Inexpensive ionic liquids: [HSO4]--based solvent production at bulk scale

Article abstract of DOI:10.1039/c4gc00016a

Through more than two decades' intensive research, ionic liquids (ILs) have exhibited significant potential in various areas of research at laboratory scales. This suggests that ILs-based industrial process development will attract increasing attention in the future. However, there is one core issue that stands in the way of commercialisation: the high cost of most laboratory-synthesized ILs will limit application to small-scale, specialized processes. In this work, we evaluate the economic feasibility of two ILs synthesized via acid-base neutralization using two scenarios for each: conventional and intensification processing. Based upon our initial models, we determined the cost price of each IL and compared the energy requirements of each process option. The cost prices of triethylammonium hydrogen sulfate and 1-methylimidazolium hydrogen sulfate are estimated as $1.24 kg ^-1 and $2.96-5.88 kg^-1, respectively. This compares favourably with organic solvents such as acetone or ethyl acetate, which sell for $1.30-$1.40 kg^-1. Moreover, the raw materials contribute the overwhelming majority of this cost and the intensified process using a compact plate reactor is more economical due to lower energy requirements. These results indicate that ionic liquids are not necessarily expensive, and therefore large-scale IL-based processes can become a commercial reality. This journal is the Partner Organisations 2014.

Full text of DOI:10.1039/c4gc00016a

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Inexpensive ionic liquids: [HSO4]−-based solvent
production at bulk scale†
Cite this: Green Chem., 2014, 16,
3098
Long Chen,‡a,b Mahdi Sharifzadeh,‡c,d Niall Mac Dowell,e Tom Welton,b Nilay Shahc,d
and Jason P. Hallett*c
Through more than two decades’ intensive research, ionic liquids (ILs) have exhibited significant potential
in various areas of research at laboratory scales. This suggests that ILs-based industrial process develop-
ment will attract increasing attention in the future. However, there is one core issue that stands in the way
of commercialisation: the high cost of most laboratory-synthesized ILs will limit application to small-
scale, specialized processes. In this work, we evaluate the economic feasibility of two ILs synthesized via
acid–base neutralization using two scenarios for each: conventional and intensification processing. Based
upon our initial models, we determined the cost price of each IL and compared the energy requirements
of each process option. The cost prices of triethylammonium hydrogen sulfate and 1-methylimidazolium
hydrogen sulfate are estimated as $1.24 kg−1 and $2.96–5.88 kg−1, respectively. This compares favourably
with organic solvents such as acetone or ethyl acetate, which sell for $1.30–$1.40 kg−1. Moreover, the
raw materials contribute the overwhelming majority of this cost and the intensified process using a
compact plate reactor is more economical due to lower energy requirements. These results indicate that
ionic liquids are not necessarily expensive, and therefore large-scale IL-based processes can become a
commercial reality.
Received 6th January 2014,
Accepted 2nd April 2014
DOI: 10.1039/c4gc00016a
www.rsc.org/greenchem
single component molecular solvents. Consequently, ILs have
been applied in a broad range of areas, such as fuel cells, bat-
Introduction
Ionic liquids (ILs) have been generating rising interest over the teries, capacitors, thermal fluids, plasticizers, lubricants and
last two decades with a diversified range of applications. There solvents in analysis, synthesis and catalysis1–3 and more
are a number of properties which make ILs attractive both in recently in separations (for example, carbon capture).4–6
academic and industrial fields. For example, they generally
Despite all of these advantages and potential applications,
exhibit very low vapour pressures under ambient conditions, ILs currently suffer from clear and significant disadvantages
which also leads to most ILs being non-flammable and that stand in the way of many commercial applications. Most
reduces exposure risk. Therefore, much of the interest of ILs significant and frequently cited among these is the high cost
has centred on the use of these solvents as alternatives to vola- of most ILs. For example, ILs have been applied as solvents for
tile organic solvents. Moreover, ILs are claimed to be ‘designer a biomass deconstruction process which is believed to be a
solvents’1 based upon their being composed of two distinct nascent pre-treatment technology and holds great promise.7–9
parts, resulting in a synthetic flexibility that is not available for Klein-Marcuschamer et al.10 have conducted techno-economic
analysis of this ILs-based biomass pre-treatment process, and
report that in order to make this process a practical reality,
aState Key Laboratory of Chemical Engineering, School of Chemical Engineering,
East China University of Science and Technology, 130 Meilong Road,
Shanghai 200237, China
three key factors should be addressed: reducing IL cost, redu-
cing IL loading and increasing IL recycling. Close inspection
reveals that the latter two items are also associated with the
cost of the IL employed. If the purchase price of ILs can be
reduced, this process will be placed in a competitive position
with other conventional pre-treatment process. However, at the
time of writing, the Sigma-Aldrich website (selling ILs manu-
factured by BASF) quotes the price of acetate ILs at
bDepartment of Chemistry, Imperial College London, London, SW7 2AZ, UK
cDepartment of Chemical Engineering, Imperial College London, London, SW7 2AZ,
UK. E-mail: j.hallett@imperial.ac.uk
dCentre for Process Systems Engineering, Imperial College London, London,
SW7 2AZ, UK
eCentre for Environmental Policy, Imperial College London, London, SW7 2AZ, UK
†Electronic supplementary information (ESI) available: Details of IL1 and IL2
ca. $700 kg−1 and chloride ILs at ca. $300 kg−1 11
Although
.
synthesis, simulation results and economical analysis method are given here.
See DOI: 10.1039/c4gc00016a
prices for small quantities should not be used as a guide to
commercial utility, it is believed that ILs normally fall in the
‡Both of these authors contributed equally to this work.
3098 | Green Chem., 2014, 16, 3098–3106
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
range of 5–20 times more expensive than molecular solvents.12
However, if an ionic liquid is being considered as a component
of an industrial process (for example, as the solvent for a
biomass pre-treatment process), it is important to investigate
and optimise in terms of both cost and environmental impact
the synthetic route (at manufacturing scale) leading to that IL.
For example, the only two required starting materials for
synthesising the IL triethylammonium hydrogen sulfate
([HNEt3][HSO4]) are triethylamine and sulfuric acid. Neither
costs more than $2 kg−1 in tonne quantities, and the synthetic
route is limited to simple mixing and stoichiometric reaction.
However, to the best of our knowledge, there have been no
reports of the techno-economic impacts of IL production at
scale, although it is commonly held that ILs are currently too
expensive to be utilized at industrial scale. The most common
criticism of ILs that the authors encounter is that of the
‘severe’ limitations placed upon their large-scale deployment
by their high cost. But are ionic liquids inherently expensive,
or is this opinion a consequence of the specific ILs that are
historically prominent (dialkylimidazolium cations with poly-
fluorinated anions)? The answer to this question requires
techno-economic analyses of the IL manufacturing processes,
involving detailed process engineering and analyses such as
those applied in many early-stage analyses of novel processes,
for example the aforementioned biofuels production.
Fig. 1 Schematic diagram of two IL synthetic processes (a) IL1; (b) IL2.
released from the exothermic reaction, making the tempera-
ture and reaction rate easier to control. In the present work,
two kinds of IL (Fig. 1) which were made by this method are
investigated, namely, triethylammonium hydrogen sulfate
([HNEt3][HSO4], IL1) and 1-methylimidazolium hydrogen
sulfate ([C1Him][HSO4], IL2). These ILs have previously been
proposed as solvents, both for their interesting physical pro-
perties15 and their potential use as acid catalysts.16 The reac-
tion details and products characterizations are included in the
ESI.†
Process modelling and design
Implementation methods. The modelling and simulation of
the IL production process was conducted using Aspen Plus
In this paper, we evaluate the commercialization potential V7.3 with full details reported in the ESI.† The basis of the
of two ‘protic’ ILs (acid–base complexes), which have great conventional industrial scale process involves a rather literal
potential in the biomass deconstruction field.13 In order to scale-up of the lab process and constitutes a large continuous
achieve this aim, we have developed conceptual process stirred tank reactor which requires significant dilution to avoid
models of IL production processes and analysed the key thermal runaway; followed by flashing of the dilution water.
factors (process indicators including process configurations We also evaluated an intensified process which takes advan-
and operating conditions) that impact the cost price of ILs. tage of developments in high surface area flow reactors which
The results indicate that the cost of starting materials is the allow high heat transfer rates and effective cooling. The inten-
largest contributor to the cost price of the ILs studied. Our sified process was modelled as a reactor train with interstage
models also reinforce the conclusion that some ILs are not cooling. This process does not need any additional dilution.
necessarily too expensive for large volume applications, and The results of the simulations including the specifications
even can be as inexpensive as conventional organic solvents.
(pressure, temperature, composition, etc.) of the process
streams, the required heat removal in each stage, and the
required power for pumping are provided in the ESI.† These
results were later used as inputs to Aspen Exchanger Design
and Rating in order to design the compact plate reactors,
again with associated results provided in the ESI.†
Results and discussion
Ionic liquids synthesis
The ILs studied below are made by combination of a Brønsted
As discussed earlier the associated reactions are highly
acid with a Brønsted base (‘protic’ ILs14). In this preparation exothermic and occur very fast. Therefore, tight control of the
process, stoichiometric amounts of acid and base are mixed temperature of the reaction mixture is crucial. However,
together to form the salt. ILs are formed when a proton from because of the large amount of reaction heat, it is not poss-
the acid is transferred onto the base. Generally, in protic ionic ible to remove the heat using an exchanger embedded in the
liquids the heteroatom (N, P or S) of the cation is charged by reactor. Therefore, the reaction mixture is conventionally
protonation. This reaction releases extreme heat and typically diluted by a large volume (often several times larger than the
is very fast, making this type of reaction difficult to control original reaction mixture) of a diluting medium in order to
using standard batch procedures. On a laboratory scale, the control the temperature rise. Such a process configuration is
acid agent is usually added drop-wise to the amine base in a shown in Fig. 2.
vessel designed to avoid hot-spots and to ensure a constant
Under this process option, the reactants are fed in stoichio-
reaction temperature. In the authors’ laboratory, the reagents metric amounts, according to eqn (a) and (b) in Fig. 1.
were also diluted in water and the reactions were cooled in an In addition, a large volume of diluting water is added in
ice bath. The presence of water removes the excess heat order to control the temperature rise. A fraction of the diluting
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 3098–3106 | 3099

View Article Online
Paper
Green Chemistry
Fig. 2 Process flow diagram for the conventional process.
water (about 20 wt%) is necessary in order to reduce the The aqueous IL is then heated (using steam) in a heat exchan-
viscosity of the IL product for storage and transportation. ger, before being fed into a flash drum. Inside the flash drum,
However, the extra diluting water needs to be separated and the pressure is lowered by volume expansion, leading to the
evaporated from the mixture. Then, the IL product is cooled excess water being boiled off as steam. The IL product is recov-
and sent to the storage and the evaporated water is condensed ered from the bottom of the flash drum, while the steam is
and recycled for reuse in the process.
cooled in a heat exchanger (using cooling water) and then fed
Process flow diagrams (PFDs) are a simple diagrammatic back into the initial dilution mixer for the acid (it is recycled).
representation of all of the unit operations contained within a Fig. 3 (see description below) is similar to Fig. 2, except that a
process. Fig. 2 and 3 show two PFDs for IL synthesis routes 4-stage reactor is employed (see ESI† for reactor details).
(discussed below), each containing slightly different levels of
The study of the conventional process, shown in Fig. 2,
complexity. Fig. 2 shows the sulfuric acid reagent being suggests that separation and recycling of the diluting water
diluted with water in a vented mixing drum (vented to relieve imposes significant energy penalties, which will represent a
pressure build-up from excess heat of mixing), and this significant added cost. Therefore, in the present research with
mixture is fed into a reactor where it is mixed with the amine. the aim of reducing the energy requirements and simplifying
Fig. 3 Process flow diagram for the intensified process.
3100 | Green Chem., 2014, 16, 3098–3106
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
the temperature control, an intensified process flow diagram Process economics assessment
was developed and compared to the aforementioned process.
The ultimate purpose for developing this process design and
simulation model is to estimate the production cost at indus-
trial scale of ILs and to evaluate the economic feasibility of IL
production. We therefore performed an economic assessment
of each proposed process in order to evaluate both the capital
investment and manufacturing costs required to produce these
ILs at bulk scale.
The process is evaluated for a 10-year project life time
(selected as a short capital repayment time, with no interest),
assuming the plant to be operational for 330 days per year,
equivalent to 7920 operating hours per year. The plant capacity
is designed as 144 000 tons per year, which is a suggested
design capacity for an IL-based biomass pre-treatment
process.10 The construction year is assumed to be 2013. In
accordance with common practice in most process economic
evaluations in the public domain, all the costs provided in this
study are given in 2013 US$.
The configuration of the intensified process is shown in Fig. 3.
In this process the sulfuric acid is considered as the limiting
reactant and fed gradually to the reaction mixture. We
assumed that this reaction is fast in comparison to the rate of
addition, as there is no kinetic data reported in the open litera-
ture and it is difficult to measure any finite reaction rate for an
acid–base neutralization. Therefore, the new process diagram
was simplified and each reaction stage is assumed to consist
of two steps: an adiabatic reactor and a cooler. In the reactor,
all of the sulfuric acid (i.e., the limiting reactant) is completely
consumed and the evolved heat of the exothermic reaction
causes a temperature rise. Since the reactor is assumed to be
adiabatic, the temperature of the reactant effluent is the
highest temperature rise that can occur in each stage. By
choosing the correct value for the maximum allowable temp-
erature it is possible to ensure that no phase separation or
degradation of products will occur, based on the knowledge of
the system phase behaviour. The main problem of phase sep-
aration arises from the relatively high melting point of triethyl-
ammonium hydrogen sulfate (85 °C);15 therefore the
operating temperature was kept higher than the melting
point of this IL in order to ensure that no phase separation
could occur.
Total capital investment
The total capital investment (TCI) consists of the fixed capital
investment (FCI), the working capital cost and start-up cost.
The latter two items are dependent upon the FCI. TCI esti-
mation has been described in a number of engineering
textbooks.17–19 The matter of which method is “correct” is of
course open for debate. However, at the level of early-stage esti-
mation employed here, no single methodology has a clear
Process design specifications
Table 1 shows the process design specifications. In the present advantage. In this study, we used the methodology from Peters
research, the maximum allowable temperature was set to be et al.19 There are many costs required to estimate the TCI;
95 °C. The process pressure is considered to be 4 bar through- however, most of these costs can be related directly to cost of
out the process diagram. Therefore, the temperature and equipment. Therefore, the cost of equipment was determined
pressure specifications provide the safe margins from any first. Note that each piece of equipment has a purchase cost
phase separation or runaway reaction. The outlet temperature dependent upon time; the methodology estimates costs based
of the inter-stage coolers was set to 50 °C in order to maintain on 2002 prices. Since we set the construction year as 2013,
the desirable mixture viscosity, which is below 3 cP.
these costs required adjustment for inflation. This was
The heat removal capacity (heat transfer area) is over- achieved by using the following expression:
ꢀ
ꢁ
designed to be 100% larger than the values calculated by the
model. It is proposed that this level of over-design will com-
pensate for any uncertainties in the thermo-physical pro-
perties, which are often difficult to obtain for ionic liquids.5
I2
I1
C2 ¼ C1
where C is the purchase cost, I is the cost index, subscript 1
In the simplified flow sheet each stage consists of an adia- refers to the base time when the initial cost was calculated and
batic reactor and an inter-stage cooler; the aim of this was to subscript 2 refers to the desired time for the cost to be esti-
determine the maximum temperature rise and the number mated. Equipment costs were adjusted using Chemical Engin-
of required stages. In reality, the reaction and heat removal eering Plant Cost Index (CEPCI). The information about the
can occur in the same piece of equipment. In the present size of each item was obtained from the simulation results
model, a compact plate reactor for each stage is employed, as and cost calculations based on Peters et al.19 The result is
explained in the next section and detailed in the ESI.†
shown in Table 2 for both process options. The reactor cost is
Table 1 Design specifications
Description
Specification
Description
Specification
Production rate kg h−1 (ionic liquid kg h−1
Concentration of water in IL product
Interstage cooler outlet temperature
Reactor outlet temperature
)
22 730 (18 184)
20% wt
>50
Over design (heat removal)
Process pressure
Allowable pressure drop in each stage
Mixture viscosity
100%
<4 bar
−0.3 bar
<3 cP
<95
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 3098–3106 | 3101

View Article Online
Paper
Green Chemistry
Table 2 Equipment costs calculation
Cost (millions of dollars)
Item
Description
IL1 intensified
IL1 conventional
IL2 intensified
IL2 conventional
Reactor
Storage tank
Pumps
IL preparation
Storage of starting materials and ILs
Pumping water and raw materials
1.23 × 10−1
2.10 × 10−1
3.26 × 10−2
3.66 × 10−1
1.59 × 10−1
2.10 × 10−1
4.35 × 10−2
4.13 × 10−1
1.16 × 10−1
2.10 × 10−1
2.18e × 10−2
3.48 × 10−1
1.59 × 10−1
2.10 × 10−1
4.35 × 10−2
4.13 × 10−1
Cost
30% lower for the intensified process than for the convention- Manufacturing costs
al process, leading to a lower capital investment for the inten-
sified process.
The cost of manufacturing (COM) associated with the day-to-
day operation of a plant is the other cost source. The elements
that influence COM can be divided into three categories: direct
manufacturing costs (DMC), fixed manufacturing costs (FMC)
and general expenses (GE).20 DMC represents operating
expenses that vary with production rate. For example, raw
materials and operating labour will be lowered when the pro-
duction rate is reduced. FMC are independent of changes in
production rate. It includes depreciation, taxes, insurance and
overhead costs. GE includes management, sales, financing and
research functions, all of which are necessary to carry out
business functions. These three items constitute the total COM:
Once the total equipment cost is determined, TCI can be
calculated through the use of various factors. Techno-econ-
omic reports usually draw on market research in order to esti-
mate a competitive selling price for products. The suggested
selling price is then set in order to draw conclusions about the
economic viability of the selected technology or a new product.
However, since there is neither a market for IL1 or IL2 at
present, nor
a commercial IL1 or IL2 source with a
specified price, our analysis instead takes the approach of esti-
mating the production cost based on a minimum acceptable
economic result – the return on investment for the plant.
Table 3 summarizes the selected categories, additional factors
and costs information. It is clear that the TCI of the intensified
process is lower than for the conventional process, mainly due
to the significantly lower equipment costs for the intensified
process.
COM ¼ DMC þ FMC þ GE
COM can be calculated when the following costs are
determined:
(a) fixed capital investment (FCI); (b) cost of operating
labour (COL); (c) cost of utilities (CUT); (d) cost of waste treat-
Table 3 Total capital investment and start-up costs calculation
Cost (millions of dollars)
Factors (% of purchased
of equipment)
IL1
intensified
IL1
conventional
IL2
intensified
IL2
conventional
Cost items
Direct costs
Onsite costs
Purchased equipment
Purchased-equipment installation
Instrumentation and control
Piping
100%
40%
18%
20%
12%
3.66 × 10−1
1.46 × 10−1
6.59 × 10−2
7.32 × 10−2
4.39 × 10−2
4.13 × 10−1
1.65 × 10−1
7.43 × 10−2
8.26 × 10−2
4.96 × 10−2
3.48 × 10−1
1.39 × 10−1
6.26 × 10−2
6.96 × 10−2
4.18 × 10−2
4.13 × 10−1
1.65 × 10−1
7.43 × 10−2
8.26 × 10−2
4.96 × 10−2
Electrical equipment and materials
Offsite costs
Buildings (process and auxiliary buildings,
maintenance shops, building services)
Yard improvements
Service facilities
Total direct costs
30%
1.10 × 10−1
1.24 × 10−1
1.04 × 10−1
1.24 × 10−1
6%
20%
246%
2.20 × 10−2
7.32 × 10−2
9.00 × 10−1
2.48 × 10−2
8.26 × 10−2
1.02
2.09 × 10−2
6.96 × 10−2
8.56 × 10−1
2.48 × 10−2
8.26 × 10−2
1.02
Indirect costs
Engineering and supervision
Construction expenses
Contractor’s fee
Contingency
Total indirect costs
26%
15%
16%
30%
87%
9.52 × 10−2
5.49 × 10−2
5.86 × 10−2
1.10 × 10−1
3.18 × 10−1
1.07 × 10−1
6.19 × 10−2
6.61 × 10−2
1.23 × 10−1
3.59 × 10−1
9.05 × 10−2
5.22 × 10−2
5.57 × 10−2
1.04 × 10−1
3.03 × 10−1
1.07 × 10−1
6.19 × 10−2
6.61 × 10−2
1.23 × 10−1
3.59 × 10−1
Fixed capital investment (FCI)
333%
1.22
1.38
1.16
1.38
Working capital (15% of the total capital investment)
Total capital investment (TCI)
Start-up costs (10% of FCI)
59%
392%
33.3%
2.16 × 10−1
1.43
2.44 × 10−1
1.62
2.05 × 10−1
1.36
2.44 × 10−1
1.62
1.22 × 10−1
1.38 × 10−1
1.16 × 10−1
1.38 × 10−1
3102 | Green Chem., 2014, 16, 3098–3106
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
ment (CWT); (e) cost of raw materials (CRM). The cost for each It was also found that the intensified process uses less electri-
of the three categories can be determined as follows:
city and cooling water than the conventional process. This is
mainly due to the plate reactors that are used possessing
higher heat removal efficiency.
DMC ¼ CRM þ CWT þ CUT þ 1:33COL þ 0:069FCI þ 0:03COM
Table 6 summarizes each individual item of COM and the
calculation information for these. As it can be seen, the DMC
is the largest part of the manufacturing cost and the reason for
this is the high cost of raw materials. Fig. 4 exhibits the cost
distribution of each component, i.e. DMC, FMC and GE and
TCI. It shows that DMC dominates the total cost in both scen-
arios, representing 82.4% and 83%, respectively. Moreover, raw
materials accounts for almost 99% of DMC. As discussed
before, this corresponds to the simplicity of the ILs process,
resulting in low cost of process equipment and therefore small
TCI value (0.8% and 0.3%, respectively).
FMC ¼ 0:708COL þ 0:068FCI þ depreciation ð0:1FCIÞ
GE ¼ 0:177COL þ 0:009FCI þ 0:16COM
Therefore, COM = 0.28FCI + 2.73COL + 1.23(CUT + CWT + CRM).
FCI determination is outlined above. A description of the
COL calculation methods is provided in the ESI.†
CUT is directly influenced by the cost of electricity and
cooling water in the current system. The cost values of electri-
city and cooling water listed in Table 4 are adopted from
Turton et al.,20 as recently described by Benavides et al.21
Table 5 illustrates the annual operating cost determined on
the basis of the simulation results. The prices of the raw
materials were obtained from ICIS and from estimates pro-
vided by BASF. It can be seen that raw materials costs contrib-
ute the most to the total IL synthesis cost. This is due to the
simplicity of these IL synthesis routes and therefore a low cost
of utilities and operating labour. In reality, most of the ILs can
be synthesized via one or two steps; for example, only one reac-
tion step is involved in the quaternisation and alkylation pro-
cesses, and the metathesis process would introduce two steps.
For the ILs purification (separation) process, most can be
easily purified via extraction or recrystallization. In this case,
the aim of lower operating costs can be achieved as long as
affordable starting materials are utilized, as demonstrated by
the dominant role of amine prices in the final cost estimates.
Ionic liquids cost
The production cost of ILs, calculated on the basis of the
above investments, has been used in this study to represent a
final cost price. It is defined as the selling price of ILs required
to ensure the net present value of the ILs production process
equals zero within a return period over the life of the plant (10
years in the present study). It therefore refers to the ILs price
at the break-even point where annual costs and income are
equal at this price. In the intensification scenario, the cost
prices of IL1 and IL2 are $1.24 kg−1 and $2.96–5.88 kg−1
,
respectively (the price of 1-methylimidazole raises consider-
able uncertainty as it is not presently produced at this scale).
On the basis of the above modelling and economic assessment
results, one can estimate the cost prices for other types of ILs
which are made via acid–base neutralization and quaterniza-
tion reactions as follows:
Table 4 Operating costs
M1P1 þ M2P2
Cost item
Cost (U.S. $)
ILs price ¼
ꢀ 1:25
M1 þ M2
Raw materials
Triethylamine
1-Methylimidazole
Sulfuric acid
Water
Operators wage
Utilities
Electricity
1920 ton−1
where M1 and M2 are the molecular weights of the two starting
materials while P1 and P2 are the price of the two starting
materials. This assumes that the raw materials costs will domi-
nate the final cost price of the ILs, as in the present example.
It also highlights that, due to the 1 : 1 stoichiometry inherent
to salt formation, the cost price of protic ILs will always be
determined by the molecular weight of the more expensive com-
5300–10 640 ton−1
80 ton−1
0.067 ton−1
50 000 per year
0.06 kWh−1
Cooling water
Wastewater treatment
0.354 GJ−1 (14.8/1000 m3)
56/1000 m3
Table 5 Annual operating costs calculation
Annual cost (millions of dollars)
Description
IL1 intensified
IL1 conventional
IL2 intensified
IL2 conventional
Raw materials
Triethylamine (1-Methylimidazole)
Sulfuric acid
Water
Operation labour
Utilities
Electricity
Cooling water
Wastewater treatment
146.08
140.41
5.67
146.08
140.41
5.67
353.99–704.34
347.72–698.07
6.27
353.99–704.34
347.72–698.07
6.27
2.41e × 10−3
6.00 × 10−1
6.81 × 10−2
4.69 × 10−3
6.34 × 10−2
0
2.41e × 10−3
6.00 × 10−1
8.57 × 10−1
7.81 × 10−3
8.49 × 10−1
0
2.42 × 10−3
6.00 × 10−1
6.83 × 10−2
3.82 × 10−3
6.45 × 10−2
0
2.42 × 10−3
6.00 × 10−1
8.57 × 10−1
7.81 × 10−3
8.49 × 10−1
0
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 3098–3106 | 3103

View Article Online
Paper
Green Chemistry
Table 6 Manufacturing costs calculation
Cost (millions of dollars)
Cost item
Multiplying factor
IL1
intensified
IL1 non-
intensified
IL2
intensified
IL2 non-
intensified
1. Direct manufacturing costs (DMC)
Raw materials
Waste treatment
Utilities
Operating labour
Direct supervisory and clerical labour
Maintenance and repairs
Operating supplies
Laboratory charges
Total direct manufacturing cost
CRM
CWT
CUT
COL
0.18COL
0.06FCI
0.009FCI
0.15COL
146.08
0
146.08
0
353.99–704.34
0
353.99–704.34
0
6.81 × 10−2
6.00 × 10−1
1.08 × 10−1
7.32 × 10−2
1.10 × 10−2
9.00 × 10−2
147.03
8.57 × 10−1
6.00 × 10−1
1.08 × 10−1
8.28 × 10−2
1.24 × 10−2
9.00 × 10−2
147.83
6.83 × 10−2
6.00 × 10−1
1.08 × 10−1
6.96 × 10−2
1.04 × 10−2
9.00 × 10−2
354.93–705.93
8.57 × 10−1
6.00 × 10−1
1.08 × 10−1
8.28 × 10−2
1.24 × 10−2
9.00 × 10−2
355.74–706.09
CRM + CWT + CUT + 1.33COL
0.069FCI + 0.03COM
+
2. Fixed manufacturing costs (FMC)
Depreciation
Local taxes and insurance
Plant overhead costs
0.1FCI
0.032FCI
0.708COL + 0.036FCI
0.708COL + 0.068FCI + depreciation
1.22 × 10−1
3.90 × 10−2
4.68 × 10−1
6.29 × 10−1
1.38 × 10−1
4.42 × 10−2
4.75 × 10−1
6.56 × 10−1
1.16 × 10−1
3.71 × 10−2
4.26 × 10−1
5.79 × 10−1
1.38 × 10−1
4.42 × 10−2
4.75 × 10−1
6.56 × 10−1
Total fixed manufacturing costs
3. General manufacturing expenses (GE)
Administration costs
Distribution and selling costs
Research and development
Total general manufacturing costs
Total manufacturing costs
0.177COL + 0.009FCI
0.11COM
0.05COM
0.177COL + 0.009FCI + 0.16COM
DMC + FMC + GE
1.17 × 10−1
19.99
9.08
29.19
176.85
1.19 × 10−1
19.99
9.09
29.20
177.69
1.17 × 10−1
1.19 × 10−1
48.08–95.31
21.86–43.32
70.05–138.75
425.59–845.26
48.23–95.64
21.92–43.47
70.27–139.23
426.67–845.98
Cost sensitivity
There are many factors that can affect our cost estimate; we
identified two parameters likely to exert significant influence
on the final cost price of these ILs: plant capacity and water
concentration in the IL product. In order to estimate the
impact of these two variables on the final cost of the ILs, we
calculated the impact of changes in these variables on the final
IL price and the influence of raw materials costs under each scen-
ario. The results of these sensitivity calculations are presented
in Table 7, for the intensified process only. For these calcu-
lations, we altered the model to include drying each IL to 1%
final water content (vs. 20% water) for three different plant
Fig. 4 Cost distribution of the intensified processes for IL1 (left) and IL2
(right) preparation. (DMC: direct manufacturing costs; FMC: fixed manu-
facturing costs; GE: general manufacturing expenses; TCI: total capital
investment).
ponent. For these [HSO4]-based examples, this inevitably leads capacities (144 000, 14 400 and 1440 tons per year).
to the conclusion that smaller amines – not always cheaper It is clear from the table that the extra drying (to 1% water
ones – will yield less expensive ILs. Thus, triethylammonium content) has no noticeable effect on the final cost price of IL
hydrogen sulfate will cost less to produce than triethanolammo- production. Whilst it is unlikely that ILs would be dried to this
nium hydrogen sulfate, despite the lower cost (per kg) of the level in an actual process (the viscosity penalty would be prohi-
latter amine. This will obviously reverse if the acid is the more bitive), we feel this is an important variable to analyse, and
expensive component (i.e., triflic acid); in such cases, a larger our calculations will represent a conservative estimate of
amine will yield a less expensive IL, in the majority of cases.
actual costs. The drying step under our scenario is actually a
The low production cost of the triethylammonium hydro- more extensive flash process (we are above 100 °C here), and
gen sulfate IL ($1.24 kg−1) compares favourably with the therefore contributes negligible energy and cost (less than
selling price of conventional organic solvents such as acetone $20 kg−1 of water removed). The plant capacity has a much
($1.32 kg−1) or ethyl acetate ($1.39 kg−1) according to ICIS.11 more marked effect – a small plant (1% capacity of the original
These ILs are much less expensive than higher-end organic model) will have a much higher relative operating cost (raw
solvents, such as acetonitrile ($1.54 kg−1) and are close to the materials drop from contributing 82.6% to 50.2% of the cost),
price range of low cost solvents such as toluene ($1.03 kg−1). leading to a 60% increase in total IL cost. This suggests that
This strongly suggests that cost considerations of bulk ionic the scale of IL production should be considered when estimat-
liquid production can be less intimidating than traditionally ing the optimal size of an IL-based processing plant, such as a
assumed.
biorefinery.
3104 | Green Chem., 2014, 16, 3098–3106
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Table 7 Cost sensitivity estimation
Plant capacity
(ton per year)
IL1 cost kg−1
(20% water)
IL1 cost kg−1
(1% water)
IL2 cost kg−1
(20% water)
IL2 cost kg−1
(1% water)
Raw material
contribution
1.44 × 105
1.44 × 104
1.44 × 103
$1.24
$1.31
$2.02
$1.24
$1.31
$2.03
$2.96–5.88
$3.11–6.21
$4.79–9.57
$2.96–5.88
$3.11–6.21
$4.80–9.59
82.6%
77.9%
50.2%
Green metrics
One final aspect of note is the overall ‘green-ness’ of protic IL
synthesis. Since these ILs are created from a one-step acid–
base neutralisation, they produce less waste than other IL
syntheses. In the present example, the atom economy for the
IL synthesis is 100% – indeed, use of excess base will ensure
that there can be no separable waste from the reaction (due to
the second acidic proton’s ability to form additional cations).
This is not possible for traditional dialkylimidazolium ILs,
which will always have lower atom economies through the pro-
duction of salt waste during the metathesis step1 (e.g.
[C2C1im][BF4] synthesized from the halide intermediate would
have an atom economy of 93%; ILs made from the methyl-
sulfate intermediate will have much lower atom economies).22
The E-factor for our process is likewise negligible – the only
waste product is distilled water from the flash drying. High
E-factors plague dialkylimidazolium IL synthesis, where
E-factors are often unity or greater for each reaction step.22
Finally, due to the exothermic nature of protic IL synthesis, the
energy inputs are also negligibly low for this process route.
These metrics for this process compare favourably to the syn-
thesis of traditional ILs, where a trade-off is often observed
between atom economy and E-factor.22
These metrics would be similar for all protic ionic liquids,
regardless of the nature of the constituent ions. However, our
selection of simple tertiary amines and sulfuric acid also
reduces the complexity of the synthesis of the reagents.
Jessop23 pointed out that the number of synthetic steps in a
solvent will be a dominant factor on the environmental impact
of the solvent, as more synthetic steps yield more waste and
larger energy usage. Jessop also points out that most dialkyl-
imidazolium ionic liquids require ca. 30 synthetic steps from
raw materials (e.g. [C4C1im][BF4] will require 32), and even
simple ILs, such as [C4C1im]Cl require 22 steps. This is not
entirely tied to the fluorination of anions – even [C2C1im]-
[OAc], which is well-studied as a solvent for biomass appli-
cations, requires 29 steps to make!
Fig. 5 Synthesis tree for IL1 from raw materials.
lium ILs would fail all nine.23 The reduced impact from chemi-
cal synthesis of reagents cannot be overlooked as a green
metric for solvent selection – reducing the size of the synthesis
tree by employing mineral acids and simple tertiary amines
can greatly improve the green credentials and reduce the total
environmental impact of an IL down to the level of common
organic solvents. It is no accident that the cost of production
should be linked to the size of the synthesis tree, as in the
current example – more synthetic steps will increase solvent
cost alongside waste production and energy usage. Green sol-
vents must be simple to manufacture!
Conclusions
We have estimated the production cost of two protic ionic
liquids – triethylammonium hydrogen sulfate and 1-methyl-
imidazolium hydrogen sulfate. The simpler ammonium salt
was determined to cost just $1.24 kg−1 to produce, while the
latter imidazolium complex would be $2.96 kg−1. This differ-
ence illustrates our finding that raw material costs (of the
amine in particular) dominate sulfuric acid-based ionic liquid
preparation.
To achieve this goal, ILs manufacturing process models
were implemented for the first time using ASPEN software. An
economic assessment of IL production plants was performed
based on the simulation models. The results show that some
ILs can be as cheap as conventional organic solvents, such as
acetone or ethyl acetate, and may even compete with low-cost
solvents such as toluene. Alongside this reduced cost, the
environmental impact of these simple ILs will be similarly
reduced. This result could direct future development of ILs for
large-scale bulk applications, where more efforts should be
concentrated on developing new ILs which can be synthesized
Fig. 5 displays the synthesis tree for making [HNEt3][HSO4].
This IL requires only 7 steps from raw materials (oil, N2, H2,
O2, S8, H2O). This is a similar number of steps to most organic
solvents (e.g. THF: 7; ethyl acetate: 8) and this large reduction
in synthetic complexity will reduce both the energy required
and waste produced during solvent manufacture. Based on an
LCA performed by Bakshi et al.,24 Jessop recommended nine
simple questions to ask about a solvent synthesis tree to assess
the ‘green-ness’ of the solvent. [HNEt3][HSO4] appears to pass
seven of these tests definitively, while most dialkyl imidazo-
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 3098–3106 | 3105

View Article Online
Paper
Green Chemistry
from affordable raw materials in very few steps. The techno-
economic analysis of other types of ILs is presently underway
in our group.
R. T. J. Porter, M. Pourkashanian, G. T. Rochelle, N. Shah,
J. G. Yao and P. S. Fennell, Energy Environ. Sci., 2014, 7,
130–189.
We also compared an intensified process model with a
more conventional process to evaluate the economic advan-
tages available through process intensification. It was found
that the intensified process reduces the cost of ILs, and should
be utilized in future development efforts.
7 A. Brandt, M. J. Ray, T. Q. To, D. J. Leak, R. J. Murphy and
T. Welton, Green Chem., 2011, 13, 2489–2499.
8 H. Tadesse and R. Luque, Energy Environ. Sci., 2011, 4,
3913–3929.
9 D. Groff, A. George, N. Sun, N. Sathitsuksanoh,
G. Bokinsky, B. A. Simmons, B. M. Holmes and
J. D. Keasling, Green Chem., 2013, 15, 1264–1267.
10 D. Klein-Marcuschamer, B. A. Simmons and H. W. Blanch,
Biofuels, Bioprod. Biorefin., 2011, 5, 562–569.
Acknowledgements
L. Chen gratefully acknowledges the China Scholarship 11 All web-based pricing (http://www.sigmaaldrich.com or
Council for providing funding. Additional support from the
http://www.icis.com) were accessed in November 2013.
UK Engineering and Physical Sciences Research Council 12 N. V. Plechkova and K. R. Seddon, Chem. Soc. Rev., 2008,
(EPSRC grant EP/K014676/1), the UK Natural Environment
37, 123–150.
Research Council (NERC grant NE/H01392X/1) and grant 13 A. Brandt, J. Gräsvik, J. P. Hallett and T. Welton, Green
309102 under the EU FP7 Agreement is also gratefully acknowl-
Chem., 2013, 15, 550–583.
edged. We would also like to thank BASF for providing esti- 14 T. L. Greaves and C. J. Drummond, Chem. Rev., 2008, 108,
mates for the bulk production price of 1-methylimidazole. JH
206–237.
is especially grateful to Professor Philip Jessop for helpful dis- 15 J.-P. Belieres and C. A. Angell, J. Phys. Chem. B, 2007, 111,
cussions about green synthesis.
4926–4937.
16 Z. S. Qureshi, K. M. Deshmukh, M. D. Bhor and
B. M. Bhanage, Catal. Commun., 2009, 10, 833–837.
17 K. K. Humphreys and P. Wellman, Project and Cost Engin-
eers’ Handbook, Marcel Dekker, New York, 3rd edn, 1996.
18 J. M. Douglas, Conceptual Design of Chemical Processes,
McGraw Hill, New York, 1988.
Notes and references
1 J. P. Hallett and T. Welton, Chem. Rev., 2011, 111, 3508–
3576.
19 M. S. Peters, K. D. Timmerhaus and R. E. Wrest, Plant
Design and Economics for Chemical Engineers, McGraw Hill,
New York, 5th edn, 2003.
20 R. Turton, R. C. Bailie, W. B. Whiting, J. A. Shaeiwitz and
D. Bhattacharyya, Analysis, Synthesis and Design of Chemical
Processes, Prentice Hall, Upper Saddle River, NJ, 4th edn,
2012.
21 P. T. Benavides, J. Salazar and U. Diwekar, AIChE J., 2013,
32, 11–24.
22 M. Deetlefs and K. R. Seddon, Green Chem., 2010, 12,
17–30.
2 H. Olivier-Bourbigou and L. Magna, J. Mol. Catal., 2002,
182–183, 419–437.
3 M. A. Gebbie, M. Valtiner, X. Banquy, E. T. Fox,
W. A. Henderson and J. N. Israelachvili, Proc. Natl. Acad.
Sci. U. S. A., 2013, 110, 9674–9679.
4 N. Mac Dowell, N. Florin, A. Buchard, J. Hallett, A. Galindo,
G. Jackson, C. S. Adjiman, C. K. Williams, N. Shah and
P. Fennell, Energy Environ. Sci., 2010, 3, 1645–1669.
5 M. Ramdin, T. W. de Loos and T. J. H. Vlugt, Ind. Eng.
Chem. Res., 2012, 51, 8149–8177.
6 M. E. Boot-Hanford, J. C. Abanades, E. J. Anthony,
M. J. Blunt, S. Brandani, N. Mac Dowell, J. R. Fernández,
M.-C. Ferrari, R. Gross, J. P. Hallett, R. S. Haszeldine,
P. Heptonstall, A. Lyngfelt, Z. Makuch, E. Mangano,
23 P. G. Jessop, Green Chem., 2011, 13, 1391–1398.
24 Y. Zhang, B. R. Bakshi and E. S. Demessie, Environ. Sci.
Technol., 2008, 42, 1724–1730.
3106 | Green Chem., 2014, 16, 3098–3106
This journal is © The Royal Society of Chemistry 2014