Indeed, we were the presence of lithium ions (Figure 2) quite similar spectra
able to prepare this rather stable oxaphosphetane 4b and to those shown in Figure 1 are registered. At very low tem-
we find for this oxaphosphetane dynamic NMR spectra as peratures the spectra are even more complex and give at
15]
[
9]
reported earlier for other oxaphosphetanes. Both lithium least four signals which are interchanging with each other.
and sodium ions are complexed, and free and complexed Subsequent addition of sodium or lithium ions changes the
form are in dynamic equilibrium. In this case, ring opening relative intensity of these signals.
to a betaine is probably prevented due to the chelating bite
Though the stereochemical consequences of carrying out
angle of the biphenyl ligand preferring fivefold coordi- the Wittig reaction under Љlithium salt freeЉ and Љlithium
salt containingЉ conditions are very well known[ , they are
13]
nation at the phosphorus atom.
A typical dynamic NMR spectrum in the presence of so- not really understood at present. Our dynamic NMR spec-
dium ions is shown for 4a in Figure 1, where we ascribe the tra show physical evidence of the interaction of metal salts
two signals at the low temperature limit to the complexed with oxaphosphetanes. Although this happens in a tempera-
and free form of the oxaphosphetane. Addition of lithium ture regime where the oxaphosphetanes are stable and the
1086
Eur. J. Org. Chem. 1998, 1085Ϫ1087

Wittig Reaction
FULL PAPER
Figure 2. Temperature-dependent 31P NMR spectra of 4b in the
dipyridyl ketone dissolved in THF was added to the NMR tube at
Ϫ60°C and directly measured.
presence of LiBr in THF
For the preparation of the oxaphosphetane 4b 1-phenyldibenzo-
phosphol was prepared according to reference[ . 4.54 g (17 m)
of this material were methylated in 30 ml of dry toluene using 5.3
ml (5 equiv.) methyl iodide for 16 h at 120°C. The precipitated
white phosphonium salt was washed with diethyl ether (5.06 g,
16]
7
4%). 0.506 (1.26 m) of 1,1Ј-biphenylmethylphenylphosphonium
iodide were suspended in 10 ml dry THF. To this suspension 1
equiv. NaHMDS in 2 ml THF (Aldrich) were added at Ϫ10°C
and stirred for 30 min. The THF was evaporated and the residue
suspended in dry petroleum ether. Undissolved solid material was
filtered off and the petroleum ether was evaporated. The residue
was redissolved in THF, cooled to Ϫ60°C and transferred to an
NMR tube under exclusion of moisture and air. Finally 1 equiv. of
2
,2Ј-dipyridyl ketone dissolved in THF was added to the NMR
tube at Ϫ60°C and directly measured.
[
[
1]
2]
E. Vedejs, M. J. Peterson, Top. Stereochem. 1994, 21, 1Ϫ157.
K. C. Nicolaou, M. W. Härter, J. L. Gunzner, A. Nadin, Liebigs
Ann. 1997, 1283Ϫ1301.
[
[
3]
4]
H. J. Bestmann, O. Vostrowsky, Top. Curr. Chem. 1983, 109,
85Ϫ157; B. Maryanoff, A. Reitz, Chem. Rev. 1989, 89,
863Ϫ927.
H. Yamataka, K. Nagareda, T. Takatsuka, K. Ando, T. Hana-
fusa, S. Nagase, J. Am. Chem. Soc. 1993, 115, 8570Ϫ8576; H.
Yamataka, T. Takatsuka, T. Hanafusa, J. Org. Chem. 1996, 61,
7
22Ϫ726; M. Nishizawa, Y. Komatsu, D. M. Garc ´ı a, Y. Nogu-
chi, H. Imagawa, H. Yamada, Tetrahedron Lett. 1997, 38,
215Ϫ1218.
1
[
[
5]
6]
G. Wittig, G. Geissler, Liebigs Ann. Chem. 1953, 580, 44Ϫ57.
M. Schlosser, H. B. Tuong, Angew. Chem. 1979, 91, 675; Angew.
Chem. Int. Ed. Engl. 1979, 18, 633.
[
[
7]
8]
second part of the Wittig reaction has not yet set in, it could
well be, that this complexation is responsible for the stereo-
chemical outcome.
E. Vedejs, K. A. J. Snoble, J. Am. Chem. Soc. 1973, 95,
5
778Ϫ5780.
M. Schlosser, H. B. Tuong, C. Tarchini, Chimia 1977, 31,
2
19Ϫ220; M. Schlosser, K. F. Christmann, Liebigs Ann. Chem.
1
967, 708, 1Ϫ35; M. Schlosser, A. Piskala, C. Tarchini, H. B.
This work was supported by the Deutsche Forschungsgemein-
schaft (SFB-260-A6) and by the Fonds der Chemischen Industrie.
Tuong, Chimia 1975, 29, 341Ϫ342.
[9]
C. Geletneky, F.-H. Försterling, W. Bock, S. Berger, Chem. Ber.
1
993, 126, 2397Ϫ2401.
[10]
C. Geletneky, Ph. D. Thesis, Marburg, 1994.
Experimental Section
[11]
I. V. Borisova, N. N. Zemlyanskii, A. K. Shestakova, Y. A. Us-
tynyuk, Mendeleev Commun. 1996, 90Ϫ92; Chem. Abstr. 1996,
. Dynamic NMR Measurements: Low temperature 31P NMR:
1
1
25, 58619z.
Bruker AM-400 NMR spectrometer in THF using a 10-m multi-
nuclear probe head under standard measurement condition and
proton decoupling.
[12]
[13]
C. Subramanyam, Tetrahedron Lett. 1995, 36, 9249Ϫ9252.
M. Schlosser, B. Schaub, J. Oleveira-Neto, H. Jeganathan, Chi-
mia 1986, 40, 244Ϫ245.
[
[
14]
15]
We are indebted to Prof. H. Günther, Siegen, for letting us use
his MSL-300 solid state NMR spectrometer.
2. Synthesis: For the preparation of the oxaphosphetane 4a 0.81
g (2.0 m) of triphenylmethylphosphonium iodide were suspended
in 10 ml dry THF. To this suspension 1 equiv. NaHMDS in 2 ml
THF (Aldrich) were added at Ϫ15°C and stirred for 30 min. The
reaction mixture was cooled to Ϫ60°C and transferred to an NMR
tube under exclusion of moisture and air. Finally 1 equiv. of 2,2Ј-
E. Vedejs, C. Marth, R. Ruggeri, J. Am. Chem. Soc. 1988, 110,
3
940Ϫ3848; E. Vedejs, C. Marth, Tetrahedron Lett. 1987, 28,
445Ϫ3448.
3
[16]
J. Cornforth, R. H. Cornforth, R. T. Gray, J. Chem. Soc. Perkin
Trans. 1 1982, 2289Ϫ2297.
[97408]
Eur. J. Org. Chem. 1998, 1085Ϫ1087
1087