Ring expansion and vinylic nucleophilic substitution competing for (tert-alkyl)2CC(Li)-Cl in carbenoid chain processes

Article abstract of DOI:10.1016/j.tet.2014.03.004

Vinylic nucleophilic substitution (S_NV) reactions of unactivated, cyclic α,α-dichloroalkenes [(tert-alkyl) ₂CCCl₂] with aryllithiums (RLi) to give (tert-alkyl) ₂CC(Cl)-R are presented to be carbenoid chain react

Full text of DOI:10.1016/j.tet.2014.03.004

Tetrahedron 70 (2014) 2703e2710
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Ring expansion and vinylic nucleophilic substitution competing for
(tert-alkyl)2C]C(Li)eCl in carbenoid chain processesq
*
Rudolf Knorr , Thomas Menke, Karsten-Olaf Hennig, Johannes Freudenreich,
€
Petra Bohrer, Bernhard Schubert
€
€
Department Chemie, Ludwig-Maximilians-Universitat, Butenandtstrasse 5-13, 81377 Munchen, Germany
a r t i c l e i n f o
a b s t r a c t
Article history:
Vinylic nucleophilic substitution (SNV) reactions of unactivated, cyclic
a,a-dichloroalkenes [(tert-al-
Received 5 February 2014
Received in revised form 27 February 2014
Accepted 3 March 2014
kyl)2C]CCl2] with aryllithiums (RLi) to give (tert-alkyl)2C]C(Cl)eR are presented to be carbenoid chain
reactions, which involve the unsaturated Cl,Li-carbenoids (tert-alkyl)2C]C(Li)eCl as transient in-
termediates. The chains are longer and the overall reactions much slower in tert-butyl methyl ether (t-
BuOMe) than in THF as the solvent. In competition with the fast SNV step of these Cl,Li-carbenoids, the
FritscheButtenbergeWiechell (FBW) ring expansion in t-BuOMe (but less so in THF) generates short-
lived cyclohexyne species, which are trapped by the accompanying RLi species to yield chlor-
ocyclohexene derivatives [(tert-alkyl)e(Cl)C]C(R)e(tert-alkyl)] as the FBW chain products.
Ó 2014 Elsevier Ltd. All rights reserved.
Available online 11 March 2014
Keywords:
Carbenoid chain reaction
FritscheButtenbergeWiechell reaction
Vinylic nucleophilic substitution
Cl/Li interchange reaction
1. Introduction
are sufficiently basic to perform the final transfer (step 3) of par-
ticles A (¼H or D) from 1a or 1b, producing 4a or 4b along with the
Two different types of carbenoid chain reactions were in-
dependently identified in the last decade. The type-1 chain
(Scheme 1) looks like a simple vinylic nucleophilic substitution
(SNV) reaction that produces 4 from 1. However, a significantly
retarding kinetic deuterium isotope effect (1b reacting slower than
1a) established that 1a was consumed through an initial scission of
chain carrier 2. An unintentional trapping of 3 by side-reactions
(‘leakage’) may interrupt a running chain reaction. For the
a,a-
dichloride 1c, this mechanism was established in a different man-
ner with methyllithium (the usual tetramer) as a strongly basic but
weakly nucleophilic reagent R1Li in competition with PhC^CLi as
a weakly basic but strongly nucleophilic reagent R2Li as follows. No
conversion of 1c took place with PhC^CLi alone, whereas the ex-
pected product 4c (R2 ¼ C^CPh) emerged as soon as substantially
less than 1 equiv of methyllithium was added as the initiator R1Li
(step 1) that was unable to compete with PhC^CLi for the in-
termediate 2 (step 2). This established for 1c that the initiating step
1 (performed by methyllithium) is different from a step 2 in which
R2Li generates the precursor 3 of product 4c, so that at least one
intermediate (namely, 2) must occur in between these steps during
the conversion of 1c to 4c. In the general protocol1,2 with R2Li ¼
R1Li, the total consumption of 1aec often required only little more
than 1 equiv (instead of 2) of a sufficiently basic nucleophile R2Li, as
demanded if the uninterrupted carbenoid chain mechanism
obtains.
the AeC(a
) bond (step 1) by a strong base R1Li with formation of the
unsaturated Cl,Li-carbenoid 2.1,2 A sufficiently reactive nucleophile
R2Li (which may be identical with R1Li) will usually intercept 2 very
quickly3 (step 2) before 2 decays.4 The resultant alkenyllithiums 3
2
2
A
α
Li
α
Cl
Cl
Li
α
A
α
R
R
step 1
+ R Li
step 2
+ LiR
step 3
Me
Me
Me
Me
Me
Me
+
1
2
Me
Me
-
LiCl
-
R A
-
A
1,4
?
1
2
3
4
a
b
c
H
D
Cl
FBW
leakage
The FritscheButtenbergeWiechell (FBW) reaction3,5 will elimi-
nate LiHal from a short-lived carbenoid R3R4C]C(Li)eHal with
formation of an alkyne R3C^CR4. For the R3R4C part embedded in
a cyclic arrangement, this means ring expansion with formation of
a cycloalkyne6 and deviation from the usual course of the above
type-1 carbenoid chain reaction if an interception of a carbenoid
Scheme 1. The type-1 carbenoid chain reaction.
q
Sterically congested molecules, Part 28. For Part 27, see Ref. 25.
* Corresponding author. E-mail address: rhk@cup.uni-muenchen.de (R. Knorr).
0040-4020/$ e see front matter Ó 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.tet.2014.03.004

2704
R. Knorr et al. / Tetrahedron 70 (2014) 2703e2710
such as 2 by R2Li in step 2 is too slow. A similar kind of deviation by
FBW ring expansion was recently detected7 in connection with the
carbenoid chain reaction type-28,9 that explained the THF-
dependent conversion of the acyclic starting material 5 into cyclic
isomers 7 (Scheme 2) in the following way. The weak basicity of
LiC^CR0 (6) sufficed for an initial activation of 5 by the transfer of
particles A (¼H, D, I, or Br)7 with formation of AC^CR0 (8) and the Li
acetylide 9. The increased nucleophilicity of the LiC^C function in 9
(compared to 5) facilitated the intramolecular SN2 reaction of its C-
catalyst 6 or the chain carrier 9. Using our carbenoid chain reaction
type 1, we will now examine sterically encumbered, unsaturated
Li,Cl-carbenoids such as 2 for their propensity to FBW ring expan-
sion and for the rate-limiting chain steps.
2. Results and discussion
2.1. Syntheses and FBW ring expansion in the cyclopentylidene
series
b
center with stereoinversion7 of the attacked carbon center of the
CeY bond, remarkably accompanied by the fixation of Y¼iodine in
the developing Li,Y-carbenoid 10.7,9 With Y¼iodine in 10, mono- or
di-iodides 7 may result through the final transfer of A (¼H or I,
respectively) in two ways: (i) delivery of A from the coproduct
AC^CR0 (8) will regenerate LiC^CR0 (6) as the catalyst8 of the
process; for A¼deuterium in the substrate 5 and LiN(i-Pr)2 (LDA) as
the catalyzing base,7 a corresponding transfer of A¼D may occur
from the byproduct DN(i-Pr)2. (These catalytic modes are unavail-
able for our carbenoid chain type-1 in Scheme 1 where R1A with
A¼H or D cannot regenerate the necessary strong base R1Li). (ii)
Alternatively, a transfer of A (¼H or I) directly from the starting
material 5 to carbenoid 10 with formation of product 7 and the
chain carrier 9 will maintain a carbenoid chain process. Both of
these processes may be outrun by the following deviating choices
made by carbenoid 10.
The rather unstable carbenoid LiCHCl2 was generated in THF
(Scheme 3) at ꢀ70 ꢁC from an excess of dichloromethane with LDA
in the presence of 2,2,5,5-tetramethylcyclopentanone11 (15), pro-
ducing the lithium alkoxide 16a. This primary product must be kept
and finally protonated at internal temperatures well below ꢀ40 ꢁC
to give 16b, because 16a is prone to eliminate LiCl with formation of
an a-chlorooxirane and thence an a-chloroaldehyde, as exempli-
fied12 previously with the related alkoxide t-Bu2C(OLi)eCHCl2 that
is burdened with even stronger internal strain. The ensuing de-
hydration of 16b with SOCl2 furnished purer (namely, crystalline)
samples of the
a,a-dichloroalkene 17 than had been obtained
previously13 by a different procedure.
CHCl2
Cl
Cl
1
O
RO
Me2
(R = H)
SOCl2
< 40 °C
-
Me2
3
Me2
Me2
2
CH2Cl2
+ LDA
Me2
Me2
C
Y
C
A
4
5
A
THF
16a: R = Li
b: R = H
X
X
R
15
Li
C
C
R
´
R
+
17
Y
6
5
Scheme 3. Synthesis of substrate 17.
7
+
6
X = oxygen
A = H, I
Y = I
The vinylic nucleophilic substitution (SNV) reaction (Scheme 4)
of the unactivated dichloroalkene 17 by 2,6-dimethylphenyllithium
(19), devised to produce the chloroalkene 23, should not be ex-
pected to proceed via an additionerotationeelimination (ARE)14
mechanism with elimination of LiCl from the intermediate 18, be-
cause the negative charge at C-2 of 18 would be strongly destabi-
lized by the two flanking tert-alkyl substituents that shield C-2
+
5, -9
or + 8, -6
A
C
C
R
´
(8)
+
C
Y
C
Li
Li
C
Li
X
R
+ 6, - LiY
X
R
X
R
X = oxygen
Y = I, OTs
Y
C
9
10
11
R
´
FBW
Me
Me
+
Me
Me
Me
Me
+ 5, -9
X = CH2
-
LiY
Cl
Cl
Cl
or +
8,
-
6
Li
Li
-
20
19
step 1
2
Li
A
C
Me2
Me2
X
+ LiR
´´
+
+
R
´´
18
Cl
Cl
Li
Cl
α
α
R
R
R
C
12
R
13
14
´
3
1Me2
5
Me2
Me2
2
Me2
Scheme 2. Examples of the type-2 carbenoid chain reaction.
4
21
17
19 + 21
With X¼CH2 in 10 (Y¼I or O3Searyl),7,10 the cyclohexyne de-
rivatives 13 were generated through FBW ring expansion.7,9 These
short-lived intermediates 13 were intercepted through cycloaddi-
tion10 to 1,3-diphenylisobenzofuran or through nucleophilic addi-
tion of R00Li (¼6 or 9 or n-butyllithium);7,10 the latter additions
afforded cyclohexenyllithiums 12, which may reimburse the cata-
lyst 6 (from 8) or the chain carrier 9 (from 5) through the A transfer
reactions as above. With the FBW process retarded by X¼oxygen,
carbenoid 10 may choose to be trapped by nucleophiles such as
reagent LiC^CR0 (6); the resultant alkenyllithium 11 can likewise
undergo the A transfer reaction to create product 14 along with the
step 2
fast
chain
21
-
LiCl
-
Me
Me
Me
α
α
Li
Cl
3
H
3
Me
Me
Me
step 3
17
1
1
H3C
H C
CH3
CH
H3C
H C
CH3
H3C
H C
CH3
2
2
+
CH
CH
4
5
4
5
22
23
24
Scheme 4. The carbenoid chain reaction in THF as the solvent.

R. Knorr et al. / Tetrahedron 70 (2014) 2703e2710
2705
against stabilizing solvation. Instead, a Cl/Li interchange reaction
Cl
Li
α
step 2
fast
Me
Me
(step 1) will initiate a carbenoid chain reaction (type 1) by the
transfer of A¼chlorine from 17 to 19 with formation of the Cl,Li-
carbenoid 21, as witnessed by the detection of chloro-2,6-
dimethylbenzene (20) as a coproduct. SNV reactions at Li,Hal-
carbenoids usually1,3 are very rapid, so that 19 can intercept (step
2) the short-lived carbenoid 21 before it decays. The resultant,
strongly basic alkenyllithium 22 can propagate the chain (step 3)
through acceptance of a chlorine particle from the substrate
dichloroalkene 17, creating the product 23 and the chain carrier
21 (The coproduct 20 did not transfer its chlorine to 22 and sur-
vived, therefore.). Clearly, step 3 rather than the much faster1,3 step
2 will limit the velocity of running carbenoid chain cycles.
An undisturbed carbenoid chain reaction in Scheme 4 would
require steps 2 and 3 to be much faster than step 1, so that only
slightly more than 1 equiv of reagent 19 would suffice1 for the total
conversion of dichloroalkene 17 into product 23 (chain steps 2 and
3) along with merely a trace of coproduct 20 (step 1). In contrast,
a reverse rate sequence (steps 1 and 2 being faster than step 3)
would lead to the consumption of up to 2 equiv of reagent 19 per
mol of 17 (steps 1 and 2) and an accumulation of the intermediate
alkenyllithium 22 along with one entire equivalent of 20: product
23 would then not be obtained from 22 unless a residual amount of
17 was available for a delayed step 3. Hence, the observation of
substantially more product 23 than coproduct 20 can provide evi-
dence for the carbenoid chain mechanism (faster step 3).
chain:
step 3
Li
Me2
α
Cl
Me
Me2
Me2
Me
+
19
3
Me2
1
Me2
2
-
LiCl
+
17
21
Me2
21
-
4
5
22
FBW
LiCl
23
2 x
step 2
´
-
-
2 Li
~H
1
6
2
3
m
Me2
Me2
Me2
1
p
5
Me o
+
19
Me
26
4
4
5
m
Li
α
fast
3
o
Me
2
Me
Me2
25
p´
Me2
Me2
m´
o´
m´
chain: step 3´
o´
Me
leakage
Me
+
17
α´
Me2
2´
H
-
21
1´
3´
Me2
Me
Me
5´
Cl
H
4´
Me
Me
1
2
1
2
27
3
6
3
6
Me2
Me2
Me2
Me2
4
5
4
5
28
29
Scheme 5. The carbenoid chain reaction (steps 2 and 3) competing in t-BuOMe as the
solvent with FBW ring expansion (step 20) whose chain is propagated in step 30.
In THF as the solvent, the dichloroalkene 17 was consumed by
reagent 19 (1.4 equiv) at room temperature (rt) within less than
40 min. The crude mixture of nonacidic products was shown by 1H
NMR to contain 20, 23, and 24 in a ca. 2:6:1 ratio. The olefin 24
arose through chain terminations by the slow proton transfer
(within ca.1 h at rt) from THF to 22 in the aprotic milieu maintained
by reagent 19; it did not arise during the aqueous workup, as
confirmed through final quenching with solid CO2, which served to
detect residual portions of reagent 19 in the form of 2,6-
dimethylbenzoic acid but uncovered only small portions of the
carboxylic acid that derived from the equally small amounts of
surviving 22. Therefore, the main product 23 revealed the opera-
tion of a carbenoid chain type 1 process with a rather slow chain-
propagating step 3 that was disturbed through competition by
step 1 (coproduct 20 augmented) and by slow chain termination
(‘leakage’).
was then closed through a chlorine particle transfer from 17 to 26
(step 30), which furnished the FBW product 28 and the chain carrier
21. Step 30 occurred in close competition with a proton transfer
(‘leakage’) that afforded the ring-expanded olefin 29, as shown by
the product ratio of ca. 1:6:2:1 for 20, 23, 28, and 29 (Control ex-
periments confirmed that 29 was not formed from the chlor-
oalkene 28 in an organolithium milieu.). Due to its poor yield, 29
could not be purified but was identified in the mixtures by means of
its NMR spectra. The structural discrimination of the isomeric
chloroalkenes 28 and 23 was simple on the basis of 1H and com-
plete 13C NMR signal assignments: selective proton decoupling by
irradiation of 3-CH3 simplified the 13C multiplet of C-2 in 28 to
a triplet with a 3JC,H¼4.5 Hz coupling to CH2-4 (Scheme 5), whereas
the chlorine-bearing C-
a in 23 cannot have and did not show such
a
3JC,H splitting. The small amount of coproduct 20 permitted to
The carbenoid chain mechanism developed in Scheme
remained valid in tert-butyl methyl ether (t-BuOMe) as the sol-
vent, using the dichloroalkene 17 and equiv of 2,6-
4
conclude that both 23 and 28 were formed via carbenoid chain
pathways, which implies that steps 3 (23) and 30 (28) were faster
than step 1 (coproduct 20) in their competition for the substrate 17.
This demonstrates a relative rate sequence of step 2[(steps 3 and
30)>step 1.
2
dimethylphenyllithium (19): the chloroalkene 23 was the main
product and again accompanied by some chloro-2,6-
dimethylbenzene (20) as an indicator of step 1. However, warm-
ing to 42 ꢁC during at least 23 h was now necessary for the com-
plete consumption of 17. One of the reasons for this substantial
retardation in t-BuOMe may be seen in the differing aggregational
states of reagent 19: since 19 in THF is known to be monomeric15 at
ꢀ125 ꢁC and dimeric16 at ꢀ60 ꢁC with both forms in a mobile
equilibrium, one can expect 19 in the weaker donor solvent t-
BuOMe to be more extensively aggregated and hence presumably
less reactive in steps 1 (initiation) and 2 (SNV), which depend on the
reactivity of 19. Such a deceleration of step 2 apparently encour-
aged an FBW ring expansion (step 20 in Scheme 5) of 21, which
created the strained 3,3,6,6-tetramethylcyclohexyne (25). Never-
theless, step 2 must still be very fast so as to trap a portion of the
short-lived carbenoid 21 before it expands to give 25. Although
more shielded than the ‘parent’ cyclohexyne (13 if R¼H in Scheme
2), 25 was intercepted by 19 to afford the alkenyllithium 26 at
a high rate, as judged from our failure to find the ring-opened ‘di-
mer’17 of 25 that was reported18 to be formed from 25 in a matrix
milieu even at ꢀ228 ꢁC. The extended (FBW) carbenoid chain cycle
Is step 1 the only kinetic ‘bottleneck’ in Scheme 4 (and in the
prelude to Scheme 5)? The above-mentioned (6þ2):1 ratio of
chlorine-bearing products (23 and 28) and the leakage product 29
points to short reaction chains whose termination occurred on
leakage of the intermediate 26 and implied a corresponding loss of
the possibility to propagate the chains via step 30 with regeneration
of the chain carrier 21 for further passes through both steps 2 and
20. Without such occasional withdrawal of 26 from the circulation,
the parallel reaction chains of intermediate 22 would have been
substantially longer: Because 22 is persistent in t-BuOMe and only
a very small amount of the nonexpanded olefin 24 was detected
during experiments in the spirit of Scheme 5, 22 in t-BuOMe ob-
viously can wait for the substrate 17 with which to generate 23 in
step 3. In such a situation of an undisturbed chain reaction, we
would have been allowed to conclude that an observed sluggish
overall rate was codetermined by a slow step 3 as a second kinetic
‘bottleneck’. Regrettably, the withdrawal of 26 impaired a correct
assessment of step 3; but direct evidence for a slow step 3 in
a similar system will be presented below in Section 2.2. In the

2706
R. Knorr et al. / Tetrahedron 70 (2014) 2703e2710
meantime, a certain accumulation of 22 (caused presumably by
a retarded step 3) may be inferred from the detection of a small
amount of the hydrocarbon 27 that can be thought to arise through
an oxidative dimerization of accumulated 22, followed by a formal
p0-to-a0 hydrogen shift.19 The molecular formula (C36H50) of 27
became evident through its prominent Mþ mass spectral peak,
which was accompanied by an Mþþ1 satellite (2H plus 13C) of 40.8%
intensity (calculated 40.35%). Starting with the only olefinic proton
The crude product acquired from 1c with p-tolLi contained the
chloroalkene 33 (isolated yield 29%)22 and residual 1c but no
‘leakage’ olefin23 31, which points to barely disturbed chain re-
actions. In addition, the ring-expanded FBW product (36) (yield 4%)
was detected in the previously20 ignored last chromatographic
fractions; it is now analyzed and shown in the sequel to be gen-
erated with the help of a slow step 3 of the carbenoid chain reaction
type 1. An X-ray diffraction analysis at 23 ꢁC revealed the solid-state
topology of 36 (but not the mode of its formation) as displayed in
NMR resonance (a
0-H), all 12 proton signals of 27 were assigned; on
this basis, the constitution of 27 followed from complete structural
assignments of all 26 13C resonance positions (6ꢂCH3, 4ꢂCH2,
4ꢂquaternary sp3, 4ꢂsp2-CH, and 8ꢂquaternary sp2).
Fig. 1: The plane of the central double-bond C2/C14 (C-2/C-a) en-
closes interplanar angles of 69ꢁ and 73ꢁ with its two substituents at
C14 (p-tolyl and the 2-naphthyl-derived chloroalkene, re-
spectively). Internal strain created through repulsion of these two
C-a
substituents by the 1,1- and 3,3-dimethyl groups (C10e13) may
/C-
2.2. FBW ring expansion in the indan-2-ylidene series
be read from the shrinked sp2-angle C15/C14/C23 (C-ipso/C-
a
20¼112.0ꢁ) and from the unequal angular distortions of the two 3-
CH3 groups (C12 and C13) from local tetrahedral symmetry. Dis-
solved 36 maintains its chirality (Fig. 1) on the 400-MHz NMR
timescale, as shown by the nonequivalence of all nine methyl
groups in the fully assigned 1H and 13C NMR spectra.
Here we are completing the description of a previous20 exper-
iment (steps 1, 2, and 3 in Scheme 6) that produced the chlor-
oalkene 33 from dichloroalkene 1c (1.5 equiv) with the reagent p-
methylphenyllithium (ptolLi) via the Cl,Li-carbenoid 2. This con-
sumption of p-tolLi in t-BuOMe as the solvent took 2e4 h at 37 ꢁC,20
which reaction period is practically equal to that of 2,6-
dimethylphenyllithium (19) with 1c, namely,21 3 h at 33 ꢁC with
production of the benzo derivative21 of 23. Comparison with the
sluggish conversion of substrate 17 in t-BuOMe, as reported in the
preceding section, reveals 1c to be more reactive than 17, perhaps
due to stronger steric shielding of C-
are more inclined toward C- because of the longer C-4/C-5 bond in
17 as compared to the shorter C-3a/C-7a bond in 1c (Scheme 6).
a by the methyl groups, which
a
ptol
Me2
ptol
Me2
Cl
Br
H
Cl
α
Me2
Me2
Me2
3a
Me2
7a
1c
31
30
Fig. 1. X-ray structure of 36 whose nonhydrogen atoms are shown with their crys-
tallographic numbering.
+ tBuBr ?
slower + ptolLi
+ tBuLi
fast
-
tBuBr
-
ptolCl
step 1
Li
Cl
Me
m
ptol
Me2
Li
Scheme 6 depicts the route from 1c to the FBW product 36
(steps 1, 20, and 30). In contrast to the interception of cyclohexyne
derivative 25 (Scheme 5) by 19, the similarly shielded cycloalkyne
34 was not trapped by the aryllithium reagent p-tolLi. This may be
due to a retarded ring expansion of the indanylidene carbenoid 2
(step 20) as compared with the more flexible cyclopentylidene
carbenoid 21. The alternative possibility of an accelerated carbenoid
SNV reaction (step 2) with the sterically unencumbered p-tolLi
appears less obvious in consideration of the higher aggregation
states16 of p-tolLi in THF (dimeric) or in Et2O (tetrameric), as
compared with 19,16 so that p-tolLi might be expected to be more
highly aggregated (and perhaps less reactive) than 19 also in the
present solvent t-BuOMe. With p-tolLi concentrations dwindling
anyway in later reaction stages, however, step 2 will slow down so
that step 20 can begin to compete: The emerging, short-lived
cycloalkyne 34 added quickly to the meanwhile accumulated
alkenyllithium 32,24 creating the strongly basic intermediate 35
that completed (step 30) the deviating FBW chain cycle through
a chlorine particle transfer from the surplus dichloroalkene 1c to 32
with formation of 36 and the chain carrier 2. This pathway was
confirmed through an independent generation of the intermediate
32 from the known25 bromoalkene 30 with t-BuLi (2 equiv) in t-
BuOMe, followed by addition of 1c (1 equiv) and observation of the
consumption of 32 in situ26 for 2 h. After quenching with solid CO2
and aqueous workup, the crude material (no carboxylic acids
p
o
m
fast
i
Me2
Me2
chain:
slow
step 3
α
Cl
o
Me2
step 2
1
Me2
3Me2
2
+ ptolLi
-
LiCl
+
1c
2
2
-
32
fast
step 2
FBW
LiCl
´
-
33
fast
34
7´
6´
Me2
Me2
8´
Me
m
p
8a´
o
5´
Me2
Me2
34
m
1´
4a´
4´ Me2
i
o
ptol
Me2
α
Me2
2´
3´
Cl
FBW
chain:
step 3´
Li
Me2
Me21
3Me2
2
7a
7
3a
4
+
1c
2
-
6
5
36
35
Scheme 6. Dichloroalkene 1c (step 3) and cycloalkyne 34 reacting with the alke-
nyllithium 32 in t-BuOMe as the solvent (p-tol¼p-methylphenyl).

R. Knorr et al. / Tetrahedron 70 (2014) 2703e2710
2707
detected) was a mixture of 1c, 31, 33, and 36 (ca. 44:30:22:4) in
rough agreement with the 33/36 ratio of 29:4 cited above. The
unexpected olefin23 31 is thought to have resulted through proton
transfer to 32 from part of the coproduct t-BuBr (Scheme 6). We
estimated the first half-reaction time of this separated step 3 to be
in the order of 20 min at rt in t-BuOMe; the corresponding overall
reaction period resembles that of the above-mentioned overall
reaction (120e240 min at 37 ꢁC) of p-tolLi with 1c. Hence, this
chain-propagating step 3 was not fast and had limited the rate of
the overall process in (partial) cooperation with the slower step 1.
Of course, the addition reaction of 32 to the short-lived cycloalkyne
34 must be fast; but this cannot suppress the slow step 3 because of
the low concentration of 34. Comparably slow steps 3 were pre-
viously27 encountered for the carbenoid chain reactions of 1c with
an excess of several other aryllithiums in t-BuOMe or Et2O as the
solvents, where alkenyllithium intermediates (corresponding to
32) were also observed (1H NMR in situ) but did not permit an
unequivocal interpretation.
argon gas cover, then tightly closed with a solvent-stable soft
rubber stopper, sealed with ParafilmÒ that was wrapped around
the stopper, and stored at ꢀ18 ꢁC in a large Schlenk tube filled with
argon. 1H and 13C NMR chemical shifts
d were referenced to tetra-
methylsilane as an internal standard. Chromatographic separations
ꢀ
used silicagel (60 A, 100e200
mm). Combustion analyses were
carried out in the departmental service unit.
4.2. 1-(Dichloromethyl)-2,2,5,5-tetramethylcyclopentanol (16b)
A freshly prepared solution of lithium N,N-diisopropylamide
(LDA, 214 mmol) in anhydrous THF (35 mL) was added dropwise
within 35 min to the stirred solution of dichloromethane (18.3 mL,
285 mmol) and 2,2,5,5-tetramethylcyclopentanone11 (15, 10.00 g,
71.3 mmol) in anhydrous THF (50 mL) at ꢀ70 ꢁC. The orange-
brownish mixture was stirred for another 60 min at up to ꢀ40 ꢁC
internal temperature and then quenched at ꢀ40 ꢁC by the slow
addition of glacial acetic acid (25 mL) in THF (25 mL). On warming
up to 0 ꢁC, the mixture deposited a crystalline precipitate was
poured into ice-cold aqueous hydrochloric acid (2 M, 500 mL), and
was shaken with Et2O (4ꢂ100 mL). The combined aqueous layers
were washed with distilled water until neutral, dried over Na2SO4,
and concentrated to leave a brown liquid (14.86 g) containing 16b
3. Conclusion
The unactivated, cyclic a,a-dichloroalkenes (tert-alkyl)2C]CCl2
(1c and 17) can react with organolithiums via unsaturated Cl,Li-
carbenoids (tert-alkyl)2C]C(Li)Cl (2 and 21) in carbenoid chain
cycles of alternating SNV (very fast3 step 2) and Cl/Li interchange
(slower step 3) events. The solvent-dependent velocity of step 3
appears to be at least partially determinative for the preparative
productiveness and the overall reaction periods: the primary SNV
products (tert-alkyl)2C]C(Li)earyl (22 and 32) in the solvent THF
were labile (‘leakage’) but sufficiently reactive toward the sub-
strates (17 and 1c, respectively) despite steric shielding, whereas
they were durable in t-BuOMe and revealed that product formation
by the chain-propagating step 3 can be a strikingly sluggish pro-
cess. There may even be other cases in which the initiating carbe-
noid formation (step 1) remains competitive with (or faster than)
step 3, so that it may partially (or totally) thwart the chain mech-
anism by means of an exhaustive consumption of the substrate
(tert-alkyl)2C]CCl2 that should have carried out step 3. For pre-
parative purposes, this nonchain carbenoid possibility (steps 1 and
2 only) may be either ameliorated by an excess of the dichlor-
oalkene substrate or made productive through quenching with
another finally introduced external organic chlorine supplier.
Ring expansions are the main reaction mode of unsaturated
Hal,K-carbenoids that were generated from (halogenomethylidene)
cyclobutane derivatives3,28 in a nonpolar milieu. We have pre-
sented here an example of our indan-2-ylidene Cl,Li-carbenoids as
being reasonably resistant against this undesired FBW reaction. In
contrast, one of our cyclopentylidene Cl,Li-carbenoids revealed that
the very facile SNV step can be partially disturbed by competing
FBW ring expansion in t-BuOMe (but less so in THF) as the solvent.
The FBW products can result from trapping of the cycloalkynes by
the organolithium reagent (ca. 20% of 28) or by the chain-
propagating (or an independently generated) alkenyllithium such
as 32 (ca. 4% of 36). It follows that substantially less reactive nu-
cleophiles than aryllithiums may not always undergo clean SNV
reactions (namely, avoiding the FBW complication) with our un-
saturated Hal,Li-carbenoids; examples were already described29 for
the corresponding Cl,K-carbenoids (tert-alkyl)2C]C(K)eCl.
together with the corresponding spirocyclic
a-chlorooxirane
(11%) (The latter side-product would have been avoided through
working at less than ꢀ40 ꢁC and with a shorter dwell time). Dis-
tillation at 35e52 ꢁC (bath temp 75e95 ꢁC)/0.015 mbar furnished
16b as a yellowish liquid (10.85 g, 68%) that contained a trace of the
corresponding 1-chloro-1-formyl-2,2,5,5-tetramethylcyclopentane
and crystallized slowly. Mp 38.5e41 ꢁC (from pentane at ꢀ70 ꢁC);
1H NMR (CDCl3, 400 MHz)
d
1.19 and 1.21 (2s, 2ꢂ6H, 2ꢂCH3), 1.48
and 1.72 (2m, 2ꢂ2H, CH2eCH2), 2.34 (s, 1H, OH), 6.07 (s, 1H,
CHCl2) ppm; 13C NMR (CDCl3, 100.6 MHz)
d
25.77 (qm, 1J 126 Hz,
apparent 3J 2.5 Hz, 2ꢂCH3), 25.92 (qm, 1J 126 Hz, 3J unresolved,
2ꢂCH3), 40.03 (tsept, 1J 130 Hz, 3J 4.5 Hz, 2ꢂCH2), 48.41 (un-
resolved, C-2/-5), 79.80 (sharp d, 1J 174 Hz, CHCl2), 85.44 (un-
resolved, C-1) ppm; IR (KBr) n 3589 and 3568 (2 sharp OeH), 2951,
2878, 1469, 1382, 1123, 1028, 992, 771, 747, 735, 718 cmꢀ1. Anal.
Calcd for C10H18Cl2O (225.16): C, 53.35; H, 8.06; Cl, 31.49. Found: C,
53.61; H, 8.15; Cl, 31.75.
4.3. 2-(Dichloromethylidene)-1,1,3,3-tetramethylcyclopentane
(17)
A solution of the alcohol 16b (15.0 g, 66.6 mmol) in distilled
pyridine (75 mL) was stirred in an ice-bath during the slow,
dropwise addition of thionyl chloride (13.10 mL, 180 mmol). After
stirring at rt overnight, the dark brown suspension was poured into
ice-cold aqueous HCl (2 M, 450 mL) and shaken with Et2O
(4ꢂ150 mL). The combined organic layers were washed with
aqueous HCl (2 M, 2ꢂ100 mL) and then with distilled water until
neutral, dried over Na2SO4, and concentrated at rt to afford almost
pure 17 (8.91 g, 65%). This orange-brown, volatile liquid was filtered
through silicagel (80 g) with low-boiling petroleum ether/Et2O
(98:2) and concentrated at rt to leave a colorless liquid (7.41 g, 54%)
that crystallized slowly in a refrigerator: colorless needles with mp
23.5e26 ꢁC (pentane at ꢀ70 ꢁC; Ref. 13: no mp); 1H NMR (CDCl3,
400 MHz)
NMR (CDCl3, 100.6 MHz)
d
1.27 (s, 12H, 4ꢂCH3), 1.62 (s, 4H, CH2eCH2) ppm; 13C
d
26.13 (4ꢂCH3), 41.01 (2ꢂCH2), 46.99 (C-
4. Experimental section
4.1. General information
1/-3), 113.57 (CCl2), 155.06 (C-2) ppm, assigned through compari-
son with 2-(dichloromethylidene)-1,1,3,3-tetramethylindane (1c,
which is compound 2c in Ref. 1), but in disagreement with the eight
(rather than five) 13C signals reported13 previously for 17; IR (film)
n
All organometallic procedures were carried out under a cover of
slowly streaming, dry argon gas. The in situ experiments were
performed in dried NMR tubes (5 mm) that were charged under
2958, 2869, 1605, 1591, 1460, 1366, 908, 849 cmꢀ1. Anal. Calcd for
10H16Cl2 (207.14): C, 57.98; H, 7.79; Cl, 34.23. Found: C, 58.04; H,
7.74; Cl, 33.80.
C

2708
R. Knorr et al. / Tetrahedron 70 (2014) 2703e2710
4.4. 2-(
cyclopentane (23)
a
-Chloro-o,o-dimethylbenzylidene)-1,1,3,3-tetramethyl
selective {1H} decoupling as follows: {o-CH3}/o-CH3 as a dis-
turbed q with d 3J 4.1 Hz, C-m as a sharp dd with 1J 156 and 3J
7.6 Hz, C-ipso as a td with 3J 6.6 and 2J 1 Hz, C-o simplified to
a peculiar dm (doublet next to a dt) that was successfully simu-
lated with the following set of coupling constants: d 3J 7.6 Hz to p-
A solution of pure bromo-2,6-dimethylbenzene (4.50 mL,
34.5 mmol) in anhydrous THF (75 mL) was stirred at ꢀ70 ꢁC
during the slow addition of tert-butyllithium (t-BuLi, 68 mmol) in
hexane (39.8 mL). The solution was stirred at rt for 30 min so as to
destroy the byproduct t-BuBr30 and then recooled to ꢀ70 ꢁC for
the dropwise addition of dichloroalkene 17 (5.00 g, 24.1 mmol)
dissolved in anhydrous THF (25 mL). The black solution was stir-
red at rt for 40 min and poured onto solid CO2. The warmed-up
mixture was treated with aqueous NaOH (2 M, 500 mL) and
shaken with Et2O (4ꢂ75 mL). The combined organic layers were
washed with distilled water until neutral, dried over Na2SO4, and
concentrated at rt to afford a yellow liquid (6.65 g) containing 23,
the olefin 24, and chloro-2,6-dimethylbenzene (20, dH 2.37 ppm,
boiling at ca. 30 ꢁC/0.02 mbar) in a molar ratio of 6:1:2 together
with a tiny trace of 29 as the only products. Distillation at
76e92 ꢁC (bath temp 95e145 ꢁC)/0.05 mbar furnished a mixture
of 23 (49%) and 24 (11%), which was redistilled to provide pure 23
H, d 3J 2.5 Hz to
a
-H, d 2J 0.9 Hz to one of the m-H, d 4J ꢀ1.2 Hz to
3
4
the other m-H, JH,H 7.3 Hz, JH,H þ1.5 Hz; {1-CH3}/t (ca. 4 Hz)
coupling of 1-CH3 with CH2-5, C-1 as a dm with 3J 8.6 Hz trans to
a
-H, CH2-5 simplified; {3-CH3}/t (ca. 4 Hz) coupling of 3-CH3
with CH2-4, C-3 as a dm with 3J 5.5 Hz cis to
-H, CH2-4 simplified;
-H}/C-o as a pseudo-qi that could be simulated with the above
set of JC,H and JH,H values, but including 2J 5.8 Hz to o-CH3 and
excluding -H; {all aryl-H}/o-CH3 as a sharp q, C-o as a broad-
a
{
a
a
ened q with 2J 5.8 Hz to o-CH3, C-ipso as an m with 3J 4.2 Hz. The
unusual appearance of several proton-coupled 13C NMR reso-
nances was confirmed by simulations with the following appro-
priate coupling constants: for C-o with the above set of JC,H and
JH,H values and again including 2J 5.8 Hz to o-CH3; for o-CH3, 1J
126.2 Hz, 3J 4.1 Hz to one of the m-H, 5J þ0.7 Hz to the other m-H,
4J þ0.3 Hz to p-H, 3JH,H 7.3 Hz, 4JH,H þ1.4 Hz; for C-m, d 1J 156.0 Hz,
as a yellow liquid. 1H NMR (CDCl3, 400 MHz)
d
0.78 (s, 6H, 2ꢂ1-
d
3J 7.6 Hz, q 3J 4.9 Hz to one of the two o-CH3, 2J 1.2 Hz, JH,H
4
CH3), 1.48 (s, 6H, 2ꢂ3-CH3), 1.53 (t, 3J 7 Hz, 2H, CH2-5), 1.66 (t, 3J
7 Hz, 2H, CH2-4), 2.33 (s, 6H, 2ꢂo-CH3), 7.05 and 7.10 (AA0B sys-
tem, 3J 7.5 Hz, 2Hþ1H, 2ꢂm-H and p-H) ppm, assigned through
the following {1H} decoupled NOE difference spectra: {1-CH3}/
CH2-5 and o-CH3, {3-CH3}/CH2-4 only; 13C NMR (CDCl3,
þ1.4 Hz; for C-ipso, t 3J 6.6 Hz, sept 3J 4.3 Hz, d 2J 1 Hz, 4J 0 Hz, 3JH,H
7.3 Hz; IR (film)
n .
3062 (w), 2954, 2865, 1459, 1362, 767 cmꢀ1
Anal. Calcd for C18H26 (242.41): C, 89.19; H, 10.81. Found: C, 88.64;
H, 11.19.
100.6 MHz)
d
19.97 (qm with reduced intensity, 1J 126.5 Hz, 2ꢂo-
4.6. 20-{o0,o0-Dimethyl-p0-[o,o-dimethyl-
a-(1,1,3,3-tetramethyl
CH3), 26.85 (qsext, 1J 126 Hz, 3J 4.5 Hz, 2ꢂ1-/3-CH3), 41.08 (tm, 1J
129 Hz, CH2-4), 41.52 (tm, 1J 129 Hz, CH2-5), 46.54 (m, apparent J
3.8 Hz, C-3), 46.58 (m, apparent J 3.8 Hz, C-1), 123.53 (sharp s, C-
cyclopent-2-ylidene)benzyl]benzylidene}-10,1,030,30-tetrame-
thylcyclopentane (27)
a
), 127.30 (ddq, 1J 157.5 Hz, 3J 7.5 Hz, 3J 4.8 Hz, 2ꢂC-m), 127.86
Although this product of an oxidative dimerization of 22 was not
visible by 1H NMR in the crude product mixture described further
below, it was traced down as follows. The pot residue that remained
after a distillation of the chloroalkene 28 was prepurified by
chromatography on silicagel with low-boiling petroleum ether,
affording a mixture (82 mg) of 23, 28, and 27 that deposited small
needles (30 mg, 1%) of 27 from hot methanol (5 mL): mp
(sharp d, 1J 159 Hz, C-p), 136.49 (dq, 3J 7.4 Hz, 2J 6.2 Hz, 2ꢂC-o),
139.41 (m, C-ipso), 152.94 (m, 3J 3.3 Hz, C-2) ppm, assigned
through the following selective {1H} decoupling experiments: {1-
CH3}/1-CH3 simplified; {3-CH3 and CH2-4/-5}/3-CH3 simpli-
fied, CH2-4 decoupled from 3-CH3, CH2-5 as a narrow m, and C-3
as a sharp s; IR (film)
n 2954, 2868, 1635 (w), 1461, 1363, 770,
723 cmꢀ1. Anal. Calcd for C18H25Cl (276.85): C, 78.09; H, 9.10; Cl,
12.81. Found: C, 78.59; H, 9.16; Cl, 11.39.
213e214 ꢁC (3ꢂ from methanol); 1H NMR (CDCl3, 400 MHz)
d 0.88
(s, 6H, 2ꢂ10-CH3), 0.91 (s, 6H, 2ꢂ1-CH3), 0.99 (s, 6H, 2ꢂ3-CH3), 1.19
(s, 6H, 2ꢂ30-CH3), 1.51 (m, 4H, CH2-4 and CH2-50), 1.54 (m, 4H, CH2-
40 and CH2-5), 2.13 (s, 6H, 2ꢂo0-CH3), 2.45 (s, 6H, 2ꢂo-CH3), 5.98 (s,
4.5. 2-(2,6-Dimethylbenzylidene)-1,1,3,3-tetramethylcyclo
pentane (24)
1H,
a
0-H), 6.95 (s, 2H, 2ꢂm0-H), 6.96 and 6.98 (AA0B system, 3H,
2ꢂm-H and p-H) ppm, assigned through the NOESY correlations a0
-
A
crude sample of the
a
-chloroalkene 23 (450 mg, ca.
H430-CH34CH2-40, 0-H4o0-CH3410-CH34CH2-50, o0-CH34m0-
a
1.6 mmol) in tert-butyl alcohol (20 mL) was refluxed with ele-
mental Liꢁ (2ꢂ140 mg, 2ꢂ20 mmol) for 2 h (all Liꢁ dissolved). The
mixture was diluted with Et2O (60 mL) and water (30 mL), and the
aqueous layer was extracted with more Et2O (10 mL). The com-
bined Et2O layers were washed with distilled water until neutral,
dried over Na2SO4, and concentrated to yield the crude liquid 24
(330 mg, ca. 1.36 mmol); bp 75e85 ꢁC (bath temp)/0.002 mbar. 1H
H43-CH34CH2-4, and m0-H4o-CH341-CH34CH2-5; 13C NMR
(CDCl3, 100.6 MHz)
d
20.97 (qm, 1J 126 Hz, 2ꢂo0-CH3), 22.56 (qm, 1J
126 Hz, 2ꢂo-CH3), 27.69 (qsext, 1J 125.2 Hz, 2ꢂ10-CH3), 28.28 (qsext,
1J 125.2 Hz, 2ꢂ1-CH3), 30.31 (qm, 1J 125.2 Hz, 2ꢂ3-CH3), 30.45 (qm,
1J 125.2 Hz, 2ꢂ30-CH3), 38.48 (tsept, J 128.4 Hz, CH2-40), 40.85
1
(tsept, 1J 128.4 Hz, CH2-50), 41.64 (tsept, 1J 128.4 Hz, CH2-5), 42.58
1
(tsept, J 128.4 Hz, CH2-4), 43.33 (unresolved, C-10), 44.81 (un-
NMR (CDCl3, 400 MHz)
d
0.87 (s, 6H, 2ꢂ1-CH3), 1.22 (s, 6H, 2ꢂ3-
resolved, C-30), 45.00 (unresolved, C-1), 45.79 (unresolved, C-3),
CH3), 1.52 (AA0 part of an AA0BB0 system, 2H, CH2-5), 1.57 (BB0
118.51 (sharp d, 1J 149 Hz, C- 0), 125.90 (sharp d, 1J 158.5 Hz, C-p),
a
part, 2H, CH2-4), 2.20 (s, 6H, 2ꢂo-CH3), 6.05 (s, 1H,
a
-H), 6.98 and
127.44 (ddq, 1J 155 Hz, 2ꢂC-m), 128.46 (ddq, 1J 155.5 Hz, 2ꢂC-m0),
7.04 (AA0B system, 3J 7.3 Hz, 3H, 2ꢂm-H and p-H) ppm, assigned
133.75 (t, J 3.5 Hz, C-p0), 134.65 (qd, 2J 5.4 Hz, J 2.3 Hz to m0-H,
2
2
through comparison with compound 38 in Ref. 31; 13C NMR
2ꢂC-o0), 135.73 (overlaid m, C-ipso0), 135.90 (apparent qi, 2ꢂC-o),
(CDCl3, 100.6 MHz)
d
20.79 (qm, see the {1H} decoupling and the
141.02 (sharp s, C-a), 142.69 (m, C-ipso), 154.24 (unresolved, C-2),
computer-simulated multiplets further below, 2ꢂo-CH3), 27.65
(qsext, 1J 125.3 Hz, 3J 4.2 Hz to 1-CH3 and CH2-5, 2ꢂ1-CH3), 30.53
(qsext, 1J 125.3 Hz, 3J 4.2 Hz to 3-CH3 and CH2-4, 2ꢂ3-CH3), 38.47
(tm, 1J 128.7 Hz, apparent J 4.8 Hz, CH2-4), 40.78 (tm, 1J 128.7 Hz,
apparent J 4.8 Hz, CH2-5), 43.37 (m, for 2J and 3J see {1H}, C-1),
44.77 (m, for 2J and 3J see {1H}, C-3), 118.47 (sharp d, 1J 149.2 Hz, C-
159.65 (unresolved, C-20) ppm, assigned through HETCOR and the
following selective {1H} decoupling experiments: {3-CH3}/CH2-4
as a t 126 Hz, C-3 and C-2 sharpened; {1-CH3}/CH2-5 as a t 128 Hz,
C-1 and C-2 sharpened; {30-CH3}/CH2-40 as a t 123 Hz, C-30 nar-
rowed; {o0-CH3}/C-m0 as a dd with 1J 155.5 Hz and 3J 6.8 Hz, C-o0 as
a d 2J 2.3 Hz, C-ipso as a t 3J 6.7 Hz; {o-CH3}/C-m as a ddd with 1J
ca. 155 Hz, 3J 7 Hz, and 2J 1.1 Hz, C-o as a d 3J 6.9 Hz, C-ipso as a sharp
t; {all tetramethylcyclopentylidene protons}/C-10 as a d 3J 9.3 Hz
a
), 126.05 (sharp d, 1J 158.0 Hz, C-p), 126.72 (ddq, see {1H}, 2ꢂC-
m), 136.29 (m, see {1H}, 2ꢂC-o), 137.91 (m, see {1H}, C-ipso), 159.64
(m, apparent J 3.4 Hz, C-2) ppm, assigned as above and through
(trans to a a
0-H), C-30 as a d 3J 5.8 Hz (cis to 0-H), C-1/-2/-3 as three

R. Knorr et al. / Tetrahedron 70 (2014) 2703e2710
2709
sharp s, C-20 as a d 2J 3.8 Hz to
a
0-H; {all aromatic H}/o-CH3 and o0-
4.9. 2-[
20-yl)-4-methylbenzylidene]-1,1,3,3-tetramethylindane (36)
a
-(30-Chloro-10,40-dihydro-10,10,40,40-tetramethylnaphth-
CH3 as two sharp q with 1J 126 Hz, C-m0 as a q 3J 4.5 Hz, C-o as a q 2J
5.7 Hz, C-p0 as a sharp s, C-ipso and C-ipso0 simplified; EIMS (70 eV)
m/z¼482.79 (Mþ) with (1Hþ13C) satellite 40.8% (calcd 40.35% for
This ring-expanded side-product was obtained20 from the car-
benoid chain reaction of dichloroalkene 1c (1.5 equiv) with 4-
methylphenyllithium (19) in t-BuOMe (not in THF) as the solvent.
After column chromatography on silicagel had afforded the ex-
C
36H50). Anal. Calcd for C36H50 (482.80): C, 89.56; H, 10.44. Found:
C, 89.01; H, 10.83.
pected
a
-chloroalkene 33,22 further elution of the column with
4.7. 2-Chloro-1-(o,o-dimethylphenyl)-3,3,6,6-tetramethylcyclo
hexene (28)
Et2O furnished an oily mixture (918 mg) that deposited cotton-like
needles of 36 (138 mg, yield 4%) from ethanol (25 mL); mp
228e229 ꢁC. 1H NMR (CDCl3, 400 MHz)
d
0.63 (s, 3H, 10a-CH3), 1.13
A
solution of pure bromo-2,6-dimethylbenzene (3.22 mL,
(s, 3H, 1a-CH3), 1.42 (s, 3H, 3a-CH3), 1.50 (s, 3H, 40a-CH3), 1.52 (s, 3H,
1b-CH3), 1.61 (s, 3H, 3b-CH3), 1.67 (s, 3H, 40b-CH3), 1.85 (s, 3H, 10b-
CH3), 2.36 (s, 3H, p-CH3), 7.08 (m, 2H, 4-/7-H), 7.11 (broadened d, 3J
7.8 Hz, 2H, 2ꢂm-H), 7.18 (m, 2H, 5-/6-H), 7.22 (m, 2H, 60-/70-H), 7.33
(m, 1H, 80-H), 7.36 (br, 1H, 1ꢂo-H), 7.40 (m, 1H, 50-H), 7.74 (br, 1H,
1ꢂo-H) ppm, assigned through HETCOR and COLOCS correlations
24.1 mmol) in anhydrous t-BuOMe (41 mL) was stirred at ꢀ70 ꢁC
during the slow addition of t-BuLi (48.3 mmol) in hexane (28.4 mL),
then kept at rt for 30 min so as to destroy the byproduct t-BuBr,30
and stirred at ꢀ70 ꢁC during the dropwise addition of the
dichloroalkene 17 (2.50 g, 12.1 mmol) in t-BuOMe (8 mL) within
20 min. The mixture was warmed at 40e45 ꢁC for ꢃ23 h, poured
onto solid CO2, warmed up, treated with aqueous NaOH (2 M,
150 mL), and shaken with Et2O (3ꢂ50 mL). The combined organic
layers were extracted with NaOH (2ꢂ), washed with distilled water
until neutral, dried over Na2SO4, and concentrated to yield a yellow
liquid (2.41 g) that contained a roughly 1:6:2:1 mixture of 20, 23,
28, and 29 accompanied by traces of further products. Column
chromatography on silicagel with low-boiling petroleum ether
afforded first 23 (1535 mg, ca. 46% containing some 29), followed
by the isomer 28 (490 mg, ca. 15%), which crystallized slowly from
pentane at ꢀ70 ꢁC or from methanol at rt: transparent rods, mp
58e59.5 ꢁC; bp 80e85 ꢁC/0.04 mbar. 1H NMR of 28 (CDCl3,
(see below); 13C NMR (CDCl3, 100.6 MHz)
d
21.24 (qt, 1J 126 Hz, 3J
4.4 Hz, p-CH3), [28.06 (3b-CH3), 28.22 (10b-CH3), 29.41 (1a-CH3),
29.80 (40b-CH3), 30.32 (40a-CH3), 34.09 (3a-CH3), 35.42 (1b-CH3),
38.75 (10a-CH3), altogether 8ꢂqq with 1J ca. 126 and 3J ca. 4.6 Hz],
41.49 (m, apparent J 3.5 Hz, C-10), 42.15 (m, apparent J 3.5 Hz, C-40),
49.09 (unresolved, C-3), 49.19 (unresolved, C-1), 121.11 and 122.25
(2 dm, 1J 156 Hz, C-7/-4), 124.48 (dm, 1J 157 Hz, C-80), 126.06 (dd, 1J
160 Hz, 3J 7.5 Hz, C-60), 126.48 (dm, 1J 157 Hz, C-50), [126.68 (C-6 or
C-5), 126.69 (C-70), 126.74 (C-5 or C-6), altogether 3ꢂdm with 1J
159 Hz], [127.14 (C-o), 129.02 (C-m), 129.78 (second C-o), 132.25
(second C-m), altogether 4ꢂbroadened d with 1J 158 Hz], 134.83 (t,
3
3J 3.5 Hz, C-
a
), 136.37 (>qi, J 4.8 Hz to 40-CH3, C-30), 136.86 (tq, 3J
400 MHz)
d
0.98 (s, 6H, 2ꢂ6-CH3), 1.25 (s, 6H, 2ꢂ3-CH3), 1.67 and
7.7 Hz, 2J 5.5 Hz, C-p), 140.58 (t, 3J 7.5 Hz, C-ipso), 141.62 (m, C-4a0),
1.84 (AA0MM0 system, 2Hþ2H, CH2-5 and CH2-4), 2.18 (s, 6H, 2ꢂo-
CH3), 7.03 and 7.07 (AA0B system, 2Hþ1H, 2ꢂm-H and p-H) ppm,
assigned through the NOESY correlations m-H4o-CH346-
CH34CH2-54CH2-443-CH3; 13C NMR (CDCl3, 100.6 MHz)
3
143.05 (m, J 4.5 Hz, C-20), 144.35 (m, C-8a0), [149.79 (C-3a) and
150.11 (C-7a), 2ꢂtm with 3J 5.5 and 3J 3.8 Hz], 156.23 (m, coupled to
CH3 only, C-2) ppm, assigned through the 1JC,H values, off-resonance
{1H} decoupling, and selective {1H} decoupling experiments as
follows: {p-CH3}/C-p as a t 3J 7.7 Hz to m-H; {broadened m-H}/p-
CH3 as a sharp q 1J 126 Hz, C-ipso as a s; {broadened o-H at dH
d
20.66 (qm, 1J 126.0 Hz, 2ꢂo-CH3), 28.35 (qsext, 1J 126.6 Hz, 3J
4.4 Hz, 2ꢂ3-CH3), 28.80 (qqt, 1J 126.3 Hz, 3J 4.8 Hz, 3J 4.5 Hz, 2ꢂ6-
CH3), 35.75 (tm, 1J 123 Hz, CH2-4), 36.00 (tm, 1J 123 Hz, CH2-5),
37.98 (m, C-3), 39.86 (m, C-6), 126.19 (sharp d, 1J 158.5 Hz, C-p),
127.28 (ddq, 1J 156.5 Hz, 3J 7.5 Hz, 3J 5 Hz, 2ꢂC-m), 136.19 (apparent
qi, 3J¼2J¼6.0 Hz, 2ꢂC-o with reduced intensity due to the pro-
longed relaxation time T1¼12ꢄ3 s), 138.67 (m, C-ipso), 139.35 (tm, 3J
3.5 Hz, C-1), 140.21 (tm, 3J 4.5 Hz, C-2) ppm, assigned through
HETCOR and the following selective {1H} decoupling experiments:
{6-CH3}/C-1 as a t 3J 3.5 Hz, C-6 sharpened, CH2-5 simplified; {3-
CH3}/C-2 as a t 3J 4.5 Hz, C-3 sharpened, CH2-4 simplified; {CH2-
5}/6-CH3 as a qq, C-6 sharpened; confirmed through HMBC ex-
periments showing the 3JC,H correlations 6-CH34C-14CH2-54C-
3, 3-CH34C-24CH2-44C-6, C-443-CH343-CH34CH2-4, C-
7.74 ppm}/C-
a as a s; {all aromatic H centered at dH 7.25 ppm}/
C-5/-6/-60/-70/-80/-
a
/-ipso as seven sharp s, C-2 and C-20 fully cou-
pled, C-50/-4/-7 as three shrinked d, C-10 and C-40 simplified, C-p as
a q 2J 5.5 Hz, p-CH3 as a sharp q, C-4a0 and C-8a0 as two m 3J 3.6 Hz to
10-/40-CH3, C-3a and C-7a as two m 3J 3.8 Hz to 1-/3-CH3, only C-o at
1
dC 129.8 ppm as a shrinked d with residual JC,H to o-H at dH
7.74 ppm; {10a-CH3 at dH 0.63 ppm}/10a-CH3 disturbed, 10b-CH3 as
a sharp q with 1J 126 Hz, C-10 as a dq coupled to 80-H and 10b-CH3, C-
20 as a q 3J 4.5 Hz to 10b-CH3; {all CH3 centered at dH 1.62 ppm}/C-
2/-20/-30 as 3s, C-7a as a t 3J 5.5 Hz to 4-/6-H, C-3a as a t 3J 5.6 Hz to
5-/7-H, C-8a0 as a broadened t (overlaid by residual 3J to 10a-CH3 at
dH 0.63 ppm) but C-4a0 as a t 3J 6 Hz to 60-/80-H, C-p as a broadened t
(overlaid by residual 2J to p-CH3), C-1/-3 narrowed, C-10 broadened
546-CH346-CH34CH2-5,
m-H4C-m4o-CH34C-ipso4m-
H4o-CH3, p-H4C-o, and the 2JC,H correlations 6-CH34C-64CH2-
5, 3-CH34C-34CH2-4, and o-CH34C-o. Anal. Calcd for C18H25Cl
(276.85): C, 78.09; H, 9.10. Found: C, 78.20; H, 9.00.
2
3
(residual J to 10a-CH3) but C-40 as a d J 3 Hz to 50-H. Final as-
signments were made by two-dimensional COLOCS experiments
3
(window 4 Hz) as follows. (a) JC,H correlations: 10a-CH34C-
20410b-CH3, 40a-CH34C-30440b-CH3, 10a-CH3410b-CH3, 10b-
CH3410a-CH3, 40a-CH3440b-CH3, 40b-CH3440a-CH3, 1a-CH341b-
CH3, 1b-CH341a-CH3, 3a-CH343b-CH3, 3b-CH343a-CH3, 4-
H4C-6, 7-H4C-5, C-445-/6-H4C-7, 50-H4C-70, 60-H4C-80,
4.8. 1-(o,o-Dimethylphenyl)-3,3,6,6-tetramethylcyclohexene
(29)
2
This ring-expanded olefin was measured as an enriched com-
70-H4C-50, 80-H4C-60; (b) JC,H correlations: p-CH34C-p, 1a-
ponent in an inseparable mixture as obtained from preparations of
CH34C-141b-CH3, 10a-CH34C-10410b-CH3, 3a-CH34C-343b-
23 and 28. 1H NMR (CDCl3, 400 MHz)
d
0.96 (s, 6H, 2ꢂ6-CH3), 1.05
CH3, 40a-CH34C-40440b-CH3; IR (KBr)
n 2971, 2927,1621 (w), 1486,
(s, 6H, 2ꢂ3-CH3), 1.64 (quasi-s, 4H, CH2-4/-5), 2.23 (s, 6H, 2ꢂo-CH3),
1459, 1383, 1362, 823, 755 (s), 586 cmꢀ1. Anal. Calcd for C35H39Cl
(495.1): C, 84.90; H, 7.94; Cl, 7.16. Found: C, 84.34; H, 7.33; Cl, 6.92.
The transparent, orthorhombic rodlets suitable for X-ray dif-
fraction analysis at 20 ꢁC (Fig. 1) were obtained through slow
evaporation of a CDCl3 solution. Crystallographic data (excluding
structure factors) for the structure in this paper have been de-
posited with the Cambridge Crystallographic Data Centre as
5.07 (s, 1H, 2-H), ca. 7.01 (overlaid, 2ꢂm-H and p-H) ppm; 13C NMR
(CDCl3, 100.6 MHz)
d
21.4 (2ꢂo-CH3), 28.9 (2ꢂ6-CH3), 29.8 (2ꢂ3-
CH3), 37.2 and 39.0 (CH2-4/-5), C-3/-6 not assigned, 125.7 (CH-p),
127.2 (2ꢂCH-m), 136.8 (2ꢂC-o), 137.4 (CH-2), 141.2 and 141.5
(2ꢂquart., C-ipso and C-1) ppm, assigned through HETCOR and
comparison with 24.

2710
R. Knorr et al. / Tetrahedron 70 (2014) 2703e2710
8. Harada, T.; Kitano, C.; Mizunashi, K. Synlett 2007, 1130e1132.
9. Harada, T.; Imaoka, D.; Kitano, C.; Kusakawa, T. Chem.dEur. J. 2010, 16,
9164e9174.
supplementary publication no. CCDC 259731. Copies of the data can
be obtained, free of charge, on application to CCDC, 12 Union Road,
Cambridge CB2 1EZ, UK (fax: þ44 (0)1223 336033 or e-mail:
deposit@ccdc.cam.ac.Uk).
10. Harada, T.; Iwazaki, K.; Otani, T.; Oku, A. J. Org. Chem. 1998, 63, 9007e9012.
€
€
11. Knorr, R.; Mehlstaubl, J.; Bohrer, P. Chem. Ber. 1989, 112, 1791e1793 and quoted
literature.
€
12. Knorr, R.; Hennig, K.-O.; Schubert, B.; Bohrer, P. Eur. J. Org. Chem. 2010,
Acknowledgements
6651e6664 compounds 17e19 therein.
ꢁ
ꢁ
13. Mloston, G.; Romanski, J.; Swiatek, A.; Heimgartner, H. Helv. Chim. Acta 1999,
82, 946e956 compound 4f on p 955 therein.
This work is dedicated to the memory of Professor Manfred
Schlosser. We are indebted to the Deutsche Forschungsgeme
inschaft for generous financial support. Thanks are due to Dr. Kurt
Polborn for performing the X-ray diffraction analysis.
14. Rappoport, Z. Acc. Chem. Res. 1992, 25, 474e479 and quoted literature.
€
15. Reich, H. J.; Sikorski, W. H.; Gudmundsson, B. O.; Dykstra, R. R. J. Am. Chem. Soc.
1998, 120, 4035e4036 note 12 therein.
16. Wehman, E.; Jastrzebski, J. T. B. H.; Ernsting, J.-M.; Grove, D. M.; van Koten, G. J.
Organomet. Chem. 1988, 353, 133e143.
17. Bush, C. N.; Applequist, D. E. J. Org. Chem. 1977, 42, 1076e1079.
18. Sander, W.; Chapman, O. L. Angew. Chem. 1988, 100, 402e403.
19. For the corresponding oxidative ‘dimer’ (C40H42) and its allene-type precursor
References and notes
€
in the 2-indanylidene series, see: Knorr, R.; Lattke, E.; Rapple, E. Liebigs Ann.
1. Knorr, R.; Pires, C.; Behringer, C.; Menke, T.; Freudenreich, J.; Rossmann, E. C.;
Chem. 1980, 1207e1215 compounds 17 and 16 therein.
20. Knorr, R.; Menke, T.; Behringer, C.; Ferchland, K.; Mehlstaubl, J.; Lattke, E. Or-
€
Bohrer, P. J. Am. Chem. Soc. 2006, 128, 14845e14853.
€
2. Knorr, R.; Pires, C.; Freudenreich, J. J. Org. Chem. 2007, 72, 6084e6090.
3. Knorr, R. Chem. Rev. 2004, 104, 3795e3849.
4. The following important publications on alkylidenecarbenoids were extracted
from the review in Ref. 3 (references 242, 8, 6, and 213 therein) to be men-
tioned here as recommended by one of the referees: (a) Kobrich, G.; Ansari, F.
Chem. Ber. 1967, 100, 2011e2020; (b) Kobrich, G. Angew. Chem. 1972, 84,
ganometallics 2013, 32, 4070e4081.
21. See compound 4n in Ref. 1.
22. Compound 6c in Ref. 20.
23. Compound 9c in Ref. 20.
24. Compound 8c in Ref. 20.
€
€
25. Knorr, R.; Menke, T.; Freudenreich, J.; Pires, C. Beilstein J. Org. Chem. 2014, 10,
557e570; Angew. Chem., Int. Ed. Engl. 1972, 11, 473e485; (c) Topolski, M.; Du-
307e315.
ꢁ
raisamy, M.; Rachon, J.; Gawronski, J.; Gawronska, K.; Goedken, V.; Walborsky,
26. For 1H NMR data of 32, see Table S19 of Ref. 20.
27. See Section 4 of Ref. 1.
€
H. M. J. Org. Chem. 1993, 58, 546e555; (d) Boche, G.; Marsch, M.; Muller, A.;
Harms, K. Angew. Chem. 1993, 105, 1081e1082; Angew. Chem., Int. Ed. Engl. 1993,
32, 1032e1033.
5. For a presentation of the impressive contemporary progress in FBW applica-
tions to polyyne syntheses, see: Jahnke, E.; Tykwinski, R. R. Chem. Commun.
2010, 3235e3249.
6. See Section 3.3 in Ref. 3 or Section 8.1.3 in Hoffmann, R. W. Dehydrobenzene and
Cycloalkynes; Chemie: Weinheim, Germany, 1967.
28. Du, Z.; Erickson, K. L. J. Org. Chem. 2010, 75, 7129e7140 and earlier publications
cited therein.
29. Detailed in the Supplementary data of Ref. 25.
30. Seebach, D.; Neumann, H. Chem. Ber. 1974, 107, 847e853.
31. Knorr, R.; von Roman, T.; Freudenreich, J.; Hoang, T. P.; Mehlstaubl, J.; Bohrer, P.;
€
€
Stephenson, D. S.; Huber, H.; Schubert, B. Magn. Reson. Chem. 1993, 31, 557e565.
7. Harada, T.; Muramatsu, K.; Mizunashi, K.; Kitano, C.; Imaoka, D.; Fujiwara, T.;
Kataoka, H. J. Org. Chem. 2008, 73, 249e258.