Table 4 Leaching of tin under different conditions of use of resin P1
Tin content of
the product
(ppm)
Leaching of
initial tin
content (%)
Temperature
(T/ЊC)
Stirring
(yes/no)
Entry
1
2
3
4
65
65
95
95
N
Y
N
Y
220
270
410
800
0.3
0.4
1.2
0.6
may release oligomers into the solution; when used, the stirring
was performed with a suspended magnetic bar at low speed (50
rpm) as described above. The results are reported in Table 4. It
appears from these data that the main parameter influencing
the leaching of tin is the temperature of use of the resin. Some-
what unexpectedly the use of an ether–maleimide copolymer
instead of the usual styrene–DVB system does not seem to
increase the temperature range of use of these tin-containing
resins. At 95 ЊC, a significant thermal degradation occurs lead-
ing to the release of tin-bearing oligomers or fragments into
solution. The effect of the stirring is much less dramatic (Table
Fig. 2 Catalytic activity of the polymer-supported organotin hydrides.
Reduction of 1-bromoadamantane (1 equiv.) with sodium borohydride
(
5 equiv.) at 95 ЊC in 1,2-dimethoxyethane using 0.1 equiv. of polymer-
supported Sn-Cl.
reactors is the abrasion of the polymer caused by constant
12
magnetic stirring. Seebach described recently an apparatus
13
using the “tea bag” approach. This system looks efficient but
requires a rather large amount of solvent.
4
, entries 1 and 2) probably due to the gentleness of the method
We considered that the main problem was associated with the
we adopted.
“
millstone effect” of the stirring bar in contact with the bottom
4
. Recycling tests. It is very important, when developing a
of the reactor. We thus designed a magnetic bar suspended by a
flexible nylon wire thread above the container base. The wire,
being attached to a swivel, turns torsionlessly at the same speed
as the bar. This latter, spinning freely in the solvent without
contact with the walls of the reactor, produces a gentle move-
ment of the beads allowing the heterogeneous reaction to occur
without degradation of the beads even in a rather low volume
of solvent.
polymer-supported catalyst, to investigate its stability during
successive runs in a batch reactor. This study was performed
with resin P1 using the reduction reaction described previously.
Two kinds of experiments were performed: the first one
14
(
method A) with washing of the beads successively with water,
ethanol and then diethyl ether between two runs; the other
method B) without washing. Each run was conducted for 2 h
(
after which the polymer was filtered off, then used again with
a fresh solution of 1-bromoadamantane, sodium borohydride
and AIBN. The results obtained are reported in Fig. 3. From
these, it can be concluded that the evolution of the activity on
recycling can be divided in three parts; (i) an increase of activ-
ity of the resin during the first 3 runs, then (ii) a series of runs
with maximum activity (close from 100%) followed by (iii) a
slow decrease in activity from run 7. The washing of the beads
between two runs clearly reduces the deactivation of the cat-
alyst, probably by removing the water-soluble borate derivatives
obstructing the pores of the support.
An estimation of the evolution of the loss of tin from the
support during successive runs, using method A, was made
using the procedure described above. The results are reported
in Table 5. A slow decrease of leaching can be observed with
successive runs, with a levelling at ~0.5% per run (the reactions
were performed at 95 ЊC because at 65 ЊC the rate of reduction
was very low). During the 8 runs performed the average conver-
sion in 2 h was 89%; thus 1.85 mmol of 1-bromoadamantane
2
. Catalytic reduction of 1-bromoadamantane. The different
polymer-supported organotin hydrides synthesised were evalu-
ated as catalysts in a reaction often proposed in the literature
14
as a test of reactivity, i.e. the reduction of 1-bromoadaman-
tane by sodium borohydride. The molar amount of tin chloride
was fixed at 10% with respect to the 1-bromoadamantane, the
sodium borohydride being in a 5-fold excess. The conversions to
adamantane reached in 2 h at 95 ЊC in 1,2-dimethoxyethane for
the different resins tested are reported in Fig. 2.
The results obtained are the average of at least two duplicate
experiments. The data are quite difficult to explain in term of a
structure–activity relationship but it seems that the nature of
the porogen used plays an important role: The resins P1 and P2,
prepared with a 1:1 volumic ratio of a mixture of toluene and
N-methylformanilide, display good activity (almost 80%),
whereas, resins P3, P4 and P6, prepared with a ratio 4:1 (there-
fore more rich in toluene), are somewhat less active. Resin P7,
prepared using N-methylformanilide as sole porogen, is, as well,
less efficient. The specific surface area, as estimated by the BET
method, seems not to be a good indication of activity in these
species. There was almost no reaction in the absence of an
organotin catalyst.
was reduced by NaBH in adamantane using 1.5% (molar) of
4
polymer-supported organotin chloride. The total leaching of
tin during these experiments (48 h at 95 ЊC) was estimated to be
~20% (molar) of the initial loading.
3
. Estimation of the leaching. One of the major claims for
using polymer-supported organotin species is the possible
Experimental
Fourier transform IR (FTIR) spectra were recorded on a
Perkin-Elmer Paragon 100 spectrometer; H NMR spectra were
recorded on a Bruker AC 250 spectrometer with CDCl solu-
reduction in contamination of the final product by organotin
15
species. Therefore, it seemed important to us to try to estimate
the level of tin leaching from our resins under different con-
ditions of use. Hence, a sample of resin P1 (amount necessary
to reach 0.1 equiv. of Sn) was suspended in 1,2-dimethoxy-
1
3
119
tions and Me Si as internal standard. Sn NMR spectra were
4
3
ethane (15 cm ) containing adamantane (150 mg, 1.1 mmol)
recorded on a BRUKER AC 250 spectrometer operating at
89.15 MHz. Elemental microanalytical data were provided by
the Service central d’analyse du CNRS, Vernaison, France, with
an error of 0.3% for C, H, N and 0.5% for Cl, Br and Sn.
Nitrogen adsorption isotherms were obtained using a Micro-
meritics Acusorb 2100E apparatus, University of Strathclyde,
Glasgow, UK and the resultant data were subjected to a Brun-
auer, Emmet and Teller (BET) treatment for the determination
of the specific surface area.
for 6 h. After cooling, the polymer was filtered off, then washed
with diethyl ether. The collected filtrate was concentrated to
dryness then the tin content was analysed by Atomic Absorp-
16
tion Spectroscopy (AAS).
The influence of 2 parameters was investigated: (i) the tem-
perature of the reaction; an increase of temperature may
increase the decomposition of the resin; tests were performed at
6
5 and 95 ЊC; (ii) the stirring which can break the beads and
1
40 J. Chem. Soc., Perkin Trans. 1, 1999, 137–142