On the reactivity of naphthalene and biphenyl dianions: Tying up loose ends concerning an SN2-ET dichotomy in alkylation reactions

Article abstract of DOI:10.1002/poc.3423

Naphthalene and biphenyl dianions are interesting compounds that can be obtained by double reduction of the corresponding arenes in solution with certain alkali metals. These dianions are highly reactive and rather elusive species with very high laying and highly delocalized electrons. They share many aspects of the reactivity of the alkali metal they originated from and consequently behave primarily as strong electron transfer (ET) reagents. We report here kinetic evidence for a different type of reactivity in their alkylation reactions with alkyl fluorides. By using cyclopropylmethyl fluoride (c-C₃H₅CH₂F) as a very fast radical probe, we were able to settle that this alkylation does not involve the classical electron transfer reaction followed by radical coupling between diffusing radicals, but supports the alternative S_N2 concerted mechanism, discerning thus this mechanistic S_N2-ET dichotomy.

Full text of DOI:10.1002/poc.3423

Research Article
Received: 7 October 2014,
Revised: 14 December 2014,
Accepted: 18 January 2015,
Published online in Wiley Online Library
(wileyonlinelibrary.com) DOI: 10.1002/poc.3423
On the reactivity of naphthalene and biphenyl
dianions: tying up loose ends concerning an
S 2-ET dichotomy in alkylation reactions
N
a
a
a
Inmaculada Blasco , Henoc Pérez and Albert Guijarro *
Naphthalene and biphenyl dianions are interesting compounds that can be obtained by double reduction of the correspond-
ing arenes in solution with certain alkali metals. These dianions are highly reactive and rather elusive species with very high
laying and highly delocalized electrons. They share many aspects of the reactivity of the alkali metal they originated from
and consequently behave primarily as strong electron transfer (ET) reagents. We report here kinetic evidence for a different
type of reactivity in their alkylation reactions with alkyl fluorides. By using cyclopropylmethyl fluoride (c-C H CH F) as a very
3
5
2
fast radical probe, we were able to settle that this alkylation does not involve the classical electron transfer reaction followed
by radical coupling between diffusing radicals, but supports the alternative S 2 concerted mechanism, discerning thus this
N
mechanistic S 2-ET dichotomy. Copyright © 2015 John Wiley & Sons, Ltd.
N
Keywords: biphenyl dianion; cyclopropylmethyl fluoride; naphthalene dianion; S 2/ET dichotomy; very fast radical probe
N
powerful electron transfer (ET) reagents. In spite of that, new pat-
terns of reactivity (other than ET) that are ascribed solely to the
INTRODUCTION
dianion (and not to the radical anion) were recently identified.
Under certain conditions, lithium naphthalene (LiC10
well as lithium biphenyl (LiC12 10, LiBp) can be over-reduced with
Li(s) to afford even more reduced species, i.e. the corresponding
8
H , LiNp) as
Lithium dianions of naphthalene and biphenyl have displayed un-
expected nucleophilic behavior versus simple alkenes, affording
H
[
8]
[
1,2]
carbolithiation products with propene, isobutene, etc. Also, in
its reaction with n-, s- and t-alkyl fluorides, competitive kinetics re-
vealed for dilithium naphthalene a reactivity profile consistent
2 10 8 2 2 12 2
dianions Li C H (Li Np) and Li C H10 (Li Bp), respectively.
There are many fundamental aspects that remain to be explored
in these compounds, so we became involved some time ago in
the study of the structure and reactivity of these species. The struc-
[9]
with an S (nucleophilic substitution) reaction pathway. Alkylated
N
dihydronaphthalenes were apparently obtained in these reactions
2 10 8
ture of lithium naphthalene dianion Li C H obtained in this way
through a mechanism (S ) different from the classical mechanistic
N
was elucidated only recently, revealing an ionic contact triple (tri-
[10]
[
3]
view that involves ET and coupling between diffusing radicals
a classical mechanism valid otherwise for the rest of alkyl halides
RCl, RBr and RI). In this work, we use cyclopropylmethyl fluoride
,
ple meaning three ions) in the ground state. In the case of biphe-
nyl, the experimental crystal structure is yet unknown, but ab initio
and DFT calculations foresee a similar type of structure and bond-
(
[4]
(1a) as a fast radical probe to shed light on this S /ET mechanistic
N
ing, i.e. an ionic contact triple in the ground state (Fig. 1). In spite
of a profound change in the electron configuration of the hydro-
carbon during its reduction, going from an open-shell radical an-
ion, to a closed-shell configuration in the dianion, these changes
occur while preserving the structural identity of the arene. This
provides an opportunity to study the reactivity of these species
from a unique perspective. Both in the radical anion and dianion,
the vacant π-LUMO of the parent hydrocarbon has been occupied
by one or two extra electrons coming from the lithium metal. The
resulting anions have both very high-lying and highly delocalized
electrons. This dictates much of their reactivity, which is often rem-
iniscent of the alkali metal they originated from. Still, over-
reduction to the dianion is not a trivial issue for these small arenes.
dichotomy, and also raise attention to the fact that this issue
was not settled beyond doubt in the past by using slower
radical probes such as 5-hexenyl fluoride (1b).
RESULTS AND DISCUSSION
Cyclopropylmethyl fluoride as a very fast radical probe
The synthesis of cyclopropylmethyl fluoride (1a) with an ade-
quate level of purity, in particular, free from rearranged isomers,
The second reduction potentials (E° ) of naphthalene and biphe-
2
+
[5]
* Correspondence to: Albert Guijarro, Departamento de Química Orgánica and
Instituto de Síntesis Orgánica. Universidad de Alicante. Campus de Sant Vicent
del Raspeig, Apdo. 99, E-03080 Alicante, Spain.
nyl rival with the Li /Li pair, leaving only a narrow window
(s)
for a feasible (exergonic) double reduction of these hydrocar-
[6]
bons. Indeed, the double reduction of naphthalene with Li(s) in
THF is expected to be unfeasible (ΔG° > 0) if formulated solely
on electrochemical grounds, which does not take into account
the lithium ion pairing stabilization, and it is also so for biphenyl
E-mail: aguijarro@ua.es
a I. Blasco, H. Pérez, A. Guijarro
Departamento de Química Orgánica and Instituto de Síntesis Orgánica,
Universidad de Alicante, Campus de Sant Vicent del Raspeig, Apdo. 99, E-
03080, Alicante, Spain
[
7]
under the same premises. As a result, these compounds are
J. Phys. Org. Chem. 2015
Copyright © 2015 John Wiley & Sons, Ltd.

I. BLASCO, H. PÉREZ AND A. GUIJARRO
cyclopropylmethyl derivative initially confronted us with an
additional problem: the easy cationic isomerization of this
structure. It occurs via the cyclopropylmethyl carbocation triad,
which includes the cyclobutyl and 3-butenyl cations. Details of this
cationic rearrangement can be found in the supplementary
information (Scheme S1 and S2). After a scrutiny of several alter-
natives, cyclopropylmethyl tosylate (2a) was chosen as the
starting material to obtain 1a. 2a has been described as a liquid
[
19]
too unstable to purify by distillation.
It has been obtained in
[
20]
[21]
variable yields (when reported) ranging from 66%, 85–93%,
8.5%. It is often overlooked in the synthetic literature that 2a
[
22]
6
is a rather unstable compound that should be carefully manipu-
lated to avoid the above mentioned skeletal rearrangements as
well as further decomposition. To prevent this isomerization, we
have successfully optimized the synthesis of 2a using low tem-
r
perature, and solvents of low ionizing power (ε < 10) during the
synthetic process. Obtained in this way, 2a is a low melting point
white solid (mp = 16–17 °C, uncorrected). As a nucleophilic
fluoride source, tetrabutylammonium fluoride (TBAF) turned out
to be the most indicated for avoiding unwanted structural
rearrangements. Thus, reaction with TBAF (1M in THF) at 5 °C,
and collection of the volatiles in a cold trap afforded a ca. 25%
of cyclopropylmethyl fluoride (1a), which could be easily purified
to >95% pure product, suitable for our purposes (Scheme 1 and
experimental).
Figure 1. Dilithium naphthalene (Li
2
Np) from X-ray crystallography (co-
[3]
ordinating TMEDA omitted for simplicity) ; and dilithium biphenyl
[4]
2
(Li Bp) from DFT calculations
was necessary for our purposes. It proved to be not an easy task.
Cyclopropylmethyl fluoride is a reagent conspicuously missing
from the chemical literature, except for a few instances. In
general, an adequate description of its synthesis and isolation
Antecedents with other radical probes
[
11]
is lacking while many details are omitted, as noticed by Percy.
In previous studies we examined the reaction of LiBp and Li
with a radical probe (hex-5-enyl fluoride, 1b) in THF and THP at
°C, paying special attention to the distribution of products ob-
2
Bp
We started an extensive search for synthetic methodologies to
obtain pure cyclopropylmethyl fluoride. In general, nucleophilic
fluorination of cyclopropylmethanol derivatives proved to be
the method of choice, while other approaches such as
cyclopropanation of allyl fluoride (CH = CHCH F) failed to give
0
[2]
served. It was significant that, among the alkylated products
>80% overall yield for Li Bp), no rearranged products were ob-
(
2
2
2
served (see Schemes 2 and 3 and the structures of R· and
rearranged R′· below). The concentration of 4b′ remained zero
de desired compound, affording only homoallyl fluoride
CH = CHCH CH F). There are two main synthetic procedures
(
2
2
2
(
i.e. undetectable by GLC, contrasted using an authentic synthe-
sized sample of 4b′) in experiments of dilution ([Li Bp] = 0.1, 0.05
and 0.01 M). The same conclusions were reached for the naph-
described in the literature. First, the synthesis of 1a was de-
scribed by reaction of cyclopropylmethanol with DAST (Et NSF
in diglyme at ꢀ50 °C and low temperature-low pressure fraction-
2
2
3
)
[9]
thalene dianion Li
2
Np. In its reaction with the same radical
Np afforded a 76% overall yield of
[12]
ation column.
There is no evidence of the purity of the
probe (1b) in THP at 0 °C, Li
2
distilled fraction, nor were yields reported. In our hands, the
reaction product always turned out to be an inseparable mixture
of two isomers of cationic rearrangement (cyclobutyl fluoride as
mayor component and cyclopropylmethyl fluoride in a 1.6:1
ratio). Our attempts to improve these results modifying the reac-
tion conditions (T, solvent) and equipment, as well as using alter-
native fluorinating reagents such as Yarovenko’s and Ishikawa’s
alkylation products, from which 0% were rearranged (also con-
firmed by GLC after synthesis of the rearranged products), and
2
remained so in experiments of dilution ([Li Np] = 0.1, 0.05 and
ꢀ2
0
.01 M). Preparation of solutions below 10 M was not consid-
ered reliable enough by the method of direct reduction with Li(s)
lithium powder obtained by milling of a lithium bar), so no fur-
(
ther experiments at higher dilution were considered. These facts
led us to think that the alkylation reaction of Li Bp or Li Np with
[
13]
reagents, were unsuccessful. A second published approach
involves the nucleophilic substitution of cyclopropylmethyl
2
2
primary fluororalkanes was likely to proceed through an SN2
[14]
+
ꢀ
tosylate (2a),
2 4 2
with TBAF·3H O (Bu N F ·3H O), which had
been previously partially dehydrated by a thermal treatment un-
[15]
der vacuum.
According to the authors, the volatiles of the
reaction were then vacuum transferred into an empty ampule,
giving two layers, an aqueous and an organic one, this last one
consisting of ~50% of cyclopropylmethyl fluoride along with
[
13]
olefins and other unidentified products.
Then that crude or-
ganic mixture was further purified (to >95%) by treating it with
Cp* ZrH , but no spectroscopic evidence of purity was either re-
2
2
16]
[
ported.
Finally, two recent articles with syntheses inspired in
[
17]
this paper have also been published, including a detailed ex-
perimental section in the most recent one, although no evidence
[
18]
of its isomeric purity.
The synthesis of cyclopropylmethyl
fluoride by nucleophilic fluorination of an appropriate
Scheme 1. Synthesis of cyclopropylmethyl fluoride (1a)
wileyonlinelibrary.com/journal/poc
Copyright © 2015 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2015

THE DIANIONS OF NAPHTHALENE AND BIPHENYL ARE AMONG THE STRONGEST ET REAGENTS AVAILABLE
not by nucleophilic substitution (i.e. we set kSN = 0). Should this
scenario be impossible, the alternative scenario, i.e. alkylation
of the arene taking place by nucleophilic substitution, is very
likely to be true (reductio ad absurdum), since alternative mech-
anisms of alkylation seem unlikely. From the rate laws derived
from Scheme 4 (Eqns 1 and 2), Eqn 3 is obtained. Applying the
steady state approximation (SSA) to a short lived intermediate
(equation 4), the ratio between radicals [R·]/[R′·] is obtained,
which substituted in equation 3 affords the ratio between
non-rearranged (RH + RAH) to rearranged products (R′H + R′AH)
2
Scheme 2. Reaction of Li Np with RF (R = radical probe, see structure of
R below)
(Eqn 5). The ratio expressed in Eqn 5 is, as expected, solely de-
pendent on the ratio of reaction rates immediately following
[
23,24]
the generation of the primary radical R· by dissociative ET.
Since we are only interested in the coupling products, we math-
[
25]
ematically split Eqn 5 in its two components,
which are the
reduction step (Eqn 6) and the coupling step (Eqn 7), and focus
in the last one.
d½RH þ RAHꢁ
À
À
Á
¼
¼
kred þ kcoupl ½LixAꢁ½R·ꢁ
(1)
(2)
(3)
dt
d½R’H þ R’AHꢁ
Á
kred þ kcoupl ½LixAꢁ½R’·ꢁ
dt
d½RH þ RAHꢁ
½R·ꢁ
hence;
¼
d½R’H þ R’AHꢁ ½R’·ꢁ
2
Scheme 3. Reaction of Li Bp with RF (R = radical probe, see structure of
R below)
and from the SSA,
d½R’·ꢁ
À
Á
transition state, although being rigorous the potential role of rad-
icals in the reaction could not be discarded.
¼ k ½R·ꢁ ꢀ k þ kcoupl ½Li Aꢁ½R’·ꢁ ¼ 0
(4)
(5)
r
red
x
dt
therefore,
½
RH þ RAHꢁ
kredþkcoupl
kr
¼
½LixAꢁ
½
R’H þ R’AHꢁ
which splits into
½
RHꢁ
kredþkcoupl
kr
¼
½LixAꢁ; and
(6)
(7)
Structure of R (unrrearranged) and R′, R″ (rearranged).
½
R’Hꢁ
Let us evaluate now quantitatively the reason for this last
statement.
½
RAHꢁ
kredþkcoupl
¼
½LixAꢁ
½
R’AHꢁ
kr
The kinetic scheme of the reaction
The distribution of the reduction products shown in Eqn 6 was
already studied in the past and is consistent with two consecu-
The complete set of reactions has been represented in Scheme 4.
We shall consider first the hypothetical scenario in which alkyl-
ation of the arene takes place only by radical coupling, and
[
10]
tive ET steps.
The rate constants of reduction of the primary
hex-5-enyl radical (R·) with NaNp and NaBp were estimated as
9
–1 –1
9
–1 –1
k
red (NaNp) = 10 M
s
and kred (NaBp) = 1.6 10 M
s
in
[
10]
DME at 25 °C.
The rate constants for the radical coupling
(kcoupl) were not directly evaluated in that work, which was fo-
[
26]
cused only on reduction products.
of reduction (RH + R′H) was ca. 50 % (±10%),
0% being described as mixtures of alkylated dihydroarenes
RAH + R′AH). The coupling of the radical anions with R· and R′·
But the combined yields
[
10]
the remaining
5
(
to eventually afford alkylation products RAH and R′AH is in direct
competition with the reduction to carbanions and eventual for-
mation of RH and R′H. Since both types of compounds (reduction
and alkylation) are observed in the reaction crudes, the corre-
sponding kcoupl must be of the same order as kred; therefore
the rate of radical–radical anion coupling is straightforward,
Scheme 4. Overall reaction network, including both S
reaction pathways, as well as the rest of events following ET. The ET reagent
has been represented as Li A, with A = Np, Bp, and 1 ≤ x ≤ 2, indicating an
undetermined mixture of radical anion and dianion. R = non rearranged
framework, R′ = rearranged framework. The corresponding hydrocarbon
N
(kSN) and ET (kET)
x
9
–1 –1 [27]
kcoupl ≅ 10 M s . We have to point out here that these rate
products are generated in a final protonation step (SH = H
2
O or solvent).
constants have not been explicitly determined for the lithium
J. Phys. Org. Chem. 2015
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

I. BLASCO, H. PÉREZ AND A. GUIJARRO
Table 1. Calculated and experimental ratio of non-rearranged (RAH) to rearranged (R′AH) alkylation products under different Li
concentrations
x
A
a
a
b
½RAHꢁ
R’AHꢁ
c
½RAHꢁ
½R’AHꢁ
d
½RAHꢁ
½R’AHꢁ
e
f
RF
[Li A] (M)
Compounds
x
½
calc:
diffus:
expt:
0
0
0
.1 M Li Np
2000:1
1000:1
200:1
7200:1
3600:1
720:1
3b/3b′ > 100:1
>100:1
>100:1
x
.05 M Li Np
x
.01 M Li Np
x
0
0
0
.1 M Li
.05 M Li
.01 M Li
x
Bp
3200:1
1600:1
320:1
7200:1
3600:1
720:1
4b/4b′ > 100:1
>100:1
>100:1
x
Bp
Bp
x
0
0
0
.1 M Li
.05 M Li
.01 M Li
x
Np
2:1
1:1
0.2:1
7.2:1
3.6:1
0.7:1
3a/3a′ > 100:1
>100:1
>100:1
x
Np
Np
x
0
0
0
.1 M Li
.05 M Li
.01 M Li
x
Bp
3.2:1
1.6:1
0.32:1
7.2:1
3.6:1
0.7:1
4a/4a′ > 100:1
>100:1
>100:1
x
Bp
Bp
x
a
The reactions were performed by addition of probes 1b and 1a to an excess of Li A of a given formal concentration in THF (Li Bp)
x
x
or THP (Li Np) both at 0 °C and 25 °C followed by hydrolysis. Molar ratio Li A to RF = 10:1; an excess is used to simplify the
x
x
calculation of the theoretical limiting ratios [RAH]/[R′AH] under pseudo-first-order conditions. About 4–5 experiments of each
kind were run with nearly identical ratios.
b
Formal concentration of [Li A], with A = Np, Bp, and 1 ≤ x ≤ 2, indicating an undetermined mixture of radical anion and dianion.
x
c
Calculated using Eqn 7 as explained in the text.
d
Ratio assuming diffusion controlled rates for kred and kcoupl at 25 °C.
e
We set a sensitivity threshold for the simultaneous measurement of both isomers by GLC as ≤100:1. In all cases, no traces of
rearranged product were detected by GLC. This was contrasted with a sample of the actual rearranged alkylation product (3a′,
4a′, 3b′ and 4b′) synthesized for that purpose (see experimental). The cyclobutyl derivatives 3a″ and 4a″ are not produced in
these type of reactions (scheme 2 and 3).
f
1
13
Complete structural elucidation ( H-NMR, C-NMR, IR, MS and HRMS) of all alkylation compounds can be found in the
experimental and supplementary information. From the distribution of reduction and alkylation products at 25 °C, kET ≅ kSN for
b and kET ≅ 7kSN for 1a.
1
wileyonlinelibrary.com/journal/poc
Copyright © 2015 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2015

THE DIANIONS OF NAPHTHALENE AND BIPHENYL ARE AMONG THE STRONGEST ET REAGENTS AVAILABLE
derivatives (LiNp and LiBp) due to excessive dispersion of
not easy to rationalize the amount of rearranged/non rearranged
products in reactions involving metallic surfaces and radical
1
0b,27
data.
for the sodium derivatives as a good estimate for our purposes
Table 1, tablenote c). Notice that these rate constants are very
Still, we decided to use the rate constants measured
[
32]
probes.
To circumvent this problem, reactions in this work
(
were carried out after centrifugation of the excess metal from
the reaction media, so only the homogeneous phase containing
Li A in solution in the absence of Li(s) is responsible for the
x
large, near diffusion control, and even larger rate constants are
expected for the dianions. In our case, we treated kred as the
overall outcome of the two possible reduction processes (carried
reported results. It should be pointed out that the reaction fails
to afford alkylation products with cyclopropylmethyl chloride or
bromide. Only low mw hydrocarbons are obtained by an ET
2
out by LiA and Li A) which take place in parallel in the reacting
system. One further step forward would be considering the most
demanding scenario, i.e. a diffusion-controlled reaction. Let’s an-
alyze this scenario in detail. The theoretical estimate of a bimolec-
ular rate constant for a diffusion controlled reaction is given by:
[
33]
process.
k
D
= 4πN
A
(D
A
+ D
B
)(r
and r
A
+ r
B
), where D
are the reaction radii of the molecules A
A B
and D are the diffusion co-
CONCLUSIONS
efficients and r
A
B
and B. In the absence of diffusion coefficient data, it is common
to use the Stokes–Einstein equation relating the diffusion coeffi-
A kinetic study of the alkylation of the dianions of naphtha-
lene and biphenyl was carried out with the help of the very
fast radical probe cyclopropylmethyl fluoride (1a). To investi-
gate in detail the composition of the reaction crudes, the
whole set of potential rearranged and unrrearranged alkylation
products was synthetized and were used as controls. Analysis
by GLC showed that, to the limit of detection, the reaction
crudes contained only unrearranged alkylated arenes. The ob-
served distribution of products anticipates a mechanism of al-
kylation different from the classical one (which involves ET
followed by coupling between diffusing radicals and arene
radical-anions), and supports a concerted mechanism such as
cient to the viscosity η of the solvent: D
πηr , and to identify the diffusion radii with the radii for reaction
of (A + B). Combination of these two equations affords a simple
expression: k = 8RT/3η. For radical coupling, (i.e. for kcoupl = k
A B B B
= k T/6πηrA, D = k T/
6
B
D
D
)
only radical pairs which are in the singlet state during the en-
counter lead to the reaction products; therefore, there is an addi-
tional statistical factor of σ = 1/4 to be accounted for too. In our
9
ꢀ1 ꢀ1
case for THF at 25 °C (η = 0.456 mPa·s), k = 3.62 · 10 M
s ,
D
and likewise
k = 3.63 · 10 M
D
for
s
DME
at
25 °C
(η = 0.455 mPa·s),
The conclusions
9
ꢀ1 ꢀ1
[28]
(Table 1, tablenote d).
drawn in this paper are based on the lack of rearranged products.
As we shall evidence quantitatively below (Table 1), they are in-
controvertible for the cyclopropylmethyl probe, since the oppo-
site would require rates of coupling way beyond the diffusion
control limit. All together reduction and coupling make out of
the naphthalene and biphenyl radical anions and dianions some
excellent radical traps, keeping the concentration of [R·] and [R′·]
low and thus validating the steady state approximation used in
Eqn 4. The remaining required kinetic data is well reported. The
unimolecular rearrangement of R· → R′· is barely dependent on
the S 2. These results overtake inconclusive former studies
with slower radical probes such as hex-5-enyl fluoride (1b)
that, although pointed out to the same conclusions, would
not stand a rigorous quantitative test.
N
EXPERIMENTAL SECTION
Synthesis and spectroscopic description of cyclopropylmethyl
4-methylbenzenesulfonate (2a)
the solvent and occurs with a rate constant k
r
(hex-5-enyl)
A
solution of cyclopropylmethanol (7.247 g, 100 mmol) and
5
–1
[29]
=
10 s at 25 °C.
For the cyclopropylmethyl radical, the rear-
dimethylaminopyridine (0.617 g, 5 mmol) in triethylamine (50 ml) was
added to a solution of p-toluenesulfonyl chloride (22.98 g, 120 mmol) in
dichloromethane (60 ml) at 0 °C, under a dry atmosphere (Ar). The mixture
was stirred at 0 °C for 12 h. After this time, the reaction crude was poured
on a mixture of crushed ice and concentrated sulfuric acid, and was ex-
tracted with dichloromethane (3 × 30 ml) maintaining ice-cold tempera-
tures at all times. The organic phase was dried with sodium sulfate, and
the solvent was removed under reduced pressure at 15 °C. The resulting
crude cyclopropylmethyl tosylate is pure and can be used without further
purification. Addition of a carefully weighed amount of an internal stan-
dard (hexamethylbenzene) revealed a yield of 96% of pure 2a (>99%),
8
–1
rangement is much faster, k
r
(cyclopropylmethyl) = 10 s at
The application of Eqn 7 to our set of experimental
[
30,31]
2
5 °C.
conditions is collected in Table 1.
The results shown in Table 1 make clear a number of things
about the reactions of Schemes 2 and 3. The experiments
reported so far using hex-5-enyl fluoride (1b) as a radical probe
proved to be inconclusive. For this probe, the calculated RAH/R′
AH ratio would be too large to be measured experimentally
under any of the experimental conditions assayed. In the most
1
x
favorable case (0.01 M Li Np whit 1b) the ratio (200:1) is twice
as determined by 400 MHz H NMR.
the detection limit. The scenario is very different for the
cyclopropylmethyl fluoride (1a). The calculated RAH/R′AH ratio
assuming radical coupling should be easily measured by GLC,
or even by a less precise quantitative techniques such as
Cyclopropylmethyl 4-methylbenzenesulfonate (2a). Obtained in
this way, 2a is a low melting point white solid (mp = 16–17 °C,
1
[13]
H-NMR spectroscopy.
Moreover, the rearranged compound
uncorrected), R
f
= 0.56 (hexane/ethyl acetate 95:5); IR (film): ν
ꢀ
1
(
cm ) = 3045, 1597, 1356, 1171, 1096, 1026, 924, 842, 813,
R′AH should be the major component in most instances. However
this is not the case, since rearranged products R′AH remain
undetectable under all the experimental conditions. Even in the
limiting case of a hypothetical diffusion controlled reaction of
1
7
81; H NMR (300 MHz, CDCl
3 H
, 25 °C, TMS): δ = 0.20–0.29 (m,
2
H, 2 × CHH of cyclopropane), 0.53–0.63 (m, 2H, 2 × CHH of cyclopro-
3
pane), 1.04–1.19 (m, 1H, CH of cyclopropane), 3.90 (d, J(H,H) = 7.4 Hz,
3
3
2
=
(
7
H, OCH
8.3 Hz, 2H, aromatic H); C NMR (75 MHz, CDCl
2 × CH of cyclopropane), 10.02 (CH of cyclopropane), 21.79 (CH
6.03 (OCH ), 127.99 (arom. CH), 129.92 (arom. CH), 133.63 (CCH
2
CH), 7.35 (d, J(H,H) = 8.3 Hz, 2H, aromatic H), 7.81 (d, J(H,H)
Li A with the free radical R·, the expected ratio RAH/R′AH
is
13
x
diffuss.
3 C
, 25 °C, TMS): δ = 3.86
expected to fall within the normal measurable range of most
techniques (from 7.2:1 to 0.7:1 depending on the assayed con-
centrations). Concerning unwanted metal–solution interface
effects, Mattalia et al. raised attention to the issue that it was
2
3
),
),
2
3
+
+
144.72 (CS); MS (70 eV): m/z (%): 228 (0.14) [M + 2], 227 (0.35) [M + 1],
226 (1.80) [M ], 198 (24), 155 (95), 92 (15), 91 (100), 65 (21).
+
J. Phys. Org. Chem. 2015
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

I. BLASCO, H. PÉREZ AND A. GUIJARRO
Synthesis and spectroscopic description of cyclopropylmethyl fluo-
ride (1a)
with HCl 3 M. Hydroquinone (10 mg) was added as stabilizing, the crude
was extracted with diethyl ether (3 × 20 ml) and dried over Na SO4(s). In
2
the case of naphthalene derivatives, they were rearomatized with a slight
excess of DDQ at room temperature overnight, washed with 1 M NaOH
2
and dried over Na SO4(s). The solvent was removed by rotatory evapora-
tion (15 Torr), and the resulting residue was purified by column chroma-
tography (silica gel doped with a 5% of hidroquinone, hexane/ethyl
acetate). Yields: 3a 50 %; 3a′ 55%; 4a 55 %; 4a′ 63%.
The synthesis was performed in a dry two-neck 100-ml flask containing a
Teflon-coated stirrer bar and connected to a cold trap. One neck was
capped with a rubber septum, the second was glass connected to a cold
trap using conical joints and this last to a bubbler. The apparatus was
evacuated with argon; the flask was cooled to 0 °C and the trap at
ꢀ
80 °C. Solid cyclopropylmethyl p-toluenesulfonate (2a, 3.620 g,
6 mmol) was placed in the flask, and a solution of tetrabutylammonium
1
1
-Cyclopropylmethylnaphthalene (3a).
R
f
= 0.49 (hexane); IR (film): ν
fluoride (TBAF 1 M in THF at 0 °C, 20 ml) was added with stirring while
maintaining everything between 0 and 5 °C. A gentle stream of argon
was bubbled through the reaction mixture using a needle, and the vola-
tile material 1a was allowed to distil from the reaction mixture and col-
lected in the cold trap. The distilled fluorinated material was 93%
cyclopropylmethyl fluoride (1a), along with 4.4% of cyclobutyl fluoride
ꢀ
1
(cm ) = 3074, 3009, 2923, 2850, 1639, 1598, 1513, 1463,
1
1
399, 1272, 1093, 1020, 914, 832, 791, 771, 734, 645;
H
NMR (300 MHz, CDCl , 25 °C, TMS): δ = 0.14–0.38 (m, 2H,
3
H
2 × CHH of cyclopropane), 0.48–0.66 (m, 2H, 2 × CHH of cy-
clopropane), 1.08–1.27 (m, 1H, CH of cyclopropane), 3.01
3
(1a″) and a 2.6% of homoallyl fluoride (1a′) in THF in an overall 22–25%
(d, J(H,H) = 6.7 Hz, 2H, CH
2
CHcyclopropane), 7.33–7.60 (m,
4
H, CHar), 7.73 (app d, J = 7.8 Hz, 1H, CHar), 7.78–7.93 (m, 1H CHar), 8.00–
13
yield. The crude distillate was treated with a slight excess of bromine (3
mole %) and distilled again to afford >95% pure 1a in the strict absence
8
.14 (m, 1H, CHar); C NMR (75 MHz, CDCl
of cyclopropane), 11.13 (CH of cyclopropane), 37.28
CHcyclopropane), 123.96, 125.53, 125.64, 125.74, 125.82, 126.68,
28.81 (7 × CHar), 132.23, 133.84 (2 × Car), 138.18 (CarCH ); MS (70 eV): m/
3 C
, 25 °C, TMS): δ = 5.26 (2C,
of unwanted 1a′. Alternatively, CHCl
3
could be used as solvent instead of
2 × CH
CH
2
(
2
THF (see provided spectrum in the supporting information).
1
2
+
+
+
z (%): 185 (0.02) [M + 3], 184 (0.43) [M + 2], 183 (6.19) [M + 1], 182
1
+
Cyclopropylmethylfluoride (1a). H NMR (300 MHz, CDCl
3
, ꢀ20 °C,
(
(
1
39.51) [M ], 167 (29), 166 (6), 165 (14), 154 (18), 153 (35), 152 (19), 142
12), 141 (100), 115 (22); HRMS: (EI) m/z calcd for C14
82.1089.
TMS): δ
H
= 0.18–0.35 (m, 2H, 2 × CHH of cyclopropane), 0.49–
H14 182.1095, found
0
.66 (m, 2H, 2 × CHH of cyclopropane), 1.09–1.30 (m, 1H, CH of
2
3
cyclopropane), 4.19 (dd, J(F,H) = 48.9 Hz, J(H,H) = 7.5 Hz, 2H,
1
3
3
FCH
C) = 6.9 Hz, 2C, 2 × CH
of cyclopropane), 88.51 (d, J(F,C) = 165.7 Hz, FCH
2
CH); C NMR (75 MHz, CDCl
3 C
, 25 °C, TMS): δ = 2.82 (d, J(F,
ꢀ1
2
1
-(3-Butenyl)naphthalene (3a′).
R
f
= 0.48 (hexane); IR (film): ν (cm )
2
of cyclopropane), 10.97 (d, J(F,C) = 25.2 Hz, CH
1
=
3070, 3005, 2948, 2846, 1440, 1260, 1085, 1032, 914,
2
).
1
7
91, 771;
H
NM3R (300 MHz, C3DCl
3
, 25 °C, TMS):
δ
H
= 2.51 (app td, J(H,H) = 7.9 Hz, J(H,H) = 7.2 Hz, 2H,
3
Kinetic experiments: reaction of the fluorinated probes 1a and 1b
with Li Np and Li Bp. Solutions of the given formal concentrations
CH
2
CH
2
CH = CH
2
), 3.17 (app t,3 J(H,H) = 7.9 Hz, 2H,
J(H,H) = 10.2 Hz, 1H,
CH = CHH), 5.10 (app d, J(H,H) = 17.1 Hz, J(H,H) = 1.6 Hz, 1H, CH = CHH),
x
x
C CH CH ); 5.02 (app d,
ar
2
2
[
9,2,8b]
[2,8a]
3
3
of Li
x
Np in THP,
x
and Li Bp in THF,
were prepared as usual under
3
3
3
argon using a large excess of lithium powder (ca. 10:1 Li:A molar ratio)
and the corresponding weighted arene with stirring at 0 °C during 1 h.
These suspensions were centrifuged (2500 rpm, 2 min) at the same tem-
perature to remove the excess metal and the corresponding probe 3a or
5.97 (app ddt, J(H,H) = 16.8 Hz, J(H,H) = 10.2 Hz, J(H,H) = 6.6 Hz, 1H,
3
CH CH = CH ), 7.30–7.56 (m, 4H, CH ), 7.72 (d, J(H,H) = 8.1 Hz, 1H, CH ),
7.82–7.89 (m, 1H, CH ), 8.04 (d, J(H,H) = 7.9 Hz, CH ); C-RMN (75 MHz,
2
2
ar
ar
3
13
ar
ar
CDCl , 25 °C, TMS): δ = 32.62 (CH CH CH = CH ), 34.96 (C CH CH ),
3
C
2
2
2
ar
2
2
3
b diluted in THF (ca. 10:1) was slowly injected into the deeply colored
solution of the dianions (dark purple for Li Np and dark greenish-blue
for Li Bp; molar ratio Li A:RF ≈ 10:1) at 25° and 0 °C. After 30 min, the re-
action was hydrolized with acetonitrile (1 ml) followed by water (5 ml)
115.07 (CH = CH), 123.88, 125.57, 125.66, 125.90, 126.08, 126.78, 128.93
2
x
(7 × CH ), 131.97, 134.01 (2 × C ), 138.07 (C CH ), 138.40 (CH = CH );
ar
ar
ar
2
2
+
+
x
x
MS (70 eV): m/z (%): 185 (0.01) [M + 3], 184 (0.25) [M + 2], 183 (3.61)
+ +
[M + 1], 182 (23.55) [M ], 142 (12), 141 (100), 139 (6), 115 (20); HRMS:
(EI) m/z calcd for C14H14 182.1096, found 182.1095.
[1c]
and neutralized. Hydroquinone (10 mg) was added as stabilizing, and
it was extracted with diethyl ether (3 × 20 ml) and dried over Na
2
SO4(s).
Samples were submitted to GLC analysis. In the case of naphthalene
derivatives, they were rearomatized with a slight excess of DDQ at
room temperature overnight, washed with 1 M NaOH and submitted to
GLC analysis. In all cases, pure isolated 3a, 3a′, 3a″, 3b, 3b′, as well as
1
-(Cyclobutyl)naphthalene (3a″).
R
f
= 0.51 (hexane); IR (film):
ν
ꢀ1
(
1
cm ) = 3054, 2960, 2932, 2854, 1676, 1594, 1513, 1464,
1
264, 1097, 795, 775; H NMR (300 MHz, CDCl
= 1.84 2.00 (m, 1H, CHH of cyclobutane), 2.05–2.42 (m,
H, 3 × CHH of cyclobutane), 2.47–2.62 (m, 1H, CHH of
3
, 25 °C, TMS):
δ
H
4a, 4a′, 4a″, 4b and 4b′ were used as control.
3
cyclobutane), 4.06–4.25 (m, 1H, CH of cyclobutane), 7.33–
3
Synthesis and spectroscopic description of alkylated compounds 3–
7.40 (app d, J(H,H) = 7.1 Hz, 1H, CH ), 7.41–7.53 (m, 3H,
ar
(′)–(″)
3
4a
. General procedure
3 × CHar), 7.70 (d, J(H,H) = 8.1 Hz, 1H, CHar), 7.81–7.88 (m, 1H, CHar),
13
7
(
.93–8.00 (m, 1H, CHar); C-RMN (75 MHz, CDCl
CH of cyclobutane), 29.32 (2 × CH of cyclobutane), 38.22 (CH of
cyclobutane), 122.56, 124.41, 125.53, 125.62, 125.65, 126.42, 128.76
3 C
, 25 °C, TMS): δ = 18.83
x x
For the alkylation of Li Np and Li Bp and subsequent isolation of prod-
ucts, we used a procedure similar to that described in Kinetic experiments
section above, but using a stoichiometric ratio of reagents at a lower re-
2
2
(
(
[
(
7 × CHar), 131.65, 133.87 (2 × Car), 141.78 (CarCH of cyclobutane); MS
70 eV): m/z (%):185 (0.02) [M + 3], 184 (0.40) [M + 2], 183 (5.48)
M + 1], 182 (36.03) [M ]; 155 (12), 154 (100), 153 (82), 152 (33); HRMS:
EI) m/z calcd for C14 14 182.1096, found 182.1099.
action temperature to enhance S
ium powder (Li(s), 70 mg, 10 mmol) and the corresponding arene
naphthalene, 256.4 mg, 2 mmol; or biphenyl, 308.5 mg, 2 mmol) in a
N
over ET products. A suspension of lith-
+
+
+
+
(
H
dry ethereal solvent (10 ml) (tetrahydropyran was used in the case of
naphthalene while tetrahydrofuran was used in the case of biphenyl)
was allowed to stir magnetically for 1 h at 0 °C under Ar atmosphere.
These deeply colored suspensions were further cooled down to ꢀ40 °C
(1-Cyclopropylmethylcyclohexa-2,5-dienyl)benzene (4a).
f
R = 0.47
ꢀ
1
(hexane); IR (film): ν (cm ) = 3017, 2947, 2914, 2844, 1470,
1463, 1356, 1238, 1022, 944, 911, 735, 692; H NMR
(300 MHz, CDCl3, 25 °C, TMS): δH = 0.03–0.09 (m, 2H,
2 × CHH of cyclopropane), 0.36–0.44 (m, 2H, 2 × CHH of cy-
clopropane), 0.61–0.76 (m, 1H, CH of cyclopropane), 1.78 (d,
1
(
1
for Li
x
Np) or ꢀ80 °C (for Li
x
Bp), and the corresponding alkyl fluoride
a, 1a′ or 1a″ (2 mmol) diluted in THF (1 ml) was slowly injected to the
[34]
suspension of the dianions at the same temperature.
the reaction was hydrolyzed with acetonitrile (1 ml) followed by water
5 ml), the temperature was raised to rt, and the mixture was neutralized
After 30 min,
(
wileyonlinelibrary.com/journal/poc
Copyright © 2015 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2015

THE DIANIONS OF NAPHTHALENE AND BIPHENYL ARE AMONG THE STRONGEST ET REAGENTS AVAILABLE
3
J(H,H) = 6.5 Hz, 2H,CCH
2
CH), 2.67–2.72 (m, 2H, (CH = CH)
C(Ph)), 5.84 (app dm,
C(Ph)), 7.12–7.21 (m, 1H, Ph), 7.26–7.40
m, 4H, Ph); C NMR (75 MHz, CDCl , 25 °C, TMS): δ = 5.22 (2C, 2 × CH
of cyclopropane), 7.26 (CH of cyclopropane), 26.25 (CCH CH), 44.94
), 123.09 (2C, (CH = CH) CH ),
25.91 (CHar), 126.79 (2C, CHar), 128.37 (2C, CHar), 133.35 (2C, CH = CH)
2
CH
2
), 5.75 (app
CRC Press: Boca Raton, FL., 2004, 8–33; b) In THF, a reported half-peak
+ +
3
3
dt, J(H,H) = 10.4 Hz, J(H,H) = 1.8 Hz, 2H, (CH = CH)
2
½
potential for the pair Li(s)/Li(THF) is E (Li/Li ) = –3.07 V vs. Ag/AgCl: J.
3
Mortensen, J. Heinze, Tetrahedron Lett. 1985, 26, 415–418; c) The
J(H,H) = 10.4 Hz, 2H, (CH = CH)
2
+
f
1
3
formal Li /Li
potencial was measured recently, E = ꢀ3.48 V
(
s) (THF)
0
(
3
C
2
+
6–
vs. Fc/Fc PF : C. A. Paddon, S. E. Ward Jones, F. L. Bhatti, T. J.
Donohoe, R. G. Compton, J. Phys. Org. Chem. 2007, 20, 677–684.
6] Among polycyclic aromatic hydrocarbons, naphthalene, holds the
2
((CH = CH)
2
C(Ph)), 45.90 ((CH = CH)
2
CH
2
2
2
[
1
highest—yet to be measured—second reduction potential E°
2 8
(C10H ,
+
2
C(Ph)), 148.34 (CPh); MS (70 eV): m/z (%): 212 (0.03) [M + 2], 211 (0.33)
E° = ꢀ2.53 V; E° = beyond the experimental limit). Biphenyl on the
1
2
+ +
[
(
M + 1], 210 (1.93) [M ], 156 (13), 155 (100), 154 (31), 153 (12); HRMS:
EI) m/z calcd for C16 18 210.1408, found 210.1395.
1
other hand, has the highest first reduction potential E° , giving rise to
H
the radical anion, and the second higher E°
2
, affording the dianion
NH vs.
(
C
12
H
10, E° = –3.18 V). All measured in Me
1
= ꢀ2.68 V, E°
2
2
Ag/AgCl: K. Meerholz, J. Heinze, J. Am. Chem. Soc. 1989, 111, 2325–2326.
7] Taking E°(Li/Li+) ≈ ꢀ3.08 V vs. Ag/AgCl as a reasonable value in THF
and comparing it to the second reduction wave of biphenyl, E°2
= ꢀ3.18 V vs. Ag/AgCl, it turns out that the lithium double
(1-(3-Butenyl)cyclohexa-2,5-dienyl)benzene (4a′).
R
f
= 0.61 (hexane);
[
ꢀ
1
IR (film): ν (cm ) = 3028, 2924, 2847, 2809, 1634, 1602,
1
1
487, 1441, 908, 744, 700; H NMR (300 MHz, CDCl
TMS): δ 1.82–1.94 (m, 2H, (Ph)CCH CH ), 1.99–2.11 (m,
CH CH = CH ), 2.63–2.73 (m, 2H, (CH = CH) CH ),
.94 (app dm, J(H,H) = 10.2 Hz, 1H, CH = CHH), 5.02 (app
3
, 25 °C,
(Bp–2/Bp–1)
reduction of biphenyl is slightly endergonic under standard condi-
tions (ΔG° = ꢀnFE°). The situation is even more unfavorable for
naphthalene. Although this is only approximate since electrochem-
ical data comes from different sources under different conditions (e.
g. solvent), it is still indicative of the reducing strength of the
dianions considered in this work.
H
2
2
2
4
H, CH
2
2
2
2
2
3
3
3
dm, J(H,H) = 17.1 Hz, 1H, CH = CHH), 5.61 (app dt, J(H,H)
4
=
10.5 Hz, J(H,H) = 2.0 Hz, 2H, (CH = CH)
CH , CH = CH ), 7.14–7.21 (m, 1H, Ph), 7.27–7.40 (m, 4H, Ph);
C NMR (75 MHz, CDCl , 25 °C, TMS): δ = 26.26 ((CH = CH) CH ), 29.78
CH CH CH = CH ), 39.29 ((Ph)CCH CH ), 44.04 ((CH = CH) C(Ph)), 114.18
CH = CH), 123.87 (2C, (CH = CH) CH ), 126.05 (CHPh), 126.77 (2C,
× CHPh), 128.42 (2C, 2 × CHPh), 132.55 (2C, (CH = CH) C(Ph)), 139.44
CH = CH ), 148.26 (CPh); MS (70 eV): m/z (%): 211 (0.08) [M + 1], 210
2
C(Ph)), 5.80–5.96 (m, 3H,
(
CH = CH)
2
2
2
[
8] a) C. Melero , A. Guijarro , M. Yus , Tetrahedron Lett. 2006, 47, 6267–6271;
b) C. Melero , A. Guijarro , V. Baumann , A. J. Perez-Jimenez , M.
Yus , Eur. J. Org. Chem. 2007, 5514–5526.
1
3
3
C
2
2
(
(
2
2
2
2
2
2
2
2
2
2
[9] R. P. Herrera , A. Guijarro , M. Yus , Tetrahedron Lett. 2003, 44,
1309–1312.
2
+
[
[
[
[
10] a) J. F. Garst , F. E. Barton II , Tetrahedron Lett. 1969, 7, 587–590; b) J. F.
Garst , F. E. Barton II , J. Am. Chem. Soc. 1974, 96, 523–529.
11] J. M. Percy, in: Science of Synthesis (Ed: J. M. Percy), Thieme: Stuttgart,
(
(
2
+
0.36) [M ], 156 (13), 155 (100), 154 (15), 153 (12); HRMS: (EI) m/z calcd
18 210.1408, found 210.1401.
for C16
H
2
006, Vol. 34, 267–268.
12] J. R. During , Z. Yu , C. Zheng , G. Guirguis , J. Phys. Chem. A. 2004,
108, 5353–5364.
13] B. M. Kraft , R. J. Lachicotte , W. D. Jones , J. Am. Chem. Soc. 2001,
123, 10973–10979.
14] L. Mattew , J. Warkentin , J. Am. Chem. Soc. 1986, 108, 7981–7984.
15] D. P. Cox , J. Terpinski , W. Lawrynowicz , J. Org. Chem. 1984, 49,
(1-Cyclobutyl)cyclohexa-2,5-dienyl)benzene (4a″).
R
f
= 0.46 (hex-
ꢀ
1
ane); IR (film): ν (cm ) = 3021, 2964, 2928, 2850, 2814, 1484,
1
1
2
1
448, 1203, 1130, 938, 759, 698, 681; H NMR (300 MHz, CDCl
5 °C, TMS): δ 1.62–1.99 (m, 6H, 3 × CH of cyclobutane),
.99–2.11 (m, 2H, CH CH CH = CH ), 2.66–2.76 (m, 2H,
CH ), 2.85–3.02 (m, 1H, CH of cyclobutane), 5.65
3
,
[
[
H
2
2
2
2
3
216–3219.
(CH = CH)
2
3
2
[
16] Cp* ZrH was introduced in this paper as a reagent of
2
2
4
2
(app dt, J(H,H) = 10.4 Hz, J(H,H) = 2.0 Hz, 2H, (CH = CH) C
hydrodefluorination. Its reactivity against monofluorinated aliphatic
C―F bonds was described to decrease in the order 1° > 2° > 3. In
our hands, the organic layer consisted of 74% of cyclopropylmethyl
fluoride, 24% of cyclobutyl fluoride and 2% of 3-butenyl fluoride.
The purification step was apparently against the reported order of
reactivity given by the authors, and was not carried out.
3
4
(Ph)), 5.90 (app dt, J(H,H) = 10.5 Hz, J(H,H) = 3.3 Hz, 2H,
13
(
CH = CH)
2
CH
2
), 7.10–7.21 (m, 1H, CHPh), 7.26–7.33 (m, 4H, 4 × CHPh);
, 25 °C, TMS): δ = 17.67, 23.69 (2C, 2 × CH
cyclobutane), 26.61, 43.15 ((CH = CH) CH ), 46.25 ((CH = CH) C(Ph)),
24.06 (2C, (CH = CH) CH ), 125.92 (CHPh), 127.30 (2C, 2 × CHPh), 128.29
2C, 2 × CHPh), 130.98 (2C, 2 × (CH = CH) C(Ph)), 147.49 (CPh); MS (70 eV):
C
NMR (75 MHz, CDCl
3
C
2
of
2
2
2
1
(
2
2
[
17] Z. Mo , Q. Zhang , L. Deng , Organometallics. 2012, 31, 6518–6521.
2
+
+
[18] S. Samdal , H. Mollendal , J.-C. Guillemin , J. Phys. Chem. A. 2014, 118,
344–2352.
m/z (%): 211 (0.08) [M + 1], 210 (0.54) [M ], 156 (13), 155 (100), 154
27), 153 (14), 152 (10); HRMS: (EI) m/z calcd for C16 18 210.1408, found
10.1391.
2
(
2
H
[
[
19] D. D. Roberts , J. Org. Chem., Vol. 1970, 35, 4059–4062.
20] S. L. Bogen , S. Ruan , R. Liu , S. Agrawal , J. Pichardo , A. Prongay , B.
Baroudy , A. K. Saksena , V. Girijavallabhan , F. G. Njoroge , Bioorg.
Med. Chem. Lett. 2006, 16, 1621–1627.
Acknowledgements
[
[
[
21] D. D. Roberts , J. Org. Chem. 1965, 30, 23–28.
22] L. Mathew , J. Warkentin , J. Am. Chem. Soc. 1986, 108, 7981–7984.
23] a) J. K. Kochi , J. W. Powers , J. Am. Chem. Soc. 1970, 92, 137–146;
b) C. L. Jenkins , J. K. Kochi , J. Org. Chem. 1971, 36, 3103–3111.
This work was generously supported by the Spanish Ministerio de
Educación y Ciencia (CTQ2011-24165) and the Universidad de Ali-
cante. IB thanks the Instituto de Síntesis Orgánica for financial
support.
[24] In this approach, the anionic rearrangement (k′ ) in Scheme 4 have
r
been purposely set to zero. For the hex-5-enyllithium, the anionic
8
cyclization k′
r
is known to be ca. 10 times slower than the corre-
sponding radical cyclization: W. F. Bayley, J. J. Patricia, V. C.
DelGobbo, R. M. Jarret, P. J. Okarma, J. Org. Chem. 1985, 50, 2000–
REFERENCES
2
003. For the cyclopropylmethyllithium k′
mined, to our best knowledge.
25] Notice that this splitting is mathematically exact given that [RAH]/
r
has not been deter-
[
1] The involvement of these dianions has been proposed only in a lim-
ited number of studies: Li : a) J. Smid, J. Am. Chem. Soc. 1965,
7, 655–656; b) R. Benken, H. Günther, Helv. Chim. Acta 1988, 71,
94–702; c) M. Yus, R. P. Herrera, A. Guijarro, Chem. Eur. J. 2002, 8,
574–2584. Li 10; d W. Huber , A. May , K. Müllen , Chem. Ber.
981, 114, 1318–1336; e J. J. Eisch , J. Org. Chem. 1963, 28, 707–710.
2] C. Melero, R. P. Herrera, A. Guijarro, M. Yus, Chem. Eur. J. 2007, 13,
0096–100107.
2 10 8
C H
[
8
6
2
[
[
R′AH] = [RH]/[R′H], which is the case in here since both equal to
R·]/[R′·]. The chemical meaning of this is more subtle and entails
2 12
C H
that both kred and kcoupl are the same for R· and R′·. This is cus-
tomarily accepted in radical probe experiments, since the final
products are never the radical themselves, but some final stable
products derived from them. Dealing with only one set of con-
stants (kred and kcoupl) is an acceptable approximation implicit in
Scheme 4 that simplifies much the model. In any case, the
toughest test for the model comes at the limit of diffusion con-
1
[
1
[
[
3] C. Melero , A. Guijarro , M. Yus , Dalton Trans. 2009, 1286–1289.
4] a S. Irle , H. Lischka , J. Chem. Phys. 1995, 103, 1508–1522; b M. de la
Viuda , M. Yus , A. Guijarro , J. Phys. Chem. B. 2011, 115, 14610–14616.
5] a) The pair Li(s)/Li(H2O) is the most negative among alkali metals,
+
[
x
trolled reactions of R· and R′· with Li A. This limiting case is also
with a standard potential E°(Li/Li+) = –3.04 V, in water relative to
SHE: CRC Handbook of Chemistry and Physics, 85th ed. (Ed: DR Lide),
considered and explained later in this work.
J. Phys. Org. Chem. 2015
Copyright © 2015 John Wiley & Sons, Ltd.
wileyonlinelibrary.com/journal/poc

I. BLASCO, H. PÉREZ AND A. GUIJARRO
[
[
[
26] The corresponding coupling products (RAH + R′AH) were isolated
and characterized only recently. Full characterization of these com-
pounds can be found in ref. 2.
27] We have also observed mixtures of both rearranged and non-
rearranged products in the reaction of hex-5-enyl fluoride with
LiNp: see ref. 9.
[33] However, if the arene bears electron withdrawing groups capable of
further dianion stabilization (such as ―CN), the carbanion is soft-
ened and alkylation becomes a main reaction outcome with inter-
esting synthetic applications: T. A. Vaganova, E. V. Panteleeva, V. D.
Shteingarts, Russ. Chem. Bull. 2008, 57, 768–779.
[34] a) 4-Fluoro-1-butene (1a′) was prepared by reaction of 3-butenyl
tosylate with KF in triethyleneglycol following the procedure de-
scribed in: J. San Filippo Jr., L. J. Romano, J. Org. Chem. 1975, 40,
1514–1515; b) Cyclobutyl fluoride (1a″) was prepared by reaction
of cyclobutanol with DAST: T. B. Patrick, L. Zhang, Q. Li, J. Fluorine
Chem. 2000, 102, 11–15.
28] Data from: CRC Handbook of Chemistry and Physics, Internet Version
2007, 87th Edition, (Ed: DR Lide), Taylor and Francis: Boca Raton,
FL, 2007, 15–21 and 15-16.
[
[
29] D. J. Carlsson , K. U. Ingold , J. Am. Chem. Soc. 1968, 90, 7047–55.
30] a) C. L. Jenkins , J. K. Kochi , J. Org. Chem. 1971, 36, 3103–3111; b) B.
Maillard , D. Forrest , K. U. Ingold , J. Am. Chem. Soc. 1976, 98, 7024–7026.
[
31] This rearrangement is reversible, but the equilibrium heavily favors
4
the 3-butenyl radical by a factor >10 at 25°C, so the reverse reaction
can be ignored for our purposes: A. Effio, D. Griller, K. U. Ingold, A. L. J.
Beckwith, K. J. Serelis, J. Am. Chem. Soc. 1980, 102, 1734–1736.
32] H. Hazimeh, J.-M. Mattalia, C. Marchi-Delapierre, R. Barone, N. S.
Nudelman, M. Chanon, J. Phys. Org. Chem. 2005, 18, 1145–1160.
SUPPORTING INFORMATION
Additional supporting information may be found in the online
[
version of this article at the publisher’s website.
wileyonlinelibrary.com/journal/poc
Copyright © 2015 John Wiley & Sons, Ltd.
J. Phys. Org. Chem. 2015