Microsolvation, aggregation, and pseudomonomolecular, ionic sp 2-stereoinversion mechanism of two exocyclic β,β-di-tert- alkyl-α-arylvinyllithiums This article is dedicated to Professor Paul Knochel in recognition of his kind support.

Article abstract of DOI:10.1016/j.jorganchem.2014.05.031

A THF-solvated, crystalline, exocyclic alkenyllithium, (tert-alkyl) ₂CC(Li)-Ph, was synthesized and shown to be a disolvated dimer in the solid state and in toluene as the solvent; increasing amounts of the monomeric species emerged on cooling the toluene solution. In THF as the solvent, only the trisolvated monomer was present and identified as a contact ion pair (CIP) through its scalar ¹³C,⁶Li NMR coupling. This ground-state needs only one further THF ligand as a catalyst for breaking the C-Li bond with formation of a tetrasolvated, solvent-separated ion pair (SSIP) on the way to the transition state of cis/trans sp²-stereoinversion. The ensuing pseudomonomolecular, ionic mechanism is confirmed by low pseudoactivation parameters: enthalpy ΔH_ψ^? = 6.9(3) kcal mol^-1; entropy ΔS_ψ^? = -23.3(9) cal mol^-1 K^-1. Similar parameters were found with 2,6-dimethylphenyl in place of Ph.

Full text of DOI:10.1016/j.jorganchem.2014.05.031

Accepted Manuscript
2
Microsolvation, aggregation, and pseudomonomolecular, ionic sp -stereoinversion
mechanism of two exocyclic β,β-di-tert-alkyl-α-arylvinyllithiums
Rudolf Knorr , Karsten-Olaf Hennig , Petra Böhrer , Bernhard Schubert
PII:
S0022-328X(14)00277-0
10.1016/j.jorganchem.2014.05.031
JOM 18603
DOI:
Reference:
To appear in: Journal of Organometallic Chemistry
Received Date: 29 March 2014
Revised Date: 21 May 2014
Accepted Date: 27 May 2014
Please cite this article as: R. Knorr, K.-O. Hennig, P. Böhrer, B. Schubert, Microsolvation, aggregation,
2
and pseudomonomolecular, ionic sp -stereoinversion mechanism of two exocyclic β,β-di-tert-alkyl-α-
arylvinyllithiums, Journal of Organometallic Chemistry (2014), doi: 10.1016/j.jorganchem.2014.05.031.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
transition
state
R
α
R
α
.
.
.
(THF)3Li
.
Li(THF)3
R
R
+ 1 THF
1
-
1 THF
1
3
3
H3C
H3C
CH3
CH3
solvent-
separated
ion pairs
CH3
CH3
H3C
H3C
SSIP
SSIP
´
5
4
4
5
contact ion pair (CIP)
contact ion pair (CIP´)
The pseudomonomolecular (THF-catalyzed) rate constants kψ of cis/trans interconversion of
the trisolvated, monomeric ground-states CIP and CIP´ in THF via solvent–separated ion pairs
(SSIP) can be measured through line shape analyses of the pairwise “coalescing” (averaging)
1
NMR signals of diastereotopic H and 13C nuclei at positions 1/3 and 4/5.

ACCEPTED MANUSCRIPT
Microsolvation, aggregation, and pseudomonomolecular, ionic
sp2-stereoinversion mechanism of two exocyclic β,β-di-tert-alkyl-α-
arylvinyllithiums§
Rudolf Knorr*, Karsten-Olaf Hennig, Petra Böhrer, Bernhard Schubert
Address: Department Chemie, Ludwig-Maximilians-Universität München, Butenandtstrasse
5-13 (Haus F), 81377 München, Germany
Email: Rudolf Knorr* - rhk@cup.uni-muenchen.de
* Corresponding author
Abstract
A THF-solvated, crystalline, exocyclic alkenyllithium, (tert-alkyl)2C=C(Li)–Ph, was
synthesized and shown to be a disolvated dimer in the solid state and in toluene as the solvent;
increasing amounts of the monomeric species emerged on cooling the toluene solution. In
THF as the solvent, only the trisolvated monomer was present and identified as a contact ion
pair (CIP) through its scalar 13C,6Li NMR coupling. This ground-state needs only one further
THF ligand as a catalyst for breaking the C–Li bond with formation of a tetrasolvated,
solvent-separated ion pair (SSIP) on the way to the transition state of cis/trans sp2-
stereoinversion. The ensuing pseudomonomolecular, ionic mechanism is confirmed by low
pseudoactivation parameters: enthalpy ∆Hψ = 6.9(3) kcal mol–1; entropy ∆Sψ = –23.3(9)
‡
‡
cal mol–1 K–1. Similar parameters were found with 2,6-dimethylphenyl in place of Ph.
This article is dedicated to Professor Paul Knochel in recognition of his kind support.
Keywords.
1

ACCEPTED MANUSCRIPT
Monomeric alkenyllithiums; 13C,6Li NMR coupling; ionization mechanism;
pseudoactivation parameters; sp2-stereoinversion; THF catalysis
1. Introduction
The characterization of organolithium compounds in solution is incomplete without
knowledge of the microsolvation numbers d, which state how many electron-donor ligands
(other than the carbanionic centers) are coordinated to a lithium cation. For example, the
monomeric α-phenyl-α-lithioalkene 1a is trisolvated (d = 3) in tetrahydrofuran (THF) as the
solvent but disolvated (d = 2) in Et2O or in tert-butyl methyl ether (tBuOMe) [1]. The neglect
of such microsolvation or the use of incorrect d data can lead to wrong entropy values, as
exemplified for the dimerization equilibria [2] of 1a and 1c in toluene as the solvent.
Unfortunately, microsolvation numbers at Li are most often not accessible for monodentate
(nonchelating) ethereal ligands (Et2O, THF, etc.), because NMR spectroscopy usually cannot
directly distinguish the ligand fraction coordinated at Li from the noncoordinated (“free”)
portion if ligand interchange between these two situations is very fast on the NMR timescale,
so that only averaged ligand signals can be detected. Sufficiently slow ligand scrambling was
observed for C–Li compounds with the very strong donor ligand (Me2N)3PO (HMPA) [3],
with THF at the endocyclic Li of dimeric Me2CuLi&LiCN [4], with intramolecular
(chelating) donor functions [5], and with tBuOMe, Et2O, or THF at our sterically congested
model system 1 [1].
2

ACCEPTED MANUSCRIPT
m
R´
p
m
o
p
R
o
R
i
α
i
.
.
Li(THF)d
α
R
.
.
Li
(THF)d
R
1
3
H3C
2
CH3
1
3
H3C
2
CH3
CH3
H3C
CH3
H3C
3a
7a
5
4
4
7
1,2
R
R´
2a,b
6
5
a
b
c
H
H
H
CH3
H
1a,b,c
SiMe3
Scheme 1. The α-arylalkenyllithiums 1 and 2.
This inability of measuring microsolvation numbers d directly may be circumvented by
means of an empirical [1] relationship (eq 1) of d with the scalar, one-bond spin-spin
1
6
6
coupling constant JC,Li between a carbanionic 13C nucleus and a Li atom; the Li isotope
is known [6] to be generally preferable to the more abundant isotope 7Li in 13C NMR
measurements. In eq 1, n is the number of Li cations in direct contact with the inspected
(13C NMR) carbanionic center and a is the number of carbanionic centers bound directly to a
Li cation, so that n = a = 1 for monomers, n = a = 3 for cubane-like tetramers [7], and n = a
= 2 for dimers or for “static” (nonfluxional) [1] cyclooligomers. The sensitivity factors L in
eq 1 apply to both the monomers and the aggregates within a family of C–Li compounds [8]:
L = 63 (± 3) Hz for n-alkyllithiums (but 68.4 Hz for Me4Li4), 60 (± 2) for neopentyllithium,
56 (± 1) for s-alkyllithiums, 47.5 (± 1) for tert-butyllithium (tBuLi), 42 (± 2) for 1-aryl-1-
alkenyllithiums [but 69 (± 2) for dimeric and tetrameric H2C=CH–Li], 62 (± 3) for
phenyllithium, and 69 (± 3) Hz for alkynyllithiums.
1JC,Li = L × [n × (a + d)]–1
(1)
Generally, scalar coupling between the nuclear spins in 13C–6Li bonds can only be observed
under conditions of a sufficiently slow intermolecular scrambling of the Li cations (usually at
6
low temperatures). Due to the spin quantum number I = 1 of the Li nucleus, the coupling
6
with one (n = 1) Li will split a 13C NMR signal into 2nI+1 = 2n+1 = 3 components with a
3

ACCEPTED MANUSCRIPT
1:1:1 intensity pattern; such a triplet provides evidence for a monomeric state unless the
molecule bears a Lewis-basic heteroatom X that can coordinate to the Li cation of a second
molecule, thus forming aggregates which carry C–Li–X units. The contact of a 13C center
6
with two Li nuclei (n = 2) will become evident through a 1:2:3:2:1 quintet splitting which
establishes the CLi2 units of a dimer and also those of a “static” cyclooligomer, namely,
without intraaggregational (“fluxional”) scrambling [9] of more than two Li cations. With
these reservations, the 13C NMR splitting multiplicity can identify n and hence the
1
aggregational state (since normally a = n in eq 1); if so, the magnitude of JC,Li (to be read
from the frequency intervals in the multiplets) can provide the microsolvation numbers d.
The relationship of eq 1 was discovered [1] because steric shielding in 1a and 1c impeded the
scrambling of coordinated with free monodentate donor ligands (THF, Et2O, tBuOMe) to such
an extent that d could be measured for the dimers and for some of the monomers of 1 at low
1
temperatures by H and/or 13C NMR integration. On this footing and with the above-
mentioned sensitivity factor of L = 42 (± 2) Hz in eq 1, the microsolvation numbers d of the
α-aryl-β,β-di-tert-alkylvinyllithiums 2a and 2b studied in this work should be indicated by the
1
following prospective JC,Li values: 10.5 (± 0.5) Hz for a trisolvated monomer (d = 3), 14.0
(± 0.7) Hz for a disolvated monomer (d = 2), 7.0 (± 0.4) Hz for disolvation of a CLi2
structural motif (d = 1 per lithium), or 5.3 (± 0.3) Hz for a tetrasolvated CLi2 motif (d = 2).
The results should enable us to decide whether the recently [10] established
pseudomonomolecular, ionic mechanism of sp2-stereoinversion applies also to 2a and 2b.
2. Results and discussion
2.1. Synthesis and ground-state properties of alkenyllithium 2a
Pursuing a strategy described [11] for the precursors of 1a, we heated the known [12]
alcohol 3 (Scheme 2) for a short time with an excess of N-bromosuccinimide (NBS) to
4

ACCEPTED MANUSCRIPT
produce the α-monobromide 4, taking care to avoid overbromination that would form the
corresponding α,α-dibromide whose rapidly ensuing ring expansion was shown to generate
the bromoketone 7. The analogous, slower ring expansion of monobromide 4 to give the
halogene-free ketone 8 [13] is expecially threatening for solvent-free 4, so that crude 4 should
be dehydrated by SOCl2 immediately to furnish the crystalline bromoalkene 5 which had
previously [14] been obtained as a liquid on a different route. (A further alternative synthesis
of 5 from the olefin 9a is reported in the Supplementary Material.)
p
p
Br
OH
α
m
m
Ph
i
α
i
Ph
α
LiSnMe3
SOCl2
α
OH
o
NBS
Me2
o
SnMe3
Br
Me2
Me2
pyridine
Me2
1
5
3
1
5
3
2
2
Me2
Me2
Me2
Me2
4
4
4
3
6
NBS
5
ClSnMe3
-
HBr
+ nBuLi
-
HBr
9
X
H
-
nBuBr
+ nBuLi
nBuSnMe3
a
-
b CO2H
c
p
nBu
m
i
α
α
o
X
Ph
Li
Ph
Br
Ph
O
O
2
2
H
1
1
1
5
3
1
3
2
2
Me2
Me2
Me2
Me2
Me2
Me2
Me2
Me2
4
5
4
4
5
4
5
2a
9a-c
7
8
Scheme 2. Synthesis and derivatives of 2a.
The Br/Li interchange reaction between bromoalkene 5 and n-butyllithium (nBuLi)
generated the alkenyllithium 2a (Scheme 2) along with 1-bromobutane (nBuBr); this
interchange occurred in cyclopentane solution even in the absence of donor molecules at room
temperature (r.t.), as witnessed after one week by quenching with solid CO2 and isolation of
the acid 9b. Prepared from 5 more rapidly in tBuOMe or Et2O as the solvents with ca. two
equivalents of nBuLi, 2a decayed slowly (in > 19 hours at r.t.) to furnish the “parent” olefin
5

ACCEPTED MANUSCRIPT
9a, whereas 2a in THF reacted quickly with the concomitant nBuBr to afford the α-butyl
derivative 9c (not detected in tBuOMe or Et2O). Generated from 5 in tBuOMe with nBuLi,
2a was also trapped after 20 min at 20 °C by Me3SnCl to give the α-SnMe3 derivative 6
(39%) and olefin 9a (44%). In a slightly more productive alternative process, a mixture of 6
and 9a (78:22) was obtained from bromoalkene 5 with LiSnMe3 [15] in THF as the solvent.
These two stannylation reactions opened a route to cleaner and more stable samples of 2a: the
subsequent treatment of pure 6 with nBuLi furnished 2a along with nBuSnMe3, nBu2SnMe2,
and methyllithium, thus avoiding the formation of disturbing coproducts (LiBr or nBuBr).
After this practically instantaneous Sn/Li interchange reaction, 2a in THF decayed through
proton transfer from the solvent to give olefin 9a and ethylene with a first half-life time of ca.
5.5 hours at 23 °C: quenching with D2O after 6 hours afforded 9a and [α-D]9a (58:42)
exclusively. The Sn/Li interchange was slower in tBuOMe as the solvent; it did not occur in
pentane alone (> 24 hours), but it proceeded slowly (overnight) in pentane after the addition of
THF (three equiv) and produced of (2a&THF)2 as single crystals which could be purified
through washing with (cyclo)pentane and were well suited for NMR or X-ray analyses. An
analogous run with Et2O (three equiv) in place of THF required three days in pentane at 20 °C
and produced crystals that were unsuitable because they dissolved quickly above –70 °C in
pentane.
6

ACCEPTED MANUSCRIPT
Figure 1. Solid-state structure of (2a&THF)2 at –100 °C, with the
crystallographic numbering and thermal ellipsoids of 25% probability.
The X-ray analysis revealed a centrosymmetric, disolvated dimer (2a&THF)2 with one THF
ligand (d = 1) at each Li atom. The planar four-membered ring Li2(C14)2 and the plane of
the C2/C14 (C-2/C-α) double bond (Figure 1) enclose an interplanar angle of 104°. This
angle and many of the other geometric parameters [13] resemble those of crystalline
(1a&THF)2 [1]; however, all crystalline dimers [1] of 1a and 1c had practically planar indan-
2-ylidene moieties, whereas the corresponding cyclopent-2-ylidene part of the carbanion in
Figure 1 exhibits a vexing twist of 40.4° along the C-4/C-5 single bond which is even stronger
than that (30.9°) in an isoelectronic imine [16]. Repulsive strain between the two 1-CH3
groups and the phenyl ring may be recognized from the enlarged angle C1/C2/C14 = 128.7°
{as also in (1c&THF)2 [1])}, and the concomitant weak pyramidalization of C15 (= C-ipso)
is clearly visible in Figure 1. However, the rotational orientation of the phenyl group is
almost orthogonal (ca. 85°) with respect to the C2/C14 double bond, thus providing for a
nearly optimum overlap of the σ-type carbanionic electron pair at C-α (= C14, as shown in
Scheme 1) with the aromatic π system [17]. Each THF ligand has one of its OCH2 protons
close to one of the six 3-CH3 protons (distance down to 2.3 Å) or to one of the two aromatic
ortho-protons (2.6 Å). The two oxygen atoms are in a trans relation at the Li2C2 core with
angles of O1/Li1/Li1A = 163.2°.
The various short distances [13] were essentially retained when the crystals of (2a&THF)2
were dissolved in [D8]toluene, as established by two-dimensional interproton NOESY
experiments at 25 °C. At –70 °C (Figure 2), the six-proton NMR singlet of 3-CH3 was
broadened by the adjacent Li cations, whereas the six-proton singlet of 1-CH3 remained
sharp. Ultimate evidence for the dimeric nature of this dissolved species was obtained at –70
7

ACCEPTED MANUSCRIPT
1
°C (Figure 3) through a 1:2:3:2:1 quintet splitting of the 13C-α NMR resonance with JC,Li
= 7.5 Hz, as anticipated for the disolvated (d = 1) CLi2 motif in the Introduction. Because
this splitting pattern did not change over the temperature range of –96 to –15 °C (Table S12
[13]), one can dismiss the possible structure of a higher cyclooligomer which would have
displayed the contracted “fluxional” 13C-α multiplet that would result through
intraaggregational scrambling of more than two Li cations [9] on warming up. Therefore, the
persistent splitting pattern established that the disolvated solid-state topology of (2a&THF)2
1
had passed over to the [D8]toluene solution. Importantly, the H and 13C NMR OCH2
signals of THF broadened on cooling and split at –96 °C into separate absorptions for OCH2
of coordinating and free THF (for instance, δC = 69.0 and 67.9 ppm, respectively). Although
these absorptions (Figures 4 and 5) were not baseline-separated, the signal area assigned to
coordinated OCH2 was compatible with d = 1 in (2a&THF)2.
1
Figure 2. H NMR spectrum (400 MHz, –70 °C) of 2a and THF (2.6
equiv) in [D8]toluene; M = monomer, D = dimer, o = olefin 9a.
8

ACCEPTED MANUSCRIPT
Figure 3. 13C NMR spectrum (100.6 MHz, –70 °C) of dimeric 2a and THF (2.6 equiv) in
1
[D8]toluene. Insert: C-α as a quintet with JC,Li = 7.5 Hz; o = olefin 9a.
Figure 4. THF signal splitting in the 1H NMR spectrum (400 MHz, –96 °C) of 2a and THF
(2.6 equiv) in [D8]toluene; M = monomer, D = dimer, o = olefin 9a, frei = free, koord. =
coordinated. Compare M and D with Fig. 2.
9

ACCEPTED MANUSCRIPT
Figure 5. THF signal splitting in the 13C NMR spectrum (100.6 MHz, –96 °C) of 2a and
THF (2.6 equiv) in [D8]toluene; M = monomer, D = dimer, o = olefin 9a, frei = free, koord.
= coordinated. Compare M and D with Fig. 3.
An increasing portion of trisolvated, monomeric 2a emerged in this [D8]toluene solution
(Tables S9 and S12 [13]) at temperatures below –32 °C (up to 28% at –96 °C in Figures 4
and 5, to be compared with Figures 2 and 3, respectively), which established the
deaggregation of (2a&THF)2 to be exothermic (and the reverse dimerization of monomeric
2a to be endothermic) in agreement with previous evidence [2] for 1a and 1c. Although this
1
monomeric component of 2a in toluene did not exhibit a resolved JC,Li splitting, it was
identified by its other NMR data (Tables S9 and S12 [13]) which closely resembled those of
monomeric 2a&3THF in THF as the solvent (Figures 6 and 7). The latter monomeric
contact ion pair (CIP with n = a = 1) was recognized at –82 °C in THF (Figrure 7) by virtue
of its 13C-α triplet (1:1:1) splitting, and its microsolvation by d = 3 THF ligands followed
1
from the magnitude of JC,Li = 10.7 Hz, as anticipated in the Introduction. Therefore, the
10

ACCEPTED MANUSCRIPT
ground-state of 2a&3THF cannot be a solvent-separated ion pair (SSIP) since that would
lack the direct 13C–6Li contact. This triplet pattern was wiped out on warming to above –66
°C, because an accelerating C–Li bond breaking destroyed the 13C,6Li spin coherence,
whereas all aromatic resonance positions did not change significantly between –82 and +55
°C (Tables S7 and S10 [13]), which established the presence of a sole species (namely,
monomeric 2a) all over this range. On further warmup, the accelerating cis/trans
stereoinversion led to NMR line-broadening and signal coalescences of the aliphatic nuclei as
analyzed in the sequel. However, the above-mentioned monomeric portion of 2a&3THF in
[D8]toluene (Tables S9 and S12) showed no significant line broadening (hence no signs of a
rapid cis/trans stereoinversion) [13].
1
Figure 6. H NMR spectrum (400 MHz, –66 °C) of monomeric 2a in THF; o = olefin 9a.
11

ACCEPTED MANUSCRIPT
Figure 7. 13C NMR spectrum (100.6 MHz, –66 °C) of monomeric 2a in THF; o = olefin 9a.
1
Insert: C-α as a triplet with JC,Li = 10.7 Hz at –82 °C.
2.2. The pseudomonomolecular, ionic sp2-stereoinversion of 2a in THF
The cis/trans stereoinversion of the monomer 2a&3THF, depicted in Scheme 3 as an
interconversion of the CIP ground-states 10 and 10´, cannot proceed over the very high energy
barrier that opposes a simple rotational motion of the C-2/C-α double bond. Instead, a
transitory cleavage of the C–Li bond is necessary in order to provide for migration of the
lithium cation during stereoinversion of the carbanion. The rapidity of this
diastereotopomerization [18] became perceptible on the NMR time scales (400 and 100.6
MHz, Tables S7 and S10 [13]) through coalescences of the NMR signals of pairwise
interconverting diastereotopic (namely, constitutionally but not stereochemically equivalent)
1H and 13C nuclei: 1-CH3/3-CH3 at –20 °C, 1-CH3/3-CH3 at –5 °C, CH2-5/CH2-4 at +25
°C, and C-1/C-3 at +40 °C.
12

ACCEPTED MANUSCRIPT
Li+
Li+
transition
(THF)3
(THF)3
Li
(THF)4
(THF)4
state with Li+
migration
α
+
THF
α
aryl
aryl
Li
aryl
α
-
-
THF
.
.
-
α
aryl
.
.
.
.
.
.
3
3
1
2
2
2
2
Me2
1Me2
Me2
Me2
Me2
Me2
Me2
Me2
12
4
5
4
5
10 (CIP)
11 (SSIP)
11´ (SSIP)
10 (CIP)
´
Scheme 3. Proposed mechanism of sp2-stereoinversion of the monomeric contact ion pairs
(CIP) 10 and 10´ of 2a&3THF and 2b&3THF.
Full line shape analyses [19] of the broadened NMR signals furnished the pseudo-first-order
rate constants kψ of stereoinversion as a function of the temperature: As shown in Figure 8,
these kψ values (Table S1 [13]) do not depend on the concentration of the monomer
2a&3THF. This excludes the possibility of a bimolecular stereoinversion mechanism via an
aggregated intermediate; it excludes also a dissociative ionic mechanism via the free ions. It
leaves the possibility of a nondissociative ionisation with the transitory formation of an NMR-
invisible amount of a solvent-separated ion pair 11 (SSIP); in this intermediate, the required
C–Li bond heterolysis is accomplished at the expense of the usual immobilisation of solvent
at lithium. Since the ground-state 10 of 2a&3THF is already trisolvated, only one further
THF molecule needs to be “frozen” to generate Li+(THF)4 in 11 with a corresponding
entropy contribution of ca. –11 cal mol–1 K–1 as delineated in ref [2]. After migration of
Li+(THF)4 on the way to the transition state 12 and beyond with formation of 11´ (SSIP) in
Scheme 3, one of the four THF ligands will be released to generate the product 10´ of
stereoinversion, so that the total process is pseudomonomolecular, namely, monomolecular
(kinetically of first order) in the monomer 2a&3THF and catalyzed by free THF. Thus, the
measured pseudo-first-order rate constants kψ = ko[free THF] are the products of the “true”
second-order rate constants ko and the concentrations of free THF. We did not strive for an
experimental confirmation of this kinetically first-order catalysis by THF because a significant
(at least 2-fold) decrease of the concentration [free THF] would entail a serious change of the
13

ACCEPTED MANUSCRIPT
solvent polarity with questionable kinetic evidence. As previously [10] with 1a&3THF, we
prefer to leave the constant factor of [free THF] latent in kψ, admitting that the “true”
activation entropy ∆S‡ (as derived from the ko data) would be more negative than the
measured pseudoactivation entropy of ∆Sψ = –23.3 cal mol–1 K–1 (from kψ) by a
‡
mathematical correction of roughly [2] R × ln[free THF] = ca. 5 cal mol–1 K–1 (where R =
1.986 cal mol–1 K–1).
ln kψ
5.5
5.0
4.5
4.0
3.5
3.0
3.4
3.6
3.8
1000/T
4.0
4.2
Figure 8. Arrhenius diagram of the natural logarithms of pseudo-first-order sp2-
stereoinversion rate constants kψ [s–1] versus 1000/T [K–1] for 2a in THF as the solvent.
Concentrations of 2a: open symbols, 0.13 M; hatched, 0.06 M.
The pseudomonomolecular, ionic mechanism of 2a and 2b in Scheme 3 is an independently
deduced example of the sp2-stereoinversion pathway that had been established [10] for
‡
monomeric 1a and 1b. Even the numerical values of the pseudoactivation parameters ∆Hψ
‡
‡
‡
‡
(enthalpy), ∆Sψ (entropy), and ∆Gψ (0 °C) = ∆Hψ – (273 K) × ∆Sψ (free enthalpy) of 2a
14

ACCEPTED MANUSCRIPT
(entry 1 in Table 1) are almost identical with those of 1a (entry 2). Therefore, the conclusions
to be drawn for 2a should be similar to those presented earlier [10] for 1a, for which we
established “the temporary development of considerably more negative π-charge density in
the 4´-position” of the α-aryl substituent in the transition state of stereoinversion. In
particular, the obvious absence of a kinetic influence of the obstructive benzo part of 1a (as
compared with 2a) appears to support the suggestion [10] that “Li+(THF)4 migrates along the
charge gradient” (C-meta → C-para) that was shown to increase on the way to the transition
state.
15

ACCEPTED MANUSCRIPT
Table 1. Pseudoactivation parameters ∆Gψ (kcal mol–1 at 0 °C), ∆Hψ (kcal mol–1), and ∆Sψ (cal mol–1 K–1) of cis/trans
‡
‡
‡
1
diastereotopomerization rates of the trisolvated, monomeric 1-aryl-1-alkenyllithiums 1 and 2 in THF, and 13C,6Li coupling constants J.
1J (Hz)
reference
‡
‡
‡
entry cpd. no. aryl substituent
∆Gψ (0 °C)
∆Hψ
∆Sψ
1
2
3
4
p-H
10.7
10.7
10.7
10.7
this work
[10]
2a
1a
2b
1b
13.259 ± 0.002 6.91 ± 0.25 –23.3 ± 0.9
p-H
13.35 ± 0.03
12.71 ± 0.02
12.47 ± 0.01
6.63 ± 0.24 –24.6 ± 1.0
6.87 ± 0.14 –21.4 ± 0.6
6.77 ± 0.18 –20.8 ± 0.7
o,o´-Me2
o,o´-Me2
this work
[10]
16

ACCEPTED MANUSCRIPT
2.3. Synthesis and properties of alkenyllithium 2b
The known [20] α-chloroalkene 13 (Scheme 4) and nBuLi in THF generated nBuCl and the
alkenyllithium 2b within ca. two hours at r.t. (or at –4 °C with a first half-life time of ca. 75
min). Unfortunately, 2b and nBuCl recombined at a comparable rate to give LiCl and the α-
butyl product 15c along with the “parent” olefin [21] 15a (ca. 2:1). This Cl/Li interchange
reaction of 13 in THF was faster at –70 °C with tBuLi, of which a second equivalent was
applied as usual [22] in the hope of destroying the coproduct tBuCl. Regrettably, the
formation of alkenyllithium 2b and its protonation by tBuCl (affording 15a) also occurred at
comparable rates and in competition with the intended proton transfer from tBuCl to tBuLi, so
that only ca. 30% of 2b survived, as shown through trapping with ClSnMe3 (providing 14) or
with D2O which furnished 15a and 15b (7:3). The similar generation of 2b from 13 and
tBuLi in tBuOMe as the solvent did not take place at –70 °C but was complete at r.t. within
10 minutes: A major portion of 2b was again protonated by tBuCl, while the surviving
portion could be observed (1H NMR) in situ at r.t. and was trapped with ClSnMe3 or, after 4.5
hours at r.t., with solid CO2 to afford 15a and the acid 15d. The reaction of 13 with nBuLi in
tBuOMe required many hours at r.t., producing 2b and nBuCl which coupled more slowly to
give 15c and 15a. All of these observations illustrate the impeding influence of steric
congestion in 13 or in 2b on the attack by nucleophiles (R–Li) or electrophiles (nBuCl or
tBuCl), respectively, in ethereal solvents.
17

ACCEPTED MANUSCRIPT
p
Me
Me
m
+ nBuLi
nBuCl
i
α
α
o
Cl
Me2
Li
Me
Me
-
1
3
2
Me2
Me2
Me2
5
4
-
nBuSnMe3
+ nBuLi
13
2b
+ LiSnMe3
+ ClSnMe3
p
p
Me
Me
m
m
i
i
α
α
o
o
SnMe3
X
15
X
Me
Me
a
H
1
3
1
3
2
2
Me2
Me2
Me2
Me2
b
c
d
D
5
4
5
4
nBu
CO2H
14
15a-d
Scheme 4. Preparation and derivatives of 2b.
The recommendable method of preparing the α-SnMe3 derivative 14 from 13 used LiSnMe3
1
in THF [15] at r.t. Well-resolved H,119Sn couplings were detected for the 3-CH3 (ca. 3 Hz
5
[23]) and the o,o´-CH3 protons in 14 (4.4 Hz = J or through space?), whereas such long-
range coupling was not observed for 1-CH3 (5J) and 1-CH3 (4J). Pure 14 and nBuLi in THF
produced the alkenyllithium 2b and nBuSnMe3 at 35 °C within less than two minutes. In
contrast to the behavior of 6 (Scheme 2), this Li/Sn interchange reaction occurred neither in
tBuOMe (slow formation of MeLi only) nor in Et2O; but it proceeded within 72 hours at r.t.
in Et2O containing one equivalent of THF, accompanied by some deterioration that furnished
the olefin 15a. Therefore, 2b could not be crystallized and purified (like 2a) for studies with
donor molecules other than THF. As outlined in Section 2.1 for the alkenyllithium 2a, the
1:1:1 triplet splitting of 13C-α (Figure 9) identified the monomeric state of [6Li]2b in THF
as the solvent below –59 °C, while the microsolvation number d = 3 followed through eq 1
1
from the magnitude of the NMR coupling constant JC,Li = 10.7 Hz with the same sensitivity
factor of L = 42 that was used before. As an independent confirmation of trisolvation in
THF, Scheme 5 compares lithiation NMR shifts ∆δ = δ(R–Li) – δ(R–H) for 2b (16) and the
18

ACCEPTED MANUSCRIPT
bona-fide monomer 1b&3THF (17) [24]. Because ∆δ data are known [1] to be clearly
different for tri- and disolvated monomers and disolvated dimers and the unsolvated trimer of
the indan-2-ylidene family [25] that includes 1a and 1b, the closely similar values for
comparable positions in 16 and 17 certified the proposed structure of 2b&3THF. This
monomer remained the only detectable species of 2b in THF between –85 and +25 °C, as
shown (Tables S8 and S11 [13]) by the practically temperature-independent chemical shift
1
values of the aromatic H and 13C nuclei that do not participate in the cis/trans
diastereotopomerization. As all of the diastereotopic atoms belong to the cyclopent-2-ylidene
part of 2b, their NMR signals (Figures 9 and 10) tended to be disturbed by the aliphatic
contaminations which could not be removed since 2b did not crystallize. Therefore, only the
C-4/C-5 and 1-CH3/3-CH3 pairs of 13C nuclei proved suitable for line shape analyses [19] of
sp2-stereoinversion at merely four temperatures and only one concentration (Table S2 [13]).
Nevertheless, the resulting pseudoactivation parameters of 2b (entry 3 of Table 1) and their
error limits were very similar to those of 1b (entry 4), suggesting once more that migration of
Li+(THF)4 in the transition state is hardly influenced by the obstructive benzo part of 1b
‡
which is absent in 2b. The free pseudoactivation enthalpies ∆Gψ (0 °C) revealed that
2b&3THF inverts a little faster than 2a&3THF, perhaps due to a somewhat lower internal
mobility (less negative activation entropy) in the ground-state of 2b.
(-0.88)
H (-0.46)
H
-12.4
H (-0.40)
(-0.84)
+1.4
H
-12.6
CH3(-0.26)
+21.0
+1.4
-1.3
-11.5
CH3(-0.17)
+20.8
-1.2
-11.5
Li&3THF
CH3
+66.0
-17.3
Li&3THF
+1.3
CH3
+64.0
+1.2 H3C+1.1
-1.0
3
-17.4
CH3
+1.4
1
+2.0 H3C+1.1
CH3
H3C
(ca. -0.06)
-0.6
3
CH3
(-0.14)
1
+4.0
+3.7
-0.1
CH3
H3C
(+0.06)
(-0.12)
-0.3
+0.9
+1.3
-1.7 -1.9
16: 2b in THF, -85 °C
17:
1b in THF, -95 °C
19

ACCEPTED MANUSCRIPT
1
Scheme 5. Lithiation shifts ∆δ = δ(R–Li) – δ(R–H) for 13C and (in parentheses) H NMR
signals of the trisolvated monomers 2b and 1b in THF as the solvent.
Figure 9. 13C NMR spectrum (100.6 MHz, –85 °C) of monomeric 2b in THF; o = olefin
1
15a, v = contamination. Insert: C-α as a triplet with JC,Li = 10.7 Hz at –85 °C.
20

ACCEPTED MANUSCRIPT
1
Figure 10. H NMR spectrum (400 MHz, –85 °C) of monomeric 2b in THF; o = olefin
15a.
3. Conclusion
The pseudomonomolecular, ionic sp2-stereoinversion mechanism was established for two
new examples (2a,b) in THF as the solvent through some of the criteria that had been shown
1
[10] to provide conclusive evidence for 1a and 1b: (i) the J(13C-α,6Li) magnitudes and
splitting patterns characterize monomeric, trisolvated CIP ground-states of 2a and 2b; (ii) the
concentration–independent pseudo-first-order rate constants of sp2-stereoinversion exclude
both bimolecular and dissociative mechanisms for 2a; (iii) the strongly negative
‡
pseudoactivation entropies ∆Sψ of 2a and 2b resemble those of 1a and 1b (Table 1). This
points as previously [10] to nondissociative ionization via an SSIP intermediate with THF
catalysis through immobilization of only one THF ligand as the common cause, inclusive of a
‡
similar entropic penalty of a ca. 49% portion of the measured ∆Sψ values of ca. –23 cal
mol–1 K–1. The inherent kinetic privilege [10] of THF over Et2O is confirmed by an
extrapolation of the rate constants of 2a to –55 °C that yields a pre-inversion life time of ca.
0.16 seconds in THF, to be compared to > 1800 seconds reported [26] for the Z → E
stereoinversion of 18 to give 19 (Scheme 6) in Et2O as the solvent at –55 °C.
p
Ph
Li
Li
Ph
Et2O
slow
+
Li (THF)4
H3C
H3C
H3C
H3C
o
i
18 (Z)
19 (E)
.
.
-
α
H3C
H3C
CH3
CH3
2
m
THF
fast
p
o
+
Li
(THF)4
-
i
20 (SSIP)
α
.
.
possible
H3C
2
CH3
transition
CH3
H3C
state with Li+
migration
11 (SSIP)
21

ACCEPTED MANUSCRIPT
Scheme 6. Slow Z → E interconversion of 18 in Et2O [26]; proposed transition state 20 of
the much faster cis → trans sp2-stereoinversion via the SSIP intermediate 11 in THF.
Compared with their more backside-shielded benzo relatives 1a and 1b, the two new
examples (2a,b) perform the stereoinversion in THF with almost the same enthalpy barriers,
respectively. This suggested that the migration of Li+(THF)4 was not influenced by the
benzo moiety in 1 but occurred along the π-charge gradient within the α-aryl substituents that
is roughly illustrated by the ∆δ values of 16 in Scheme 5 and intensifies [10] on the way to
the transition state 20.
4. Experimental
4.1. General remarks
Organolithium samples were prepared and handled under a stream of dry argon cover gas as
detailed in ref [1]. Pseudo-first-order stereoinversion rate constants kψ were determined as
1
described in ref [10]. All H and 13C NMR chemical shifts are referenced to internal TMS.
4.2. 2-(α-Lithiobenzylidene)-1,1,3,3-tetramethylcyclopentane (2a)
A dry NMR tube (5 mm) was charged under argon cover gas at r.t. with the stannyl
compound 6 (70 mg, 0.19 mmol), dry pentane (0.50 mL), anhydrous THF (0.047 mL, 0.57
mmol), and after short mixing finally with [6Li]nBuLi (0.54 mmol) [27] in cyclopentane (0.31
mL). The tube was closed with a soft rubber stopper that was wrapped with a layer of
parafilm . After gentle tilting without wetting the stopper, the tube was kept overnight at 22
°C under argon cover gas in a big Schlenk tube for completion of the Li/Sn interchange
reaction. On subsequent cooling to 2 °C for five hours, the first crystals appeared after ca. 80
min and grew over the next three hours. The tube was then cooled at –70 °C for 15 min,
whereupon the supernatant was withdrawn by syringe. The block-shaped crystals were
22

ACCEPTED MANUSCRIPT
suspended in dry cyclopentane (0.3 mL) at r.t. and re-cooled at –70 °C during the withdrawal
of this supernatant. After further three of such washing procedures, these crystals of
(2a&1THF)2 were stored under cyclopentane in an inclined-lying, big Schlenk tube filled
with argon gas at –18 °C. For NMR measurements, the cyclopentane was withdrawn and the
colorless crystals were blown dry in a stream of dry argon gas at r.t. for five seconds, then
finally cooled under argon to –70 °C and dissolved in an anhydrous solvent (Tables S7, S9,
1
S10, and S12 [13]). H NMR of the monomer 2a&3THF (THF, –82 °C, 400 MHz) δ 1.03
(s, 6H, 2 × 1-CH3), 1.07 (s, 6H, 2 × 3-CH3), 1.39 (t, 2H, CH2-5 or CH2-4), 6.17 (t, 1H, 3J 7.0
Hz, p-H), 6.44 (d, 2H, 3J 8.0 Hz, 2 × o-H), 6.73 (t, 2H, 3J 7.5 Hz, 2 × m-H) ppm, assigned
1
through comparison with monomeric 1a&3THF [1]; H NMR of the monomer 2a&3THF
([D8]toluene, –70 °C, 400 MHz) δ 1.43 (s, 6H, 2 × 1-CH3), 1.65 (s, 6H, 2 × 3-CH3), 1.93 (t,
2H, CH2-5), 2.08 (t, 2H, CH2-4), 6.69 (t, 1H, 3J 7.4 Hz, p-H), 6.90 (d, 2H, 2 × o-H), 7.21 (t,
1
2H, 3J 7.5 Hz, 2 × m-H) ppm; H NMR of the dimer (2a&1THF)2 ([D8]toluene, 25 °C, 400
MHz) δ 0.85 (broadened s, 6H, 2 × 3-CH3), 1.26 (sharp s, 6H, 2 × 1-CH3), 1.60 (m, 2H, CH2-
5), 1.63 (m, 2H, CH2-4), 6.67 (d, 2H, 3J 7.5 Hz, 2 × o-H), 6.75 (t, 1H, 3J 7.3 Hz, p-H), 7.14 (t,
2H, 3J 7.5 Hz, 2 × m-H) ppm, assigned through the NOESY correlations (25 °C) p-H ↔ m-H
↔ o-H (the only two-proton doublet) ↔ 1-CH3 ↔ 5-H, and 1-CH3 ↔ OCH2 of THF ↔ 3-
CH3; 13C NMR of the monomer 2a&3THF (THF, –82 °C, 100.6 MHz) δ 32.1 (2 × 1-CH3),
32.9 (2 × 3-CH3), 40.1 (CH2-4), 42.2 (CH2-5), 44.0 (C-3), 47.2 (C-1), 113.3 (C-p), 121.4 (2 ×
C-o), 126.8 (2 × C-m), 146.6 (C-2), 163.2 (C-ipso), 186.9 (t, 1JC,Li 10.7 Hz, C-α) ppm,
assigned through comparison with monomeric 1a&3THF [1]; 13C NMR of the monomer
2a&3THF ([D8]toluene, –83 °C, 100.6 MHz) δ 32.4 (2 × 1-CH3), 33.1 (2 × 3-CH3), 40.0
(CH2-4), 42.0 (CH2-5), 44.1 (C-3), 47.3 (C-1), 113.9 (C-p), 121.4 (2 × C-o), 127.2 (2 × C-m),
147.7 (C-2), 162.9 (C-ipso), 185.2 (broad, C-α) ppm; 13C NMR of the dimer (2a&1THF)2
23

ACCEPTED MANUSCRIPT
([D8]toluene, 25 °C, 100.6 MHz) δ 31.64 (sharp q, 2 × 1-CH3), 31.68 (broadened q, 2 × 3-
CH3), 39.8 (tm, 1J 127.5 Hz, CH2-4), 41.9 (tm, 1J 127.5 Hz, CH2-5), 44.4 (m, C-3), 47.3 (m,
C-1), 118.5 (broadened d, 1J 158 Hz, C-p), 122.7 (broadened d, 1J 153 Hz, 2 × C-o), 128.3
(obscured dd, 1J ca. 153 Hz, 2 × C-m), 158.6 (broadened t, 3J ca. 6 Hz, C-ipso), 161.6 (very
broad, C-2), 173.3 (very broad, C-α) ppm, assigned through HSQC and the following two-
2
3
dimensional JC,H and JC,H correlations (25 °C, window 6 Hz): 1-CH3 ↔ 1-CH3 ↔ CH2-5,
CH2-4 ↔ 3-CH3 ↔ C-3 ↔5-H ↔ 1-CH3, and 1-CH3 ↔ C-1; 13C NMR of the dimer
(2a&1THF)2 ([D8]toluene, –83 °C, 100.6 MHz) δ 31.0 (2 × 3-CH3), 31.5 (2 × 1-CH3), 39.5
(CH2-4), 41.4 (CH2-5), 44.4 (C-3), 47.1 (C-1), 118.4 (C-p), 122.4 (2 × C-o), 128.4 (2 × C-m),
158.5 (C-ipso), 160.9 (C-2), 172.9 (sharp qi, 1JC,Li 7.5 Hz, C-α) ppm. For further NMR data,
see Tables S7, S9, S10 and S12 [13].
A similar procedure with unlabeled nBuLi furnished the colorless, block-shaped single
crystal that precipitated from pentane at –20 °C and was washed only once with pentane for
the X-ray diffraction analysis at –100 °C.
4.3. 2-(α-Lithio-o,o´-dimethylbenzylidene)-1,1,3,3-tetramethylcyclopentane (2b)
The α-stannyl precursor 14 (50 mg, 0.123 mmol), anhydrous THF (0.50 mL), and
[D6]benzene (0.080 mL) were placed in a dry NMR tube (5 mm) and cooled to –70 °C under
argon gas cover for the addition of [6Li]nBuLi (0.185 mmol) in cyclopentane (0.11 mL). The
tube was closed with a soft rubber stopper, sealed with parafilm, shaken gently for 5 min at
r.t., and stored at –70 °C in a big Schlenk tube filled with argon gas. This THF solution of
monomeric 2b&3THF (Tables S8 and S11 [13]) will deteriorate within 60 min at 35 °C with
formation of the “parent” olefin 15a. We did not succeed in crystallizing 2b or studying it in
1
other solvents. H NMR (THF, –85 °C, 400 MHz) δ 0.93 (s, 6H, 2 × 1-CH3), 1.14 (s, 6H, 2 ×
3-CH3), 2.06 (s, 6H, 2 × o-CH3), CH2-5 and CH2-4 overlaid, 6.20 (t, 1H, 3J 7.0 Hz, p-H), 6.58
(d, 2H, 3J 7.0 Hz, 2 × m-H) ppm, assigned through comparison with monomeric 1b&3THF
24

ACCEPTED MANUSCRIPT
[10]; 13C NMR (THF, –85 °C, 100.6 MHz) δ 22.6 (2 × o-CH3), 30.0 (2 × 1-CH3), 31.9 (2 × 3-
CH3), 40.1 (CH2-4), 41.9 (CH2-5), 44.8 (C-3), 44.9 (C-1), 114.3 (C-p), 124.9 (2 × C-o), 126.2
(2 × C-m), 142.1 (C-2), 158.9 (C-ipso), 183.3 (t, 1JC,Li 10.7 Hz, C-α) ppm, assigned as above.
4.4. 1-(α-Bromobenzyl)-2,2,5,5-tetramethylcyclopentanol (4)
A round-bottomed flask (1000 mL) was charged with 1-benzyl-2,2,5,5-
tetramethylcyclopentanol [12] (3, 15.00 g, 64.55 mmol), N-bromosuccinimide (NBS, 17.30 g,
96.8 mmol), CCl4 (500 mL), azobis(isobutyronitrile) (AIBN, 1.65 g, 10.05 mmol), and a
magnetic stirring bar. The vessel was fitted with a reflux condenser and heated up to 95 °C.
1
A H NMR spectrum showed 85% conversion after a reflux period of 35 min. More of NBS
(3.45 g, 19.4 mmol) and AIBN (500 mg, 3.04 mmol) was added and refluxing continued with
stirring for 30 min at 95 °C. This controlled addition served to avoid overbromination that
would have formed the α,α-dibromide with a very rapidly ensuing ring expansion. After
cooling in an ice bath, the supernatant succinimide powder was removed by filtration, and the
solvent was distilled off in a rotary evaporator without heating to leave crude 4 as a yellow
liquid (22.7 g) which should immediately be processed to give 5 before the slower ring
1
expansion of 4 would form the ketone 8. H NMR of 4 (CCl4, 80 MHz) δ 0.33, 0.93, 1.28,
and 1.46 (4 s, 4 × 3H, 4 × CH3), ca. 1.71 (overlaid m, 2 × CH2), 5.25 (s, 1H, CHBr), 7.19 (m,
5H, phenyl) ppm.
4.5. 2-(α-Bromobenzylidene)-1,1,3,3-tetramethylcyclopentane (5)
The crude 1-(α-bromobenzyl)-2,2,5,5-tetramethylcyclopentanol (4, 22.7 g, maximum 64.55
mmol) was dissolved in distilled pyridine (175 mL) and cooled in an ice bath. Thionyl
choride (10.0 mL, 138 mmol) was added slowly (exothermic reaction!) with stirring. After
continued stirring at r.t. overnight, the black suspension was poured into aqueous HCl (2 M,
750 mL) and shaken with Et2O (3 × 250 mL). The combined Et2O layers were shaken with
aqueous HCl (2 M, 3 ×) and washed with distilled water until neutral, then dried over Na2SO4
25

ACCEPTED MANUSCRIPT
and concentrated to give crude 5 as a pale yellow liquid (15.4 g, < 81%). Pure 5 was obtained
through distillation (after a forerun of white needles of AIBN, δH = 1.72 ppm) with bp 95–110
°C/0.004 mbar (ref [14]: 130–150 °C/1 Torr) and subsequent crystallization from small
1
amounts of methanol at –20 °C (2 ×, mp 38–39 °C, no mp in ref [14]). H NMR (CDCl3, 400
MHz) δ 0.80 (s, 6H, 2 × 1-CH3), 1.47 (s, 6H, 2 × 3-CH3), 1.51 (pseudo-t, AA´ part of an
AA´MM´ system, 2H, CH2-5), 1.70 (pseudo-t, MM´ part, 2H, CH2-4), 7.23 (m, 3H, 2 × o-H
and overlaid p-H), 7.30 (tm, 2H, 3J ca. 7 Hz, 2 × m-H) ppm, assigned through SCS [12] in
accord with NOE difference spectra that exhibited intensification of 4-H by irradiation at {3-
CH3} and of 5-H and o,o´-H by {1-CH3}; 13C NMR (CDCl3, 100.6 MHz) δ 26.90 (qsext, 1J
126 Hz, 3J 4.6 Hz, 2 × 3-CH3), 29.00 (qsext, 1J 126 Hz, 3J 4.6 Hz, 2 × 1-CH3), 41.43 (tm, 1J
ca. 127 Hz, CH2-5), 41.45 (tm, 1J ca. 127 Hz, CH2-4), 47.00 (m, 3J ca. 3.5 Hz, C-3), 47.75 (m,
3J ca. 3.8 Hz, C-1), 115.47 (t, 3J 4.3 Hz, C-α), 127.51 (dt, 1J 160.6 Hz, 3J 7.5 Hz, C-p), 127.77
(dd, 1J 160 Hz, 3J 7.5 Hz, 2 × C-m), 129.37 (dt, 1J 160.5 Hz, 3J 6.7 Hz, 2 × C-o), 143.31
3
(sharp t, 3J 7.5 Hz, C-ipso), 156.85 (m, apparent J ca. 4 Hz, C-2) ppm, assigned through SCS
[12] in accord with selective {1H} decoupling experiments as follows: {1-CH3} → 1-CH3 as
3
a sharp t with residual J to 5-H, and C-1 narrowed; {3-CH3 and CH2-5} → 3-CH3 and
CH2-5 decoupled, and C-3 narrowed; IR (KBr) ν 2950, 2867, 1459, 1365, 715 (s) cm–1.
Anal. calcd for C16H21Br (293.25): C, 65.53; H, 7.22; Br, 27.25. Found: C, 65.65; H, 7.17;
Br, 26.56.
Alternatively, pure 5 was obtained via epoxidation of olefin 9a [13]. 5 was not present in
the complex product mixture that was formed from 9a [12] with elemental bromine in CCl4.
4.6. 2-(α-Trimethylstannylbenzylidene)-1,1,3,3-tetramethylcyclopentane (6)
a) From 5 with LiSnMe3. A green solution of LiSnMe3 [15] (ca. 6.14 mmol) in THF (8.77
mL) was added under argon cover gas to the bromoalkene 5 (1.00 g, 3.41 mmol) in anhydrous
THF (4 mL) with stirring at –72 °C. The mixture was stirred for another 15 min and then at
26

ACCEPTED MANUSCRIPT
r.t. for 45 min, poured onto solid CO2, warmed up, diluted with 2 M aqueous NaOH (25 mL),
and shaken with Et2O (3 × 10 mL). The combined Et2O extracts were washed with distilled
water until neutral, dried over Na2SO4, and concentrated to give a yellowish mixture (1.27 g)
of 6, olefin 9a, and Sn2Me6 (6:2:2). Microdistillation separated 9a [bp 86–92 °C (bath
temp.)/0.02 mbar, 118 mg] from 6 [bp 102–118 °C (bath temp.)/0.02 mbar, 607 mg, 47%
1
yield]; the latter fraction crystallized from ethanol at –30 °C: mp 27.5–29 °C; H NMR
(CDCl3, 400 MHz) δ –0.03 (s, 9H; 119Sn satellites, 2J 51.6 Hz; α-SnMe3), 0.84 (s, 6H; 119Sn
satellites, |5J| < 1.5 Hz; 2 × 1-CH3), 1.22 (s, 6H; 119Sn satellites, 5J (?) 3.1 Hz; 2 × 3-CH3),
3
1.47 (t, 2H, 3J 6.8 Hz, CH2-5), 1.60 (t, 2H, J 6.8 Hz, CH2-4), 6.86 (dd, 2H, 3J 8.2 Hz; 119Sn
satellites, 4J ca. 4 Hz; 2 × o-H), 7.04 (tt, 1H, 3J 7.4 Hz, 4J 1.4 Hz, p-H), 7.19 (tm, 2H, 3J 7.8
Hz, 2 × m-H) ppm, assigned through NOE difference spectra that exhibited intensification of
CH2-4 by irradiation at {3-CH3} and of CH2-5 and o,o´-H by {1-CH3}; 13C NMR (CDCl3,
100.6 MHz) δ –5.23 (q; 119Sn satellites, 1J 336.5 Hz; α-SnMe3), 29.91 (q; 119Sn satellites, 4J
7.6 Hz; 2 × 1-CH3), 30.65 (q; 119Sn satellites, |4J| ≤ 4 Hz; 2 × 3-CH3), 40.42 (t, CH2-5), 41.18
(t, CH2-4), 45.22 (quart.; 119Sn satellites, 3J 21.0 Hz; C-3), 47.53 (quart.; 119Sn satellites, 3J
61.0 Hz; C-1), 124.16 (d; 119Sn satellites, 5J 12.2 Hz; C-p), 127.17 (d; 119Sn satellites, 4J 9.5
Hz; 2 × C-m), 128.47 (d; 119Sn satellites, 3J 18.7 Hz; 2 × C-o), 135.91 (quart.; 119Sn satellites
not detected; C-α), 146.00 (t, 3J 7.2 Hz; 119Sn satellites, 2J 23.6 Hz; C-ipso), 167.64 (quart.;
119Sn satellites, 2J 16.6 Hz; C-2) ppm, assigned through the 119Sn coupling constants and the
following selective {1H} decoupling experiments: {1-CH3} → 1-CH3 as a narrow t through
residual coupling to CH2, C-1 as a narrow t coupled to CH2; {3-CH3 + CH2-5 + CH2-4} → 3-
CH3, C-3, and CH2-5 as three sharp s; IR (KBr) ν 2954, 2864, 1598, 1362, 769, 708, and 520
cm–1. Anal. calcd for C19H30Sn (377.16): C, 60.51; H, 8.02. Found: C, 61.09; H, 8.00.
b) From 2a with ClSnMe3. A dry Schlenk flask (25 mL) was charged with a magnetic
stirring bar and nBuLi (3.55 mmol) in hexanes (1.42 mL) under argon cover gas. The flask
27

ACCEPTED MANUSCRIPT
was cooled to –40 °C and treated first with anhydrous tBuOMe (10 mL) and then with the
bromoalkene 5 (800 mg, 2.73 mmol). The resultant solution of the alkenyllithiium 2a and
residual nBuLi was stirred at r.t. for 20 min, whereupon ClSnMe3 (707 mg, 3.55 mmol;
Caution: very poisenous!) was added portionwise within 10 min. The mixture was stirred
overnight, then poured into iced water (50 mL) and extracted with Et2O (3 ×). The combined
Et2O extracts were washed with distilled water, dried over K2CO3, and concentrated to give a
yellow liquid (805 mg) containing product 6 and olefin 9a (47:53). Distilled 6 (405 mg, 39%)
was stable against aqueous NaOH (2 M) or HCl (2 M) for two days (very slow formation of
olefin 9a).
4.7. 2-Bromo-3,3,6,6-tetramethyl-2-phenylcyclohexanone (7)
The colorless needles 7 were obtained (together with the monobromide 4) through
overbromination of the alcohol 3 with NBS (2.4 equiv): Mp 146–147 °C (pentane); 1H NMR
(CDCl3, 400 MHz) δ 0.83 (s, 3H, 1 × 6-CH3), 1.14 (s, 3H, 1 × 6-CH3), 1.25 (s, 3H, 1 × 3-
CH3), 1.45 (ddd, 1H, |2J| 14.1 Hz, 3J 4.2 Hz, 3J 3.3 Hz, equat. 4-H), 1.61 (s, 3H, 1 × 3-CH3),
1.72 (ddd, 1H, |2J| 14.1 Hz, 3J 4.0 Hz, 3J 3.3 Hz, equat. 5-H), 1.91 (td, 1H, |2J| 14.1 Hz, 3J 4.2
Hz, ax. 5-H), 2.48 (td, 1H, |2J| 14.1 Hz, 3J 4.2 Hz, ax. 4-H), 7.24–7.29 (m, 3H, 2 × m-H and p-
H), 7.40 (dm, 2H, 3J 7.2 Hz, 2 × o-H) ppm; 13C NMR (CDCl3, 100.6 MHz) δ 25.1, 27.7, and
29.3 (3 × CH3), 33.2 and 34.2 (CH2-4/-5), 34.5 (1 × CH3), 41.2 and 42.8 (2 × quart., C-3/-6),
82.3 (C-2), 126.4 (2 × C-m), 127.5 (C-p), 131.3 (2 × C-o), 136.2 (C-ipso), 209.4 (C-1) ppm;
IR (KBr) ν 2961, 2933, 2869, 1699 (s), 1462, 1444, 744, 702 cm–1. Anal. calcd for
C16H21BrO (309.25): C, 62.14; H, 6.84; Br, 25.84. Found: C, 62.21; H, 6.76; Br, 25.88.
4.8. 3,3,6,6-Tetramethyl-2-phenylcyclohexanone (8)
See ref [13].
4.9. 2-Benzylidene-1,1,3,3-tetramethylcyclopentane (9a)
28