Low toxicity functionalised imidazolium salts for task specific ionic liquid electrolytes in dye-sensitised solar cells: A step towards less hazardous energy production

Article abstract of DOI:10.1039/c3gc42393j

Novel solvent free task specific ionic liquid (TSIL) electrolytes for dye sensitised solar cells (DSSC) were synthesised and tested. Of great concern is the replacement of low-moderate toxicity second generation ILs, with high toxicity third generation TSILs. As most 1-butyl-3-methylimidazolium (Bmim) and especially 1-ethyl-3-methylimidazolium (Emim) based ILs have low toxicity, the designing of replacement TSILs of comparable toxicity is a challenge. Structural features of TSIL investigated herein were incorporation of heteroatoms into the side chain of imidazolium cations (i.e. ether, ester and amide) and anion (bromide, iodide, and triflimide [NTf₂]). Preliminary toxicity screening against 20 microorganisms (8 bacteria and 12 fungi) found that all ILs, imidazolium salts, N-butylbenzimidazole (NBB) and guanidinium thiocyanate (GNCS) do not exhibit high antimicrobial toxicity. However NBB and a pentyl ester substituted IL displayed moderate toxicity to several strains of bacteria and fungi. Further toxicity testing to establish IC₅₀ values shows several novel TSIL compounds and imidazolium salts are in fact less toxic to microorganisms (e.g. bacteria) than commonly used 1-ethyl-3-methylimidazolium iodide (EmimI) and 1,3-dimethylimidazolium iodide (DmimI). We have demonstrated that the presence of ether and either ester or amide groups in the structure of the cation of the TSIL and imidazolium salts reduces antimicrobial toxicity, which is consistent with the lowering of the lipophilicity of ILs. Iodide and bromide analogues have lower toxicity than the NTf₂ examples in this study. The DSSC performance using these "greener" ILs in place of the standard EmimI compare quite favourably. Two low antibacterial toxicity iodide examples exhibit photocurrents of 9.27 mA cm^-2 and 8.85 mA cm ^-2, respectively, achieving promising efficiencies of 3.39% and 3.31%, respectively (EmimI = 4.94%). DSSC performance is further improved by 15% minimum to 66% maximum, depending on IL chosen, by the presence of small amounts of moisture and DSSCs employing a low antibacterial toxicity iodide TSIL or imidazolium salt can surpass the performance of dry EmimI. Of note the DSSC containing TSIL NTf₂ examples, performed poorly compared to the halide analogues, with the outcome that the most toxic TSILs under investigation are also the least preferred based on performance. the Partner Organisations 2014.

Full text of DOI:10.1039/c3gc42393j

Green Chemistry
View Article Online
View Journal | View Issue
PAPER
Low toxicity functionalised imidazolium salts for
task specific ionic liquid electrolytes in
dye-sensitised solar cells: a step towards less
hazardous energy production†
Cite this: Green Chem., 2014, 16,
2252
Mukund Ghavre,a,b Owen Byrne,c Lena Altes,c Praveen K. Surolia,c Marcel Spulak,d
Brid Quilty,e K. Ravindranathan Thampi*b,c and Nicholas Gathergood*a,b
Novel solvent free task specific ionic liquid (TSIL) electrolytes for dye sensitised solar cells (DSSC) were
synthesised and tested. Of great concern is the replacement of low-moderate toxicity second generation
ILs, with high toxicity third generation TSILs. As most 1-butyl-3-methylimidazolium (Bmim) and especially
1-ethyl-3-methylimidazolium (Emim) based ILs have low toxicity, the designing of replacement TSILs of
comparable toxicity is a challenge. Structural features of TSIL investigated herein were incorporation of
heteroatoms into the side chain of imidazolium cations (i.e. ether, ester and amide) and anion (bromide,
iodide, and triflimide [NTf2]). Preliminary toxicity screening against 20 microorganisms (8 bacteria and 12
fungi) found that all ILs, imidazolium salts, N-butylbenzimidazole (NBB) and guanidinium thiocyanate
(GNCS) do not exhibit high antimicrobial toxicity. However NBB and a pentyl ester substituted IL displayed
moderate toxicity to several strains of bacteria and fungi. Further toxicity testing to establish IC50 values
shows several novel TSIL compounds and imidazolium salts are in fact less toxic to microorganisms (e.g.
bacteria) than commonly used 1-ethyl-3-methylimidazolium iodide (EmimI) and 1,3-dimethylimidazolium
iodide (DmimI). We have demonstrated that the presence of ether and either ester or amide groups in the
structure of the cation of the TSIL and imidazolium salts reduces antimicrobial toxicity, which is consistent
with the lowering of the lipophilicity of ILs. Iodide and bromide analogues have lower toxicity than the
NTf2 examples in this study. The DSSC performance using these “greener” ILs in place of the standard
EmimI compare quite favourably. Two low antibacterial toxicity iodide examples exhibit photocurrents of
9.27 mA cm−2 and 8.85 mA cm−2, respectively, achieving promising efficiencies of 3.39% and 3.31%,
respectively (EmimI = 4.94%). DSSC performance is further improved by 15% minimum to 66% maximum,
depending on IL chosen, by the presence of small amounts of moisture and DSSCs employing a low anti-
bacterial toxicity iodide TSIL or imidazolium salt can surpass the performance of dry EmimI. Of note the
DSSC containing TSIL NTf2 examples, performed poorly compared to the halide analogues, with the
outcome that the most toxic TSILs under investigation are also the least preferred based on performance.
Received 21st November 2013,
Accepted 30th January 2014
DOI: 10.1039/c3gc42393j
www.rsc.org/greenchem
Introduction
Over the last decade, ionic liquids (ILs) have been extensively
aSchool of Chemical Sciences and National Institute of Cellular Biotechnology,
Dublin City University, Glasnevin, Dublin 9, Ireland. E-mail: nick.gathergood@dcu.ie
bSFI Strategic Research Cluster in Solar Energy Conversion, UCD School of Chemical
and Bioprocess Engineering, University College Dublin, Belfield, Dublin 4, Ireland
cSchool of Chemical & Bioprocess Engineering, Engineering Building,
University College Dublin (UCD), Belfield, Dublin 4, Ireland.
investigated as potential replacements for volatile organic com-
pounds (VOCs) for use as (inter alia) as tuneable reaction
media.1 Much of this interest has been focussed on the develop-
ment of ILs as alternative, ‘green’ materials; with applications
in processes as diverse as ionic compressors, the BASILTM
process and electroplating.1q Immense interest in the environ-
mental impact of ILs2 has led to a plethora of papers dealing
with three assays: (1) their toxicity3 (for example antibacterial
and antifungal),4 (2) the importance of biodegradation studies5
(something only recognised since 2002),6 and recently (3) bio-
accumulation and metabolite identification studies.7
E-mail: ravindranathan.thampi@ucd.ie
dFaculty of Pharmacy, Charles University, Heyrovského 1203, 500 05 Hradec Králové,
Czech Republic
eSchool of Biotechnology and National Institute of Cellular Biotechnology, Dublin
City University, Glasnevin, Dublin 9, Ireland
†Electronic supplementary information (ESI) available. See DOI:
10.1039/c3gc42393j
2252 | Green Chem., 2014, 16, 2252–2265
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
The design of a ‘green’ compound, be it as a solvent,1 (including Gathergood4,6,15 and Stephens16) as the starting
reagent or catalyst8 should ideally address issues such as low point for screening ILs, due to the high toxicity of many analo-
toxicity and ready biodegradability without the generation of gous Quarternary Ammonium Compounds (QACs), especially
toxic, persistent metabolites. Of equal importance is the func- surfactants.3c,17,18 Antibacterial and antifungal toxicity studies
tional performance of the environmentally benign material. of novel ILs are a rapid and convenient approach to investigate
The decision to replace a ‘toxic’ chemical with a ‘greener’ the IL effect on microorganisms important to our natural
alternative is easier if a performance benefit is also attained. environment.
The role of a green chemist (in our view) is to make this
In this study, we have designed novel ester and amide
decision as easy as possible and to avoid the ‘gray area’ where based ILs with two aims (1) efficient synthesis of third gen-
environmental protection comes at a performance cost. One eration TSILs containing functional groups known to reduce
application of ILs is a solvent in electrolytes for dye sensitised antimicrobial toxicity;4,6,15,16 and (2) study the effect on DSSC
solar cells (DSSC).1s,t
properties of replacing a second generation IL with a low anti-
A typical DSSC consists of a TiO2 layer electrode chemi- microbial toxicity third generation TSIL. The attempts to eli-
sorbed with a monolayer of dye molecules absorbing the minate liquid or gel electrolytes completely from high
visible light spectrum. Up on absorbing light the excited dye performance and stable DSSC is not yet successful. Therefore,
injects electrons into the conduction band of the TiO2, which first generation commercial DSSC, which requires good stabi-
are then routed through an external circuit and a counter elec- lity, will rely mostly on IL based electrolytes and hence its
trode into an electrolyte containing a suitable redox species toxicity and environmental acceptance assumes utmost impor-
(typically, iodide/tri-iodide couple). The electrolyte is an essen- tance. As the introduction of heteroatoms into the IL structure
tial component, performing charge transport between the two is fundamental to lowering antimicrobial toxicity, we can
electrodes. Iodide/tri-iodide electrolytes are the best performing assess the effect due to the presence of these H-bond donor/
examples utilising volatile solvents (e.g. >11% power conver- acceptor groups on DSSC performance.
sion efficiency achieved in acetonitrile–valeronitrile),9,10 with
the current DSSC record of 12.3% efficiency which involves a
Co(II/III)tris(bipyridyl) tetracyanoborate complex as the redox
Experimental
couple in acetonitrile.11 A significant problem of DSSC sealant
failure occurs facilely due to the volatile organic solvent’s large IL synthesis
vapour pressures. Electrolytes with very low vapour pressures
All starting chemicals were purchased from Sigma Aldrich,
can be envisaged from ILs or eutectic melts of ionic com-
pounds. These classes of salts solve this problem, to a signifi-
cant extent, and show excellent stability under light soaking at
60 °C for up to 1000 h.12,13 Record efficiencies of 8.2%13 and
7.6%12 are achieved with DmimI/EmimI/EMITCB/I2/NBB/
GNCS eutectic melt and PMII/EMITCB/I2/NBB/GNCS,
respectively.
with the exceptions of lithium bis(trifluoromethanesulfonyl)
imide (LiNTf2) which was obtained from Solvionic. Methanol,
hexane and triethylamine were dried over molecular sieves,
diethyl ether was dried over sodium metal wire, dichloro-
methane (DCM) was dried over calcium hydride. All dry sol-
vents were distilled before use. Riedel de Haën silica gel was
used for flash and thin layer chromatography. Sodium carbon-
ate, ammonium chloride, calcium chloride anhydrous were
obtained from Riedel de Haën and magnesium sulphate
heptahydrate from Fluka.
The design of ILs has progressed through three gener-
ations.1 The first generation included examples with reactive
−
and/or water sensitive ions (e.g. AlCl4 salts). Stability was
improved in the second generation (e.g. BmimBF4 and
BmimOAc), while TSILs (Task Specific Ionic Liquids) are the
third generation. When conventional electrolyte media were
replaced with second generation ILs, discussion about IL toxi-
city, ecotoxicity and biodegradation was often lacking. This is
more apparent with third generation TSILs and their role in
DSSCs, where toxicity data for novel ‘tailored’ ILs is often not
reported as a part of the green chemistry assessment. As ILs
are considered as a ‘greener’ alternative to conventional sol-
vents, it is important to investigate the toxicity and ecotoxicity
of these chemicals on ecological systems and the fate of ILs
due to potential accidental release or other exposures to
environment.14 Of great concern is the replacement of low-
ILs: 1,3-dimethylimidazolium iodide (1, DmimI), 1-ethyl-3-
methylimidazolium iodide (2, EmimI), 1-ethyl-3-methylimida-
zolium tetracyanoborate (3, EMTCB), and N-butylbenzimida-
zole (4, NBB) were purchased from Merck as Solarpur grade
(>99.5% by HPLC). Guanidinium thiocyanate (5, GNCS) and I2
were purchased from Sigma-Aldrich.
Eleven ester and amide based imidazolium salts were syn-
thesized according to procedures developed by Gathergood
et al.4,15i,19 (Fig. 1, 6–9a; 6b, 8–9b, 6–9c) (see ESI† for experi-
mental procedure).
Karl Fischer analysis
moderate toxicity second generation ILs, with high toxicity Mettler Toledo V20 Compact Volumetric KF Titrator using
third generation TSILs. As most Bmim and especially Emim Hydranal 5 titrant was employed for water content estimation.
based ILs have low toxicity, the challenge of designing replace- For each sample, a water content value was obtained by Karl
ment TSILs of comparable toxicity is great. Antimicrobial toxi- Fischer titration according to the general procedure outlined
city studies have been previously proposed by several groups below:
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2252–2265 | 2253

View Article Online
Paper
Green Chemistry
Fig. 1 Ester and amide based ILs used for screening in DSSC.
Solid samples: 100 mg of the sample to be analysed was Antifungal toxicity screening procedure (broad spectrum)
weighed on an analytical balance in a weighing boat. The
Four ATCC strains (CA1: Candida albicans ATCC 44859, CA2:
sample was transferred to the automatic titrator’s reaction
Candida albicans ATCC 90028, CP: Candida parapsilosis ATCC
vessel from the weighing boat without the aid of any solvents
22019, CK1: Candida krusei ATCC 6258) and eight clinical iso-
lates of yeasts (CK2: Candida krusei E28, CT: Candida tropicalis
or spatula to avoid introduction of additional moisture or acci-
dental sample loss via the spatula. The weight of the boat after
156, CG: Candida glabrata 20/I, CL: Candida lusitaniae 2446/I,
the transfer was recorded to get the exact sample mass used
TA: Trichosporon asahii 1188) and filamentous fungi (AF: Asper-
gillus fumigatus 231, AC: Absidia corymbifera 272, TM: Tricho-
phyton mentagrophytes 445) were studied from the collection of
fungal strains cultured at the Department of Biological and
for the analysis. The moisture content results from titration
were expressed in % (w/w).
Liquid samples: 0.1 mL of the sample to be analysed was
drawn into a 1 mL syringe and weighed on an analytical
Medical Sciences, Faculty of Pharmacy, Charles University,
balance. The sample was transferred from the syringe to the
Hradec Králové, Czech Republic. Three of the above ATCC
automatic titrators reaction vessel. The weight of the syringe
strains (C. albicans ATCC 90028, C. parapsilosis ATCC 22019,
C. krusei ATCC 6258) also served as the quality control strains.
after the transfer gave the exact sample mass being analysed.
The water content results were expressed in % (w/w).
All the isolates were maintained on Sabouraud dextrose
agar prior to being tested. MICs were determined by the micro-
dilution format of the NCCLS M27-A guidelines.20 Dimethyl
sulfoxide (100%) served as a diluent for all compounds; the
final concentration did not exceed 2%. RPMI 1640 (Seva-
pharma, Prague) medium supplemented with L-glutamine and
buffered with 0.165 M morpholinepropanesulfonic acid (Serva)
to pH 7.0 by 10 N NaOH was used as the test medium. The
wells of the microdilution tray contained 100 μL of the RPMI
1640 medium with 2-fold serial dilutions of the compounds
(2000 or 1000 to 0.48 μmol L−1) and 100 μL of inoculum sus-
pension. Fungal inoculum in RPMI 1640 was prepared to give
a final concentration of 5 × 103 0.2 cfu mL−1. The trays were
incubated at 35 °C and MICs were read visually for filamentous
fungi and photometrically for yeasts as an absorbance at
540 nm after 24 h and 48 h. The MIC/IC50 values for the der-
matophytic strain (T. mentagrophytes) were determined after
72 h and 120 h and for A. fumigatus, A. corymbifera after 24 and
48 h. For all other strains MIC/IC80 values were evaluated. The
MICs were defined as 50% or 80% inhibition of the growth of
control. MICs were determined twice and in duplicate. The devi-
ations from the usually obtained values were no higher than the
nearest concentration value up and down the dilution scale.
Antibacterial toxicity screening procedure (broad spectrum)20
Three CCM strains (SA: Staphylococcus aureus CCM 4516, EC:
Escherichia coli CCM 4517, PA: Pseudomonas aeruginosa CCM
1961) and five clinical isolates (MRSA: Staphylococcus aureus
MRSA H 5996/08, SE: Staphylococcus epidermidis H 6966/08,
EF: Enterococcus sp. J 14365/08, KP-E: Klebsiella pneumoniae
ESBL J 14368/08, KP: Klebsiella pneumoniae D 11750/08) were
studied from the collection of bacterial strains cultured at the
Department of Biological and Medical Sciences, Faculty of
Pharmacy, Charles University, Hradec Kralove, Czech Republic.
The aforementioned CCM strains also served as the quality
control strains.
All the isolates were maintained on Mueller-Hinton dex-
trose agar prior to being tested. Dimethyl sulfoxide (100%)
served as a diluent for all compounds with the final concen-
tration never exceeding 2%. Mueller-Hinton agar (MH,
HiMedia, Čadersky-Envitek, Czech Republic) buffered to pH 7.4
( 0.2) was used as the test medium. The wells of the micro-
dilution tray contained 200 μL of the Mueller-Hinton medium
with 2-fold serial dilutions of the compounds (2000 or 1000 to
0.48 μmol l−1) and 10 μL of inoculum suspension. Inoculum in
MH medium was prepared to give a final concentration of 0.5
McFarland scale (1.5 × 108 cfu mL−1). The trays were incubated
at 36 °C and Minimum inhibitory concentrations (MICs) were
Antibacterial toxicity screening method (IC50 determination,
5 strains)
read visually after 24 h and 48 h. The MICs were defined as Mueller-Hinton broth was purchased from Oxoid. Five bacteria
95% inhibition of the growth of control. MICs were deter- strains were used in this study: the Gram-positive bacterium
mined twice and in duplicate. The deviations from the usually Bacillus subtilis DSMZ 10 (B. subtilis) and the Gram-negative
obtained values were no higher than the nearest concentration bacteria Escherichia coli DSMZ 498 (E. coli), Pseudomonas
value up and down the dilution scale.
fluorescens DSMZ 270 50090 (P. fluorescens), Pseudomonas
2254 | Green Chem., 2014, 16, 2252–2265
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
putida CP1 (P. putida CP1) and Pseudomonas putida KT2440 All preparation steps were performed under argon atmosphere,
(P. putida KT2440). All strains were purchased at DSMZ since most of the compounds are highly hygroscopic.
(German Collection of Microorganisms and Cell Cultures)
except P. putida CP1 which was sourced in the laboratory at
Dublin City University.
Results and discussion
IC50 values for the compounds were determined using a
modification of the broth microdilution method described by Imidazolium derived ionic liquids were selected for study: (1)
Amsterdam.21 Strains were grown in nutrient broth overnight, due to outstanding performance of this class as electrolytes in
washed with 0.01 M sodium phosphate buffer (pH 7) and the DSSC,12,13 and (2) a plethora of toxicity data of imidazolium
cell number adjusted to give an optical density reading of 0.07 ILs,14–18 including known high toxicity for some long alkyl
at 660 nm. The antimicrobial activity of the ILs was tested in chain examples,3a,25 thus assisting a ‘benign by design’ ideal.
96 well round bottom microplates. 180 μL of Mueller-Hinton
Considering the toxicity results of previously reported ILs
broth was pipetted into column 1 of the wells and 100 μL into from our group4,15i and Boethling’s rules of thumb,24 we
the other wells. 20 μL of the chemical solution was transferred designed ILs herein (Fig. 1) which incorporate ester and amide
into column 1 giving a concentration of 200 mM. 100 μL of the functionalities and additional heteroatoms (e.g. ether groups)
solution from column 1 was then transferred to the next in the structure. Long hydrocarbon side chains were avoided
column and mixed. The procedure was repeated to give a as Pernak et al.3a and Kanjilal et al.25 have disclosed that com-
series of two-fold dilutions. Each well was inoculated with 5 μL pounds with high antimicrobial toxicity include this structural
of bacterial culture. Wells containing medium only were used motif.
as blanks and wells containing medium and culture only were
7a–c was chosen to enable study of the inclusion of an ester
used as positive controls. All the toxicity tests were carried out group into the alkyl imidazolium substituent, while 6a–c has
in triplicate. The microplates were incubated overnight at in addition 2 ether functional groups. ILs 8a–c are amide
30 °C. The presence or absence of growth was determined by derivatives with 2 ether groups, and 9a–c incorporate an
measuring the optical density of the wells at a wavelength of amide and single ether, conformationally restricted due to
405 nm using a plate reader. The IC50 values were determined 6 membered ring). Iodides (6c–9c) were selected, so study of
as the concentration or range of concentrations that caused a the effect of the imidazolium cation (cf. Emim iodide, in high
50% reduction in growth.
performing DSSC) was possible, while bromide derivatives
(6a–9a) were also prepared as this is a frequently studied
alternative in this field of work. NTf2 examples (6b–9b) were
Fabrication of DSSC
Full DSSC fabrication and measurement details are available selected due to the low viscosity observed for this class of ILs
in ESI.† Briefly, DSSC working electrodes22,23 (WE) consisted and stability in the presence of the iodide/tri-iodide redox
of a compact blocking layer of TiO2, seven layers of a transpar- couple, a fundamental component of the DSSC.
ent TiO2 paste and two layers of a TiO2 scattering paste fol-
The designed ILs (6a–c to 8a–c and 9b) and imidazolium
lowed by a compact TiO2 over-layer yielding a total TiO2 salts (9a and 9c) were synthesized according to our reported
thickness of 15 micron, after sintering, with an active area of procedures with moderate overall yields (i.e. 6a–c = 47%,
0.28 cm2. The WEs were placed in a dye bath of N719 overnight 7a–c = 42%, 8a–c = 35%, 9a–c = 39%).4,15i,19
in order to sensitise the TiO2 surface. A platinum counter elec-
Synthesis of ILs 6a, 7a and 8a were successfully completed
trode was then sandwiched together with the WE and heat according to methods reported in our previous papers4,15i
sealed using a Bynel® polymer gasket (50 micron thick). The whereas novel NTf2 ILs 6b, and 8b were prepared according to
specified electrolyte was then filled into the space between the other patented procedures.19 ILs 6a–8b were synthesized in
two electrodes through a hole in the counter electrode via two or three steps: (i) preparation of ester or amide alkylating
vacuum back filling. The back hole was sealed with a thin agent (10, 98%; 11 64%; 12, 76%) (Fig. 2); (ii) synthesis of
piece of glass heat sealed with Bynel®.
bromide IL (6a, 90%; 7a 97%; 8a, 83%) and (iii) halide anion
exchange on reaction with LiNTf2. (6b, 80%; 8b, 74%) (see
ESI† for experimental procedure).
Electrolyte fabrication
All ILs and imidazolium salts were dried for 24 h under high
Scheme 1 shows a representative synthesis of the bromide
vacuum (0.05 mbar) and water content determined by Karl and iodide imidazolium salts and NTf2 IL, starting with mor-
Fischer analyses. The electrolyte composition DmimI/X/ pholine. α-Bromoamide 13 was formed in good yield (85%) via
EMITCB/I2/NBB/GNCS was prepared in the molar ratio 12/12/ a condensation reaction of morpholine with bromo acetyl
16/1.67/3.33/0.67 as a homogenous mixture using the follow- bromide. Synthesis of bromide 9a, a 100% atom economy26
ing procedure. Initially, DmimI (1) was dissolved in EMITCB addition reaction of methylimidazole and 13, is achieved in
(3) at 60 °C on a hot plate while stirring. After approximately 85% yield. Counterion metathesis with LiNTf2 in water gave
1 h a clear yellow solution was obtained. Then the respective the NTf2 based IL 9b in 62% yield.
novel ILs (X = 2, 6a–c to 9a–c) were added. After 30 min NBB
Iodides 6c, 8c and 9c were prepared by anion exchange of
(4), GNCS (5) and I2 were added successively when dissolved. A corresponding [NTf2] salts (i.e. 6b, 8b and 9b) respectively on
dark red homogenous solution of low viscosity was obtained. treatment with tetrabutylammonium iodide (TBAI). For the
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2252–2265 | 2255

View Article Online
Paper
Green Chemistry
Fig. 2 α-Bromoester, α-bromoamide and α-iodoester halide IL precursors.
Scheme 1 Synthesis of IL 9c from morpholine.
morpholino based imidazolium salt 9c, the procedure involves 1500–6000 ppm, except for 9a (10 700 ppm). Increasing the
addition of a solution of TBAI in dichloromethane to a drying time for 9a did not further reduce the water content.
dichloromethane solution of the [NTf2] salt 9b.27 After stirring
No thermal degradation was observed by NMR under these
the reaction mixture for 3 h at room temperature the product drying conditions. Seo et al. reported the use of peptide based
was extracted into an aqueous layer. Removal of water affords TSILs which were stable up to 200 °C in DSSCs.29 Also no
the corresponding iodide 9c in 87% yield (39% yield, from degradation was observed when ester based imidazolium
α-bromoamide 13). This procedure has an unsatisfactory high TSILs were screened in DSSCs by Wang et al.30 By analogy we
E-factor,28 therefore, an alternative greener procedure was also do not expect significant thermal degradation of ester and
developed where the iodide based alkylating agent (14) was amide based TSIL and imidazolium salts (Fig. 1) under the
synthesized from 2-iodoacetyl chloride (yield 68%), and in the test conditions, however, the eutectic mixture may behave
final step the target iodide IL 6c was prepared from 14, in 82% differently at elevated temperature in the cell.
yield.4 IL synthesis was also demonstrated on a large scale
with 155 g (ca. 0.2 Mol) of 6b prepared in 40% overall yield
from the alcohol starting material via the alkylating agent 10 =
Toxicity
In our study of DSSCs, we have replaced the 2° generation low-
68%, Br IL 6a = 73%, and NTf2 IL 6b = 80% yield. During the
moderate toxicity ILs with 3° generation TSILs which have a
work up, ILs were warmed to 60 °C under high vacuum
comparatively high molecular weight. This is a potential ‘catch
(0.05 mbar) to reduce moisture content.
22’ situation – many TSIL tailored properties are due to the
presence of functional group(s) in the side chain, which are
either an integral part of the side chain (i.e. amino acid
Karl Fischer analysis
The lowest water content after drying for 24 h under high sequence,15k or functional groups appended to a hydrocarbon
vacuum (0.05 mbar) was attained with the commercially avail- chain. If one assumes Emim (or even Bmim) as the model
able iodides EmimI (2) and DmimI (1) (200 and 700 ppm, compound for IL studies, addition of functional groups to
respectively) (Table S1†). NBB (4) under the same conditions convert them into TSILs will increase MW. The toxicity of ILs
had a water content of 5100 ppm. Novel ILs and imidazolium analogues with a long hydrocarbon side chain (i.e. without het-
salts 6–9c water content after drying was between eroatoms) were found to exhibit high antimicrobial activity
2256 | Green Chem., 2014, 16, 2252–2265
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
due to high lipophilicity.3c,18 Hence, by the introduction of
heteroatoms into the hydrocarbon alkyl chain, the lipophilicity
of ILs is reduced (with expected decrease in antimicrobial toxi-
city).4 Our hypothesis was supported by the recent work of
Samori et al.,31 Gathergood4 and Costa Gomez et al.5e who state
that ‘the lipophilicity of ILs decreases when the alkyl side-
chain contain oxygen functionalities, and simultaneously
reduces the toxicity of ILs’. 6b and 7b were two of the ILs
under investigation.5e Over the previous decade the link
between antimicrobial toxicity and side chain length of imida-
zolium ILs has been extensively reported.3c,4,15i,18,32 However,
as this is related to the lipophilicity of the IL, introduction of
heteroatoms would be expected to reduce antimicrobial toxi-
city via this mode of action. Every TSIL is unique (and often
complex compared to DmimI) and toxicity to microorganisms
via a different mode of action is feasible. Therefore, antimicro-
bial screening of all ILs utilised in this project was performed.
Furthermore, the authors proposed that the introduction of
oxygen-functionalised side chains (ester) can increase the bio-
degradability of ILs via hydrolysis.5e This offers a potential
advantage when future ecological and environmental impact
studies of TSILs are performed.
Table 1 Antibacterial toxicity study results for ILs, imidazolium salts,
4 and 5
MIC IC95 (mM)
1–3, 5, 6a, 6c,
7a, 8a–c, 9a–c
Strainsa
SA
Time (h)
4
6b
7c
24
48
24
48
24
48
24
48
24
48
24
48
24
48
24
48
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
2
2
2
2
2
2
1
2
>2
>2
>2
>2
>2
>2
>2
>2
>1
>1
>1
>1
>2
>2
>2
>2
>2
>2
2
>2
>2
>2
>2
>2
>2
>2
>2
>2
MRSA
SE
0.5
>1
>1
>1
>1
>1
1
>1
>1
>1
>1
>1
EF
EC
KP
KP-E
PA
a SA: Staphylococcus aureus CCM 4516, EC: Escherichia coli CCM 4517,
PA: Pseudomonas aeruginosa CCM 1961) and five clinical isolates
(MRSA: Staphylococcus aureus MRSA H 5996/08, SE: Staphylococcus
epidermidis H 6966/08, EF: Enterococcus sp. J 14365/08, KP-E: Klebsiella
pneumoniae ESBL J 14368/08, KP: Klebsiella pneumoniae D 11750/08.
Antibacterial toxicity preliminary screen
In the DSSC screening, we have utilised eleven amide and ester
based ILs and imidazolium salts (Fig. 1) and three commer-
cially available ILs. Eight strains of bacteria were selected for
the study to represent a wide range of different classes of bac-
teria. In vitro antibacterial activities20 (IC95) of ILs and addi-
tives NBB (4) and GNCS (5) were evaluated.
Table 2 Antifungal toxicity study results for ILs, imidazolium salts,
4 and 5
MIC IC80/IC50 b (mM)
1–3, 5, 6a, 6c, 7a,
Strainsa
CA1
CA2
CP
Time (h)
8a–c, 9a–c
4
6b
7c
Table 1 shows that eleven ILs (1–3, 6a, 6c, 7a, 7c, 8a–c and
9b) and two imidazolium salts (9a and 9c) were found to not
have high antibacterial toxicity, up to the maximum test con-
centration validated (2 mM). IL 7c was moderately toxic to one
of the eight bacteria strains screened (Enterococcus sp.). IL 6b
was tested at 1 mmol due to limited solubility in the broth. 6b
does not have high toxicity (>1 mM) except for S. epidermidis.
N-Butylbenzimidazole (4, NBB) and guanidinium thiocyanate
(5, GNCS) are used as additives in the electrolyte. NBB shows
moderate toxicity for all four Gram-positive bacteria (IC95
2 mM) while not exhibiting high toxicity to Gram-negative
strains (IC95 > 2 mM). This is a lipophilic compound and can
cross the membrane of cell, leading to enhanced biological
activity.33 Overall, Table 2 shows that the ILs and imidazolium
salts utilised for the DSSC screening are not highly toxic to a
wide range of strains of bacteria.
24
48
24
48
24
48
24
48
24
48
24
48
24
48
24
48
24
48
24
48
24
48
72
120
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
2
2
2
2
2
2
2
2
1
2
2
2
2
2
2
2
1
1
2
2
2
2
2
2
1
>1
>1
>1
>1
>1
>1
>1
>1
>1
1
>1
>1
>1
>1
>1
>2
>2
>2
>2
>2
>2
2
CK1
CK2
CT
>2
2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
>2
CG
CL
TA
0.5
>1
>1
>1
>1
>1
>1
>1
AF
AC
Antifungal toxicity preliminary screen
TM
Fungi belongs to a large group of eukaryotic micro-organisms
and are distinguished among this group due to their chitin
containing cell wall. Fungi toxicity is often assessed in tandem
with bacteria toxicity to investigate compounds antimicrobial
biological effect.5
The antifungal toxicity results (Table 2) indicate that ten ILs
(1, 2, 3, 6a, 6c, 7a, 8a–c and 9b), two imidazolium salts (9a and
9c) and 5 do not exhibit high toxicity to the twelve strains of
a (CA1: Candida albicans ATCC 44859, CA2: Candida albicans ATCC
90028, CP: Candida parapsilosis ATCC 22019, CK1: Candida krusei
ATCC 6258) and eight clinical isolates of yeasts (CK2: Candida krusei
E28, CT: Candida tropicalis 156, CG: Candida glabrata 20/I, CL: Candida
lusitaniae 2446/I, TA: Trichosporon asahii 1188) and filamentous fungi
(AF: Aspergillus fumigatus 231, AC: Absidia corymbifera 272, TM:
Trichophyton mentagrophytes 445). b IC50 values were assessed for AF,
AC and TM. For all other fungi strains IC80 values were evaluated.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2252–2265 | 2257

View Article Online
Paper
Green Chemistry
fungi, up to maximum validated test concentration limit 50–100 mM, while even lower toxicity >200 mM for the Gram
(2 mM). 6b was tested up to 1 mM concentration due to positive bacteria strain Bacillus subtilis was found.
limited solubility in RPMI 1640 broth. 6b has antifungal
A general trend is the iodide series 6c–9c is more toxic to
activity towards T. asahii (IC80 0.5 mM), and moderate activity bacteria than the bromides 6a–9a, although the difference is
towards C. albicans, C. tropicalis (IC80 1 mM). Table 3 also marginal (only for 1 out of 5 strains screened) for 8c and 9c.
shows NBB (4) exhibits moderate activity to all strains of fungi. Significantly, the bromides (8a and 9a) and iodides (8c and 9c)
Although, the cell wall of fungi is different (composed of have lower antibacterial toxicity than commercially available
chitin) to bacteria, we submit that as NBB is a lipophilic widely applied ILs 1 and 2. TSIL 6a exhibits similar antibacter-
neutral compound, it can penetrate this barrier more easily ial toxicity to 1 and 2.
than the ILs (Fig. 1).
The [NTf2]− series 6b, 8b and 9b is more toxic to the 5 bac-
ILs and imidazolium salts selected for DSSC screening teria strains than the bromide analogues, with the increase in
(Fig. 1), including 1, 2 and 3 do not show high antimicrobial toxicity more prominent for the amide TSIL 8b and 9b than
toxicity (Tables 1 and 2). However, these preliminary results do ester 6b. While the amide examples 8a,c and 9a,c have very
not establish the toxicity limit for ILs, hence the toxicity study low antibacterial toxicity (IC50 values >200 mM), exchanging
was expanded to higher concentrations i.e. 200 mM, solubility the halide for the [NTf2]− lead to ILs with the greatest toxicity
limits withstanding. All compounds were screened against five to bacterial, within the scope of this study (Fig. 1 and Table 4).
bacteria strains: Gram-positive (Bacillus subtilis) and Gram- This demonstrates the importance of toxicity screening of indi-
negative (Escherichia coli, Pseudomonas fluorescens, Pseudomo- vidual compounds, as the increase in antibacterial toxicity of
nas putida CP1, Pseudomonas putida KT2440). Compound toxi- the ester TSIL 6a to the 6b is only slight (i.e. no clear trend
city determination was based on bacterial growth inhibition in when comparing the effect of anion on toxicity for the ester
a 24 hour assay and was expressed as IC50 value range. The TSILs to the amide TSILs series). However results in Table 4,
maximum test concentration for ILs (1, 2, 3, 6a, 6c, 7a, 7c, 8a, from an intelligent design perspective, highlight lipophilic ILs
8c and 9b) and imidazolium salts (9a and 9c) was 200 mM (NTf2 examples and alkyl ester TSILs) in general are more toxic
whereas ILs (6b and 8b) were tested at 50 mM. The lower test to bacteria than the hydrophilic ILs (halide examples and
concentration for (6b and 8b) was necessary due to the limited incorporating ether function groups into sidechain).
solubility of ILs in the broth. The IC50 value ranges are pres-
DSSC performance (dry)
ented in Table 3.
Table 3 shows that of the commercially available ILs (1–3)
and electrolyte additives (4 and 5), 4 has the highest toxicity to
Gram positive and Gram negative bacteria. These results are
consistent with the preliminary antibacterial toxicity study
(Table 2). The iodide based ILs 1 and 2 have IC50 values in the
ranges 50–100 or 100–200 mM depending on the bacteria
strains. EMTCB (3) has moderate toxicity (IC50 values
6.25–25 mM) to the 5 bacteria strains. GNCS (5) shows low
activity towards Gram negative bacteria in the range
The low toxicity of some the TSILs and imidazolium salts pre-
pared (particularly 6a, 8a, 8c, 9a and 9c) compared to standard
2 suggested they would be more suitable, greener and safer for
use as DSSC electrolytes. We fabricated DSSC with the best
reported IL combination13 [1/2/3/I2/4/5 in molar ratio 12/12/16/
1.67/3.33/0.67] electrolyte to act as a standard comparison for
the new IL electrolyte materials tested here.
We achieve an efficiency of 4.94% with this standard
EmimI (2) combination, which is lower than the 8.2% reported
Table 3 Antibacterial toxicity study at higher concentrations. IC50 values
IC50 value (mM)
Compound
E. coli
B. subtilis
P. fluorescens
P. putida (CP1)
P. putida (KT2440)
1
2
3
4
100–200
100–200
12.5–25.0
0.78–1.56
50–100
50–100
25–50
25–50
25–50
25–50
>200
12.5–25
>200
>200
6.25–12.5
>200
50–100
50–100
6.25–12.5
1.56–3.12
>200
50–100
25–50
50–100
25–50
6.25–12.5
>200
25.50
>200
>200
6.25–12.5
>200
50–100
50–100
6.25–12.5
1.56–3.12
50–100
50–100
25–50
50–100
25–50
6.25–12.5
>200
6.25–12.5
100–200
>200
6.25–12.5
100–200
50–100
50–100
12.5–25
6.25–12.5
50–100
50–100
25–50
50–100
25–50
25–50
>200
>50
>200
>200
6.25–12.5
>200
100–200
50–100
12.5–25
6.25–12.5
50–100
100–200
>50
50–100
25–50
50–100
>200
>50
>200
>200
12.5–25
100–200
5
6a
6b
6c
7a
7c
8a
8b
8c
9a
9b
9c
2258 | Green Chem., 2014, 16, 2252–2265
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Table 4 Photovoltaic parameters of the DSSC devices made with DmimI/X/EmimTCB/I2/NBB/GNCS electrolyte measured at 1 sun (1000 W m−2
)
incident intensity of AM1.5 solar light. Their performance when freshly made and after one month exposed to ambient i.e. following moisture
ingress, are listed
As fabricated (dry)
After one month back seal open to allow moisture entry
Compound (X)
Jsc (mA cm−2
)
Voc (mV)
FF (%)
η (%)
Jsc (mA cm−2
)
Voc (mV)
FF (%)
η (%)
2
11.5
8.95
7.23
7.19
10.4
8.83
8.85
6.30
6.97
9.27
6.97
5.15
669
622
638
625
675
655
620
662
630
667
640
660
64
53
52
62
56
65
62
60
60
53
52
72
4.94
2.92
2.38
2.79
3.92
3.76
3.39
2.51
2.61
3.31
2.30
2.46
12.5
12.4
9.74
10.2
11.8
13.0
11.6
8.80
10.7
11.0
9.95
9.56
725
663
661
656
661
698
669
703
689
675
665
716
70
57
60
55
60
48
66
59
59
58
53
57
6.33
4.71
3.87
3.69
4.63
4.33
5.08
3.67
4.34
4.34
3.53
3.90
6c
6a
6b
7c
7a
8c
8a
8b
9c
9a
9b
for similar electrolytes and it is due to the use of different TiO2 superior performance, as can be observed in Table 4. Iodide
paste, dye and fabrication procedures employed here. TSILs better performance is due to significantly higher cur-
−
However, within this study it can act as a benchmark for com- rents than the corresponding bromide or triflimide (NTf2
)
paring the performance of the new electrolytes. The perform- salts. Cross couple redox systems such as these are compli-
−
ance results of DSSC fabricated with either our new ILs or cated. Interhalogen redox systems based on the IBr2 and the
imidazolium salts replacing 2 are given in Table 4 and Fig. 3.
I2Br− anions are difficult to characterize due to the complex
All iodide salts (6c to 9c) achieve current densities greater equilibrium between various interhalogen anions in the elec-
than 8 mA cm−2. The only bromide or triflimide salt that trolyte solution.34
reaches above 8 mA cm−2 is 7a with a value of 8.83 mA cm−2
.
Furthermore, the triflimide anion is considerably more
The two best performing ILs achieve efficiencies of 3.92% and bulky than I− and Br− and shows high resonance-stability
3.76% for 7c and 7a respectively. 7c and 7a compare well to the resulting in low Lewis basicity. The formation of the iodide-
efficiency achieved with the EmimI (2) standard. 7a exhibits a triflimide and iodide-bromide cross couples is complicated
Jsc of 8.83 mA cm−2 which is much lower than the EmimI (2) and much slower than tri-iodide formation which occurs more
value of 11.5 mA cm−2 but its power conversion efficiency is readily in the iodide ILs. Thus the higher iodide ILs cell cur-
aided by a relatively better fill factor (FF). 7c yields a notably rents may be rationalised by their enhanced tri-iodide
high current density of 10.4 mA cm−2 which is comparable to diffusion via a Grotthus-like bond exchange. It may be
−
2 as displayed in Fig. 4. Slightly lower power conversion expected that in general the NTf2 anion is not as involved in
efficiency is mostly due to the lower FF value compared to 2 cross couple formation, compared to Br−, which could lead to
(64 vs. 56). Both involve the long chain alkyl cation suggesting lower photocurrents and lower charge transfer resistances at
this cation structure performs best of the four tested here, the counter electrode.
although these are also two of the more toxic TSILs. Two of the
other novel imidazolium derived compounds also exhibit note-
Improved performance by presence of water
worthy performance, the lower antibacterial toxicity iodide
TSIL 8c showed 3.39% and imidazolium salt 9c gave 3.31%
power conversion efficiencies. 9c exhibits a high photocurrent
of 9.27 mA cm−2, which is the second highest Jsc value of the
new TSILs and imidazolium salts tested (Fig. 4) suggesting this
structure may be the key towards TSIL for electrolytes with
high current and low toxicity. 8c exhibits a lower Jsc value of
8.85 mA cm−2 but coupled with a good FF results in a similar
power conversion efficiency as 9c. The J–V curves of these best
performing ILs are compared to 2 in Fig. 4(left). Clearly, 7c is
the best, it achieves 80% the performance of 2 with similar
voltage and comparable current density values. 8c and 9c
achieve almost 70% the performance of EmimI (2), but with
much lower toxicity.
It is known that the performance of IL electrolytes in DSSC can
be improved by presence of small amounts of moisture.35
Improvements are brought about by enhancements of Jsc, FF
and Voc which are associated with a decrease in charge transfer
resistances at the counter electrode and increases in ionic con-
ductivities and an increase in the difference between redox
potentials of I−/I3 and the Fermi level of TiO2.35 ILs consist
−
entirely of ion pairs, which differ in degrees of dissociation, on
a case by case basis. As a consequence, formation of aggregates
occurs. With the presence of water, ions can be separated by
the neutral solvent molecules, the number of mobile charge
carriers increases, resulting in an increase of conductivity.36 In
this study we opened the back seal of our DSSCs and stored in
ambient conditions for a period of up to one month to allow
moisture ingress from the atmosphere and observed consider-
able improvement in cell performance. The standard EmimI
We also compared the effect of different counter-ions on
the performance. In all cases the iodide salts (6c–9c) yield
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2252–2265 | 2259

View Article Online
Paper
Green Chemistry
Fig. 3 J–V curves of DSSC fabricated with different electrolyte compositions measured under standard AM1.5 (1000 W m−2) simulated solar light
conditions. Electrolyte compositions consisted of dried DmimI/X/EmimTCB/I2/NBB/GNCS in the molar ratio 12/12/16/1.67/3.33/0.67 (X = 6a–c
to 9a–c).
Fig. 4 (left) J–V curves of DSSC using the best performing novel IL electrolytes compared with the standard EmimI based electrolyte measured
under AM1.5, 1000 W m−2 simulated solar light conditions. (Right) The best performing novel ILs compared to electrolyte containing the standard
EmimI (2) in the presence of small amounts of moisture.
(2) electrolyte improved from 4.94% to 6.33% with improved Seven of the eleven TSILs and imidazolium salts (Fig. 1) now
FF, Jsc and Voc (Table 4). For all compounds tested here, the show efficiencies greater than 4% (Table 4 and Fig. 5).
presence of a small amount of moisture causes significant
The percentage increase in cell parameters in the presence
increase in Jsc with values approaching or exceeding 10 mA of moisture is given in Table S2.† For the iodide salts a slight
cm−2 observed and efficiencies greater than 3.5% in all cases. increase in FF (less than 10%) is observed in each case. This is
2260 | Green Chem., 2014, 16, 2252–2265
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
Fig. 5 J–V curves measured with AM1.5, 1000 W m−2 illumination for DSSC fabricated with our novel electrolytes in the presence of small amounts
of moisture.
coupled with considerable increase in short circuit currents
Recent studies37 reported that water molecules seemed to
with values increasing by 13.5% up to 38.5%. These resulted interact mainly with trifluoromethane-sulfonyl-imide anion
in improved performance efficiencies of up to 60% compared (TFSI−) and Emim and with other water molecules, but not
to their initial dry analogue. The bromides also all showed with I− or I3 implying ILs containing bulky anions such as
−
−
improved Jsc upon presence of moisture with increases of 35% NTf2 should show a higher anion-moisture interaction com-
to 47% observed. This is coupled with small increases in Voc pared to our I− and Br− TSILs. This higher interaction with
values. In some cases FF improved (6a and 9a) but with moisture could result in higher charge transfer resistances at
(7a and 8a) FF decreased. The improved Jsc and Voc increases the counter electrode and thus the observed decreased FF for
−
cell performance efficiencies in all cases by amounts of 15% to the NTf2 ILs tested here. Small amounts of water were found
63%. For the triflimide salts, decreased cell FF is observed but to increase the entropy and mobility37 of Emim and since I−
−
large increases in cell short circuit currents (42% to 85%) lead and I3 are coordinated by several Emim, it leads to higher
to higher power conversion efficiencies with the presence of local fluctuation which should increase the rate of Grotthus-
moisture.
like electron transfer which is diffusion-controlled and can
One might suggest that the improved performance is lead to an improving performance of the DSSC. Therefore, we
caused by the IL cation being attacked by water molecules, expect the crucial factor determining the amount of atmos-
degrading the IL cation. For example, hydrolysis of the ester pheric moisture absorbed and improved DSSC performance
group of 6 and 7 would produce imidazolium carboxylate by- relates to the hydrophilic or hydrophobic nature of the cation.
products which may be the cause of improved performance. Molecular interaction with the cation can include hydrogen
However, if this imidazolium carboxylate was the source of bonding of the acidic proton of the imidazolium ring to the
increased performance then similar improvements for both 6 oxygen atom of water as well as interactions with the substitu-
and 7 would be expected, due to comparable rates of hydrolysis ent on the 1-position of the imidazolium ring, since atoms
for the ester groups. This however is not the case. For with high electro-negativity (oxygen and nitrogen) are present.
example 7c increases in current by 1.4 mA cm−2 but for 6c it is Increasing the number of oxygen and nitrogen atoms,
dramatically different, increasing by more than twice as much, increases the hydrophilicity of the IL via enhanced hydrogen
3.5 mA cm−2
.
bonding interactions with water molecules. The tendency of
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2252–2265 | 2261

View Article Online
Paper
Green Chemistry
the performance of the ILs and imidazolium salts in relation cation reduces antimicrobial toxicity, which is consistent with
to their hydrophilicity increases in the order 7, 9, 8, 6. The the lowering of the lipophilicity of ILs. Iodide and bromide
more oxygen atoms that are incorporated into the chemical TSILs and imidazolium salts have lower toxicity than the NTf2
structure of the TSIL cation (and thus more hydrophilic), the examples in this study.
greater the increase of the performance of the DSSC after
Furthermore, we report the toxicity limits against five bac-
exposure to air. Of the ILs and imidazolium salts in this study, teria strains of common DSSC electrolyte components namely:
7a and 7c show the smallest performance improvement of EmimI, DmimI, 1-ethyl-3-methylimidazolium tetracyanoborate
15% and 18% respectively (Fig. 4) which can be related to their (EmimTCB) and additives N-butylbenzimidazole (NBB) and
greater hydrophobic nature (due to pentyl ester side chain) as guanidinium thiocyanate (GNCS). NBB has the highest toxicity
longer alkyl chains are known to lead to increased hydrophobi- of all compounds in the study with IC50 values in the range
city.38 The standard EmimI (2) with slightly shorter ethyl side 0.78–1.56 mM observed. This significant finding identifies the
chain shows a greater improvement in performance of 28%. additive NBB as the compound of concern based on the scope
The most hydrophilic TSIL cation (6c and 6a) containing four of the toxicity assessment of ILs, imidazolium salts and addi-
oxygen atoms in the side chain shows the greatest increases in tives screened in this study.
performance (61% and 63% respectively).
In addition, while none of the novel TSILs herein, lead to
This trend applies to the I− and Br− containing ILs and imi- exceptional improvement in DSSC performance, they compare
−
dazolium salts, but for NTf2 the trend is not so clear cut. We favorably with EmimI, without a significant increase in anti-
propose this is due to NTf2 interactions with moisture,34 microbial toxicity. In dry conditions the novel TSIL 8c and imi-
−
likely leading to increased Lewis basicity, due to decreased dazolium salt 9c show promise due to their lower toxicity
resonance-stability, which can cause large increases in current compared to EmimI (2) and good performance in terms of Jsc
between the dry and moisture enriched analogues e.g., up to and FF respectively. Upon the addition of moisture to the elec-
85% for 9b.
trolyte system, low toxicity 8c continues to show promise exhi-
Two TSILs of noteworthy performance are 6c and 7c which biting the highest efficiency of the 11 TSIL and imidazolium
show efficiencies of 4.71% and 4.63% respectively. Both salts investigated. This is again due to its excellent FF
exhibit similar Voc and FF values, with 6c displaying a slightly suggesting future TSIL with improved performance should be
higher Jsc as outlined in Fig. 4(right). 7a exhibits a large photo- based on these low toxicity cation structures (8 and 9). In both
current value of 13.0 mA cm−2 which is higher than 2 (12.5 mA dry and moisture containing forms, 70–80% the performance
cm−2) and would be expected to give an outstanding power of EmimI (2) standard is possible but with much lower toxicity.
conversion efficiency except for its poor FF value of 48% but These results show how IL structure influences toxicity and
offers no advantage in terms of toxicity. The best performing performance in DSSC. Also of note, the DSSC containing TSIL
TSIL is the low toxicity iodide salt 8c, which exhibits a per- NTf2 examples, performed poorly compared to the halide ana-
formance of 5.08% which compares favourably to the 6.33% logues, with the outcome that some of the most toxic TSILs
value achieved with 2 and surpasses the performance of the under investigation are also the least preferred based on
dry EmimI (2) analogue.
performance.
The attempts to eliminate liquid or gel electrolytes comple-
tely from high performance and stable DSSC is not yet success-
ful. Therefore, first generation commercial DSSC, which
requires good stability, will rely mostly on IL based electrolytes
Conclusion
As most second generation Bmim and especially Emim based and hence its toxicity and environmental acceptance assumes
ILs have low toxicity, the challenge of designing replacement utmost importance.
third generation TSILs of comparable toxicity is great. Struc-
tural features of TSIL investigated herein (which are known to
reduce antimicrobial toxicity compared to alkyl derivatives)
were incorporation of heteroatoms into the sidechain of imida-
zolium cations (i.e. ether, ester and amide) and anion
(bromide, iodide, and NTf2).
Acknowledgements
Preliminary toxicity screening against 20 microorganisms (8 This work was supported by Enterprise Ireland NILCT project
bacteria and 12 fungi) found that all ILs, imidazolium salts (NG and MG), OB and PKS received support from European
and additives NBB (4) and GNCS (5) do not exhibit high anti- Commission’s FP7 SMARTOP project under the Grant Agree-
microbial toxicity. However NBB (4) and 6b displayed moderate ment number: 265769. K.R. Thampi acknowledges the SFI-Air-
toxicity to several strains of bacteria and fungi. Further toxicity tricity Stokes professorship grant. Lena Altes acknowledges
testing to establish IC50 values shows several novel TSIL and support from PRTLI-5 funding under the UCD Earth Institute’s
imidazolium salts are less toxic to microorganisms (e.g. bac- structured Ph.D. programme. Mukund Ghavre thanks the SFI
teria) than commonly used EmimI and DmimI. We have RC SEC for support. The antibacterial and antifungal screen-
demonstrated that the presence of ether and either ester or ing was supported by the Czech Science Foundation (MS,
amide groups in the structure of the substituted imidazolium project no. P207/10/2048).
2262 | Green Chem., 2014, 16, 2252–2265
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
R.-N. Dou, H.-L. Liu and J. Zhang, Chemosphere, 2011, 82,
1024–1029; (n) E. Liwarska-Bizukojc, Water, Air, Soil Pollut.,
2011, 221, 327–335; (o) Y. Tong, Q. Wang, Y. Bi, M. Lei,
Y. Lv, Y. Liu, J. Liu, L. Lu, Y. Ma, Y. Wu and S. Zhu, Open
Biotechnol. J., 2012, 6, 1–4; (p) S. P. M. Ventura,
C. S. Marques, A. A. Rosatella, C. A. M. Afonso,
F. Gonçalves and J. A. P. Coutinho, Ecotoxicol. Environ. Saf.,
2012, 76, 162–168.
References
1 Selected reviews: (a) T. Welton, Chem. Rev., 1999, 99, 2071–
2084; (b) P. Wassercheid and W. Keim, Angew. Chem., Int.
Ed., 2000, 39, 3772–3789; (c) R. Sheldon, Chem. Commun.,
2001, 2399–2407; (d) J. S. Wilkes, Green Chem., 2002, 4, 73–
80; (e) J. H. Davis and P. Fox, Chem. Commun., 2003, 1209–
1212; (f) R. A. Sheldon, Green Chem., 2005, 7, 267–278;
(g) A. Riisager, R. Fehrmann, M. Haumann and
P. Wasserscheid, Eur. J. Inorg. Chem., 2006, 695–706;
(h) C. Hardacre, J. D. Holbrey, M. Nieuwenhuyzen and
T. G. A. Youngs, Acc. Chem. Res., 2007, 40, 1146–1155;
(i) A. A. H. Padua, M. F. Costa Gomes and J. N. A. Canongia
Lopes, Acc. Chem. Res., 2007, 40, 1087–1096; ( j) X. Han and
D. W. Armstrong, Acc. Chem. Res., 2007, 40, 1079–1086;
(k) J. Ranke, S. Stolte, R. Störmann, J. Arning and
B. Jastorff, Chem. Rev., 2007, 107, 2183–2206; (l) F. van
Rantwijk and R. A. Sheldon, Chem. Rev., 2007, 107, 2757–
2785; (m) M. Smiglak, A. Metlen and R. D. Rogers, Acc.
Chem. Res., 2007, 40, 1182–1192; (n) V. I. Pârvulescu and
C. Hardacre, Chem. Rev., 2007, 107, 2615–2665;
(o) T. L. Greaves and C. J. Drummond, Chem. Rev., 2008,
108, 206–237; (p) M. A. P. Martins, C. P. Frizzo,
D. N. Moreira, N. Zanatta and H. G. Bonacorso, Chem. Rev.,
2008, 108, 2015–2050; (q) Y. Gua and G. Li, Adv. Synth.
Catal., 2009, 351, 817–847; (r) N. V. Plechkova and
K. R. Seddon, Chem. Soc. Rev., 2008, 37, 123–150;
(s) Y. Tomohiro and M. Watanabe, J. Ion Exch., 2011, 22,
58–64; (t) M. Watanabe, Mod. Chem., 2010, 474(9), 50–53;
(u) H. Xu, C. Gros, S. Brandes, P. Ge, H. Girault and
J. Barbe, J. Porphyrins Phthalocyanines, 2011, 15, 560–574.
2 (a) J. Ranke, S. Stolte, R. Stormann, J. Arning and
B. Jastorff, Chem. Rev., 2007, 107, 2183–2206; (b) J. Ranke
and B. Jastorff, Environ. Sci. Pollut. Res. Int., 2000, 7, 105–
114.
4 S. Morrissey, B. Pegot, D. Coleman, M. T. Garcia,
D. Ferguson, B. Quilty and N. Gathergood, Green Chem.,
2009, 11, 475–483.
5 (a) D. Coleman and N. Gathergood, Chem. Soc. Rev., 2010,
39, 600–637; (b) S. Stolte, S. Steudte, A. Igartua and
P. Stepnowski, Curr. Org. Chem., 2011, 15, 1946–1973;
(c) R. S. Boethling, E. Sommer and D. DiFiore, Chem. Rev.,
2007, 107, 2207–2227; (d) M. Markiewicz, S. Stolte,
Z. Lustig, J. Łuczak, M. Skup, J. Hupka and C. Jungnickel,
J. Hazard. Mater., 2011, 195, 378–382; (e) Y. Deng, P. Besse-
Hoggan, M. Sancelme, A.-M. Delort, P. Husson and
M. F. Costa Gomes, J. Hazard. Mater., 2011, 198, 165–174;
(f) M. Petkovic, K. R. Seddon, L. P. N. Rebelo and
C. S. Pereira, Chem. Soc. Rev., 2011, 40, 1383–1403;
(g) G. Quijano, A. Couvert, A. Amrane, G. Darracq,
C. Couriol, P. Le Cloirec, L. Paquin and D. Carrie, Chem.
Eng. Sci., 2011, 66, 2707–2712; (h) G. Quijano, A. Couvert,
A. Amrane, G. Darracq, C. Couriol, P. Le Cloirec, L. Paquin
and D. Carrié, Chem.–Eng. J., 2011, 174, 27–32; (i) C. Zhang,
S. V. Malhotra and A. J. Francis, Chemosphere, 2011, 82,
1690–1695; ( j) C. Abrusci, J. Palomar, J. L. Pablos,
F. Rodriguez and F. Catalina, Green Chem., 2011, 13, 709–
717.
6 (a) N. Gathergood and P. J. Scammells, Aust. J. Chem., 2002,
55, 557–560; (b) M. T. Garcia, N. Gathergood and
P. J. Scammells, Green Chem., 2005, 7, 9–14.
7 Selected examples. (a) S. Stolte, S. Abdulkarim, J. Arning,
A. Blomeyer-Nienstedt, U. Bottin-Weber, M. Matzke,
J. Ranke, B. Jastorff and J. Thoeming, Green Chem., 2008, 10,
214–224; (b) T. P. Pham, C. W. Cho, C. O. Jeon, Y. J. Chung,
M. W. Lee and Y. S. Yun, Environ. Sci. Technol., 2009, 43,
516–521; (c) K. M. Docherty, M. V. Joyce, K. J. Kulacki and
C. F. Kulpa, Green Chem., 2010, 12, 701–712.
3 Selected examples. (a) J. Pernak, K. Sobaszkiewicz and
I. Mirska, Green Chem., 2002, 5, 52–56; (b) J. Pernak, I. Goc
and I. Mirska, Green Chem., 2004, 6, 323–329; (c) L. Carson,
P. K. W. Chau, M. J. Earle, M. A. Gilea, B. F. Gilmore,
S. P. Gorman, M. T. McCann and K. R. Seddon, Green
Chem., 2009, 11, 492–497; (d) J. Arning and M. Matzke,
Curr. Org. Chem., 2011, 15, 1905–1917; (e) T. P. Pham;,
C. W. Cho and Y. S. Yun, Water Res., 2010, 44, 352–372;
(f) R.-N. Dou, S.-S. Liu, L.-Y. Mo, H.-L. Liu and F.-C. Deng,
Environ. Sci. Pollut. Res. Int., 2011, 18, 734–742;
(g) Y. V. Nancharaiah and A. J. Francis, Bioresour. Technol.,
2011, 102, 6573–6578; (h) S. Viboud, N. Papaiconomou,
A. Cortesi, G. Chatel, M. Draye and D. Fontvieille, J. Hazard.
8 H. Xie, T. Hayes and N. Gathergood, Catalysis of Reactions
by Amino Acids, in Amino Acids, Peptides and Proteins in
Organic Chemistry, Origins and Synthesis of Amino Acids, ed.
A. B. Hughes, Wiley-VCH, Weinheim, Germany, vol. 1,
2009, ISBN: 978-3-527-32096-7.
9 Y. Chiba, A. Islam, Y. Watanabe, R. Komiya, N. Koide and
L. Han, Jpn. J. Appl. Phys., 2006, 45, L638–L640.
Mater., 2012, 215–216, 40–48; (i) M. H. Fatemi and 10 (a) S. Ito, M. K. Nazeeruddin, P. Liska, P. Comte,
P. Izadiyan, Chemosphere, 2011, 84, 553–563; ( j) M. Alvarez-
Guerra and A. Irabien, Green Chem., 2011, 13, 1507–1516;
(k) M. I. Hossain, B. B. Samir, M. El-Harbawi, A. N. Masri,
M. I. A. Mutalib, G. Hefter and C.-Y. Yin, Chemosphere,
2011, 85, 990–994; (l) S. P. M. Ventura, R. L. Gardas,
R. Charvet, P. Péchy, M. Jirousek, A. Kay, S. M. Zakeeruddin
and M. Grätzel, Prog. Photovoltaics: Res. Appl., 2006, 14,
589–601; (b) E. Gibson, L. Pleux, J. Fortage, Y. Pellegrin,
E. Blart, F. Odobel, A. Hagfeldt and G. Boschloo, Langmuir,
2012, 28, 6485–6493.
F. Gonçalves and J. A. P. Coutinho, J. Chem. Technol. Bio- 11 A. Yella, H.-W. Lee, H. N. Tsao, C. Yi, A. K. Chandiran,
technol., 2011, 86, 957–963; (m) J. Zhang, S.-S. Liu,
M. K. Nazeeruddin, E. W.-G. Diau, C.-Y. Yeh,
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2252–2265 | 2263

View Article Online
Paper
Green Chemistry
S. M. Zakeeruddin and M. Grätzel, Science, 2011, 334, 629–
634.
J. Ferguson, K. R. Seddon and G. Stephens, Green Chem.,
2009, 11, 402–408.
12 (a) D. Kuang, C. Klein, Z. Zhang, S. Ito, J.-E. Moser, 17 (a) P. Thi and C. Cho, Water Res., 2010, 44(2), 352–372;
S. M. Zakeeruddin and M. Grätzel, Small, 2007, 3, 2094–
2102; (b) J. Kroon, N. Bakker, H. Smit, P. Liska, K. Thampi,
(b) E. Urbanik, J. Pernak, A. Skrzypczak, J. Zabielska and
M. Jadwiga, PL 186744 B1 20040227, 2004.
P. Wang, S. Zakeeruddin, M. Grätzel, A. Hinsch, S. Hore, 18 J. Ranke, K. Molter, F. Stock, U. Bottin-Weber, J. Poczobutt,
U. Wurfel, R. Sastrawan, J. Durrant, E. Palomares,
H. Pettersson, T. Gruszecki, J. Walter, K. Skupien and
J. Hoffmann, B. Ondruschka, J. Filser and B. Jastorff, Eco-
toxicol. Environ. Saf., 2004, 58, 396–404.
G. Tulloch, Prog. Photovoltaics: Res. Appl., 2007, 15, 1–18; 19 N. Gathergood, S. Morrissey and B. Pegot, Biodegradable
(c) A. Hinsch, J. Kroon, R. Kern, I. Uhlendorf, J. Holzbock,
A. Meyer and J. Ferber, Prog. Photovoltaics: Res. Appl., 2001,
9(6), 425–438.
ionic liquids for chemical industry, PCT Int. Appl, WO
2009024607 A1 20090226, 2009.
20 Clinical and Laboratory Standards Institute, Methods for
Dilution Antimicrobial Susceptibility Tests for Bacteria
That Grow Aerobically, Approved Standard, Seventh
Edition, CSLI document M07-A7, 940 West Valley Road,
Suite 1400, Wayne, PA, 19087–898, 2006.
13 (a) Y. Bai, Y. Cao, J. Zhang, M. Wang, R. Li, P. Wang,
S. M. Zakeeruddin and M. Gratzel, Nat. Mater., 2008, 7,
626–630; (b) A. Hagfeldt, Ambio, 2012, 41(2), 151–155.
14 (a) R. F. M. Frade and C. A. M. Afonso, Hum. Exp. Toxicol.,
2010, 29(12), 1038–1054; (b) D. Zhao, Y. Liao and Z. Zhang, 21 D. Amsterdam, Susceptibility testing of antimicrobials in
Clean, 2007, 35(1), 42–48; (c) M. North and C. Vulpe,
Int. J. Mol. Sci., 2010, 11(12), 4796–4813; (d) B. Gaytán,
A. Loguinov, S. Lantz, J. Lerot, N. Denslow and C. Vulpe,
Toxicol. Sci., 2013, 132(2), 347–358.
liquid media, in Antibiotics in Laboratory Medicine, ed.
V. Lorian and W. Wilkins, V. Baltimore, 3rd edn, 1991,
pp. 72–78.
22 S. Ito, P. Chen, P. Comte, M. K. Nazeeruddin, P. Liska,
P. Péchy and M. Grätzel, Prog. Photovoltaics: Res. Appl.,
2007, 15, 603–612.
15 (a) S. P. M. Ventura, M. Gurbisz, M. Ghavre,
F. M. M. Ferreira, F. Gonçalves, I. Beadham, B. Quilty,
J. A. P. Coutinho and N. Gathergood, ACS Sustainable Chem. 23 S. Ito, T. N. Murakami, P. Comte, P. Liska, C. Grätzel,
Eng., 2013, 1, 393; (b) N. Ferlin, M. Courty, S. Gatard,
M. Spulak, B. Quilty, I. Beadham, M. Ghavre, A. Haiß,
M. K. Nazeeruddin and M. Grätzel, Thin Solid Films, 2008,
516, 4613–4619.
K. Kümmerer, N. Gathergood and S. Bouquillon, Tetrahe- 24 (a) R. Boethling, E. Sommer and D. Fiore, Chem. Rev., 2007,
dron, 2013, 69, 6150–6161; (c) I. Hemeon, N. Barnett,
N. Gathergood, P. J. Scammells and R. Singer,
Aust. J. Chem., 2004, 2, 125–130; (d) L. Myles, R. Gore,
M. Špulák, N. Gathergood and S. J. Connon, Green Chem.,
2010, 12, 1157–1162; (e) L. Myles, N. Gathergood and
S. J. Connon, Chem. Commun., 2013, 49, 5316–5318;
107, 2207–2227; (b) R. Boethling, Designing Safer Chemi-
cals, ACS Symp. Ser., 1996, 640, 156–171; (c) P. Howard,
R. Boethling, W. Stiteler, W. Meylan and J. Beauman, Sci.
Total Environ., 1991, 109/110, 635–641; (d) R. Boethling,
Cationic Surfactants, Surfactant Science Series 53, Marcel
Dekker, New York, 1994, pp. 95–135.
(f) S. Bouquillon, T. Courant, D. Dean, N. Gathergood, 25 S. Kanjilal, S. Sunitha, P. Reddy, K. Kumar, U. Murty
S. Morrissey, B. Pegot, P. J. Scammells and R. D. Singer,
and R. Prasad, Eur. J. Lipid Sci. Technol., 2009, 111, 941–
Aust. J. Chem., 2007, 60, 843–847; (g) S. Morrissey,
948.
I. Beadham and N. Gathergood, Green Chem., 2009, 11, 26 (a) B. M. Trost, Science, 1991, 254, 1471–1477;
466–474; (h) Y. Deng, S. Morrissey, N. Gathergood,
A.-M. Delort, P. Husson and M. F. Costa Gomes, Chem-
(b) B. M. Trost, Angew. Chem., Int. Ed. Engl., 1995, 34, 259–
281.
SusChem, 2010, 3(3), 377–385; (i) M. T. Garcia, 27 H. Nguyen, P. Kirilov, H. Matondo and M. Baboulène,
N. Gathergood and P. J. Scammells, Green Chem., 2004, 6, J. Mol. Catal. A: Chem., 2004, 218, 41–45.
166–175; ( j) N. Gathergood, P. J. Scammells and 28 R. Sheldon, Pure Appl. Chem., 2000, 72(7), 1233–1246.
M. T. Garcia, Green Chem., 2006, 8, 156–160; 29 D. Seo, S. Sarker, N. Nath, S. Choi, A. Ahammad, J. Lee and
(k) D. Coleman, M. Spulak, M. T. Garcia and
N. Gathergood, Green Chem., 2012, 14, 1350–1356; 30 H. Wang, X. Zhang, F. Gong, G. Zhou and Z. Wang, Adv.
(l) R. G. Gore, T.-K.-T. Truong, M. Pour, L. Myles, Mater., 2012, 24, 121–124.
S. J. Connon and N. Gathergood, Green Chem., 2013, 15, 31 C. Samori, D. Malferrari, P. Valbonesi, A. Montecavalli,
W. Kim, Electrochim. Acta, 2010, 55, 1483–1488.
2727; (m) L. Myles, R. G. Gore, N. Gathergood and
S. J. Connon, Green Chem., 2013, 15, 2740; (n) R. G. Gore,
L. Myles, M. Spulak, I. Beadham, T. M. Garcia, S. J. Connon
F. Moretti, P. Galletti, G. Sartor, E. Tagliavini, E. Fabbri
and A. Pasteris, Ecotoxicol. Environ. Saf., 2010, 73, 1456–
1464.
and N. Gathergood, Green Chem., 2013, 15, 2747; 32 (a) C. Cho, T. Pham, Y. Jeon, K. Vijayaraghavan, W. Choe
(o) N. Ferlin, M. Courty, A. N.-V. Nhien, S. Gatard, M. Pour,
B. Quilty, M. Ghavre, A. Haiß, K. Kümmerer, N. Gathergood
and S. Bouquillon, RSC Adv., 2013, 3, 26241.
and Y. Yun, Chemosphere, 2007, 69, 1003–1007;
(b) C. Pretti, C. Chiappe, I. Baldetti, S. Brunini,
G. Monni and L. Intorre, Ecotoxicol. Environ. Saf., 2009,
72, 1170–1176; (c) R. Frade and C. Afonso, Hum. Exp.
Toxicol., 2010, 29(12), 1038–1054; (d) A. Romero,
16 (a) N. Wood and G. Stephens, Phys. Chem. Chem. Phys.,
2010, 12, 1670–1674; (b) M. Rebros, H. Q. N. Gunaratne,
2264 | Green Chem., 2014, 16, 2252–2265
This journal is © The Royal Society of Chemistry 2014

View Article Online
Green Chemistry
Paper
A. Santos and J. Rodriguez, J. Hazard. Mat., 2008, 151, 35 (a) S. Mikoshiba, S. Murai, H. Sumino, T. Kado, D. Kosugi
268–273; (e) M. Matzke, S. Stolte, K. Thiele, T. Juffernholz,
J. Arning, J. Ranke, U. Welz-Biermann and B. Jastorff,
Green Chem., 2007, 9(11), 1198–1207.
and S. Hayase, Curr. Appl. Phys., 2005, 5, 152–158;
(b) A. Olaya, P. Ge and H. Girault, Electrochem. Commun.,
2012, 19, 101–104.
33 (a) M. Kürschner, K. Nielsen, C. Andersen, V. Sukhorukov, 36 M. Galiński, A. Lewandowski and I. Stępniak, Electrochim.
W. Schenk, R. Benz and U. Zimmermann, Biophys. J., 1998, Acta, 2006, 51, 5567–5580.
74(6), 3031–3043; (b) J. Ranke, M. Cox, A. Müller, 37 J. Jeon, H. Kim, W. A. Goddard, T. A. Pascal, G.-I. Lee and
C. Schmidt and D. Beyersmann, Toxicol. Environ. Chem.,
2006, 88(2), 273–285.
34 J. Cong, X. Yang, L. Kloo and L. Sun, Energy Environ. Sci.,
2012, 5, 9180–9194.
J. K. Kang, J. Phys. Chem. Lett., 2012, 3, 556–559.
38 J. G. Huddleston, A. E. Visser, W. M. Reichert,
H. D. Willauer, G. A. Broker and R. D. Rogers, Green Chem.,
2001, 3, 156–164.
This journal is © The Royal Society of Chemistry 2014
Green Chem., 2014, 16, 2252–2265 | 2265