Mandelic acid derived ionic liquids: synthesis, toxicity and biodegradability

Article abstract of DOI:10.1039/c6ra25562k

A series of ten ionic liquids (ILs) was synthesised from the renewable resource mandelic acid. The ILs showed low antimicrobial activity towards the thirteen bacterial and twelve fungal strains they were screened against. A general trend of increasing bacterial toxicity in the order methyl ester < ethyl ester < n-butyl ester/amide was observed. IL biodegradability was evaluated using the Closed Bottle test (OECD 301D). Biodegradation increased in the order of increasing alkyl chain length for the ester ILs (methyl < ethyl < n-butyl). Despite none of the ILs presenting as readily biodegradable (>60% in 28 days), a series of biodegradation transformation products has been proposed based on the degradation of the ester/amide alkyl chain. This data has allowed for an assessment of the effect of IL structural features on toxicity and biodegradation, particularly allowing for a comparison to earlier work where additional oxygen atoms were present to facilitate biodegradation and attenuate toxicity. The mandelic acid derived ILs did not pass the Closed Bottle test (OECD 301D) and can be included in the rules of thumb for the design of biodegradable ILs.

Full text of DOI:10.1039/c6ra25562k

RSC Advances
View Article Online
View Journal | View Issue
PAPER
Mandelic acid derived ionic liquids: synthesis,
toxicity and biodegradability†
Cite this: RSC Adv., 2017, 7, 2115
Hannah Prydderch,a Annette Haib,b Marcel Spulak,c Brid Quilty,d Klaus Kummerer,b
¨
e
f
*
Andreas Heise and Nicholas Gathergood
A series of ten ionic liquids (ILs) was synthesised from the renewable resource mandelic acid. The ILs
showed low antimicrobial activity towards the thirteen bacterial and twelve fungal strains they were
screened against. A general trend of increasing bacterial toxicity in the order methyl ester < ethyl ester <
n-butyl ester/amide was observed. IL biodegradability was evaluated using the Closed Bottle test (OECD
301D). Biodegradation increased in the order of increasing alkyl chain length for the ester ILs (methyl <
ethyl < n-butyl). Despite none of the ILs presenting as readily biodegradable (>60% in 28 days), a series
of biodegradation transformation products has been proposed based on the degradation of the ester/
amide alkyl chain. This data has allowed for an assessment of the effect of IL structural features on
toxicity and biodegradation, particularly allowing for a comparison to earlier work where additional
oxygen atoms were present to facilitate biodegradation and attenuate toxicity. The mandelic acid derived
ILs did not pass the Closed Bottle test (OECD 301D) and can be included in the rules of thumb for the
design of biodegradable ILs.
Received 20th October 2016
Accepted 25th November 2016
DOI: 10.1039/c6ra25562k
www.rsc.org/advances
There have been numerous efforts towards this aim with several
works on the biodegradability and toxicity of ILs published
since the early 2000's.4–16
Introduction
Ionic liquids (ILs) are poorly coordinated salts most oen with
melting points <100 C.1 The myriad of properties that ILs can
ꢀ
The utilisation of renewable resources for the synthesis of
ILs has also been of great interest towards the aim of creating
more sustainable ILs. Examples in the literature include the use
of cholinium,17,18 prolinate,19 L-lactate20 and succinate;20 where
low toxicity and readily biodegradable ILs from renewable
resources have been identied.17 More recently renewable pyr-
idinium ILs have been synthesised from pyridinium zwitterions
derived from amino acids and furfural.21
Mandelic acid is an aromatic a-hydroxy acid that is found
naturally in small quantities in plant tissues such as wheat
leaves and grapes.22 A wide range of organisms (e.g. Penicillium
chrysogenum23 and phytoplankton24) can convert the naturally
occurring compounds L-phenylalanine and phenyl acetic acid
into mandelic acid.22 Mandelic acid can also be obtained by
heating the glycoside amygdalin with concentrated hydro-
chloric acid.25 Amygdalin can be isolated from the seeds of the
bitter almond tree26,27 and is also found in apricots27,28 and
peach pits.22 In industry mandelic acid is made from the acid
or enzyme catalysed hydrolysis of mandelonitrile.29 However,
as mandelonitrile is oen produced from the reaction between
benzaldehyde and sodium cyanide, alternative methods
avoiding the use of this toxic reagent to prepare mandelic acid
and its derivatives have been developed.30 Optically pure (R)-
and (S)-mandelic acid have also been produced from the
renewable feedstock glucose using a genetically modied
strain of E. coli.31
possess (such as highly tuneable, low vapour pressure, large
liquid range, low ammability and high solvating capacity)
oen makes them attractive as reaction additives and alterna-
tive solvents for the replacement of volatile organic
compounds.2 However, possession of these properties does not
necessarily make them a greener alternative. An important
consideration is the toxicity and biodegradability assessment of
chemicals,3 which can lead to the design of safer (low toxicity
and high biodegradability, i.e. complete mineralisation) ILs.
aSchool of Chemical Sciences, Dublin City University, Glasnevin, Dublin 9, Ireland
bInstitute of Sustainable and Environmental Chemistry, Leuphana University
¨
¨
Luneburg, DE-21335 Luneburg, Germany
cDepartment of Inorganic and Organic Chemistry, Charles University, Faculty of
´
´
´
Pharmacy, Heyrovskeho 1203, CZ-500 03 Hradec Kralove, Czech Republic
dSchool of Biotechnology, National Institute for Cellular Biotechnology, Dublin City
University, Glasnevin, Dublin 9, Ireland
eDepartment of Pharmaceutical & Medicinal Chemistry, Royal College of Surgeons in
Ireland, 123 St. Stephen's Green, Dublin 2, Ireland
fDepartment of Chemistry, Faculty of Science, Tallinn University of Technology,
Akadeemia tee 15, 12618 Tallinn, Estonia. E-mail: nicholas.gathergood@ttu.ee; Tel:
+372 620 4381
† Electronic supplementary information (ESI) available: Includes additional
antimicrobial screening results, experimental procedures for the synthesis of
ILs and reaction intermediates with characterization data (NMR, IR, mp, MS),
as well as a description of the antimicrobial and biodegradation screening
methods. See DOI: 10.1039/c6ra25562k
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 2115–2126 | 2115

View Article Online
RSC Advances
Paper
Studies on the toxicity of ILs that contain phenyl rings as an conditions to the previously published work, a direct compar-
alternative to long alkyl chains have shown that it is possible to ison of IL toxicity between the two series was possible. The study
incorporate a lipophilic substructure into an IL whilst still presented herein additionally expands on the previous work by
maintaining low levels of toxicity.32,33 Colemen et al.34 and more providing an evaluation of the biodegradability of the newly
recently Jordan et al.35 additionally showed that the introduc- synthesised ILs. Overall, this study aims to further investigate
tion of a phenyl group in the form of phenylalanine in IL the effect of structural modications on IL toxicity and
structures did not result in high levels of antimicrobial activity biodegradation, in particular the effect on toxicity of removing
towards the bacterial or fungal strains screened, in the majority oxygen from the previous class of ILs. This work intends to
of cases. The impact of phenyl substitution was assessed in satisfy the 4th, 7th and 10th principles of green chemistry:
terms of IL cytotoxicity by Ranke et al. in 2007.33 Here the designing safer chemicals, the use of renewable feedstocks, and
authors studied the inuence of cation lipophilicity on IL designing for degradation (respectively).
cytotoxicity in IPC-81 leukemia cells and determined that lip-
ophilicity was a dominating factor for IL cytotoxicity. Relatively
low toxicity of the phenyl IL 1-methyl-3-(phenylmethyl)-
Results and discussion
imidazolium chloride was demonstrated (EC50 > 1000 mM)
Synthesis
compared to the 1-heptyl-3-methyl-1H-imidazolium chloride
A series of ten mandelic acid derived ILs (1–10) was synthesised
with specically designed structural motifs to allow for the
investigation of their effect on toxicity and biodegradation
(Fig. 1). These structural features included: three alkyl chain
lengths, n-butyl, ethyl and methyl; both ester and amide moie-
ties; the halide anion bromide; and various N-heterocyclic
headgroups, 1-methylimidazolium, pyridinium, 3-methoxy
pyridinium and 3-(ethoxycarbonyl)pyridinium.
Other than the research carried out on this particular class of
compounds directly preceding this work,32 there have been
several mentions of this class of compounds in the litera-
ture.43–45 However, the only exact compound match for this new
series that could be found in the literature was for the pyr-
idinium ethyl ester IL (4), where compound characterisation
could be found, but not biodegradation and toxicity analysis.46
All the ILs were synthesised as the racemate from the starting
material DL-mandelic acid as both the D(ꢁ) and L(+) isomers are
known to occur in nature.42 The synthesis of the ILs consisted of
three steps (Schemes 1 and 2): esterication or amidation of
mandelic acid; halogenation of the a-position of the ester/
amide with thionyl bromide; and nucleophilic substitution of
(same number of carbons in cation side chain as the phenyl IL)
(EC50 ¼ 399 mM). The decrease in toxicity was attributed to the
decrease in IL lipophilicity due to the presence of the phenyl
group compared to the heptyl side chain. This result demon-
strates that it is possible to attenuate IL cytotoxicity by replacing
alkyl chains (e.g. heptyl) with phenyl groups. Ranke et al. also
reported that decreasing the IL alkyl chain length from heptyl to
n-butyl resulted in lower cytotoxicity (e.g. 1-butyl-3-methyl-1H-
imidazolium chloride, EC50 ¼ 3548 mM).
The introduction of aromatic rings can however have
a marked effect on compound biodegradability (as determined
from biodegradation in CO2 Headspace test ISO 14593), as in
the case of Gore et al. where the introduction of benzyl groups
signicantly reduced IL biodegradability.36 In 2012 Coleman
et al. showed that amino acid ILs containing phenyl rings can
present as readily biodegradable.34 The assumption was that the
aromatic ring of the phenylalanine moiety was converted to
CO2, although no transformation product analysis was per-
formed in that study to conrm this breakdown. More recently
Jordan et al.35 and Haib et al.37 have further shown that a fully
mineralisable IL can be produced from an amino acid
(phenylalanine) IL containing an aromatic side chain.
Although, the introduction of a phenyl group in the form of
phenylalanine did not lead to a signicant increase in biodeg-
radation in all cases. This demonstrates the challenge of iden-
tifying “rules of thumb”38 to assist in the design of
biodegradable ILs.12,13
Previous work in the Gathergood group includes the design
and synthesis of low toxicity ILs from mandelic acid deriva-
tives.32 Herein we present a novel series of ILs synthesised
directly from the renewable building block mandelic acid in
a shorter and more efficient synthesis. Mandelic acid was
selected as the backbone for the synthesis of the ILs as there are
numerous reports in the literature detailing the microbial
metabolic pathways for the degradation of mandelic acid and
mandelates.22,39–42 Thus, by utilising
a structural moiety
susceptible to microbial breakdown we aim to better under-
stand the ready biodegradability of the ILs.
It is desired that the ILs reported in the study will present
with low toxicity similar to the previous class of mandelic acid
derived ILs.32 By carrying out toxicity screening under the same Fig. 1 Mandelic acid derived bromide ILs (1–10).
2116 | RSC Adv., 2017, 7, 2115–2126
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
Scheme 2 General synthesis of mandelic acid derived amide ILs (9 and
10).
Scheme 1 General synthesis of mandelic acid derived ester ILs (1–8).
the a-halo ester with an aromatic N-heterocycle. The synthesis
of the previously reported mandelic acid derived ILs required an
additional synthetic step to produce the IL starting materials, as
they were not synthesised directly from mandelic acid.32 Thus,
the ILs described herein (1–10) were produced in a shorter and
more efficient synthesis directly from the renewable resource.
Additionally, mandelic acid production from benzaldehyde
using sodium cyanide is avoided in this route by utilising the
natural mandelate.30
Esterication of mandelic acid was carried out by the well-
known Fischer–Speier method47 to produce methyl, ethyl and
n-butyl mandelic acid esters 11–13 (Scheme 1) in excellent yields
(92–97%). Modied literature procedures were employed for the
synthesis of previously reported methyl ester 11 48 and ethyl ester
12,49 with a new procedure described for the synthesis of n-butyl
ester 13 50 (see ESI†). The n-butyl amide 14 (Scheme 2) was syn-
thesised with acetyl chloride (AcCl), according to a modied
version of the literature procedure,51 in high yield (89%) (see
ESI†). Initial attempts to synthesise 14 by greener methods did
not proceed in high yields (<9%).52,53
thionyl bromide (SOBr2) to the mandelic acid esters 11–13 and
amide 14. The reaction occurs via a meso substitution of the
alcohol by bromine and was utilised to give alkylating agents
15–18 (Schemes 1 and 2) in moderate yields (51–67%). Previ-
ously reported a-bromoester 15 was synthesised by a modica-
tion of the literature procedure,54 with a new procedure
described for a-bromoester 16 55 (see ESI†).
ILs 1–10 (Schemes 1 and 2) were synthesised via a nucleo-
philic substitution reaction between an N-heterocycle and the a-
bromoester/a-bromoamide of mandelic acid (15–18) (see ESI†).
N-heterocycles 1-methylimidazole, pyridine, 3-methoxy pyridine
and ethyl nicotinate (required for the 3-(ethoxycarbonyl)pyr-
idinium headgroup) were selected for the IL synthesis. Previ-
ously reported IL 4 46 was prepared according to a new synthetic
procedure (see ESI†).
Seven of the ILs (1–5, 9 and 10) were produced in excellent
yields (84–99%), with the three n-butyl esters with pyridinium
(6), 3-methoxy pyridinium (7) and 3-(ethoxycarbonyl)pyridinium
(8) headgroups produced in moderate yields (56, 74 and 63%
respectively). All ten ILs were obtained as crystalline solids with
a wide range of melting points observed (48–169 ꢀC) (see ESI†).
At the COIL-6 conference in South Korea, Professor Kenneth
Synthesis of the a-bromoester and a-bromoamide alkylating
agents 15–18 (Schemes 1 and 2) were carried out by addition of
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 2115–2126 | 2117

View Article Online
RSC Advances
Paper
Table 1 Antifungal screening results for ILs 1–10 (MIC, IC80 or IC50
)
the n-butyl ester ILs, which is expected due to their additional
hydrogen bond donor ability.
IL MICb (mM)
Straina
CA1
CA2
CP
Time (h)
1, 8, 9, 10
2, 4, 7
3
5, 6
Microbial toxicity screening
Our strategy regarding toxicity screening of chemicals as part of
a tandem toxicity and biodegradation assessment has been
previously published.34–37 One reason a compound can fail
a biodegradation test is due to high toxicity to critical micro-
organisms in the inoculum required for breakdown. For
instance, previous studies have shown that high toxicity to
bacteria, e.g. Pseudomonas putida, can be linked to inhibited
biodegradation.57 Thus, completing a microbial toxicity study
prior to biodegradation screening provides valuable data to
assist analysis of biodegradation results. Compounds are rst
evaluated against twenty bacterial and fungal strains to identify
high antimicrobial activity examples. Secondly, IC50 values at
higher IL concentrations are measured for compounds which
do not exhibit high antimicrobial activity. Compounds which
have high antimicrobial activity are identied as ‘hit’
compounds as part of our group's medicinal chemistry and
biocide research interests. Low antimicrobial activity
compounds are prioritised for green chemistry applications.
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
72 h
120 h
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>1.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
0.5
0.5
1.0
1.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
>2.0
CK1
CK2
CT
CG
CL
TA
AF
AC
TM
Fungal toxicity results
a
The in vitro antifungal activities of ILs 1–10 (Table 1) were
CA1: Candida albicans ATCC 44859, CA2: Candida albicans ATCC
90028, CP: Candida parapsilosis ATCC 22019, CK1: Candida krusei evaluated against four ATCC strains (Candida albicans ATCC
ATCC 6258, CK2: Candida krusei E28, CT: Candida tropicalis 156, CG:
44859, Candida albicans ATCC 90028, Candida parapsilosis
Candida glabrata 20/I, CL: Candida lusitaniae 2446/I, TA: Trichosporon
ATCC 22019, Candida krusei ATCC 6258) and eight clinical
asahii 1188, AF: Aspergillus fumigatus 231, AC: Absidia corymbifera 272,
b
TM: Trichophyton mentagrophytes 445. IC50 values were assessed for isolates of fungi, ve yeasts (Candida krusei E28, Candida tro-
AF, AC and TM and IC80 for all other strains.
picalis 156, Candida glabrata 20/I, Candida lusitaniae 2446/I,
Trichosporon asahii 1188) and three lamentous fungi (Asper-
gillus fumigatus 231, Absidia corymbifera 272, Trichophyton
mentagrophytes 445). Three ATCC strains were also used as the
quality control strains. All the isolates were maintained on
Sabouraud dextrose agar prior to being tested. Minimum
ꢀ
Seddon proposed that the arbitrary use of 100 C as a melting
point cut-off for the denition of an IL was inexible and con-
straining and should no longer be used as a prerequisite for
classication as an IL.56 Thus, the terminology of IL is used
inhibitory concentrations (MICs) were determined by modied
CLSI standard of the microdilution format of the M27-A3 58 and
herein despite the majority of the ILs described possessing
M38-A2 59 documents for yeasts and lamentous fungi respec-
melting points >100 ꢀC. ILs 1 and 6 had the lowest melting
tively. MIC values were calculated aer 72 and 120 h for T.
points (61–63 and 48–50 ꢀC respectively) despite any similar
structural features. The highest melting points for the series
mentagrophytes, and for all other strains MIC values were
calculated aer 24 and 48 h. The MICs were dened as 50%
inhibition (IC50) of the control growth for yeasts or 80% inhi-
bition (IC80) of the control growth for lamentous fungi. The
were observed for the pyridinium ILs 4 and 10 (161–162 and
167–169 ꢀC respectively). ILs 2 and 9 had melting points around
140 ꢀC (140–142 and 143–145 ꢀC respectively), again with no
maximum test concentration of the IL was dependant on the IL
similar structural features. ILs 3 and 5 had melting points
aqueous solubility, with the highest IL concentration in the test
around 100 ꢀC (101–102 and 99–101 ꢀC respectively), both with
being 2.0 mM. See ESI† for full experimental details.
1-methylimidazolium headgroups. ILs 7 and 8 had moderately
The results from Table 1 show that the 1-methylimidazolium
high melting points (127–128 and 112–113 ꢀC respectively),
methyl ester IL (1), the 3-(ethoxycarbonyl)pyridinium n-butyl
both n-butyl esters with 3-methoxy pyridinium and 3-(ethox-
ester IL (8) and the 1-methylimidazolium and pyridinium n-
ycarbonyl)pyridinium headgroups respectively. It is observed
that the class of headgroup has a more marked effect on the IL
melting points than the alkyl chain length. In all but one case
the pyridinium ILs had higher melting points than the 1-
methylimidazolium ILs with the same ester/amide alkyl chain,
with this trend only reversed for the n-butyl ester ILs 5 and 6.
The n-butyl amides ILs 9 and 10 had higher melting points than
butyl amide ILs (9 and 10) did not have high fungal toxicity, up
to the maximum IL concentration limit, with IC50/IC80 values
>2.0 mM. The pyridinium methyl and ethyl ester ILs (2 and 4)
and the 3-methoxy pyridinium n-butyl ester IL (7) did not have
high toxicity, up to the maximum IL concentration limit, with
IC50/IC80 values >1.0 mM. The 1-methylimidazolium and pyr-
idinium n-butyl ester ILs (5 and 6) inhibited C. tropicalis at
2118 | RSC Adv., 2017, 7, 2115–2126
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
Table 2 Antibacterial screening results for ILs 1–10 (MIC, IC95
IL MIC (mM)
2, 8
)
Straina Time (h)
1
3
4, 7
5
6
9
10
SA
24 h
48 h
2.0 >2.0 0.50 >1.0 2.0 0.50 >2.0 2.0
2.0 >2.0 0.50 >1.0 >2.0 2.0 >2.0 2.0
MRSA 24 h
48 h
>2.0 >2.0 0.50 >1.0 >2.0 2.0 >2.0 >2.0
>2.0 >2.0 0.50 >1.0 >2.0 >2.0 >2.0 >2.0
2.0 2.0 0.25 >1.0 2.0 0.50 >2.0 2.0
2.0 2.0 0.25 >1.0 >2.0 1.0 >2.0 2.0
>2.0 >2.0 0.50 >1.0 >2.0 >2.0 >2.0 >2.0
>2.0 >2.0 0.50 >1.0 >2.0 >2.0 >2.0 >2.0
>2.0 >2.0 2.0 >1.0 >2.0 >2.0 >2.0 >2.0
>2.0 >2.0 2.0 >1.0 >2.0 >2.0 >2.0 >2.0
>2.0 >2.0 2.0 >1.0 >2.0 >2.0 >2.0 >2.0
>2.0 >2.0 2.0 >1.0 >2.0 >2.0 >2.0 >2.0
2.0 >2.0 1.0 >1.0 1.0 >2.0 >2.0 >2.0
2.0 >2.0 1.0 >1.0 >2.0 >2.0 >2.0 >2.0
>2.0 >2.0 >2.0 >1.0 >2.0 >2.0 >2.0 >2.0
>2.0 >2.0 >2.0 >1.0 >2.0 >2.0 >2.0 >2.0
SE
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
24 h
48 h
EF
EC
KP
KP-E
PA
Fig. 2 Mandelic acid derived ILs 19–20.32
2.0 mM aer 24 h but showed no inhibition up to the test
concentration of 2.0 mM aer 48 h, suggesting that the fungal
strain was able to overcome the initial observed inhibition at
2.0 mM aer 24 h and recover. However, this result is not
signicant when compared to other data in Table 1, especially
as it appears for only one fungal strain. The concentration is
also well above the level used in the biodegradation studies. The
overall ndings are that none of the ILs demonstrated high
toxicity to the range of fungal strains screened against. The 1-
methylimidazolium ethyl ester IL (3) inhibited C. krusei ATCC
6258 at 0.50 mM aer both 24 and 48 h, and C. krusei E28 at
1.0 mM aer both 24 and 48 h. Higher levels of fungal toxicity
are usually observed as the length of the alkyl chain increases.7
Although it is unexpected that the ethyl ester IL (3) is more toxic
a
SA: Staphylococcus aureus ATTC 6538, EC: Escherichia coli ATTC 8739,
PA: Pseudomonas aeruginosa ATTC 9027, MRSA: Staphylococcus aureus
MRSA HK5996/08, SE: Staphylococcus epidermidis HK6966/08, EF:
Enterococcus sp. HK14365/08, KP: Klebsiella pneumoniae HK11750/08,
KP-E: Klebsiella pneumoniae ESBL HK14368/08.
comparative purposes despite containing a chloride anion, as
the halide anions bromide and chloride have previously been
reported to have similar toxic effects in imidazolium ILs.63
Bacterial toxicity results
that the n-butyl analogue (5), one must be careful not to over- The in vitro antibacterial activities of ILs 1–10 (Table 2) were
interpret the data based on two compounds. High toxicity was evaluated against three ATTC strains (Staphylococcus aureus
not observed for the mandelic acid derived ILs (1–10) in this ATTC 6538, Escherichia coli ATTC 8739, Pseudomonas aeruginosa
screening, which is a signicant result for the design of safer ATTC 9027) and ve clinical isolates (Staphylococcus aureus
ILs. As no compounds with high fungal toxicity were found, all MRSA HK5996/08, Staphylococcus epidermidis HK6966/08,
were classed as suitable substrates for biodegradation Enterococcus sp. HK14365/08, Klebsiella pneumoniae HK11750/
screening based on fungal toxicity data.
08, Klebsiella pneumoniae ESBL HK14368/08). The ATCC
Previous work evaluated the antifungal activity of a series of strains were also used as the quality control strains. MIC values
mandelic acid derived ILs similar to those described herein, but were calculated aer 24 and 48 h. The MICs were dened as
which incorporated a higher oxygen content (e.g. ether, phenol 95% inhibition (IC95) of the control growth. The maximum test
and catechol groups are present).60,61 When comparing the IC80 concentration of the IL was dependant on the IL aqueous
or IC50 values of the 1-methylimidazolium and pyridinium solubility, with the highest IL concentration in the test being
methyl and n-butyl ester ILs (1, 2, 5 and 6) (Table 1) to the 2.0 mM. All the isolates were maintained on Mueller–Hinton
mandelic acid derived ILs previously published with the same dextrose agar prior to being tested. See ESI† for full experi-
ester alkyl chain length and headgroup32 (19–22) (Fig. 2), only IL mental details.
1 had the same low level of inhibition as the previous ILs (19–
The results from Table 2 show that the ILs 1, 2, 4 and 7–10
22) with no high toxicity up to the maximum IL concentration did not have high toxicity towards the bacterial strains screened
limit (IC50/IC80 > 2.0 mM)60,61 (Table S1†). This result is not against. The pyridinium ethyl ester IL (4) and the 3-methoxy
surprising as a lower level of inhibition is expected for the pyridinium n-butyl ester (7) did not have high toxicity, up to the
previous ILs (19–22) due to the presence of the oxygen func- IL maximum concentration limit, with IC95 values >1.0 mM. The
tionality on the aromatic ring which decreases IL lipophilicity, 1-methylimidazolium n-butyl amide IL (9) did not have high
increases IL hydrophilicity, and in turn decreases IL toxicity.62 toxicity, up to the IL maximum concentration limit, with IC95
However, despite a higher fungal toxicity observed for the new values >2.0 mM. The 1-methylimidazolium and pyridinium
ILs (1, 2, 5 and 6) the results herein do not represent a large methyl ester ILs (1 and 2), the 3-methoxy pyridinium n-butyl
increase in the level of inhibition compared to the previously ester IL (8) and the pyridinium n-butyl amide IL (10) all
published ILs (19–22). IL 19 has been included here for inhibited S. epidermidis at a concentration of 2.0 mM, with IL 10
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 2115–2126 | 2119

View Article Online
RSC Advances
Paper
additionally inhibiting S. aureus at 2.0 mM and IL 1 addition-
In all cases the previously synthesised mandelic acid derived
ally inhibiting S. aureus and K. pneumoniae ESBL at 2.0 mM. ILs32 (19–22) (Fig. 2) had lower levels of inhibition than the new
Overall, the results for ILs 1, 2 and 8–10 are not signicantly ILs with the same ester alkyl chain length and headgroup (1, 2, 5
different. The 1-methylimidazolium n-butyl ester IL (5) and 6) for every bacterial strain that was screened against, with
inhibited S. aureus and S. epidermidis at 2.0 mM and K. pneu- no high toxicity observed up to the maximum IL concentration
moniae ESBL at 1.0 mM aer 24 h, but aer 48 h no longer limit (IC95 values > 2.0 mM) (Table S2†).60,61 Despite a higher
inhibited S. aureus, S. epidermidis and K. pneumoniae ESBL up bacterial toxicity observed for the new ILs (1, 2, 5 and 6), the
to the test concentration of 2.0 mM, suggesting that the results presented herein (IC95 values in the range 0.50–2.0 mM)
bacterial strains were able to overcome the initial observed do not represent a large increase in the level of inhibition. This
inhibition aer 24 h and recover. The pyridinium n-butyl ester is a signicant positive result considering the new class of ILs
IL (6) inhibited S. aureus MRSA at 2.0 mM aer 24 h, but aer are synthesised directly from a renewable resource in a shorter
48 h no longer inhibited S. aureus MRSA up to the test and more efficient synthesis.
concentration of 2.0 mM. IL 6 also inhibited S. aureus and S.
The antibacterial activities of ILs 1–10 were also evaluated
epidermidis at 0.50 mM aer 24 h, which reduced to inhibition against one Gram-positive bacterial strain (Bacillus subtilis
at 2.0 mM to S. aureus and 1.0 mM to S. epidermidis aer 48 h. DSMZ 10) and four Gram-negative bacteria (Escherichia coli
The 1-methylimidazolium ethyl ester IL (3) showed the greatest DSMZ 498, Pseudomonas uorescens DSMZ 50090, Pseudomonas
level of inhibition to the bacterial strains across the board with putida CP1, and Pseudomonas putida KT 2440). Of note, the
the highest level of inhibition observed against S. epidermidis Pseudomonas strains have been reported as important for
(IC95 ¼ 0.25 mM) and against S. aureus, S. aureus MRSA and biodegradation.57,64 The bacterial toxicity of the ILs was deter-
Enterococcus sp. (IC95 ¼ 0.50 mM). ILs 3 and 6 generally showed mined by bacterial growth inhibition in a 24 hour assay and are
greater inhibition towards the Gram-positive bacterial strains expressed as IC50 values. Stock solutions of the ILs were
(S. aureus, S. aureus MRSA, S. epidermidis and Enterococcus sp.) prepared in deionised water to a maximum concentration of
than towards the Gram-negative bacterial strains (E. coli, K. 2000 mM. The concentration of each stock solution was deter-
pneumoniae, K. pneumoniae ESBL and P. aeruginosa). This mined by the maximum aqueous solubility of each IL (Table
observation is most pronounced for IL 3, where the greatest S3†). ILs 2, 4, 7 and 9 had a solubility limit lower than 2000 mM.
level of inhibition of a Gram-negative bacterial strain was The maximum test concentration of each IL was a factor of 10
observed against K. pneumoniae ESBL (IC95 ¼ 0.50 mM). The dilution of the IL stock solution. All the toxicity tests were
greatest level of inhibition by IL 3 for a Gram-positive bacterial carried out in triplicate. See ESI† for full experimental details.
strain was observed against S. epidermidis (IC95 ¼ 0.25 mM);
The results from Table 3 show that the majority of the ILs
however, this is a much higher concentration than used in the have low toxicity (IC50 values > 25–50 mM) towards the bacterial
biodegradation testing (CBT). The overall ndings of this work strains screened against, with the lowest toxicity observed for
show that none of the ILs demonstrated high toxicity to the the pyridinium ethyl ester IL (4) against P. putida KT 2440 (IC50
>
range of bacterial strains screened. IL 3 presented with the 100 mM). The highest levels of inhibition were observed for the
highest toxicity; however, this was still only observed as 1-methylimidazolium n-butyl ester IL (5) which inhibited E. coli
moderate toxicity against several bacterial strains.
and P. putida CP1 with IC50 values of 6.25–12.5 mM, the n-butyl
Table 3 Antibacterial screening results for ILs 1–10 and 19–22 (IC50
)
IC50 (mM)
IL
E. coli
B. subtilis
P. uorescens
P. putida CP1
P. putida KT 2440
1
2
3
4
5
6
7
8
50–100
25–50
25–50
50–100
25–50
50–100
50–100
12.5–25
25–50
50–100
25–50
25–50
50–100
12.5–25
25–50
12.5–25
25–50
12.5–25
6.25–12.5
100–200
100–200
100–200
25–50
50–100
25–50
25–50
100–200
50–100
50–100
>100a
12.5–25
25–50
25–50
25–50
25–50
12.5–25
100–200
100–200
100–200
25–50
25–50
50–100
6.25–12.5
12.5–25
12.5–25
12.5–25
12.5–25
12.5–25
100–200
100–200
100–200
25–50
6.25–12.5
12.5–25
12.5–25
12.5–25
6.25–12.5
6.25–12.5
50–100
100–200
50–100
25–50
12.5–25
25–50
9
12.5–25
12.5–25
50–100
50–100
50–100
25–50
10
19b
20b
21b
22b
a
Solubility limit; IC50 value greater than solubility in media. b Previously reported in ref. 32. Bacillus subtilis DSMZ 10 (B. subtilis), Escherichia coli
DSMZ 498 (E. coli), Pseudomonas uorescens DSMZ 270 50090 (P. uorescence), Pseudomonas putida CP1 (P. putida CP1), Pseudomonas putida KT2440
(P. putida KT2440).
2120 | RSC Adv., 2017, 7, 2115–2126
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
Table 4 Calculation of theoretical ester/amide biodegradation for ILs 1–10 with a comparison to the experimental results obtained in the CBT
(OECD 301D) after 28 days
Total carbon
number
Ester/amide carbon
number
Theoretical biodegradation due to
share of carbon in side chain (%)
Experimental
biodegradation (%)
Absolute
difference (%)
IL
1
2
3
4
5
6
7
8
9
13
14
14
15
16
17
18
20
16
17
1
1
2
2
4
4
4
6
4
4
8
7
5
8
+3
ꢁ1
ꢁ6
ꢁ3
ꢁ2
+3
14
13
25
24
22
30
25
24
20
16
27
21
22
31
0
+1
ꢁ1
+25
+22
10
1
amide pyridinium IL (10) which inhibited E. coli and P. uo- lower levels of inhibition towards each bacterial strain that was
rescens with IC50 values of 6.25–12.5 mM, and the 1-methyl- screened against. Despite the maximum inhibition for the
imidazolium n-butyl amide (9) which inhibited E. coli with IC50 previous class being lower than the new class (highest toxicity,
values of 6.25–12.5 mM. A general trend of increasing IL toxicity IC50 ¼ 25–50 mM for IL 22 across all bacterial strains), it is
with increasing alkyl chain length was observed across all a signicant positive result that the new class of ILs do not
bacterial strains in the order methyl ester < ethyl ester < n-butyl represent a large increase in the level of inhibition (highest
ester/amide. In general the methyl ester ILs 1 and 2 had lower toxicity, IC50 ¼ 6.25–12.5 mM for IL 5, 10 and 9 to several
toxicity (IC50 ¼ 25–200 mM) than the ethyl ester ILs 3 and 4 (IC50 bacterial strains).
¼ 25–100 mM), which in turn had lower toxicity than the n-butyl
The microbial toxicity of the mandelic acid derived ILs 1–10
ester/amide ILs 5–8 (IC50 ¼ 6.25–50 mM). This observed trend is is clearly inuenced by the lipophilicity of the IL structures. An
expected as the literature shows increasing the length of alkyl increase in microbial toxicity over the previous class of man-
groups of ILs leads to higher antimicrobial activity.7 This is also delic acid derived ILs (19–22) is observed due to the removal of
consistent with microbial toxicity data for linear alkylbenzene the phenyl ring substitution and subsequent increase in IL
sulfonates and quaternary ammonium compounds.65,66 ILs with lipophilicity. It has been shown in the literature that decreasing
longer alkyl chains have a higher lipophilicity, which increases the lipophilicity of imidazolium ILs by introducing ether groups
their ability to derange/interact or permeate through the into alkyl side chains can reduce IL microbial toxicity.68 In an
bacterial cell membrane and disrupt it leading ultimately to cell effort to decrease the microbial toxicity of the mandelic acid
death.62,67 IL 3 did not present with the same inhibition towards derived ILs 1–10, the replacement of the alkyl side chains with
E. coli (IC50 ¼ 25–50 mM) as observed in the previous bacterial ethereal side chains was considered. However, as it is recom-
toxicity run (IC95 ¼ 2 mM) (Table 2), which is attributed to two mended that both methoxy- and ethoxy-terminal substitution
different strains of E. coli used. There was no observed trend for are avoided with the aim of preventing any concurrent decrease
the level of bacterial inhibition between 1-methylimidazolium
ILs and pyridinium ILs across the various alkyl chain lengths.
When comparing the IC50 values of the n-butyl ester ILs with
pyridinium (6), 3-methoxy pyridinium (7) and 3-(ethoxy-
carbonyl)pyridinium (8) headgroups, a difference in the inhi-
bition was only observed for IL 7, which showed the same
increase in inhibition towards B. subtilis and P. uorescens
(IC50 ¼ 12.5–25 mM). When comparing the inhibition of the
Gram-positive bacteria E. coli with the Gram-negative bacteria
B. subtilis, seven ILs (3–6 and 8–10) inhibited E. coli to a greater
extent than B. subtilis. The remaining ILs (1, 2 and 7) inhibited
both E. coli and B. subtilis to the same extent. High susceptibility
of Gram-positive bacteria over Gram-negative bacteria has
previously been highlighted in the literature and has been
linked to IL lipophilicity.16
When comparing the IC50 values of the 1-methyl-
imidazolium and pyridinium methyl and n-butyl ester ILs (1, 2,
5 and 6) to the mandelic acid derived ILs previously published Fig. 3 Percentage biodegradation results for pyridinium ILs with
benzyl (23) and methyl (24) analogues synthesised by Gore et al.36 and
with the same ester alkyl chain length and headgroup32 (19–22)
the PTPs (23 PTP and 24 PTP). Biodegradation values assessed by the
CO2 Headspace test ISO 14593.
(Fig. 2 and Table 3), in all cases the previous class of ILs had
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 2115–2126 | 2121

View Article Online
RSC Advances
Paper
Biodegradation studies
The aerobic biodegradability of the ten mandelic acid derived
ILs 1–10 (biodegradation data, Table 4) were evaluated based on
the Closed Bottle test (CBT) (OECD 301D) (see ESI† for
method).71 This test is the most stringent of the OECD series as
it works with the lowest density and functional diversity of
bacteria. It was selected to make sure that any compounds
which are classied as readily biodegradable are biodegraded
everywhere in the aquatic environment. The validity of the test
run was conrmed by satisfying the criteria for validity
according to the test guidelines, which includes the reference
substance sodium acetate reaching >60% biodegradation
within 14 days and the concentration of oxygen in the bottles
not decreasing below 0.5 mg Lꢁ1 at any time.
A percentage of theoretical ester/amide biodegradation can
be calculated for a compound by assuming that all the ester/
amide carbons in the molecule are completely mineralised,
i.e. broken down to carbon dioxide and water during biodeg-
radation (oxidation of nitrogen by nitrication has to also be
taken into account when calculating theoretical values of
oxygen demand). The calculated theoretical ester/amide
biodegradation values can then be compared to the experi-
mental biodegradation values. If the values are similar it is
reasonable to assume that biodegradation of the compound is
probably occurring at the ester/amide. Signicantly lower
theoretical values than the experimental ones would suggest
that other parts of the molecule are also likely to be bio-
degrading. Theoretical values that are signicantly higher than
the experimental values would suggest the ester/amide bond is
likely not biodegrading, or only partially biodegrading. The
theoretical ester/amide biodegradation calculations for ILs 1–10
were performed and the values were compared to the experi-
mental results obtained in the CBT (Table 4).
Fig. 4 Percentage biodegradation results for pyridinium ILs (25 and
26) synthesised by Coleman et al.34 Biodegradation values assessed by
the CO2 Headspace test ISO 14593.
in IL biodegradability,68 those ether derivatives were not
selected as target ILs for this study.
The increase in IL toxicity that is observed with increasing
alkyl chain length of the IL cation has been widely pub-
lished.7,10,11,15,16,69,70 This observation presented for the mandelic
acid derived ILs 1–10, where IL toxicity increased in the order
methyl ester < ethyl ester < n-butyl ester/amide. Again, this is
linked to an increase in IL lipophilicity where ILs with longer
alkyl chain lengths have a higher lipophilicity and a greater
interaction with cell membranes.
A
quantitative structure–property relationship (QSPR)
assessment carried out by Couling et al. for the marine bacteria
Vibrio scheri suggested that imidazolium ILs are more toxic
than pyridinium ILs.69 They propose this is due to an increased
number of nitrogen atoms with two aromatic and one single
bond in the imidazolium ring. However, in the work presented
herein no clear conclusion could be drawn on the effect of the
imidazolium and pyridinium headgroups on the microbial
toxicity of the mandelic acid derived ILs 1–10.
Biodegradation screening results
The results show that none of the ILs reached >60% biodegra-
dation in 28 days and thus cannot be classied as readily
biodegradable by the OECD 301D CBT (Table 4). The 3-(ethox-
ycarbonyl)pyridinium n-butyl ester IL (8) had the highest level of
biodegradation (31% in 28 days). The 1-methylimidazolium n-
butyl amide IL (9) and pyridinium n-butyl amide IL (10) did not
biodegrade at all under the test conditions (0–1% in 28 days).
The percentage biodegradation reached by each IL in the test
Fig. 5 Percentage biodegradation results for fully mineralisable pyr-
idinium IL (27) synthesised by Jordan et al.35 and Haib et al.37 Biodeg-
radation values assessed by the CBT OECD 301D.
Fig. 6 Percentage biodegradation results for pyridinium (28) and nicotinic acid derived (29 and 30) ILs synthesised by Harjani et al.73 and
mandelic acid derived ILs 4 and 8. Biodegradation values assessed by the CBT OECD 301D.
2122 | RSC Adv., 2017, 7, 2115–2126
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
level of biodegradation (CO2 Headspace test ISO 14593) was
published by Gore et al. for the benzylimidazolium IL 23 (6%),
compared to the methylimidazolium IL 24 (31%). We propose
this is in part due to a higher share of carbon content of the
proposed transformation product (PTP) of IL 23 (23 ¼ 18C vs. 23
PTP ¼ 14C, 22% decrease in C) compared to 24 PTP (24 ¼ 12C
vs. 24 PTP ¼ 8C, 33% decrease in C), which are both assumed to
persist (Fig. 3). While hydrolysis of all three ester groups in both
ILs 23 and 24 to give the fully hydrolysed 23 PTP and 24 PTP is
reasonable under the test conditions, only the biodegradation
value for IL 24 (31%) supports this complete ester hydrolysis.
The biodegradation of IL 23 (6%) is so low that signicant biotic
ester hydrolysis cannot be stated. A comprehensive LC-MS study
would be required to conrm the structure of transformation
products and was beyond the scope of the study. In contrast to
the result reported herein are the high levels of biodegradation
reported by Coleman et al.34 (Fig. 4, CO2 Headspace test ISO
14593), Jordan et al.35 and Haib et al.37 (Fig. 5, CBT) for several
phenyl containing amino acid ILs.
The effect of a phenyl group attached to the same carbon
atom as the headgroup can be evaluated by comparing the
biodegradation of ILs 4 and 8 to previously reported ILs with
pyridinium headgroups (28) and nicotinic acid derived head-
groups (29 and 30) (Fig. 6).73 The nicotinic acid derived IL (8)
presented herein gave signicantly lower levels of biodegrada-
tion (31%) than the readily biodegradable nicotinic acid derived
ILs 29 and 30 (72 and 84% respectively). Also, pyridinium IL 4
gave signicantly lower levels of biodegradation (16%) than the
comparative readily biodegradable pyridinium IL 28 (87%). This
suggests that a-substitution of the N-heterocycle with a phenyl
group (from the mandelic acid starting material) has a signi-
cant negative impact on the biodegradability of the ILs. We
propose this is a new rule of thumb to be considered when
designing biodegradable compounds. It should also be noted
that even lower levels of biodegradation were recorded for the
amide ILs 9 and 10, which were expected for reasons previously
described.
The comparison of the theoretical ester/amide biodegrada-
tion values to the experimental values for ILs 1–10 (Table 4)
show that a good prediction for the ester ILs 1–8 is obtained.
This analysis allows for the conclusion that the biodegradation
of the ester ILs is most likely occurring only at the ester bonds. A
clear trend is observed across the series in the order of methyl
ester ILs (5–8% biodegradation), ethyl ester ILs (16–20%
biodegradation), n-butyl ester ILs (21–27% biodegradation) and
ethyl/n-butyl ester IL 8 (31% biodegradation). When consid-
ering the condence levels of the test, even a prediction of 6%
lower theoretical biodegradation compared to that observed
experimentally (IL 3) is likely not signicant enough to suggest
that additional parts of the molecule (i.e. the 1-methyl-
imidazolium ring) are breaking down. This conclusion is
additionally supported by the literature which shows that 1-
methylimidazolium based ILs are oen poorly biodegradable.74
The calculated theoretical biodegradation values for the 1-
methylimidazolium n-butyl amide IL (9) and the pyridinium n-
butyl amide IL (10) are much higher than the biodegradation
results obtained experimentally (+25 and +22% respectively).
Fig. 7 Structures of the PTPs for ILs 1–8 (31–34).
can be rationalised as biodegradation of the ester/amide groups
(side chain) of the ILs. It is well known that ester groups can
facilitate high levels of biodegradability and are oen intro-
duced into compounds to increase their levels of biodegrad-
ability,68,72 although there are some exceptions.35 IL 8 contains
the largest percentage of ester groups per molecule, with both
an ethyl and n-butyl ester, thus it is not surprising that it shows
the highest level of biodegradation in the CBT. The CBT results
also show, as expected, that the level of biodegradation
increased as the ester chain length increased. This trend is
observed for the 1-methylimidazolium ILs 1, 3 and 5 (5, 20, 27%
biodegradation in 28 days, for methyl, ethyl and n-butyl esters
respectively) and the pyridinium ILs 2, 4 and 6 (8, 16, 21%
biodegradation in 28 days, for methyl, ethyl and n-butyl esters
respectively). Amide moieties have been shown in the literature
to have poor levels of biodegradation,12,13,68 thus the low levels of
biodegradation observed for the 1-methylimidazolium (9) and
a pyridinium (10) n-butyl amide ILs in the CBT (0 and 1%
biodegradation in 28 days, respectively) are expected. The CBT
results of the n-butyl ester ILs 5–8 were in the range 21–31%.
This narrow range of values does not support a detailed analysis
of the effect of the headgroup (pyridinium, 3-methoxy pyr-
idinium, 1-methylimidazolium and 3-(ethoxycarbonyl)pyr-
idinium) on biodegradation.
This class of compounds (1–10) presents with low levels of
biodegradation similar to the benzylimidazolium ILs previously
reported by Gore et al.36 Benzylimidazolium ILs containing
amide groups showed very low levels of biodegradation (0–4%),
a similar result to the amide ILs 9 and 10 reported herein.
Biodegradation of a benzylpyridinium IL was also shown to
increase with the introduction of an ethyl ester group (by 23%),
a result which was also observed in this study for IL 8. A lower
Table 5 PTPs for ILs 1–8 (31–34)
Parent IL
PTPs
1, 3, 5
2, 4, 6
7
8
31
32
33
34
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 2115–2126 | 2123

View Article Online
RSC Advances
Paper
The calculation is made assuming that biodegradation of the transformation products is also always recommended as they
amide is occurring; however, from the experimental values we may be more toxic than the parent compounds.77 For toxicity
can see that for IL 9 and 10 no biodegradation occurred at all studies to be undertaken, synthesis of the individual PTPs
under the conditions of the CBT in the 28 day period. This is in needs to be carried out. However, with the same PTPs pre-
accordance with literature that reports amides to have poor senting for multiple ILs, only a small number of compounds
levels of biodegradation.68 Thus, the calculation overestimates would need to be synthesised to obtain toxicity/biodegradation
the level of biodegradation expected for the amide ILs as the data on the transformation products for multiple parent
experiments show that the amides did not biodegrade under the compounds.
CBT conditions over 28 days.
The 3-(ethoxycarbonyl)pyridinium n-butyl ester IL (8)
contains two ester groups which may have different rates of
biodegradation in the CBT, leading to the production of two
additional transformation products (35 and 36) (Scheme 3). A
comparison of the theoretical ester/amide biodegradation of IL
8 to the experimental biodegradation (30 and 31% respectively)
shows that they are very similar. This suggests that the
biodegradation of IL 8 did not stop at either of the mono ester
transformation products (35 and 36) during the 28 day period of
the CBT. Both biodegradation pathways may have the same
kinetic rate leading to the predominant transformation product
34 (Scheme 3) at the end of the CBT. However, a more detailed
LC-HRMS study would be required to determine which of the
pathways are followed.
Biodegradation proposed transformation products
When a compound biodegrades a number of biodegradation
transformation products are produced that might not be
accessible to further biodegradation (so called persistent
transformation products).14 The PTPs of ILs 1–8 formed by ester
breakdown are the carboxylic acid derivatives of the parent ILs
(23–26) (Fig. 7 and Table 5). Amide ILs 9 and 10 give the same
PTPs as the ester ILs (31 and 32 respectively), although no
biodegradation was observed in the CBT. Alkylamines have
previously shown high levels of degradation in biodegradation
tests (oxygen consumption, colorimetric, total organic carbon
(TOC) and CO2 production measurements).64 We propose that if
amide hydrolysis had occurred the biodegradation of n-butyl-
amine would have been detected by the CBT. In our test there is
no evidence of this breakdown. Transformation products can be
detected aer biodegradation studies by carrying out LC-MS,
MS/MS, LC-HRMS and/or NMR analysis on the resultant
biodegradation stream.37,75,76 An analysis of the toxicity of the
Conclusion
A series of ten bromide ILs (1–10) derived directly from the
biorenewable natural product mandelic acid was synthesised.
The series was selected based on diverse structural features
(including n-alkyl chains, ester/amide bonds and various N-
heterocyclic headgroups), thus allowing for the investigation
of their effect on IL toxicity and biodegradation. None of the ILs
showed high antimicrobial activity to the twenty bacterial and
fungal strains screened as a preliminary toxicity assay. However,
moderate toxicity was observed for the 1-methylimidazolium
ethyl ester (3) against several of the bacterial strains. Subse-
quent IC50 determination showed that the majority of the ILs
presented with low toxicity towards one Gram-positive bacterial
strain and four Gram-negative bacterial strains when high IL
concentrations (up to 200 mM) were evaluated. IL toxicity across
all bacterial strains increased as the alkyl chain length
increased in the order methyl ester < ethyl ester < n-butyl ester/
amide. Despite observing in general a higher antimicrobial
activity for the ILs (1–10) compared to the previously syn-
thesised highly oxygenated mandelic acid derived ILs (19–22), it
is a substantial positive result that the ILs herein do not
represent a large increase in the level of inhibition. Low toxicity
ILs (1–10) were synthesised in a shorter and more efficient
synthesis directly from mandelic acid (c.f. 19–22) satisfying the
4th and 7th principles of green chemistry: designing safer
chemicals and the use of renewable feedstocks (respectively).
The biodegradation studies of the ILs in the CBT showed
biodegradation in the range of 5–31% in 28 days for the man-
delic acid derived ester ILs (1–8). The mandelic acid derived
amide ILs 9 and 10 demonstrated no biodegradation in the CBT
(0 and 1% in 28 days, respectively). The CBT results of the n-
butyl ester ILs 5–8 were in the narrow range (21–31%) which
does not support a detailed analysis of the effect of the
Scheme 3 Proposed biodegradation pathways for IL 8.
2124 | RSC Adv., 2017, 7, 2115–2126
This journal is © The Royal Society of Chemistry 2017

View Article Online
Paper
RSC Advances
headgroup. The biodegradation pathway proposed for the ILs is
through ester alkyl chain biodegradation to a carboxylic acid
9 M. Petkovic, K. R. Seddon, L. P. N. Rebelo and C. S. Pereira,
Chem. Soc. Rev., 2011, 40, 1383.
derivative. This conclusion is based on previous results in the 10 J. Ranke, S. Stolte, R. Stormann, J. Arning and B. Jastorff,
literature, and was reinforced by comparing the results of Chem. Rev., 2007, 107, 2183.
a theoretical ester/amide biodegradation calculation to the 11 T. P. T. Pham, C.-W. Cho and Y.-S. Yun, Water Res., 2010, 44,
results obtained in the CBT. 352.
An overall goal is to develop ILs which biodegrade fully (i.e. 12 D. Coleman and N. Gathergood, Chem. Soc. Rev., 2010, 39,
mineralise completely) by breaking down via transient trans- 600.
formation products, and to expand our understanding of rules 13 A. Jordan and N. Gathergood, Chem. Soc. Rev., 2015, 44, 8200.
of thumb to design biodegradable chemicals. As we and others 14 S. Stolte, S. Steudte, A. Igartua and P. Stepnowski, Curr. Org.
work towards this goal, persistent transformation products
have been proposed and identied.13,35,37 It is an ongoing chal- 15 K. S. Egorova and V. P. Ananikov, ChemSusChem, 2014, 7,
lenge to design from the outset a series of compounds (which 336.
includes predicted transformation products) that lead to the 16 M. Amde, J.-F. Liu and L. Pang, Environ. Sci. Technol., 2015,
Chem., 2011, 15, 1946.
greatest optimisation of a structure activity relationship (SAR)
for biodegradation (mineralisation).
49, 12611.
17 X. D. Hou, Q. P. Liu, T. J. Smith, N. Li and M. H. Zong, PLoS
One, 2013, 8, e59145.
Biodegradation studies of ILs containing a phenyl ring
demonstrate that examples are known which can pass readily 18 Q.-P. Liu, X.-D. Hou, N. Li and M.-H. Zong, Green Chem.,
biodegradable tests (25, 26, 28, 29, 30), can be classed as min- 2012, 14, 304.
eralisable (27), or not pass the readily biodegradable tests (23). 19 N. Ferlin, M. Courty, A. N. Van Nhien, S. Gatard, M. Pour,
The mandelic acid derived ILs (1–10) fall into the latter category
and can be included in the rules of thumb for the design of
biodegradable ILs.
B. Quilty, M. Ghavre, A. Haiss, K. Kummerer,
N. Gathergood and S. Bouquillon, RSC Adv., 2013, 3, 26241.
20 N. Ferlin, M. Courty, S. Gatard, M. Spulak, B. Quilty,
I. Beadham, M. Ghavre, A. Haiss, K. Kummerer,
N. Gathergood and S. Bouquillon, Tetrahedron, 2013, 69,
6150.
Acknowledgements
The authors HP, AHeise and NG have received funding from the 21 S. Kirchhecker, S. Troger-Muller, S. Bake, M. Antonietti,
European Union's Seventh Framework Programme for research, A. Taubert and D. Esposito, Green Chem., 2015, 17, 4151.
technological development and demonstration under grant 22 C. A. Fewson, FEMS Microbiol. Lett., 1988, 54, 85.
agreement No. 289253. NG acknowledges funding from the 23 D. J. D. Hockenhull, A. D. Walker, G. D. Wilkin and
European Union's Seventh Framework Programme for research,
technological development and demonstration under grant 24 A. F. Landymore, N. J. Antia and G. H. N. Towers, Phycologia,
agreement #621364 (TUTIC-Green). Antimicrobial screening 1978, 17, 319.
(MS) was supported by Project No. 15-07332S of the Czech 25 J. W. Walker and V. K. Krieble, J. Chem. Soc., Trans., 1909, 95,
Science Foundation. The authors also wish to also acknowledge
1369.
the contribution of Florian Gross for antimicrobial screening 26 E. Micklander, L. Brimer and S. B. Engelsen, Appl. Spectrosc.,
F. G. Winder, Biochem. J., 1952, 50, 605.
carried out at Dublin City University.
2002, 56, 1139.
27 P. D. A. Singh, J. R. Jackson and S. P. James, Biochem.
Pharmacol., 1985, 34, 2207.
28 M. J. R. Nout, G. Tuncel and L. Brimer, Int. J. Food Microbiol.,
1995, 24, 407.
29 K. Yamamoto, K. Oishi, I. Fujimatsu and K. Komatsu, Appl.
Environ. Microbiol., 1991, 57, 3028.
30 Z. Wang, in Comprehensive Organic Name Reactions and
Reagents, John Wiley & Sons, Inc., 2010, vol. 2, ch. 408,
pp. 1816–1819.
References
1 N. V. Plechkova and K. R. Seddon, Chem. Soc. Rev., 2008, 37,
123.
2 S. Sowmiah, V. Srinivasadesikan, M. C. Tseng and Y. H. Chu,
Molecules, 2009, 14, 3780.
3 P. T. Anastas and J. C. Warner, Green Chemistry: Theory and
Practice, Oxford University Press, New York, 1998.
4 M. T. Garcia, N. Gathergood and P. J. Scammells, Green 31 Z. Sun, Y. Ning, L. Liu, Y. Liu, B. Sun, W. Jiang, C. Yang and
Chem., 2005, 7, 9. S. Yang, Microb. Cell Fact., 2011, 10, 71.
5 N. Gathergood, P. J. Scammells and M. T. Garcia, Green 32 S. P. M. Ventura, M. Gurbisz, M. Ghavre, F. M. M. Ferreira,
Chem., 2006, 8, 156.
6 N. Gathergood, M. T. Garcia and P. J. Scammells, Green
Chem., 2004, 6, 166.
7 J. Pernak, K. Sobaszkiewicz and I. Mirska, Green Chem., 2003,
5, 52.
F. Goncalves, I. Beadham, B. Quilty, J. A. P. Coutinho and
N. Gathergood, ACS Sustainable Chem. Eng., 2013, 1, 393.
33 J. Ranke, A. Muller, U. Bottin-Weber, F. Stock, S. Stolte,
J. Arning, R. Stormann and B. Jastorff, Ecotoxicol. Environ.
Saf., 2007, 67, 430.
8 J. Ranke, K. Molter, F. Stock, U. Bottin-Weber, J. Poczobutt, 34 D. Coleman, M. Spulak, M. T. Garcia and N. Gathergood,
J. Hoffmann, B. Ondruschka, J. Filser and B. Jastorff,
Ecotoxicol. Environ. Saf., 2004, 58, 396.
Green Chem., 2012, 14, 1350.
This journal is © The Royal Society of Chemistry 2017
RSC Adv., 2017, 7, 2115–2126 | 2125

View Article Online
RSC Advances
Paper
35 A. Jordan, A. Haiß, M. Spulak, Y. Karpichev, K. Kummerer 58 M27-A3, Reference Method for Broth Dilution Antifungal
and N. Gathergood, Green Chem., 2016, 18, 4374.
36 R. G. Gore, L. Myles, M. Spulak, I. Beadham, T. M. Garcia,
S. J. Connon and N. Gathergood, Green Chem., 2013, 15,
2747.
37 A. Haiß, A. Jordan, J. Westphal, E. Logunova, N. Gathergood
and K. Kummerer, Green Chem., 2016, 18, 4361.
38 R. S. Boethling, E. Sommer and D. DiFiore, Chem. Rev., 2007,
107, 2207.
Susceptibility Testing of Yeasts; Approved Standard Institute,
Clinical and Laboratory Standards Institute, Wayne, PA,
3rd edn, 2008.
59 M38-A2, Reference Method for Broth Dilution Antifungal
Susceptibility Testing of Filamentous Fungi; Approved
Standard Institute, Clinical and Laboratory Standards
Institute, Wayne, PA, 2nd edn, 2008.
60 M. Ghavre, PhD thesis, Dublin City University, 2012.
39 S. I. T. Kennedy and C. A. Fewson, J. Gen. Microbiol., 1968, 53, 61 M. Gurbisz, PhD thesis, Dublin City University, 2012.
259.
62 M. C. Bubalo, K. Radosevic, I. R. Redovnikovic, J. Halambek
and V. G. Srcek, Ecotoxicol. Environ. Saf., 2014, 99, 1.
63 C.-W. Cho, Y.-C. Jeon, T. P. T. Pham, K. Vijayaraghavan and
Y.-S. Yun, Ecotoxicol. Environ. Saf., 2008, 71, 166.
64 K. Yoshimura, S. Machida and F. Masuda, J. Am. Oil Chem.
Soc., 1980, 57, 238.
40 M. Jamaluddin, P. V. S. Rao and C. S. Vaidyanathan, J.
Bacteriol., 1970, 101, 786.
41 C. S. Harwood and R. E. Parales, Annu. Rev. Microbiol., 1996,
50, 553.
42 S. G. Bhat and C. S. Vaidyanathan, J. Bacteriol., 1976, 127,
¨
¨
1108.
65 H. Sutterlin, R. Alexy, A. Coker and K. Kummerer,
´
43 A. Kowalkowska and A. Jonczyk, Synth. Commun., 2011, 41,
Chemosphere, 2008, 72, 479.
3308.
66 T. Ikeda, H. Yamaguchi and S. Tazuke, J. Bioact. Compat.
Polym., 1990, 5, 31.
44 B. Riegel and H. Wittcoff, J. Am. Chem. Soc., 1946, 68, 1913.
45 C. Zhang, H. Ito, Y. Maeda, N. Shirai, S.-i. Ikeda and Y. Sato, 67 D. B. Zhao, Y. C. Liao and Z. D. Zhang, Toxicity of ionic
J. Org. Chem., 1999, 64, 581.
liquids, Clean: Soil, Air, Water, 2007, 35(1), 42–48.
68 S. Morrissey, B. Pegot, D. Coleman, M. T. Garcia,
D. Ferguson, B. Quilty and N. Gathergood, Green Chem.,
2009, 11, 475.
69 D. J. Couling, R. J. Bernot, K. M. Docherty, J. K. Dixon and
E. J. Maginn, Green Chem., 2006, 8, 82.
46 Y. Su, L. Zhang and N. Jiao, Org. Lett., 2011, 13, 2168.
47 Z. Wang, in Comprehensive Organic Name Reactions and
Reagents, John Wiley & Sons, Inc., New Jersey, 2010, vol. 1,
ch. 239, pp. 1095–1097.
˜ ´
48 O. Kuisle, E. Quinoa and R. Riguera, J. Org. Chem., 1999, 64,
8063.
70 X. F. Wang, C. A. Ohlin, Q. H. Lu, Z. F. Fei, J. Hu and
P. J. Dyson, Green Chem., 2007, 9, 1191.
71 Organization for Economic Cooperation and Development
(OECD), OECD Guideline for Testing of Chemicals; Ready
Biodegradability, Paris, July 1992.
49 N. Shakya, K. K. Roy and A. K. Saxena, Bioorg. Med. Chem.,
2009, 17, 830.
50 S.-S. Weng, H.-C. Li and T.-M. Yang, RSC Adv., 2013, 3, 1976.
51 G. C. Lloyd-Jones, P. D. Wall, J. L. Slaughter, A. J. Parker and
D. P. Laffan, Tetrahedron, 2006, 62, 11402.
72 S. Morrissey, PhD Thesis, Dublin City University, 2008.
52 B. R. Kim, H.-G. Lee, S.-B. Kang, G. H. Sung, J.-J. Kim, 73 J. R. Harjani, R. D. Singer, M. T. Garcia and P. J. Scammells,
J. K. Park, S.-G. Lee and Y.-J. Yoon, Synthesis, 2012, 44, 42. Green Chem., 2009, 11, 83.
53 L. J. Gooßen, D. M. Ohlmann and P. P. Lange, Synthesis, 74 S. Stolte, S. Abdulkarim, J. Arning, A. K. Blomeyer-Nienstedt,
2009, 2009, 160.
54 P. C. Astles, T. J. Brown, F. Halley, C. M. Handscombe,
U. Bottin-Weber, M. Matzke, J. Ranke, B. Jastorff and
J. Thoming, Green Chem., 2008, 10, 214.
N. V. Harris, T. N. Majid, C. McCarthy, I. M. McLay, 75 K. M. Docherty, M. V. Joyce, K. J. Kulacki and C. F. Kulpa,
A. Morley, B. Porter, A. G. Roach, C. Sargent, C. Smith and
R. J. A. Walsh, J. Med. Chem., 2000, 43, 900.
55 K. Ohno, Y. Ma, Y. Huang, C. Mori, Y. Yahata, Y. Tsujii,
T. Maschmeyer, J. Moraes and S. Perrier, Macromolecules,
2011, 44, 8944.
Green Chem., 2010, 12, 701.
76 Y. Deng, I. Beadham, M. Ghavre, M. F. Costa Gomes,
N. Gathergood, P. Husson, B. Legeret, B. Quilty,
M. Sancelme and P. Besse-Hoggan, Green Chem., 2015, 17,
1479.
´
56 K. R. Seddon, Presented in part at the COIL-6, Jeju, South 77 A. B. A. Boxall, C. J. Sinclair, K. Fenner, D. Kolpin and
Korea, 2015. S. J. Maud, Environ. Sci. Technol., 2004, 38, 368a.
¨
57 K. Kummerer, A. Al-Ahmad and A. Henninger, Acta
Hydrochim. Hydrobiol., 2002, 30, 171.
2126 | RSC Adv., 2017, 7, 2115–2126
This journal is © The Royal Society of Chemistry 2017