Polyoxotungstate Photoinduced Alkylation of Electrophilic Alkenes by Cycloalkanes

Article abstract of DOI:10.1002/chem.200305285

Alkyl radical obtained by irradiation of tetrabutylammonium decatungstate in acetonitrile in the presence of cycloalkanes (C₅H₁₀, C₆H₁₂, C₇H₁₄) are efficiently trapped by electrophilic alkenes (acrylonitrile, isopropylydenmalonitrile, isopropylydencyanoacetate) to give the corresponding alkylated aliphatic nitriles. The reaction can be carried out up to complete conversion of the alkene with reasonable (in most cases 60-65%) yields. Addition of the radicals to the alkene is followed by electron transfer from reduced decatungstate regenerating the sensitizer (turn over number up to 60). Steady-state measurements, EPR evidence, deuteration experiments and attempted intramolecular trapping of the adduct radical support the mechanistic proposal.

Full text of DOI:10.1002/chem.200305285

FULL PAPER
Polyoxotungstate Photoinduced Alkylation of Electrophilic Alkenes by
Cycloalkanes
Daniele Dondi,[a] Maurizio Fagnoni,[a] Alessandra Molinari,[b] Andrea Maldotti,[b] and
Angelo Albini*[a]
Abstract: Alkyl radical obtained by ir-
radiation of tetrabutylammonium deca-
tungstate in acetonitrile in the presence
of cycloalkanes (C5H10, C6H12, C7H14)
are efficiently trapped by electrophilic
alkenes (acrylonitrile, isopropylyden-
be carried out up to complete conver-
sion of the alkene with reasonable (in
most cases 60 65%) yields.Addition
of the radicals to the alkene is followed
by electron transfer from reduced dec-
atungstate regenerating the sensitizer
(turn over number up to 60).Steady-
state measurements, EPR evidence,
deuteration experiments and attempted
intramolecular trapping of the adduct
radical support the mechanistic propos-
al.
Keywords: alkylation
activation photocatalysis
photochemistry ¥ radicals
¥
C H
malonitrile,
isopropylydencyanoace-
¥
¥
tate) to give the corresponding alkylat-
ed aliphatic nitriles.The reaction can
Introduction
radical (and possibly other radical intermediates formed in
the course of the reaction) remains bonded to the polyoxo-
tungstate.[2,7] As for the reactions of the alkane, it has been
demonstrated that alkyl radicals add to ethylene, 1-hexene,
acetylene,[8] carbon monoxide,[9] nitriles[10] and cyanofor-
mate[11] forming a CꢀC bond, besides adding to oxygen to
give hydroperoxides and alcohols and ketones from them.[7,12]
Somewhat surprising, the most typical reaction of alkyl radi-
cals, addition to electrophilic alkenes[13] has not been ex-
plored.Since the most useful synthetic paths for organic
compounds via alkyl radicals are based on the last reaction,
due to the well known nucleophilic character of such spe-
cies, we deemed it worthwhile to explore the potential of
decatungstate photocatalysis in this direction.
Photoexcited polyoxometalate anions, such as the decatung-
4ꢀ [1]
state W10O32
,
react with organic molecules in solution.
Much work has been devoted to the mechanistic aspects of
these photochemical processes[2] or to applications such as
photomicrolithography[3] and the photocatalytic decontami-
nation of toxic materials.[4] However, these reactions are not
destitute of synthetic interest, in particular when applied to
the generation of alkyl radicals from alkanes [Eq.(1)], [5]
thus contributing a mild activation path for the activation of
aliphatic CꢀH bonds.[6]
W10O432ꢀ þ RH ! H W10O35ꢀ2 þ RC
þ
*
ð1Þ
The reaction as represented in Equation (1) actually is a
gross oversimplification, and much work has been devoted
to clarify mechanistic questions, such as which of the excited
states of the decatungstate participates into the chemical
step, which is the protonation state of the reduced species,
most often designed as H+ W10O325ꢀ, and whether the alkyl
Results
The experiments were carried out by irradiating 0.002m tet-
rabutylammonium decatungstate (TBADT) in acetonitrile.
The solutions were degassed by freeze-degas-thaw cycles.
When a 0.5m cyclohexane solution was used, 0.025m each of
both bicyclohexyl and cyclohexyl methyl ketone were
formed by a 3 h irradiation.An analogous experiment in the
presence of 0.1m 1-hexene gave some hexylcyclohexane (1,
4% with respect to the starting alkene) and a reduced yield
of the above products (Scheme 1, Table 1).Both experi-
ments duplicated the results previously obtained by Hill
under essentially identical conditions and, again as in his
case, prolonged irradiation only marginally increased the
yield of the products.[8]
[a] D.Dondi, Dr.M.Fagnoni, Prof.A.Albini
Department of Organic Chemistry, University of Pavia
via Taramelli 10, 27100 Pavia (Italy)
Fax : (+39)0382-507-323
E-mail: angelo.albini@unipv.it
[b] Dr.A.Molinari, Prof.A.Maldotti
Department of Chemistry, University of Ferrara
Via Borsari 46, 44100 Ferrara (Italy)
Fax : (+39)0532-240-709
E-mail: mla@dns.unife.it
142
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
DOI: 10.1002/chem.200305285
Chem. Eur. J. 2004, 10, 142 148

142 148
We then used further electrophilic alkenes.With isopropy-
lydenmalononitrile (IPMN) alkylation was successful giving
malononitrile 7 and similarly adduct 8 was obtained from
isopropylydencyanacetate (IPCA).On the other hand, using
tetracyanoethylene (TCNE, 0.001m) caused the rapid devel-
opment of yellow color and no alkylated product was
formed.
The course of the alkylation was further probed by using
a w-alkenyl derivative of IPMN, 2-cyano-3-methylhept-6-di-
enecarbonitrile, CMHN.A single product was obtained in
this case and recognized as the 3-(w-alkenyl)-3-cyclohexyl-
hept-6-enecarbonitrile (9).On the contrary, cyclized dinitrile
10 that we had previously obtained through radical addition
onto CMHN under different conditions (see Scheme 2)[14]
was not formed in the present case.
Scheme 1.
Table 1.Alkylation of unsaturated nitriles by cycloalkanes upon deca-
tungstate photocatalysis.
Alkane, 0.5m Trap, 0.1m
Products (% Yield)[a]
RꢀR
5
RCOMe
Adducts
Scheme 2.
cyclohexane
none
5
1-hexene
AN
AN (1m)
IPMN
IPCA
1
1.5
1 (4)
2 (63)
2 (21), 3 (31), 4 (11)
7, 66
8, 58
9, 45
5, 30
6, 51
The origin of the a-hydrogen atom in the alkylated ni-
triles obtained was tested.Thus, using perdeuterocyclohex-
ane in the reaction with AN led to no detectable deutera-
tion, while adding 0.5% D2O led to substantial deuterium
incorporation (50%, see Table 2).Duplication of the experi-
ment with IPMN was precluded by the fast proton exchange
of the resulting dinitrile.[15a]
CMHN
cyclopentane AN
cycloheptane AN
[a] After 3 h irradiation with 0.002m TBADT, see Experimental Section.
Table 2.Deuteration in the alkylated nitriles.
However, the use of an electrophilic alkene such as acryl-
onitrile (AN, 0.1m) led to a much more efficient reaction,
yielding 63% of cyclohexylpropionitrile (2, Scheme 1) in
3 h.Furthermore, very little ( <1%) of bicyclohexyl and the
ketone were formed under this condition and no other prod-
uct was detected.Preparative experiments for products sep-
aration and identification were similarly carried out in argon
flushed solutions (see Experimental Section).Competition
experiments with 1-hexene and AN both present (both 0.1
m) gave >100 AN trapping versus cyclohexene trapping.
When a 1m rather than 0.1m AN concentration was used,
a more complex product distribution resulted, and products
containing two and three AN units (compounds 3 and 4)
were among the major products, with 3 actually the most
abundant one.
Reagents
Solvent
a-Deuteration
C6D12, AN
C6H12, AN
CH3CN
CH3CN, 0.5% D2O
a[D]2, <5%
a[D]2, 50%
In the above experiments with TBADT and cyclohexane
the solution turned to the blue color characteristic of the
H+ W10O325ꢀ/H2W10O32
mixture, as previously describe-
4ꢀ
d.[10a,15b] The color developed also in the presence of the
electrophilic alkenes and the course of the reaction was fol-
lowed by determining the amount of reduced polyanion and
alkylated product on the same sample under degassed con-
ditions at various irradiation times.The results are reported
in Figure 1.In order to have a mechanistic significance, the
data are presented in the form of number of electrons ac-
cepted by the polyoxotungstate to yield the blue reduced
form (left-hand ordinate) and moles of alkylated nitrile
formed (right-hand ordinate).As it appears from the figure,
reduced polyoxotungstate developed in minutes when irradi-
The reaction could be easily extended to other cycloal-
kanes, as shown in Scheme 1 and Table 1, at least for those
sufficiently soluble in acetonitrile, such as cyclopentane and
cycloheptane (0.1m), which yielded nitriles 5 and 6.
143
Chem. Eur. J. 2004, 10, 142 148
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim

A.Albini et al.
FULL PAPER
Figure 2.EPR spin trapping spectra obtained irradiating ( l> 300 nm)
TBADT (0.002m) in a nitrogen-flushed solution containing cyclohexane
(0.3m), AN (0.1m) and PBN (0.05m).A) After a few seconds irradiation
(aN =13.9 G, aH =2.2 G, light line); B) after 20 min irradiation (aN =14.3
G, aH =3.4 G, heavy line).
Figure 1.Results from the irradiation of 00. 02 m tetrabutylammonium
decatungstate (TBADT) in acetonitrile in the presence of 0.5m cyclohex-
ane.Filled symbols (left-hand scale): moles of electrons accepted by the
decatungstate as measured by the formation of the two reduced forms.
Empty symbols (right-hand scale): moles of alkylated products formed.
~ ~
&
*
^
(
) No additive; ( ) 0. 1m 1-hexene; ( , ) 0. 1m AN; ( ,^) 0. 1m IPMN.
ated in the presence of cyclohexane in MeCN.The color de-
veloped faster when acrylonitrile was added, while it grew
at a somewhat slower rate in the presence of IPMN, at least
in the first part of the conversion.In all cases, a plateau was
reached and this was not much different in the three experi-
ments, corresponding to 60 70% conversion of the polyoxo-
tungstate into the reduced form (using the published e
values for the two reduced forms, see below).[15b] The alky-
lated adducts with both AN and IPMN developed at rough-
ly the same rate and in a molar amount much larger than
that of reduced TBADT accumulating.
It is apparent from Table 1 that TBADT is acting as a
photocatalyst, since about 0.06 moles products are formed
when 0.002 moles of it are used. For mechanistic indication,
we considered as turn-over number the ratio between the al-
kylated products formed and the electrons accepted by the
photocatalyst, calculated from Figure 1.In the alkylation of
AN, the ratio was 6 at the beginning of the reaction (4%
AN converted) and grew to 46 in the advanced phase (60%
AN converted) and it similarly grew from 12 to 60 with
IPMN.
EPR experiments were also carried out.Irradiation of
TBADT in an oxygen-equilibrated 0.3m cyclohexane solu-
tion in MeCN in the presence of 0.05m a-phenyl N-tert-bu-
tylnitrone (PBN) gave rise to a triplet of doublets attributa-
ble to the trapping of the cyclohexyl radical, in agreement
with previous work.[16] When a similar experiment was car-
ried out in a nitrogen-flushed solution to which 0.1m AN
had been added, the same signal was initially observed but
after 20 min was superseded by a new triplet of doublets,
corresponding to the trapping of a different C-centered radi-
cal (Figure 2).Using 01. m IPMN gave a similar result, with
the second radical being revealed earlier (5 min, Figure 3).
Figure 3.EPR spin trapping spectra obtained irradiating ( l>300 nm)
TBADT (0.002m) in a nitrogen-flushed solution containing cyclohexane
(0.3m), IPMN (0.1m) and PBN (0.05m).A) After a few seconds irradia-
tion (aN =13.9 G, aH =2.2 G, light line); B) after 5 min irradiation (aN =
15.0 G, aH =2.9 G, heavy line).
Indeed, this signal appeared also with a 1î10ꢀ3 m concentra-
tion of IPMN, though the reaction was slower and the
second radical developed at a longer time (ca.20 min).The
spectra of the second radical had a different structure with
the two alkenes, as showed by the different coupling con-
stants.No significant signals were detected when omitting
TBADT or cyclohexane or when leaving the solution in the
dark.With 1î10 ꢀ3 m TCNE the nine-lines signals corre-
sponding to the TCNE radical anion was present from the
beginning along with the triplet of doublets resulting from
the trapping of the cyclohexyl radical.
The role of oxygen was also investigated.Repeating the
small-scale experiments with cyclohexane and AN omitting
degassing showed that, while the deep blue color of reduced
polyoxotungstate did not develop, alkylation to yield prod-
uct 2 proceeded steadily, though at rate about one forth
than in oxygen-free solution.In preparative experiments, ir-
144
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 142 148

Alkylation of Electrophilic Alkenes
142 148
radiation of TBADT in non de-aerated solutions containing
cyclohexane led to small amounts of cyclohexanone and cy-
clohexanol, but in the presence of AN or IPMN, alkylated
nitriles 2 and 5 were formed in good yield also when omit-
ting degassing.
and has been suggested to involve previous reduction of the
radical to the carbanion.[10] All of the products were ob-
tained in a low yield, in keeping with the literature.[8] When
present, 1-hexene is alkylated, again in a low yield.This is
due to the fact that reaction of the cyclohexyl radical with a
nucleophilic alkene is not favored and thus the persistent
radical effect limits the conversion, just as in the absence of
traps.
Discussion
The present experiments show that with an electrophilic
alkene the TBADT photoinduced reaction of cyclohexane is
rather efficient (F 0.25) and can be carried out up to elevate
conversion with no declining of the yield (Figure 1).As one
may expect from the well known behavior of alkyl radicals
in solution,[13] addition to AN is more than 100 times as ef-
fective as addition to 1-hexene.The rate constant for the ad-
dition of cyclohexyl radical to methyl acrylate has been re-
cently measured as 3.3î106 mꢀ1 sꢀ1 at 258C.[18]
The detailed mechanism of the formation of radicals has
been previously explored and both a hydrogen abstraction
and an electron transfer-radical cation deprotonation mech-
anism have been considered, the former one appearing
more likely.[2a,7a] The present trapping corresponds to the
first step of the classical conjugate radical addition,[13] but
since there is no hydrogen donor in solution, the stabilized
The basic photochemical reactions of polyoxotungtate com-
plexes have been fully clarified by extensive mechanistic in-
vestigations.[2,7] In essence, electron and hydrogen transfer
generate alkyl radicals according to Equation (1).Reverse
hydrogen atom transfer [Eq.(2)] then competes with cou-
pling [Eq.(3)] or trapping [by a suitable reagent X, Eq.(4)]
of the radicals.Hydrogen transfer to the adduct radicals
formed in the latter case may occur according to Equa-
tion (5).Furthermore, the reduced polyoxotungstate dispro-
portionates on a relatively long (up to seconds) time scale
yielding the diprotonated two-electron reduced anion
[Eq.(6)].
Hþ W10O352ꢀ þ RC ! W10O342ꢀ þ RH
2 RC ! products ðR2; alkenes etc:Þ
RC þ X ! R-XC
ð2Þ
ð3Þ
ð4Þ
ð5Þ
(and much less electrophilic) adduct radical 11 is reduced by
5ꢀ
the persistent H+ W10O32
species to anion 12 (see
Scheme 3).Reduction of the a-cyano radical 11, character-
ized by a high electron affinity,[19] is efficient, while this is
Hþ W10O532ꢀ þ R-XC ! W10O34ꢀ2 þ R-XH
Hþ W10O35ꢀ2 ! W10O432ꢀ þ H2W10O342ꢀ
ð6Þ
Formation of alkyl radicals according to Equation (1) has
been well documented, for example by spin trapping,[16] as
has their trapping by aliphatic alkenes and alkynes.[8] Such
addition processes are interesting as a way for the direct
functionalization of inactivated CꢀH bonds, but occur
slowly (F <0.05) and only up to a low extent. As an exam-
ple, Hill reported that ethylcyclohexane is formed from cy-
clohexane and ethene with 83% selectivity and 8 9% con-
version after 78 h, and conversion could not be brought over
10%.[8] This is probably due to the ™persistent radical ef-
fect∫,[7b,17] namely to the fact that, while highly reactive alkyl
radicals are consumed through various paths (dimerization,
disproportionation to alkenes and other unidentified), a
Scheme 3.
5ꢀ
more persistent species such as H+ W10O32 builds up and
not the case for the starting alkyl radical (compare Ered for
TBADT, ꢀ1.2 V,[2a] and for sec-butyl radical, ca. ꢀ1.35 V vs
SCE,[20] and take into account that other reactions including
this in turn increases the role of reverse reactions such as
that in eq.2, thus inhibiting the progress of the addition.
Therefore, the process has limited synthetic interest, except
when a good trap for the nucleophilic radical is available, as
in the case for CO or cyanoformates.In particular, the reac-
tion in the presence of oxygen has been studied in detail.
This leads rather efficiently to alcohols or ketones (F 0.35).
alternative paths, such as Equations (3), (4), limit its steady
[10]
state concentration).As mentioned above,
reduction of R.
to Rꢀ has a role in the absence of traps and probably ex-
plains the formation of small amounts of methyl ketone via
attack of the carbanion to the solvent acetonitrile and hy-
drolysis of the resulting imine.
Dioxygen traps the alkyl radicals and the resulting peroxyl
5ꢀ [7a,12a]
radicals are reduced to hydroperoxides by H+ W10O32
.
Deuteration experiments (Table 2) support the role of
anion 12 by the positive result with D2O.On the other hand,
the negative result with C6D12 indicate that neither hydrogen
back donation from reduced polyoxotungstate, nor hydrogen
Table 1 shows that in the absence of any trap cyclohexyl
radicals couple to bicyclohexyl, disproportionate or add to
acetonitrile.The last process gives cyclohexyl methyl ketone
145
Chem. Eur. J. 2004, 10, 142 148
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim

A.Albini et al.
FULL PAPER
abstraction from the alkane by the adduct radical (at any
rate thermodynamically unfeasible) have any role.In neat
MeCN, the a proton reasonably arises from traces of water
present.
that is 60 moles of the alkene are alkylated for every elec-
tron accepted by TBADT to give H+ W10O324ꢀ.The prepara-
tive significance of the method is indicated by alkylation
yields attaining 66% with the photocatalyst used in a 1 to
50 proportion with respect to the alkene.
The mechanism requires that the electrophilic alkene
added functions as a trap, not interfering with the sensitiza-
tion cycle (Scheme 3).Indeed, this holds true for AN ( Ered
= ꢀ2.6 V vs SCE)[22] and IPMN (ꢀ1.75),[22] but the limits of
the method are clearly shown by the fact that TCNE (+
0.23)[23] is reduced to the radical anion [Eq.(7), as indicated
by the EPR and by the yellowing of the solution], making
alkylation impossible.
In keeping with this hypothesis, incorporation of further
2
AN units becomes significant (accounting for = of the iso-
3
lated products) at 1m AN, when concentration is high
enough to make relatively slow addition by nucleophilic rad-
ical 11 to AN (rate constant ca.1î10 4 mꢀ1 sꢀ1) competitive
with reduction.Addition yields radical 13 (Scheme 3).Incor-
poration of two or three alkene units has been previously
detected in some conjugate alkylation reactions of radicals
to acrylonitrile, though the products have not been isolat-
ed.[21]
The intermediacy of both cyclohexyl radical and of adduct
radicals 11 with both AN and IPMN is also evidenced by
trapping by phenyl-tert-butylnitrone to give the correspond-
ing nitroxyl radicals 14 (Figures 2 and 3).EPR experiments
also show that the highly stabilized a,a-dicyano radical
adduct from IPMN builds up to a larger concentration than
in the case of the monocyano radical from AN.
Hþ W10O532ꢀ þ TCNE ! W10O342ꢀ þ TCNECꢀ þ Hþ
ð7Þ
Importantly, this is a stoichiometric process in terms of
light absorbed, not a photoinitiated chain process, that
would be impossible due to the stability of the radical
adduct.This makes the process less oxygen-sensitive.In fact,
oxygen slows down the alkylation but does not inhibit it,
which is not quite surprising when one takes into account
that this is no chain process.Thus, the oxygen effect is limit-
ed to the competition with the alkene (the latter being a
hundred times more concentrated) for the radical, but the
multiplicative effect of chain reactions is lacking.Actually,
we found that the preparative alkylation could be carried
out to a satisfactory conversion also when omitting de-aera-
tion.Apparently, the slowing down of the alkylation due to
radical trapping by oxygen is in part compensated for by the
much lower conversion of the polyoxometallate into the re-
duced form, which is re-oxidized by oxygen (the solution re-
mains of a pale blue), and thus by the greater light fraction
absorbed by the active species.
Further support to the mechanism comes from the fact
that when using CMHN as the trap, no cyclization takes
place, while this is the main process (cyclized 10 vs open
chain 9 3 to 1)[14] when the cyclohexyl radical is generated
by an alternative method, for example hydrogen abstraction
from cyclohexane by triplet benzophenone.Under the pres-
ent conditions, on the contrary, the 5-hexenyl radical 15 is
Conclusion
reduced by H+ W10O32 accumulated to a significant con-
5ꢀ
Photocatalysis by polyoxoanions, and in particular by
TBADT, can be considered a mild and efficient method for
the generation of alkyl radicals from alkanes.The mecha-
nism of the primary steps in the TBADT alkane MeCN
system has been studied in detail elsewhere.For the present
purpose, the significant point is that radicals are trapped by
electrophilic alkenes in a reaction (Scheme 3) that follows a
path quite similar to that proved for the trapping by oxygen
to give hydroperoxides.In this way, conjugate alkylations
have been obtained that extend the classical radical alkyla-
tion reactions since the radical precursor is an alkane, rather
than a bromide or iodide, and represent a mild method for
the activation of aliphatic CꢀH bond with CꢀC bond forma-
tion.This significantly adds to the synthetic appeal of poly-
oxoanions photoinduced reactions.
centration to give anion 16, so that cyclization is completely
inhibited (Scheme 2).
Electron transfer to the radical adduct regenerates the
starting polyoxometalate that is used in catalytic amounts
and operates with an initial turn-over number of 6 with AN
and 12 with IPMN.The latter trap scavenges alkyl radicals
more efficiently and also yields a more easily reduced
adduct radical.Regeneration of TBADT is not complete
5ꢀ
and part of H+ W10O32
undergoes disproportionation
(known rate constant 1î107 mꢀ1 sꢀ1)[7f] to give the two-elec-
trons reduced form H2W10O32 (see Figure 1).As men-
4ꢀ
tioned above, the efficiency of this step is lower in the reac-
tion with IPMN due to the faster reoxidation.During irradi-
ation, an increasing fraction of the polyoxoanion is accumu-
2
lated in the reduced form, arriving at about = with the elec-
3
4
trophilic alkenes (up to = in their absence, see Figure 1),
5
but a sufficient amount of it remains to make the photocata-
lytic action continue, also thanks to the choice of the irradia-
tion wavelength at 310 nm, where TBADT absorbs strongly.
Thus, the turn over number at high conversion is about 60,
Experimental Section
Materials: Tetrabutylammonium decatungstate (TBADT) was prepared
according to a published procedure.[24] The electrophilic alkenes IPMN,
146
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 142 148

Alkylation of Electrophilic Alkenes
142 148
IPCA[25] and CMHN[14] were prepared as previously reported.The other
materials and the solvents were of commercial origin.
trogen flushing and put in the EPR cell under inert atmosphere.An anal-
ogous experiment has been carried out with an oxygen equilibrated solu-
tion in the absence of the alkene.Then, photochemical excitation was
carried out with light of wavelength higher than 300 nm directly inside
the instrument cavity, using a flat quartz cell as the reaction vessel.Anal-
ogous experiments were performed irradiating in the absence of TBADT,
or in the absence of C6H12, or in the dark.
Preparative irradiations: These were carried out in 1 cm diameter quartz
tubes containing a solution (6 mL) of TBADT (40 mg, 0.002m), the
alkane (0.5m) and the electrophilic alkene (0.1m).These were purged
with purified Argon for 10 min, serum capped and irradiated for 15 20 h
in a multilamp apparatus fitted with six 15 W phosphor-coated lamps
(center of emission 310 nm).The irradiated solution was flushed with
oxygen until colorless, passed through a 4 cm layer of neutral alumina
and evaporated.The residue was first examined by GC/MS and then
chromatographed on silica gel by using cyclohexane and cyclohexane/
ethyl acetate 98:2.With IPMN essentially the same results were obtained
also when omitting degassing.
Acknowledgements
Support of this work by MURST, Rome, through a PRIN Program is
gratefully acknowledged.DD. .thanks Solchem for a fellowship.
Compounds 2, 5, 6 and 9 were identical to samples prepared through al-
ternative procedures.[14] Hexylcyclohexane was recognized by comparison
of the GC/MS spectrum with literature data,[26] bicyclohexyl and cyclo-
hexylmethylketone by comparison of the GC/MS spectrum with that of
authentic samples.
2-Cyclohexylmethylpentanedicarbonitrile (3): colorless syrup; 1H NMR
(300 MHz, CDCl3): d=0.8 1.8 (m, 13H), 1.95 (m, 2H; 3-H2), 2.6 (m, 2H;
4-H2), 2.85 (m, 1H; 2-H); 13C NMR (75 MHz, CDCl3, TMS): d = 15.8
(CH2; 4-C), 26.2 (CH2), 26.3 (CH2), 26.6 (CH2; cyclohexyl ring), 28.7
(CH), 29.0 (CH2), 32.7 (CH2), 33.8 (CH2), 35.8 (CH), 39.8 (CH2), 118.0
(CN), 120.4 (CN); GC/MS: m/z (%): 190 (68), 150 (25), 135 (62), 122
(18), 108 (43), 95 (37), 82 (48), 65 (100).
[1] C.L. Hill, C.M. Prosser-McCartha, Coord. Chem. Rev. 1995, 143,
407 455.
[2] a) R.F. Renneke, M. Pasquali, C.L. Hill,
J. Am. Chem. Soc. 1990,
112, 6585 6594; b) C.Tanielian, Coord. Chem. Rev. 1998, 178 180,
1165 1181; c) I. Texier, J.F. Delouis, J.A. Delaire, C. Giannotti, P.
Plaza, M.M.Martin, Chem. Phys. Lett. 1999, 311, 139 145.
[3] a) H.Okamoto, A.Ishikawa, T.Kudo,
J. Photochem. Photobiol. A
1989, 49, 377 385; b) C.L.Hill, M.Kozik, J.Winkler, Y.Hou, C.M.
Prosser-McCartha, Adv. Chem. Ser. 1993, 238, 243 259.
4-Cyano-2-cyclohexylmethylheptanedicarbonitrile (4): (diastereoisomers
mixture; the sample obtained by chromatography was somewhat conta-
minated by dinitrile 3), colorless syrup; 1H NMR (300 MHz, CDCl3,
TMS): d=0.8 1.9 (m, 13H), 2.05 (m, 2H; 6-H), 2.6 (m, 2H; 3-H), 2.9 3.1
(two m, 2H; 2-H, 4-H ; 13C NMR (75 MHz, CDCl3): d=15.6 (CH2), 26.3
(CH2), 26.6 (CH2), 27.0 (CH), 28.0 (CH2), 29.0 (CH), 30.3 (CH), 32.7
(CH2), 33.9 (CH2), 40.2 (CH2), 118.5 (CN), 120.2 (CN), 120.3 (CN); GC/
MS (two peaks with almost identic pattern): m/z (%): 243 (8), 203 (12),
188 (15), 150 (18), 137 (20), 122 (12), 108 (24), 82 (23), 55 (100).
[4] a) A.Mylonas, A.Hiskia, E.Papaconstantinou,
J. Mol. Catal. A
Chem. 1996, 114, 191 200; b) R.R. Ozer, J.L. Ferry, Environ. Sci.
Technol. 2001, 35, 3242 3246; c) A.Hiskia, M.Ecke, A.Troupis, A.
Kokorakis, H.Hennig, E.Papaconstantinou, Environ. Sci. Technol.
2001, 35, 2358 2364.
[5] a) R.F. Renneke, C.L. Hill,
J. Am. Chem. Soc. 1988, 110, 5461
5470; b) C.L.Hill, Synlett 1995, 127 132; R.F.Renneke, C.G.Hill,
J. Am. Chem. Soc. 1986, 108, 3528 3529; C.L. Hill, C.M. Prosser-
McCartha, Cat. Met. Compl. 1993, 14, 307 330; d) L.A. Combs-
Walker, C.L.Hill, J. Am. Chem. Soc. 1992, 114, 938 946; e) E.Pa-
paconstantinou, Chem. Soc. Rev. 1989, 18, 1 31; f) C.L. Hill, R.F.
Renneke, L.A.Combs, New J. Chem. 1989, 13, 701; g) M.K.Awad,
A.B. Anderson, J. Am. Chem. Soc. 1990, 112, 1603 1606; h) N.Z.
Murazov, M.I.Rustamov, Kinet. Katal. 1989, 30, 248; i) M.Graetzel,
K.R.Thampi, J Kiwi, J. Phys. Chem. 1989, 93, 4128 4132.
[6] a) A.A.Fokin, P.R.Schreiner, Chem. Rev. 2002, 102, 1551; b) Acti-
vation and Functionalization of Alkanes (Ed.: C. L. Hill), Wiley,
New York, 1989.
2-(1-Cyclohexyl-1-methylethyl)propanedicarbonitrile (7): colorless solid;
m.p. 48 508C; elemental analysis calcd (%) for C 75.7, H 9.7, N 14.5;
found: C 75.74, H 9.54, N 14.72; 1H NMR (300 MHz, CDCl3, TMS): d=
1.2 (s, 6H; Me2), 1.0 1.95 (several m, 11H; cyclohexyl), 3.7 (s, 1H; 2-H);
13C NMR (75 MHz, CDCl3): d = 21.7 (CH3), 25.9 (CH2), 26.3 (CH2),
27.1 (CH2), 33.5 (CH), 41.0, 44.1 (CH), 112.1 (CN); IR (melt): n˜
=
2253 cmꢀ1; GC/MS: m/z (%): 190 (1) [M+], 175 (1), 125 (4), 107 (3), 83
(100).
Ethyl 2-cyano-3-cyclohexyl-3-methylbutanoate (8): colorless syrup;
1H NMR (300 MHz, CDCl3, TMS): d =1.1 (s, 3H; Me), 1.3 (s, 3H, Me),
1.0 1.9 (several m, 11H; cyclohexyl), 1.35 (t, 3H; Me), 3.6 (s, 1H; 2-H),
4.25 (q, 2H; CH2); 13C NMR (75 MHz, CDCl3): d =13.4 (CH3), 20.9
(CH3), 21.7 (CH3), 25.6 (CH2), 25.9 (CH2), 26.1 (CH2), 26.4 (CH2), 26.6
(CH2), 40.3, 44.9 (CH), 46.5 (CH), 61.6 (CH2), 116.3 (CN), 165.6 (CO);
IR (neat): n˜ = 2246, 1742 cmꢀ1; GC/MS: m/z (%): 237 (10) [M+], 164
(18), 140 (21), 124 (65), 114 (42), 82 (100).
[7] a) C.Tanielian, R.Seghrouchni, C.Schweiter,
J. Phys. Chem. A
2003, 107, 1102 1110; b) T.Kohte, R.Martschke, H.Fischer,
Chem. Soc. Perkin Trans. 2 1998, 503 507; c) H.Duclusaud, S.A.
Borshch, Chem. Phys. Lett. 1998, 290, 526 534; d) C.Tanielian, K.
Duffy, A.Jones, J. Phys. Chem. B 1997, 101, 4276 4282; e) D.C.
Duncan, M.A.Fox, J. Phys. Chem. A 1998, 102, 4559 4567; f) L.P.
Ermolenko, J.A.Delaire, C.Giannotti, J. Chem. Soc. Perkin Trans.
J.
2 1997, 25 30; g) DC. .Duncan, TL. .Netzel, CL. .Hill,
Chem. 1995, 4640 4646.
Inorg.
Steady state measurements: Solutions (2 mL) of 0.002m TBADT in
MeCN containing the appropriate additives in 1 cm optical path spectro-
photometric couvettes were degassed by five freeze-degas-thaw cycles (to
10ꢀ6 Torr).The samples were irradiated by means of a focalised Osram
150 W high-pressure mercury arc fitted by an interference filter at
313 nm or a band-pass filter (the first one for quantum yield measure-
ments, the latter for comparing reactivities, see the results after 3 h irradi-
ation in Table 1).Light absorbed was determined by using a photometer
and the light flux was measured through ferrioxalate actinometry.Forma-
tion of the alkylated products was determined by GC on the basis of cali-
bration curves by using an internal standard (biphenyl).Formation of the
two reduced forms of TBADT was monitored by spectroscopy in the visi-
ble on the basis of the known molecular extinction coefficient of such
molecules.[15b]
[8] B.S.Jaynes, C.L.Hill, J. Am. Chem. Soc. 1993, 115, 12212 12213.
[9] B.S.Jaynes, C.L.Hill, J. Am. Chem. Soc. 1995, 117, 4704 4705.
[10] a) R.F.Renneke, M.Kadkhodayan, M.Pasquali, C.L.Hill,
J. Am.
Chem. Soc. 1991, 113, 8357 8367; b) C.M.Prosser-McCartha, C.L.
Hill, J. Am. Chem. Soc. 1990, 112, 3671 3673.
[11] Z.Zeng, C.L.Hill, Chem. Commun. 1998, 2467 2468.
[12] a) L.P.Ermolenko, C.Giannotti, J. Chem. Soc. Perkin Trans. 2 1996,
6, 1205 1210; b) C.Giannotti, C.Richter, Int. J. Photoenerg. 1999,
1, 69 73; c) C.Giannotti, C.Richter, Trends Photochem. Photobiol.
1997, 4, 43 54; d) A.Molinari, R.Amadelli, V.Carassiti, A.Mal-
dotti, Eur. J. Inorg. Chem. 2000, 91 96; e) A.Maldotti, A.Molinari,
G.Varani, M.Lenarda, L.Storaro, F.Bigi, R.Maggi, A.Mazzacani,
G.Sartori, J. Catal. 2002, 209, 210 216; f) A.Molinari, R.Amadelli,
EPR Measurements: EPR spectra were recorded with a X-band Bruker
220 SE spectrometer that was calibrated by using a,a’-diphenylpicrylhy-
drazyl.EPR spin trapping experiments were performed using a-phenyl,
A.Mazzacani, G.Sartori, A.Maldotti,
5405; g) Y.Guo, C.Hu, S.Jiang, C.Guo, Y.Yang, E.Wang,
Catal. B 2002, 36, 9 17.
Langmuir 2002, 18, 5400
Appl.
N-tert butyl nitrone (PBN) as a spin trap.In a typical experiment CH CN
solution containing TBADT (0.002m), C6H12 (0.3m), PBN (0.05m) and an
alkene (AN or IPMN, 0.1m, or TCNE 0.001m) were deoxygenated by ni-
[13] a) B.Giese, Radicals in Organic Synthesis: Formation of Carbon-
Carbon Bonds, Pergamon, Oxford, 1986; b) D.P.Curran in Compre-
hensive Organic Synthesis, Vol. 4 (Eds.: B. M. Trost, I. Fleming), Per-
3
147
Chem. Eur. J. 2004, 10, 142 148
www.chemeurj.org
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim

A.Albini et al.
FULL PAPER
gamon, Oxford, 1991, pp.415 440; W.B.Motherwell, D.Crich, Free
Radicals Chain Reactions in Organic Synthesis, Academic Press,
London, 1992.
[19] G.N.Merrill, G.D.Dahlke, S.R.Kass,
4462 4468.
[20] a) C.P.Andrieux, I Gallardo, J.M.Savÿant, J. Am. Chem. Soc. 1989,
J. Am. Chem. Soc. 1996, 118,
111, 1620 1627; b) D.D.M. Wayner, A. Houman,
Scand. 1998, 52, 377 384.
Acta Chem.
[14] A.M.Cardarelli, M.Fagnoni, M.Mella, A.Albini,
J. Org. Chem.
2001, 66, 7320.
[21] a) H.I. Tasahtcush, R. Sustmann, Chem. Ber. 1992, 125, 287 289;
b) P.A. Baguley, J.C. Walton, J. Chem. Soc. Perkin Trans. 1 1998,
2073 2082.
[15] a) F.Hibbert, F.A.Long, E.A.Walters,
J. Am. Chem. Soc. 1971, 93,
2829 2835; b) T.Yamase, N.Takabayashi, M.Kaji,
Dalton Trans. 1984, 793 799.
[16] A.Maldotti, R.Amadelli, V.Carassiti, A.Molinari,
Acta 1997, 256, 309 312.
[17] a) H.Fischer, J. Am. Chem. Soc. 1986, 108, 3925; b) B.E. Daikh,
R.G. Finkle, J. Am. Chem. Soc. 1992, 114, 2938; c) A.Studer,
Angew. Chem. 2000, 112, 1157 1160; Angew. Chem. Int. Ed. 2000,
39, 1108 1111.
J. Chem. Soc.
[22] D.R.Arnold, R.M.Borg, A.Albini,
J. Chem. Soc. Chem. Commun.
Inorg. Chim.
1981, 138 139.
[23] K.L¸cking, M.Rese, R.Sustmann, Liebigs Ann. Chem. 1995, 1129
1138.
[24] T.Yamase, T.Usami, J. Chem. Soc. Dalton Trans. 1988, 183 190.
[25] M.S¸sse, S.Johne, J. Prakt. Chem. 1981, 323, 647 653.
[26] M.S.Schlegel, G.Schaefer, Chem. Ber. 1984, 117, 1400 1425.
[18] A.L.J.Beckwith, J.S.Poole,
9497.
J. Am. Chem. Soc. 2002, 124, 9489
Received: June 27, 2003 [F5285]
148
¹ 2004 Wiley-VCH Verlag GmbH & Co.KGaA, Weinheim
www.chemeurj.org
Chem. Eur. J. 2004, 10, 142 148