Decision support towards agile eco-design of microreaction processes by accompanying (simplified) life cycle assessment

Article abstract of DOI:10.1039/c1gc15054e

Continuously running syntheses in microstructured reactors offers novel ways to intensify conventional chemical processes. An outstanding advantage of microreaction technology is the high surface-to-volume-ratio which enables intensive mixing phenomena as well as high mass and heat transfer rates. Thus, microstructured reactors may be a suitable means to improve multiphase reactions by increasing the interfacial area and the intensification of internal mixing. This improvement in reaction performances may lead to reduced environmental burdens of the process under consideration. The method of simplified life cycle assessment (SLCA) is a suitable tool to evaluate the environmental burdens caused by chemical processes. It has been applied already in research and development to identify the key parameters for a deliberate green process design of two biphasic reactions, the esterification of phenol and benzoyl chloride resulting in phenyl benzoate and the synthesis of one of the corresponding phase transfer catalysts, [BMIM]Cl. Further, SLCA is complemented by a simple cost estimation to investigate the main cost drivers relevant for possible industrial application of the syntheses investigated.

Full text of DOI:10.1039/c1gc15054e

View Article Online / Journal Homepage / Table of Contents for this issue
Dynamic Article Links
Green Chemistry
Cite this: Green Chem., 2011, 13, 1694
www.rsc.org/greenchem
PAPER
Decision support towards agile eco-design of microreaction processes by
accompanying (simplified) life cycle assessment
Sabine Huebschmann,a Dana Kralisch,*a Holger Loewe,b,c Denis Breuch,b Jan Hauke Petersen,b
Thomas Dietrichd and Ralf Scholzd
Received 14th January 2011, Accepted 29th March 2011
DOI: 10.1039/c1gc15054e
Continuously running syntheses in microstructured reactors offers novel ways to intensify
conventional chemical processes. An outstanding advantage of microreaction technology is the
high surface-to-volume-ratio which enables intensive mixing phenomena as well as high mass and
heat transfer rates. Thus, microstructured reactors may be a suitable means to improve multiphase
reactions by increasing the interfacial area and the intensification of internal mixing. This
improvement in reaction performances may lead to reduced environmental burdens of the process
under consideration. The method of simplified life cycle assessment (SLCA) is a suitable tool to
evaluate the environmental burdens caused by chemical processes. It has been applied already in
research and development to identify the key parameters for a deliberate green process design of
two biphasic reactions, the esterification of phenol and benzoyl chloride resulting in phenyl
benzoate and the synthesis of one of the corresponding phase transfer catalysts, [BMIM]Cl.
Further, SLCA is complemented by a simple cost estimation to investigate the main cost drivers
relevant for possible industrial application of the syntheses investigated.
Due to there inherent large surface-to-volume ratio, which
Introduction
may lead to intensive mixing as well as to very efficient
The need to create more sustainable processes, to use raw
mass and heat transfers, microstructured reactors are gaining
materials more efficiently and to save energy as well as money
increasing interest not only in academic but also in industrial
has become an important issue in the chemical industry. Several
applications. Further, hazardous and toxic substances can be
concepts have been developed to reduce the environmental
burdens of chemical processes, e.g., by implementing catalysts,
substituting petrochemical substances by renewable feedstocks,
handled more safely because of the small inner volumes of the
reaction channels. A high flexibility is obtained by the modular
assembling of the microreactor plant including pumps and
employing alternative energy sources or implementing novel
controlling instruments.4 The residence times can be precisely
reaction media. Another possibility to obtain more sustainable
adapted to the reaction under investigation by controlling the
processes is the intensification of the process itself which can
flow rates to avoid side and subsequent reactions.5
be achieved, e.g., by the transmission of conventional batch
processes to continuously running processes in microstructured
Phase transfer catalyses are often difficult to handle in con-
ventional reactor equipment, because these reactions are mostly
reactors while maintaining or improving the reaction yield
mass transfer controlled, requiring a high and reproducible
and/or selectivity.1–3
interfacial area. In microstructured reactors, an increased and
well-defined interfacial area is achieved by realizing a segmented
flow.6,7 Further, the internal mixing of reactants can be increased
aInstitute of Technical Chemistry and Environmental Chemistry,
by a circular flow within the single plugs of the dispersed phase. A
dispersion as well as a coalescence of both immiscible fluids can
be minimized by regular-sized individual plugs in a continuous
phase.8,9
However, there may also be some difficulties in handling
multiphase reactions in microstructured reactors. For example,
the flow rates have to be precisely adapted to the system under
investigation. This also includes the flow rate ratios of the
process fluids in order to avoid parallel flow of both fluids,
which decreases the interfacial area and the mass transfer.7
Friedrich-Schiller-University Jena, D-07443, Jena, Germany.
E-mail: dana.kralisch@uni-jena.de; Fax: +49-3641-9484-02;
Tel: +49-3641-9484-57
bInstitute of Organic Chemistry, Johannes Gutenberg University Mainz,
Duisbergweg 10-14, D-55128, Mainz, Germany.
E-mail: loewe@uni-mainz.de; Fax: +49-6131-392-6677;
Tel: +49-6131-392-6667
cInstitut fuer Mikrotechnik Mainz GmbH, Carl-Zeiss-Str. 18-20,
D-55129, Mainz, Germany
dmikroglas chemtech GmbH, Galileo-Galilei-Str. 28, D-55129, Mainz,
Germany. E-mail: T.Dietrich@mikroglas.com;
Fax: +49-6131-55550-52; Tel: +49-6131-55550-50
1694 | Green Chem., 2011, 13, 1694–1707
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Furthermore, even small fabrication defects may lead to a
completely different flow behaviour.10 Since continuously op-
erating devices for the separation of two liquid–liquid phases
are still under development, there may also be some difficulties
in separating both phases after the reaction as well as in the
isolation of the product.11–14
method by Kolleret al.32 or an algorithm proposed by Sugiyama
et al.33 starting with an estimation based on stoichiometry and
becoming more detailed with every iteration step. The latter
approach introduced also the ELI (energy loss index) as a new
indicator for estimating energy costs in early process design,
which was further refined by Bumann et al.34 Some authors
have also suggested the combined evaluation of environmental
burdens and process costs during R&D, e.g., the ECO method
proposed by Kralischet al.,25 as well as a combination of the
environmental indicator MIPS and an economic parameter
TAPPS suggested by Hoffmann et al.35
Nevertheless, the continuous processing of phase transfer
catalysis in microreactors bears the potential for significant yield
enhancements and reductions in energy requirements which
may lead to more sustainable chemical processes compared
to conventional reaction equipment. This has been proven in
the examples of the phase-transfer catalysed phenyl benzoate
synthesis starting from phenol and benzoyl chloride. Several
ionic liquids have been tested as phase transfer catalysts for this
reaction. Thus, they were included in the comparative evaluation
of the resulting environmental and cost impacts. Exemplarily, the
synthesis of the ionic liquid [BMIM]Cl from 1-methylimidazole
and butyl chloride will be discussed in more detail, since the
benefits of microprocessing for biphasic reactions have become
visible here as well. In both cases, an accompanying simplified
life cycle assessment (SLCA) and cost estimation approach gave
the direction for a deliberate design of a green chemical process.
In the past decades, several metrics and concepts have been
developed to evaluate the environmental burdens of a product
or process, e.g., the E-factor by Sheldon,15 the concept of atom
economy by Trost,16 EATOS by Eissen and Metzger17 as well as
E-Green by Hossain et al.18 However, the most comprehensive
method is the life cycle assessment (LCA) standardized in the
ISO norms 1404019 and 14044.20 According to these norms, the
whole life cycle of a product or process from the extraction
of raw materials, the manufacture or synthesis of the product
until the use and disposal are to be investigated. In the first
step, the boundaries of the system under investigation and the
functional unit are defined. Afterwards, all mass and energy
flows within the system boundary are comprehended referring
to the functional unit (life cycle inventory analysis – LCI) and
the environmental impacts of these mass and energy flows are
assessed (life cycle impact assessment – LCIA). In the last step,
all results are summed up and interpreted to derive potential
improvements or to support decisions between different options.
Usually, LCAs are performed at the end of process de-
velopment because detailed information of the process under
investigation is required according to the ISO norms. This
information is often not sufficiently available during research
and development (R&D) stages. Data gaps and uncertainties are
typical problems during this stage of process design and impede
a sufficient evaluation. Further, R&D stages are regarded as
making only a minor contribution to the overall environmental
impact of a process.21 But environmental burdens of a pro-
cess become mostly predefined by the decisions made during
R&D and high efforts are required to reduce these impacts
subsequently.22–24 Thus, several authors25–27 have already pointed
out that evaluation methods most beneficially are applied during
R&D. To address the environmental burdens in these early stages
of process development, qualitative criteria have been proposed
by Anastas and Warner,28 Anastas and Zimmerman,29 Biwer and
Heinzle,22 as well as by Graedel.30 Several quantitative metrics
have also been developed going into further detail, e.g., euroMat
by Fleischer and Schmidt,24 the MIPS concept,31 the EHS
We have found that a simplified LCA (SLCA) provides a
suitable compromise between single metrics and a holistic life-
cycle-based analysis, which provides an insight into the environ-
mental impacts of a chemical process under development and for
decision support between several alternative reaction pathways.
SLCA can either comprise simple input/output analyses or
hybrid analyses, but may also consist of a horizontally limited
approach by neglecting some life cycle stages. But, a horizontally
limited approach may lead to a problem shift to the life cycle
stages not taken into consideration.21 Because of this, vertically
bounded SLCA, which consider all life cycle stages but in
a superficial manner, are more recommended. These SLCA
approaches are regarded as suitable for screening purposes36
as they are less cost- and time-consuming than detailed LCAs.37
In the following, a research accompanying SLCA is described.
It has been performed in order to evaluate the environmental
burdens of the continuously running phase transfer catalysis
of the biphasic esterification of benzoyl chloride and phenol
resulting in phenyl benzoate in microstructured reactors, which
already starts in research and development stages. We also
applied this method to the biphasic synthesis of the ionic liquid
[BMIM]Cl in laboratory scale carried out both in batch as
well as in continuously running syntheses. Further, the costs
of the syntheses in laboratory scale were estimated to predict the
economic competitiveness of the reactions under investigation
compared to the actual market price. The main cost drivers of
the syntheses and thus the key figures for process improvement
could be identified by this approach.
Evaluation approach
In this section, the applied methods for performing the SLCA
as well as for cost estimation are described. The strategy for
research accompanying environmental as well as economic as-
sessments are illustrated in Fig. 1. As we had already successfully
applied it in other case studies, we started our investigations with
SLCA during R&D, when the knowledge about the process was
still very low.
During process design there are vast amounts of process
alternatives to choose from. Thus, scenario analyses have been
performed to support the decisions regarding these alterna-
tives. Additionally, sensitivity analyses were used to identify
parameters with major influences on the environmental impact
of the process. At the end of process development, when the
system is known in detail, a holistic LCA will be performed
to validate the decisions taken. During all process development
stages, the costs were also considered, again starting already
at the laboratory scale. The results were adapted according to
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1694–1707 | 1695
©

View Article Online
as sensitivity analyses were performed by varying significant
parameters, e.g., the amount of solvents used, the reaction yield,
the type of micromixer or the flow rates.
We chose Umberto38 as the software tool for process mod-
elling during life cycle inventory (LCI) analysis and for the
implementation of the life cycle impact assessment (LCIA).
The Ecoinvent39 database was used to account for the up-
stream processes of the substances involved as well as the
generically calculated supply of electrical energy. The environ-
mental burdens of the process alternatives were assessed by
means of the cumulative energy demand (CED)40 as well as
the life cycle impact categories proposed by CML.41 By this, the
environmental burdens of the up-stream processes of the supply
of the substances involved in the syntheses, the syntheses itself
and the recycling of the organic solvents were investigated.
As the experimental work in the phase transfer catalytic syn-
thesis of phenyl benzoate is still going on, no work-up procedures
have been applied yet. Nevertheless, we made some prognostic
calculations to guide further process design, since the work-up
procedures may have a significant influence on the overall SLCA
result. In many cases, the efficiency of the work-up procedure
decides the superiority of one process alternative against others.
In the synthesis of [BMIM]Cl, two work-up procedures were
established depending on the processing mode, which were fully
included in the SLCA investigations as well as in the cost
estimation. However, the provision of the reaction periphery is
not included in the environmental evaluation, because a previous
study42 showed that the environmental burdens of the technical
equipment are only of minor importance.
Fig. 1 Beneficial application of evaluation methods during process
development.
the knowledge about the process obtained during experimental
work, which was performed mainly at the Johannes-Gutenberg-
University Mainz, to identify major environmental hot-spots
and cost drivers.
Simplified life cycle assessment
We decided to evaluate the environmental burdens of the
processes under investigation by a vertically bounded SLCA,21
including all life cycle stages from source to gate to avoid
problem shifts to down-stream processes. For this, we followed
the desimplification approach proposed by the Society of
Environmental Toxicology and Chemistry (SETAC Europe).37
These assessments were performed in accordance to the ISO
norms 14040 and 14044.
Frequently occurring data gaps, which are inherent to R&D
stages, may occur, for example, when:
∑ reaction yields are directly analyzed from the reaction
mixture without the application of work-up procedures;
∑ the energy consumption is not measured;
Before the establishment of a suitable reaction set-up for
microreaction processes, some batch syntheses are usually
conducted as internal benchmarks. Afterwards the reaction is
transferred to the microreactor plant, where several process
alternatives are tested and optimized during the ongoing process
development. Thus, the batch syntheses as well as the continu-
ously running processes have been evaluated by SLCA in both
case studies and the results of these screenings were compared
with each other to derive further optimization potential.
∑ only major solvent fractions are recycled, whereas smaller
amounts of catalysts or solvents are mostly disposed.
In our approach, these data gaps were overcome by the
hypothetical calculation of work-up procedures and the com-
pletion of experimental data by generic data which enabled
the performance of mass and energy balances (Fig. 2). The
assessments are refined iteratively by close collaboration with
the researchers in the laboratory as more and more experimental
data become available. Scenario-based calculations as well
Economic evaluation during process development
Conjointly to the SLCA investigations, a screening concerning
economic hot-spots has been performed. In case study 1, the
costs of one batch reaction as well as the continuously running
processes were estimated based on the reaction protocols.
Furthermore, information about manpower requirements as
well as the investment and maintenance costs for the reactor
equipment were included in the estimation. All costs were
estimated at laboratory scale.
Although these costs are significantly higher than the actual
market price of the product, the comparison should be per-
formed on one scale.
The general method applied was as following:
(1) calculation of annual production capacities based on
reaction yields;
(2) calculation of annual material and energy costs based on
the annual production capacities;
(3) determination of the annual investment and maintenance
costs based on a 10 year period;
Fig. 2 Components of SLCA and cost evaluation during process
development.
1696 | Green Chem., 2011, 13, 1694–1707
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
(4) calculation of annual labour costs based on annual
production capacities;
(5) summation of all costs on an annual basis and assessment
of total costs per kg of product.
We decided not to perform cash flow analyses or to calculate
return on investment, since this would also require the consid-
eration of cost fluctuations.
The proposed method for cost estimation in laboratory scale
was also applied to case study 2. For calculation of fixed costs
of the continuously running syntheses, we considered the same
microreactor plant as in case study 1.
As the microreactor plant operates fully automatically, stop-
ping only in case of a disaster, the annual operating time was
calculated on the basis of 350 days per year and 24 hours per
day. Contrary to the continuously running reactions, the batch
processes can only be conducted under permanent attendance
of a laboratory assistant. For the sake of comparability between
both processing modes, we assumed a three-shift operation for
the batch reactions which can also be realized by the operation
of 3 batch reaction set-ups at one time. Thus, 6000 working
hours per year were considered for the batch process. Additional
to the mere reaction times, a cleaning procedure carried out
once a week was assumed in the case of the continuously
running reactions as well as a cleaning procedure after each
batch reaction.
Case study 1: synthesis of phenyl benzoate
As mentioned above, multiphase reactions may be carried out
beneficially in microreactors. Thus, the phase transfer catalysis
of phenol and benzoyl chloride giving phenyl benzoate (Scheme
1) was investigated by means of SLCA and cost estimation.
Phenyl benzoate is an industrially relevant chemical substance,
which is used for polymer modification and as an anti-oxidant,44
as well as an intermediate in the production of liquid crystals.45
The continuously running reactions were performed utilizing
micromixers with different mixing structures and partly applying
ionic liquids as phase transfer catalysts.
The material and energy costs were calculated based on the re-
action protocols linearly extrapolated to the annual production
capacity of each scenario considered. Current catalogue prices
as well as prices for electricity and water were used to estimate
the annual material and energy costs.
A lifetime of 10 years was assumed for the microreaction plant
although not all components were considered to work all this
time. Thus, components with a shorter lifetime were considered
to be replaced during the 10 year period. The current market
prices of these components were added to the investment costs
of the microreactor plant itself. These total investment costs
were divided by 10 according to a usual depreciation period.
Additionally, the annual maintenance costs were included in
these fix costs.
In the case of the batch synthesis, we considered standard
laboratory equipment consisting of round bottom flasks, mag-
netic stirrer with heating function, mechanical stirrer, and
reflux condenser. The purchasing costs of these items were also
calculated to a 10 year period taking under consideration the
(estimated) lifetime of every item.
Scheme 1 Phase transfer catalysis of benzoyl chloride and phenol
giving phenyl benzoate.
The experimental data used in the research accompany-
ing economical and ecological assessments were provided by
the working group of H. Loewe at the Johannes-Gutenberg-
University Mainz, Germany. All continuously running processes
where performed in a microreaction plant designed by mikroglas
chemtech GmbH, Mainz, Germany. This lab bench scaled plant
was custom-designed in close collaboration with Loewe and co-
workers to meet the needs of multiphase reactions.
Two batch syntheses, without a catalyst and with the ionic
liquid [BMIM]Cl in catalytic amounts, were evaluated by SLCA
to establish an internal benchmark. These two experiments
already revealed the phase transfer catalytic potential of ionic
liquids.
The continuously running experiments in the microreactor
plant were carried out to investigate the influence of increasing
the interfacial area by different micromixers (interdigital mixer,
herringbone structured mixer and emulsification mixer), which
were also provided by mikroglas chemtech GmbH. These
glass micromixers were used to generate uniform microbubbles
while simultaneously enabling a precise determination of the
interfacial area. Due to the mixing structure chosen, the
residence times differed from 0.83 s to 1.355 s. The reaction
conditions investigated are summarized in Table 1. The only
side reaction which may occur in the syntheses of phenyl
benzoate from benzoyl chloride and phenol is the hydrolysis
to benzoic acid. But, as the reaction solution was analyzed both
by IR spectroscopy and gas chromatography the amount of side
product was always less than 1%.
In the next step, the labour costs were calculated based on the
annual operation time of the continuously running processes.
Further, the gross hourly labour costs of a laboratory assistant
were estimated, including the employer’s contribution to social
insurance.
Finally, all these costs were summed up on an annual basis
and related to 1 kg of product.
As the purpose of the cost estimation was a relative
comparison of main cost drivers in batch and continuously
running syntheses, we did not consider any inflation rates
which may affect the estimation of fixed costs during the 10
year period. Further, we did not estimate any fluctuations
in substance prices or variations in labour costs which may
be due, for example, to legal measurements or salary adjust-
ments.
Cost estimation for the adaptation of the developed processes
to industrial scale were based on the reaction protocols already
applied for cost estimation in laboratory scale. The amounts
of substances used were extrapolated to typical industrial
production capacities (1000 tons per year) and multiplied with
prices for large purchase quantities.43
As phase transfer catalysts, the ionic liquids [C18MIM]Br and
[MIM]BuSO3 were used in the continuously running syntheses
(Scheme 2). In the batch synthesis, [BMIM]Cl was implemented
as a phase transfer catalyst. The three ionic liquids used in the
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1694–1707 | 1697
©

View Article Online
Table 1 Reaction conditions used for SLCA investigating the phase transfer catalysis of phenol and benzoyl chloride giving phenyl benzoate
Entry
Catalyst
T/◦C
Mixing structure
Residence time [s]
Yield [%]
1
2
3
4
5
6
7
8
No catalyst
No catalyst
[C18MIM]Br
[C18MIM]Br
[C18MIM]Br
[MIM]BuSO3
No catalyst
[BMIM]Cl
57
57
62
68
75
30
66
45.5
Interdigital mixer
Herringbone mixer
Herringbone mixer
Emulsification mixer
Interdigital mixer
Interdigital mixer
Batch synthesis
0.83
18.8
18.8
1.355
0.83
0.83
5400 (90 min)
2700 (45 min)
10
26
45
51
62
70
66
72
Batch synthesis
as for continuously running reactions. Here, an efficiency of
heat transfer under laboratory conditions of 40% (based on
experimental studies) was assumed. In contrast, the energy
consumption of the periphery in the case of the continuously
running syntheses, e.g. pumps and control systems, were based
on measurements.
As the process design is still under development, there has not
yet been a work-up procedure applied, since the reaction yields
were analyzed directly from the reaction mixture by gas chro-
matography. Thus, we considered the recycling of the organic
solvent toluene by assuming a redistillation and estimated the
energy consumption by using thermodynamic data. This was
completed by the calculation of two (hypothetical) scenarios for
the work-up of the reaction mixtures in microreactors including
ionic liquids as catalysts (entries 5 and 6). Here, we considered
a phase separation (in batch mode) first. The product remained
in the organic phase and was washed several times and dried
over sodium sulfate. Afterwards, the toluene was removed◦by
distillation since the boiling point of phenyl benzoate (315 C)
is much higher than that of toluene (111 ◦C). The ionic liquids
as well as non-reacted benzoyl chloride (as sodium benzoate
by reaction with sodium hydroxide) and phenol are soluble
in the aqueous phase and can be separated easily from the
product.
Scheme 2 Ionic liquids used as phase transfer catalysts: (a) [BMIM]Cl,
(b) [C18MIM]Br, (c) [MIM]BuSO3.
In order to avoid disposal problems for the contaminated
aqueous phase due to the remaining ionic liquid and non-reacted
reactants, we assumed a distillation of the water and a disposal of
the residual organic substances including ionic liquid as organic
waste
reactions as phase transfer catalysts were synthesized in the
working group of H. Loewe in the following procedures:
A continuously running microreactor plant provided by the
Institut fu¨r Mikrotechnik, Mainz, (IMM) was used to synthesize
the ionic liquid [BMIM]Cl from 1-methylimidazole and butyl
chloride (5% molar excess) at 145 ◦C and 6 bar. The product
was obtained by distillation of non-reacted butyl chloride and
recrystallization of the crude product from acetone.46 As the
neat reaction of butyl chloride and 1-methylimidazole is also a
biphasic reaction, the synthesis of [BMIM]Cl is subject to case
study 2.
Results
The functional unit chosen for both the SLCA as well as the
economic evaluation was 1 kg of the product phenyl benzoate.
All results are related to this unit.
[C18MIM]Br was obtained by reacting equimolar amounts of
octadecyl bromide and 1-methylimidazole in a round bottom
flask for 48 h at 90 ◦C. Afterwards the crude product was
recrystallized from tetrahydrofuran.10
As the synthesis of [MIM]BuSO3 is highly exothermic,
the reactants 1-methylimidazole and 1,4-butane sulfone were
premixed in a T-junction in situ dropped into hot vapors of
toluene. By this method, no further work-up was needed to
purify the ionic liquid after the distillation of toluene, since
toluene also functions as a recrystallization medium.10
The energy consumed during the reaction was estimated
based on thermodynamics both for batch syntheses as well
Simplified life cycle assessment
We investigated the environmental burdens of the process
simultaneously during the ongoing process development by
the SLCA approach mentioned above. By this method, the
continuously running syntheses were compared to the reference
batch syntheses with and without the implementation of a phase
transfer catalyst.
By means of the SLCA we addressed the following questions:
I. Are the continuously running syntheses more environmentally
benign than the batch syntheses?
1698 | Green Chem., 2011, 13, 1694–1707
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
II. What is the influence of the periphery in the energy
consumption of the continuously running syntheses com-
pared to the batch reactions?
III. Is the higher yield obtained by the implementation of
ionic liquids as phase transfer catalysts compensated by
higher environmental burdens caused by the supply of the
ionic liquids?
IV. Which mixing structure is the most efficient?
V. Which ionic liquid is more beneficial from an ecological point
of view?
VI. How does the consideration of work-up procedures affect
the environmental burdens of the process?
In Fig. 3 some toxicity potentials proposed by CML41 are
shown, presenting the most environmentally benign scenarios in
the centre of the chart.
Fig. 4 Further environmental impacts of the biphasic reaction of
benzoyl chloride and phenol giving phenyl benzoate (POCP – photo-
chemical ozone creation potential).
Fig. 3 Overview of various toxicity potentials of the biphasic reaction
of benzoyl chloride and phenol giving phenyl benzoate (HTP – human
toxicity potential, MAETP – marine aquatic ecotoxicity potential,
FAETP – fresh water aquatic ecotoxicity potential, TETP – terrestrial
ecotoxicity potential, MSETP – marine sedimental ecotoxicity potential,
FSETP – fresh water sedimental ecotoxicity potential).
Fig. 5 HTP of the biphasic reaction of benzoyl chloride and phenol
giving phenyl benzoate.
Further, reactions without the implementation of a phase
transfer catalyst (entries 1 and 2) reveal a higher HTP than
the reactions implementing ionic liquids. However, a reduction
of the HTP by more than 60% in the case of the uncatalyzed
continuously-running reaction can be obtained by simply sub-
stituting the interdigital mixer by the herringbone-structured
mixer. In contrast, the interdigital mixer performed best under
phase-transfer catalytic conditions.
The GWP indicates the greenhouse gas emissions caused
by the process as well as the up-stream processes. In Fig. 6,
presenting the GWP, the same trends as in the analysis of
the HTP could be observed. But, the energy consumed by
the periphery of the continuously-running processes during the
reaction exhibits the highest influence.
The highest toxicity potentials are caused by the continu-
ously running synthesis in an interdigital mixer without the
implementation of a phase transfer catalyst (entry 1). These
environmental impacts of the uncatalyzed reaction can be
reduced significantly by substituting the interdigital mixer
with a herringbone structured mixer (entry 2). Because of its
considerably higher environmental burdens, entry 1 was used
as a reference for scaling the environmental impacts. For the
sake of better clarity, in Fig. 4 further environmental impacts
are presented without the indication of entry 1. Here, a strong
reversed correlation between reaction yields obtained and the
resulting environmental impacts is observed due to less process
waste and a more efficient use of materials and energy.
In the following, the human toxicity potential (HTP) and
the global warming potential (GWP) as well as the cumulative
energy demand (CED) are chosen for a more detailed analysis
of environmental impacts. The HTP reflects emissions of the
process under investigation including up-stream processes which
are harmful to human health due to acute or chronic toxicity as
well as cancer-causing effects.
In both impact categories, the lowest environmental im-
pacts of the continuously-running phase transfer catalysis are
caused by the reaction in the interdigital mixer implementing
[MIM]BuSO3 (entry 6).
This reaction is conducted at 30 ◦C resulting in a low energy
demand for the reaction itself, which is mostly consumed
by the plant periphery. Surprisingly, the ionic liquids have
only a minor influence on the environmental impact, which
is counterbalanced by the significantly higher reaction yields
obtained in the catalyzed reactions. This is opposite to other
investigations where the environmental impacts were dominated
As can be seen from Fig. 5, the supply of phenol as well as
the energy consumption of the periphery in the continuously
running syntheses give the highest contribution to the HTP.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1694–1707 | 1699
©

View Article Online
Fig. 6 GWP of the biphasic reaction of benzoyl chloride and phenol
giving phenyl benzoate.
Fig. 8 HTP for the syntheses of the three ionic liquids used as phase
transfer catalysts.
more or less by the supply of these compounds if they were used
as a solvent in chemical syntheses.26,47,48 However, in the reactions
discussed here, [BMIM]Cl, [C18MIM]Br and [MIM]BuSO3 were
implemented only in catalytic amounts instead of being used
more commonly as solvents. Nevertheless, [MIM]BuSO3 has
been found to be more beneficial from an ecological point of view
than [C18MIM]Br, because it is synthesized in an exothermic
reaction, whereas the latter one is obtained after 48 h of
heating to 90 ◦C. But, [BMIM]Cl used in the batch synthesis
reveals the lowest overall environmental impacts due to its plain
reactants and the easy synthesis, as can be seen from Fig. 7.
Merely in the impact categories Ozone Depletion Potential and
Terrestrial Ecotoxicity Potential, the environmental burdens of
[MIM]BuSO3 exceed that of [C18MIM]Br.
of the ionic liquid syntheses. It can be seen from this figure that
the environmental burden of the supply of [C18MIM]Br is mainly
caused by the energy consumption during the reaction. Against
this, the HTP of [MIM]BuSO3 is mainly caused by the provision
of the acylating agent 1,4-butane sulfone. Please pay attention
to the fact that, contrary to the comparison of the alternative
reaction procedures of the phase transfer catalysis on a kilogram
scale, the ionic liquids were compared on a molar basis as the
catalysts were added to the reaction mixture in a fixed molar
amount.
Further we found that the batch syntheses with and without
the implementation of a catalyst (entries 7 and 8) show similar
environmental impacts as the continuously running processes
implementing [C18MIM]Br or [MIM]BuSO3 under the employ-
ment of an interdigital mixer (entries 5 and 6) in all impact
categories (Fig. 3–6). This results from the energy demand of
the plant periphery in the continuously-running syntheses which
reveals a considerable impact on the environmental burdens of
the processes.
In contrast, the batch reactions can be carried out without
any energy for pumping or controlling at the laboratory scale.
However, this apparent advantage of the batch processes is
suppressed when the processes are transferred to the industrial
scale. At industrial scale processing, pumps and controlling
instruments are needed for both continuously running as well as
batch reaction processing modes. To illustrate this issue in more
detail, the CED for the continuously running esterification with
the implementation of [C18MIM]Br and [MIM]BuSO3 (entries
5 and 6) is shown in Fig. 9. The CED comprises all energy
consumed during the whole processing chain from the supply
of raw materials and electrical current, the reaction itself, as
well as the work-up and the disposal of wastes. In this figure
the continuously running syntheses are presented both, with
the consideration of the energy demand of the plant periphery
and without its consideration. These results are opposite to the
CED of the batch syntheses (entries 7 and 8) indicating, that
under similar conditions, i.e. under the same energy demand for
pumping and controlling, the continuously running syntheses
would be 2 to 3 times more beneficial from an ecological point
of view (see also Huebschmann et al.49).
Fig. 7 Overall life cycle impacts for the three ionic liquids implemented
as phase transfer catalysts.
The result of lower environmental impacts caused by
[BMIM]Cl in comparison to [C18MIM]Br and [MIM]BuSO3 can
also be confirmed by Fig. 8, which exemplarily shows the HTP
Although no experimental work has been performed so far
to introduce an efficient work-up procedure of the ionic liquids
1700 | Green Chem., 2011, 13, 1694–1707
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
of ionic liquids are comparable to the reactions in batch mode
at laboratory scale.
II. The energy demand of the plant periphery is the main
influence on the environmental impacts of the continuously
running syntheses. This leads to the conclusion that the batch
reactions may be more environmentally benign because less
energy is needed. However, this advantage becomes neutraliz
ed at industrial scale as pumps and controlling systems are
needed for batch processes as well. Then, the continuously run-
ning processes should result in significantly lower environmental
impacts.
III. The environmental burdens of the ionic liquids used as phase
transfer catalysts are considerably smaller compared to the
higher yields obtained by their usage. Thus, the higher yields
obtained outweigh the environmental impacts of the
supply of ionic liquids.
IV. Without the implementation of a catalyst, the herringbone
structured mixer is more advantageous than the interdigital
mixer. In the case of the implementation of [C18MIM]Br, the
interdigital mixer performs best, followed by the emulsification
mixer and the herringbone structured mixer.
V. From an ecological point of view, [BMIM]Cl is the most
beneficial ionic liquid used as phase transfer catalyst followed
by [MIM]BuSO3, since [C18MIM]Br is synthesized under
energy demanding conditions at 90 ◦C for 48 h.
VI. The work-up procedure considered increases the environ-
mental burdens of the process considerably. Thus, this step is
crucial in the SLCA to avoid problem shifts. Future work is
focused on the substitution of the aqueous phase by a suitable
ionic liquid allowing for an automated phase separation and
recycling of the ionic liquid.
Fig. 9 CED for the continuously running biphasic reaction of benzoyl
chloride and phenol under consideration of the energy needed by the
reaction periphery (left) and without the consideration (middle) vs. the
CED of the biphasic reaction in batch mode (right) (scaled impact with
respect to entry 1).
and to isolate the product, this interesting issue was addressed
by means of hypothetical work-up steps considered in the
environmental assessment. The theoretical work-up procedure
for the isolation of the product and the adequate disposal of the
ionic liquids mentioned above leads to significantly increased
environmental impacts compared to the reactions without the
consideration of a work-up procedure, as shown in Fig. 10. This
result is mainly due to the evaporation of water from the aqueous
phase prior to the disposal of the organic residues including the
ionic liquids. In further experimental results, the aqueous phase
will be substituted by an ionic liquid, being both a solvent and
a phase transfer catalyst, to overcome the drawbacks of this
laborious work-up procedure. By using the ionic liquid as a
solvent, an automated phase separation and recycling of the
ionic liquid may be possible.
Economic evaluation
Similar or even better economic efficiencies are a premise for
newly developed chemical processes to enter the market and
compete with conventional processes. Thus, in addition to the
SLCA investigations, cost analyses have been performed already
at the laboratory scale to address the following issues:
I. Is the biphasic esterification in microstructured reactors
already economically competitive to the conventional produc-
tion procedure of phenyl benzoate?
II. What kind of costs mainly affect the overall costs?
III. How will these main components be reduced by transferring
the reaction to the production scale?
The continuously running synthesis under implementation of
[MIM]BuSO3 was chosen for the cost assessment. This reaction
resulted in the highest reaction yield and the lowest environmen-
tal impacts within their respective continuous processing modes.
The analysis, based on laboratory scale processing, was
performed on a superficial level, considering the categories of
fixed costs and variable costs of main cost items. The fixed
costs were estimated by including plant purchase, maintenance
and replacement of worn-out components. These costs were
summed up over a time period of 10 years and divided by 10 to
obtain the annual depreciation. Further, within the calculation
of variable costs, the costs of materials and energy utilized in
the reaction as well as in the hypothetically calculated work-up
procedure were considered based on the experimental protocols
Fig. 10 GWP for the continuously running esterification of phenol
and benzoyl chloride under the implementation of [C18MIM]Br and
[MIM]BuSO3 both without (left) and with (right) the consideration of a
hypothetical work-up procedure.
In summary, the questions raised at the beginning of the
process design can be answered as following:
I. From an ecological point of view, the continuously running
esterifications in an interdigital mixer under the implementation
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1694–1707 | 1701
©

View Article Online
and multiplied by actual catalogue prices.50 In addition, the
labour was estimated by multiplying the working time based
on the reaction protocols with the hourly labour costs of a
laboratory assistant. This included also the time which may
be needed to work-up the reaction solution once per day. As
a cautious estimate we considered a lower average gross wage of
€18 per hour including also the employer’s contribution to the
social insurance.
These costs were summed up on an annual basis and related
to the production capacity of phenyl benzoate to obtain the
costs per kg of product. As the highest yield was obtained in the
interdigital mixer by implementation of [MIM]BuSO3 as phase
transfer catalyst, this reaction was selected to discuss the cost
analysis.
As was foreseeable at the laboratory scale, the total costs for
producing 1 kg of phenyl benzoate by continuously running
phase transfer catalysis in microstructured reactors under im-
plementation of the ionic liquid have been found to be higher
than the selling price of this substance. However, considering a
batch process on the same laboratory scale (applying [BMIM]Cl
as phase transfer catalyst) leads to a total cost 4.5 times higher
due to very high labour costs. The total cost of the continuously
running synthesis at the laboratory scale are composed of 43.8%
fixed costs, 43.7% labour costs and 12.5% material costs. The
material costs are mainly influenced by the hypothetical work-
up procedure, the provision of toluene (although recycling of
the solvent was assumed), the synthesis of [MIM]BuSO3 at
laboratory scale, the energy consumption of the microreactor
plant as well as the provision of the reactants benzoyl chloride
and phenol (Fig. 11).
capacity assumed may be achieved by enhancing the flow rates
under retention of yield and selectivity.
This calculation demonstrates that the compensation for the
fixed and labour costs is feasible by higher production rates
as the material becomes more influential. Usually, the share of
the raw material costs is about 50 to 60% of the total costs.51
Taking into account market prices for large purchase quantities
of solvents and basic substances for the same set-up, an overall
reduction in 99% of total costs seems to be possible for a future
industrial scale processing.43
By estimating the costs for the continuously running synthesis
under implementation of [MIM]BuSO3 (entry 6) and compari-
son to the batch reaction under implementation of [BMIM]Cl
as a phase transfer catalyst (entry 8) we found that:
I. The biphasic esterification carried out in the microstructured
reactor at laboratory scale is more economically preferable than
the batch reaction at laboratory scale.
II. The main cost drivers in the base case are fixed and labour
costs, which may become negligible by enhancing the production
capacities.
III. A reduction in 99% of total costs seems to be possible by
transferring the continuously running biphasic esterification
from laboratory scale to industrial scale.
Case study 2: synthesis of [BMIM]Cl
In the second case study, the environmental burdens and total
costs of the synthesis of the ionic liquid [BMIM]Cl were
investigated. This substance can be used, for example, as a phase
transfer catalyst as was shown in case study 1. Although many
ionic liquids are still under investigation, this is a very well-
known representative of this class of substances. [BMIM]Cl
was chosen as the second model substance for the comparative
evaluation of batch and continuous processing, because in the
solventless reaction (Scheme 3) of butyl chloride (BuCl) and 1-
methylimidazole (MIM) a biphasic system is formed after the
conversion of at least 8% 1-methylimidazole.52 As this is also an
exothermic reaction, it was regarded as being beneficially carried
out in microstructured reactors.
Scheme
3
Synthesis of the ionic liquid [BMIM]Cl from
1-methylimidazole and butyl chloride.
Fig. 11 Variable costs (without labour costs of the esterification) for the
continuously running synthesis of phenyl benzoate at laboratory scale.
As a benchmark, two batch◦ syntheses were investigated
◦
carried out at 80 C53 and 75 C52 by D. Kralisch and A.
Große Bo¨wing & A. Jess, respectively. Both batch reactions
were conducted in a 500 mL round bottom flasks equipped
with a reflux condenser and a mechanical stirrer. These results
were compared to four continuously running synthesis scenarios
at temperatures of 145 ◦C and 150 ◦C under the employment
of a Knauer SmartmixTM mixer, which were conducted by
Loewe and co-workers.46 All reaction conditions considered are
summarized in Table 2.
A reduction in the total cost may be achieved by increasing
the production capacity by a factor of at least 75 under the
same reaction conditions (reaction temperature 30 ◦C, yield
70%, constant labour and fixed costs), which would be realizable
in the lab scale plant. By this, the total costs can be decreased by
90% relative to the formerly discussed base case. The total costs
consist of 5.5% fixed costs, 5.5% labour costs and 89% material
costs (under consideration of lower unit costs for larger amounts
of substances required for the synthesis). The higher production
The reaction mixtures were heated up by oil bath heating in
all cases. Contrary to case study 1, the estimation of the energy
1702 | Green Chem., 2011, 13, 1694–1707
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
Table 2 Reaction conditions used for SLCA investigation of the biphasic reaction of butyl chloride and 1-methylimidazole giving [BMIM]Cl46,52,53
Entry
Processing mode
Molar ratio (BuCl : MIM)
T/◦C
Residence time [min]
Yield [%]
1
2
3
4
5
6
Batch
Batch
Continuous
Continuous
Continuous
Continuous
1.2
1.4
1.2
2
2
1
75
80
145
145
150
150
2820 (47 h)
1305 (21.75 h)
31.7
23.4
12.3
96
70
87
70
77
82
12.1
needed during the syntheses by means of thermodynamic data
was not suitable, since this is a highly exothermic reaction.
Thus, the energy consumption of the oil bath heating was
estimated based on energy measurements. This can be regarded
as a conservative approach because thermal effects of the
synthesis were mainly neglected. Furthermore, in the case of
the continuously running reactions the energy of the periphery
(pumps, controlling instruments) were estimated based on the
data investigated in case study 1. Here, we assumed that the same
microreaction plant could be used for the synthesis of the ionic
liquid, too.
Two different work-up procedures were considered depending
on the processing mode: in the batch syntheses, a dilution in
water and extraction with methyl tert-butyl ether was assumed
as described by Kralisch.53 Afterwards, the ether extracts were
collected and the ether was distilled off with a recovery rate
of 90%. In case of the continuously running synthesis, the
residual butyl chloride was distilled off and the crude product
was recrystallized from acetone.46
Fig. 12 Overview of environmental impacts of the synthesis of
[BMIM]Cl without solvent.
This outcome is independent, whether the reaction is carried out
as a batch reaction or as a continuously running reaction.
The environmental impacts, e.g. the HTP (Fig. 13), of the
syntheses are dominated by the provision of the reactants
1-methylimidazole and butyl chloride. Thus, the molar ratio of
both substances is of great importance for the environmental
performance of the reaction. Because of this, the continuously
running reactions with a molar ratio BuCl : MIM of 1.2 : 1
and 1 : 1 (entries 3 and 4) resulted in the lowest environmental
impacts.
Results
As the functional unit, 1 kg [BMIM]Cl was chosen to investigate
the environmental burdens as well as the total costs of this
biphasic reaction.
Simplified life cycle assessment
The reactions were carried out under varying reaction conditions
in two different processing modes (see Table 2). Thus, by means
of SLCA the following aspects were investigated:
I. Which reaction component shows the main influence on the
environmental impacts?
II. Which molar ratio of both reactants is the most environmen-
tally benign?
III. Do the reaction temperature and the residence time influence
the environmental performance of the system under investi-
gation?
IV. How does the applied work-up procedure affect the environ-
mental burdens of the reaction?
V. Which processing mode is more beneficial from an ecological
point of view?
In Fig. 12, an overview of the environmental impacts of the
reactions investigated is given. As the results obtained by Große
Bo¨wing and Jess were considered as a reference, entry 1 was
chosen to scale the environmental impacts. It can be seen that
processes with a reaction yield lower than 80% (entries 2, 4,
5) are less environmentally benign than the reference synthesis.
Fig. 13 HTP of the synthesis of [BMIM]Cl from butyl chloride and 1-
methylimidazole without solvent in batch (left) and continuously (right)
running reactions.
In addition, the energy consumed during the synthesis
strongly influences the overall GWP of the reaction procedures
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1694–1707 | 1703
©

View Article Online
III. Reaction temperatures and residence times influence the
environmental burdens of the process only indirectly. Due to the
exothermic character of this process, high reaction temperatures
can only be handled safely in continuously running reactors with
short residence times.
IV. The work-up procedures considered have only a minor
influence on environmental burdens (<10%).
V. From an ecological point of view, the continuously running
processes are more environmentally benign than the ionic liquid
syntheses in batch mode. This result is only valid if high reaction
yields can be obtained, low molar ratios are applied and short
residence times can be realized.
Economic evaluation
In addition to the SLCA investigations, a rough cost estimation
was applied to the solventless synthesis of [BMIM]Cl in labora-
tory scale. By this, the following questions were examined:
I. What kind of costs mainly influence the total costs?
II. Whichprocessing mode should be preferred froman economic
point of view?
Fig. 14 GWP of the synthesis of [BMIM]Cl from butyl chloride and 1-
methylimidazole without solvent in batch (left) and continuously (right)
running reactions.
considered (Fig. 14). In consequence, higher GWPs have been
found for the syntheses under batch conditions compared to
III. Which reaction component has the main influence on the
variable costs?
◦
◦
the continuously running reactions at 145 C and 150 C with
molar ratios BuCl : MIM of 1.2 : 1 and 1 : 1 (entries 3 and 5).
In microstructured reactors, the synthesis can be performed at
higher temperatures to activate the reaction. Due to the high
surface-to-volume ratio of the microreactor, the generated heat
can be removed more efficiently to avoid a thermal runaway.
Thus, the residence times can be shortened to several minutes
instead of a couple of hours. Clear ecological advantages for the
continuously running syntheses in a microreaction set-up can be
gained from that, as shown in the examples of HTP and GWP
in Fig. 13 and 14.
According to a lack of more detailed information, the fixed
costs for the continuously operating microreaction plant as
well as labour costs were estimated based on the information
gathered in case study 1. This was assumed to be an appro-
priate approach, since the microreaction plant was specifically
designed for biphasic reactions.
In Fig. 15, the total costs of the syntheses of [BMIM]Cl in
laboratory scale are presented, scaled to the estimation of total
costs of the batch synthesis published by Große Bo¨wing and
Jess.52 Due to the high labour costs, the total costs of the batch
syntheses are considerably higher than the total costs of the
syntheses performed in a fully automated microreaction plant.
In the case of the continuously running syntheses, labour costs
and variable cost have about equal shares in the total costs.
The share of the respective work-up procedures are compara-
bly low in all cases and more or less independent from the work-
up procedure assumed. The HTP (Fig. 13) of the respective
procedures are 3.5% in the case of the batch syntheses and
1.7 to 2.8% of the overall environmental impact depending on
the reaction yield obtained during the continuously running
syntheses. In the GWP (Fig. 14), the share of the work-up
procedure in the overall environmental burden is 4 to 4.5% for
the batch syntheses and 4 to 6.7% for the continuously running
reactions. These results are basically caused by the provision of
the organic solvent for extraction or recrystallization and the
energy consumed by distillation of the solvent. The reuse of the
organic solvents is already included in this estimation. These
simple procedures for product isolation can be applied because
of the solventless reaction.
The questions raised at the beginning of the the SLCA
investigations could be answered as follows:
I. Both reactants, 1-methylimidazole and butyl chloride, con-
siderably influence the environmental impacts of the syntheses
under investigation. Furthermore, the energy consumed during
the reaction may also have a strong influence on the environ-
mental burdens.
Fig. 15 Composition of total costs of the synthesis of [BMIM]Cl from
butyl chloride and 1-methylimidazole without solvent in batch (left) and
continuously running (right) reactions.
II. The molar ratio of both reactants should be kept as low
as possible. The ideal ratio is 1 : 1. Thus, a continuously running
synthesis at 150 ◦C employing equal amounts of reactants
has been found to be the most environmentally benign synthesis
pathway considered in this investigation.
The lowest total costs are caused by the continuously running
synthesis at 150 ◦C with a molar ratio BuCl:MIM of 1 : 1,
resulting in a reaction yield of 82%. According to Fig. 16, the
variable costs of this synthesis mainly consist of the purchase
costs for the reactants, 1-methylimidazole and butyl chloride.
1704 | Green Chem., 2011, 13, 1694–1707
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
to a simplified product isolation by distillation from the organic
phase as well as a recovery of the ionic liquid for further
processing.
In the case of the synthesis of the ionic liquid [BMIM]Cl,
the environmental burdens were influenced by both the re-
action yields obtained and the molar ratio of the reactants
1-methylimidazole and butyl chloride. Thus, the most environ-
mentally benign synthesis was carried out in the continuously
◦
operated microreaction plant at 150 C and equal amounts of
reactants.
Work-up procedures were considered for both processing
modes. In case study 2, the reaction was conducted without
an additional solvent. Thus, the removal of a solvent prior to
product isolation was not necessary. This led to simplified work-
up procedures compared to other reaction types implementing
organic solvents, e.g., case study 1. By this, the product isolation
had a minor influence on the environmental burdens of the
reactions under investigation.
Additionally, the economic potential of both processes was
demonstrated although at a screening level. The superficial
cost estimation at laboratory scale of the synthesis of phenyl
benzoate showed that, under laboratory scale conditions, the
batch syntheses result in higher costs due to higher manpower
requirements. Under industrial scale conditions, syntheses with
the highest yields will be the most cost efficient, since the total
costs will be influenced mainly by the costs of raw materials.
Fixed costs, especially investment costs, will have a minor share.
In conclusion, helpful decision support could be gained from
these evaluations, paving the way for a deliberately green process
design.
Fig. 16 Composition of variable costs of the synthesis of [BMIM]Cl
from butyl chloride and 1-methylimidazole without solvent.
From the comparison of batch and continuously running
syntheses in laboratory scale it was concluded, that:
I. The total costs of the batch syntheses consist mainly of
labour costs, whereas the total costs of the continuous reactions
are equally dominated by labour and variable costs. Fixed costs
have only a minor effect on the total costs in both processing
modes.
II. As the continuous reactions in a fully automated microre-
action plant lead to considerably lower total costs compared to
batch reactions, this processing mode should be preferred from
an economic point of view for further process design studies.
III. The variable costs are mainly influenced by the provision of
reactants.
Conclusions
Experimental
We have presented our research accompanying SLCA com-
plemented by a superficial cost analysis which is illustrated
by discussing the two case studies: continuously running
phase transfer catalysis of benzoyl chloride and phenol giving
phenyl benzoate, and the solventless synthesis of the ionic
liquid [BMIM]Cl. Both reactions are multiphase processes
which can be carried out beneficially in microstructured
reactors.
The environmental burdens of the synthesis of phenyl
benzoate were essentially influenced by the reaction yields
obtained, as higher yields lead to less process waste and to a
more efficient use of resources. Thus, the energy consumed by
the fully automated microreactor plant exhibited the highest
contribution to the Global Warming Potential, whereas the
Human Toxicity Potential was almost equally dominated by
the energy consumption and the provision of phenol. The
performance of different micromixing structures has been found
to be dependent on other process parameters, e.g., the utilization
of a phase transfer catalyst. In comparison, the interdigital
micromixer (IMM) in combination with [MIM]BuSO3 and
a reaction temperature of 30 ◦C have shown the lowest
environmental impact of all continuously running synthesis
alternatives.
The phase transfer catalytic reactions were carried out under
standardized two-phase conditions. The inorganic (aqueous)
phase contained 0.3 mol L-1 phenol, 0.35 mol L-1 sodium
hydroxide and 0.01 mol L-1 of the selected phase transfer catalyst
while the organic phase (toluene) contained 0.1 mol L-1 benzoyl
chloride. With the applied volume flows of 0.3 mL min-1 of the
inorganic phase (0.09 mol min-1 phenol) and 0.9 mL min-1 of
the organic phase (0.09 mol min-1 benzoyl chloride) dispersions
were formed in the micromixers chosen. To calculate kinetic
data, experiments were performed in a thermostatted water bath
◦
◦
at different temperatures ranging from 20 C to 80 C. It was
assumed that the temperature inside the glass microreactors is
equal to the outside thermostatted fluid due to their very small
inner volumes.
The concentrations of phenyl benzoate, benzoyl chloride and
benzoic acid were determined by a Thermo Nicolet 380 FT-IR
spectrometer equipped with a Smart MIR Fibre Port (Themo),
connected to a DiProbe-ATR infrared fibre sensor (ifs Aachen,
Germany). OMNIC 7.3 software was used for data acquisition
and subsequently analysed by TQ Analyst version 6.2.1.509. As
a calibration model for determining phenyl benzoate, benzoyl
chloride and benzoic acid in toluene, a PLS algorithm (partial
least squares) was applied with the assumption that concentra-
tions of each component below 0.1 mol L-1 does not influence
the concentration of toluene significantly.
As this process is still under development, there a hypothetical
work-up procedure was considered. Future experimental work
will focus on the continuously operated phase separation leading
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1694–1707 | 1705
©

View Article Online
8 G. Dummann, U. Quittmann, L. Gro¨schel, D. W. Agar, O. Wo¨rz and
K. Morgenschweis, Catal. Today, 2003, 79–80, 433–439.
9 A. Gu¨nther and K. F. Jensen, Lab Chip, 2006, 6, 1487–1503.
10 D. Breuch, Diploma Thesis, Johannes-Gutenberg-University,
Mainz, Germany, June 2009.
11 T. R. Dietrich, A. Freitag, S. Link and R. Scholz, Proceedings of
the 11th International Conference on Microreaction Technology
(IMRET 11), Kyoto, March 8–10, 2010.
12 N. Pamme, Lab Chip, 2007, 7, 1644–1659.
13 O. K. Castell, C. J. Allender and D. A. Barrow, Lab Chip, 2009, 9,
388–396.
14 J. G. Kralj, H. R. Sahoo and K. F. Jensen, Lab Chip, 2007, 7, 256–263.
15 R. A. Sheldon, Chem. Tech., 1994, 24, 38–47.
16 B. M. Trost, Angew. Chem., Int. Ed. Engl., 1995, 34, 259–281.
17 M. Eissen and J. O. Metzger, Chem.–Eur. J., 2002, 8, 3580–3585.
18 K. A. Hossain, I. F. Khan and K. Hawboldt, Ind. Eng. Chem. Res.,
2007, 46, 8787–8795.
19 EN ISO 14040: Environmental Management, Life Cycle Assessment,
Principles and Framework, DIN Deutsches Institut fu¨r Normung e.
V., Beuth Verlag, Berlin, 2006.
20 EN ISO 14044: Environmental Management, Life Cycle Assessment,
Requirements and Guidelines, DIN Deutsches Institut fu¨r Normung
e. V., Beuth Verlag, Berlin, 2006.
21 G. Rebitzer, T. Ekvall, R. Frischknecht, D. Hunkeler, G. Norris, T.
Rydberg, W.-P. Schmidt, S. Suh, B. P. Weidema and D. W. Pennington,
Environ. Int., 2004, 30, 701–720.
Abbreviations
[BMIM]Cl
[C18MIM]Br
1-Butyl-3-methylimidazolium chloride
1-Octadecyl-3-methylimidazolium bromide
[MIM]BuSO3 1-Butylsulfonate-3-methylimidazolium
BuCl
CED
CML
Butyl chloride
Cumulative Energy Demand
Institute of Environmental Science, Univer-
siteit Leiden, The Netherlands
Environmental Assessment Tool for Organic
Syntheses
Ecological and Economic Optimisation
Environmental factor
Environmental Health and Safety
Energy Loss Index
EATOS
ECO
E-Factor
EHS
ELI
FAETP
FSETP
GWP
HTP
IMM
ISO
Fresh Water Aquatic Ecotoxicity Potential
Fresh Water Sedimental Ecotoxicity Potential
Global Warming Potential
Human Toxicity Potential
Institut fuer Mikrotechnik Mainz
International Standard Organisation
Life Cycle Assessment
22 A. Biwer and E. Heinzle, Chem. Ing. Tech., 2001, 73, 1467–1472.
23 S. Kheawhom and M. Hirao, Comput. Chem. Eng., 2002, 26, 747–
755.
LCA
LCI
Life Cycle Inventory
LCIA
MAETP
MSETP
MIM
MIPS
PTC
POCP
R&D
SETAC
Life Cycle Impact Assessment
Marine Aquatic Ecotoxicity Potential
Marine Sedimental Ecotoxicity Potential
1-Methylimidazole
Material intensity per service unit
Phase transfer catalysis
Photochemical Ozone Creation Potential
Research and Development
Society of Environmental Chemistry and Tox-
icology
24 G. Fleischer and W.-P. Schmidt, Int. J. Life Cycle Assess., 1997, 2,
20–24.
25 D. Kralisch, D. Reinhardt and G. Kreisel, Green Chem., 2007, 9,
1308–1318.
26 S. Huebschmann, D. Kralisch, V. Hessel, U. Krtschil and C. Kompter,
Chem. Eng. Technol., 2009, 32, 1757–1765.
27 Y. Zhang, B. R. Bakshi and E. S. Demessie, Environ. Sci. Technol.,
2008, 42, 1724–1730.
28 P. T. Anastas and J. Warner, in Green Chemistry: Theory and Practice,
1st edn, Oxford University Press, New York, 1998.
29 P. T. Anastas and J. Zimmerman, Environ. Sci. Technol., 2003, 37,
94A–101A.
30 T. E. Graedel, Int. J. Life Cycle Assess., 1996, 1, 85–89.
31 F. Hinterberger and F. Schmidt-Bleek, Ecol. Econ., 1999, 29, 53–56.
32 G. Koller, U. Fischer and K. Hungerbu¨hler, Ind. Eng. Chem. Res.,
2000, 39, 960–972.
33 H. Sugiyama, U. Fischer, K. Hungerbu¨hler and M. Hirao, AIChE
J., 2008, 54, 1037–1053.
SLCA
TETP
TAPPS
Simplified Life Cycle Assessment
Terrestrial Ecotoxicity Potential
Total annualized profit per service unit
34 A. A. Bumann, S. Papadokonstantakis, H. Sugiyama, U. Fischer and
K. Hungerbu¨hler, Energy, 2010, 35, 2407–2418.
35 V. H. Hoffmann, K. Hungerbu¨hler and G. J. McRae, Ind. Eng. Chem.
Res., 2001, 40, 4513–4524.
Acknowledgements
36 T. Hur, J. Lee, H. Ryu and E. Kwon, J. Environ. Manage., 2005, 75,
229–237.
The authors thank the German Federal Environmental Foun-
dation (Deutsche Bundesstiftung Umwelt – DBU) for funding
this project (AZ 25599-31). Furthermore, the precious advice
provided by S. Haertner, Merck KG aA, during the project time
is gratefully acknowledged.
37 Simplifying LCA: just a cut? – Final report of the SETAC-Europe
Screening and Streamlining Working-Group, Society of Environ-
mental Chemistry and Toxicology (SETAC), Brussels, Belgium, 1997.
38 Umberto, v. 5.5, ifu Institut fu¨r Umweltinformatik, Hamburg,
Germany; ifeu Institut fu¨r Energie- und Umweltforschung,
Heidelberg, Germany, 2006.
39 R. Frischknecht, N. Jungbluth, H.-J. Althaus, C. Bauer, G. Doka, R.
Dones, R. Hischier, T. Nemecek, A. Primas and G. Wernet, Ecoinvent
database v. 2.1, Swiss Centre for Life Cycle Inventories, Du¨bendorf,
Switzerland, 2009.
40 Cumulative Energy Demand – Terms, Definitions, Methods of
Calculation, VDI Guideline 4600, Du¨sseldorf, Germany, 1999.
41 J. B. Guine´e, Life Cycle Assessment – An Operational Guideline to
the ISO standards – Parts 1–3, Ministry of Housing, Spatial Planning
and the Environment and the Centre of Environmental Science, ed., M.
Gorree and R. Heijungs, Leiden University, The Netherlands, 2001.
42 D. Kralisch and G. Kreisel, Chem. Eng. Sci., 2007, 62, 1094–1100.
43 Industrial pricing of basic chemicals available under: http://www.
chemweek.com or http://www.icispricing.com, accessed: Dec.
2010.
References
1 V. Hessel, C. Hofmann, P. Lo¨b, J. Lo¨hndorf, H. Lo¨we and A. Ziogas,
Org. Process Res. Dev., 2005, 9, 479–489.
2 V. Hessel, D. Kralisch and U. Krtschil, Energy Environ. Sci., 2008, 1,
467–489.
3 J.-C. Charpentier, Chem. Eng. J., 2007, 134, 84–92.
4 D. Schmalz, M. Ha¨berl, N. Oldenburg, M. Grund, H. Muntermann
and U. Kunz, Chem. Ing. Tech., 2005, 77, 859–866.
5 B. P. Mason, K. E. Price, J. L. Steinbacher, A. R. Bogdan and D. T.
McQuade, Chem. Rev., 2007, 107, 2300–2318.
6 S. Aljbour, H. Yamada and Y. Tagawa, Chem. Eng. Process., 2009,
48, 1167–1172.
44 US Pat., US 3 772 389, 1971.
45 Ger. Pat., DE 302696, 1981.
7 H. Song, J. D. Tice and R. F. Ismagilov, Angew. Chem. Int. Ed., 2003,
42, 767–772.
1706 | Green Chem., 2011, 13, 1694–1707
This journal is
The Royal Society of Chemistry 2011
©

View Article Online
46 J.-H. Petersen, Diploma Thesis, Johannes-Gutenberg-University,
Mainz, Germany, 2008.
47 A. Stark, S. Huebschmann, M. Sellin, D. Kralisch, R. Trotzki and B.
Ondruschka, Chem. Eng. Technol., 2009, 32, 1730–1738.
48 D. Reinhardt, F. Ilgen, D. Kralisch, B. Ko¨nig and G. Kreisel, Green
Chem., 2008, 10, 1170–1181.
50 Price lists of chemical substances: (a) http://www.sigmaaldrich.com;
(b) http://www.carl-roth.de, accessed: Nov. 2009.
51 A. C. Dimian, Integrated Design and Simulation of Chemical Pro-
cesses, Elsevier Science B. V., Amsterdam, 1st edn, 2003, ch. 15,
pp. 571–604.
52 A. Große Bo¨wing and A. Jess, Green Chem., 2005, 7, 230–235.
53 D. Kralisch, PhD thesis, Friedrich-Schiller-University Jena,
Germany, 2007.
49 S. Huebschmann, D. Kralisch, D. Breuch, H. Loewe, R. Scholz and
T. Dietrich, Chem. Eng. Trans., 2010, 21, 655–660.
This journal is
The Royal Society of Chemistry 2011
Green Chem., 2011, 13, 1694–1707 | 1707
©