Abstract- Ionic liquids are compounds of high interest for industry because of their attractive properties as solvents, but the water solubility of these compounds may lead to aquatic pollution and related risks. Experimental toxicity evaluation (Daphnia magna EC50) is a measurement of aquatic toxicity but there are theoretically over 1 trillion ionic liquids, which makes it necessary to estimate their properties by means of quantitative structure-activity relationships (QSARs). In this work, a novel QSAR based on multiple linear regression method is applied to estimate the ecotoxicity of ionic liquids. A data set of Daphnia magna EC50 was assembled to develop a novel group contribution method for estimating the EC50 of ionic liquids. The results illustrated that the data range covered for logEC50 values in between 2.07 and -4.33. From the results the contributions of anion, cation and alkyl substitutions has been established and found a good fitting value for predicting the EC50 by using SPSS software 11.5.
I. INTRODUCTION
Ionic Liquids (ILs) are, a group of molten salts containing only cations and anions, having low melting points, less than 100°C [1]. Generally, ILs possess a number of interesting properties, including negligible vapour pressure, high thermal, chemical and electrothermal stabilities, and excellent solvent behavior for a wide range of inorganic and organic materials [2, 3]. Due to these unique properties, ILs has been used for different industrial application, such as catalysis, media for liquid crystals, extraction in electrochemistry, and separation processes [4].
As ILs are nonvolatile, they are comparatively harmless to the atmosphere, but due to their high solubilities in water, it must be considered that a release of ILs from industrial processes into aquatic environments may lead to water pollution and related potential risks [5]. The effects of ionic liquids on aquatic organisms have recently been reported for the marine bacterium vibrio fischeri [1, 6], algae [7, 8], the freshwater crustacean daphnia magna [9], the freshwater snail physa acuta [10], and the zebra fish danio rerio [11]. Some of these studies also revealed that ILs toxicity increased with increasing alkyl chain length [1, 10]. In view to the applications and current interest of ILs, it is important to develop new procedures for the estimations of the toxicity of ILs.
The use of bioassays on standard test organisms represents fundamental approach in the definition of ecological risk in the aquatic environment for promising chemicals as ILs [12]. The environmental hazard assessment of chemicals consists of the identification of the effects that chemicals may have on organisms in the environment and the determination of the concentration of the chemical below which adverse effects in the environmental sphere of concern (e.g., aquatic) are not expected to occur. Development of safer and environmentally friendly processes and products is needed to achieve sustainable production and consumption patterns. Experimental determination of ecotoxicity is required to evaluate the environmental fate of a compound to improve the information related to the factors leading to environmental risks. Several standard procedures with biological systems have been established to evaluate the aquatic toxicity [5]. However ecotoxicological assays are time and material consuming and can significantly differ. Furthermore, there are theoretically over a trillion of ILs, although fewer than one thousand have been reported. For this reason, a method is needed to estimate environmental fate parameters of these agents. Fortunately, it may be possible to predict some of the important properties by means of structure-activity correlations [13]. Structure-activity relationships and quantitative structure-activity correlations, referred to as SARs and QSARs, respectively, are models that can be used to predict the physicochemical and biological properties of molecules. The basis for any (Q)SAR is that the biological activity of a new or untested chemical can be inferred from the molecular structure or other properties of similar compounds whose activities have already been assessed. In toxicology, changes in the structure of a chemical may influence the type and potency of its toxic action; thus models such as structure-activity relationships are used to represent, explain, and, most importantly, predict phenomena of interest. QSARs are employed as scientifically credible tools for predicting the acute toxicity of chemicals when few empirical data are available [14].
Group contribution methods have been applied to estimate physicochemical properties of interest in the environmental behavior of chemicals. These include melting point, boiling point, vapor pressure, Henry's Law constant, octanol-water partition coefficient, and water solubilities [15]. Some studies have been performed on the relationship between toxicity and chemical structure for several compounds [16, 17], several QSARs which predict toxicity values for Daphnia, green algae, and fish [18,19] have been developed and some QSARs estimate Vibrio fischeri toxicity for specific groups of compounds using molecular and physicochemical descriptors [20, 21]. Ionic liquids are becoming a priority solvent and they are being evaluated in order to achieve useful correlations for toxicity [5, 22].
A detail study for the prediction of EC50 values have been done by Luis et al., [5], with the help of group contribution method using Polymath software to develop the QSAR, a database of log EC50 (logarithm of 15 or 30 minute toxicity to Vibrio fischeri) for ILs had been established from experiments and literature. According to group contribution methods, properties of a molecule can be assumed to be the summation of the contributions of its atoms and/or fragments. The main contributions: anions (A), cations (C), and substitutions (S) were studied. Likewise contribution was observed where, cation have much effect and anion are having negligible effect [1, 6]. The cations were imidazolium, pyridinium, and pyrrolidinium (Imida, Pyrid, and Pyrrol) respectively.
Daphnia magna is one of the most widely used water crustacean around the world, it is also an important link in the food web of freshwater communities and has been the international standard model animal for toxicity testing studies [23]. Therefore, D. magna was used as a model organism to evaluate the toxicity of Ionic liquids, and the related biological responses induced by the imidazolium based ionic liquids. Bernot et al., [9] reported the letheal concentration values of some imidazolium ionic liquids with daphnia, where they compared these data with some common solevents and salts, and found the contribution of the cations for toxic effects. Data accumulation was being done for daphnia from the literatures [3, 10, 12, 24-34].
The aim of this work was to establish a knowledge based model to predict the toxicity in most widely used aquatic organism daphnia magna. A novel group contribution method has been developed and the contribution of the groups has been calculated from EC50 data for imidazolium-, pyridinium-, choline, sulfonium, phosphonium, ammonium and pyrrolidinium-based ionic liquids, to discern their influence and contributions to toxicity. During this study, multilinear regression analysis has been followed [35]. Besides that, the software SPSS 11.5 has been explored for multilinear regression analysis application.
II. EXPERIMENT AND METHODS
EC50 data
The EC50 values for Daphnia magna was collected from different literatures [9, 12, 24, 25 and 26]. All the data have been converted to the EC50 considering the molecular weight and the structure of ILs used. The 48-h acute toxicity test was conducted by the standard procedure of OECD guideline 202.
Theory
Multilinear regression analysis will be used to calculate the group (anions, cations and substitutions) contribution on toxicity. Regression analysis is a conceptually simple method for investigating functional relationships among variables. The relationship is expressed in the form of an equation or a model connecting the response or dependent variable and one or more explanatory or predictor variables. In this research, the response or dependent would be dimensionless toxicity Y* which is expressed in Eq. (1) while the explanatories or predictors are anions, cations and substitutions.
We denote the response variable by Y and the set of predictor variables by X1, X2, ..., XP, where p denoted the number of predictor variables. The true relationship between Y and X1, X2, ..., XP can be approximated by the regression model:
(1)
Where, ε is assumed to be a random error representing the discrepancy in the approximation. It accounts for the failure of the model to fit the data exactly. The function f(X1, X2, ..., XP) describes the relationship between Y and X1, X2, ..., XP. An example is the linear regression model [Eq. (2)]:
(2)
Where, , is called the regression parameters or coefficients, are unknown constant to be determined (estimated) from the data. We follow the commonly used notational convention of denoting unknown parameter by Greek letters.
The predictor or explanatory variables are also called by other names such as independent variables, covariates, regressors, factors, and carriers. The name independent variable, though commonly used, is the least preferred because in practice the predictor variables are rarely independent of each other.
Methodology
This paper will focus on Daphnia magna as tested organism. The multilinear regression analysis method has been used to predict toxicity of ionic liquids. The analysis follows the following steps:
Collect log EC50 data for ionic liquid toxicity on Daphnia magna.
2. To avoid the influence of the absolute values of the data to group contribution method calculations, the dimensionless toxicity, Y*, between 0 and 1 is defined for all ionic liquids [Eq. (3)].
(3)
Where logEC50_max and logEC50_min are, respectively, the maximum and minimum values of logEC50 in the database showing the application range of the model.
3. Analyze structure of ionic liquids in a table form (Table 1).
Table 1: ILs toxicities in µmol/L (log EC50), dimensionless toxicity (Y), and group contribution descriptors.
Compound
LogEC50 (µmol/L)
Y
Br-
Cl-
BF4-
PF6-
TF2N-
Imida
Pyrid
Ammon
Pyrrol
R
R1
bmimCl
1.85
0.05
0
1
0
0
0
1
0
0
0
4
1
bmimBr
1.56
0.11
1
0
0
0
0
1
0
0
0
4
1
bmimBF4
1.79
0.06
0
0
1
0
0
1
0
0
0
4
1
Where Br-, Cl-, BF4-, PF6- and TF2N- are set of anions; Imida, Pyrid, Ammon and Pyrro are Imidazolium, Pyridinium, Ammonium and Pyrrolidinium respectively; R is long-alkyl-chain substitution and R1 is an additional short chain (methyl). These variables have a nonzero value when the group is present in the molecule.
4. Estimate QSAR by summation of the contribution of each group as shown in Eq. (4).
(4)
Where Ai,, Cj and Sk are the molecular descriptors for ionic liquids, ai, cj, and sk are the contribution of each group of the toxicity, and the summation is taken over all group, subscripts indicate anions (i), cations (j), and substitutions (k).
To estimate QSAR, we need to apply multilinear regression analysis on the data. The multilinear regression analysis is well described by Chatterjee and Hadi, [35]. As it is complex and time consuming to do the manual calculations, software such as Polymath and SPSS 11.5 can be used to analyze the data. For this research, SPSS 11.5 has been chosen to estimate EC50. The steps of estimating EC50 using SPSS 11.5 software is shown by
Fig 1.
III. RESULTS & DISCUSSION
LogEC50 values for the ionic liquids included in this study are shown in Table 2. According to group contribution methods, properties of a molecule can be assumed to be the summation of the contributions of its atom and/or fragments. In this work, the structure of ionic liquids has been based on three main contributions: anions (A), cations (C) and substitutions (S). Anions were Br-, Cl-, BF4-, PF6- and Tf2N-. Since BF4- and PF6- have similar contribution values, they are grouped. Cations were imidazolium, pyridinium, ammonium, pyrrolidinium, dimetylamino pyridinium and piperinido pryridinium (Imida, Pyrid, Ammon, Pyrrol, C5 and C6 respectively). Substitutions were R which is long n-alkane chain and R1 which is an additional short chain (methyl).
Figure 1: Steps of estimating EC50 using SPSS 11.5 software.
Table 2: Data for Ionic liquid toxicities in µmol/L (logEC50), dimensionless toxicity (Y*), and group contribution descriptor.
No
Compd
Log EC50 (µmol/L)
Y
Br-
Cl-
BF4-
PF6-
TF2N-
Imida
Pyrid
Ammon
Pyrrol
C5
C6
R
R1
1
bmimCl
1.85
0.05
0
1
0
0
0
1
0
0
0
0
0
4
1
2
bmimCl
1.57
0.11
0
1
0
0
0
1
0
0
0
0
0
4
1
3
bmimCl
1.93
0.03
0
1
0
0
0
1
0
0
0
0
0
4
1
4
bmimCl
1.93
0.03
0
1
0
0
0
1
0
0
0
0
0
4
1
5
bmimBr
1.84
0.05
1
0
0
0
0
1
0
0
0
0
0
4
1
6
bmimBr
1.78
0.07
1
0
0
0
0
1
0
0
0
0
0
4
1
7
bmimBr
1.56
0.11
1
0
0
0
0
1
0
0
0
0
0
4
1
8
bmimBr
1.56
0.11
1
0
0
0
0
1
0
0
0
0
0
4
1
9
bmimBF4
1.79
0.06
0
0
1
0
0
1
0
0
0
0
0
4
1
10
bmimBF4
1.67
0.09
0
0
1
0
0
1
0
0
0
0
0
4
1
11
bmimBF4
1.67
0.09
0
0
1
0
0
1
0
0
0
0
0
4
1
12
bmimPF6
1.95
0.03
0
0
0
1
0
1
0
0
0
0
0
4
1
13
bmimPF6
1.85
0.05
0
0
0
1
0
1
0
0
0
0
0
4
1
14
bmimPF6
1.85
0.05
0
0
0
1
0
1
0
0
0
0
0
4
1
15
bmimPF6
1.93
0.03
0
0
0
1
0
1
0
0
0
0
0
4
1
16
bmimTf2N
1.65
0.09
0
0
0
0
1
1
0
0
0
0
0
4
1
17
hmimCl
1.09
0.22
0
1
0
0
0
1
0
0
0
0
0
6
1
18
hmimBr
1.06
0.23
1
0
0
0
0
1
0
0
0
0
0
6
1
19
hmimBF4
1.13
0.21
0
0
1
0
0
1
0
0
0
0
0
6
1
20
omimCl
0.54
0.35
0
1
0
0
0
1
0
0
0
0
0
8
1
21
omimBr
0.54
0.35
1
0
0
0
0
1
0
0
0
0
0
8
1
22
omimBF4
0.66
0.32
0
0
1
0
0
1
0
0
0
0
0
8
1
23
C10mimBr
-0.31
0.54
1
0
0
0
0
1
0
0
0
0
0
10
1
24
C12mimCl
-1.82
0.88
0
1
0
0
0
1
0
0
0
0
0
12
1
25
C12mimbR
-0.82
0.66
1
0
0
0
0
1
0
0
0
0
0
12
1
26
C18mimCl
-2.00
0.92
0
1
0
0
0
1
0
0
0
0
0
16
1
27
C18mimCl
-2.34
1.00
0
1
0
0
0
1
0
0
0
0
0
18
1
28
bpyCl
2.07
0.00
0
1
0
0
0
0
1
0
0
0
0
4
0
29
bpyTf2N
0.62
0.33
0
0
0
0
1
0
1
0
0
0
0
4
0
30
bmpyr Tf2N
1.94
0.03
0
0
0
0
1
0
0
0
1
0
0
4
1
31
C12C3NBr
0.09
0.45
1
0
0
0
0
0
0
1
0
0
0
12
3
32
C14C3NBr
-0.38
0.56
1
0
0
0
0
0
0
1
0
0
0
14
3
33
C16C3NBr
-0.45
0.57
1
0
0
0
0
0
0
1
0
0
0
16
3
34
ompyBr
-2.6
0.73
1
0
0
0
0
0
1
0
0
0
0
8
1
35
hmpyBr
-2.41
0.70
1
0
0
0
0
0
1
0
0
0
0
6
1
36
bmpyBr
-1.24
0.52
1
0
0
0
0
0
1
0
0
0
0
4
1
37
omimBr
-4.33
1.00
1
0
0
0
0
1
0
0
0
0
0
8
1
38
hmimBr
-2.22
0.67
1
0
0
0
0
1
0
0
0
0
0
6
1
39
bmmpyBr
-1.01
0.48
1
0
0
0
0
0
1
0
0
0
0
4
2
40
hpyBr
-1.93
0.63
1
0
0
0
0
0
1
0
0
0
0
6
0
41
hmmimBr
-2.19
0.67
1
0
0
0
0
1
0
0
0
0
0
6
2
42
hPiPyBr
-3.66
0.90
1
0
0
0
0
0
1
0
0
0
1
6
0
43
HDMAPyBr
-3.28
0.84
1
0
0
0
0
0
1
0
0
1
0
6
0
44
HMDMAPyBr
-2.79
0.76
1
0
0
0
0
0
1
0
0
1
0
6
1
The data set has been fitted to the QSAR model by multilinear regression analysis using SPSS software. The contributions (ai, cj, and sk) of each group are shown in Table 3. According to Eq. (2) a good fitting was achieved, with n = 44, r2 = 0.934, radj2 = 0.910 and standard error of estimate = 0.149.
A comparison between the calculated data (from logEC50) for the dimensionless toxicity Y* of ionic liquids and the prediction based on the novel estimation method was performed. Fig. 2 shows the parity plot.
From the results, it can be assumed that imidazolium, pyridinium, ammonium, pyrrolidinium, dimethilamino pryridinium and piperidino pyridinium groups contribute about 20% decrease, 5% increase, 86% decrease, 31% decrease, 24% increase and 40% increase to the toxicity, respectively. In addition, contributions of anions and cations to the toxicity are negative, which means that the presence of anions leads to a decrease in the toxic effect of cation. The anion group Cl- and PF6- show a negligible difference between their contributions. For each carbon atom added to R chains, an increase in toxicity of about 7% is produced. An additional methyl group in the molecule decreases the toxicity by about 14%. Fig. 3 shows a schematic way to infer the least and most toxic ionic liquids.
Table 3: Group contribution to the dimensionless toxicity.
Group
Molecular descriptor
Comments
Contribution
95% Confidence Interval
Lower
Anion (A)
Br-
Influence of Br-anion. Value = 1 if it exists and 0 if not.
a
Cl-
Influence of Cl- anion. Value = 1 if it exists and 0 if not.
-0.171
-0.303
BF4-
Influence of BF4- anion. Value = 1 if it exists and 0 if not.
-0.152
-0.317
PF6-
Influence of PF6- anion. Value = 1 if it exists and 0 if not.
-0.176
-0.358
TF2N-
Influence of TF2N- anion. Value = 1 if it exists and 0 if not.
-0.075
-0.317
Cation (C)
Imida
Influence of Imidazolium cation. Value = 1 if it exists and 0 if not.
-0.197
-0.420
Pyrrid
Influence of Pyrridinium cation. Value = 1 if it exists and 0 if not.
0.053
-0.138
Ammon
Influence of Ammoinium cation. Value = 1 if it exists and 0 if not.
-0.862
-1.391
Pyrrol
Influence of Pyrrolidinium cation. Value = 1 if it exists and 0 if not.
-0.308
-0.758
C5
Influence of dimethylamino pyridinium. Value = 1 if it exists and 0 if not.
0.240
-0.010
C6
Influence of piperidino pyridinium. Value = 1 if it exists and 0 if not.
0.397
0.049
Substitution (S)
R
Influence of number of carbons in long chain (R:1 to 18)
0.075
0.059
R1
Influence of additional short chain in the molecule (R1:1 to 3)
0.113
-0.034
a By using SPPS, Bromide is excluded variable which means more data that have different anions (rather than Br) should be add.
FINAL GRAPH BR
Figure 2: Calculated data (from logEC50) for Y* using Eq. (1) and predicted Y* using SPSS software.
+
Anions
PF6 < Cl < BF4 < TF2N
Cations
Ammonium<Pyrrolidium<Imidazolium<Dimethylamino Pyridinium<Piperidino Pyridinium
Substitutions
R < R1
Figure 3: Ecotoxicity order of anions, cations and substitutions.
IV. CONCLUSION
A novel group contribution method, QSAR, has been developed for ionic liquids to estimate the EC50 for Daphnia magna. The method is based on the prediction of dimensionless ecotoxicity, Y, by the summation of the group contributions: cations, anions, and substitutions. The data range for log EC50 values is found in between 2.07 and -4.33, which is the experimental range of ecotoxicity covered by experimental data and the literature results. The results were well correlated (r2 = 0.934). The contributions to the QSAR allow one to estimate the influence each group (cations, anions, alkyl chains) has on the EC50. Further investigations are necessary to increase the number of data in the training set in order to reduce the confidence range of some group contributions (e.g., pyrrolidinium based ionic liquids). In addition, other cations and anions need to be studied to increase the application of the novel group contribution method.
VI. REFERENCES
J. Ranke, K. Moelter, F. Stock, U. Bottin-Weber, J. Poczobutt, J. Hoffmann, B. Ondruschka, J. Filser, B. Jastorff, "Biological effects of imidazolium ionic liquids with varying chain lengths in acute Vibrio fischeri and WST-1cell viability assays," Ecotoxicology and Environmental Safety, vol. 58, pp. 396-404, 2004.
N. Gathergood, M.T. Garcia, P.J. Scammells, "Biodegradable ionic liquids: part I. Concept, preliminary targets and evaluation," Green Chemistry, vol. 6, pp. 166-175, 2004.
D.J. Couling, R.J. Bernot, K.M. Docherty, J.K. Dixon, E.J. Maginn, "Assessing the factors responsible for ionic liquid toxicity to aquatic organisms via quantitative structure-property relationship modeling," Green Chemistry, vol. 8, pp. 82-90, 2006.
H. Luo, S. Dai, P.V. Bonnesen, A.C. Buchanan, "Separation of fission products based on ionic liquids: task-specific ionic liquids containing an aza-crown ether fragment," Journal of Alloys and Compounds, vol. 418, pp. 195-199, 2006.
P. Luis, I. Ortiz, R. Aldaco, A. Irabien, "A novel group contribution method in the development of a QSAR for predicting the toxicity (Vibrio fischeri EC50) of ionic liquids," Ecotoxicology and Environmental Safety, vol. 67, pp. 423-429, 2006.
K.M. Docherty, C.F. Kulpa, "Toxicity and antimicrobial activity of imidazolium and pyridinium ionic liquids," Green Chemistry, vol. 7, pp. 185-189, 2005.
A. Latala, P. Stepnowski, M. Nedzi, W. Mrozik, "Marine toxicity assessment of imidazolium ionic liquids: acute effects on the Baltic algae Oocystis submarina and Cyclotella meneghiniana," Aquatic Toxicology, vol. 73, pp. 91-98, 2005.
C.W. Cho, Y.C. Jeon, T.P.T. Pham, K. Vijayaraghavan, Y.S. Yun, "The ecotoxicity of ionic liquids and traditional organic solvents on microalga Selenastrum capricornutum," Ecotoxicology and Environmental Safety, vol. 71, pp.166-171, 2008.
R.J. Bernot, M.A. Brueseke, M.A. Evans-White, and G.A. Lamberti, "Acute and Chronic toxicity of Imidazolium Based Ionic Liquids on Daphnia magna," Environmental Toxicology and Chemistry, vol. 24(1), pp. 87-92, 2005.
R.J. Bernot, E.E. Kennedy, G.A. Lamberti, "Effects of ionic liquids on the survival, movement, and feeding behavior of the fresh water snail Physa acuta," Environmental Toxicology and Chemistry, vol. 24, pp. 1759-1765, 2005.
C. Pretti, C. Chiappe, D. Pieraccini, M. Gregori, F. Abramo, G. Monni, L. Intorre, "Acute toxicity of ionic liquids to the zebra fish (Danio rerio)," Green Chemistry, vol. 8, pp. 238-240, 2006.
C. Pretti, C. Chiappe, I. Baldetti, S. Brunini, G. Monni, L. Intorre, "Acute toxicity of ionic liquids for three fresh water organisms: Pseudokirchneriella subcapitata, Daphnia magna and Danio rerio," Ecotoxicology and Environmental Safety, 72, 1170-1176, 2009.
J.S. Wilkes, "Properties of ionic liquid solvents for catalysis," Journal of Molecular Catalysis A: Chemical, vol. 214, pp. 11-17, 2004.
T.W. Schultz, M.T.D. Cronin, T.I. Netzeva, "The present status of QSAR in toxicology," Journal of Molecular Structure: THEOCHEM, vol. 622, pp. 23-38, 2003.
D.T. Allen, D.R. Shonnard, "Green Engineering: Environmentally Conscious Design of Chemical Processes," Prentice-Hall, Upper Saddle River, NJ ISBN 0-13-061908-6, 2002.
A. Beteringhe, A.T. Balaban, "QSAR for toxicities of polychlorodibenzofurans, polychlorodibenzo-1,4-dioxins and polychlorobiphenyls," Arkivoc, pp. 163-182, 2004.
O. Vajragupta, P. Boonchoong, Y. Wongkrajang, "Comparative quantitative structure-activity study of radical scavengers," Bioorganic & Medicinal Chemistry, vol. 8, pp. 2617-2628, 2000.
ECOSAR (Ecological Structure Activity Relationships) Program, 2000, U.S. Environmental Protection Agency. Available at /http://www.epa.gov/oppt/newchems/tools/21ecosar.htmS.
X. Liu, B. Wang, Z. Huang, S. Han, L. Wang, "Acute toxicity and quantitative structure-activity relationships of a-branched phenylsulfonyl acetates to Daphnia magna," Chemosphere, vol. 50, pp. 403-408, 2003.
V.K. Agrawal, P.V. Khadikar, "QSAR study on narcotic mechanism of action and toxicity: A molecular connectivity approach to Vibrio fischeri toxicity testing," Bioorganic & Medicinal Chemistry, vol. 10, pp. 3517-3522, 2002.
R.L. Yu, G.R. Hu, Y.H. Zhao, "Comparative study of four QSAR models of aromatic compounds to aquatic organisms," Journal of Environmental Sciences, vol. 14, pp. 552-557, 2002.
D.J. Couling, D.M. Eike, J.F. Brennecke, E.J. Maginn, "Quantitative structure-property modelling of ionic liquid aquatic toxicity: Insight into structural features that determine toxicity for Daphnia magna and Vibrio fischeri," Book of Abstracts, First International Congress on Ionic Liquids (COIL), Salzburg, 2005.
M.P. Vega, R.A. Pizarro, "Oxidative stress and defence mechanisms of the fresh water cladoceran Daphnia longispina exposed to UV radiation," Journal of Photochemistry and Photobiology B: Biology, vol. 54, pp. 121-125, 2000.
A.S. Wells, and V.T. Coombe, "On the Freshwater Ecotoxicity and Biodegradation Properties of Some Common Ionic Liquids," Organic Process Research & Development, vol. 10, pp. 794−798, 2006.
M.T. Garcia, N. Gathergood, and P.J. Scammells, "Biodegradable ionic liquids Part II. Effect of the anion and toxicology," Green Chemistry, vol. 7, pp. 9-14, 2005.
M. Yu, S-H. Wang, Y-R. Luo, Y-W. Han, X-Y. Li, B-J. Zhang, J-J. Wang, "Effects of the 1-alkyl-3-methylimidazolium bromide ionic liquids on the antioxidant defense system of Daphnia magna," Ecotoxicology and Environmental Safety, vol. 72, pp. 1798-1804, 2009.
P. Nockemann, B. Thijs, K. Driesen, C.R. Janssen, K.V. Hecke, L.V. Meervelt, S. Kossmann, B. Kirchner, and K. Binnemans, "Choline Saccharinate and Choline Acesulfamate: Ionic Liquids with Low Toxicities," Journal of Physical Chemistry B, vol. 111, pp. 5254-5263, 2007.
M. Kaniewska-Prus, "The effect of ammonia, chlorine, and chloramines toxicity on the mortality of Daphnia magna Straus," Polish Archives of Hydrobiology, vol. 29, pp. 607-624, 1982.
D.I. Mount, T.J. Norberg, "A seven-day life cycle cladoceran toxicity test," Environmental Toxicology and Chemistry, vol. 3, pp. 425-434, 1984.
U.M. Cowgill, D.P. Milazzo, "The sensitivity of Ceriodaphnia dubia and Daphnia magna to seven chemicals utilizing the three-brood test," Archives of Environmental Contamination and Toxicology, vol. 20, pp. 211-217, 1991.
G.A. LeBlanc, "Acute toxicity of priority pollutants to water flea (Daphnia magna)," Bulletin of Environmental Contamination and Toxicology, vol. 24, pp. 684-691, 1980.
J.H. Canton, D.M.M. Adema, "Reproducibility of short-term and reproduction toxicity experiments with Daphnia magna and comparison of the sensitivity of Daphnia magna with Daphnia pulex and Daphnia cucullata in short-term experiments," Hydrobiologia, vol. 59, pp. 135-140, 1978.
Z. Tong, Z. Huailin, J. Hongjun, "Chronic toxicity of acrylonitrile and acetonitrile to Daphnia magna in 14-d and 21-d toxicity tests," Bulletin of Environmental Contamination and Toxicology, vol. 57, pp. 655-659, 1996.
L. Guilhermino, T. Diamantino, M.C. Silva, A.M.V.M. Soares, "Acute toxicity test with Daphnia magna: An alternative to mammals in the prescreening of chemical toxicity?," Ecotoxicology and Environmental Safety, vol. 46, pp. 357-362, 2000.
S. Chatterjee, A.S. Hadi, Regression analysis by example, 4th edition, Willey Interscience, 2006.
