Organic reactivity of alcohols in superheated aqueous salt solutions: An overview

Article abstract of DOI:10.1039/c2nj21038j

The low dielectric constant and high self-dissociation constant of water in a temperature range between 150 and 250 °C make it a very appealing solvent for synthesis. Surprisingly, while organic chemistry in water at low temperature or around its critical point has been investigated in detail, very little seems to be known about the behaviour of organic molecules under hydrothermal conditions. The present work thus aims at shading some light on this field. As a start, we decided to investigate the reactions in which alcohols can undergo in water in the above-mentioned temperature range. Knowing that very strong salt effects on organic reactions have already been observed in super critical water, the impact of salt on the outcome of our tests was also investigated in detail.

Full text of DOI:10.1039/c2nj21038j

View Article Online / Journal Homepage / Table of Contents for this issue
New Journal of Chemistry
A journal for new directions in chemistry
www.rsc.org/njc
Volume 36 | Number 8 | August 2012 | Pages 1529–1692
ISSN 1144-0546
PAPER
Frédéric Goettmann
Organic reactivity of alcohols in
superheated aqueous salt solutions:
an overview
1144-0546(2012)36:8;1-C

Dynamic Article Links
NJC
View Article Online
Cite this: New J. Chem., 2012, 36, 1568–1573
www.rsc.org/njc
PAPER
Organic reactivity of alcohols in superheated aqueous salt solutions: an
overvieww
a
b
c
a
´ ´
Sabine Avola, Frederic Goettmann,* Markus Antonietti and Werner Kunz
Received (in Montpellier, France) 13th December 2011, Accepted 7th March 2012
DOI: 10.1039/c2nj21038j
The low dielectric constant and high self-dissociation constant of water in a temperature range
between 150 and 250 1C make it a very appealing solvent for synthesis. Surprisingly, while
organic chemistry in water at low temperature or around its critical point has been investigated in
detail, very little seems to be known about the behaviour of organic molecules under
hydrothermal conditions. The present work thus aims at shading some light on this field.
As a start, we decided to investigate the reactions in which alcohols can undergo in water in the
above-mentioned temperature range. Knowing that very strong salt effects on organic reactions
have already been observed in super critical water, the impact of salt on the outcome of our tests
was also investigated in detail.
on this topic.3 Unfortunately, SCW is only obtained under
harsh (and thus at some point expensive) conditions. It also
imposes high requirements on the used reaction vessels. All
these features restrict the industrial applicability of SCW.
The intermediate range between ‘‘normal’’ and supercritical
water is often referred to as sub-critical, superheated (SHW), hot
compressed water (HCW) or hydrothermal water (HTW).4
Notwithstanding the advantages or drawbacks of these labels,
we decided, for convenience, to employ the terms ‘‘super-heated
water’’ and ‘‘hydrothermal conditions’’ equivalently in this work.
In the hydrothermal domain, the physicochemical properties of
water are evolving rapidly and minor temperature changes have
dramatic effects (as can be seen in Fig. 1 adapted from ref. 5).
Around 250 1C the dielectric constant of water is already
very low (which makes it a good solvent for apolar species), its
self-dissociation constant reaches a maximum, while the
corresponding autogenic pressure does not exceed 40 bars.
The corrosion behaviour of water at this temperature is known
and it is possible to deal with them industrially (see the working
conditions of numerous power plants).
A
Introduction
The search for greener solvents is one of the key issues for
achieving more sustainable chemical processes.1 This purpose
renewed a long time interest for water as a reaction medium
for organic synthesis. Actually, the properties of (liquid) water
are strongly impacted by temperature. This explains why
researchers differentiate between two domains of working
conditions, namely: (i) ‘‘standard’’ water at normal pressure
between 0 and 100 1C and (ii) supercritical water (SCW)
(the critical point of water lying at 374 1C and 218 bars).
Water under normal conditions, on the one hand, is well
known to most people and chemists, in particular. It is able
to solubilise numerous very polar species, hydrogen bond
donors or acceptors and most ionic species. What is also well
known to chemists is that a majority of organic compounds are
not soluble in water between 0 and 100 1C (which explains a
recurring controversy on the question whether organic reactions
featuring water as their solvent are actually occurring ‘‘in’’ water
or ‘‘in the presence of’’ water).2 Super critical water, on the other
hand, is a very good solvent for organic molecules (alkanes are,
for example, infinitely soluble in SCW) and could thus appear as
an ideal reaction medium for many organic reactions. Actually
numerous review articles are summarising the works undertaken
These working conditions are familiar to most material
scientists. Indeed, materials as important as zeolites,6 oxide
nanoparticles7 or mesoporous oxides8 are often synthesized
under hydrothermal conditions. More recently even hybrid or
purely organic materials have been successfully synthesised
under such conditions, as can be exemplified by metal organic
frameworks9 or hydrothermal carbon.10 Despite these achieve-
ments and the above-mentioned advantages, super-heated
water still is unfamiliar to organic chemists (except some
pioneering works by Kopetzki et al.11 and contributions from
the own group).12 The aim of the present work was to start
paving the way to a real use of hydrothermal conditions in
organic synthesis.
a Institute of Physical and Theoretical Chemistry,
University of Regensburg, Universitatsstr. 31, 93053 Regensburg,
¨
Germany. Fax: +49 941 943 4532
b Institut de Chimie Se´parative de Marcoule, UMR 5257,
`
Site de Marcoule BP 17171, 30207 Bagnols sur Ceze, France.
E-mail: frederic.goettmann@cea.fr
c Max-Planck-Institute of Colloids and Interfaces Scientific Campus
Golm, 14476 Potsdam, Germany. Fax: +49 331 567 9502
w Electronic supplementary information (ESI) available. See DOI:
10.1039/c2nj21038j
c
1568 New J. Chem., 2012, 36, 1568–1573
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Fig. 1 Evolution of the dielectric constant (e), the density (r) and
self-dissociation constant (kw) of water with temperature. Reprinted
from ref. 5 by permission of Turpion Ltd.
Scheme 1 Reaction of various alcohols in water at 180 1C.
(ii) In all cases, the tested alcohols underwent (at least in a
first step) a foreseeable dehydration reaction in moderate to high
yields. This is consistent with the fact that the self-dissociation
constant of water is already very high at 180 1C. Indeed, these
dehydration reactions can be promoted by protons, which is the
reason why they usually are run in strongly acidic media.16
(iii) In a second step, some of the dehydration products or
intermediates underwent further transformations. Pinacol 4
and meso-hydrobenzoin 5 for example yielded the expected
pinacolic rearrangements (4a and 5a in, respectively, 98 and
73%). 1-Phenylethan-1,2-diol yielded 1-phenylethanone
(which is a tautomer of the expected enolic dehydration
product) but part of it condensed to a dimer through an aldol
reaction. This opens the perspective of more complex multi-
step reactions in superheated water.
As most alcohols are hydrotropic molecules we thought that
it is convenient to start our investigation with this family of
molecules. In addition, biomass, which is very rich in hydroxy-
lated molecules, will, in a foreseeable future, more and more be
used as a feedstock for the chemical industry.13 But existing
chemical plants have been designed to deal with molecules
issuing from oil processing (such as alkanes and alkenes),
which have a very low content of heteroelements. It thus is
of great importance to find sustainable ways to convert
biomass derivatives into molecules, with which the chemical
industry can deal. We also decided, knowing that salts strongly
impact the properties of near critical water (see for example
ref. 14), to investigate how very simple salts impacted the
reactivity of our alcohols.
2
Screening of the impact of salts on the reactivity of various
B
Results and discussion
alcohols in super heated water
1
Screening of the reactivity of various alcohols in super
As previously mentioned, we also were interested in determining
how strongly the presence of salts would impact the outcome of
the reactions; we thus not only used pure (MiliQ) water as a
solvent but also a 1 M NaCl solution and a 1 M Na2SO4 solution.
Sodium chloride was selected because it probably is the most
abundant salt on earth (sea water being basically a 0.5 M NaCl
solution) and sodium sulfate because its impact on organic
reactions in water is already documented.11a The results of these
tests are displayed in Table 1. The previously described results in
pure water are also reported there, for comparison purpose.
Here again the obtained results lead to some observations:
(i) Compared to pure water, the addition of NaCl resulted
in increased conversions (The conversion of phenylethandiol
increased from 43 to 93%, see Table 1, entries 7 and 8), while
the addition of Na2SO4, in most cases, drastically reduced
the conversion rates (see again for phenylethandiol Table 1,
entry 9, a conversion reduced to 2%).
heated water
At first, we wanted to screen the behaviour of various (activated)
alcohols in aqueous solutions, just to get access to a mapping of
possible organic reactions in this medium. We thus heated up
5 mmol of phenylpropanol and 200 mg of some other alcohols in
10 ml of aqueous medium to 180 1C for 16 h under autogeneous
pressure. In preliminary runs, simple alcohols, like heptanol,
failed to undergo any reaction in high yields even at higher
temperatures (see Table S1 in the ESIw). We thus selected a range
of activated alcohols: 1-phenylethanol 1 as a secondary alcohol
(which, in addition, is activated by its aromatic ring), 1-methyl-
cyclohexan-1-ol alcohol 2 as a tertiary, phenylethan-1,2-diol 3 as
a simple diol, pinacol 4 and meso-hydrobenzoin 5 as two diols
known to undergo pinacolic rearrangements. The results of these
reactions are displayed in Scheme 1.
These results led us to the following observations:
(i) Despite the fact that most tested alcohols are sensitive to
oxidation, no detectable amount of oxidation product was
formed. This is a fundamental difference between superheated and
supercritical water, where radical reactions (and thus oxidations)
can be strongly favoured.15
(ii) In the cases where secondary reactions are observed,
these reactions are also promoted by the addition of
sodium chloride. In the case of phenylethandiol the previously
observed aldol condensation product is formed with 64%
yield. In a series of tests at 200 1C (see ESIw, Table S1) pinacol
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1568–1573 1569

View Article Online
Table 1 Comparative investigation of the effect of salt on the transformation of various alcohols under hydrothermal conditionsa
Selectivity for the
major productd (%)
Entry
Substrate
Solvent
Conversionb (%)
Product 1c (%)
Product 2c (%)
1
2
3
H2O
1 M NaCl
1 M Na2SO4
82
96
43
75
88
39
7
8
4
91
92
91
4
5
6
H2O
1 M NaCl
1 M Na2SO4
57
71
77
53
70
62
2
1
15
93
99
81
7
8
9
H2O
1 M NaCl
1 M Na2SO4
43
93
2
22
29
1
21
64
0
53
69
—
10
11
12
13
H2O
99
98
98
33
98
98
32
32
0
0
54
0
99
100
55
1 M NaCl
1 M NaCle
1 M N2SO4
97
14
15
16
17
H2O
81
100
100
8
73
89
96
7
8
11
4
90
89
96
1 M NaCl
1 M NaCle
1 M Na2SO4
1
—
a
For these reactions, 5 mmol of phenyl ethanol or 200 mg of the desired alcohol were added to 10 ml of solvent and heated to 180 1C for 16 h.
Conversions were determined by GC-FID with an internal standard as the ratio between consumed and initial amounts of substrate. The
b
c
product yields were also determined by GC-FID as the molar ratio between the detected amount of product and the initial amount of substrate.
d
Selectivities were calculated as the ratio between the formed amount of product and the consumed amount of substrate. e Reactions run at 200 1C.
3
Investigation of salt effects on the dehydration of
1-phenylpropan-1-ol
The above-described promoting, respectively, blocking effects
of NaCl and Na2SO4 are relatively surprising. We thus decided
to study these salts effects in more detail on one model reaction:
the dehydration of 1-phenylpropan-1-ol (Scheme 1A).
Scheme 2 Mechanism of the dehydration/Diels–Alder reaction of
pinacol.
not only yielded the rearrangement product, but also a more
complex Diels–Alder condensation product displayed in
Scheme 2.
3.1 The effect of sodium chloride. First we investigated the
influence of NaCl concentration on this reaction at 180 1C for
16 hours. The results are displayed in Fig. 2. As previously
observed, pure super heated water is able to give a reasonable
conversion, but that conversion is improved by adding even
an ‘‘inert’’ salt. At moderate salt concentrations an increase of
the conversion was seen, which leveled into a plateau, and then
(iii) The selectivities remained good (i.e. a limited number of
products was obtained). It is worth noticing that no inter-
molecular dehydration (be it a Friedel–Crafts reaction or the
formation of an ether) could be detected even at higher alcohol
concentration (see Table S2 in the ESIw).
c
1570 New J. Chem., 2012, 36, 1568–1573
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
Fig. 4 Time dependence of the 1-phenylpropene yield in a 1 M NaCl
solution at 180 1C.
Fig. 2 Influence of the NaCl concentration on the dehydration of
1-phenyl-1-propanol at 180 1C after 16 hours.
Fig. 3 Dependence of the water dissociation constant on temperature
Fig. 5 Trans-phenylpropene yield after 16 hours as a function of the
and NaCl concentration. Adapted from ref. 18.
temperature of the reaction and the NaCl concentration.
dropped again. Torry et al. observed a similar phenomenon
when investigating the rate of hydrolysis of dibenzyl ether and
benzyl phenyl amine in near-critical aqueous NaCl.17 The
initial acceleration of the reaction was assigned to an initial
increase of the polarity of the solvent favouring the polar
transition state of the reaction. This may apply to our case.
The authors attributed the decrease of the reaction rate at
higher salt concentration to a decrease in water availability
for the reaction. This may be reasonable in their case, but
seems less plausible in ours (where on the contrary water is
generated). Another, more plausible, interpretation draws on
the in situ proton concentration. According to Busey and
Mesmer,18 the water dissociation constant not only varies
with temperature but also with NaCl concentration (Fig. 3),
showing a broad maximum between 1 M (molar) and 3 M
ionic strength (Àlog Kw reaching 10.8 at 1 M and 175 1C). This
provides an increased proton concentration in this range and,
thus, an accelerated dehydration reaction.
Fig. 6 Molar fraction of trans-phenylpropene after dehydration at
1 M chloride concentration and 180 1C reaction temperature, with
variations of the cation.
In a second step, we tested the time dependence of the yield of
trans-phenylpropene at 180 1C at constant salt concentration.
The results we obtained are displayed in Fig. 4. As can be seen,
the equilibrium is almost attained after 16 h.
is smoothed out. At intermediate temperatures a strong yield
increase is observed which is consistent with the data reported
in Table 1.
The influence of temperature on the outcome of the reaction
was also tested at different salt concentrations. Fig. 5 shows
the yield of 1-phenylpropene after 16 hours as a function of
temperature and salt concentration. At 200 1C the reaction
proceeds so fast that no effect of the salt could be observed. On
the opposite at 140 1C the reaction is so slow that the salt effect
3.2 Cation variation. In the following series of tests, we
intended to investigate the action of a broad range of salts on
the dehydration of 1-phenylpropan-1-ol (still at 1 M anion
concentration, 180 1C for 16 h). Fig. 6 displays the results
obtained when chloride was kept as an anion and the cations
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1568–1573 1571

View Article Online
driving force of the reaction. In contrast, ordering the anions
according to their basicity (via the pKa of the corresponding
acids, Fig. 8) seems to be more conclusive. All anions, which are
the conjugated base of strong acids (like HCl, HNO3 or H2SO4),
favoured the dehydration reaction, while the anions, which are
the conjugated base of weak acids, slowed the dehydration
reaction down. This leads us again to the conclusion that the
amount of free protons in the aqueous medium at 180 1C is the
main reason for changes in the trans-phenylpropene yield.
Anyhow, we cannot exclude additional chemical effects of the
anions on the reaction as the data still show minor variations
within the group of acidic and basic anions.
C
Conclusion
Fig. 7 Molar fraction of trans-phenylpropene after dehydration at
The aim of this paper was to investigate both the reactivity of
alcohols under hydrothermal conditions and the impact of
simple salts on this reactivity. The first screening test shows
that alcohols and their reaction products can undergo a broad
range of reactions including:
1 M salt concentration, with variations of the anion.
were varied. As can be seen, cation variation had only little
impact on the 1-phenylpropene yield (between 85 and 95% of
alkenes are obtained as long as chlorine is the anion), at least
for the examples presented here. This is somewhat unusual as
catalytic behaviour is very often ascribed to the cations and
more specifically to metals and their incomplete coordination
sphere. But in our case, as the observed reactions of our
alcohols are usually proton catalysed, these observations just
indicate that the employed cations neither directly interfere in
the reactions nor impact the in situ proton concentration.
– dehydration
– tautomerism
– pinacol rearrangement
– dehydrative ringclosure (see Table S1, ESIw, norborneol)
– aldol condensation
– Diels–Alder reactions
While the reactions directly related to dehydration were not
big surprises, the C–C bond forming reactions were unexpected
and are of great synthetic interest. Interestingly, despite the
numerous possible reactions under hydrothermal conditions,
the chemoselectivity of our tests remained relatively high. These
preliminary results thus open the path for further investigations
of specific reactions in superheated water.
3.3 Anion variation. Fig. 7 displays the results obtained
when sodium was used as the cation and the anions were
varied. As was to be expected, drawing on the results reported
in Table 1, the nature of the anions had a pronounced impact
on the phenylpropene yield. Interestingly, anions classically
assumed to feature some activity (e.g. fluoride, phosphate,
trifluoroacetate) show almost no conversion, while the highest
conversion is found for nitrate. The observations at hand
allow us to exclude that the anion effects on the reaction rate
can be ordered according to a classical Hofmeister series,19
i.e. the influence of the anion on water structuration is not the
The screening of the impact of simple salts on the outcome
of a model dehydration reaction shows that some families of
salts promoted the tested organic reactions in the medium
while other slowed them down. The exact mechanism of these
effects is still to be understood. A putative interpretation of
our results can be that all the tested salts had a positive impact
on water dissociation at high temperature (as evidenced by
Busey and Mesmer on NaCl)18 but that anions corresponding
to weak acids were partially protonated in the medium thus
reducing the availability of protons in the medium, and
slowing down the reactions. Such an interpretation has two
main consequences:
– It must be possible to control the outcome and/or selectivity
of organic reactions in superheated water just by selecting the
right simple salt (which could be of interest to tune the outcome
of hydrothermal carbonisation reactions).
– It should be also possible to promote OHÀ catalysed reactions
by using the above-mentioned blocking salts. Unfortunately, very
few organic reactions are catalysed by bases only and, with the
exception of the above-mentioned works of Kopetzki et al., when
such reactions are at hand, their reactants are often sensible to
hydrolysis (consider the typical base catalysed example of the
Knoevenagel condensation of benzaldehyde with malononitrile:
the first reactant can undergo a Cannizzaro reaction and the nitrile
group can hydrolyse to carboxylic acids).
Fig. 8 Conversion ratio of 1-phenyl-1-propanol to trans-phenyl-
propene vs. basicity of the added salt anion.
Both aspects are, thus, still under investigation.
c
1572 New J. Chem., 2012, 36, 1568–1573
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012

View Article Online
7 (a) M. Henry, J. P. Jolivet and J. Livage, Struct. Bonding, 1992,
77, 153; (b) B. L. Cushing, V. L. Kolesnichenko and
C. J. O’Connor, Chem. Rev., 2004, 104, 3893.
8 (a) A. Taguchi and F. Schuth, Microporous Mesoporous Mater.,
2005, 77, 1; (b) A. Corma, Chem. Rev., 1997, 97, 2373.
9 (a) G. Ferey, Chem. Mater., 2001, 13, 3084; (b) G. Ferey, Chem.
Soc. Rev., 2008, 37, 191.
10 M. M. Titirici, A. Thomas and M. Antonietti, Adv. Funct. Mater.,
2007, 17, 1010.
11 (a) D. Kopetzki and M. Antonietti, Green Chem., 2010, 12, 656;
(b) D. Kopetzki and M. Antonietti, New J. Chem., 2011, 35, 1787.
12 B. Smutek, W. Kunz and F. Goettmann, C. R. Chim., 2012,
15, 96.
Experimental section
The experiments were done in stainless steel autoclaves with PTFE
inlets and an inner volume of 21 mL. The pressure during the
reaction is the autogeneous pressure of the solution at reaction
temperature. Reaction mixtures consisted of 10 mL Millipore water,
an inorganic salt, and 5 mmol 1-phenylpropanol, or 200 mg of a
different alcohol. After sixteen hours standard reaction time the
autoclaves were cooled down to room temperature, and then the
reaction mixture was extracted three times with 5 mL diethyl ether
each time. From the extracts GC-MS and GC-FID analysis were
done, usually with toluene as internal standard. For the test reaction
the response factor was determined by reference standards,
subjected to the same extraction procedure. FID peak areas during
the alcohol screening were evaluated using the response factor
prediction model from Jorgensen, refined by Kallai.20
13 (a) J. N. Chelda, G. W. Huber and J. A. Dumesic, Angew. Chem.,
Int. Ed., 2007, 46, 7164; (b) A. Corma, S. Iborra and A. Velty,
Chem. Rev., 2007, 107, 2411; (c) P. Maki-Arvela, B. Holmbom,
¨
T. Salmi and D. Y. Murzin, Catal. Rev. Sci. Eng., 2007, 49, 197;
(d) T. Werpy and G. Petersen, Top Value Added Chemicals from
Biomass: Volume 1—Results of Screening for Potential Candidates
from Sugars and Synthesis Gas, National Renewable Energy
Laboratory, 2004.
14 (a) L. P. M. De Leo, H. L. Bianchi and R. Fernandez-Prini,
J. Chem. Thermodyn., 2005, 37, 499; (b) A. G. Dickson,
D. J. Wesolowski, D. A. Palmer and R. E. Mesmer, J. Phys.
Chem., 1990, 94, 7978.
15 M. J. Antal, A. Brittain, C. Dealmeida, S. Ramayya and J. C. Roy,
ACS Symp. Ser., 1987, 329, 77.
Notes and references
1 M. Poliakoff, J. M. Fitzpatrick, T. R. Farren and P. T. Anastas,
Science, 2002, 297, 807.
2 Y. Hayashi, Angew. Chem., 2006, 118, 8281.
3 (a) P. E. Savage, Chem. Rev., 1999, 99, 603; (b) A. R. Katritzky,
D. A. Nichols, M. Siskin, R. Murugan and M. Balasubramanian,
Chem. Rev., 2001, 101, 837; (c) M. Siskin and R. Kaminsky, Chem.
Rev., 2001, 101, 825; (d) M. Wanatabe, T. Sato, H. Inomata,
R. L. Smith, K. Arai, A. Ktuse and E. Dinjus, Chem. Rev., 2004,
104, 5803; (e) S. A. Nolen, C. L. Liotta, C. A. Eckert and
16 A. Chenede, N. AbdRahman and I. Fleming, Tetrahedron Lett.,
1997, 38, 2381.
17 J. Penninger, R. Kersten and H. Baur, J. Supercrit. Fluids, 2000,
17, 215.
18 R. Busey and R. Mesmer, J. Chem. Eng. Data, 1978, 23, 175.
19 (a) M. Chaplin, Water Structure and Science, http://www1.lsbu.ac.uk/
water, accessed 06.12.2009; (b) W. Kunz, J. Henle and B. Ninham,
Curr. Opin. Colloid Interface Sci., 2004, 9, 19; (c) H. Wang, J. Wang,
S. Zhang and X. Xuan, J. Phys. Chem. B, 2008, 9, 19.
20 (a) A. D. Jorgensen, K. C. Picel and V. C. Stamoudis, Anal. Chem.,
1990, 62, 683; (b) M. Kallai, Z. Veres and J. Balla, Chromatographia,
2001, 54, 511.
R. Glaser, Green Chem., 2003, 5, 663; (f) A. Kruse and
¨
E. Dinjus, J. Supercrit. Fluids, 2007, 39, 362.
4 P. E. Savage, J. Supercrit. Fluids, 2009, 47, 407.
5 A. A. Galkin and V. V. Lunin, Usp. Khim., 2005, 74, 24.
6 (a) C. S. Cundy and P. A. Cox, Chem. Rev., 2003, 103, 663; (b) C. S.
Cundy and P. A. Cox, Microporous Mesoporous Mater., 2005, 82, 1.
c
This journal is The Royal Society of Chemistry and the Centre National de la Recherche Scientifique 2012
New J. Chem., 2012, 36, 1568–1573 1573