Monitoring Method of Hydroxyl Functionalized Imidazolium Ionic Liquids in Complex Environment Water Samples

With the increasing emphasis on the toxicity of ionic liquids, it is imperative to develop detection methods for ionic liquids in complex environments. A new method for the analysis of hydroxyl functionalized imidazolium ionic liquids in complex environmental water samples by using ion chromatography and solid-phase extraction was developed. Under the selected chromatographic conditions, the complete separation of the two analytes was achieved in 14.0 min. The linear and repetitive data of the approach for the analysis of the two cations by ion chromatography meet the requirements of quantitative analysis. The extraction of [HEMIm] + and [HPMIm] + in water samples by ion-exchange solid-phase extraction and reversed-phase solid-phase extraction was compared. The results show that the enrichment and purication of two target cations can be better achieved using ion-exchange solid-phase extraction approach, and the enrichment multiple was 50 times. The method was used to determine the target cations in two river waters. The recovery of standard addition was between 82.5% and 96.0%, the detection limit was less than 0.01 mg/L, and inter-assay RSD was less than 2.5%. The method is simple, accurate and reliable. It is suitable for the determination of hydroxyl functionalized imidazolium ionic liquid cations in environmental water samples.


Introduction
Ionic liquids (ILs) are special liquid molten salts at room temperature. As new green solvent, ILs have many unique physicochemical properties, such as wide range of liquid temperature, low vapor pressure, strong solubility and recyclable use. At present, ILs have been widely used in organic synthesis,  Hossain et al. 2013). In addition, ILs are stable, non-volatile and soluble in water, which may lead to environmental pollution. Therefore, the determination of trace IL cations in environmental water samples is very important, which can provide reference for the environmental risk assessment of ILs.
Because the content of IL cations in the environment is very low, it can't reach the detection limit of the chromatographic instrument. Therefore, the samples should not be directly detected by chromatograph.
In addition, environmental sample systems are complex and may interfere with the analysis of target compounds, so sample pretreatment is needed for the puri cation and enrichment of the targets. Solidphase extraction (SPE) is a kind of sample pretreatment technology developed in recent years. It can reduce matrix interference and improve the detection sensitivity of analytes and its operation is simple, time-saving and labor-saving. SPE has been widely used in medicine, food, environment, commodity inspection, chemical industry and other elds (Capriotti et  hydroxyl functionalized imidazolium IL cations in reversed-phase ion-pair chromatography and reversedphase chromatography were investigated. However, how to enrich and purify hydroxyl functionalized imidazolium ILs in environmental water samples has not been studied. In addition, whether other chromatographic methods, such as ion chromatography, can analyze these ionic liquids still need to be explored. On this basis, we will continue to explore a new method for the trace analysis of hydroxyl functionalized imidazolium IL cations in environmental water samples.
The purpose of this work is to develop an analytical method for the determination of hydroxyl functionalized imidazolium ILs in river water samples by combining ion chromatography with SPE. Ion chromatography separation and SPE enrichment of the functionalized ILs in water samples were studied, which provided a new reference method for separation and detection of trace hydroxyl functionalized imidazolium ILs in environmental water samples.

Preparation of Solution
The standard stock solutions of hydroxyl functionalized imidazolium IL cations [HEMIm] + and [HPMIm] + with mass concentration of 1000.0 mg/L were prepared by using ultrapure water and stored in refrigerator. When in use, the stock solutions were diluted to the standard solutions of the required concentration for the experiment, and then ltered through 0.22 μm lter membrane before chromatographic injection.
According to the experimental requirement, the mobile phase with a certain concentration was prepared by ultrapure water, ltered by 0.22 μm lter membrane, degassed by vacuum for 15 min and then used.

Instruments
A model LC-20A chromatographic system (Shimadzu, Japan) was used, which was equipped with a model SPD-20A ultraviolet detector, a LC-20AT mobile phase infusion pump, a SIL-20A automatic sampler, a CTO-20A column temperature box and Labsolutions workstation. A model ASE-24 SPE instrument (Tianjin OTSEENS Instrument Company Limited, China) equipped with a model AP-9950 oil-free vacuum pump was used for the SPE of water samples. A model Millipore Milli-Q water puri cation system (Millipore, USA) was used to produce the ultrapure water with a resistivity of 18. 2

Analysis conditions
The optimum ion chromatographic conditions were using the Shodex IC YK-421 cation-exchange column for the separation of the analytes, 2.0 mM MSA aqueous solution/ACN (87/13, v/v) as mobile phase, ultraviolet detection wavelength at 210 nm, column temperature at 40 °C, ow rate 1.0 mL/min, injection volume 20 μL.
The optimum SPE conditions were as follows. The UF-SCX SPE column was used for the enrichment and puri cation of the targets in water samples. On the SPE column, 400 mL water samples were enriched at a ow rate of 2.0 mL/min. The rinsing solvent was 3 mL 5% methanol aqueous solution. The eluent was 10 mL 0.17 M H 3 PO 4 -NaH 2 PO 4 buffer solution-55% methanol with the ow rate of 2.0 mL/min. The results were that the chromatographic peaks of the two analytes were still completely overlapped.

Results And
These investigations show that the separation of the two cations cannot be effectively improved on the Baseline WCX column. Therefore, the Shodex IC YK-421 column was selected to continue the investigation.

Effect and selection of acids in mobile phase
The Shodex IC YK-421 column was used to determine  (Table 1), compared with methanol, when ACN was used, the retention times of [HEMIm] + and [HPMIm] + were shorter, the noise values and detection limits were lower. This may be due to the lower viscosity and lower mass transfer resistance of ACN than that of methanol, which is bene cial to improve the column e ciency.
Considering the retention times and detection limits of [HEMIm] + and [HPMIm] + , ACN was selected. In ultraviolet detection, in order to enhance the detection signal, the maximum absorption wavelength of analyte is usually selected as the detection wavelength. The maximum absorption wavelength of imidazolium cations is about 210 nm (Katoh 2007 [HPMIm] + was further investigated under the previously selected chromatographic conditions. The effect of temperature is usually described by the Van't Hoff curve equation (Yu and Mou 2006), where k is the retention factor, T is the thermodynamic temperature (K), ΔH and ΔS are the change of free enthalpy and the free energy, respectively, R is the molar gas constant, and φ is the ratio of the two phases. The linear regression equation between the lnk of [HEMIm] + and the column temperature T is , and the correlation coe cient is 0.9961. The linear regression equation between the lnk of [HPMIm] + and the column temperature T is , and the correlation coe cient is 0.9980. It can be seen that both correlation coe cients are good and the slopes are positive, which indicates that the retention was an exothermic process, and retention time of the two cations was shortened with the increase of temperature. Since the retention time of the analytes at 40 ℃ was moderate and the baseline is stable, 40 ℃ was selected.
3.1.6. Veri cation of interference with common imidazolium IL cations Common imidazolium cations and hydroxyl functionalized imidazolium cations are structurally similar. In order to verify whether imidazolium cations interfere with the determination of hydroxyl functionalized imidazolium cations, these cations were analyzed and compared under the previously selected chromatographic conditions. (Fig. 2) shows that the retention times of the imidazolium cations were longer and the retention times of hydroxyl functionalized imidazolium cations were shorter, and these cations can be separated from the baseline. Therefore, common imidazolium cations do not interfere with the determination of the target cations.  (Table 2). The results show that the detection limits are less than 0.20 mg/L and the quantitative limits are less than 0.65 mg/L. Linearity and reproducibility meet the requirements of quantitative analysis.

Reversed-phase SPE of hydroxyl functionalized imidazolium IL cations
The selection of adsorbents in SPE is mainly based on the properties of target compounds. The more similar the polarity of target compound is to that of adsorbent, the better the retention is. For the effective extraction of analytes in water samples, two extraction modes were studied in this work, namely reversedphase SPE mode and ion-exchange SPE mode. The reversed-phase SPE of hydroxyl functionalized imidazolium IL cations are as follows.

Selection of sample volume for the reversed-phase SPE
The UF-C18 reversed-phase SPE column was used for the enrichment and puri cation of [HEMIm] + and [HPMIm] + . Firstly, 5 mL methanol was used for activation of the SPE column. The purpose of activation was to remove impurities in the column and create a certain solvent environment. Secondly, 5 mL ultrapure water was used for equilibrium. Then, the penetration volume was investigated. The ow rate was 2 mL/min and the concentration of analytes was 1.0 mg/L. In this investigation, a small amount of [HEMIm] + and [HPMIm] + were detected in the e uent after 1.0 mL water sample was uploaded; while [HEMIm] + and [HPMIm] + were not detected in the e uent after 0.5 mM sodium decanesulfonate was added in the water sample. The possible reason is that the retention of analytes in the reversed-phase SPE column is mainly affected by the hydrophobic and hydrophilic force, which increases with the increase of hydrophobicity and decreases with the increase of hydrophilicity.
[HEMIm] + and [HPMIm] + are hydrophilic compounds, which are di cult to retain on the reversed-phase SPE column, so the reversedphase ion-pair SPE method is chosen. It is generally understood that in the ion-pair model, the analytes and the ion-pair reagents form neutral ion-pairs, which are then retained on the reversed-phase SPE column. Ion-pair reagents with longer carbon chain have higher hydrophobicity, which enhance the retention of measured ions on reversed-phase SPE column (Stepnowski and Nichthauser 2008). Therefore, the retention of analyte in the reversed-phase SPE column can be enhanced by adding ion-pair reagent, and the longer the carbon chain of ion-pair reagent is, the stronger the retention of analyte is. Sodium decanesulfonate with a concentration of 0.5 mM was selected as the ion-pair reagent in this experiment.
After selecting the ion-pair SPE mode, then, the sample volume needs to be chosen. In SPE, the penetration volume is equivalent to the maximum allowable sample volume of the target compound through the SPE column without obvious loss. The larger the penetration volume of SPE column is, the larger the sample volume can be treated. In order to investigate the penetration volume of SPE and select the optimum sampling volume, the effects of sample volumes of 5, 10, 15, 20, 25, 30, 35, 40, 45, 50 and 55 mL on SPE of [HEMIm] + and [HPMIm] + were investigated. The results showed that [HEMIm] + could be detected in e uent when the sample volume was 55 mL. This shows that the sample loading has exceeded the column capacity at this time. Therefore, a sample volume of 50 mL was selected.

Selection of rinsing solvent for the reversed-phase SPE
In order to reduce the interference of impurities on target compounds, the impurities should be removed as far as possible without affecting the recovery of target compounds (Fan et al. 2018a). The effects of rinsing solvents on SPE of [HEMIm] + and [HPMIm] + were investigated by using 3 mL water, 3 mL 5% methanol and 3 mL 5% ACN, respectively. The results showed that the target analytes were eluted when the rinsing solvent was 5% methanol or 5% ACN. This affected the enrichment of the targets by the SPE column. When the rinsing solvent was water, no target cations were detected in the e uent. Therefore, water is chosen as rinsing solvent.

Selection of eluent and elution volume for the reversed-phase SPE
In SPE, the eluent should be selected according to the relevant parameters of the target compounds, such as the solubility and structure of the target compounds. Solubility is one of the important factors for selecting eluent. In reversed-phase SPE, the organic solvent with high solubility to the target substance is generally preferred as eluent. According to the structure and functional group of the target compounds and the principle of similar miscibility, the eluent with similar polarity is selected. Owing to the strong polarity of [HEMIm] + and [HPMIm] + , and the greater polarity of methanol than ACN, methanol was chosen as the organic phase in eluent. The effects of 70%, 80% and 90% methanol aqueous solutions as eluent on SPE of [HEMIm] + and [HPMIm] + were compared. The results show that the recovery was the highest when the volume fraction of methanol was 90%. The reason may be that the elution capacity of eluent increases with the increase of methanol ratio in the elution solutions. Therefore, the volume fraction of methanol in the elution solution is 90%.
Next, the volume of eluent was selected. More eluents will elute the matrix together and affect the analysis of the targets. So, the eluent should be used as little as possible. The effects of elution volumes of 1 mL and 2 mL on SPE of [HEMIm] + and [HPMIm] + were investigated. The results showed that the recoveries were higher when the elution volume was 2 mL. Therefore, the elution volume 2 mL was selected.

Ion-exchange SPE of hydroxyl functionalized imidazolium IL cations
3.3.1. Selection of sample volume for the ion-exchange SPE The UF-SCX ion-exchange SPE column was used for enrichment and puri cation. Firstly, 5 mL methanol was used for activation, then ultrapure water of the same volume was used for equilibrium. Secondly, the penetration volume was investigated by sampling. The ow rate of sampling was 2 mL/min, and the concentration of analytes was 0.05 mg/L. time. When choosing the sample volume, it is also necessary to consider the adsorption of interfering substances in the water samples on the extraction column. In order to achieve good sampling and elution effect, the sample volume 400 mL was selected.

Selection of rinsing solvent for the ion-exchange SPE
The effects of rinsing solvents on SPE of [HEMIm] + and [HPMIm] + were investigated using 5 mL water, 5 mL 5% methanol and 5 mL 5% ACN, respectively. The results showed that no target cations were detected in the e uent of the three rinsing solvents, which indicated that any of the three rinsing solvents could be used. Considering that the interference impurities in the real samples may be more complex, in order to remove the interference impurities as much as possible, the washing effect of the mixed solution of water and organic solvent will be better. At the same time, compared with ACN, methanol has stronger polarity to wash impurities more effectively. Therefore, 5% methanol is chosen as the rinsing solvent.

Selection of eluent and elution volume for the ion-exchange SPE
In ion-exchange SPE, the eluent must have the ability to resolve the ion-exchange force and non-polar force, and the target should have good solubility in eluent. Because of the low pK a of the benzenesulfonic acid functional group in the UF-SCX SPE column, it is impossible to neutralize the benzenesulfonic acid by adjusting the pH to release the target compounds.  (Table 3) shows that with the increase of methanol concentration, the recovery increases and the elution effect is improved. The possible reason is that methanol has a high enough polarity to weaken the force between the target and the SPE column. When methanol concentration was 60%, partial crystallization occurred in buffer solution. The reason is that phosphate is insoluble in methanol. When a large amount of methanol was added, the buffer property of the solution was affected and phosphate precipitates. Therefore, 0.17 M H 3 PO 4 -NaH 2 PO 4 -55% methanol was selected. When the elution volume was 10 mL, the enrichment factor was 40 times, and the recoveries of [HEMIm] + and [HPMIm] + were 91.3% and 96.1% (), respectively. Although the elution volume of 8 mL had a higher enrichment factor, the recovery was relatively low. In order to obtain a higher recovery, the elution volume of 10 mL was selected.

Analysis of environmental water samples
Water samples from Songhua River and Hulan River were collected for the analysis. At room temperature, the two water samples were stationary for 48 h, centrifuged and ltered by 0.22 μm lter membrane. River water samples were treated by two SPE methods, and then ltered by 0.22 μm lter membrane for ion chromatography analysis.
When the reversed-phase UF-C18 SPE column was used, it was found that the recoveries of [HEMIm] + and [HPMIm] + in both water samples were lower than 27.5%. The possible reason is that the matrix in environmental water samples is usually complex. In SPE, the adsorbent of reversed-phase SPE not only adsorbed the target compounds, but also adsorbed some impurities. Most of the target compounds can't be adsorbed during the sample loading process, so the recoveries were very low. This indicates that the reversed-phase UF-C18 SPE column can't be applied to the analysis of the environmental water samples.
The UF-SCX ion-exchange SPE column was selected and the recoveries of the standard addition method were tested. The chromatograms of sample analysis are shown in (Fig. 5) and analysis results are shown in ( the target cation is adsorbed. Some cation matrix with weak adsorption capacity will be washed by rinsing solvent, which will not affect the determination of targets. This method is accurate, reliable and simple. Compared with previous studies, this method has lower detection limit, higher anti-interference performance and the experiment without ion-pair reagent is more convenient. It is suitable for monitoring the hydroxyl functionalized imidazolium IL cations in environmental water, and provides a reference for environmental risk assessment of hydroxyl functionalized ILs.