Solvent and salt effects on the reactions of allyltin (IV) compounds with singlet oxygen, 4-phenyl-1,2,4-triazoline-3,5-dione and diethyl azodicarboxylate

Article abstract of DOI:10.1016/S0022-328X(00)00362-4

The effect of the increase of polarity of the solvent binary mixture methanol-benzene on the selectivity and the rate of metalloene reactions of different allyltin compounds with singlet oxygen, 4-phenyl-1,2,4-triazoline-3,5-dione (PTAD) and diethyl azodicarboxylate (DEAD) was studied. Analogous comparative studies were carried out in Et₂O and 4 mol dm^-3 solutions of LiClO₄ in diethyl ether. In the reaction with PTAD and ¹O₂ the more polar solvent favoured the production of the M-ene product, whereas the cycloaddition reaction was favoured by a non-polar solvent.

Full text of DOI:10.1016/S0022-328X(00)00362-4

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 608 (2000) 49–53
Solvent and salt effects on the reactions of allyltin (IV)
compounds with singlet oxygen, 4-phenyl-1,2,4-triazoline-3,5-dione
and diethyl azodicarboxylate
Wojciech J. Kinart a,*, Cezary M. Kinart b, Izabela Tylak a
a Department of Organic Chemistry, Uni6ersity of L*o´dz´, 90-136 L*o´dz´, Narutowicza 68, Poland
b Department of Chemical Education, Uni6ersity of L*o´dz´, 90-236 L*o´dz´, Pomorska 163, Poland
Received 7 January 2000; received in revised form 29 May 2000
Abstract
The effect of the increase of polarity of the solvent binary mixture methanol–benzene on the selectivity and the rate of
metalloene reactions of different allyltin compounds with singlet oxygen, 4-phenyl-1,2,4-triazoline-3,5-dione (PTAD) and diethyl
azodicarboxylate (DEAD) was studied. Analogous comparative studies were carried out in Et2O and 4 mol dm−3 solutions of
LiClO4 in diethyl ether. In the reaction with PTAD and 1O2 the more polar solvent favoured the production of the M-ene product,
whereas the cycloaddition reaction was favoured by a non-polar solvent. © 2000 Elsevier Science S.A. All rights reserved.
Keywords: Metalloene reaction; Solvent effect; Singlet oxygen; Azo enophile; Lithium perchlorate
It seems likely that the mechanisms of the M-ene and
cycloaddition reactions are related to that of the H-ene
reactions of allylic hydrocarbons [1]. A reasonable
model is that they involve the prior formation of a
complex between the ene and the enophile (XꢀX) (1),
which may be represented as an oxirane oxide or an
aziridine imine when XꢀXꢀOꢀO or RNꢀNR respec-
tively. Cyclic transfer of the metal or of hydrogen may
then lead to the M-ene or H-ene products respectively,
and cyclization with migration of the metal results from
b-interaction of the metal with the carbonium ion cen-
tre [2]. Different authors have shown that the stereose-
lectivity of the ene reaction of singlet oxygen with
allylic alcohols and amines depends on the polarity of
the solvent [3–6]. Davies [7,8] studied the effect of the
solvent CH2Cl2 and MeOH–C6H6 on the reaction of
allyltin compounds with 4-phenyl-1,2,4-triazoline-3,5-
dione (PTAD) and 1O2, and showed that the more
polar solvent favoured the M-ene reaction. Recently we
have observed a considerable influence of a solvent on
the chemoselectivity and the rate of reactions of allyl
1. Introduction
Allyltin compounds react with singlet oxygen or azo
enophiles to give the M-ene, H-ene and cycloaddition
products as shown in Eq. (1) The relative yields depend
on the structural and environmental conditions [1,2].
(1)
1
stannanes with PTAD and O2 in CHCl3–CH3CN [9],
and have also shown that these reactions are strongly
catalysed by lithium perchlorate [10].
* Corresponding author.
0022-328X/00/$ - see front matter © 2000 Elsevier Science S.A. All rights reserved.
PII: S0022-328X(00)00362-4

50
W.J. Kinart et al. / Journal of Organometallic Chemistry 608 (2000) 49–53
We report here a study of the rates and chemoselec-
2. Results
tivities of the reactions of some allylstannanes with
PTAD, (more briefly) with diethyl azodicarboxylate
(DEAD), and with O2 in MeOH, C6H6 and C6H6–
2.1. PTAD and DEAD: sol6ent effects
1
MeOH solvents, and in Et2O in the absence and pres-
ence of LiClO4.
The reaction of allyl stannanes with equimolar
amounts of PTAD (0.314 mmol) or DEAD (0.159
mmol) in MeOH, C6H6 and C6H6–MeOH mixtures,
was followed visually by the fading of the colour of the
azo compound (times of reactions for PTAD are shown
Table 1
Reactions of allyl–Sn compounds with PTAD
in Table 1). For more dilute solutions, the rates were
followed by UV–vis spectroscopy (down to 0.0046 mol
dm−3 for PTAD and 0.0254 mol dm−3 for DEAD) by
measuring half-lives of reactions (times corresponding
to the decrease of the initial absorbance by 50%).
Products were identified by NMR spectroscopy. The
results with PTAD are shown in Table 1.
Table 2
Reactions of allyl–Sn compounds with DEAD
In MeOH, the reactions were too fast to be moni-
tored spectroscopically, and were followed visually, but
the sequence of rates lay in the order 1\2\3\4.
This corresponds with the order of values of l(13C) for
the allylic methylene groups, which provides a measure
of the relative Lewis-acid character of the tin. Com-
pound 2 provides an exception to this rule, where, as
shown before [8], the presence of more than one allyl
group confers a considerably enhanced reactivity.
Table 3
Reactions of allyltin derivatives with PTAD in Et2O

W.J. Kinart et al. / Journal of Organometallic Chemistry 608 (2000) 49–53
51
Table 4
corresponding to the decrease of the initial absorbance
by 1% (t1/100) were measured for them in benzene. All four
compounds gave only the M-ene reaction.
Reactions of allyltin derivatives with DEAD in Et2O (0.0254 mol
dm−3
)
Table 5
Reactions of allyltin derivatives with singlet oxygen in MeOH and
C6H6
Table 6
Reactions of allyltin derivatives with singlet oxygen in 4 mol dm−3
solutions of LiClO4
The chemoselectivity between the M-ene, H-ene, and
cycloaddition paths is shown in Table 1. The H-ene
reaction (15%) was observed only with allyltriphenyltin
(3) in benzene. Compound 1 showed only the M-ene
reaction whatever the solvent composition. Compounds
2, 3 and 4 showed both the M-ene and cycloaddition
routes in pure C6H6, but in pure MeOH, all four
compounds gave only the M-ene reaction.
2.2. PTAD and DEAD: salt effects
The salt effect on the rates (0.0046 mol dm−3) and
chemoselectivities of the reactions were studied with
LiClO4 added to diethyl ether, the reactions being
monitored by absorbance at 536 nm. The results with
PTAD as the enophile are given in Table 3.
The results with DEAD were parallel to those with
PTAD, again the DEAD reacting more slowly (see Table
4).
The results with DEAD were broadly parallel to those
with PTAD, the DEAD reacting more slowly (see Table
2).
The reactions of allyltriphenyltin (3) and allyltricyclo-
hexyltin (4) in benzene were very slow. Therefore, times
Again the reactions of allyltriphenyltin (3) and allyltri-
cyclohexyltin (4) in Et2O were very slow and times

52
W.J. Kinart et al. / Journal of Organometallic Chemistry 608 (2000) 49–53
corresponding to the decrease of the initial absorbance
by 1% (t1/100) were measured for them. All four com-
pounds gave only the M-ene reaction.
If the formation of the polar ene–enophile complex
as shown in (1) is rate determining, the effect of an
increasingly polar medium can be seen to result in
stabilization of the transition state to this intermediate,
in which the charges are developing. The chemoselectiv-
ity will then depend on the partition of this intermedi-
ate between the three possible paths of intermolecular
SN2 reaction at the metal (the M-ene reaction), or at
hydrogen (the H-ene reaction), or at carbon (the cy-
cloaddition reaction). Presumably a polar medium has
the effect of reducing the activation energy of the first
process more than that of the latter pair, and the M-ene
reaction becomes dominant. From an experimental
point of view, this work shows that a polar solvent such
as methanol, and/or a salt such as lithium perchlorate
should be added, if an ene–enophile reaction is to be
selective for the metalloene process.
The effect of 4 mol dm−3 LiClO4 on the nature of
the products was determined by NMR spectroscopy
and is shown in Table 3. With each compound, the
effect of adding LiClO4 is parallel to that of adding
MeOH to C6H6 solvent: an increase in the rate of
reaction and a chemoselectivity favouring the M-ene at
the expense of the H-ene and/or cycloaddition reac-
tions.
1
2.3. O2: sol6ent effects
The reactions involving singlet oxygen in MeOH,
C6H6 and C6H6–MeOH mixtures were carried out un-
der standard conditions for 3 h and monitored by
NMR spectroscopy. In C6H6 the overall yield of prod-
ucts was low but it increased as MeOH was added:
yields in pure C6H6 and in pure methanol respectively
were, for allyldibutyltin chloride, 20 and 45% and for
diallyldibutyltin, 0 and 20% (see Table 5). Allyl-
triphenyltin and allyltricyclohexyltin were ureactive un-
der the experimental conditions, whereas illuminations
carried out for more than 8 h led to their decomposi-
tion.
4. Experimental
NMR spectra were recorded using a Varian Gemini
200 BP spectrometer. UV spectra were recorded on a
Specord spectrometer (Carl Zeiss Jena) using 10 mm
cells. Allyltin compounds were prepared by a modified
Grignard method [8] in which a solution of trialkyltin
chloride and allyl bromide (5 mol equivalent) in ether
was added to magnesium (equivalent to the allyl bro-
mide) with vigorous stirring under argon at such a rate
as to maintain gentle refluxing.
1
2.4. O2: salt effects
Details of the NMR spectra of the products from
reactions of singlet oxygen and PTAD were described
before [8].
The effect of added salt was studied with all four
allyltin derivatives in ether under standard conditions.
The results are given in Table 6.
In the case of the reactions of allyl stannanes (1–4)
with PTAD the soluble product was chromatographed
using light petroleum (b.p. 30–40°C)-diethyl ether (1/1
v/v) as eluent, and the products were purified by recrys-
tallization from benzene–hexane, or by further chro-
matography. The N-metalotriazolidines underwent
hydrolysis on the column to give the corresponding
protic compound.
Products from the H-ene and cycloaddition reactions
of singlet oxygen decomposed on silica gel, and were
identified in the reaction mixture by high resolution
NMR.
The allylperoxytin compounds produced by M-ene
reactions were converted to allyl hydroperoxide during
chromatography on silica gel; this was reduced by
triphenylphosphine to allyl alcohol.Isolation of the
products from the reactions with DEAD by gradient
chromatography gave diethyl N-allylhydrazodicarboxy-
late as an oil [10a].
In pure ether, the yields of compounds 2, 3, and 4
after 3 h was less than 5%. The yield for allyldibutyltin
chloride was equal to 30%. Illuminations carried out for
a longer time led to decomposition.
The addition of 4 mol dm−3 LiClO4 gave a consider-
able increase in the yield of the reaction; no cycloaddi-
tion products were detected, and the ratio of the M-ene
and H-ene reactions varied from ca. 1:1 for allyltricy-
clohexyltin to 1:0 for allyltriphenyltin.
As observed with the azo enophiles, an increase in
polarity of the medium resulted in an increase in the
overall rate of the reaction, and a chemoselectivity
favouring the M-ene reaction.
3. Discussion
The general picture which emerges is that as the
medium is made more polar, either by varying the
solvent or by adding a salt, the overall rate of the
reaction increases, and the chemoselectivity becomes
biased towards the M-ene reaction, so that it is often
the only product.
Typical examples of metalloene reactions were as
follows:
“ A solution of tricyclohexyl (prop-2-enyl)tin (133 mg,
0.325 mmol) and tetraphenylporphin (3 mg) in di-
ethyl ether (5 cm3) and a 4 mol dm−3 solution of

W.J. Kinart et al. / Journal of Organometallic Chemistry 608 (2000) 49–53
53
LiClO4 in Et2O, or Rose Bengal (in CH3OH), was
vigorously stirred under oxygen while being irradi-
ated with light from a 400 W sodium lamp for 3 h.
The solvent was removed and the product was iden-
tified by NMR. All illuminations of the allyltin
compounds were carried out under the same condi-
tions (pressure of oxygen above the solution, inten-
sity of irradiation of the sample, temperature (298
K) etc. were constant).
K was monitored by measuring the absorbance at 410
nm. The analogous studies with PTAD were carried out
with 0.0046 mol dm−3 solutions of the azo compound
and allyltin derivative. The progress of the reaction at
298 K was monitored by measuring the absorbance at
536 nm. We measured times corresponding to the de-
crease of the initial absorbance by 50% for reactions of
PTAD and DEAD with chosen allyltin derivatives and
assume that their ratio corresponds to a certain degree
with the ratio of the reaction rate constants.
“ Tricyclohexyl(prop-2-enyl)tin (66 mg, 0.1615 mmol)
and DEAD (27.7 mg, 0.1585 mmol) were added to
CH3OH (1 cm3). The progress of the reaction was
monitored by TLC (using light petroleum-ethyl ace-
tate (7/3 v/v) as eluent) and by NMR spectroscopy,
which showed that only diethyl N-allyl-N%-tricyclo-
hexylstannylhydrazodicarboxylate was formed in
quantitative yield. Isolation of the product by gradi-
ent chromatography (light petroleum–ethyl acetate)
gave diethyl N-allylhydrazodicarboxylate as an oil.
“ PTAD (55 mg, 0.314 mmol) and allyltricyclohexyltin
(133 mg, 0.325 mmol) were added to C6H6 (1 cm3).
After discharge of the colour of the azo compound,
the solvent was removed and the product was iden-
tified by NMR.
References
[1] H.-S. Dang, A.G. Davies, J. Chem. Soc. Perkin Trans. 2 (1991)
721.
[2] H.-S. Dang, A.G. Davies, Tetrahedron Lett. 32 (1991) 1745.
[3] (a) W. Adam, B. Nestler, J. Am. Chem. Soc. 114 (1992) 6549. (b)
ibid 115 (1993) 5041.
[4] M. Stratakis, M. Orfanopoulos, C.S. Foote, Tetrahedron Lett.
37 (1996) 7159.
[5] W. Adam, H.-G. Brunker, J. Am. Chem. Soc. 115 (1993) 3008.
[6] H.-G. Brunker, W. Adam, J. Am. Chem. Soc. 117 (1995) 3976.
[7] H.-S. Dang, A.G. Davies, J. Chem. Soc. Perkin Trans. 2 (1991)
2011.
[8] H.-S. Dang, A.G. Davies, J. Organomet. Chem. 430 (1992) 287.
[9] W.J. Kinart, C.M. Kinart, I. Tylak, J. Organomet. Chem. 590
(1999) 258.
[10] (a) A.G. Davies, W.J. Kinart, J. Chem. Soc. Perkin Trans 2
(1993) 2281. (b) W.J. Kinart, J. Chem. Res. (S) (1994) 486. (c)
W.J. Kinart, I. Tylak, C.M. Kinart, J. Chem. Res. (S) (1999) 46.
Kinetic measurements with PTAD and DEAD were
carried out in a 1 cm3, UV cell. The concentration of
DEAD and allyltin derivative chosen was equal to
0.0254 mol dm−3. The progress of the reaction at 298