Energy efficiency in chemical reactions: A comparative study of different reaction techniques

Article abstract of DOI:10.1021/op0498060

Metrics for chemists are much argued, but the energy usage of a reaction has to be one of the most simple and effective methods of comparing technologies. In this concept paper, the energy consumed in preparing one mole of a chemical compound is compared for a variety of technologies. Data are gathered for traditional oil bath, supercritical CO₂ (sc-CO ₂), and microwave reactors. Two different Suzuki couplings, a Knoevenagel condensation, and a Friedel-Crafts acylation are all compared in both the microwave and oil bath, as this is where the most noticeable differences were observed. The most notable result was an 85-fold reduction in energy demand on switching from oil bath to microwave reactor for a Suzuki reaction.

Full text of DOI:10.1021/op0498060

Organic Process Research & Development 2005, 9, 516−518
Technical Notes
Energy Efficiency in Chemical Reactions: A Comparative Study of Different
Reaction Techniques
Mark J. Gronnow, Robin J. White, James H. Clark,* and Duncan J. Macquarrie
Clean Technology Centre, Department of Chemistry, UniVersity of York, Heslington, York, YO10 5DD UK
Scheme 1. Suzuki coupling of phenyl boronic acid and
bromotoluene
Abstract:
Metrics for chemists are much argued, but the energy usage of
a reaction has to be one of the most simple and effective methods
of comparing technologies. In this concept paper, the energy
consumed in preparing one mole of a chemical compound is
compared for a variety of technologies. Data are gathered for
traditional oil bath, supercritical CO2 (sc-CO2), and microwave
reactors. Two different Suzuki couplings, a Knoevenagel
condensation, and a Friedel-Crafts acylation are all compared
in both the microwave and oil bath, as this is where the most
noticeable differences were observed. The most notable result
was an 85-fold reduction in energy demand on switching from
oil bath to microwave reactor for a Suzuki reaction.
supply and reactor to measure the energy consumed during
the course of the reaction. To normalise the data the energy
has been calculated on the basis of moles of product formed.
The experimental methods are shown in the footnotes.
Assumptions are made to maintain the simple nature of the
calculations, and no chemical or gas formation costs are
included. As product mixtures were similar, the energy
consumed in a work-up is not included as it would be
identical for all crude mixtures (e.g., column then evaporation
of solvent).
Introduction
The subject of metrics is becoming increasingly important
in green chemistry. Numerous papers have been published
concerning the measurement of “greenness” within reac-
tions,1 however the majority of these seem to be only
concerned with the efficiency of the chemistry of the reaction.
Life cycle assessment teaches us that energy consumption
should also be considered since it consumes (mainly
nonrenewable) resources and produces waste. Escalating
energy costs have also made energy efficiency an equally
important metric in determining environmental acceptability
and economic viability. Microwave-assisted chemical reac-
tions have frequently been reported to show increased
product yield, increased selectivity, and extremely short
reaction times often in the absence of solvent.2 These would
appear to indicate good metrics, but the energy efficiency
of such processes is not clear and rarely discussed. In an
attempt to encourage the regular use of energy efficiency in
process metrics we have compared the use of three different
currently popular reaction techniques to carry out important
organic reactions.
The homogeneous palladium-catalysed Suzuki coupling
is currently a very popular reaction with the pharmaceutical
and specialty chemical industries; in this example phenyl
boronic acid and 4-bromotoluene are coupled to form the
biaryl (Scheme 1). These reaction conditions4 were selected
as they are proven to work under both sc-CO2 and conven-
tional solvent conditions. Further reactions were studied with
attempts made to cover a range of different chemistries. The
second Suzuki coupling was selected as it is a traditionally
low-yielding chloro-coupling which has been found to have
excellent yields in a microwave reactor. The aluminium
chloride catalysed Friedel-Crafts acetylation has been
studied as an example of chemistry carried out below the
“bubble point”. A Knoevenagel reaction was studied as an
example of a base-catalysed reaction; furthermore, this
methodology involved using a heterogeneous catalyst which
the chemical industry is beginning to investigate as an
alternative to homogeneous catalyst as it is simple to remove
and recycle.
A system has been devised using a domestic six-dial
electricity meter in series between the laboratory power
Results and Discussion
Table 1 shows the results obtained for the Suzuki coupling
reactions between phenyl boronic acid and bromobenzene
carried out under the different reaction conditions.
* To whom correspondence should be addressed. E-mail: jhc1@york.ac.uk.
Telephone: +441904432559.
(1) For a review, see: Constable, D. J. C.; Curzons, A. D.; Freitas dos Santos,
L. M.; Geen, G. R.; Hannah, R. E.; Hayler, J. D.; Kitteringham, J.; McGuire,
M. A.; Richardson, J. E.; Smith, P.; Webb, R. L.; Yu, M. Green Chem.
2001, 3, 7-9.
(3) Oakes, R. S.; Clifford, A. A.; Rayner, C. M. J. Chem. Soc., Perkin Trans.
1 2001, 917.
(4) Raynel, G.; Rayner, C. M. First Year Progress Report; University Of Leeds,
(2) Leadbeater, N. Chem. World 2004, 38.
2002.
516
•
Vol. 9, No. 4, 2005 / Organic Process Research & Development
10.1021/op0498060 CCC: $30.25 © 2005 American Chemical Society
Published on Web 05/24/2005

Table 1. Table of results
Scheme 4. Knoevenagel reaction
media
yield (%)a
energy (kW/mol)b
oil-bath dioxane
oil-bath toluene
microwave dioxane
microwave toluene
Sc-CO2 neat
56.6
25.1
40.7
5830
13541
1680
chemical manufacturing processes. The energy efficiency of
the microwave method is again higher than the thermal
reaction although this time the difference is less due to the
lower temperature and short reaction times required.
The heterogeneous aminopropyl silica-catalysed Knoev-
enagel reaction was selected as a base-catalysed reaction for
comparison (Scheme 4/Table 2). When powered in the
microwave, the yield could not be pushed above 18.6%; this
is probably partially due to water being formed in the reaction
system. Under thermal conditions a Dean-Stark water-
collecting system was employed to azeotropically remove
the water produced during the reaction, since according to
the literature this boosts the yield.6 Additionally, the reaction
is conducted in cyclohexane which is not a good microwave
absorbing solvent; hence, these conditions reduced the rate
of reaction during microwave heating. The good yield along
with the low boiling point of the solvent and short reaction
time make the Knoevenagel thermal reaction relatively low
in terms of energy usage.
0
n/a
11.5 (80c)
21840 (3360)
a GC yield from response factors. b Includes theoretical calculation for removal
of solvent. c Literature yield.
Scheme 2. Suzuki coupling of cyclopentyloxybenzene
boronic acid and 2-chloro-5-nitropyridine using a
heterogeneous catalyst
Table 2. Table of Results
thermal yielda/ microwave yielda/
reaction
energyb
energyb
The results indicate a significant gain in energy efficiency
using microwave activation. Microwave activation provides
highly focused energy, enabling rapid reaction and more
efficient energy consumption. This is in addition to any actual
microwave effect causing specific bond activation. We did
use typical laboratory equipment in terms of type and size,
and oil-bath reactions were not lagged. With low-boiling
solvents and comparatively short reaction times the oil bath
becomes more energy efficient. Continuation of this study
into flow systems versus batch and semi-batch processes
would be interesting; continuous processing should increase
the energy efficiency of the system further. Application of
this methodology to other systems would allow for quanti-
fication of the energy savings offered by process intensifica-
tion. The supercritical reactors shared similar energy effi-
ciency to oil bathsswhen the work-up of the product is
considered, large energy savings may be possible using the
supercritical method.
heterogeneous Suzukic
Friedel-Crafts acylationc
Knoevenageld
28.6/120700
100/13.6
68.2/132
77.4/1448
100/9.7
18.2/37
a % yield. b kW/mol. c GC yield. d GC yield corrected by Rf.
Scheme 3. Friedel-Crafts acylation of methyl salicylate
The supercritical reaction has been reported to give 80%
yield,4 which would equate to an energy usage of around
3360 kW/mol. This is still considerably higher than the
microwave reaction but is twice as energy efficient as the
traditional thermal method.
In the next reaction selected for the microwave (Scheme
2) the Suzuki coupling of 2-chloro-5-nitropyridine with the
boronic acid over Si Cat1,5 a 77% yield was obtained in 4
min which corresponds to an energy usage of 1500 kW/mols
nearly 2 orders of magnitude less than that of the oil bath
(Table 2).7 The low yields under the conventional thermal
conditions make the energy consumption per gram extremely
high compared to that calculated in the microwave. However,
even taking this into account, there is a huge energy saving
in using microwaves.
We hope that this preliminary study will encourage others
to develop this concept and report other and more thorough
calculations of the energy demands of reactions carried out
under fundamentally different conditions. In this way we can
seek to move towards a more proper consideration of energy
consumption in green chemical process metrics.8
Experimental Methods
Suzuki 1. Palladium (II) trifluoroacetate (5.00 mol %),
bis(diphenylphosphino)ferrocene (dppf) (10.0 mol %), DAB-
CO/K2CO3 (1.49 mol equiv), benzeneboronic acid (1.10 mol
equiv), and bromobenzene (1.00 mol equiv) were added to
an appropriate vessel. For thermal and microwave reactions
toluene or 1,4-dioxane were used as solvents.4 Microwave
conditions were 300 W for 1 h with simultaneous air-cooling
The aluminium chloride-catalysed Friedel-Crafts acety-
lation of methyl salicylate was studied (Scheme 3/Table 2)
as another example of a reaction highly important in many
(5) Mubofu, E. B.; Clark, J. H.; Macquarrie, D. J. Green Chem. 2001, 3, 23.
(6) Macquarrie, D. J.; Clark, J. H.; Lambert, A.; Mdoe, J. E. G.; Priest, A.
React. Funct. Polym. 1997, 35, 153.
(8) Clark, J. H.; Green Chemistry Network Newsletter; March 2005 (http://
(7) White, R. J. MRes Project Report; University Of York, 2004.
www.chemsoc.org/gcn).
Vol. 9, No. 4, 2005 / Organic Process Research & Development
•
517

under autogenous pressure (maximum temperature 180 °C).
Thermal reactions were carried out under reflux conditions
at atmospheric pressure for 24 h. The supercritical rig was
pressurised to 600 psi at room temperature, heated to 100
°C and pressurised to 2500 psi.
Suzuki 2. 2-Chloro-5-nitropyridine (1 mol equiv), boronic
acid (1.1 mol equiv), K2CO3 (2 mol equiv), Si-Cat15 (0.05
mol equiv Pd) in xylene were mixed in an appropriate vessel
and stirred at 120 °C for 8 h at atmospheric pressure or
microwaved at 300 W 4 min.
Friedel-Crafts Acylation. Methyl salicylate (1 mol
equiv), acetic anhydride (1.1 mol equiv), aluminium chloride
(3 mol equiv) were stirred in nitrobenzene in an appropriate
vessel at 60 °C for 4 h at atmospheric pressure or with a
microwave programmed to ramp for 2 min to 120 °C, then
ramp to 180 °C over 2 min, then hold for 4 min under
autogenous pressure.
Knoevenagel. As by Macquarrie et al., see ref 6.6
Hardware. Heidolf MB 3003 stirrer hotplate, CEM
Discover microwave reactor, supercritical extraction rig SFT-
150 SFE system with Julabo F12 cooler, and a Varian 3800
GC with autosampler. No computer control systems were
included in the energy calculation, only the reaction hard-
ware. On the supercritical extraction rig the cooling system
was only switched on when pressurising the vessel to
minimise energy wastage. No energy was added for the
compression of the CO2 into the cylinder and compressed
air used for simultaneous cooling of the microwave.
Acknowledgment
We thank GSK for support for this project.
Received for review November 1, 2004.
OP0498060
518
•
Vol. 9, No. 4, 2005 / Organic Process Research & Development