Is carbon dioxide able to activate halogen/lithium exchange?

Article abstract of DOI:10.1002/ejoc.201402320

The unexpected effect of carbon dioxide on halogen-lithium exchange (HLE) reactions of selected haloarenes with tBuLi was investigated. In an aliphatic hydrocarbon solvent (pentane), the HLE does not occur at ca. -70 C but, surprisingly, pouring the mixture of reactants onto dry ice and subsequent aqueous acidic hydrolysis gave carboxylic acids resulting from the quench of the first-formed aryllithiums with carbon dioxide. This suggests that CO ₂ acts as a promoter of the HLE and, subsequently, serves as an electrophile to trap the aryllithium intermediates that are generated in situ. Theoretical DFT calculations were used to develop a plausible mechanism for the reaction, which indicates that CO₂ is a much weaker donor than tetrahydrofuran (THF) so the cleavage of inert tBuLi cubic tetramers into more reactive solvated dimeric species (tBuLi)₂(CO₂) ₄ is disfavored by 42.8 kJ per mol of (tBuLi)₄. It is possible that this deaggregation process occurs to some extent when a large excess of CO₂ is used. Copyright

Full text of DOI:10.1002/ejoc.201402320

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
FULL PAPER
DOI: 10.1002/ejoc.201402320
Is Carbon Dioxide Able to Activate Halogen/Lithium Exchange?
[a]
[a]
[a]
[a]
Krzysztof Durka, Sergiusz Luli n´ ski,* Marek D a˛ browski, and Janusz Serwatowski
Keywords: Reaction mechanisms / Carbon dioxide / Carboxylation / Lithiation / Solvent effects
The unexpected effect of carbon dioxide on halogen–lithium
exchange (HLE) reactions of selected haloarenes with tBuLi
was investigated. In an aliphatic hydrocarbon solvent (pent-
ane), the HLE does not occur at ca. –70 °C but, surprisingly,
pouring the mixture of reactants onto dry ice and subsequent
aqueous acidic hydrolysis gave carboxylic acids resulting
from the quench of the first-formed aryllithiums with carbon
lithium intermediates that are generated in situ. Theoretical
DFT calculations were used to develop a plausible mecha-
nism for the reaction, which indicates that CO is a much
2
weaker donor than tetrahydrofuran (THF) so the cleavage of
inert tBuLi cubic tetramers into more reactive solvated di-
meric species (tBuLi)
of (tBuLi) . It is possible that this deaggregation process oc-
curs to some extent when a large excess of CO is used.
2 2 4
(CO ) is disfavored by 42.8 kJ per mol
4
dioxide. This suggests that CO
2
acts as a promoter of the HLE
2
and, subsequently, serves as an electrophile to trap the aryl-
Introduction
presence of a large excess of a classical reactive electrophile,
namely carbon dioxide. Moreover, we have paid special at-
Solvation effects play a fundamental role in the chemistry
of organolithiums because they have a strong impact on the
structural behavior of these compounds, which, in turn,
often leads to dramatic changes in their reactivity. It is well-
known that halogen–lithium exchange (HLE) is often very
rapid and can occur even at diffusion controlled rates, pro-
vided that deaggregation of alkyllithium oligomers is
tention to the generation of 1,2-dilithiobenzene 5-Li from
2
1,2-diiodobenzene because this would allow “green”, mer-
cury-free access to this useful reagent.
Results and Discussion
achieved through the use of donor solvents (typically Et O
In a typical experiment, a solution of a selected haloar-
2
or THF) or additives (e.g., TMEDA or tBuOK).[
1–5]
Thus, ene was added to a ca. 0.8 m solution of tBuLi in pentane
it was concluded that the rate-limiting step of HLE is the at –70 °C (Table 1). It is well-known that 2 equiv. of tBuLi
initial dissociation of oligomeric alkyllithium cages [e.g, per halogen atom are required to trap the byproduct tBuX
[
20–22]
(
tBuLi) or (nBuLi) ] that are stable in noncoordinating hy- (X = Br, I).
Thus, the reaction was performed by using
The profound effect of ethereal sol- the molar ratio of haloarene/tBuLi = 1:2.1 (or 1:4.2 for di-
4
6
[
6–11]
drocarbon solvents.
vents on the HLE kinetics is very well established
[
12]
al- iodobenzenes 5 and 6). There was no significant (if any)
though it should be stressed that in the case of more reac- exothermic effect observed during these experiments, which
tive substrates the reaction also occurs rapidly in aliphatic indicates that HLE does not proceed under these condi-
or aromatic hydrocarbons even at low temperatures. How- tions. For instance, mixing tBuLi with bromopentafluoro-
ever, it is important to note that these reactions were per- benzene (10) gave a clear solution, although it is known
formed using nBuLi as the lithiating reagent.[
13–15]
The dif- that this arene reacts readily with nBuLi in hexane at –80 °C
ferent reactivity of the two most popular alkyllithiums, to give a suspension of white insoluble pentafluorophenyl-
[
13]
tBuLi and nBuLi, in some reactions is remarkable. For in- lithium 10-Li.
We checked this experiment using 1 m
stance, 1,3- as well as 1,4-dilithiobenzene are readily access- nBuLi at –70 °C and indeed observed an exothermic reac-
ible from reactions of the corresponding dibromobenzenes tion accompanied by immediate precipitation of 10-Li,
[
16–18]
with tBuLi (Hal = Br),
whereas nBuLi works well only which decomposed vigorously in the reaction flask during
with more reactive diiodobenzenes.[ In this paper, we de- the addition of 10, in accordance with reported cautionary
scribe the results of our studies on the unexpected occur- notes. This indicates that nBuLi is much more reactive than
rence of HLE between tBuLi and selected haloarenes in the tBuLi towards haloarenes in aliphatic hydrocarbons. The
lack of reactivity of tBuLi is apparently due to the relative
19]
[
a] Warsaw University of Technology, Faculty of Chemistry,
6
inertness of tBuLi tetramers. Based on analysis of the Li–
Department of Physical Chemistry,
13
13
Noakowskiego 3, 00-664 Warsaw, Poland
E-mail: serek@ch.pw.edu.pl
C couplings in the C NMR spectra of (tBuLi) , it was
4
concluded that this compound indeed exhibits a static char-
acter below –22 °C, whereas it becomes fluxional above
http://www.lmt.ch.pw.edu.pl
Supporting information for this article is available on the
WWW under http://dx.doi.org/10.1002/ejoc.201402320.
[
23]
–5 °C. On the other hand, (nBuLi) exhibits dynamic be-
6
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
K. Durka, S. Luli n´ ski, M. D a˛ browski, J. Serwatowski
FULL PAPER
havior already at low temperatures, which would account was isolated in high yield (over 90%); in this case we also
for the higher reactivity of this species in noncoordinating confirmed that the occurrence of HLE at low temperatures
[
24]
solvents.
Pouring the mixture of haloarene and tBuLi can be amply excluded because we observed that the
onto a large excess of dry ice covered with pentane resulted formation of an insoluble white precipitate of 2-thienyl-
in an immediate reaction, which was manifested by exten- lithium 4-Li (accompanied by a distinct exothermic effect)
sive foaming and formation of a white suspension. Careful started only at ca. 10 °C. The increased reactivity of thio-
hydrolysis with aqueous H SO gave the corresponding car- phene derivatives is clear not only by comparing the behav-
2
4
boxylic acid (or a mixture of mono- and dicarboxylic acids) ior of 1 vs. 4 but also the behavior of the corresponding
together with the unreacted starting haloarene and pivalic aryl bromides. Thus, bromobenzene (7) and both dibromo-
acid byproduct resulting from the quench of unreacted benzenes 8 and 9 were found to be completely unreactive,
tBuLi (Table 1). In the case of iodobenzene 1, poor conver- which reflects the generally higher susceptibility of aryl iod-
sion into benzoic acid (ca. 20%) was observed and the ob- ides vs. bromides to HLE. In contrast, 2-bromothiophene
tained material consisted mainly of unreacted PhI and piv- (11) underwent HLE to some extent to give thiophene-2-
alic acid. Screening of starting haloarenes led to the conclu- carboxylic acid in low yield (19%), whereas 2,5-dibromo-
sion that significant activation of the aromatic ring is re- thiophene (12) proved to be more reactive because it was
quired to observe the occurrence of HLE in the presence of converted into a mixture of 5-bromothiophene-2-carboxylic
CO . This can be achieved, for instance, by substitution of (34%) and thiophene-2,5-dicarboxylic acids (5%). Finally,
2
1
with the electron-withdrawing 2-fluoro-substituent (as in the very strong activation of bromine in perfluorinated sub-
substrate 2). It is remarkable that 3-fluoroiodobenzene 3 strate 10 resulted in good yield of perfluorobenzoic acid
was much less reactive than 2, which points to the lack of (10a).
a long-range activating effect of the 3-fluoro-substituent.
For 1,2-diiodobenzene (5), we performed HLE using
For 2-iodothiophene (4), thiophene-2-carboxylic acid (4a) either pentane or, alternatively, the mixed aliphatic/aromatic
solvent mixture of pentane and toluene (2:1). The addition
of a pentane solution of 5 to a 1 m solution of tBuLi
(4.2 equiv.) at –70 °C was accompanied by a weak exother-
Table 1. Screening of HLE activation with CO
tBuLi/pentane system at –70 °C.
2
in the haloarene/ mic effect and resulted in the formation of a white precipi-
tate (Scheme 1). Thus, we initially suspected that HLE in
1,2-diiodobenzene may already occur at low temperature.
The carboxylation of the obtained suspension followed by
acidic hydrolysis gave mainly phthalic acid (5b; 63%) to-
gether with a small amount of 2-iodobenzoic acid (5a; ca.
4
%). To check whether HLE indeed proceeds in this case
Starting material
PhI
Product(s)
PhCOOH
2-F-PhCOOH
3-F-PhCOOH
2-ThCOOH
2-I-PhCOOH
phthalic acid
3-I-PhCOOH
isophthalic acid
none
Yield
%)
(
even at –70 °C, the obtained precipitate was separated by
decantation and washing with cold pentane, then treated
with (MeS) and Et O. GC/MS analysis of the crude reac-
1
1a
1a
3a
4a
5a
5b
6a
6b
20
78
17
91
4
63
53
20
0
0
0
86
19
34
5
1
1
2
1
-Fluoro-2-iodobenzene 2
2
2
-Fluoro-3-iodobenzene 3
tion mixture revealed mainly unreacted 5 and 1-iodo-2-
methylthio)benzene as a minor component. This indicates
clearly that a simple precipitation of 5 from the reaction
-Iodothiophene
4
5
(
,2-Diiodobenzene[
a]
,3-Diiodobenzene[a]
solution occurs. Similarly, when B(OMe) was added to the
mixture of 5 and tBuLi, no significant changes (e.g., no
clear temperature effect) occurred in the system. However,
the addition of a few drops of Et O resulted in a strongly
exothermic reaction that eventually gave 2-iodophenyl-
boronic acid in low yield (ca. 12%). This result shows that
in the absence of Et O, tBuLi does not undergo HLE and
1
6
3
PhBr
7
1
1
,2-Dibromobenzene
,3-Dibromobenzene
8
none
2
9
none
C
6
F
5
Br
10
11
C
6
F
5
COOH
10a
4a
2
2
-Bromothiophene
,5-Dibromothiophene 12
2-ThCOOH
5-Br-2-ThCOOH 12a
,5-Th(COOH) 12b
2
2
2
is also essentially not reactive towards B(OMe) at –70 °C.
3
[a] 4.2 equiv. tBuLi per 1 equiv. haloarene.
After addition of Et O, HLE starts but the initial product
2
2
Scheme 1. Synthesis of phthalic acid via in situ CO -induced HLE/carboxylation of 1,2-diiodobenzene.
2
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
Is Carbon Dioxide Able to Activate Halogen/Lithium Exchange?
5
-Li is very unstable so that it was not effectively trapped ing any of the steps. Bubbling of dry gaseous CO through
2
with the boron electrophile prior to benzyne-mediated de- the reaction solution was then started. However, no imme-
gradation. It should be noted that no significant symptoms diate acceleration of HLE and subsequent reactions (e.g,
of HLE and subsequent reactions were observed in the 5/ manifested by a temperature jump) was observed. The mix-
tBuLi/pentane system below 0 °C, which is consistent with ture was slowly warmed to room temperature in the stream
the aforementioned behavior of 4 towards tBuLi.
of CO . A distinct exothermic effect accompanied by the
2
For comparison, we have also investigated the reactivity extensive formation of insoluble products was observed
of 1,3-diiodobenzene (6) under standard conditions only when the temperature reached ca. 10–15 °C. This sug-
(4.2 equiv. tBuLi, –70 °C, pentane, then reaction with CO2). gests that, in this case, there was no expected activation of
Again, a mixture of two possible products of HLE/carbox- HLE with CO . There was also no apparent selectivity with
2
ylation was obtained. However, in this case, 3-iodobenzoic respect to competing electrophiles because the hydrolysis
acid (6a) was obtained as the major product (53%), whereas and workup gave the mixture of 4a and the corresponding
isophthalic acid (6b), resulting from a double HLE, was boronic acid 4b in an approximate ratio of 2:3. To summa-
formed in lower yield (20%).
rize this experiment, it seems that the activating effect of
The formation of benzoic acids can be explained by as- CO is indeed rather weak and operates only under stan-
2
suming a dual role of CO . Firstly, it should activate tBuLi dard carboxylation conditions when a large excess of CO2
2
tetramers towards HLE. This activation seems to be quite is used. It should be stressed that, under these conditions,
effective for the highly reactive haloarenes, as one could ex- the temperature of a mixture during the quench was essen-
pect that the competitive reaction of tBuLi with CO should tially constant, which means that both reaction steps (i.e.,
2
be preferred due to the presence of a large excess of the HLE and carboxylation) occur at low temperature (ca.
latter reagent. Thus, this points to the effective coexistence –70 °C). However, it is unclear how this can be rationalized
of tBuLi and CO on the microsecond timescale. In other in terms of a specific mechanism. A simple explanation
2
words, in the case of the most reactive substrates such as 2, would be that CO does not effectively split the unreactive
2
4
–6, and 10, the rate of CO -promoted HLE must be higher (tBuLi) cage due to its weak coordinating properties, so
2
4
than the rate of the formation of lithium pivalate. Secondly, the large concentration of CO is necessary to generate de-
2
CO serves as the in situ trapping reagent for the arylli- aggregated and thus more reactive tBuLi species.
2
thium species generated in HLE steps. In the case of diiodo-
Nevertheless, alternative explanations should also be
benzenes, the formation of iodobenzoic acids (products 5a considered. The first involves low levels of halogen–lithium
and 6a) suggests that both reactions proceed in a stepwise exchange already occurring at low temperature prior to car-
manner, that is, iodophenyllithium intermediates 5-Li and boxylation. The resulting aryllithium reacts rapidly with
6
-Li are first intercepted with CO to give lithium iodoben- CO in the next step to give ArCOOLi, which promotes
2
2
zoates, which are then lithiated and carboxylated to provide HLE of the remaining material, for example due to forma-
phthalate and isophthalate dianions, respectively. The much tion of more reactive mixed aggregates with tBuLi. It is also
higher yield of 5b compared with 6b could be due to a sig- plausible that HLE could be catalyzed by lithium pivalate
nificant activation provided by the ortho-COOLi group, formed in the competitive reaction of tBuLi with CO . The
2
which facilitates the HLE involving the adjacent iodo sub- experiment involving the addition of gaseous CO to the
2
stituent. This can be supported to some extent by a report mixture of 4 and tBuLi at low temperature should generate
concerning the activation of dibenzodioxin bearing the 1- at least a small amount of lithium carboxylates, which
carboxylate moiety towards the deprotonative 9-lithiation would, in turn, significantly increase the rate of HLE. How-
[
25]
with tBuLi. Similarly, the activating effect of the N-carb- ever, it should be stressed again that under these conditions
oxylate group formed in situ was used for regiospecific C- fast HLE was observed only above 10 °C. Finally, a combi-
lithiation of nitrogen-containing heterocycles.[
26–29]
It was nation of activating effects of CO and carboxylate-based
2
also shown that CO2 and related electrophiles such as catalysis cannot be excluded. For instance, CO would initi-
2
RCONMe (R = H, Ph, 4-MeOC H ) and RCO Li (R = ate the HLE reaction but, after generation of a small
2
6
4
2
Ph and Me) are able to assist a second lithiation of pheno- amount of ArCOOLi, the mechanism would change due to
[
30]
–
thiazine.
the stronger activating effect of RCOO with respect to
Finally, it should be noted that in our experiments we CO . To check whether HLE can be catalyzed by lithium
2
have used commercial dry ice containing impurities that are carboxylates, a suspension of lithium thiophene-2-carboxyl-
potentially able to activate HLE. Therefore, we repeated the ate 4a-Li and lithium pivalate in pentane [prepared by the
HLE in 4 by using resublimed dried CO . The result was reaction of 4 with tBuLi (2 equiv.) followed by carboxyl-
2
similar to that obtained using standard dry ice, which con- ation of resulting 4-Li and unreacted excess of tBuLi] was
firms that CO is indeed responsible for the activation of added to a stirred solution of 4 and tBuLi in pentane at
2
HLE. Another experiment was performed to check whether –70 °C. The ratio of 4a-Li vs. 4 was ca. 5 mol-%. However,
the HLE activation can be observed using only a small we did not observed any clear evidence of HLE such as a
amount of CO . Thus, the solution of 4 and tBuLi in pent- significant temperature jump upon addition of the sus-
2
ane at –70 °C was first prepared and this was followed by pected catalyst followed by precipitation of insoluble 4-Li.
a dropwise addition of electrophile B(OiPr) . Under these The resulting mixture was hydrolyzed by pouring it onto
3
conditions, there was no clear evidence of any reaction dur- dilute aqueous H SO . GC–MS analysis of the organic
2
4
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
3

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
K. Durka, S. Luli n´ ski, M. D a˛ browski, J. Serwatowski
FULL PAPER
phase revealed that 4 was recovered, however, it was con- the temperature. Interestingly, theoretical studies point to a
taminated by ca. 30% thiophene. The latter compound is significant thermodynamic stabilization of nonsolvated 5-
formed by protonolysis of 4-Li so its presence is a clear Li with respect to the meta and para isomers due to a fa-
2
indicator of the occurrence of HLE (tBuI was also found vorable structure with two lithium atoms bridging two car-
in the reaction mixture). However, a similar experiment per- bon atoms, which gives enough compensation for repulsive
formed without the use of 4a-Li additive gave a comparable interactions between adjacent carbanion lone pairs.[
38–40]
It
result, which means that, in fact, there is no activation of is clear, however, that a completely different situation must
HLE by the used carboxylates. We suppose that, in this be expected in ethereal solvents coordinated to lithium
case, HLE starts during the aqueous quench (performed at atoms, which leads, in turn, to weakening of Li–C interac-
room temperature), and is so rapid that it competes to some tions, whereas the ortho-repulsions of lone-pairs remain
extent with hydrolysis of tBuLi. It should be noted that hy- strong. In contrast, it was claimed that 1,2-dilithio-3,4,5,6-
drolysis is disfavored because the reagents are in different tetrafluorobenzene can be easily generated from the respec-
tive dibromo-derivative by using nBuLi at ca. –70 °C.[
This would reflect well the potent activating effect of mul-
tiple fluorine atoms on the HLE kinetics and the relative
thermodynamic stability of the resulting aryllithium, al-
41]
phases.
Formation of 1,2-Dilithiobenzene
The need for easy access to 1,2-dilithiobenzene 5-Li2 though even in this case there is some controversy over
prompted us to study the possibility of performing a double whether dilithiation occurs to a significant extent prior to
[
42]
HLE in 1,2-diiodobenzene (5). The synthesis of this valu- electrophilic quench.
It is also remarkable that the syn-
able reagent has been reported as early as 1957 in the pion- thesis of hexalithiobenzene by treatment of C Cl with a
6
6
eering work by Wittig. It involved the reductive cleavage of large excess of tBuLi in pentane/1,4-dioxane at –120 °C has
,2-phenylenemercury trimer with lithium metal in ether to been reported, whereas only 1,4-dilithiation was effected in
1
[
31]
[43]
give a deep-red solution of the desired reagent 5-Li2. This the absence of the latter solvent. Based on the successful
approach is in fact the only method available for the genera- formation of 1,3- as well as 1,4-dilithiobenzene from reac-
[
32,33]
[19]
tion of 5-Li2
because the classical approach by the di- tions of appropriate diiodobenzenes with nBuLi in Et O,
2
rect double halogen–lithium exchange in 1,2-dibromobenz- we have checked this approach. However, the conditions
ene proved unsuccessful. For instance, it was demonstrated were modified to reduce the risk of rapid decomposition of
that 5-Li2 is not an intermediate en route from 1,2-di- very unstable 5-Li through the benzyne pathway. Thus, the
bromobenzene to 1,2-bis(trimethylsilyl)benzene because reaction was carried out at the very low temperature of
only one bromine atom is directly replaced in the substrate –110 °C and the reverse addition mode was applied, that is,
with lithium using an excess of tBuLi and the electrophile 5 was added to the solution of nBuLi in the Et O/THF
2
Me SiOTf under the reaction conditions employed (THF, mixed solvent. THF was used to accelerate the second I/Li
3
–110 °C, in situ quench), whereas the second lithiation (and exchange. Unfortunately, the carboxylation and hydrolysis
subsequent silylation) occurs only after conversion of the did not afford the expected product 5b. In contrast, 9-fluor-
intermediate 2-bromophenyllithium into 1-bromo-2-(tri- enone 13 was formed to a significant extent, as evidenced
methylsilyl)benzene.[ This is apparently due to the strong by GC–MS analysis. This indicates that the reaction occurs
kinetic deactivation of the second bromine at the ortho posi- through the benzyne pathway as shown in Scheme 2. The
tion to the carbanion center towards HLE. Thus, the lithi- first I/Li exchange in 5 occurs effectively but the I atom in
34]
ation stops at the stage of 2-bromophenyllithium, which is 5-Li is not sufficiently prone to HLE using nBuLi. Thus,
stable below –110 °C in THF[
35,36]
and decomposes rapidly competitive decomposition of 5-Li occurs, finally resulting
[
44]
at slightly higher temperatures through an aryne mecha- in 13 as the carboxylation product.
nism to give biphenyl-type products.[ Hence, it is imposs-
37]
Since nBuLi proved ineffective, we again turned our at-
ible to attempt the second HLE in this reagent by increasing tention to tBuLi. In the initial experiment, the reaction con-
Scheme 2. Reactions of 1,2-diiodobenzene with nBuLi and tBuLi in ethereal solvents.
4
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
Is Carbon Dioxide Able to Activate Halogen/Lithium Exchange?
Scheme 3. Synthesis of phthalic acid through I/Mg and subsequent I/Li exchange in 1,2-diiodobenzene.
ditions were not changed with respect to the reaction de- Theoretical Calculations
scribed above. The resultant mixture was carboxylated and
We have performed theoretical calculations to gain a
deeper insight into the thermodynamics and kinetics of HLE
hydrolyzed to give benzoic acid as the sole aromatic prod-
uct, contaminated with an approximately equimolar
amount of pivalic acid resulting from the quench of uncon-
sumed tBuLi (Scheme 2). Such a result indicates that 5-Li2
is effectively generated but it immediately undergoes mono-
protonation in situ to give phenyllithium. We suppose that
performed with tBuLi in the presence of CO . The results
2
were compared to those obtained by using the tBuLi/THF
system, which can be regarded as the standard for HLE
reactions. To simplify the calculations, the HLE reactions
of monohalogenated benzenes PhX (X = Br, I) were ana-
lyzed. The computations were performed with the DFT
the high kinetic basicity of 5-Li in THF enables rapid pro-
2
ton abstraction from the byproduct tBuI producing isobut-
ene and lithium iodide. Alternatively, protonation of 5-Li2
by THF could also be considered. However, tBuI (as well
as tBuBr) seems to be a much more effective proton source
than THF considering that it already reacts with tBuLi at
low temperatures in the presence of THF as solvent. In the
[
48,49]
[50]
[51–53]
method [B3LYP
/6-31+G(d) +LANL2DZ
] em-
THF
[
54–56]
ploying Tomasi’s polarized continuum model.
and n-hexane were used as solvent for the reaction per-
formed in the presence of THF and CO , respectively. Ini-
2
tially, the stabilities of tBuLi species [(tBuLi) , (tBuLi) ,
4
2
tBuLi] and their respective solvated forms [(tBuLi) L ],
4
4
next approach, THF was replaced with the less polar Et O.
2
(
tBuLi) L , tBuLiL , (L = THF or CO ) were compared.
2 4 3 2
The lithiation was initially performed at –105 °C but the
temperature increased to –85 °C during a few minutes when
the majority of 5 was added. The resulting orange slurry
was carboxylated and hydrolyzed with aq. H SO to give
The appropriate thermodynamic cycles are shown in Fig-
ure 1. It is evident from the calculations that complexation
of tBuLi dimers and monomers with THF molecules is en-
ergetically very favorable. This effect is especially noticeable
2
4
5b in moderate yield (ca. 30%). Other products formed in
for tBuLi(THF) , which was found to be the most stable
3
substantial amounts and these were identified by GC/MS
analysis as 13 and 2-tert-butylbenzoic acid 14 (Scheme 2).
These results point to the significant contribution from re-
actions occurring through the benzyne mechanism. It seems
that 1,2-dithiobenzene is formed quite effectively at the ini-
tial stage of the reaction. However, gradually the concentra-
tion of tBuLi decreases and it appears that the competitive
species in THF solution. A different situation was observed
in hexane solution. It is well-known that in nonpolar hydro-
carbon solvents tBuLi exists as the cubic tetramer.[
This is also reflected by the calculations, which show that
16–18]
reaction of the added 5 with 5-Li producing highly un-
2
stable 5-Li becomes important. When the temperature was
rigorously kept below –110 °C during the HLE, the selectiv-
ity was not improved in favor of 5-Li . In contrast, 5b was
2
not isolated at all after carboxylation. We suppose that, un-
der these conditions, the second HLE (i.e., conversion of 5-
Li into 5-Li ) is kinetically hampered and thus the reaction
2
course becomes dominated by benzyne-mediated decompo-
sition of 5-Li.
We also attempted to prepare the partially magnesiated
analogue of 5-Li . In the first step, 2-iodophenyl Grignard
2
reagent 5-(MgCl) was generated through I/Mg exchange of
with iPrMgCl·LiCl in THF at –70 °C.[
45–47]
Then 5-
5
(MgCl) was treated with nBuLi to perform HLE. However,
this approach failed completely, and pure 5a was isolated
after addition of CO with good yield (68%). The use of
2
tBuLi gave a thick, red-orange slurry, which was also
quenched by carboxylation. In this case, 5b was formed as
the major product (55%) together with a small amount
Figure 1. (a) General scheme for the different forms of tBuLi.
(
b,c) Thermodynamic cycles compare stabilities of different tBuLi
species and their solvates with (b) THF (THF solvent field) and
c) CO (hexane solvent field) molecules. All energy value are given
in kJmol .
(8%) of 5a (Scheme 3). Thus, it can again be concluded that
nBuLi is less reactive than tBuLi because it is unable to
effect HLE in 5-Li and 5-(MgCl).
(
2
–
1
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
5

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
K. Durka, S. Luli n´ ski, M. D a˛ browski, J. Serwatowski
FULL PAPER
in hexane (tBuLi) is more stable than (tBuLi) and mono-
4
2
–
1
meric tBuLi by –19.2 and –73.9 kJmol (per one mono-
meric tBuLi unit), respectively. From the comparison be-
tween tBuLi(THF) and tBuLi(CO ) , it is evident that the
3
2 3
coordination of CO to tBuLi monomer is less exothermic
2
than coordination of THF. This may reflect a significantly
2
weaker donor ability of the sp -hybridized O atom in CO
2
3
vs. the sp -hybridized O atom in THF. Indeed, calculations
indicate that deaggregation of (tBuLi) in hexane with CO ,
4
2
resulting in solvated dimmers, is thermodynamically disfa-
vored by 42.8 kJ per mol. Nevertheless, CO ligands slightly
2
stabilize nonsolvated tBuLi dimers [by –17 kJ per mol of
(
(
tBuLi) ], and in the presence of a large excess of CO , the
2
2
tBuLi) (CO ) species should appear in the reaction mix-
2
2 4
ture to a sufficient extent, thus facilitating the HLE reaction
at low temperatures.
We next turned our attention to the HLE reaction
mechanism. At first, we analyzed the reactions with mono-
meric tBuLi(THF) and tBuLi(CO ) species as lithiating
3
2 3
agents. The reactions proceed through polar substitution of
the nucleophile at the halogen atom.[
57–62]
All obtained transition state structures possess a nearly
linear C(Ph)–X–C(tBu) moiety with elongated C–X bonds.
The calculations show that the activation energy for the
HLE of bromobenzene with tBuLi(THF) in the THF sol-
3
–
1
vent field is 22.0 kJmol (Figure 2); thus, it is much lower
than that obtained for the analogous reaction with nBuLi
–
1 [63]
(
36.8 kJmol ).
benzene is much faster and the corresponding ate complex
[Ph–I–tBu]Li(THF) } could even be considered to be a rel-
As expected, the I/Li exchange in iodo-
{
3
Figure 2. Reaction pathways for the lithiation of iodobenzene with
atively stable intermediate (energy barrier is only (a) tBuLi(THF) , (b) tBuLi(CO ) , and (c) (tBuLi) (CO ) together
3
2 3
2
2 4
–
1
2
(
.6 kJmol ). It should be noted that the related stable bis- with the geometries of transition-state structures. Hydrogen atoms
are omitted for clarity.
perfluorophenyl) complex [(C F ) I][Li(TMEDA) ] was
6
5 2
2
[64]
isolated and characterized by X-ray diffraction.
Finally,
the lithiation with tBuLi(CO ) gave an activation energy
2
3
–
1
value close to 30 kJmol . Due to the weaker donor ability
+
of CO vs. THF, the Li(CO ) cation is more strongly co-
2
2 3
ordinated to the central I atom and thus decreases the models in Figure S1 of the Supporting Information), which
thermodynamic stability of ate anion [d = 2.61 Å implies that they should be less prone to collapse resulting
I···Li
+
+
Li(CO ) , d
= 2.84 Å for Li(THF) ]. A similar effect in lithium pivalate. The calculation performed for the lithi-
2
3
I···Li
3
was computed for the [CH –I–CH ]Li complex, for which ation of PhI with (tBuLi) (CO ) gave a relatively high
3
3
2
2 4
+
–1
solvation of Li with water molecules resulted in a slight (65.8 kJmol ) activation barrier (Figure 2, c) and indicate
lengthening of the Li–I distance, thus conferring consider- that the I/Li exchange in PhI is hampered. On the other
able stability to this species.[
65]
hand, the electron-withdrawing substituents (especially lo-
The relatively low activation energy value suggests that cated in the ortho position to the iodine atom) stabilize the
the I/Li exchange reaction can proceed with tBuLi(CO ) in intermediate ate complex. The calculation shows that the
2
3
hexane, to give a highly unstable PhLi(CO ) intermediate, HLE reactions of pentafluoroiodobenzene (C F I) with
2
3
6 5
which rapidly undergoes rearrangement to lithium benzo- either tBuLi(CO ) or (tBuLi) (CO ) as lithiating agents
2
3
2
2 4
ate. Clearly, tBuLi(CO ) also easily rearranges to give lith- lead to stable intermediate complexes [C F –I–tBu]Li-
2
3
6 5
ium pivalate. The calculations show that activation barriers (CO ) or [C F –I–tBu][tBuLi (CO ) ]. The energy of such
2
3
6
5
2
2 4
for both processes are indeed very low, being only 8.3 and species is lower than the energy of the corresponding sub-
–
1
–1
1
(
4.5 kJmol for PhLi(CO ) and tBuLi(CO ) , respectively strates by –82.8 kJmol [lithiation with BuLi(CO ) ] and
2 3 2 3 2 3
Figure S3, Supporting Information). However, analysis of –44.4 kJmol–1 [(tBuLi) (CO ) ] (Figure S3, Supporting In-
2 2 4
the thermodynamic cycle (Figure 1, c) indicates that CO2 formation). This is consistent with our experimental obser-
can induce at best the formation of solvated tetrameric or vations, because iodobenzene 1 did not prove to be very
dimeric tBuLi species. These compounds should be kinet- reactive, while 2-fluoro-derivative 2 underwent in situ CO2-
ically more stable than the monomer tBuLi(CO ) due to induced I/Li exchange and subsequent carboxylation in
2
3
the better steric shielding of C–Li bonds (see space-filling good yield.
6
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
Is Carbon Dioxide Able to Activate Halogen/Lithium Exchange?
at –70 °C during ca. 10 min. The resulting solution was stirred for
Conclusions
1
3
5 min, followed by rapid pouring onto freshly crushed dry ice (ca.
0 g) covered with pentane. The resulting mixture was warmed to
We have demonstrated that, in some cases (i.e., for iodo-
and activated bromoarenes), the HLE reaction involving room temperature and hydrolyzed with 2 m aq. H SO (15 mL).
2
4
tBuLi can proceed in aliphatic hydrocarbon solvent under Et
2
O (20 mL) was added and the organic layer was separated. The
aqueous phase was extracted with Et O (2ϫ 10 mL) and solvents
2
the standard conditions used for carboxylation of organo-
lithium compounds. This is interesting and, in fact, contro-
were evaporated to give the crude product, which was washed with
water, hexane (2ϫ 5 mL) and dried in air, yield 1.09 g (78%); m.p.
versial because CO is generally recognized as an electro-
2
23–125 °C (ref.[
66]
m.p. 125–126 °C).
1
H NMR (400 MHz,
1
phile that reacts rapidly and irreversibly with organo-
lithiums, which implies that it is not suitable for the acti-
vation of the initial HLE step. However, for some haloar-
enes the occurrence of HLE is observed even though one
DMSO): δ = 13.1 (br., 1 H, COOH), 7.91 (m, 1 H, Ph), 7.67 (m, 1
H, Ph), 7.34 (m, 2 H, Ph) ppm.
2
Reaction of 1,2-Diiodobenzene (5) with tBuLi in Et O/Pentane; For-
mation of 1,2-Dilithiobenzene: A solution of 5 (9.9 g, 30 mmol) in
hexane (30 mL) was added dropwise to a stirred solution of tBuLi
could expect that the use of CO in large excess with respect
2
to tBuLi should strongly favor competitive formation of
(
1.7 m in pentane, 75.0 mL, 128 mmol tBuLi) diluted with Et
2
O
lithium pivalate. Hence, CO seems to cleave effectively in-
2
(
70 mL) at –100 °C during ca. 30 min. The temperature increased
ert tBuLi tetramers present in a nonpolar hydrocarbon sol-
to –85 °C when the addition was almost complete. The resulting
orange slurry containing 5-Li was stirred for 15 min at –90 °C then
vent into more reactive dimers, presumably by coordination
2
of the O atom to Li. Then HLE may occur rapidly followed poured onto freshly crushed dry ice covered with hexane. The yel-
by attack of the resulting aryllithium on the electrophilic lowish mixture was warmed to room temperature and hydrolyzed
with 2 m aq. H
2 4
SO (50 mL). The resultant suspension of phthalic
carbon atom in CO . However, the proposed mechanism
2
acid 5b was filtered, yield 1.70 g (31%); m.p. 202–205 °C (dec.)
of the HLE activation by CO should be treated with care
2
ref.[ m.p. 208–209 °C (dec.)]. H NMR (400 MHz, [D
67]
1
]acetone):
[
6
because, according to DFT calculations, the cleavage of
tBuLi tetramers into more reactive dimeric species solvated
δ = 7.78 (m, 2 H, Ph), 7.62 (m, 2 H, Ph) ppm. The filtrate was
concentrated to leave a dark viscous oil containing mainly fluor-
enone, 2-(tert-butyl)benzoic acid, biphenyl, and pivalic acid (prod-
ucts determined by GC/MS analysis).
with CO molecules is thermodynamically disfavored by
2
–
1
roughly 40 kJmol . This thermodynamic barrier could be
overcome by the use of a large excess of CO during the
2
Theoretical Calculations: All geometry optimizations and frequency
reaction. However, it cannot be excluded that the mecha-
nism of the observed acceleration of HLE in the presence
calculations were carried out with the GAUSSIAN 09 suite of pro-
grams[
68]
and Becke-style three-parameter density functional
of CO is quite different and, perhaps, more sophisticated.
2
method using the Lee–Yang–Parr correlation functional (B3LYP)
The calculations also show that the energy barrier for I/
[48,49]
[50]
was applied.
were used for C, H, O and Li elements, whereas Los Alamos
relatively high (65.8 kJmol ), which indicates that the reac- National Laboratory 2 Double Zeta (LANL2DZ) effective core po-
The standard all-electron 6-31+G(d) basis sets
Li exchange in PhI with dimeric species (tBuLi) (CO ) is
2
2 4
–
1
tion should not be very fast in hydrocarbon solvents. How- tentials were employed for Br and I atoms.[
51–53]
The minima were
confirmed by vibrational frequency calculations [B3LYP/6-
1+G(d)+LANL2DZ] within harmonic approximation. PhLi(CO
product is unstable and rearranges during optimization to much
more stable PhCOOLi(CO . To optimize the structures of transi-
tion states, synchronous transit-guided quasi-Newton approach
qst3) was applied. In this method, three input structures are
ever, it also points out that the transition state becomes
more stable when electron-withdrawing substituents are at-
tached to the aromatic ring, which is consistent with experi-
mental results. Finally, attempts to obtain 1,2-dilithioben-
zene revealed that it can be formed by a direct lithiation of
3
2 3
)
2 3
)
(
1,2-diiodobenzene with an excess of tBuLi in the mixture
needed: one corresponds to substrates, a second to products and a
third is a guess of the transition state. To verify the structures of
transition states, frequency calculations were carried out at the
of Et O/pentane at ca –90 °C. However, the method suffers
2
from rather low yield and poor selectivity, which is due to
the extensive formation of 1,2-benzyne intermediate and B3LYP/6-31+G(d) level of theory. One imaginary frequency was
subsequent products.
found in the case of [Ph–Br–tBu]Li(THF)
Ph–I–tBu]Li(CO [C –I–tBu][tBuLi
Li(CO [transition state for rearrangement of tBuLi(CO
ium pivalate], [Ph–CO ]Li(CO [transition state for rearrange-
ment of PhLi(CO to lithium benzoate]. The perfluorinated
–I–tBu]Li(CO and [C –I–tBu][tBuLi (CO ] complexes
were more stable than the corresponding substrates, and all fre-
quencies found were real, which confirms that optimized structures
were true minima of potential energy hypersurface. Therefore, they
should be treated as intermediates, not as transition states. Unfor-
tunately, we were unable to find reliable transition states for the
3
, [Ph–I–tBu]Li(THF)
3
,
]
[
2
)
3
,
6
F
5
2
(CO ], [tBu–CO
2 4
)
2
2
)
2
2 3
) to lith-
Experimental Section
2
2 2
)
2 3
)
General Information: All reactions were carried out under an argon
atmosphere. Solvents used for reactions were dried by heating to
reflux with sodium/benzophenone and distilled under argon atmo-
sphere. Starting haloarenes, nBuLi (2.0 m in hexanes), tBuLi (1.7 m
in pentane), Me S , and B(OR) (R = Me, iPr) were used as re-
2 2 3
ceived without further purification. GC–MS analyses were per-
formed with a Perkin–Elmer Clarus 580 and 560S apparatus. H
NMR chemical shifts are given relative to tetramethylsilane using
residual solvent resonances.
[C
6
F
5
2
)
3
6
F
5
2
2 4
)
1
formation of these two complexes. The computations were per-
[54–56]
formed with Tomasi’s polarized continuum model,
using the
polarizable conductor calculation model [SCRF(CPCM, solvent)].
In the presence of CO ligands, the calculations were performed by
Lithiation/Carboxylation with tBuLi in Pentane. Typical Procedure:
A solution of 1-fluoro-2-iodobenzene 2 (2.22 g, 10 mmol) in pent-
ane (10 mL) was added dropwise to a stirred solution of tBuLi
2
using n-hexane as solvent. Such an approach was successfully ap-
plied in our previous studies on the mechanism of HLE in selected
haloarenes.[
63,69]
(1.7 m in pentane, 12.0 mL, 21 mmol) diluted with pentane (20 mL)
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
7

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
K. Durka, S. Luli n´ ski, M. D a˛ browski, J. Serwatowski
FULL PAPER
Supporting Information (see footnote on the first page of this arti-
cle): Copies of H and C NMR spectra of obtained carboxylic
acids as well as details of theoretical calculations.
[30] J. Mortier, T.-H. Nguyen, D. Tilly, A.-S. Castanet, ARKIVOC
2007, vi, 47–54.
1
13
[
31] H. J. S. Winkler, G. Wittig, J. Org. Chem. 1957, 28, 1733–1740.
32] K. Lerstrup, M. Lee, F. M. Wiygul, T. J. Kistenmacher, D.
Cowan, J. Chem. Soc., Chem. Commun. 1982, 294–295.
33] K. Lerstrup, D. Talham, A. Bloch, T. Poehler, D. Cowan, J.
Chem. Soc., Chem. Commun. 1982, 336–337.
[
[
[
Acknowledgments
34] H. F. Bettinger, M. Filthaus, J. Org. Chem. 2007, 72, 9750–
9752.
[
[
35] H. Gilman, B. J. Gaj, J. Org. Chem. 1957, 22, 447–449.
This work was supported by the Warsaw University of Technology.
The authors thank the Interdisciplinary Centre for Mathematical
and Computational Modelling in Warsaw for providing computa-
tional facilities (G33-14). The support by Aldrich Chemical Co,
U.S.A, through the continuous donation of chemicals and equip-
ment is gratefully acknowledged.
36] L. S. Chen, G. J. Chen, C. Tamborski, J. Organomet. Chem.
1
980, 193, 283–292.
37] F. Leroux, M. Schlosser, Angew. Chem. Int. Ed. 2002, 41, 4272–
274; Angew. Chem. 2002, 114, 4447–4449.
38] S. M. Bachrach, M. Hare, S. R. Kass, J. Am. Chem. Soc. 1998,
20, 12646–12649.
39] S. M. Bachrach, J. V. Miller Jr., J. Org. Chem. 2002, 67, 7389–
398.
40] S. M. Bachrach, A. C. Chamberlin, J. Org. Chem. 2004, 69,
111–2122.
41] C. Tamborski, E. Soloski, J. Organomet. Chem. 1969, 20, 245–
50.
42] A. Maercker (Ed.), Houben-Weyl, Carbanionen, vol. E19d, p.
86, Springer, Heidelberg, Germany, 1987.
[
[
[
[
[
[
[
[
[
4
1
7
[
[
1] E. Weiss, Angew. Chem. 1993, 105, 1565–1587.
2] D. W. Slocum, D. Reed, F. Jackson III, C. Friesen, J. Or-
ganomet. Chem. 1996, 512, 265–267.
2
2
[
3] D. W. Slocum, P. Dietzel, Tetrahedron Lett. 1999, 40, 1823–
1826.
4
[
4] D. W. Slocum, A. Carroll, P. Dietzel, S. Eilerman, J. P. Culver,
43] J. R. Baran Jr., C. Hendrickson, D. A. Laude Jr., R. J. Lagow,
J. Org. Chem. 1992, 57, 3760–3763.
B. McClure, S. Brown, R. W. Holman, Tetrahedron Lett. 2006,
47, 865–868.
44] D. Tilly, J. Fu, B. Zhao, M. Alessi, A.-S. Castanet, V. Snieckus,
J. Mortier, Org. Lett. 2010, 12, 68–71.
45] A. Krasovskiy, A. Tishkov, V. del Amo, H. Mayr, P. Knochel,
Angew. Chem. Int. Ed. 2006, 45, 5010–5014; Angew. Chem.
[
5] H. Günther, D. Moskau, P. Bast, D. Schmalz, Angew. Chem.
Int. Ed. Engl. 1987, 26, 1212–1220; Angew. Chem. 1987, 99,
1242.
[
[
[
[
6] D. L. McKeever, R. Waack, J. Chem. Soc. D 1969, 750–751.
2006, 118, 5132.
7] T. L. Brown, Pure Appl. Chem. 1970, 23, 447–463.
[
[
46] J. Cvengros, D. Stolz, A. Togni, Synthesis 2009, 2818–2824.
47] M. A. Schade, S. Yamada, P. Knochel, Chem. Eur. J. 2011, 17,
8] P. v. R. Schleyer, Pure Appl. Chem. 1984, 56, 151–162.
9] W. N. Setzer, P. v. R. Schleyer, Adv. Organomet. Chem. 1985,
4232–4237.
2
4, 353–451.
10] W. Bauer, W. R. Winchester, P. v. R. Schleyer, Organometallics
987, 6, 2371–2379.
[
[
[
48] A. D. Becke, Phys. Rev. A 1988, 38, 3098–3100.
[
[
[
[
[
[
49] C. Lee, W. Yang, R. G. Parr, Phys. Rev. B 1988, 37, 785–789.
50] R. Krishnan, J. S. Binkley, R. Seeger, J. A. Pople, J. Chem.
Phys. 1980, 72, 650–654.
1
11] W. R. Winchester, W. Bauer, P. v. R. Schleyer, J. Chem. Soc.,
Chem. Commun. 1987, 177–179.
[51] P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 270–283.
[52] W. R. Wadt, P. J. Hay, J. Chem. Phys. 1985, 82, 284–298.
[53] P. J. Hay, W. R. Wadt, J. Chem. Phys. 1985, 82, 299–310.
[54] E. Canses, B. Mennucci, J. Tomasi, J. Chem. Phys. 1997, 107,
12] J. P. Clayden, Organolithiums: Selectivity for Synthesis Perga-
mon, Oxford, UK, UK, 2002.
13] M. Kuprat, M. Lehmann, A. Schulz, A. Willinger, Organome-
tallics 2010, 29, 1421–1427.
3032–3041.
14] X. Wang, P. Rabbat, P. O’Shea, R. Tillyer, E. J. J. Grabowski,
P. J. Reider, Tetrahedron Lett. 2000, 41, 4335–4338.
15] W. Li, D. P. Nelson, M. S. Jensen, R. S. Hoerrner, D. Cai, R. D.
Larsen, P. J. Reider, J. Org. Chem. 2002, 67, 5394–5397.
16] S. L. Buchwald, J. Am. Chem. Soc. 1987, 109, 4396–4397.
17] N. Rot, F. Bickelhaupt, Organometallics 1997, 16, 5027–5031.
18] J. P. Wiley, Bioconjugate Chem. 2001, 12, 240–250.
19] M. Fossatelli, R. den Besten, H. D. Verkruijsse, L. Brandsma,
Recl. Trav. Chim. Pays-Bas 1994, 113, 527–528.
[
[
55] J. Tomasi, M. Persico, Chem. Rev. 1994, 94, 2027–2094.
56] J. Tomasi, B. Mennucci, R. Camm, Chem. Rev. 2005, 105,
2999–3093.
[
[
[
[
[
[
57] P. Beak, D. J. Allen, J. Am. Chem. Soc. 1992, 114, 3420–3425.
58] H. J. Reich, W. L. Whipple, Can. J. Chem. 2005, 83, 1577–1587.
59] R. R. Rogers, J. Houk, J. Am. Chem. Soc. 1982, 104, 522–525.
60] G. Wittig, U. Schöllkopf, Tetrahedron 1958, 3, 91–93.
61] K. Ando, J. Org. Chem. 2006, 71, 1837–1850.
62] C. Rodríguez, A. J. Vázquez, N. S. Nudelman, J. Phys. Org.
Chem. 2010, 23, 978–984.
[
[
[
[
[20] E. J. Corey, D. J. Beames, J. Am. Chem. Soc. 1972, 94, 7210–
7211.
[
[
21] D. Seebach, H. Neumann, Chem. Ber. 1974, 107, 847–850.
22] D. W. Slocum, D. Kusmic, J. C. Raber, T. K. Reinscheld, P. E.
Whitley, Tetrahedron Lett. 2010, 51, 4793–4796.
[63] K. Durka, S. Luli n´ ski, J. Sm e˛ tek, M. D a˛ browski, J. Serwatow-
ski, K. Wo z´ niak, Eur. J. Org. Chem. 2013, 3023–3032.
[64] W. B. Farnham, J. C. Calabrese, J. Am. Chem. Soc. 1986, 108,
2449–2451.
[65] K. B. Wiberg, S. Sklenak, J. Org. Chem. 2000, 65, 2014–2021.
[66] H. Gilman, J. Am. Chem. Soc. 1956, 78, 2217–2222.
[67] A wide range of melting points are reported (ca. 190–230 °C),
see for example: M. T. Nuñez, V. S. Martin, J. Org. Chem. 1990,
55, 1928–1932.
[68] M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria,
M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B.
Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li,
H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Son-
nenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hase-
gawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai,
T. Vreven, J. A. Montgomery Jr., J. E. Peralta, F. Ogliaro, M.
Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Starov-
[
[
[
[
[
[
[
23] R. D. Thomas, M. T. Clarke, R. M. Jensen, C. T. Young, Orga-
nometallics 1986, 5, 1851–1857.
24] R. D. Thomas, R. M. Jensen, C. T. Young, Organometallics
1
987, 6, 565–571.
25] B. D. Palmer, M. Boyd, W. A. Denny, J. Org. Chem. 1990, 55,
38–441.
26] A. R. Katritzky, K. Akutagawa, Tetrahedron Lett. 1985, 26,
935–5938.
27] A. R. Katritzky, K. Akutagawa, Tetrahedron 1986, 42, 2571–
574.
4
5
2
28] A. R. Katritzky, L. M. Vazquez de Miguel, G. W. Rewcastle,
Synthesis 1988, 215–217.
29] A. R. Katritzky, M. Black, W.-Q. Fan, J. Org. Chem. 1991, 55,
5045–5048.
8
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
Is Carbon Dioxide Able to Activate Halogen/Lithium Exchange?
erov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell,
J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M.
Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Ad-
amo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev,
A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Mar-
tin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador,
J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B.
Foresman, J. V. Ortiz, J. Cioslowski, D. J. Fox, Gaussian09, re-
vision E.01, Gaussian, Inc., Wallingford, CT, 2010.
[69]
S. Luli n´ ski, J. Sm e˛ tek, K. Durka, J. Serwatowski, Eur. J. Org.
Chem. 2013, 8315–8322.
Received: April 4, 2014
Published Online: ᭿
Eur. J. Org. Chem. 0000, 0–0
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.eurjoc.org
9

Job/Unit: O42320
/KAP1
Date: 16-06-14 11:03:26
Pages: 10
K. Durka, S. Luli n´ ski, M. D a˛ browski, J. Serwatowski
FULL PAPER
Lithiation Reaction Mechanism
K. Durka, S. Luli n´ ski,* M. D a˛ browski,
J. Serwatowski ................................ 1–10
Is Carbon Dioxide Able to Activate Hal-
ogen/Lithium Exchange?
Keywords: Reaction mechanisms / Carbon
dioxide / Carboxylation / Lithiation / Sol-
vent effects
The unexpected effect of carbon dioxide on
the rates of halogen–lithium exchange reac-
tions of selected haloarenes with tBuLi in
aliphatic hydrocarbon solvent is reported.
Synthetic and computational studies on a
plausible mechanism of the reaction are
discussed.
10
www.eurjoc.org
© 0000 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Eur. J. Org. Chem. 0000, 0–0