SN1 reactions in supercritical carbon dioxide in the presence of alcohols: The role of preferential solvation

Article abstract of DOI:10.1039/c6ob01097k

Ethanol (3b) inhibits S_N1 reactions of alkyl halides 1 in supercritical carbon dioxide (scCO₂) and gives no ethers as products. The unexpected behaviour of alcohols 3 in the reaction of alkyl halides 1 with 1,3-dimethoxybenzene (2) in scCO₂ under different conditions is rationalised in terms of Br?nsted and Lewis acid-base equilibria of reagents, intermediates, additives and products in a singular solvent characterised by: (i) the strong quadrupole and Lewis acid character of carbon dioxide, which hinders S_N2 paths by strongly solvating basic solutes; (ii) the weak Lewis base character of carbon dioxide, which prevents it from behaving as a proton sink; (iii) the compressible nature of scCO₂, which enhances the impact of preferential solvation on carbon dioxide availability for the solvent-demanding rate determining step.

Full text of DOI:10.1039/c6ob01097k

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
PAPER
SN1 reactions in supercritical carbon dioxide
in the presence of alcohols: the role of
preferential solvation†
Cite this: Org. Biomol. Chem., 2016,
14, 6554
Thais Delgado-Abad, Jaime Martínez-Ferrer, Rafael Acerete, Gregorio Asensio,
Rossella Mello and María Elena González-Núñez*
Ethanol (3b) inhibits SN1 reactions of alkyl halides 1 in supercritical carbon dioxide (scCO2) and gives
no ethers as products. The unexpected behaviour of alcohols 3 in the reaction of alkyl halides 1
with 1,3-dimethoxybenzene (2) in scCO2 under different conditions is rationalised in terms of Brønsted
and Lewis acid–base equilibria of reagents, intermediates, additives and products in a singular solvent
characterised by: (i) the strong quadrupole and Lewis acid character of carbon dioxide, which hinders SN2
paths by strongly solvating basic solutes; (ii) the weak Lewis base character of carbon dioxide, which
prevents it from behaving as a proton sink; (iii) the compressible nature of scCO2, which enhances
the impact of preferential solvation on carbon dioxide availability for the solvent-demanding rate deter-
mining step.
Received 19th May 2016,
Accepted 7th June 2016
DOI: 10.1039/c6ob01097k
www.rsc.org/obc
Introduction
Supercritical carbon dioxide (scCO2) is an alternative solvent
for green chemistry1 characterised by zero dipole moment,
very low dielectric constant, and no hydrogen-bonding behav-
iour,2 yet is suitable for performing uncatalysed SN1 reactions
of alkyl halides 1 3a and electrophilic brominations of weakly
activated aromatics.3b The strong quadrupole and the Lewis
acid but non-basic character of carbon dioxide3,4 accounts for
the ability of scCO2 to solvate ionic species and to avoid
capture by acidic intermediates.
Scheme 1 Reaction of alkyl halides 1 with aromatic 2 in scCO2.
product distribution and ionising ability of scCO2. The results
revealed that solute–carbon dioxide interactions modify the
course of these reactions in relation to conventional solvents,
and stressed the relevance of solvation by scCO2 when design-
ing applications of this medium as a solvent for green
chemistry.
In the course of our study3a on the reaction of alkyl halides
1 with 1,3-dimethoxybenzene 2 in scCO2, which proceeds
through the solvent-promoted ionisation of 1 and capture of
carbenium ion I by the aromatic (Scheme 1), we noticed that
alcohols 3 inhibited the formation of Friedel–Crafts adducts, 4
and 5, and no ethers 6 formed as products.3a This unexpected
behaviour5 for a polar, protic and nucleophilic additive
prompted us to further explore SN1 reactions as sensitive
probes for solvation in scCO2.6 Herein we report on reactions
of alkyl halides 1, 1,3-dimethoxybenzene (2), and alcohols 3
in scCO2 under different conditions, and disclose the reaction
paths involved, as well as the impact of solvation on the
Results and discussion
The model systems selected for exploring SN1 reactions in
scCO2 were 1-chloro-1-phenylethane (1a) and benzylbromide
(1c) as ionogens, 1,3-dimethoxybenzene (2) as the aromatic,
and 1-phenylethanol (3a) and ethanol (3b) as additives. The
reactions were performed and analysed following reported pro-
cedures.3a The experiments performed in view cells showed
homogeneous reaction mixtures in all cases. Styrene and pro-
ducts derived from CO2-capture (carbonates or carboxylic
acids) were not detected in the reaction mixtures. The ESI†
provides a detailed description of the experimental procedure
and the balance of products for the isotopic tracer experiments
(Table S1†).
Departamento de Química Orgánica, Facultad de Farmacia, Universidad de Valencia,
Avda. Vicente Andrés Estellés s.n. 46100-Burjassot, Valencia, Spain.
E-mail: elena.gonzalez@uv.es
†Electronic supplementary information (ESI) available: Experimental procedures
and spectra. See DOI: 10.1039/c6ob01097k
6554 | Org. Biomol. Chem., 2016, 14, 6554–6560
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Paper
SN1 reactions of alkyl halides 1 and alcohols in scCO2:
reaction pathways
The absence of ethers 6 as products from the reactions of alkyl
halides 1 with aromatic 2 in scCO2 in the presence of ethanol
(3b) was firstly attributed to solvation by scCO2, which would
either prevent alcohol 3b to react with carbocations I, or would
modify the reaction paths of the intermediate species in
relation to those observed in conventional solvents. In order to
explore these possibilities, we designed isotopic tracer experi-
ments with 1-chloro-1-phenylethane (1a) and 1-phenylethanol
(3a) in scCO2 in order to open alternative reaction channels to
the intermediate species and track the reaction components
under these conditions. Scheme 2 depicts the ionisation
and proton-transfer equilibria, and the irreversible aromatic
electrophilic substitution paths involved in these reactions.
Fig. 1 Isotopic tracer experiments for reactions of 1-chloro-1-phenyl-
ethane (1a) (0.05 M), and 1-phenylethanol (3a) in scCO2 at 250 bar and
60 °C for 15 h. Isotopic labels were obtained from the relative intensities
of the ions [M+] for 1d-1a, 3a, and [M − 15+] for 6aa. The figures are the
average of at least three independent experiments within standard devi-
ation of 15%.
Reaction of 1-chloro-1-phenylethane (1a) with 1-phenylethanol
(3a). Isotopic tracer experiments
The reaction of equimolar amounts of 1-chloro-1-deutero-
1-phenylethane (1d-1a) 80 D-atom% and 1-phenylethanol (3a) dehydration of alcohol 3a under these conditions (Scheme 2).7
The formation of symmetrical ether D,D-6aa (entry 1, Fig. 1)
suggests that protonated ether H,D-6aaH+ ionises in the reac-
in scCO2 at 250 bar and 60 °C for 15 h gave 36% of the corres-
ponding ether 6aa as a mixture of syn- and anti-diastereomers
(run 1, Fig. 1). This result shows that solvation by scCO2 does tion medium to release alcohols 3a or 1d-3a and carbocation
intermediates Ia or 1d-Ia (Scheme 2).
not prevent alcohol 3a from capturing carbocation intermedi-
ates 1d-Ia formed by the solvent-promoted ionisation of alkyl
The reaction of 1-chloro-1-phenylethane (1a) with a twofold
halide 1d-1a (Scheme 2). The reaction of 1-phenylethanol (3a) excess of 1-deutero-1-phenylethanol (1d-3a) in scCO2 under the
same conditions (entry 2, Fig. 1) followed the same trends. In
this case, symmetrical ether D,D-6aa was the major product
with alkyl halide 1a to give ether 6aa contrasts with the reluc-
tance of ethanol (3b) to undergo an analogous transformation,
and suggests that the ability of alcohol 3 to ionise under these with conversions of alkyl halide 1a and alcohol 1d-3a of 8%
and 48%, respectively (Table S1†).
conditions may determine the reaction course.
Finally, the low isotopic dilution observed for unreacted
The formation of ether H,H-6aa in amounts larger than
ether H,D-6aa (entry 1, Fig. 1), and the higher conversion rate alkyl halide 1d-1a and alcohol 1d-3a (4–13% and 7–0%,
respectively, Table S1†) compared to the relatively high extent
of ethers 6aa formed (entries 1 and 2, Fig. 1) indicates that the
of alcohol 3a (81%) compared to alkyl halide 1d-1a (25%)
(Table S1†), indicate the involvement of acid catalysed
Scheme 2 Reaction paths proposed for 1-chloro-1-deutero-1-phenylethane (1d-1a), 1-phenylethanol (3a), and 1,3-dimethoxybenzene (2) in scCO2.
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 6554–6560 | 6555

View Article Online
Paper
Organic & Biomolecular Chemistry
chloride anion does not compete efficiently with alcohols 3a to wise the D-labelling for adducts 4a and 5a would be lower than
trap carbocation intermediates Ia in scCO2.
those observed. The results hence suggest that hydrogen chlor-
ide formed in the Friedel–Crafts reaction competes efficiently
with carbocations Ia and 1d-Ia to trap alcohol 3a and that the
H-bonding interactions8 of protonated species 3aH+ and
6aaH+ with alcohol 3a facilitate the acid-catalysed dehydration
pathway. These interactions are enhanced by the low basicity
of carbon dioxide,9 which prevents the solvent from behaving
as a proton sink in these reactions.
The reactions performed with equimolar amounts of
ionogen 1d-1a, alcohol 3a and 1,3-dimethoxybenzene (2)
(entry 3, Fig. 2) led to Friedel–Crafts adducts 4a with 37–35
D-atom%, respectively, and disubstituted adducts 5a in the
ca. 2 : 2 : 1 H,H : H,D : D,D distribution. The conversions of
1d-1a and 3a were 40% and 95%, respectively (Table S1†).
These isotopic traces reveal the increased competitiveness of
alcohol 3a in relation to 1,3-dimethoxybenzene (2) to trap car-
bocations 1d-Ia and Ia compared to the reactions performed
with a 4-fold excess of 2 (entry 2, Fig. 2).
Reaction of 1-chloro-1-phenylethane (1a) with 1,3-
dimethoxybenzene (2) in the presence of 1-phenylethanol (3a).
Isotopic tracer experiments
1-Chloro-1-phenylethane (1a) reacts quantitatively with aro-
matic 2 (4 equiv.) in scCO2 at 60 °C and 250 bar for 5 h to give
mono- and di-substituted Friedel–Crafts adducts 4a (82%) and
5a (18%) as mixtures of regioisomers (4ao,p and 4ao,o) and dia-
stereomers (5 and 5′), respectively (entry 1, Fig. 2).3a Con-
versely, the reaction of equimolar 1-chloro-1-deutero-1-
phenylethane (1d-1a) 80 D-atom% and 1-phenylethanol (3a)
with a 4-fold excess of 1,3-dimethoxybenzene (2) in scCO2 for
15 h under the same conditions (entry 2, Fig. 2) gave mono-
substituted Friedel–Crafts adducts 4a (38%) with 44–47
D-atom% and disubstituted adducts 5a (6%) with a nearly stat-
istical (1 : 2 : 1) H,H : H,D : D,D distribution (entry 2, Fig. 1).
Substrate conversions were 42% and 72% for 1d-1a and 3a,
respectively (Table S1†).
The isotopic dilutions found for the unreacted substrates in
these reactions (entries 2 and 3, Fig. 2) were 4–13% for alkyl
halide 1d-1a, and were negligible for alcohol 3a (Table S1†).
These data evidence that the Friedel–Crafts reaction
involves nearly equivalent amounts of carbocations 1d-Ia and
Ia, and indicate that protonated ether H,D-6aaH+ is the major
source of electrophilic intermediates for this reaction. There-
fore, carbocations 1d-Ia, formed by the ionisation of alkyl
Friedel–Crafts reactions of alkyl halide 1a in the presence of
ethanol (3b)
halide 1d-1a, react with alcohol 3a faster than with 1,3- The Friedel–Crafts reactions of alkyl halide 1a with aromatic 2
dimethoxybenzene (2) (Scheme 2). in scCO2 in the presence of ethanol (3b) followed a different
The isotopic traces provide further information on the reac- course (Fig. 3). Thus increasing amounts of ethanol (3b) above
tion course in scCO2. Thus the preferential formation of sym- 0.5 equiv. progressively inhibited the conversion of alkyl
metrical H,H-ethers 6aa (entries 2 and 3, Fig. 2), and the halide 1a (entries 2–5, Fig. 3). Ethyl 1-phenylethyl ether (6ab),
conversions of alkyl halide 1d-1a (42%) and alcohol 3a (7%) 1-phenylethanol (3a), and ethyl-substituted Friedel–Crafts
(Table S1†), once again indicate the involvement of the acid adducts 4b and 5b were not found as products under these
catalysed dehydration of alcohol 3a under these conditions. conditions. These results can now be interpreted with the
However, the isotopic labels found for Friedel–Crafts adducts information collected from the isotopic tracer experiments
4a (44–47 D-atom%) and 5a (H,H : H,D : D,D ca. 1 : 2 : 1) (entry (Scheme 3).
1, Fig. 2) suggest that protonated alcohol 3aH+ and ethers
Protonated ether 6abH+ is the source of carbocation inter-
6aaH+ from the acid-catalysed dehydration path (Scheme 2) do mediates Ia for Friedel–Crafts reactions in the presence of
not play a significant role in the Friedel–Crafts reaction, other- ethanol (3b) (Scheme 3). The absence of 1-phenylethanol (3a)
Fig. 2 Reactions of 1-chloro-1-phenylethane (1a) (0.05) M, 1,3-dimethoxybenzene (2), and 1-phenylethanol (3a) in scCO2 for 5 h (run 1)3a and 15 h
(runs 2 and 3). Isotopic labels were obtained from the relative intensities of the ions [M+] for 1d-1a, 3a, 4a, and 5a, and [M − 15+] for 6aa. The figures
are the average of at least three independent experiments within standard deviation of 15%.
6556 | Org. Biomol. Chem., 2016, 14, 6554–6560
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Paper
SN1 reactions of alkyl halides 1 and alcohols in scCO2:
preferential solvation
The results reported in Fig. 1–3 have shown that alcohols 3
react with carbocation intermediates Ia formed by the scCO2-
promoted ionisation of 1a faster than 1,3-dimethoxybenzene
(2), and that the irreversible Friedel–Crafts reaction with 1,3-
dimethoxybenzene (2) involves carbocations Ia formed from a
series of competitive ionisation and proton-transfer reversible
processes (Schemes 2 and 3). The reaction paths depicted in
Schemes 2 and 3 do not account, however, for the low isotopic
dilution found for unreacted substrates 1d-1a and 3a
(Table S1†), the low conversions of alkyl halide 1a (Fig. 1 and
2; Table S1†), nor for the progressive reaction inhibition pro-
moted by increasing amounts of ethanol (3b) (Fig. 3). These
Fig. 3 Reactions of 1-chloro-1-phenylethane (1a) (0.05) M and ethanol results rather suggest that alcohols 3 inhibit the ionisation of
(3b) with 1,3-dimethoxybenzene (2) (4 equiv.) in scCO2. The figures are
the average of at least three independent experiments within standard
deviation of 15%.
alkyl halide 1a in scCO2, a notion which apparently contradicts
the well-established behaviour of alcohols 3 as solvents for SN1
reactions.2,5 Nevertheless, it is known that ethanol (3b)
behaves as an electrophilic catalyst for SN1 reactions in non-
polar and non-protic solvents,5c yet inhibits SN1 reactions in
solvents which are stronger H-bond donors than itself.10 For
instance, the solvolysis of alkyl halides 1 in binary mixtures
water : 3b10a and fluorinated alcohols : 3b,10b become progress-
ively slower with increasing amounts of ethanol (3b). The
same trend is known for acetone, dimethylsulfoxide, and for
dioxane as cosolvents for SN1 reactions in aqueous medium.11
The changes in the reaction rates and equilibria promoted
by additives or cosolvents are generally attributed to com-
petitive substrate–solvent, substrate–cosolute, and cosolute–
solvent interactions.12 This concept suggests that alcohols 3
may hinder the highly solvent-demanding rate-determining
step of SN1 reactions in scCO2 by competing with alkyl halides
Scheme 3 Prevailing reaction path for 1-chloro-1-phenylethane (1a),
1,3-dimethoxybenzene (2) and ethanol (3b) in scCO2.
1 for solvation. The properties of scCO2, such as compressibil-
ity, densities lower than those for liquid solvents within the
75–250 bar range,13 and density inhomogeneities promoted by
specific solute–solvent and solvent–solvent interactions,14
as a product in these reactions (Scheme 3) shows that SN2
would enhance the effect of alcohols 3 on these reactions. In
order to explore this possibility, we performed the reaction of
benzyl bromide (1c), 1,3-dimethoxybenzene (2) and ethanol
(3b) in scCO2 under different conditions (eqn (1), Table 1).
nucleophilic displacements on the highly electrophilic primary
carbon atom of 6abH+ by either alcohol 3b or the chloride
anion do not compete with unimolecular pathways under
these conditions (Scheme 3). Solvation by carbon dioxide prob-
ably hinders nucleophiles approaching electrophilic sp3
Friedel–Crafts reactions of alkyl halide 1c in the presence of
ethanol (3b)
carbon atoms in an SN2 fashion.
This fact limits the reaction paths available for protonated
ether 6abH+ to the back-ionisation and proton transfer to Benzyl bromide (1c) was proved to be more sensitive to ethanol
ethanol (3b), and those for 3bH+ to the back proton transfer to (3b) than 1-chloro-1-phenylethane (1a) (Table 1). Thus 0.7
ether 6ab (Scheme 3) since neither 6abH+ nor 3bH+ undergoes equivalents of ethanol (3b) sufficed to suppress the reaction
SN2 or SN1 reactions at the primary carbon atom under these for 1c (entry 9, Table 1), while more than 2 equiv. were
conditions (Scheme 3). Accordingly, the reaction progresses required to achieve the same effect for 1a (entry 5, Fig. 3). This
towards the irreversible electrophilic aromatic substitution result suggests a stronger solvent-demand in the rate-deter-
with aromatic (2) (Scheme 3), which makes ethyl 1-phenylethyl mining step for benzyl bromide (1c) than for 1-chloro-1-
ether (6ab) formation unfeasible. Conversely, the reactions phenylethane (1a), in agreement with the different stabilities
with 1-phenyl-1-ethanol (3a) as an additive (Fig. 2) have uni- of primary and secondary benzylic carbocations, Ic < Ia.
molecular paths available to deplete protonated alcohol 3aH+
from the solution due to the stability of secondary benzylic on both ethanol (3b) and benzyl bromide (1c) concentrations
The results showed that reaction efficiency was dependent
carbocation Ia (Scheme 2).
(Table 1). Thus reactions with a molar ratio of 1c : 3b of 1 : 0.5
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 6554–6560 | 6557

View Article Online
Paper
Organic & Biomolecular Chemistry
rate-determining step (step 5). Steps 4 and 5 are expected to be
strongly sensitive to the cosolutes that compete with alkyl
halide 1c for solvation. By assuming that the concentration of
carbon dioxide available for alkyl halide 1 is roughly that
which remains after the solvation of alcohol 3b, the rate law
for a simplified scheme with eqn (1), (2), (4) and (5)
(Scheme 4) shows that the order for ethanol (3b) depends on
the solvation requirement of alkyl halide 1 in the rate-deter-
mining step (Scheme 4). Therefore, the impact of alcohol 3b
on the reaction depends on carbocation stability, the leaving
group ability of the halide anion, and the solvation demand of
the ion pairs. Accordingly, 1-chloro-1-phenylethane (1a) was
found to be less sensitive to ethanol (3b) than benzyl bromide
(1c) (Fig. 3 and Table 1).
Ethanol (3b) exerts stronger interactions with carbon
dioxide than alkyl halides 1 and 1,3-dimethoxybenzene (2)
given its stronger basic and H-bond donor character.8 Hence
the solvation shell for 3b would be larger than that for 1
(n > m), and the solvent exchange equilibrium would favour
the solvation of 3b (step 3, Scheme 4). In this way, ethanol (3b)
rarefies the reaction medium and renders it less efficient to
perform the solvent-demanding ionisation of 1 and the dis-
sociation of the resulting ion pair (Scheme 4).16 The ability of
a cosolute to perturb SN1 reactions in scCO2 would then
depend on its basicity. Accordingly, reactions would be insen-
sitive to phenol17 or increasing concentrations of aromatic 2
(entries 1–3, Table 1) while readily inhibited by tertiary amines
or water.3a It is noteworthy that solvation in scCO2 could
involve solute–solvent interactions that are not directly related
to basicity. For instance, fluorinated solutes are commonly
called “CO2-philic” owing to the ability of fluorine atoms
to interact with carbon dioxide.18 Actually, we found that 2,2,2-
trifluoroethanol inhibits the ionisation of alkyl halides 1 as
efficiently as ethanol (3b) or water.
Table 1 Reaction of benzyl bromide (1c) with 1,3-dimethoxybenzene
(2) in scCO2, in the presence of ethanol (3b)a
Run
Molar ratio of 1c : 3b
[1c] (M)
[3b] (M)
Conv. (%)
1b
2b
3b
4b,c
5
6
7
8
9
10
11
12
13
14d
—
—
—
—
0.1
—
—
—
—
98
98
99
99
—
—
76
—
—
76
77
83
—
59
0.05
0.025
0.05
0.1
0.1
0.1
0.05
0.05
0.05
0.05
0.05
0.025
0.025
1 : 1
0.1
1 : 0.5
1 : 0.25
1 : 1
1 : 0.7
1 : 0.6
1 : 0.5
1 : 0.25
1 : 1
0.05
0.025
0.05
0.035
0.03
0.025
0.0125
0.025
0.0125
1 : 0.5
a Reactions were performed with the molar ratio of 1c : 2 of 1 : 4, for
15 h. The figures are the average of at least three independent experi-
ments within standard deviation of 15%. b Typical product distri-
bution: 4co,p 82%, 4co,o 6%, 5c 6%, and 5c′ 6%. c Reaction at 90 bar.
d Molar ratio 1c : 2 1 : 8.
proceeded for [3b] = 0.025 M, but not for [3b] = 0.05 M
(entries 6 and 11, Table 1), while the reactions with a molar
ratio of 1c : 3b of 1 : 1 did not proceed at all, not even for [3b] =
0.025 M (entries 5, 8 and 13, Table 1). These trends strongly
suggest that ethanol (3b) competes with alkyl halide 1c for
solvation.15
Scheme 4 illustrates the solvation equilibria15 for alkyl
halide 1 and ethanol (3b) (steps 1 and 2), the exchange of
carbon dioxide molecules clustered around 1c and 3b (step 3),
and the integration of further carbon dioxide molecules into
the solvation shell of alkyl halide 1 (step 4) required in the
The reaction of benzyl bromide (1c) with 1,3-dimethoxy-
benzene (2) at 60 °C in scCO2 in the presence of ethanol (3b)
was found to be pressure-insensitive within the 100–250 bar
range for [1c] = 0.05 M and ratios 1c : 3b 1 : 0.5, 1 : 0.6 and
1 : 0.7. This result is related to the dependence of the size of
the solvation shells on pressure, which follows the same pace
as the bulk density within this range.14 Thus raising pressure
does not provide larger amounts of uncoordinated carbon
dioxide molecules available for solvation of 1c since the size of
the clusters around the solutes also increases.14
If alcohol 3 does not fully prevent the ionisation of alkyl
halide 1 (Fig. 3), the subsequent reactions generate hydrogen
halides, which readily ionise in the reaction medium (Schemes
2 and 3). Since these ionic species introduce further solvation
equilibria and displace the alkyl halide ionisation equilibrium
to the left-hand side (step 5, Scheme 4), the SN1 reactions of
alkyl halides 1 in scCO2 in the presence of alcohols 3 are
expected to autoinhibit to an extent that depends on the sol-
vation requirements of alkyl halide 1, alcohol 3, halide anions
X−, protonated intermediates, and side products (ethers,
water), in agreement with the experimental observations
(Fig. 1–3 and Table 1).
Scheme 4 Mechanistic scheme with solvation equilibria for the reac-
tion of alkyl halide 1 (X = Cl, Br) with 1,3-dimethoxybenzene (2) in scCO2
in the presence of ethanol (3b). The rate law applies to eqn (1), (2), (4)
and (5), by considering the actual concentration of CO2 available for sol-
vation of alkyl halide 1 as that remaining after solvation of ethanol (3b).
6558 | Org. Biomol. Chem., 2016, 14, 6554–6560
This journal is © The Royal Society of Chemistry 2016

View Article Online
Organic & Biomolecular Chemistry
Paper
These results show that the SN1 reactions of alkyl halides 1
in scCO2 are extremely sensitive to cosolutes that are able to
exert strong interactions with carbon dioxide. Therefore, alco-
hols 3, water, amines, phosphines, ketones, esters or fluori-
nated compounds, among others, must be rigorously excluded
from the reaction medium in order to obtain reproducible
results. In the ESI† we provide detailed experimental pro-
cedures that address avoiding reaction medium contamination
while performing these reactions.
3 (a) T. Delgado-Abad, J. Martínez-Ferrer, A. Caballero,
A. Olmos, R. Mello, M. E. González-Núñez and G. Asensio,
Angew. Chem., Int. Ed., 2013, 52, 13298–13301;
(b) T. Delgado-Abad, J. Martínez-Ferrer, J. Reig-López,
R. Mello, R. Acerete, G. Asensio and M. E. González-Núñez,
RSC Adv., 2014, 4, 51016–51021.
4 (a) P. Raveendran, Y. Ikushima and S. L. Wallen, Acc. Chem.
Res., 2005, 38, 478–485; (b) S. Kazarian, M. F. Vincent,
F. V. Bright, C. L. Liotta and C. A. Eckert, J. Am. Chem. Soc.,
1996, 118, 1729–1736.
5 (a) F.-A. Carey and R. J. Sundberg, Advanced Organic Chemi-
stry, Part A: Structure and Mechanisms, Kluwer/Plenum,
New York, 4th edn, 2000, ch. 5; (b) T. H. Lowry and
K. S. Richardson, Mechanism and Theory in Organic
Chemistry, Harper & Row Publishers, New York, 3rd edn,
1987, ch. 4 and 5; (c) K. Okamoto, Pure Appl. Chem., 1984,
56, 1797–1808.
Conclusions
In summary, the unexpected behaviour of ethanol (3b) in the
SN1 reactions of alkyl halides 1 with 1,3-dimethoxybenzene (2)
in scCO2 namely, reluctance to form ethers and reaction rate
inhibition, is the result of the Brønsted and Lewis acid–base
equilibria that take place in a peculiar reaction medium in
which: (i) the strong quadrupole and Lewis acid character of
carbon dioxide hinders SN2 paths by strongly solvating basic
and nucleophilic solutes; (ii) the weak Lewis base character of
carbon dioxide prevents it from behaving as a proton sink; (iii)
the compressible nature of scCO2 enhances the impact of pre-
ferential solvation on carbon dioxide availability for the
solvent-demanding rate-determining step. Thus the SN1 reac-
tions of alkyl halides 1 in scCO2 are inhibited in the presence
of cosolutes that are able to exert stronger interactions with
the solvent than substrates 1. The same behaviour is known
for the SN1 reactions of alkyl halides 1 in an aqueous medium
in the presence of additives. The results reported herein reveal
that scCO2 is a remarkably structured solvent2 capable of pro-
moting and sustaining ionic reactions and is, therefore, not at
all similar to n-hexane or carbon tetrachloride as commonly
regarded.
6 J. F. Brennecke and J. E. Chateauneuf, Chem. Rev., 1999, 99,
433–452.
7 W. K. Gray, F. R. Smail, M. G. Hitzler, S. K. Ross and
M. Poliakoff, J. Am. Chem. Soc., 1999, 121, 10711–10718.
8 (a) F. Lafrad, A. Idrissi and T. Tassaing, J. Supercrit. Fluids,
2014, 94, 65–70; (b) S. Reiser, N. McCann, M. Horsch and
H. Hasse, J. Supercrit. Fluids, 2012, 68, 94–103; (c) W. Xu,
J. Yang and Y. Hu, J. Phys. Chem. B, 2009, 113, 4781–4789;
(d) M. Saharay and S. Balasubramanian, J. Phys. Chem. B,
2006, 110, 3782–3790; (e) P. Lalanne, T. Tassaing,
Y. Danten, F. Cansell, S. C. Tucker and M. Besnard, J. Phys.
Chem. A, 2004, 108, 2617–2624; (f) Y. Danten, T. Tassaing
and M. Besnard, J. Phys. Chem. A, 2002, 106, 11831–
11840.
9 S. G. Lias, J. F. Liebman and R. D. Levin, J. Phys. Chem.
Ref. Data, 1984, 13, 695–808.
10 (a) T. W. Bentley and G. E. Carter, J. Am. Chem. Soc., 1982,
104, 5741–5747; (b) T. W. Bentley, G. Llewellyn and
Z. H. Ryu, J. Org. Chem., 1998, 63, 4654–4659.
Acknowledgements
11 D. J. Raber, R. C. Bingham, J. M. Harris, J. L. Fry and
P. v. R. Schleyer, J. Am. Chem. Soc., 1970, 92, 5977–
5981.
Financial support from the Spanish Ministerio de Economía y
Competitividad (CTQ2013-47180-P), and Fondos FEDER is
gratefully acknowledged. TDA thanks the Spanish Ministerio
de Educación, Cultura y Deporte for fellowships. We thank the
SCSIE (Universidad de Valencia) for access to its instrumental
facilities.
12 (a) S. Gupta, S. Rafiq, M. Kundu and P. Sen, J. Phys. Chem.
B, 2012, 116, 1345–1355; (b) J. L. Cook, S. A. Hunter,
C. M. R. Low, A. Perez-Velasco and J. G. Vinter, Angew.
Chem., Int. Ed., 2008, 47, 6275–6277; (c) J. L. Cook,
S. A. Hunter, C. M. R. Low, A. Perez-Velasco and
J. G. Vinter, Angew. Chem., Int. Ed., 2007, 46, 3706–3709;
(d) A. Bagno, G. Scorrano and S. Stiz, J. Am. Chem. Soc.,
1997, 119, 2299–2300; (e) C. H. Langford and J. P. K. Tong,
Acc. Chem. Res., 1977, 10, 258–264.
Notes and references
1 (a) Handbook of Green Chemistry Vol. 4 Supercritical Solvents,
ed. W. Leitner and P. G. Jessop, Wiley-VCH, New York, 13 (a) R. Span and W. Wagner, J. Phys. Chem. Ref. Data, 1996,
2010; (b) Green Chemistry Using Liquid and Supercritical
Carbon Dioxide, ed. J. M. DeSimone and W. Tumas, Oxford
University Press, Oxford, 2003; (c) E. J. Beckman, J. Super-
crit. Fluids, 2004, 28, 121; (d) C. M. Rayner, Org. Process Res.
Dev., 2007, 11, 121–132.
25, 1509–1596; (b) E. W. Lemmon, M. O. McLinden and
D. G. Friend, Thermophysical Properties of Fluid Systems,
in NIST Chemistry WebBook, NIST Standard Reference
Database Number 69, ed. P. J. Linstrom and W. G. Mallard,
National Institute of Standards and Technology,
2 C. Reichardt and T. Welton, Solvents and Solvent Effects in
Organic Chemistry, Wiley-VCH, Weinheim, 4th edn, 2011.
Gaithersburg,
[retrieved May 2nd, 2016].
MD,
20899,
http://webbook.nist.gov
This journal is © The Royal Society of Chemistry 2016
Org. Biomol. Chem., 2016, 14, 6554–6560 | 6559

View Article Online
Paper
Organic & Biomolecular Chemistry
14 (a) S. C. Tucker, Chem. Rev., 1999, 99, 391–418; 17 SN1 reactions in scCO2 in the presence of phenol proceed
(b) O. Kajimoto, Chem. Rev., 1999, 99, 355–389.
15 N. Nunes, C. Ventura, F. Martins and R. E. Leitao, J. Phys.
Chem. B, 2009, 113, 3071–3079.
16 This process is analogous to the salting-out effect, i.e.
the diminished solubility of non-electrolytes in the
efficiently to give Friedel–Crafts products derived from 1,3-
dimethoxybenzene (2) and phenol. These results have not
been included in the present study since phenol catalyzes
SN1 reactions in non-polar and non-protic solvents through
hydrogen bonding (see ref. 5c).
presence of ionic solutes with highly structured solvation 18 J. Peach and J. Eastoe, Beilstein J. Org. Chem., 2014, 10,
shells.2
1878–1895.
6560 | Org. Biomol. Chem., 2016, 14, 6554–6560
This journal is © The Royal Society of Chemistry 2016