Solvent-dependent changes in the triazolinedione-alkene ene reaction mechanism

Article abstract of DOI:10.1002/chem.200800920

The influence of the solvent on the triazolinedione-alkene ene reaction mechanism has been investigated. Both inter- and intramolecular kinetic isotope effects with tetramethylethylenes and 2,2,2-(trideuterio)methyl-7-methyl-2,6- octadiene-[D₃]-1,1,1 provide, for the first time, strong evidence for changes in the mechanism of the reaction on going from non-protic to polar protic solvents. In non-protic polar or apolar solvents, an aziridinium imide that equilibrates to an insignificant extent with an open intermediate (a dipolar or a polarized biradical) is formed irreversibly in the first, rate-determining step of the reaction, which is followed by fast hydrogen abstraction. On the contrary, in polar protic solvents, hydrogen abstraction is rate limiting, allowing the main dipolar intermediate to equilibrate with its open intermediate(s) as well as with the starting reagents.

Full text of DOI:10.1002/chem.200800920

FULL PAPER
DOI: 10.1002/chem.200800920
Solvent-Dependent Changes in the Triazolinedione–Alkene Ene
Reaction Mechanism
Georgios C. Vougioukalakis, Manolis M. Roubelakis, Mariza N. Alberti, and
Michael Orfanopoulos*[a]
Abstract: The influence of the solvent
on the triazolinedione–alkene ene reac-
tion mechanism has been investigated.
Both inter- and intramolecular kinetic
isotope effects with tetramethylethy-
lenes and 2,2,2-(trideuterio)methyl-7-
polar protic solvents. In non-protic
polar or apolar solvents, an aziridinium
imide that equilibrates to an insignifi-
cant extent with an open intermediate
(a dipolar or a polarized biradical) is
formed irreversibly in the first, rate-de-
termining step of the reaction, which is
followed by fast hydrogen abstraction.
On the contrary, in polar protic sol-
vents, hydrogen abstraction is rate lim-
iting, allowing the main dipolar inter-
mediate to equilibrate with its open in-
termediate(s) as well as with the start-
ing reagents.
Keywords: ene reaction · kinetic
methyl-2,6-octadiene-[D3]-1,1,1
pro-
isotope
effects
·
reaction
vide, for the first time, strong evidence
for changes in the mechanism of the re-
action on going from non-protic to
mechanisms
·
solvent effects
·
triazolinediones
Introduction
of trapping experiments.[11,12] Subsequently, isotope effect
measurements on deuterium-labeled tetramethylethylenes
(TMEs)[3,4] and 2-butenes[5,6] also suggested the formation of
an AI intermediate in the rate-determining step of the reac-
tion. Thus, large intramolecular kinetic isotope effects
(KIEs) were found in the ene reactions of RTADs with the
cis-related methyl and deuteriomethyl groups in substrates 1
and 2 (H/D isotopic competition; Scheme 1), whereas only a
small isotope effect was observed with the trans-related
groups in compound 3 (no isotopic competition). AI inter-
mediates have also been observed spectroscopically in the
reactions of biadamantylidene,[13] trans-cycloheptene,[14] and
trans-cyclooctene[15] with RTADs. Furthermore, because all
of the methyl groups of TME are symmetry equivalent, sim-
ilar isotopic competition would have been expected for sub-
strates 1, 2, and 3 in a concerted mechanism; however, this
was found not to be the case. On this basis, a one-step mech-
anism was excluded.
The ene reaction of triazolinedione[1] (RTAD, R=methyl or
phenyl), one of the most reactive neutral enophiles, with al-
kenes that bear allylic hydrogen atoms, to form N-allylura-
zoles [Eq. (1)] apart from being synthetically useful,[2] has
attracted considerable mechanistic[3–9] and theoretical atten-
tion[10] for many years. A number of mechanisms and key
intermediates (Figure 1) have been proposed for this reac-
tion. Among these, the formation of a closed, three-mem-
bered aziridinium imide (AI) intermediate (Figure 1) was
the most popular, and initially found support in the results
[a] Dr. G. C. Vougioukalakis, Dr. M. M. Roubelakis, Dr. M. N. Alberti,
Prof. Dr. M. Orfanopoulos
Department of Chemistry, University of Crete
Voutes 71003, Heraklion, Crete (Greece)
Fax : (+30)281-054-5001
E-mail: orfanop@chemistry.uoc.gr
Supporting information for this article is available on the WWW
under http://dx.doi.org/10.1002/chem.200800920.
Figure 1. Proposed intermediates for the triazolinedione ene reaction.
Chem. Eur. J. 2008, 14, 9697 – 9705
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
9697

become non-competitive in the corresponding biradicals, as
is observed. The above described biradical mechanism was
subsequently challenged by results derived from other ster-
eoisotopic and product studies on the ene reactions of
RTADs.[8,9,16–18]
More recently, photochemical formation of the two ste-
reoisomeric AI intermediates from the addition of PTAD to
cycloheptene has been reported.[19] These AI intermediates
rearrange to the corresponding ene products through their
dipolar biradical precursors. Since one of the two dipolar
biradical intermediates has the wrong conformation for H
À
abstraction, a 1808 rotation about the C N bond is required
to form the biradical conformer suitably predisposed for H
abstraction. Although in this particular system the proposed
mechanism may well rationalize the obtained results, previ-
ous computational work[10] has shown the activation barrier
À
for rotation about the C N bond in similar biradical inter-
mediates to be energetically unfavorable compared to that
of the H abstraction.
Scheme 1. Reported kinetic isotope effects for the ene reaction of
RTADs with deuterium-labeled tetramethylethylenes in CDCl3.
In the present study, we report for the first time signifi-
cant changes in the triazolinedione ene reaction energy pro-
file on going from non-protic to polar protic solvents. This
missing mechanistic information may complete the puzzle of
this otherwise very well studied reaction that has been a
topic of investigation for many years. In particular, we have
studied the triazolinedione ene reactions of gem-[D6]TME
(1), cis-[D6]TME (3), and 2,2,2-(trideuterio)methyl-7-
methyl-2,6-octadiene-1,1,1-[D3] ([D6]DMOD, 4), as well as
the intermolecular isotope effects in the reaction of
[D0]TME versus [D12]TME (5; Figure 2), in a variety of sol-
Nevertheless, the AI intermediate was challenged by Sin-
gleton and Hang on the basis of experimental and theoreti-
cally predicted KIEs, as well as transition-state energy pro-
files.[10] In their work, an open biradical key intermediate, in
rapid equilibrium with an AI, was proposed (Scheme 2).
Figure 2. [D6]DMOD and [D12]TME.
vents. Our new results triggered our long standing interest
in this reaction to discuss, along with the present results,
previous experimental and theoretical conclusions, and to
propose a “unified mechanism” for this classic ene reaction.
Scheme 2. Proposed biradical mechanism.
Furthermore, its rotation about the initial alkene double
bond was calculated to be restricted. According to this
mechanism, H or D abstraction in biradical intermediates 1H
and 1D (Scheme 2) should not be competitive, and no pri-
mary KIE should be expected in the second, fast step of the
reaction. To rationalize the large primary KIEs previously
found for gem-[D6]TME (1; Scheme 1) and cis-2-[D3]butene,
fast equilibration of the biradical intermediates 1H and 1D
with an AI intermediate was proposed. Moreover, since ro-
Results
The ene reaction of cis-[D6]TME (3) with PTAD was stud-
ied in a variety of solvents. As illustrated in Table 1, the pri-
mary intramolecular isotope effects in acetone and acetoni-
trile (entries 3 and 4) are negligible (kH/kD ꢀ1.1), as is also
the case in dichloromethane and chloroform (entries 1 and
2). On the contrary, unprecedented large intramolecular iso-
tope effects (kH/kD ꢀ2.0–4.0) were measured in EtOH,
MeOH, and MeOH/H2O (entries 6, 7, and 8). This clearly
indicates that solvent properties dictate the PTAD ene reac-
tion mechanism.
À
tation about the C N bond was also calculated and found to
be restricted, the biradical intermediate retains the stereo-
chemical integrity of the AI. Thus, the geminal methyl and
deuteriomethyl groups in cis-[D6]TME (3; Scheme 1)
9698
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 9697 – 9705

Solvent-Dependence of the Ene Reaction Mechanism
FULL PAPER
Table 1. Intramolecular kinetic isotope effects for the ene reaction of cis-
[D6]TME with PTAD.
[a,b]
Entry
Solvent
kH-ene/kD-ene
Dielectric constant
1[c]
2
3
4
5
CH2Cl2
CDCl3
1.08Æ0.1
9.1
4.8
21
37
47
24
33
N/A
1.14Æ0.03
1.13Æ0.03
1.13Æ0.03
1.78Æ0.05
2.10Æ0.06
2.99Æ0.09
3.96Æ0.12
acetone
CH3CN
DMSO
EtOH[e]
MeOH[e]
MeOH/H2O[e]
1
6
7
8[d]
1
[a] Determined by H NMR (500 MHz) spectroscopy. For accurate H in-
tegrations, a spin-lattice relaxation T1 of 5 s was used. [b] All isotope ef-
fects were measured at room temperature at 80% conversion.
[c] Ref. [4]. [d] MeOH/H2O, 4:1. [e] A product ratio of ꢀ60% ene/
ꢀ40% solvent-trapped adduct was formed.
Scheme 3. Trapping and ene adducts for the reaction of PTAD with TME
in MeOH.
Table 2. Intramolecular isotope effects in the reaction of [D6]DMOD
with PTAD.
The large isotope effects measured in protic solvents (kH/
kD ꢀ2.0–4.0) require reversion of the intermediate(s) to the
starting materials and provide, for the first time, evidence
that H(D) abstraction occurs in the rate-determining step of
the reaction (vide infra). The substantial isotope effect ob-
served in the PTAD ene reaction of 3 carried out in DMSO
(entry 5) could be the result of a partial reversion of the in-
termediate(s) to the starting materials. In an earlier study
concerning the addition of MTAD to biadamantylidene
(Ad=Ad) to form the corresponding diazetidine, reversion
to the starting reagents was proposed.[20] This reaction pro-
ceeds through an AI intermediate, which is in equilibrium
with an open intermediate (OI) in a variety of solvents.
However, in this case, the reaction does not lead to ene
products. It also has to be kept in mind that the reaction of
PTAD with TME in MeOH affords two products, the ene
adduct A and the MeOH-trapped adduct B (Scheme 3).
This reaction has been thoroughly studied in earlier work.[21]
In seeking additional evidence for a solvent-dependent al-
teration in the mechanism of the alkene ene reaction of
RTADs, we studied the reaction between PTAD and
[D6]DMOD (4) in CH2Cl2 and MeOH. The reason for pre-
paring alkene 4 (Figure 2) is that, unlike 1, it bears two pairs
of geminal methyl and deuteriomethyl groups isolated from
each other by four carbon atoms. In the case of an irreversi-
ble pathway, these pairs of methyl groups would not be
competitive in the second H or D abstraction step, which
makes it possible to measure an intramolecular isotope
effect that actually simulates the intermolecular one. When
PTAD was added to a stirred solution of [D6]DMOD in
MeOH, besides the ene adducts, the MeOH-trapped adducts
were also isolated. The obtained KIEs are summarized in
Table 2. The insignificant primary isotope effect for the re-
action in CH2Cl2 (Table 2, entries 1 and 2) suggests that no
[a,b]
[a,b]
Entry
Solvent
T [8C]
kH-ene/kD-ene
kH-trap/kD-trap
1
2
3
4
5
6
CH2Cl2
CH2Cl2
MeOH[c]
MeOH[c]
MeOH[c]
MeOH[c]
1
25
0
40
25
0
1.09Æ0.03
1.16Æ0.03
4.7Æ0.1
–
–
1.65Æ0.05
1.70Æ0.05
1.70Æ0.05
1.80Æ0.05
1
6.0Æ0.2
9.1Æ0.3
À35
23.3Æ0.7
[a] Determined by H NMR (500 MHz) spectroscopy. For accurate H in-
tegrations, a spin-lattice relaxation T1 of 5 s was used. [b] Conversion
25%. [c] The percentage ratios of ene:trap in entries 3–6 are 45:55, 37:63,
25:75, and 18:82, respectively.
termining step of this reaction. In sharp contrast, in the case
of methanol as solvent (Table 2, entries 3–6), there is a dra-
matic increase in the isotope effect (e.g., from kH-ene/kD-ene
ꢀ1.10Æ0.03 to kH-ene/kD-ene =6.0Æ0.2 at 258C). Additionally,
as can also be seen in Table 2, the intramolecular primary
isotope effect strongly depends on the reaction temperature.
À
substantial C H bond breaking was involved in the rate-de-
Chem. Eur. J. 2008, 14, 9697 – 9705
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9699

M. Orfanopoulos et al.
For example, kH-ene/kD-ene is 4.7 at 408C, whereas at lower
temperatures (entries 4–6) the kH-ene/kD-ene ratio is abnormal-
ly large, reaching a value of 23.3 at À358C, which is more
than three times the magnitude of that at room temperature.
To a first approximation, the maximum isotope effect, corre-
MeOH attackon an open zwitterionic intermediate. Howev-
er, only the stereospecific adduct 8 was observed.
In this context, the intermolecular isotope effects for the
equimolar competition between [D12]TME and [D0]TME in
dichloromethane and MeOH (ene path) were also mea-
sured. As illustrated in Table 3, the primary intramolecular
À
sponding to the zero-point energy difference for the C H
À
and C D stretches in the transition state, is calculated to be
about 7.4 at room temperature.[22] In the present system,
however, the measured kH-ene/kD-ene values are substantially
larger than 7.4 at low temperatures, and they decrease
sharply with an increase in temperature. These results may
be attributed to extensive hydrogen tunneling along the re-
action coordinate. In our earlier work, we showed that a
similar hydrogen tunneling takes place in the ene reaction
of PTAD with gem-[D6]TME (1).[23]
Table 3. Intermolecular kinetic isotope effects for the reaction of
[D0]TME/[D12]TME with MTAD.
A
To shed more light on the nature of the intermediate in
MeOH, we also studied the b-secondary KIEs in the addi-
tion reaction of MeOH and PTAD to [D6]DMOD (4). The
molar ratio of the two isomeric methanol adducts 7c and 7d
(Table 2, entries 3–6) was found to be proportional to the
[a,b]
Entry
Solvent
T [8C]
kH-ene/kD-ene
1
2
3
4
CH2Cl2
CH2Cl2
MeOH
MeOH
1
25
0
25
0
1.06Æ0.03
1.08Æ0.03
2.25Æ0.07
4.02Æ0.12
1
isotopic ratio kH-trap/kD-trap
.
Moreover, we studied the b-secondary KIEs in the addi-
tion reaction of MeOH and PTAD to gem-[D6]TME (1).
Compared to the b-secondary KIEs measured in the addi-
tion reaction of MeOH and PTAD to [D6]DMOD (4), the
isotope effects for gem-[D6]TME (1) were found to be
slightly smaller (kH-trap/kD-trap =1.60Æ0.05 and 1.64Æ0.05 at
25 and 08C, respectively).
[a] Determined by H NMR (500 MHz) spectroscopy. For accurate H in-
tegrations, a spin-lattice relaxation T1 of 5 s was used. [b] Conversion
25%.
isotope effects in dichloromethane (entries 1 and 2) are
again negligible (kH/kD ꢀ1.1). On the contrary, large intra-
molecular isotope effects (kH/kD ꢀ2.2–4.0) were measured
in MeOH (entries 3 and 4), signifying a change in the
energy profile of the reaction. It should also be noted that
besides the ene product (ꢀ60% at 258C) the known sol-
vent-trapped adduct (ꢀ40% at 258C) was formed in
MeOH.
We also examined the stereochemistry of the MeOH
adduct of PTAD and cis-[D6]TME (Scheme 4). The
Discussion
A mechanism that could account for the observed KIEs
when the PTAD ene reaction of cis-[D6]TME is carried out
in protic or non-protic solvents is presented in Scheme 5. In
aprotic solvents, the irreversible formation of the intermedi-
ates leads to ene products 6a and 6b without isotopic com-
petition. Also, in aprotic solvents, the previously proposed
equilibration of AI with an OI, biradical,[10] or polarized bir-
adical[19] cannot be excluded as a mechanistic possibility. In
this case, the requirement for restricted rotation about the
Scheme 4. Stereospecific addition of MeOH and PTAD to cis-[D6]TME
(3).
À
previous double bond or C N bond of the OI must be
met.[10] However, the extent of AI equilibration with an OI,
as well as the lifetime of the OI, depend on the particular
system and the polarity of the aprotic solvent. Our previous
results[17] concerning the PTAD ene reaction in aprotic sol-
vents suggest a rather “tight” AI intermediate with minimal,
if any, equilibration with an open intermediate. In this
recent work,[17] the vinylcyclopropyl moiety, as in substrate
10, was used as a probe to test the nature of the PTAD–
alkene ene reaction intermediate (Scheme 6). In aprotic sol-
1H NMR spectra of the methanol-trapping reactions of cis-
[D6]TME show only one methyl signal at d=1.26 ppm, at-
tributable to the methyl group adjacent to the oxygen atom
of diastereomer 8. As illustrated in Scheme 4, two methyl
resonances for the methyl group adjacent to the oxygen
atom (i.e., one for the methyl group of 8 and one for that of
9) would have been expected in the case of top and bottom
9700
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 9697 – 9705

Solvent-Dependence of the Ene Reaction Mechanism
FULL PAPER
Figure 3. Proposed energy profiles for the triazolinedione ene reaction in
non-protic (A) and protic solvents (B).
rationalized in terms of a more pronounced charge separa-
tion for the first transition state and a polar/protic solvent
stabilization. As a result, the first transition state (TS1, more
polar) in diagram B is of lower energy than the second one
(TS2, less polar), and so the latter becomes the rate-limiting
step. In the case of the reaction of cis-[D6]TME (3) with
PTAD in DMSO, the lower but still substantial KIE of 1.78
suggests an intermediate energy profile, with its two transi-
tion states TS1 and TS2 being of roughly equal energy.
Scheme 5. Proposed mechanism for the ene reaction of cis-[D6]TME with
PTAD.
Before we continue discussing our present results, it is
useful to mention the key aspects of the MeOH trapping ex-
periments. The A/B product ratio (Scheme 3) depends on
the reaction temperature. For example, at À788C the
MeOH adduct B is essentially the only product, whereas
above the isokinetic temperature, 13.88C, adduct A predom-
inates, a 78:22 ratio of ene/trapping products being obtained
°
°
at 608C. The activation parameters DDHAB and DDSAB
have been calculated as 5.2 kcalmolÀ1 and 17.9 calmolÀ1 KÀ1,
respectively.[21] The enthalpy of activation favors the forma-
tion of the MeOH adduct B, whereas the entropy favors the
formation of the ene product A. These results are rather ex-
pected because of the bimolecularity of the path leading to
the MeOH adduct B through transition state TSB versus the
monomolecular path leading to the ene product A via TSA
(Scheme 3). These earlier results indicate that the large en-
tropy factor DDSAB dictates DDGAB changes in the com-
peting reactions and, consequently, controls the remarkable
variations in the reactivity of the two competing paths. It
was proposed that the common intermediate of the two
competing pathways is the aziridinium imide; however, our
present results could also be rationalized in terms of an azir-
idinium imide in equilibrium with its open intermediate
(zwitterion or polarized biradical).
As regards the reaction between PTAD and [D6]DMOD
(4) in dichloromethane (Table 2, entries 1 and 2), a tight AI
intermediate (in partial equilibrium with an open intermedi-
ate) is most probably formed in the rate-limiting step, and
this is then followed by a second, product-determining step
leading to the ene products without isotopic competition.
The sharp increase in the isotope effect with methanol as
Scheme 6. The vinylcyclopropyl moiety as a mechanistic probe.
vents, this reaction afforded only the ene adduct 11, via the
aziridinium imide as the major intermediate, whereas in
protic solvents a dipolar intermediate was favored, which
was trapped by the cyclopropyl moiety to form the corre-
sponding cyclopropyl-rearranged, solvent-trapped adduct 12.
Had the AI been in equilibrium to a significant extent with
a dipolar or polarized biradical intermediate in non-protic
solvents, a cyclopropyl rearrangement (within the limits of
the rate of phenylcyclopropyl ring-opening, 3”1011 sÀ1)
would have been observed, contrary to the experimental
findings. These observations also support the intermediacy
of an AI in aprotic solvents, whereas in protic solvents an
OI (zwitterion or dipolar biradical) is indicated.
°
°
Proposed energy diagrams for the ene reactions of
RTADs with alkenes in protic and non-protic solvents are
presented in Figure 3. The shift from energy diagram A to
B, on going from non-protic to polar protic solvents, may be
Chem. Eur. J. 2008, 14, 9697 – 9705
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9701

M. Orfanopoulos et al.
the solvent (Table 2, entries 3–6) again suggests reversible
formation of the intermediates, revealing a change in the
energy profile on going from a non-protic (A) to a protic
solvent (B) in Figure 3. The two proposed mechanistic path-
ways, which involve: a) the reversible, in MeOH, formation
of an aziridinium imide in equilibrium with its open inter-
mediate (OIH or OID), and b) the irreversible, in CH2Cl2,
formation of an aziridinium imide through the ene reaction
of [D6]DMOD with PTAD, are presented in Scheme 7.
six deuterium atoms in TS4, are expected to give a normal
and large b-secondary KIE (kH/kD ꢀ1.05–1.10 per deuteri-
um), as is found experimentally. Similar isotope effects have
been reported in the dipolar cycloaddition of tetracyano-
ethylene to 2,4-hexadiene.[24] If the transition states in
Figure 4 were of a tight SN2 type, the measured b-secondary
isotope effects would have been unity or slightly inverted
because of a steric deuterium kinetic isotope effect (see
below).[22,25] In accordance with the above hypothesis, the re-
cently reported thermodynamic parameters DDH° and
DDS° for the reaction of PTAD with 2-methyl-2-butene in
nucleophilic solvents were found to be in favor of an “SN2-
like” transition state.[26] Moreover, the b-secondary KIEs
measured in the MeOH trapping experiments with gem-
[D6]TME again support a loose SN2 dipolar transition state.
As a final note, although the present b-secondary KIEs in
the reactions of gem-[D6]TME and [D6]DMOD can be satis-
factorily rationalized by a loose SN2 dipolar transition state
(Figure 4), an open zwitterionic or a polarized biradical
mechanism cannot be excluded.
It has also been observed that, in contrast to the reaction
of 2-methylindene (13), there is no loss of stereochemistry
in the addition of MeOH and PTAD to 2-butenes (cis and
trans), 1-methylcyclopentene, indene (14), and 2-methyl-2-
butene (15) (Scheme 8).[27] To rationalize the formation of
only one stereoisomer, a rather tight AI intermediate that
does not equilibrate to any significant extent with a zwitter-
ionic intermediate has been proposed. The loss of stereo-
chemical integrity observed in the case of methyl-substituted
indene (13; Scheme 8) was attributed to the fact that this AI
intermediate can yield a highly stable, tertiary benzylic car-
bocation. It was therefore concluded that the stability of the
AI intermediate and its equilibration with the open inter-
mediate depends on the particular system.[27]
Scheme 7. Mechanistic possibilities for the ene reaction of [D6]DMOD
with PTAD in CH2Cl2 and MeOH.
In the MeOH trapping experiments with [D6]DMOD, the
b-secondary KIE (ꢀ1.10 per deuterium) may be attributed
to a rather loose SN2 dipolar transition state (Figure 4; TS3
À
and TS4), which is characterized by a small degree of C O
À
bond making and extensive C N bond breaking. In transi-
tion state TS3, hyperconjugative effects involving the six hy-
drogen atoms of the two methyl groups, as opposed to the
Figure 4. Proposed transition states for the addition reaction of MeOH
and PTAD to [D6]DMOD.
Scheme 8. MeOH and PTAD addition reactions with 2-methylindene
(13), indene (14), and 2-methyl-2-butene (15).
9702
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 9697 – 9705

Solvent-Dependence of the Ene Reaction Mechanism
FULL PAPER
General procedure for the triazolinedione–alkene ene reactions: A fresh-
ly prepared solution of RTAD (0.025 and 0.08 mmol for inter- and intra-
molecular competitions, respectively) in the appropriate solvent (1 mL)
was added to a solution of the alkene (0.1 mmol) in 2 mL of same sol-
vent. Once the reaction was complete (as determined by the disappear-
ance of the characteristic pinkcolor of the RTAD), the solvent and the
residual alkene were removed in a rotary evaporator. The last traces of
solvent were removed under high vacuum and the 1H NMR spectrum of
the products was recorded in CDCl3 solution.
With regard to the MeOH adduct of PTAD and cis-
[D6]TME (Scheme 4), the retention of the stereochemical
integrity at the reaction center reveals the stereospecific
character of the trapping reaction, similar to previous obser-
vations for the addition of MeOH and PTAD to 1-methylcy-
clopentene, 2-methyl-2-butene (15), cis- and trans-2-butenes,
and indene.[27]
Finally, in the intermolecular competition between
[D12]TME and [D0]TME, the absence of a KIE when the re-
action is carried out in CH2Cl2 (Table 3, entries 1 and 2)
demonstrates that once the aziridinium imide or/and open
intermediate is formed, only one kind of isotope (either H
or D, non-competing) is available for abstraction in an irre-
versible mode. In contrast to the hitherto accepted mecha-
nism for the ene reaction of RTADs, the KIEs observed in
MeOH (Table 3, entries 3 and 4) again suggest reversible
generation of the aziridinium imide. The difference in the
measured isotope effects in these two solvents, which are
aprotic and protic, is again consistent with the proposed
energy profiles A and B illustrated in Figure 3.
(Z)-2,3-Dimethyl([D6]-1,1,1,4,4,4)-2-butene (cis-[D6]TME): This deuteri-
um-labeled alkene was prepared according to the procedure reported in
the literature,[28] as detailed below:
(Z)-2,3-Dimethyl-2-butenedioic acid dimethyl ester (dimethyl dimethyl-
maleate): Dimethylmaleic anhydride (4.36 g, 34.4 mmol) was refluxed in
methanol (80 mL) for 90 min. The solution was then cooled to 08C and
titrated with diazomethane in diethyl ether until a yellow color persisted.
The solvents were evaporated and the yellow oil was filtered through
silica gel (12 g) with diethyl ether to yield 5.92 g of dimethyl dimethylma-
leate (34.4 mmol, 100%). 1H NMR (500 MHz, CDCl3, 268C, TMS): d=
1.94 (s, 6H), 3.75 ppm (s, 6H).
(Z)-2,3-Dimethyl([D4]-1,1,4,4)-2-butene-1,4-diol: Aluminum chloride
(1.41 g, 10.6 mmol) was added in small portions to an ice-cooled suspen-
sion of lithium aluminum deuteride (1.34 g, 32 mmol) in absolute diethyl
ether (160 mL). After stirring the deuteride mixture at room temperature
for 30 min, it was cooled to 08C, whereupon a solution of dimethyl dime-
thylmaleate (3 g, 17.5 mmol) in diethyl ether (15 mL) was added drop-
wise. The reaction mixture was stirred for a further 2 h at room tempera-
ture. Acid work-up afforded 1.65 g of the desired alcohol (14 mmol,
80%). 1H NMR (500 MHz, CDCl3, 268C, TMS): d=1.75 ppm (s, 6H);
MS: m/z (%): 138 (4) [M+], 119 (17), 102 (52), 87 (97), 72 (100), 58 (40),
45 (54).
Conclusion
Unlike the results in non-protic solvents, the large, non-ste-
reochemically dependent kinetic isotope effects measured in
the RTAD–alkene ene reactions of TMEs and [D6]DMOD
in polar protic solvents are suggestive of rate-determining
H(D) abstraction. This is a clear reversal of what has hither-
to been known for the otherwise extensively studied
RTADs–alkene ene reaction mechanism. These results have
led us to a “unified” mechanism, whereby the stability of
the aziridinium imide intermediate and its equilibration with
the open intermediate depends on the solvent and the par-
ticular system. In aprotic solvents, a rather tight aziridinium
imide intermediate is envisaged, with insignificant equilibra-
tion to its open intermediate (in this case, retention of con-
figuration), whereas in protic solvents one may envisage a
loose aziridinium imide in extensive equilibrium with its
open intermediate (dipolar or polarized biradical) and the
starting reagents. In this case, large KIEs, independent of
the relative isotope stereochemistry, were recorded.
(Z)-1,4-Dichloro-2,3-dimethyl([D4]-1,1,4,4)-2-butene:
Me2S
(1.06 g,
16.6 mmol) was added dropwise to a solution of N-chlorosuccinimide
(2.24 g, 16.8 mmol) in CH2Cl2 (100 mL) at 08C. Upon stirring for 15 min,
a precipitate was formed. The mixture was cooled to À208C, whereupon
a solution of the above diol (1 g, 8.2 mmol) in CH2Cl2 (40 mL) was added
dropwise. The resulting mixture was kept at 08C for 2 h, and then satu-
rated aqueous NaCl solution (100 mL) and ice (100 g) were added, the
organic phase was separated, and the aqueous layer was extracted with
Et2O. The organic layers were washed separately with H2O and stripped
of solvent to yield a total of 1.04 g of the desired dichloride, which was
used in the next step without further purification. 1H NMR (500 MHz,
CDCl3, 268C, TMS): d=1.83 ppm (s, 6H); MS: m/z (%): 158 (24) [M+],
156 (39), 123 (31), 121 (100), 107 (25), 105 (80), 91 (23), 85 (38), 69 (47),
55 (26).
(Z)-2,3-Dimethyl([D6]-1,1,1,4,4,4)-2-butene (cis-[D6]TME):
A Schlenk
flask, connected to a rotaflo trap that was cooled to À788C, was charged
with lithium aluminum deuteride (290 mg, 7 mmol) in dry triglyme
(12 mL) under Ar. The flaskwas cooled to À258C, whereupon a solution
of the above dichloride (1 g, 6.6 mmol) in dry triglyme (5 mL) was added
over a period of 10 min by means of a syringe. The mixture was stirred at
À258C for 1 h and at room temperature for a further 2 h. With the help
of a slow stream of Ar, and by heating the reaction mixture to 1108C,
the alkene was collected in the trap together with some solvent. This mix-
ture was further purified by preparative GC to afford 286 mg (50%) of
cis-[D6]TME. 1H NMR (500 MHz, CDCl3, 268C, TMS): d=1.64 ppm (s,
6H); MS: m/z (%): 90 (79) [M+], 75 (84), 72 (100), 59 (19).
Experimental Section
General considerations: 1H and 13C NMR spectra were recorded on a
500 MHz (125 MHz for 13C) spectrometer from samples in CDCl3 solu-
tions. Chemical shifts are reported in ppm downfield from Me4Si by
using the residual solvent peakas an internal standard. Isomeric purities
were determined by 1H NMR and by GC on a 50 m HP-5 capillary
column connected to a 5971A MS detector. Vapor-phase chromatograph-
ic separations were performed on a GOW MAC 55 chromatograph with
a thermal conductivity detector. TLC was carried out on SiO2 (silica gel
F254). Chromatography refers to flash chromatography and was carried
out on SiO2 (silica gel 60, SDS, 230–400 mesh ASTM). During work-up
of the reaction mixtures, organic extracts were dried over anhydrous
MgSO4. Solvents were evaporated in a rotary evaporator.
2,2,2-(Trideuterio)methyl-3-methyl-2-butene-[D3]-1,1,1 (gem-[D6]TME):
This deuterium-labeled alkene was prepared according to the procedure
reported in the literature,[29] as detailed below:
2,2-Dimethyl-3-hydroxy-3-(trideuterio)methyl-butyric acid-[D3]-4,4,4:
A
flame-dried, 500 mL three-necked round-bottomed flask, equipped with
a magnetic stirrer, a reflux condenser, and an addition funnel, was
charged, under N2, with a solution of dry diisopropylamine (14 mL,
100 mmol) in dry THF (100 mL). After cooling the solution to À788C,
nBuLi (1.6m in n-hexane, 62.5 mL, 100 mmol) was added dropwise. The
mixture was left for 1 h at room temperature and then cooled to À788C
once more. Next, isobutyric acid (4.41 g, 50 mmol, 1m solution in dry
Chem. Eur. J. 2008, 14, 9697 – 9705
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9703

M. Orfanopoulos et al.
THF) was added dropwise. The resulting mixture was left for 1 h at room
temperature and then cooled to 08C, whereupon [D6]acetone (3.7 mL,
50 mmol) was added as a 2.5m solution in dry THF. After stirring at
room temperature for 12 h, the reaction mixture was poured onto ice and
transferred to a separatory funnel. After several extractions with Et2O,
the aqueous layer was acidified with 6 n HCl and extracted with Et2O
(5”50 mL). The combined extracts were dried and the solvent was
evaporated to afford the b-hydroxy acid (6.5 g, 85%), which was used in
the next step without further purification. 1H NMR (500 MHz, CDCl3,
268C, TMS): d=6.01 (brs, 2H; COOH + OH), 1.26 ppm (s, 6H).
the flaskwas heated over a water bath until no further reaction was evi-
dent (about 3 h). H2O (4 mL) was then added through the addition
funnel; the reaction mixture was heated for a further 1 h, cooled to about
508C, and filtered. The solid was returned to the flaskand heated with
fresh benzene (10 mL) to dissolve any remaining pinacol. The combined
filtrates were then concentrated to half of the original volume in order to
remove the acetone; the remaining benzene solution was treated with
H2O (6 mL) and cooled to 10–158C. Pinacol hydrate was precipitated,
which was collected by filtration and washed with benzene (6.5 g, 43%
based on the magnesium used). The pinacol hydrate was then dehydrated
and distilled to anhydrous pinacol (2.5 g). MS: m/z (%): 130 (0.1) [M+],
112 (2), 94 (5), 65 (100), 46 (26), 33 (10).
3,3-Di(trideuterio)methyl-4,4-dimethyl-b-lactone:
A
single-necked
500 mL round-bottomed flask, equipped with a magnetic stirrer, was
charged with the above b-hydroxy acid (1.52 g, 10 mmol) in dry pyridine
(60 mL). The solution was cooled to 0–58C, p-toluenesulfonyl chloride
(3.8 g, 20 mmol) was added, and the resulting mixture was stirred for
10 min. The flaskwas then sealed and left in a freezer for 12 h. There-
after, the reaction mixture was poured onto crushed ice (four to five
times greater in volume) and extracted with Et2O (5”50 mL). The com-
bined extracts were washed with saturated aqueous NaHCO3 solution
and H2O, dried, and stripped of solvent to afford the b-lactone (0.7 g,
2,3-Dimethyl-2-[D12]butene (5, [D12]TME): A Schlenkflask, which was
connected to
a
rotaflo trap cooled to À788C, was charged with
[D12]pinacol (2.2 g, 16.95 mmol) and ethyl orthoformate (2.52 g). The
flaskwas heated from 125 8C to 1408C over a period of 8 h, during which
time ethanol (1.9 mL) was distilled off. The remaining colorless liquid (2-
ethoxy-4,4,5,5-tetramethyl-1,3-[D12]dioxolan) was heated at 150–1608C
for 10 h, during which time CO2 was evolved and 2.2 mL of distillate was
collected. This distillate, which consisted mainly of [D12]TME and etha-
1
nol, was further purified by preparative GC to afford
6.15 mmol). 13C NMR (125 MHz, CDCl3, 268C, TMS): d=123.4,
19.4 ppm (septet, 1J(C,D)=19 Hz); MS: m/z (%): 96 (40) [M+], 78 (86),
5 (800 mg,
52%). H NMR (500 MHz, CDCl3, 268C, TMS): d=1.30 ppm (s, 6H).
2,2,2-(Trideuterio)methyl-3-methyl-2-butene-[D3]-1,1,1 (gem-[D6]TME):
The above b-lactone (0.5 g, 3.72 mmol) was placed in a Schlenkflask,
which was connected to a rotaflo trap cooled to À788C. The flaskwas
heated at 1608C, which resulted in decomposition of the b-lactone to the
deuterated alkene and CO2. With the help of a slow stream of N2, the
alkene (0.26 g, 77%) was collected in the trap. 1H NMR (500 MHz,
AHCTREUNG
62 (17), 46 (100), 42 (27).
2,3-Dimethylbut-2-[D0]ene ([D0]TME): This compound was purchased
from Aldrich. 1H NMR (500 MHz, CDCl3, 268C, TMS): d=1.66 ppm (s,
12H); 13C NMR (125 MHz, CDCl3, 268C, TMS): d=123.4, 20.3 ppm.
CDCl3, 268C, TMS): d=1.66 ppm (s, 6H); 13C NMR (125 MHz, CDCl3,
1
268C, TMS): d=123.4, 123.2, 20.3, 19.4 ppm (septet, J
A
Methyltriphenylphosphonium iodide: A flame-dried, 100 mL round-bot-
tomed flask, equipped with a magnetic stirrer, was charged with iodome-
thane (1.24 mL, 20 mmol), triphenylphosphine (3.93 g, 15 mmol), and tol-
uene (50 mL). The solution was stirred at room temperature for 12 h and
the white solid product was washed with hot toluene to afford 5.76 g of
methyltriphenylphosphonium iodide (14.25 mmol, 95%). 1H NMR
Acknowledgements
M.O. thanks Professor R. H. Grubbs for his generous hospitality during
his sabbatical stay at Caltech (2006). The financial support of the Greek
Secretariat of Research and Technology (IRAKLITOS 2002 and PITHA-
GORAS II, 2005) is also acknowledged.
(500 MHz, CDCl3, 268C, TMS): d=3.23 (d, 3H, 3J
(H,H)=8 Hz),
N
7.75 ppm (m, 15H).
2,2,2-(Trideuterio)methyl-7-methyl-2,6-octadiene-[D3]-1,1,1
([D6]DMOD): A flame-dried, 100 mL round-bottomed flask, equipped
with a magnetic stirrer, was charged with a solution of methyltriphenyl-
phosphonium iodide (3.32 g, 8.2 mmol) in dry THF (20 mL) under Ar.
The solution was cooled to 08C, and then 1.6m nBuLi (5.15 mL, solution
in hexanes) was added dropwise. The orange solution was left at room
temperature for 30 min, cooled to 08C, and then a solution of 5-bromo-2-
methyl-2-pentene (1.1 mL, 8.2 mmol) in THF (5 mL) was added drop-
wise. The reaction mixture was stirred at room temperature for 1 h,
cooled to 08C, and then 1.6m nBuLi solution (5.15 mL) was added drop-
wise. The resulting mixture was stirred at room temperature for 15 min,
cooled to 08C, and then [D6]acetone (3 mL, 40 mmol) was added. The so-
lution obtained was concentrated to a volume of 10 mL, pentane (40 mL)
was added, and the mixture was stirred for 30 min. Filtration, evaporation
of the solvents, and purification of the residue by preparative GC afford-
ed 472 mg of the desired 4 (3.28 mmol, 40%). 1H NMR (500 MHz,
CDCl3, 268C, TMS): d=5.12 (t, 2H), 2.73 (t, 4H), 2.02 ppm (s, 6H);
13C NMR (125 MHz, CDCl3, 268C, TMS): d=131.51, 131.34, 124.46,
[1] For a recent review article regarding the triazolinedione ene reac-
tion, see: G. C. Vougioukalakis, M. Orfanopoulos, Synlett 2005, 713–
731.
[2] For some representative examples, see: a) W. H. Pirkle, J. C. Stick-
ler, J. Chem. Soc. Chem. Commun. 1967, 760–761; b) E. J. Corey,
B. B. Snider, J. Org. Chem. 1973, 38, 3632–3633; c) R. Breslow, R. J.
Corcoran, B. B. Snider, R. J. Doll, P. L. Khanna, R. Kaleya, J. Am.
Chem. Soc. 1977, 99, 905–915; d) W. Adam, A. Pastor, T. Wirth,
Org. Lett. 2000, 2, 1295–1297; e) W. Adam, H. G. Degen, O. Krebs,
C. R. Saha-Mçller, J. Am. Chem. Soc. 2002, 124, 12938–12939;
f) P. S. Baran, C. A. Guerrero, E. J. Corey, J. Am. Chem. Soc. 2003,
125, 5628–5629; g) P. S. Baran, C. A. Guerrero, E. J. Corey, Org.
Lett. 2003, 5, 1999–2001.
[3] A. Seymour, F. D. Greene, J. Am. Chem. Soc. 1980, 102, 6384–6385.
[4] C. C. Cheng, C. A. Seymour, M. A. Petti, F. D. Greene, J. F. J.
Blount, J. Org. Chem. 1984, 49, 2910–2916.
[5] M. Orfanopoulos, C. S. Foote, I. Smonou, Tetrahedron Lett. 1987, 28,
15–18.
[6] M. Orfanopoulos, I. Smonou, C. S. Foote, J. Am. Chem. Soc. 1990,
112, 3607–3614.
[7] J. S. Chen, K. N. Houk, C. S. Foote, J. Am. Chem. Soc. 1997, 119,
9852–9855.
[8] G. Vassilikogiannakis, M. Stratakis, M. Orfanopoulos, Org. Lett.
2000, 2, 2245–2248.
[9] M. N. Alberti, G. C. Vougioukalakis, M. Orfanopoulos, Tetrahedron
Lett. 2003, 44, 903–905.
28.39, 28.36, 25.72, 24.80 (septet, 1J
(C,D)=19 Hz), 17.70, 16.80 ppm
(septet, 1J(C,D)=19 Hz); MS: m/z (%): 144 (13) [M+], 126 (2), 101 (5),
ACHTREUNG
88 (10), 75 (84), 69 (100), 45 (24).
2,3-Dimethyl-2-[D12]butene ([D12]TME): This deuterium-labeled alkene
was prepared according to the procedure reported in the literature,[30] as
detailed below:
2,3-Dimethyl-2,3-[D12]butanediol ([D12]pinacol): A flame-dried 100 mL
round-bottomed flask, equipped with a magnetic stirrer, a reflux con-
denser, and an addition funnel, was charged with magnesium turnings
(1.6 g, 65.8 mmol) and dry benzene (16 mL). A solution of mercury(II)
chloride (1.8 g) in [D6]acetone (10 mL, 138 mmol) was added gradually
through the addition funnel. When the first vigorous reaction subsided, a
mixture of [D6]acetone (5.2 mL) and dry benzene (4 mL) was added, and
[10] D. A. Singleton, C. Hang, J. Am. Chem. Soc. 1999, 121, 11885–
11893.
[11] E. K. Von Gustorf, D. V. White, B. Kim, D. Hess, J. Leitich, J. Org.
Chem. 1970, 35, 1155–1165.
9704
www.chemeurj.org
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Chem. Eur. J. 2008, 14, 9697 – 9705

Solvent-Dependence of the Ene Reaction Mechanism
FULL PAPER
[12] S. R. Turner, L. J. Guilbault, G. B. Butler, J. Org. Chem. 1971, 36,
2838–2840.
[23] Y. Elemes, M. Orfanopoulos, J. Am. Chem. Soc. 1992, 114, 11007–
11009; see also the Supporting Information.
[13] S. F. Nelsen, D. L. Kapp, J. Am. Chem. Soc. 1985, 107, 5548–5549.
[14] M. Squillacote, M. Mooney, J. De Felipis, J. Am. Chem. Soc. 1990,
112, 5364–5365.
[15] T. H. W. Poon, S. Park, Y. Elemes, C. S. Foote, J. Am. Chem. Soc.
1995, 117, 10468–10473.
[16] M. Stratakis, M. Hatzimarinaki, G. E. Froudakis, M. Orfanopoulos,
J. Org. Chem. 2001, 66, 3682–3687.
[17] M. M. Roubelakis, G. C. Vougioukalakis, Y. S. Angelis, M. Orfano-
poulos, Org. Lett. 2006, 8, 39–42.
[24] G. Vassilikogiannakis, M. Orfanopoulos, Tetrahedron Lett. 1996, 37,
3075–3078.
[25] L. S. Bartell, J. Am. Chem. Soc. 1961, 83, 3567–3571.
[26] Z. Syrgiannis, Y. Elemes, Tetrahedron Lett. 2006, 47, 2961–2964.
[27] I. Smonou, S. Khan, C. S. Foote, Y. Elemes, I. M. Mavridis, A. Panti-
dou, M. Orfanopoulos, J. Am. Chem. Soc. 1995, 117, 7081–7087.
[28] a) M. B. Grdina, M. Orfanopoulos, L. M. Stephenson, J. Org. Chem.
1979, 44, 2936–2938; b) J. Y. Roberge, P. Deslongchamps, Synth.
Commun. 1989, 19, 817–827.
[18] G. Vassilikogiannakis, Y. Elemes, M. Orfanopoulos, J. Am. Chem.
Soc. 2000, 122, 9540–9541.
[19] M. E. Squillacote, C. Garner, L. Oliver, M. Mooney, Y. L. Lai, Org.
Lett. 2007, 9, 5405–5408.
[20] S. F. Nelsen, S. J. Klein, J. Phys. Org. Chem. 1997, 10, 456–460.
[21] Y. Elemes, M. Orfanopoulos, Tetrahedron Lett. 1991, 32, 2667–2670.
[22] a) L. Melander, W. H. Saunders, Reaction Rates of Isotopic Mole-
cules, Wiley-Interscience, New York, 1980; b) B. K. Carpenter, De-
termination of Organic Reaction Mechanism, Wiley, New York,
1984.
[29] a) W. Adam, C. I. Rojas, Synthesis 1972, 616–618; b) W. Adam, J.
Baeza, J.-C. Liu, J. Am. Chem. Soc. 1972, 94, 2000–2006.
[30] a) G. Crank, F. W. Eastwood, Aust. J. Chem. 1964, 17, 1392–1398;
b) R. Adams, E. W. Adams, H. T. Clarke, R. Phillips, in Org. Synth.
Coll. Vol. I (Ed.: H. Gilman), Wiley, New York, 1932, pp. 448–450;
c) G. W. Ayers, J. Am. Chem. Soc. 1938, 60, 2957–2960.
Received: May 14, 2008
Published online: September 24, 2008
Chem. Eur. J. 2008, 14, 9697 – 9705
ꢀ 2008 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.chemeurj.org
9705