G. Kegle6ich et al. / Journal of Organometallic Chemistry 570 (1998) 49–53
51
Table 2
Phosphinic esters (4a–f) prepared by the photolysis of phosphabicyclooctene 2 in alcohols
+b
found
Product
Yield (%)
lpa
(llpit
)
M+found
(M+calc
)
[M+H]
([M+H]+calc
)
4ac
4b
4c
4d
4e
4f
90
85
84
66
86
77
44.8
42.7
42.8
41.3
42.8
50.1
(45.09)
(42.02)
170.0523
184.0678
(170.0497)
(184.0653)
199.0859
199.0862
213.1019
(199.0888)
(199.0888)
(213.1044)
156.0326
(156.0340)
a CDCl3 solution.
b FAB measurements.
c bp 80–85°C/0.4 mmHg (lit8: 106–110°C/0.6 mmHg).
the absence of an alcohol. The result of this photolysis
was polimeric precipitate, presumably
further the mechanism of the photolysis of bridged
P-heterocycles and to reexamine the mechanism of the
fragmentation of phosphabicyclooctadienes by means
of kinetic methods.
a
[PhP(O)(CH2)]n. We observed, however, that the frag-
mentation was faster in the presence of an alcohol;
moreover, the rate was dependent on the molar excess
of the protic species. A summary of the experiments
with methanol is listed in Table 3. The nature of the
alcohol also had an impact on the reaction time. The
fragmentation in the presence of ethanol, n-propanol or
n-butanol was slower than in methanol (Table 4). This
order of reactivity corresponds to the pKa values of the
alcohols [12]. No phosphorylated product was formed
when tert-butanol was the protic species. The above
observations suggest the involvement of the alcohol in
the rate-determining step of the fragmentation. Accord-
ing to this, the protic species is added on the phospho-
ryl group of the phosphabicyclooctene (either 2 or 5) to
form intermediate 7 with a pentacoordinated phospho-
rus atom. Adduct 7 is then fragmented in a fast step to
afford phosphinate 4 (mechanism B in Scheme 3).
It can be concluded that the photochemical fragmen-
tation of phosphabicyclooctene 2 takes place according
to concurrent E–A and A–E mechanisms. Proportions
of the two pathways seem to be comparable. It can be
imagined that mechanism B may also play a role in the
fragmentation of 2-phosphabicyclo[2.2.2]octa-5,7-dienes
having a bigger ring strain (hC–P–C is 99.5° [13]) than
cycloadduct 2. The above results encourage us to study
3. Experimental
1
The 31P-, 13C- and H-NMR spectra were taken on a
Bruker DRX-500 spectrometer operating at 202.4,
125.7 and 500 MHz, respectively. Chemical shifts are
downfield relative to 85% H3PO4 or TMS. Mass spectra
were obtained on a MS-902 spectrometer or on a
ZAB-2SEQ instrument at 70 eV. Photolyses were con-
ducted in a photochemical reactor equipped by a
quartz, water-cooled immersion well with a high-pres-
sure mercury lamp (125 W).
3.1. Preparation of the isomers (A and B) of
dihydrophosphinine-N-phenylmaleimide cycloadduct 2
A solution of 2.1 g (8.81 mmol) of dihydrophos-
phinine oxide 1 consisting of 75% of the A isomer and
25% of the B isomer [14] and 1.8 g (10.41 mol) of
N-phenylmaleimide in 40 ml of toluene was stirred at
boiling point for 6 days. Solvent was evaporated and
the residue so obtained purified by column chromatog-
raphy (silica gel, 3% methanol in chloroform) to give
2.7 g (75%) of 2 as the mixture of isomer A (61%) and
isomer B (39%). Fractional crystallization from ethyl
acetate-n-pentane led to 0.45 g (21%) of pure 2A; mp.
211–214°C. Found: C, 64.37%, H, 4.75%.
C22H19ClNO3P requires C, 64.16%, H, 4.62%.
2A: 31P-NMR (CDCl3) l 37.3; 13C-NMR, Table 1;
1H-NMR (CDCl3) l 1.80 (s, 3H, Me), 2.03–2.58 (m),
3.38–4.13 (m), total int. 5H, skeletal protons, 6.05 (dd,
3JPH=3JHH=7.9, 1H, CHꢀ), 7.14–7.69 (m, 10H, Ar);
MS, m/z (rel. int.) 411 (M+, 28), 376 (M-35, 63), 91
(100).
2B: 31P-NMR (CDCl3) l 37.1; 13C-NMR, Table 1;
1H-NMR (CDCl3) l 1.49 (s, 3H, Me); MS, m/z 411
(M+).
Scheme 2.