Why is cis/trans stereoinversion with Li+(THF)4 migration across the phenyl ring of α-lithiostyrene accelerated by two ortho-methyl groups?

Article abstract of DOI:10.1016/j.tet.2018.01.047

Common wisdom might anticipate that two methyl groups placed on a molecular migration route should act as an impediment. However, the “conducted tour” migration of Li⁺(THF)₄ across the aryl ring (“π-route”) during the cis/trans stereoinversion of α-arylvinyllithiums had been found to occur with practically equal velocities in the presence of either one or two ortho-alkyl substituents. We now report that the omission of both ortho-methyl groups retards the stereoinversion process. In order to arrive at an answer to the title question, we investigate the aggregation equilibria and microsolvation states of ortho, ortho′-unsubstituted α-lithiostyrenes by means of approved secondary NMR criteria. Beyond such necessary knowledge about the ground-state properties, we provide kinetic evidence showing that the retarded cis/trans stereoinversion of α-lithiostyrene proceeds by the pseudomonomolecular, ionic mechanism with Li⁺(THF)₄ migration.

Full text of DOI:10.1016/j.tet.2018.01.047

Accepted Manuscript
+
Why is cis/trans stereoinversion with Li (THF) migration across the phenyl ring of α-
4
lithiostyrene accelerated by two ortho-methyl groups?
Rudolf Knorr, Ernst Lattke, Jakob Ruhdorfer, Kathrin Ferchland, Ulrich von Roman
PII:
S0040-4020(18)30107-8
10.1016/j.tet.2018.01.047
TET 29262
DOI:
Reference:
To appear in: Tetrahedron
Received Date: 11 October 2017
Revised Date: 24 January 2018
Accepted Date: 26 January 2018
Please cite this article as: Knorr R, Lattke E, Ruhdorfer J, Ferchland K, von Roman U, Why is cis/trans
+
stereoinversion with Li (THF) migration across the phenyl ring of α-lithiostyrene accelerated by two
4
ortho-methyl groups?, Tetrahedron (2018), doi: 10.1016/j.tet.2018.01.047.
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
R
4
R = H, Cl, SiMe3
R
R
4
4
+
Li
(THF)4
2
1
1
.
.
.
.
(THF)
Li
(THF)
3 Li
H
3
1
α
α
.
.
β
β
H
H
H
-
+ THF
-
THF
α
cis
trans
trans
cis
β
H
H
1

Why is cis/trans stereoinversion with Li+(THF)4 migration across the
ACCEPTED MANUSCRIPT
phenyl ring of α-lithiostyrene accelerated by two ortho-methyl groups?
In memory of Professor Joseph Klein
Rudolf Knorr, Ernst Lattke, Jakob Ruhdorfer, Kathrin Ferchland, Ulrich von Roman
ABSTRACT
Common wisdom might anticipate that two methyl groups placed on a molecular migration
route should act as an impediment. However, the “conducted tour” migration of Li+(THF)4
across the aryl ring (“π-route”) during the cis/trans stereoinversion of α-arylvinyllithiums had
been found to occur with practically equal velocities in the presence of either one or two
ortho-alkyl substituents. We now report that the omission of both ortho-methyl groups
retards the stereoinversion process. In order to arrive at an answer to the title question, we
investigate the aggregation equilibria and microsolvation states of ortho,ortho´-unsubstituted
α-lithiostyrenes by means of approved secondary NMR criteria. Beyond such necessary
knowledge about the ground-state properties, we provide kinetic evidence showing that the
retarded cis/trans stereoinversion of α-lithiostyrene proceeds by the pseudomonomolecular,
ionic mechanism with Li+(THF)4 migration.
Key words: Aggregation, Li+ migration, Microsolvation, Reaction mechanisms,
Stereolability, Vinylanions
1. Introduction
Crystallographic studies of many organolithium compounds disclosed a multitude of solid-
state structures with various kinds of aggregation and degrees of microsolvation (namely, the
coordination of “explicit” electron-pair donating ligands at lithium).1–5 These structures may
or may not survive a transfer from the crystal into a solution; therefore, dissolved
organolithiums require additional analytic techniques for differentiating monomeric from
dimeric and higher aggregational states. Under suitable conditions, scalar NMR one-bond
1
1
coupling constants J(13C,6Li) or (less conveniently) J(13C,7Li) can provide such
differentiations. In contrast to molecular weight determinations, this NMR technique can
work even for contaminated solutions and in the presence of two or more organolithium
species.6 Due to the high mobility of Li+ cations within and between dissolved organolithium
molecules, however, such 13C,6Li spin-spin coupling must be sought at sufficiently low
6
temperatures that retard intermolecular Li+ scrambling and conserve Li–13C(α) spin
2

coherence on the NMR time scales. Since the most common monodentate (nonchelating)
ACCEPTED MANUSCRIPT
ligands such as tetrahydrofuran (THF) or Et2O often are even more mobile than Li+,
determinations of their microsolvation numbers d in solution remained difficult for a long
time. Scheme 1 displays α-arylalkenyllithiums 1 – 12 whose (non)aggregation and
1
microsolvation states were established7–15 through the above JC,Li technique: With one
exception15 (namely, dimeric 10), all of these compounds turned out to be monomeric in
THF as the solvent; all of these monomers were trisolvated with d = 3 THF ligands at Li+.
With non-THF ligands such as Et2O or tert-butyl methyl ether (t-BuOMe), the monomeric
species of 1a–e,7 5c,13 and 7b13 were only disolvated (d = 2), whereas most of the other
examples in Scheme 1 formed disolvated dimers with d = 1 monodentate ligand at Li. Steric
congestion by bulky substituents in the C-β region of 1a–e or at the ortho,ortho´-positions of
the α-aryl group (5c, 7b) disfavored dimerization of the monomers. Higher aggregated states
were detected for unsolvated species (d = 0) of 1a, 5c, and 10–12 that were sufficiently
soluble in donor-free hydrocarbon solvents; however, only the unsolvated cyclotrimeric7
species of 1a could be structurally characterized.
R
4
CH3
R
α
.
α
α
.
.
.
.
.
Li
Li
Li
.
.
Li
H3C
R
H3C
H3C
CH3
β
CH3
(H3C)2
(CH3)2
H3C
CH3
CH3
CH3
H3C
H3C
4
3
1
R
a
b
c
d
e
H
CH3
Ph
Cl
SiMe3
3a: R = H
3b: R = CH3
1a-e
2
H3C CH
R
4
3
2
R
CH3
R
1
H
H
.
.
Li
H
.
.
Li
α
H
.
.
Li
R
R
α
CH3
α
Li
β
β
H
H
Rcis
H
H
C
Rtrans
H
C
H
β
5a-c
6a,b
7a,b
5
R
6
Rcis Rtans
7
R
a
b
c
CH3
CH(CH3)2
C(CH3)3
a
b
CH3
H
H
CH3
a
b
CH3
C(CH3)3
8
4
2
2
α
Li
CH3
CH(CH3)2
C
1
.
.
Li
H
C
.
.
Li
.
.
Li
H
β
α
α
β
β
H
H
Me2
H
H
12
11
9
10
Scheme 1 α-Arylalkenyllithiums whose (non)aggregation and microsolvation
states had been established in solution: 1a–e,7,8 2,8 3,9 4,10 5a,11 5b,12 5c,13
6a and b,10 7a,11 7b,13 8,14 9,14 and 10–12.15
3

ACCEPTED MANUSCRIPT
The most direct experimental evidence for the discrete microsolvation numbers d = 1 or
2 at Li was obtained through NMR integration only from the sterically congested α-
arylalkenyllithiums 1a,7 1e,8 3a,9 and 5b,12 whose ligand scrambling was sufficiently
1
retarded: The H and 13C NMR spectra of 5b displayed separate signals for free and
coordinated (immobilized) monodentate ligands, so that the d values followed from the
NMR integrals of the immobilized ligands in comparison with suitable integrals of the
1
carbanionic part. These integral ratios and also the magnitudes of the above JC,Li couplings7
were dubbed primary NMR criteria of microsolvation; but such criteria are unavailable
without steric shielding and efficient cooling of the solutions. Therefore, the less encumbered
examples in Scheme 1 had to be analyzed by secondary NMR criteria: First, scalar two-bond
2
coupling constants JH,H between the two β-protons in the H2C=C moieties can reveal
microsolvation numbers.11 Second, chemical shift differences ∆
nuclei of an organolithium (RLi) and its “parent” substance (RH) can also serve as secondary
criteria: Some of these lithiation shifts ∆ (RLi) – (RH) may contain hints at
δ between corresponding
δ
=
δ
δ
microsolvation, aggregation, and the α-aryl conformation. Except for 10 whose α-aryl group
is conformationally fixed, all other compounds in Scheme 1 preferred a close to perpendicular
relationship of the α-aryl and the C(α)=C(β) double-bond planes; even the single small 2-
CH3 substituent in 12 did not admit greater deviations from this conformation that is
preserved by substantial electronic barriers12,14 against α-aryl rotation about the C(α)–C(1)
bond. Considering that the above conformational preference and the resistance against
aggregation in THF had been detected also for all three α-arylvinyllithiums (9, 11, and 12)
that carry only one ortho-substituent, we felt motivated to extend earlier studies16 of
examples without any ortho substituent, in particular to the apparently “simple” α-
lithiostyrene. The present kinetic results raised the title question and provided an answer that
is compatible with the “conducted tour”14 migration.
2. Results and discussion
2.1. Preparation, ground states, and aggregation of the o,o´-unsubstituted α-lithiostyrenes
The Br/Li interchange reaction (Scheme 2) of α-bromo-4-(trimethylsilyl)styrene (13a)
with n-butyllithium (n-BuLi, 1.1 equiv) to give α-[4-(trimethylsilyl)phenyl]vinyllithium (14a)
and 1-bromobutane (n-BuBr) was surprisingly fast in donor-free pentane as the solvent:
Unsolvated 14a emerged with a first half-reaction time of ca. 8 min at 32 °C and displayed a
2
considerably broadened AB-type spectrum of its β-protons with JH,H = 4.0 Hz. Regrettably,
this donor-free species of 14a began to decay with polymerization. In Et2O as the solvent, the
Br/Li interchange of 13a was much faster and furnished the Et2O-solvated species of 14a that
4

polymerized less rapidly and formed the acid 16a on carboxylation with solid CO2; however,
ACCEPTED MANUSCRIPT
the interfering reaction of 14a with its coproduct n-BuBr started soon and was complete
within two days at room temperature (rt). In order to prevent such a destruction of 14a, n-
BuBr and all other volatile impurities were removed from the Et2O solution under reduced
pressure (< 0.1 mbar), leaving 14a together with some LiBr (ca. 0.3 equiv) as an oily residue
that was dissolved in Et2O, THF, or other anhydrous solvents for the NMR studies. Since the
above-mentioned primary NMR criterion of 13C,6Li spin-spin coupling remained unavailable
for 14a and for all of the following ortho-unsubstituted α-lithiostyrenes even at the lowest
temperatures, we used the secondary NMR criteria of lithiation shifts, ∆δ = δ(RLi) – δ(RH),
as a tool for purging
(RH).
δ
(RLi) from some of the constitutional effects that are also present in
δ
3
R
R
4
3
4
2
2
1
5
Br
1
α
β
5
.
.
Li
+ n-BuLi
n-BuBr
6
6
α
-
H
H
β
H
H
13a-d
cis
trans
13-15 R
a SiMe3
14a-d
b Cl
c CH3
d H
CO2
+
H
3
6
R
4
2
1
3
R
4
α
CO2H
2
1
5
H
α
β
5
β
H
cis
H
trans
6
H
H
cis
trans
16a: R = Me3Si
16c: R = CH3
15a-d
Scheme 2. Preparation and derivatization of the α-arylvinyllithiums 14a–d.
In Et2O as the solvent at 25 °C, the value of ∆
classified 14a as predominantly dimeric on account of the similar values of the established
dimers of 5a (entry 2)11 and 12 (entry 4).15 The strongly positive ∆
(C-β) in entry 6 will turn
δ(C-1) = +20.7 ppm (entry 6 of Table 1)
δ
out to be characteristic of the other dimeric 14 fellows; but 14a was not sufficiently stable to
furnish additional evidence at and below rt. On the other hand, 14a was stable in THF as the
solvent in the absence of n-BuBr: Its ∆
nonaggregated (monomeric) state through comparison with the established monomers of 5a
(entry 1)11 and 12 (entry 3).15 The most strongly negative value of ∆
(C-4) = –14.8 ppm
(entry 5) is typical for Me3Si substitution (–17.9 ppm in 1e)7 and due to a close to
δ data for C-α, C-1, and C4 (entry 5) disclosed a
δ
5

perpendicular relationship of the α-aryl plane with respect to the C(α)=C(β) double-bond
ACCEPTED MANUSCRIPT
plane.15 Although a dimeric species of 14a could not be separately identified in this THF
solution, its growing population on warming (the usual17 endothermic dimerization) appeared
1
to be the reason behind the temperature-dependent H and 13C NMR spectra (Tables S118):
The ∆δ data became less positive for C-α and C-1 but more positive for C-β, C-4, 2-/6-H, 3-
/5-H, and the two β-H resonances. The full ∆
δ
set of monomeric 14a at –115 °C is depicted
in Figure 1b.
H(-0.36)
H(-0.41)
(-0.09)
(a)
(b)
Me3Si
Cl
-11.0
3
4
3
4
+0.5
-14.8
2
H(-0.74)
2
H(-0.80)
5
5
-4.1
-2.3
-1.5
6
6
+26.9
+27.9
α
.
α
.
&3THF
Li
-3.7
1
.
Li
.
&3THF
1
+72.7
+1.0
+73.3
-0.4
β
β
H
(-0.06)
H
H
(-0.14)
H
(-0.19)
(-0.21)
cis
trans
cis
trans
Monomeric 14b at -115 °C
in THF/hydrocarbons
Monomeric 14a at -115 °C
in THF/hydrocarbons
(-0.05)
H(-0.21)
(c)
H(-0.24)
(d)
H3C
Cl
-5.8
3
3
4
4
-0.3
-9.9
2
2
H(-0.50)
H(+0.44)
5
5
-1.8
-1.6
-1.5
6
6
+24.9
+21.4
α
.
.
α
.
.
&2THF
Li
-2.8
2&Et O
1
Li
1
2
+68.5
+5.9
+65.2
+9.8
β
β
H
(+0.30)
H
H
H
(+0.27)
(+0.43) (+0.50)
cis
trans
cis
trans
Dimeric 14c in THF: 13C
Dimeric 14b at -99 °C
(1H at 25 °C) in Et 2O
at -88 °C, 1H at 27 °C
(-0.47)
H(-0.33)
H(-0.24)
(-0.61)
(e)
(f)
H
H
-9.4
3
4
3
4
-7.7
2
H(-0.50)
2
H(-0.74)
5
-2.5
5
-1.6
-1.6
6
6
+24.5
+26.6
α
α
.
.
.
.
&2THF
Li
1
-3.6
Li
&3THF
1
+66.4
+8.1
+72.8
+1.5
β
β
H
(+0.25)
H
(+0.27)
H
(-0.04)
H
(-0.12)
cis
trans
cis
trans
Dimeric 14d at -118 °C in
THF (83%)/hydrocarbons
Monomeric 14d at -118 °C
in THF (83%)/hydrocarbons
(-0.34)
H(-0.34)
3
(-0.37)
H(-0.18)
3
(h)
(g)
H
H
4
4
-4.9
-6.7
2
H(-0.93)
2
H(-0.54)
5
5
+0.1
-1.3
6
6
+17.5
+21.3
α
α
.
.
.
.
&t-BuOMe
Li
-1.9
1
-2.5
Li
&2t-BuOMe
1
+56.2
+13.6
+66.4
+10.2
β
β
H
H
H
H
(+0.53) (+0.99)
(+0.40) (+0.55)
cis
trans
cis
trans
Agg1 of 14d at -41 °C in cyclo-
pentane/t-BuOMe (2.4 equiv)
Dimeric 14d at -101 °C
in t-BuOMe (91%)
1
Fig. 1. Lithiation H (in parentheses) and 13C NMR shifts ∆δ = δ(RLi) – δ(RH) of 14a–d.
6

ACCEPTED MANUSCRIPT
Table 1. Microsolvation numbers d, NMR coupling constants 2JH,H [Hz] of CH2-β, and lithiation shifts ∆
δ
=
δ
(RLi) –
4-H
δ(RH) of the α-
arylvinyllithiums 5a, 12, and 14a–d in four solvents.
2
agg a JH,H [Hz]
d c
temp. Ref.
°C d
α-aryl
no. substituent
∆2JH,H
chemical shifts
δ
[ppm]
C-4
entry cpd.
solvent
THF
no.
(at °C) b
[Hz]
C-1
C-α
C-β
1
2
2,6-Me2
2,6-Me2
2-Me
2-Me
4-SiMe3
4-SiMe3
4-Cl
M
6.6
3.5
6.24
–
2.9 +76.4 –10.9 +24.2 –11.0 –0.76 –89 11
1.1 +68.1 –3.6 +19.4 –7.0 –0.43 –40 11
2.7 +77.3 –4.7 +27.8 –10.4 –0.61 –38 15
5a
8.8 (≤ +2)
t-BuOMe D
5a
5.7 (≤ –7)
8.0 (+25)
–
3
THF
M
D
12
4
Et2O
THF e
–
+67.9
3.2 +73.3
1.7
+5.0 +21.8 –6.1
–0.4 +27.9 –14.8
–
–
–
–
–
–
–70 15
12
5
M
D
7.0
4.5
7.1
4.9
5.6
–
–115
f
f
f
f
f
f
f
f
f
14a
14a
14b
14b
14c
14d
14d
14d
14d
8.0 (≥ –50)
5.5 (+25)
8.0 (–44)
5.8 (+25)
6.7 (+51)
–
6
Et2O
–
+9.4 +20.7
–
+25
–115
–99
7
THF e
Et2O
M
D
3.2 +72.7
1.9 +65.2
2.3 +68.5
+1.0 +26.9 –11.0
+9.8 +21.4 –5.8
+5.9 +24.9 –9.9
8
4-Cl
9
4-Me
4-H
THF
D
–88
10
11
12
13
THF e
THF e
t-BuOMe D g
M
–
–
+73.4
+67.7
+1.5 +26.6 –9.4 –0.61 –118
+8.1 +24.5 –7.7 –0.47 –118
4-H
D
–
–
4-H
4.7
3.7
1.8 +66.4 +10.2 +21.3 –6.7 –0.37 –101
1.2 +56.2 +13.6 +17.5 –4.9 –0.34 –41
5.8 (≥ –22)
h
4-H
C5H10
Agg1 4.8 (+25)
2
c
a “M” = monomer, “D” = dimer at the temperatures denoted before the last column. b Temperatures of this JH,H value. Formal
e
value as calculated from eq 2 before rounding. d Temperature of determinations of ∆
δ
=
δ
(RLi) –
δ
(RH). With hydrocarbon
f
cosolvents. This work. g Accompanied by the unidentified aggregates Agg1 and Agg2. h C5H10 = cyclopentanewith t-BuOMe (2.4
equiv with respect to 14d).
7

α-(4-Chlorophenyl)vinyllithium (14b) was prepared from bromoalkene 13b in pentane
ACCEPTED MANUSCRIPT
under the above conditions (as for 14a) with a comparably short first half-reaction time and
2
again JH,H = ca. 4 Hz. In a similar race against polymerization, the procedure for 14a was
used to create 14b and LiBr (ca. 0.3–1.4 equiv) in Et2O solutions with subsequent evaporation
of n-BuBr. Although the interference by polymerization reduced the number of interpretable
NMR spectra drastically, the ∆δ data (entry 7 and Figure 1a) revealed 14b to be monomeric
in THF, as shown by the similar data of monomeric 14a in entry 5 and Figure 1b. In Et2O as
the solvent, 14b was predominantly dimeric (entry 8 and Figure 1c), as recognized through
comparisons with the established dimers of 5a (entry 2) and 12 (entry 4). The constitutions of
14b and 14a in Et2O before evaporation were confirmed through their clean addition reactions
1
to dialkyl ketones: H and 13C NMR analyses of the pure adducts (S8–S10)18 showed no
signs of internal overcrowding. Nevertheless, t-Bu2C=O added to both 14a and 14b only in
the absence of the more reactive diisopropyl ketone, in analogy with a corresponding
selectivity reported19 for 5a.
Unsolvated α-(4-methylphenyl)vinyllithium (14c) was obtained from bromoalkene 13c
with n-BuLi in pentane by a Br/Li interchange reaction that was complete after 60 min at rt.
The precipitation of powdery 14c was accompanied by the formation of LiBr that could not
totally be removed through subsequent washings with pentane. Formation of the acid 16c18
through carboxylation established the constitution of 14c. In THF as the solvent, 14c was
dimeric down to –88 °C, as revealed by Figure 1d and the following comparisons with
dimeric 14b and with the established15 dimer of 12 in Et2O: In the vicinity of the Li–C(α)
bonds, the ∆δ(C-α) and ∆δ(C-β) values were similar for all three compounds (entries 9, 8, and
4, respectively). Thus, 14c had overcome the usual tendency of THF to deaggregate our
dimers, presumably because the π-electron-repellent 4-CH3 disfavored the nonobserved
monomer.
α-Lithiostyrene (α-phenylvinyllithium, 14d) was prepared from α-bromostyrene (13d) in
cyclopentane or pentane by the above procedure (for 14c) with a similar problem: The
precipitating powder could be purified through washing with (cyclo)pentane, but it retained
some LiBr that was carried over to the fresh solvents and the NMR studies. The alternative
performance of this Br/Li interchange reaction directly in Et2O or t-BuOMe at –75 °C was
unprofitable since 14d was generated in competition with formation of a similar amount of
the acetylide PhCCLi. In contrast to 14a and 14b, the purified samples of 14d were
thermally stable up to 110 °C for at least a few minutes in THF as the solvent. With a
sufficient portion of saturated hydrocarbons as cosolvents, such THF solutions remained
1
liquid at and below –118 °C during the H and 13C NMR runs that disclosed the presence of
both monomeric and dimeric 14d (Tables S5a18 and S5b). The monomeric species (entry 10
8

of Table 1; Figure 1f) was identified through its lithiation shifts ∆
δ
that resembled those of
ACCEPTED MANUSCRIPT
the established monomers of 5a, 12, 14a, and 14b (entries 1, 3, 5, and 7). Dimeric 14d was
recognized at –118 °C (entry 11) in the same solution through its ∆δ data that resembled
1
those of dimeric 14c (entry 9). The separated H and 13C NMR signals of these two 14d
species coalesced pairwise in the temperature region of –104 °C; on further warm-up, the
averaged resonance positions moved according to a changing monomer/dimer population
ratio: With a total 14d concentration of 0.3 M in units of the monomer formula, ca. 40% of
the material in THF was in the monomeric form at and below –118 °C but only ca. 20% at 25
1
°C. The H NMR spectrum of such a mixture (Figure 2b) exhibited the typical upfield shifts
of 4-H (triplet) and 2-/6-H (doublet) that result through delocalization of negative electric
charge from the Li–C(α) bond into the aromatic π system (quasi-benzyllithium resonance8,15).
The monomer population of 14d in THF was higher than that of 14c (bearing 4-CH3) but
lower than those of 14b (4-Cl) and 14a (4-SiMe3). The dimeric species of 14d was present
also in the solvent t-BuOMe, as shown by comparison of the data in entries 11 and 12 or in
Figures 1e and 1g; this dimer was accompanied by two unidentified components (but not by
monomeric 14d). These two components (Agg1 and Agg2) were 14d species since all of
1
their H and 13C NMR signals formed weighted averages with those of the dimeric species
above the coalescence temperatures of about –50 °C, as shown in Tables S6a18 and S6b. For
instance, the averaged
dimer/Agg1/Agg2 ratio of ca. 3:5:2 that was measured at –118 °C. Figure 2a displays the 1H
NMR spectrum of that mixture with values that differ substantially from those of the
δ values up to 25 °C were roughly compatible with the
δ
monomer/dimer mixture in Figure 2b. (The broad hump belongs to polymerized material, and
6
the trans-H doublet is broadened through a neighbourly magnetic interaction with Li.)
Replacing the excess portion of the t-BuOMe solvent by cyclopentane, we re-encountered
with Agg1 as a highly predominant species of 14d that exhibited practically the same 13C
NMR shifts at –80 °C (Table S7a18) as in t-BuOMe as the solvent. On the other hand, the
broadened two-proton NMR resonance of 2-/6-H was peculiarly temperature-dependent
(Table S7b18), perhaps due to the shortage of t-BuOMe (ca. 2.4 equiv) donor ligands in
cyclopentane. Nevertheless, the higher than dimeric aggregational state of Agg1 was clearly
evident from entry 13 (Table 1) and Figure 1h by the typically11 diminished ∆
δ magnitudes
of C-α, C-1, C-2/-6, C-3/-5, and C-4. It will be shown in the next section that Agg1 was
solvated by d = 1 t-BuOMe ligand per carbanion unit at +25, –25, and –41 °C.
9

3
3-/5-H
4
ACCEPTED MANUSCRIPT
2-/6-H
2
5
a) in t-BuOMe
.
.
Li
6
4-H
H
H
cis
trans
x
cis
trans
2-/6-H
3-/5-H
b) in THF
4-H
x
cis
trans
1
Fig. 2. H NMR spectra (400 MHz, 25 °C) of dimeric α-lithiostyrene (14d) in rapidly
equilibrating mixtures with its congeners: a) averaged with higher 14d aggregates (ca. 70%)
in t-BuOMe as the solvent; b) averaged with monomeric 14d (ca. 20%) in THF; x =
benzene.
This section revealed that the α-lithiostyrene family (14d) consisted of a least four
members, three of which (dimer, Agg1, and Agg2) coexisted in t-BuOMe as the solvent,
while the monomeric and dimeric species alone were met in THF. A shortage of the donor
ligand t-BuOMe in cyclopentane favored the preponderant species Agg1. With electron-
withdrawing para-substituents (4-SiMe3 and 4-Cl), 14a and 14b were predominantly
monomeric in THF but remained purely dimeric in Et2O as the solvent; unfortunately, they
were inconveniently prone to polymerization. The 4-CH3 derivative 14c was (like 10) one of
our first examples with a strong inclination toward dimerization in THF.
2
2.2. Differentials of JH,H can disclose microsolvation
Microsolvation numbers d of β-unsubstituted vinyllithium derivatives (“RLi”) may be
2
recognized by their linear relationship with the two-bond NMR coupling constants JH,H of
the CH2-β protons: The empirical11 eq 1 predicted that unsolvated (d = 0) samples should
2
display JH,H = 3.9 Hz, whereas the experimental values were 4.6 Hz15 for 11 and 4.5 Hz15
for 12 (Scheme 1). As a possible explanation, “solvation” by the contaminating LiBr (often
10

ca. 0.3 equiv) in the donor-free solutions may account for these somewhat oversized
ACCEPTED MANUSCRIPT
2
magnitudes of JH,H. In fact, this interference by LiBr disappeared in ethereal solvents that
solvated both LiBr and the organolithium species separately, so that their THF solutions
1
showed H NMR data that equaled those of LiBr-free 14d in THF.20 However, the solvation-
independent term 3.9 ppm in eq 1 may contain constitutional contributions that are also
2
present in the JH,H values of the appertaining “parent” alkenes (“RH”). This was hitherto of
2
little consequence for us as long as JH,H magnitudes of RH were hardly different: 2.2 (±0.2)
Hz for 2,6-dimethylstyrene21–23 (= RH of 5a), 1.7 Hz15 for RH of 11, 1.76 Hz15 for RH of
12, 1.9 Hz14 for RH of 8, and 1.7 Hz14 for RH of 9. However, we now have to consider18
2
the diminished JH,H magnitudes of other alkenes (RH) such as 15a (1.0 Hz),24 15b (0.87
2
Hz),25 15c (1.11 Hz),26 and 15d (1.09 Hz).26 Hence, the empirical eq 2 with JH,H
13
differentials (∆2JH,H
)
is now proposed as a numerically identical re-formulation of eq 1.
The new solvation-independent part [1.7(1) Hz in eq 2] may still contain smaller contributions
2
to JH,H from residual constitution- and conformation-dependent effects that are not equal for
2
RLi (14) and RH (15). For unsolvated 14a and 14b, eq 2 (d = 0) predicts JH,H = ca. 1.7 +
0.9 ≈ 2.6 Hz, whereas the experimental values were ca. 4 Hz in Section 2.1. Of course, we
must allow again for “solvation” by the contaminating LiBr in the “donor-free” hydrocarbon
solutions. (14c and 14d were not sufficiently soluble in donor-free, saturated hydrocarbons.)
As this LiBr effect vanished in ethereal solvents, as deduced above, eq 2 provided reliable
estimates for the microsolvation numbers d in the following examples. Column “d” in Table
1 displays the (numerically over-exact) results before rounding off: The monomeric species
are trisolvated (d ≈ 3 in entries 1, 3, 5, and 7), and monomeric 12 (entry 3) appeared to
populate its dimeric species at rt in THF. As usual, dimeric 5a was disolvated (d = 1 t-
BuOMe ligand in entry 2);11 hence it came as a surprise that the dimeric ortho-unsubstituted
α-lithiostyrenes 14a–c (entries 6, 8, and 9) appeared to be tetrasolvated by d = 2 ligands per
carbanion unit, presumably due to facilitated immobilization of a second “explicit” ligand per
2
Li center. Regrettably, JH,H coupling was not resolved for the two separated species of 14d
in THF; we therefore took recourse to comparisons of the ∆δ(C-α) and ∆δ(C-1) data which
suggested that monomeric 14d was trisolvated (entry 10 compared with 7) and that dimeric
14d was tetrasolvated in THF (entry 11 compared with 9 and 12). The unidentified higher
aggregate Agg1 of 14d turned out to be solvated by d = 1 t-BuOMe ligand in cyclopentane
solution (entry 13). Consequently, the mixture of tetrasolvated dimer (d = 2) with the higher
aggregated species in t-BuOMe should provide averaged d values between 1 and 2 under
conditions of rapid interconversion; this prediction agrees with the observed value of d = 1.8
(above –22 °C in entry 12).
11

2JH,H = d × 1.67 Hz + 3.9 Hz
(1)
(2)
ACCEPTED MANUSCRIPT
∆(2JH,H) = JH,H(RLi) – 2JH,H(RH) = d × 1.67 Hz + 1.7(1) Hz
2
Hexamethylphosphoramide (HMPA) is known to coordinate much more strongly than
THF to Li cations.27 Thus, a modest concentration of HMPA (1.1 M, ca. 2 equiv) sufficed
apparently to shift the mobile monomer/dimer equilibrium (ca. 1:4) of 14d in THF at rt
2
toward the trisolvated monomer with JH,H = 8.0 Hz (formally d = 3.1 by eq 2). Actually, a
somewhat higher portion of HMPA (3 equiv) created the lithiation shift ∆δ(4-H) = –0.76 ppm
for 14d in THF; this value exceeded the magnitude of ∆ (4-H) = –0.61 ppm for 14d in the
δ
absence of HMPA (entry 10, or Figure 1f). This demonstrated that HMPA-solvated Li+ is
more electron-releasing than THF-solvated Li+, which is understandable because the oxygen
atom in HMPA carries a formal negative charge. As a consequence, this reduced
2
electronegativity of HMPA-solvated Li+ may exaggerate the magnitude of JH,H that
depends the σ-inductive effect within the double-bond plane, as explained previously.11
2
Indeed, still higher HMPA concentrations increased JH,H up to at least 8.9 Hz at –74 °C, for
which eq 2 would predict an overestimated microsolvation number of d = 3.7; this suggested
that eq 2 would require a modified parametrization for counting electrically charged donor
centers such as in HMPA or the “solvating” LiBr of Section 2.1. However, such a
modification was deemed unnecessary as far as the microsolvation numbers d = 1 – 4 of
HMPA could be determined directly through inspection27 of Li/31P NMR coupling patterns.
This section illustrated the applicability of the empirical eq 2 that infers microsolvation
numbers from “differential” two-bond inter-proton coupling constants ∆2JH,H of β-
unsubstituted vinyllithiums. Experimental problems arose occasionally through serious line
broadening of this olefinic =CH2 AB spectral system in 14.
2.3. Cis/trans stereoinversion in THF
The pseudomonomolecular, ionic mechanism8,11 of cis/trans stereoinversion (Scheme 3)
was established for the trisolvated, monomeric species of the alkenyllithiums 1a–e, 2, 3a–b,
5a–c, 9, 11, and 12 in THF as the solvent. It proceeds via an NMR-invisible, solvent-
separated ion pair with transitory immobilization of a fourth THF ligand at Li; therefore, it is
catalyzed by THF at the expense of a characteristically negative pseudoactivation entropy of
∆Sψ = ca. –23(3) cal mol–1 K–1. For α-arylvinyllithiums such as 5, 9, 11, or 12, this
‡
stereoinversion interconverted the cis and trans environments of the two diastereotopic β-H
nuclei. Depending on the applied magnetic field strength and the temperatures, increasing
rates of this “diastereotopomerization” will first broaden the four-line (AB type, Section 2.2)
proton pattern and then lead to coalescence into a singlet signal at the resonance position of
12

(
δA
+
δB)/2.28 Computer-aided total line-shape simulations29 afforded the pseudo-first-order
ACCEPTED MANUSCRIPT
rate constants kψ which depend on the THF concentration and on the Kelvin temperatures T.
The linear correlation of the logarithms of these kψ values (Tables S9–S1218) versus 1/T, as
‡
illustrated by the Arrhenius plots in Figure 3, afforded the pseudoactivation parameters ∆Gψ
‡
‡
= ∆Hψ – T×∆Sψ for the stereoinversion of 14a, b, and d in THF. The predominantly
monomeric α-[4-(trimethylsilyl)phenyl]vinyllithium (14a, entry 4 of Table 2) could be
measured at temperatures up to 62 °C and turned out to be the fastest member of this series
(line c of Figure 3, and entries 4–6 of Table 2). The more readily polymerizing α-(4-
chlorophenyl)vinyllithium (14b) inverted a little less rapidly (Figure 3, line b) and furnished
‡
practically the same pseudoactivation entropy ∆Sψ as 14a (entries 4 and 5). Both of these
‡
‡
∆Sψ values were significantly more negative than the above-mentioned benchmark of ∆Sψ
= ca. –23(3) cal mol–1 K–1 and the values in entries 1 – 3. Since these entropy values are
partially due to immobilization of the fourth THF ligand in an ion-pair intermediate, their
increased magnitudes in entries 4 and 5 may be understood as a hint at the possiblility that the
stereoinversions of 14a and 14b (Scheme 3) depended on both the trisolvated monomeric (d =
3) and the tetrasolvated dimeric (d = 2) species: The transitory immobilization of more than
one THF ligand in order to generate Li+(THF)4 from the dimeric portion would then create an
additional entropic penalty. Compared with 5a (entry 1) and 12 (entry 2), the strongly
‡
diminished pseudoactivation enthalpies ∆Hψ in entries 4 and 5 were caused partially by the
accelerating effect of the para-substituents 4-SiMe3 and 4-Cl, respectively, as expected from
the following established8 trait of the ionic stereoinversion mechanism. The appertaining
Hammett reaction constant
mechanistic evidence; as far as it is applicable also to the α-lithiostyrenes 14a–d, it would
ρ
= +5.2 (discovered8 with 1a–e) is an essential piece of
–1
‡
predict for 14d a ∆Gψ (0 °C) barrier that should be higher by ca. 1 kcal mol than those of
the more rapidly inverting 14a and 14b (ca. 14.8 kcal mol–1 in entries 4 and 5). Such an
estimate would approach the barrier of ca. 16.2 kcal mol–1 (entries 1 and 2) that could have
been expected for 14d because ortho-CH3 groups do not14 change the barriers of 5a (with
ortho,ortho´-dimethyl) and 12 (with one free ortho-position). Instead, the observed value
(entry 6) was even higher by almost 1.5 kcal mol–1, which suggests to look more closely at the
kinetic problems with 14d in the sequel.
13

R
R = H, Cl, SiMe3
4
ACCEPTED MANUSCRIPT
R
4
R
R
4
4
+
Li (THF)4
2
1
1
1
1
.
.
.
.
(THF)
(THF)
Li
.
Li
(THF)3 Li
H
.
3
0.5
2
-
THF
α
α
α
.
.
β
β
H
β
H
H
H
H
-
+ THF
-
THF
+ THF
α
cis
trans cis
trans
cis
trans
β
H
H
2
Transition state
Monomer in THF
Dimer in THF
Monomer in THF
Scheme 3. The rapidly reversible formation of tetrasolvated dimers withdraws
fractions of the trisolvated monomers of 14a, b,or d from their
pseudomonomolecular, ionic cis/trans stereoinversion process in THF as the solvent.
–1
–1
‡
‡
Table 2. Pseudoactivation parameters ∆Gψ (kcal mol at 0 °C), ∆Hψ (kcal mol ), and
–1 –1
‡
∆Sψ (cal mol K ) for the cis/trans diastereotopomerization rates of α-arylvinyllithiums
5a, 9, 12, 14a, 14b, and 14d in THF.
entry compd aryl group a
HT b Ref.
+5.0 11
‡
‡
‡
∆Gψ (0 °C)
∆Hψ
∆Sψ
1
2
3
4
5
6
7
2,6-Me2Ph
2-MePh
5a
16.3 (±0.1) 10.2 (±0.7) –22.2 (±2.2)
16.14 (±0.05) 10.6 (±0.3) –20.2 (±0.8)
12.94 (±0.01) 6.9 (±0.1) –22.3 (±0.4)
14.57 (±0.05) 6.4 (±0.4) –29.8 (±1.2)
15.06 (±0.03) 7.3 (±0.4) –28.3 (±1.3)
17.68 (±0.05) 7.9 (±0.3) –35.9 (±1.1)
13.95 (±0.00) 6.1 (±0.3) –28.9 (±0.9)
+74.0 15
+28.0 14
12
naphthyl-1
4-Me3SiPh
4-ClPh
9
+62.0
+35.5
+80.0
+28.0
c
c
c
c
14a
14b
14d d phenyl
14d e
phenyl
a Me = methyl, Ph = phenyl. b HT = highest temperature (°C) of the rate measurements.
e
c This work. d [14d] = 0.6 M. [14d] = 0.57 M, [HMPA] = 1.1 M, [LiBr] = 0.7 M,
2JH,H = 8.5 Hz.
14

10log kψ
2.0
AcC) CR E= SPiTMeE3D MANUSCRIPT
d) R = H with HMPA
1.6
b) R = Cl
1.2
0.8
0.4
0.0
a) R = H
R
R
.
.
.
.
Li
Li
H
H
H
H
-0.4
-0.8
in THF
3.0
2.8
4.0
3.2
3.6
3.8
3.4
1000/T
Fig. 3. Temperature-dependent decadic logarithms (10log) of the pseudo-first-order
stereoinversion rate constants kψ (s–1) of the α-lithiostyrenes 14a, 14b, and 14d in THF
(Tables S9–S1218): (a) 14d; (b) 14b; (c) 14a; (d) 14d with HMPA (2 equiv).
Our stereoinversion studies with α-lithiostyrene (14d) posed the theoretical problem that it
would not be possible to obtain detailed knowledge about the individual contributions of the
monomeric and dimeric components even though dimeric22 and higher aggregates are
known11 to invert more slowly (by unidentified mechanisms) than the monomers.
‡
Furthermore, experimental confirmations of the above-considered increase of ∆Gψ (0 °C) for
14d were foreshadowed by the practical problem of measuring stereoinversion rate constants
kψ that were often not fast enough to generate suitable line-broadening in the lower
temperature regions even at NMR machine frequencies as small as 100 or 60 MHz.
Therefore, we used also the double-resonance NMR-relaxation technique that was described
by Forsén and Hoffman30a and improved by Anet and Bourn.30b We applied this saturation-
transfer method to the equilibrating mixture (47:53) of (Z)- and (E)-[β-D1]14d that had been
prepared from (E,Z)-[β-D1]13d18 (Scheme 4). With deuterium decoupling of the two =CHD
1
proton singlets of (Z,E)-[β-D1]14d in THF, a sudden selective H irradiation (“saturation”) of
the Z singlet reduced the intensity of the E singlet (or vice versa) at a rate that depended on
the Z → E (or E → Z, respectively) stereoinversion rate. For simplicity, we report (Table
S1118) kψ values that are weighted averages of the slightly differing Z → E and E → Z rate
constants (whose ratio accounts for the above equilibrium mixture). The temperature
dependence of kψ (line a of Figure 3) afforded the pseudoactivation parameters for the
15

monomer/dimer mixture listed in entry 6 of Table 2: Compared with monomeric 5a (entry 1),
ACCEPTED MANUSCRIPT
the dimeric fraction of 14d appeared to have lowered both ∆Hψ (by more than 2 kcal mol–1)
‡
and ∆Sψ (by almost 14 units); these trends agree with the usual17 exothermic and entropy-
‡
consuming deaggregation of dimeric α-arylalkenyllithiums in THF.
Br
D
Ph
H
Ph
D
Br
H
E/Z = 63:37
(E)-[β-D1]13d
(Z)-[β-D1]13d
+ n-BuLi
-
n-BuBr
Li
D
Ph
H
Ph
D
Li
H
Z/E = 47:53
(Z)-[β-D1]14d
(E)-[β-D1]14d
Scheme 4. Preparation and Z,E-stereoinversion equilibrium of [β-D1]14d.
Should we interpret these 14d properties on the basis of the ionic stereoinversion
mechanism that applies to 5a? Admixtures of dry LiBr to 14d in THF appeared to cause
modest (up to 6-fold) accelerations that resembled those observed31 for 5a, whereas additions
of dry (aprotic) LiOt-Bu did not change the rate-broadened line shapes; this absence of an
equal-cation-caused retardation effect agreed with the absence of an ionic dissociation step
from the established8,11 ionic stereoinversion mechanism. On the other hand, additions of
increasing portions of aprotic KOt-Bu (up to 1 equiv) accelerated not only the stereoinversion
dramatically (up to > 200-fold) but also the decomposition of 14d. Thus, an increased ionic
character of the (presumably monomeric) alkenylpotassium compound facilitated the
stereoinversion. An even stronger acceleration by almost 3 orders of magnitude can be
visualized by comparing the log kψ data of 14d in line a of Figure 3 with those in line d of
Figure 3 for 14d in the presence of HMPA (2 equiv). The latter system diastereotopomerized
rapidly enough at low temperatures to permit the use of a few line-shape simulations that
yielded concordant kψ values for [β-D1]14d and unlabeled 14d (hence, no perceptible kinetic
isotope effect). The pseudoactivation parameters of such a run (entry 7 of Table 2) suggested
‡
a twofold effect of HMPA: Compared with entry 6, the further decrease of ∆Hψ may be due
to stronger energetic stabilization of both the ion-pair intermediate and the stereoinversion
transition state than of the ground state of 14d. Energetic ground state stabilization by HMPA
(2 equiv) was also supposed in Section 2.2 to decrease the dimer population; this second
‡
effect can account for the reduced magnitude of ∆Sψ (entry 7) in the direction toward the
entropies of the monomers in entries 1–3. Regrettably, an attempted closer approach to the
16

entropy bench mark values of –23(3) cal mol–1 K–1 was unattainable because higher HMPA
ACCEPTED MANUSCRIPT
concentrations (> 3 equiv) led to the increasingly rapid decomposition of 14d and to ill-
defined activation parameters.16 However, the observed pseudoactivation parameters with
HMPA (entry 7) appeared to be still compatible with the ionic mechanism.32
This section illustrated that an equilibrating system of monomeric and dimeric species of
the α-lithiostyrene 14d can impair numerically more precise interpretations of the kinetic
results. Nevertheless, we could demonstrate that 14d satisfies several criteria for the ionic
mechanism: The rate constants kψ were higher in THF than in Et2O or t-BuOMe as the
solvents; they stayed unchanged with LiOt-Bu as an equal-cation additive in THF but were
substantially increased by added HMPA ligands or by π-electron-withdrawing para-
substituents, while a para-methyl group in dimeric 14c retarded the inversion so much that 1H
NMR line-broadening could not be detected up to 73 °C in THF. However, all observed kψ
data were averages of those of the smaller monomeric fraction (≤ 20%) and of the
significantly lower rate constants22 of a major dimeric fraction. As a consequence, kψ must
decrease when the monomeric population decreases with increasing concentrations of 14d.
–1
‡
Purely monomeric 14d should invert with the same ∆Gψ (0 °C) ≈ 16.2 kcal mol as 5a and
–1
‡
12, whereas the experimental value (entry 6) was ∆Gψ (0 °C) = 17.68(5) kcal mol ; the rise
by 1.5 kcal mol–1 can be ascribed to the presence of dimeric 14d that participated indirectly
in the stereoinversion process through deaggregation.
3. Conclusion
(i) The structural assignments by secondary NMR criteria (∆δ values) and the conclusions
in this work were based on ample background knowledge that had been collected from the
analytically more suitable systems shown in Scheme 1. The more complicated (because
rapidly scrambling) systems of 14d in THF or t-BuOMe might be suited for additional
reactivity studies by the Rapid Injection NMR33 technique.
(ii) In contrast to the purely monomeric nature of α-(2,6-dimethylphenyl)vinyllithium (5a)
and several other alkenyllithiums (Scheme 1), the ortho,ortho´-unsubstituted α-lithiostyrene
(14d) formed a monomer/dimer mixture (ca. 1:4 at 25 °C in units of the monomer formula) in
THF as the solvent. This very mobile dimerization equilibrium required temperatures below
–101 °C for detecting and assigning the two components. The equilibrium ratios changed
with the nature of the para-positioned substituents: The electron-withdrawing 4-SiMe3 and
4-Cl groups preferred an increased monomer population in THF, whereas the electron-
repelling 4-CH3 group tolerated only the dimeric species. Modest portions of HMPA (2–3
equiv) appeared to deaggregate the dimeric species of 14d in THF.
17

(iii) In t-BuOMe as the solvent, the equilibrating mixture of dimeric 14d (ca. 30%) and
ACCEPTED MANUSCRIPT
two unidentified aggregates (ca. 50 and 20%) “froze” already below –60 °C and provided no
NMR-evidence for the presence of a monomeric species. Only this major, t-BuOMe-solvated
aggregate was found in cyclopentane as the solvent.34
(iv) In analogy with the examples depicted in Scheme 1, the monomeric species of 14a, b,
and d were detected to be trisolvated in THF. These microsolvation numbers (d = 3) of
electrically uncharged donor centers (excluding HMPA) could not be measured through NMR
integration; instead, they were calculated from the magnitudes of the H2C= interproton
2
coupling constants JH,H (eq 1) or from the numerically equivalent empirical relationship (eq
2
2
2) that uses the difference ∆2JH,H = JH,H(RLi) – JH,H(RH) = d × 1.67 Hz + 1.7(1) Hz of
RLi and its “parent” RH. Unlike all disolvated dimeric ortho-substituted examples in Scheme
1, the dimers of the o,o´-unsubstituted α-lithiostyrenes 14a–d were found to be tetrasolvated
with the microsolvation number d = 2 electron-pair donor ligands per Li. Eq 2 served also to
recognize d = 1 t-BuOMe ligand per carbanion unit of the preponderant higher aggregate of
14d in cyclopentane.
(v) The interconversion of the cis and trans environments of the H2C= protons of the α-
lithiostyrenes 14a, b, and d occurred most rapidly in THF as the solvent. In accord with the
established pseudomonomolecular, ionic stereoinversion mechanism,8 the para-substituents
SiMe3 and Cl accelerated this cis/trans interconversion, and the absence of ortho-substituents
did not facilitate this process. Although the very mobile dimerization equilibrium in THF
prevented detailed interpretations of the rate constants, their apparent pseudoactivation
parameters in Table 2 (entries 4 – 7) were compatible with the ionic mechanism8 The
addition of KOt-Bu or HMPA accelerated both the cis,trans stereoinversion and the
decomposition of 14d strongly.
(vi) Since the above cis,trans stereoinversion mechanism8 applies to monomeric α-
lithiostyrene (14d), the pseudoactivation parameters in entry 6 of Table 2 can provide the
following answer to the title question. As far as dimeric 14d in THF is much less stereolabile
‡
than the monomer, its rapidly reversible deaggregation lowered ∆Sψ (by 13.7 units) more
than ∆Hψ (by 2.3 kcal mol–1), both relative to purely monomeric 5a in entry 1. This
‡
–1
‡
retardation by an increase of ∆∆Gψ (0 °C) ≈ 1.4 kcal mol can be ascribed to the kinetically
inactive fraction (namely, dimeric 14d) that reduced the active monomeric population of 14d.
In short, the stereoinversion rate of the partially monomeric 14d was accelerated by two
ortho-methyl groups in 5a because they made their ground state of 5a fully monomeric.35
18

ACCEPTED MANUSCRIPT
4. Experimental
4.1. General information
The previously7,8,12 described methods were applied to perform the preparations and
1
reactions of organolithium samples in NMR tubes under argon gas cover. All H and 13C
chemical shifts were referenced to internal Me4Si; abbreviations for the signal multiplicities
are as follows: d = doublet, dd = doublet of doublets, ddt = doublet of doublets of triplets, dm
= doublet of multiplets, dtd = doublet of triplets of doublets, m = multiplet, q = quartet, qi =
quintet, quat = quaternary, s = singlet, sept = septet, t = triplet. All concentrations and
populations were counted in units of the monomer formula.
4.2.
α-Bromo-4-(trimethylsilyl)styrene (13a)
1,1-Dibromo-1-[4´-(trimethylsilyl)phenyl]ethane (388 mg, 1.15 mmol), prepared18 from
4-(trimethylsilyl)acetophenone,36 was dissolved in anhydrous N,N-dimethylformamide (2.5
mL) and heated to 90 °C under dry inert gas for 18 min (method A2 in Ref. 37). The cooled
mixture was dissolved in Et2O (25 mL), washed with distilled water (5 × 6 mL), dried over
MgSO4, filtered, concentrated under dry inert gas, and dried for 1 h under reduced pressure at
rt. The crude material (290 mg) became oxidized easily on contact with air to give α-bromo-
4-(trimethylsilyl)acetophenone (δH = 0.30 and 4.20 ppm) that distilled together with the
known38 liquid product 13a (bp 47 °C/0.005 mbar).
1H NMR of 13a (CDCl3, 400 MHz)
δ 0.27 (s, 9H, Me3Si), 5.77 and 6.12 (AB spectral
system, 2J = 2.0 Hz, 1 + 1H, olefinic CH2-β), 7.50 and 7.56 (AA´BB´ system, 3J = 8.4 Hz, 2 ×
2H, 2-/3-/5-/6-H) ppm;
13C NMR of 13a (CDCl3, 100.6 MHz)
δ
–1.2 (qsept, 1J = 119.5 Hz, 3J = 2.0 Hz, Me3Si),
1
117.6 (dd, apparent J = 167.1 and 158.4 Hz, CH2-β), 126.5 (ddt, 1J = 159.8 Hz, 3J = 5.2 Hz,
4
3
apparent J = ca. 2 Hz, C-2/-6), 131.1 (pseudo-qi, apparent J = 5 Hz, C-α), 133.3 (sharp dd,
1J = 158.5 Hz, 3J = 8.6 Hz, C-3/-5), 138.8 (dtd, 3J = 8.4, 7.8, and 3.6 Hz, C-1), 142.0 (≥
3
dodecet through J to 2-/6-H and the Me3Si protons, C-4) ppm; assigned through the C,H
coupling patterns.
4.3.
α-Bromo-4-chlorostyrene (13b)
The above protocol for 13a was applied to crude 1,1-dibromo-1-(4´-chlorophenyl)ethane18
(590 mg, 1.98 mmol) in anhydrous N,N-dimethylformamide (2.0 mL), yielding 1H-NMR-
spectroscpically almost pure 13b39 (390 mg, 91%) if kept always under inert gas cover. It
became oxidized to α-bromo-4-chloroacetophenone (mp 88–92 °C, δH = 4.23 ppm) on contact
1
with air. H NMR of 13b (CCl4, 80 MHz)
δ
5.70 and 6.02 (AB spectral system, 2J = 1.8 Hz,
19

1 + 1H, olefinic CH2-β), 7.20 and 7.45 (AA´BB´ system, 3J = 8.5 Hz, 2 × 2H, 2-/3-/5-/6-H)
ACCEPTED MANUSCRIPT
ppm. This material was dried under reduced pressure of inert gas for the subsequent
preparation of 14b.
4.4.
α-[4-(Trimethylsilyl)phenyl]vinyllithium (14a)
4.4.1. In donor-free pentane. A dry NMR tube (5 mm) was charged with the α-
bromoalkene 13a (47 mg, 0.18 mmol) in pentane (0.5 mL), cooled to –78 °C under argon gas
cover, and treated with n-Bu6Li (0.18 mmol) in hexane (0.074 mL). After 9 min at 32 °C, the
1H NMR spectra (80 MHz) showed 13a and the product 14a in a ca. 45:55 ratio;
polymerization began to interfere after 20 min. Nevertheless, the moderately broadened
spectrum of 14a could be analyzed after 15 h at rt despite the presence of its coproduct n-
1
BuBr. H NMR of 14a (pentane, 80 MHz)
δ
0.18 (s, Me3Si), 5.54 (d, 2J = 4.0 Hz, β-H trans
to α-aryl), 5.97 (d, 2J = 4.0 Hz, β-H cis), 6.66 (d, 3J = 7.5 Hz, 2-/6-H), 7.24 (d, 3J = 7.5 Hz, 3-
/5-H) ppm.
4.4.2. In Et2O. A dry NMR tube (5 mm) was charged with the α-bromoalkene 13a (44
mg, 0.17 mmol) in anhydrous Et2O (ca. 0.5 mL), cooled to –78 °C under argon gas cover,
and treated with n-Bu6Li (0.21 mmol) in hexane (0.23 mL). Placed in a large Schlenk tube
that was filled with argon gas and held in a tilted position (with a ca. 20° angle above
horizontal), the opened NMR tube was cautiously evacuated down to < 0.1 mbar at rt. After
venting with dry argon gas, the NMR tube was cooled while its liquid contents were dissolved
in anhydrous Et2O or THF together with 0.060 mL of [D12]cyclohexane.
1H NMR (Et2O, 400 MHz, 25 °C)
δ 0.21 (s, ca. 9H, Me3Si), 5.64 (broadened, 1H, β-H
trans to α-aryl), 6.09 (broadened d, 2J = 5.5 Hz, 1H, β-H cis), 6.82 (d, 3J = 7.5 Hz, 2H, 2-/6-
H), 7.23 (d, 3J = 7.5 Hz, 2H, 3-/5-H) ppm;
1H NMR (THF, 400 MHz, 25 °C)
δ 0.17 (s, 9H, Me3Si), 5.25 (broad d, 1H, β-H trans to
α-aryl), 5.82 (broadened d, 2J = 8.0 Hz, 1H, β-H cis), 6.71 (d, 3J = 7.9 Hz, 2H, 2-/6-H), 7.11
(d, 3J = 7.9 Hz, 2H, 3-/5-H) ppm;
13C NMR (Et2O, 100.6 MHz, 25 °C)
δ –0.6 (Me3Si), 123.1 (CH2-β), 124.2 (C-2/-6),
133.7 (C-3/-5), 159.6 (C-1) ppm, C-4 and C-α not detected, assigned through comparison
with 14a in THF;
13C NMR (THF, 100.6 MHz, 25 °C)
δ
–0.5 (qsept, 1J = 118.5 Hz, 3J = 2 Hz, Me3Si),
117.8 (broadened t, 1J = 145 Hz, CH2-β), 123.2 (dd, 1J = 154 Hz, 3J = ca. 6 Hz, C-2/-6), 128.1
(quat, C-4), 132.7 (dm, 1J = 151 Hz, 3J = ca. 8 Hz, C-3/-5), 164.7 (quat, C-1), 207.1 (quat, C-
α) ppm, assigned through the C,H coupling constants and comparison with 1e.
4.5.
α-(4-Chlorophenyl)vinyllithium (14b)
20

The above protocol for 14a was used to generate 14b in Et2O solution from α-
ACCEPTED MANUSCRIPT
bromoalkene 13b and n-Bu6Li, followed by evaporation of n-BuBr together with all other
volatiles. The remaining liquid 14b and the accompanying LiBr were dissolved in either
anhydrous Et2O or THF.
2
1H NMR (Et2O, 400 MHz, 25 °C)
δ
5.69 (broadened d, J = 5.8 Hz, 1H, β-H trans to α-
aryl), 6.11 (d, 2J = 5.8 Hz, 1H, β-H cis), 6.86 (d, 3J = 8.4 Hz, 2H, 2-/6-H), 7.03 (d, 3J = 8.4
Hz, 2H, 3-/5-H) ppm, assigned through comparison with 14a;
1H NMR (THF, 400 MHz, –44 °C)
δ
5.12 (d, 2J = 8.0 Hz, 1H, β-H trans to α-aryl), 5.72
(d, 2J = 8.0 Hz, 1H, β-H cis), 6.61 (d, 3J = 8.4 Hz, 2H, 2-/6-H), 6.86 (d, 3J = 8.4 Hz, 2H, 3-/5-
H) ppm, assigned as above;
13C NMR (Et2O, 100.6 MHz, 25 °C)
δ
124.3 (broadened t, 1J = 146 Hz, CH2-β), 126.1
(dd, 1J = 158.5 Hz, 3J = 6.8 Hz, C-2/-6), 128.3 (dd, 1J = 163.0 Hz, 3J = 4.7 Hz, C-3/-5), 128.5
(quat, C-4), 157.6 (quat, C-1), 199.4 (broad, C-α) ppm, assigned through the C,H coupling
constants (especially those of C-3/-5);
13C NMR (THF, 100.6 MHz, –44 °C)
δ
117.1 (dd, apparent 1J = ca. 152 and 138 Hz,
CH2-β), 124.0 (broadened t, 3J = 10.5 Hz, C-4), 124.5 (sharp dd, 1J = 156 Hz, 3J = 6.8 Hz, C-
2/-6), 127.1 (dm, 1J = 159 Hz, C-3/-5), 163.3 (quat, C-1), 207.4 (broadened, C-α) ppm,
assigned as above (especially for C-4) and comparison with 14a.
4.6.
α-(4-Methylphenyl)vinyllithium (14c)
A stirred solution of α-bromo-4-methylstyrene (13c, 890 mg, 4.51 mmol) in pentane (2.0
mL) was cooled to –78 °C under argon gas cover and treated with n-Bu6Li (5.4 mmol) in
isopentane (2.1 mL). This Br/Li interchange reaction proceeded smoothly during the first 5
min and was complete within 60 min at rt. Without delay, the precipitated powder of 14c was
washed with pentane, then dried through evacuation down to < 0.1 Torr, vented with dry
argon gas, and dissolved in anhydrous THF. After the NMR studies and later aqueous work-
up of the colored solution, a potentiometric titration indicated that the sample had contained
0.23 mmol of LiBr (0.13 equiv).
1H NMR (THF, 400 MHz, 25 °C)
δ
2.18 (s, 3H, 4-CH3), 5.34 (d, 2J = 7.5 Hz, 1H, β-H
trans to α-aryl), 5.85 (d, 2J = 7.5 Hz, 1H, β-H cis), 6.65 (d, 3J = 8.0 Hz, 2H, 2-/6-H), 6.71 (d,
3J = 8.0 Hz, 2H, 3-/5-H) ppm, assigned through comparison with 14a;
13C NMR (THF, 25.2 MHz, –44 °C)
δ
21.1 (q, 1J = 126 Hz, 4-CH3), 118.4 (dd, apparent
1J = 146 and 136 Hz, CH2-β), 124.1 (dd, 1J = 155 Hz, 3J = 4.5 Hz, C-2/-6), 127.7 (broadened
d, 1J = 151 Hz, C-3/-5 with unresolved 3J couplings to CH3 and 5-/3-H), 127.7 (broadened m,
apprent J = ca. 5 Hz, C-4), 160.3 (quat, C-1), 206.2 (quat, C-α) ppm, assigned through the
C,H coupling patterns.
21

ACCEPTED MANUSCRIPT
4.7.
α-Lithiostyrene (14d)
α-Bromostyrene (13d, 150 mg, 0.82 mmol) and cyclopentane (0.10 mL) were placed in a
dry NMR tube (5 mm) under argon gas cover, cooled to –60 °C, and treated with n-Bu6Li
(0.85 mmol) in cyclopentane (0.94 mL). After up to 15 h at rt, the colorless precipitate of 14d
and some LiBr was washed with dry cyclopentane (3 × 0.3 mL), then cooled to –30 °C and
dissolved in either anhydrous THF or t-BuOMe. Caution: Dry 14d may decompose
exothermically in contact with only a trace of air! Therefore, it should not be stored under a
reduced pressure of an inert gas.
1H NMR (THF, 400 MHz, 25 °C, Figure 2b)
δ
5.34 (d, 2J = 7.5 Hz, 1H, β-H trans to α-
aryl), 5.90 (d, 2J = 7.5 Hz, 1H, β-H cis), 6.62 (tt, 3J = 7.5 Hz, 1H, 4-H), 6.77 (dm, 3J = 7.5 Hz,
2H, 2-/6-H), 6.94 (t, 3J = 7.5 Hz, 2H, 3-/5-H) ppm, assigned through HOESY correlation of
6Li with 2-/6-H and β-H trans;
1H NMR (t-BuOMe, 400 MHz, 25 °C, Figure 2a)
δ
5.70 (d, 2J = 5.8 Hz, 1H, β-H trans to
α-aryl), 6.16 (sharp d, 2J = 5.8 Hz, 1H, β-H cis), 6.80 (tt, 3J = 7.5 Hz, 1H, 4-H), 6.86 (d, 3J =
7.5 Hz, 2H, 2-/6-H), 7.02 (t, 3J = 7.5 Hz, 2H, 3-/5-H) ppm, assigned as above;
1H NMR (cyclopentane with 2.4 equiv of t-BuOMe, 400 MHz, 25 °C)
δ 5.96 (broad, 1H,
β-H trans to α-aryl), 6.19 (broadened d, 2J = 4.8 Hz, 1H, β-H cis), 6.63 (broadened d, 2H, 2-
/6-H), 6.80 (t, 3J = 7.3 Hz, 1H, 4-H), 6.91 (m, 2H, 3-/5-H) ppm, assigned as above;
1H NMR (THF, 60 MHz, –17 °C) of 14d with HMPA (0.47 M, 3 equiv)
δ 6.39 (4-H),
6.55 (2-/6-H), 6.75 (3-/5-H) ppm, assigned through spectral simulation (Figure S6).18 The
olefinic AB spectum was already in coalescence at this temperature and magnetic field
strength, showing a very broad resonance centered at
δ = 5.33 ppm;
13C NMR (THF, 100.6 MHz, 25 °C) 119.5 (t, 1J = ca. 144 Hz, CH2-β), 120.3 (dt, 1J =
δ
156 Hz, 3J = 7.5 Hz, C-4), 123.9 (dt, 1J = 156 Hz, 3J = 7 Hz, C-2/-6), 127.5 (dd, 1J = 155 Hz,
3J = 7.6 Hz, C-3/-5), 163.1 (ill-resolved m, apparent 3J = ca. 6 Hz, C-1), 205.6 (obscured t, C-
α) ppm, assigned through the C,H coupling patterns;
13C NMR (t-BuOMe, 100.6 MHz, 25 °C)
δ
122.6 (dt, 1J = 155.8 Hz, C-4), 123.2
(broadened t, 1J = 146 Hz, CH2-β), 124.8 (dt, 1J = 155.7 Hz, 3J = 7 Hz, C-2/-6), 128.1 (dm, 1J
= 156 Hz, C-3/-5), 157.8 (unresolved m, C-1), 199.7 (broad, C-α) ppm, assigned as above;
13C NMR (cyclopentane with 2.4 equiv of t-BuOMe, 100.6 MHz, 25 °C)
δ
123.4 (dt, 1J =
ca. 160 Hz, 3J = 7 Hz, C-4), 124.9 (dm, 1J = 158 Hz, C-2/-6), 126.9 (observed at and below –
25 °C, CH2-β), 128.9 (dd, 1J = 155 Hz, 3J = 8 Hz, C-3/-5), 156.7 (unresolved m, C-1), 196.2
(unresolved, C-α) ppm, but C-α at 194.0 ppm/–41 °C, assigned through the C,H coupling
constants.
22

ACCEPTED MANUSCRIPT
4.8.
α-[4-(Trimethylsilyl)phenyl]acrylic acid (16a)
A solution of the α-bromoalkene 13a (170 mg, 0.67 mmol) in anhydrous Et2O (2.5 mL)
was cooled to –65 °C, treated with n-BuLi (0.67 mmol) in hexane (0.27 mL), and after 1 h at
–65 °C poured onto solid CO2. The warmed-up mixture was dissolved in Et2O (40 mL) and
aqueous NaOH (2 M, 15 mL); the Et2O layer was extracted with aqueous NaOH (2 M, 4 × 15
mL) and set aside. The combined aqueous extracts were acidified with conc. HCl and shaken
with Et2O (4 × 20 mL). These Et2O extracts were combined, washed with distilled water (2 ×
20 mL), dried over MgSO4, filtered, and evaporated to give the crude acid 16a (39 mg). The
above first Et2O layer was washed with distilled water until neutral, dried over MgSO4,
filtered, and concentrated to yield a mixture (61 mg) of the “parent” alkene 15a (ca. 45%) and
polymerized material. Analytically pure 16a was obtained through recrystallization from
high-boiling petroleum ether (2 ×) and had mp ca. 114 °C.
1H NMR (CDCl3, 400 MHz)
δ
0.27 (s, 9H, Me3Si), 6.03 and 6.54 (AB spectral system, 2J
= 1.1 Hz, 1 + 1H, 2 × β-H), 7.42 and 7.53 (AA´BB´ system, 3J = 8.2 Hz, 2 + 2H, 2-/3-/5-/6-
H), 10.4 (s, 1H, CO2H) ppm;
13C NMR (CDCl3, 100.6 MHz)
δ
–1.2 (qsept, 1J = 119.2 Hz, 3J = 2.1 Hz, Me3Si), 127.7
1
(dd, 1J = 159.8 Hz, 3J = 6.1 Hz, C-2/-6), 129.4 (dd, apparent J = 163.6 and 160.2 Hz, CH2-
β), 133.2 (dd, 1J = 158.3 Hz, 3J = 8.7 Hz, C-3/-5), 136.4 (broad m, C-1), 140.62 (narrow m,
C-α), 140.84 (broad m, C-4), 172.0 (dd, 3J = 12.5 and 6.5 Hz, CO2H) ppm, assigned through
the C,H coupling constants and comparison with the alkene 15a;
IR (KBr) ν 3300–2800 (broad O–H), 2956, 1700, 1680 (s), 1251, 1226, 1077, 843, 827
cm–1. Anal Calcd for C12H16O2 (220.3): C, 65.41; H, 7.32. Found: C, 65.82; H, 7.27.
Appendix A. Supplementary Data
NMR data of the styrenes 15a-d and the acid 16c; α-bromostyrenes 13a,b via 1-aryl-1,1-
dibromoethanes; deuteriated α-bromostyrenes (E)- and (Z)-[β-D1]14d; dialkyl ketone
adducts S8–S10; LiBr-free 14d; tabulated primary NMR data (Tables S1a – S7b);
1
simulations of some H NMR spectra of 14d (Table S8); Rate Measurements (Tables S9–
S12).
5. References
1. a) Lithium Compounds in Organic Synthesis – From Fundamentals to Applications,
Luisi, R., Capriati, V., Eds, Wiley-VCH, 2014;
b) Capriati V, Perna FM, Salomone A. J Chem Soc. Dalton Trans. 2014;43:14204–
14210.
23

2. Reich HJ. Chem Rev. 2013;113:7130–7178.
ACCEPTED MANUSCRIPT
3. Gawley RE. Topics Stereochem. 2010;26:93–133.
4. Gessner VH, Däschlein C, Strohmann C. Chem Eur J. 2009;15:3320–3334.
5. a) Lithium Chemistry, Sapse, A-M, Schleyer, PvonR. Eds. Wiley: New York, 1995;
b) Lambert C, Schleyer PvonR. Angew Chem. 1994;106:1187–1199; Angew Chem Int
Ed. 1994;33:1129–1140.
6. For a recent, improved method of determining molecular weights with deviations down to
9% through diffusion-ordered NMR spectroscopy (DOSY), see: Neufeld, R, Stalke D.
Chem Sci. 2015;6:3354–3364.
7. Knorr R, Menke T, Ferchland K, Mehlstäubl J, Stephenson DS. J Am Chem Soc.
2008;130:14179–14188.
8. Knorr R, Menke T, Behringer C, Ferchland K, Mehlstäubl J, Lattke E. Organometallics.
2013, 32, 4070–4081.
9. Knorr R, Hennig K-O, Böhrer P, Schubert B. J Organomet Chem. 2014;767:125–135.
10. Knorr R, Behringer C, Knittl M, von Roman U, Lattke E. J Am Chem Soc.
2017;139:4690–4703.
11. Knorr R, Behringer C, Lattke E, von Roman U, Knittl M. J Org Chem. 2015;80:6313–
6323.
12. Knorr R, Ruhdorfer J, Böhrer P. Organometallics. 2015;34:1038–1045.
13. Knorr R, Knittl M, Rossmann EC. Beilstein J Org Chem. 2014;10:2521–2530.
14. Knorr R, Behringer C, Ruhdorfer J, von Roman U, Lattke E, Böhrer P. Chem Eur J.
2017;23:12861–12869.
15. Knorr R, Lattke E, Ruhdorfer J, von Roman U, Firl J, Böhrer P. J Organomet Chem.
2016;824:61–72.
16. Knorr R, Lattke E. Tetrahedron Lett. 1977;18:3969–3972, line 7 of Table 1 therein.
17. Knorr, R.; Menke, T.; Ferchland, K. Organometallics 2013, 32, 468–472, Table 1
therein.
18. See the Supplementary Data.
19. Knorr R, Knittl M, Behringer C, Ruhdorfer J, Böhrer P. J Org Chem. 2017;82:2843–
2854.
20. Section 5.1 in the Supplementary Data.
21. Gurudata, Stothers JB, Talman JD. Can J Chem. 1967;45:731–737. Table 1 therein.
22. Knorr R, Behringer C, Nöth H, Schmidt M, Lattke E, Räpple E. Chem Ber/Recueil.
1997;130:585–592.
23. Wright JA, Gaunt MJ, Spencer JB. Chem Eur J. 2006;12:949–955, compound 1d
therein.
24

24. Reynolds WF, Hamer GK, Bassindale AR. J Chem Soc Perkin Trans 2. 1977:971–974,
ACCEPTED MANUSCRIPT
Table 1 therein.
25. Hamer GK, Peat IR, Reynolds WF. Can J Chem. 1973;51:897–914.
26. Reynolds WF, Peat IR, Hamer GK. Can J Chem. 1974;52:3415–3423.
27. For a short NMR desciptin of HMPA-solvated organolithiums, see Figure 1 in: Reich
HJ, Kulicke KJ. J Am Chem Soc. 1996;118:273–274, and literature quoted therein.
28. An illustration may be found in Ref. 12.
29. For this method and some relevant literature quotations, see Ref. 13.
30. a) Forsén S, Hoffman RA. J Chem Phys. 1963;39:2892–2901;
b) Anet FAL, Bourn AJR. J Am Chem Soc. 1967;89:760–768.
31. Figure 2 of Ref. 11.
32. Studies of some other (mostly unsettled) stereoinversion mechanisms under different
conditions may be found via the following publications: a) Ref. 22; b) Ref. 11 (Section
C and Ref. 19 therein); c) Ref. 8 (Refs. 1–3 and 26–29 therein).
33. For some literature quotations, see Ref. 19.
34. An impressive example of solvent-dependent (THF versus toluene), stereospecifically
divergent reactions of a lithiated aziridine was reported by: de Ceglia MC, Musio B,
Affortunato F, Moliterni A, Altomare A, Florio S, Luisi R. Chem Eur J. 2011;17:286–
296.
35. A superficially similar ortho-methyl effect prevented a lithiated azetidine from forming a
self-coupled (quasi-dimeric) intermediate and derivative; see: Parisi G, Capitanelli E,
Pierro A, Romanazzi G, Clarkson GJ, Degennaro L, Luisi R. Chem Commun.
2015;51:15588–15591.
36. Eaton C, Ghose BN, Walton DRM. J Organomet Chem. 1974;65:169–179, on p 176
therein.
37. von Roman U, Ruhdorfer J, Knorr R. Synthesis. 1993;1993:985–992, and quoted
literature.
38. Wu J, Yoshika N. Angew Chem Int Ed. 2016;55:336–340, compound 1c on p S2 in the
Supporting Information thereof.
39. Gassman PG, Harrington CF. J Org Chem. 1984;49:2258–2278, compound 24d on p
2268 therein.
25

ACCEPTED MANUSCRIPT
Scheme 1. α-Arylalkenyllithiums whose (non)aggregation and microsooation
states had been established in solution: 1a–e,7,8 2,8 3,9 4,10 5a,11 5b,12 5c,13
6a and b,10 7a,11 7b,13 8,14 9,14 and 10–12.15
Scheme 2. Preparation and derivatization of the α-arylvinyllithiums 14a–d.
Scheme 3. The rapidly reversible formation of tetrasolvated dimers withdraws
fractions of the trisolvated monomers of 14a, b,or d from their
pseudomonomolecular, ionic cis/trans stereoinversion process in THF as the solvent.
Scheme 4. Preparation and Z,E-stereoinversion equilibrium of [β-D1]14d.
1
Fig. 1. Lithiation H (in parentheses) and 13C NMR shifts ∆δ = δ(RLi) – δ(RH) of 14a–d.
1
Fig. 2. H NMR spectra (400 MHz, 25 °C) of dimeric α-lithiostyrene (14d) in rapidly
equilibrating mixtures with its congeners: a) averaged with higher 14d aggregates (ca. 70%)
in t-BuOMe as the solvent; b) averaged with monomeric 14d (ca. 20%) in THF; x =
benzene.
Fig. 3. Temperature-dependent decadic logarithms (10log) of the pseudo-first-order
stereoinversion rate constants kψ (s–1) of the α-lithiostyrenes 14a, 14b, and 14d in THF
(Tables S9–S1218): (a) 14d; (b) 14b; (c) 14a; (d) 14d with HMPA (2 equiv).
Table 1. Microsolvation numbers d, NMR coupling constants 2JH,H [Hz] of CH2-β, and
lithiation shifts ∆δ = δ(RLi) – δ(RH) of the α-arylvinyllithiums 5a, 12, and 14a–d in four
solvents.
–1
–1
‡
‡
Table 2. Pseudoactivation parameters ∆Gψ (kcal mol at 0 °C), ∆Hψ (kcal mol ), and
–1 –1
‡
∆Sψ (cal mol K ) for the cis/trans diastereotopomerization rates of α-arylvinyllithiums 5a,
9, 12, 14a, 14b, and 14d in THF.
* * *
This research did not receive any specific grants from funding agencies in the public,
commercial, or not-for-profit sectors.
26