Application of Molecularly Imprinted Polymers (MIP) and Flowing Atmospheric-Pressure Afterglow Mass Spectrometry (FAPA-MS) to Analysis of Nonsteroidal Anti-Inﬂammatory Drugs (NSAIDs)

: In recent years, the rapid development of the pharmaceutical industry and the extensive and illicit use of painkillers have led to increased levels of nonsteroidal anti-inﬂammatory drugs (NSAIDs) in the environment. In view of the signiﬁcant impact of NSAIDs on living organisms, including humans, their presence in the environment needs to be continuously monitored at trace levels. For this purpose, a combination of molecularly imprinted solid-phase extraction (MISPE) and HPLC-MS analysis is commonly used. MISPE has been utilized in direct, fast, and ecological analysis of drugs using a ﬂowing atmospheric-pressure afterglow ion source for mass spectrometry (FAPA-MS). The new method was applied herein in the determination of naproxen, diclofenac, and ibuprofen. The linear dependence of the intensity of analytical signals on the amount of drugs is in the range of 0.2 µ g to 1 g and the method detection limit (MDL) for all drugs is 0.2 µ g in environmental samples. The new method also decreased the number of analytical stages, the time and cost of analysis, and the organic solvent consumption, besides being environmentally friendly. ﬁrst of measurements, observed a low-intensity signal from naproxen (N).


Introduction
The recent increase in the human population and the development of industry have been the main contributors to the growing number of pollutants in the natural environment. The introduction of harmful substances into the air, water, and soil, in addition to having a negative impact on ecosystems and climate, is a direct threat to human health. Furthermore, these compounds can be easily transported between water reservoirs, like from wastewater to river water, to dam water, consequently reaching drinking water and becoming a serious threat to human health. These new emerging pollutants are considered very dangerous, even in trace amounts, due to their serious ecological side effects. Therefore, water pollution by nonsteroidal anti-inflammatory drugs (NSAIDs) has become a subject of global interest and of great interest to major international organizations such as the World Health Organization, United States State Agency for Environmental Protection, and the European Commission [1][2][3]. Among several emerging pollutants, NSAIDs are the most common compounds detected in environmental compartments due to their large consumption throughout the world [4]. Pain is a serious public health problem worldwide. About 27% of the adult population suffers from chronic pain [5,6]. If left untreated, it may be associated with the risk of harmful health effects (e.g., delayed wound healing, immunity disorders, or cardiovascular problems). In addition, that allow MS analysis at atmospheric pressure. They provide rapid, direct, and high-throughput analyses with no or only minimal sample pretreatment. Moreover, substances can be analyzed directly from surfaces or other matrixes. FAPA-MS techniques involve the generation of a direct current or radiofrequency electrical discharge between a pair of electrodes in contact with a flowing inert gas, creating a stream of ionized molecules, radicals, excited state neutral atoms, and electrons. The plasma species are directed towards the sample, resulting in desorption and ionization of the analyte. Ambient plasma MS techniques have many advantages due to their simple instrumentation, rugged construction of the measuring system, no need for solvents, and generation of singly charged analyte species that are easier to identify than multiply charged ions and various adducts produced by the spray-based techniques [30][31][32][33][34][35].
We propose the use of commercially available AFFINILUTE TM MIP-NSAIDs columns to pre-isolate and preconcentrate NSAIDs from solutions, followed by direct analysis using the FAPA ionization source. The obtained results are presented and compared with results obtained via the standard way of analysis by the HPLC-MS method.

Chemicals
All reagents used were commercial products. Ibuprofen, naproxen, diclofenac, and AFFINILUTE TM MIP-NSAIDs columns were obtained from Merck (Germany). Ammonium formate, formic acid (FA), and all organic solvents were also purchased from Merck (Germany).

High-Performance Liquid Chromatography-Mass Spectrometry (HPLC-MS)
HPLC-MS spectra were recorded on a quadrupole mass spectrometer with electrospray ionization (ESI). As a standard, an Impact HD mass spectrometer was connected to an ultra-HPLC (UHPLC) liquid chromatograph (UltiMate 3000) Thermo Scientific/Dionex with a fast (200 Hz) diode detector. The mass spectrometer was operated in the mass range 177-400 m/z. The capillary voltage was set at −4.5 kV and the endplate offset at −500 V. The nebulizer pressure was 1.5 bar, and the dry gas flow was 9 Lmin −1 . In the HPLC analyses, a Kinetex 2.6 u C18 column (100 × 2.10 mm) was used. The column temperature was 35 • C, the injection volume was 50 µL, and the mobile phase flow was 0.3 mLmin −1 . A concentration gradient was used. The change in the composition of the mobile phase is shown in Table 1.

FAPA-MS
FAPA-MS spectra were recorded using an AmaZon SL ion trap (Bruker, Bremen, Germany) equipped with a FAPA ambient plasma source (ERTEC, Wroclaw, Poland) ( Figure 1). The apparatus was operated in the so-called "enhanced resolution mode" (mass range: 50-2200 m/z, scanning rate: 8100 m/z per second). The capillary voltage was set at −4.5 kV and the endplate offset at −500 V. The source temperature was 80 • C and the desolvation temperature was 250 • C. Helium was used as the cone gas and desolvating gas (nitrogen) at flow rates of 50 Lh −1 and 800 Lh −1 , respectively.
FAPA was used to generate plasma, allowing the ionization of analyte particles released thermally from a heated crucible, with temperature regulation from 50 to 500 °C at a temperature increase rate of 3 °C•s −1 . The desorption temperature of the analytes was close to the melting points of the analytes and varied between 50 and 300 ± 10 °C. The mini crucible allowing the temperature-controlled desorption was placed ca. 10 mm below the ion stream. The distance between the inlet of the mass spectrometer and the FAPA ion source should be about 10 cm.

HPLC-MS Method
Three different series of aqueous solution samples were prepared, containing ibuprofen, naproxen, and diclofenac in the concentration range from 0.02 μgL −1 to 3 mgL −1 . All samples were subjected to HPLC-MS analysis. The mass spectra as the intensity of signals versus mass-to-charge ratio were recorded. The dependence of signal intensity [M+H] + on the concentrations of individual analytes was plotted. From these data, the range of measurement linearity and limit of detection (LOD) were determined.
In the next stage, all high-dilution samples with an analyte amount below the LOD were subjected to the following procedure: Samples were analyzed according to the commercial procedure using AFFINILUTE TM MIP-NSAIDs columns and HPLC-MS. For this purpose, an AFFINILUTE TM MIP-NSAIDs column was conditioned using 1 mL acetonitrile, 1 mL methanol, and 1 mL 10 mM ammonium formate (pH = 3). Next, 5 mL of the prepared NSAIDs sample was loaded on the column. In the next step, the column was washed with 1 mL water and 1 mL acetonitrile/water at 40:60 (v/v). Then, the NSAIDs were eluted by 2 × 1 mL 1% acetic acid diluted in 20:80 (v/v) acetone/methanol. Finally, the eluate was evaporated to dryness, and 150 μL of the mobile phase of HPLC was added. The HPLC-MS spectra were recorded. This procedure is represented in Figure 2. The concentration of the NSAIDs in solution was determined. The process was repeated three times. FAPA was used to generate plasma, allowing the ionization of analyte particles released thermally from a heated crucible, with temperature regulation from 50 to 500 • C at a temperature increase rate of 3 • C·s −1 . The desorption temperature of the analytes was close to the melting points of the analytes and varied between 50 and 300 ± 10 • C. The mini crucible allowing the temperature-controlled desorption was placed ca. 10 mm below the ion stream. The distance between the inlet of the mass spectrometer and the FAPA ion source should be about 10 cm.

HPLC-MS Method
Three different series of aqueous solution samples were prepared, containing ibuprofen, naproxen, and diclofenac in the concentration range from 0.02 µgL −1 to 3 mgL −1 . All samples were subjected to HPLC-MS analysis. The mass spectra as the intensity of signals versus mass-to-charge ratio were recorded. The dependence of signal intensity [M+H] + on the concentrations of individual analytes was plotted. From these data, the range of measurement linearity and limit of detection (LOD) were determined.
In the next stage, all high-dilution samples with an analyte amount below the LOD were subjected to the following procedure: Samples were analyzed according to the commercial procedure using AFFINILUTE TM MIP-NSAIDs columns and HPLC-MS. For this purpose, an AFFINILUTE TM MIP-NSAIDs column was conditioned using 1 mL acetonitrile, 1 mL methanol, and 1 mL 10 mM ammonium formate (pH = 3). Next, 5 mL of the prepared NSAIDs sample was loaded on the column. In the next step, the column was washed with 1 mL water and 1 mL acetonitrile/water at 40:60 (v/v). Then, the NSAIDs were eluted by 2 × 1 mL 1% acetic acid diluted in 20:80 (v/v) acetone/methanol. Finally, the eluate was evaporated to dryness, and 150 µL of the mobile phase of HPLC was added. The HPLC-MS spectra were recorded. This procedure is represented in Figure 2. The concentration of the NSAIDs in solution was determined. The process was repeated three times.

MIP-NSAIDs columns
Apply a maximum 5 mL of sample

FAPA-MS Method
In the FAPA-MS method, a sample ionized at atmospheric pressure can be introduced into a mini crucible in solution or solid form. To determine the range of measurement linearity and detection limits of the method, three aqueous solutions of ibuprofen, naproxen, and diclofenac at a concentration of 3 mgL −1 were prepared. Then, different volumes (10-1000 µL) of solution were added to the mini crucible and the mass spectra were recorded. The relationship between the signal intensity [M+H] + and the amount of analyte was plotted.
Analysis of NSAIDs by the FAPA-MS method with an AFFINILUTE TM MIP-NSAIDs column was realized as follows: After conditioning the AFFINILUTE TM MIP-NSAIDs column using 1 mL acetonitrile, 1 mL methanol, and 1 mL 10 mM ammonium formate (pH = 3), a specific amount of analyte (maximum of 25 mg NSAIDs in the total volume of sample solution) was loaded directly into the column. Twenty-five milligrams of NSAIDs was the column capacity. The column was washed with 1 mL of 50:50 (v/v) acetonitrile/water and dried. In the next step, MIP (15 mg) from the column was transferred to a mini-cup of the spectrometer, and a measurement was made by recording the mass spectrum of the sample. The analytical procedure is presented in Figure 3. The concentration of NSAIDs was determined. The process was repeated three times.

MIP-NSAIDs columns
Apply a maximum of 25 mg NSAIDs in sample In the FAPA-MS method, a sample ionized at atmospheric pressure can be introduced into a mini crucible in solution or solid form. To determine the range of measurement linearity and detection limits of the method, three aqueous solutions of ibuprofen, naproxen, and diclofenac at a concentration of 3 mgL −1 were prepared. Then, different volumes (10-1000 μL) of solution were added to the mini crucible and the mass spectra were recorded. The relationship between the signal intensity [M+H] + and the amount of analyte was plotted.
Analysis of NSAIDs by the FAPA-MS method with an AFFINILUTE TM MIP-NSAIDs column was realized as follows: After conditioning the AFFINILUTE TM MIP-NSAIDs column using 1 mL acetonitrile, 1 mL methanol, and 1 mL 10 mM ammonium formate (pH = 3), a specific amount of analyte (maximum of 25 mg NSAIDs in the total volume of sample solution) was loaded directly into the column. Twenty-five milligrams of NSAIDs was the column capacity. The column was washed with 1 mL of 50:50 (v/v) acetonitrile/water and dried. In the next step, MIP (15 mg) from the column was transferred to a mini-cup of the spectrometer, and a measurement was made by recording the mass spectrum of the sample. The analytical procedure is presented in Figure 3. The concentration of NSAIDs was determined. The process was repeated three times.

Results and Discussion
The HPLC-MS method was used to analyze (Figures S1-S3) standard NSAIDs solutions consisting of naproxen (0.229 mgL −1 ), diclofenac (0.295 mgL −1 ), and ibuprofen (2.041 mgL −1 ). The retention time for analytes in HPLC was determined, and using the mass spectrometer, the value of m/z signals of analytical significance was determined. For each of the compounds, the analytical signal from the proton attachment was selected. For naproxen, m/z 231 [M+H] + ; for diclofenac, m/z 296 [M+H] + ; and for ibuprofen, m/z 207 [M+H] + (Figure 4).  (Figure 4). The signals in the spectrometry mass apparatus were recorded in two modes: total ion chromatogram (TIC) and extracted ion chromatogram (EIC). Next, using five different concentrations of analytes, the range of linearity of intensity signals [M+H] + versus the concentration of drugs and the method detection limit (MDL) were determined ( Table 2). The MDL is defined as the minimum measured concentration of a substance that can be reported with 99% confidence that the measured concentration is distinguishable from method blank results. The concentration of a substance that can be reported with 99% confidence indicates that the measured concentration is distinguishable from blank samples.  The signals in the spectrometry mass apparatus were recorded in two modes: total ion chromatogram (TIC) and extracted ion chromatogram (EIC). Next, using five different concentrations of analytes, the range of linearity of intensity signals [M+H] + versus the concentration of drugs and the method detection limit (MDL) were determined ( Table 2). The MDL is defined as the minimum measured concentration of a substance that can be reported with 99% confidence that the measured concentration is distinguishable from method blank results. The concentration of a substance that can be reported with 99% confidence indicates that the measured concentration is distinguishable from blank samples. is necessary to pre-concentrate the samples before further analysis. The analytes were concentrated using AFFINILUTE TM MIP-NSAIDs columns. The samples were prepared according to the previously described analytical procedure; in these experiments, we used solutions in the concentration range from 0.02 µgL −1 to 3.0 mgL −1 . The HPLC-MS spectra of the eluent were recorded. Table 3 presents the obtained results regarding analytical parameters. The obtained results indicate that it is a well-functioning and useful analytic tool. However, sample preparation consists of many stages that can accumulate various errors. These errors can significantly affect the result, especially in the case of trace analysis. In addition, sample preparation is time-and cost-consuming, which are crucial factors for large-scale analysis. An additional threat to the environment is the amount of waste generated during sample preparation. Particular attention should be paid to the amount of organic solvents used. In response to this problem, other analytical options using the known analytical techniques and commercially available means were considered.
In the first step of direct analysis by FAPA-MS, different amounts of analytes were placed on a mini crucible and their mass spectra were recorded. For ibuprofen, two signals corresponding to        Figure S4). The range of measurement linearity and method detection limits for all compounds were determined. The linear dependence of the intensity of analytical signals on the amount of drugs is in the range of 0.2 µg to 1 g and the MDL for all drugs is 0.2 µg. In the next step, the scope of applicability of the FAPA-MS analytical method for the determination of NSAIDs was defined for samples containing a mix of drugs. During the measurement, the heating was started at 50 • C with a gradual increase to 300 • C. In Figure 8, the first signal from ibuprofen appears when the temperature of the heating source is around 70 • C because the melting point of ibuprofen varies between 75 • C and 78 • C. Then, as the temperature rises, a signal from naproxen appears, which is intensively released at about 150 • C, because its melting point is 152-155 • C. The release of diclofenac occurs at the same time, due to its melting point of 156-158 • C.
samples containing a mix of drugs. During the measurement, the heating was started at 50 °C with a gradual increase to 300 °C. In Figure 8, the first signal from ibuprofen appears when the temperature of the heating source is around 70 °C because the melting point of ibuprofen varies between 75 °C and 78 °C. Then, as the temperature rises, a signal from naproxen appears, which is intensively released at about 150 °C, because its melting point is 152-155 °C. The release of diclofenac occurs at the same time, due to its melting point of 156-158 °C. Then, the FAPA-MS spectra obtained directly from the AFFINILUTE TM MIP-NSAIDs column after absorbing NSAIDs were recorded. In the initial heating phase, we observed signals from ibuprofen (I), which was released first due to it having the lowest melting point (Figure 9). In the first phase of measurements, we also observed a low-intensity signal from naproxen (N). In the second phase of measurements, when the crucible temperature per sample reached approximately 150 °C, we recorded the most intense signals from naproxen at m/z 231 and m/z 248. Then, with increasing temperature, the signals from ibuprofen begin to weaken, which indicates that most of this analyte had already been desorbed. A low-intensity signal appears from diclofenac (D) m/z 296, which desorbs the latest due to it having the highest melting point (Figure 10). Then, the FAPA-MS spectra obtained directly from the AFFINILUTE TM MIP-NSAIDs column after absorbing NSAIDs were recorded. In the initial heating phase, we observed signals from ibuprofen (I), which was released first due to it having the lowest melting point (Figure 9). In the first phase of measurements, we also observed a low-intensity signal from naproxen (N). Then, the FAPA-MS spectra obtained directly from the AFFINILUTE TM MIP-NSAIDs column after absorbing NSAIDs were recorded. In the initial heating phase, we observed signals from ibuprofen (I), which was released first due to it having the lowest melting point (Figure 9). In the first phase of measurements, we also observed a low-intensity signal from naproxen (N). In the second phase of measurements, when the crucible temperature per sample reached approximately 150 °C, we recorded the most intense signals from naproxen at m/z 231 and m/z 248. In the second phase of measurements, when the crucible temperature per sample reached approximately 150 • C, we recorded the most intense signals from naproxen at m/z 231 and m/z 248. Then, with increasing temperature, the signals from ibuprofen begin to weaken, which indicates that most of this analyte had already been desorbed. A low-intensity signal appears from diclofenac (D) m/z 296, which desorbs the latest due to it having the highest melting point ( Figure 10). Then, with increasing temperature, the signals from ibuprofen begin to weaken, which indicates that most of this analyte had already been desorbed. A low-intensity signal appears from diclofenac (D) m/z 296, which desorbs the latest due to it having the highest melting point (Figure 10). In the next analytical phase, we recorded one spectrum of all analytical signals from the three compounds analyzed (Figure 11). It confirms that ibuprofen, naproxen, and diclofenac were bound in the MIP structure and underwent successive thermodesorption during the analysis. In the last stage of measurements, as expected, the signal from ibuprofen disappears, the signal from naproxen decreases significantly, and the signal from diclofenac is the most intense for the longest time ( Figure 12). In the next analytical phase, we recorded one spectrum of all analytical signals from the three compounds analyzed (Figure 11). It confirms that ibuprofen, naproxen, and diclofenac were bound in the MIP structure and underwent successive thermodesorption during the analysis. most of this analyte had already been desorbed. A low-intensity signal appears from diclofenac (D) m/z 296, which desorbs the latest due to it having the highest melting point (Figure 10). In the next analytical phase, we recorded one spectrum of all analytical signals from the three compounds analyzed (Figure 11). It confirms that ibuprofen, naproxen, and diclofenac were bound in the MIP structure and underwent successive thermodesorption during the analysis. In the last stage of measurements, as expected, the signal from ibuprofen disappears, the signal from naproxen decreases significantly, and the signal from diclofenac is the most intense for the longest time ( Figure 12). In the last stage of measurements, as expected, the signal from ibuprofen disappears, the signal from naproxen decreases significantly, and the signal from diclofenac is the most intense for the longest time ( Figure 12).  Szymonik et al. [34] reported that NSAIDs in wastewater reached 0.005-1.59 μgL −1 for diclofenac, 0.05-3.35 μgL −1 for ibuprofen, and 0.03-0.7 μgL −1 for naproxen. Kosjek et al. [35] reported that the diclofenac content in river water was 0.009-0.282 μgL −1 and the naproxen content was 0.021-0.313 μgL −1 . The presented analytical method allows for the determination of the concentration of individual NSAIDs at these levels. The sample was concentrated by adding the appropriate volume of river/lake or sewage water (minimum amount of NSAIDs: 0.2 g in the whole volume) to the column, but not exceeding the total volume of 100 mL. Using a larger volume of water sample applied to the column can reduce the method detection limit by 20 times. To validate the method described in the experimental section, the FAPA-MS analytical technique, the molecularly imprinted polymer from the AFFINILUTE TM MIP-NSAIDs columns, and drug solutions with concentrations of 0.5-1.50 µg (Table 4) were used. Szymonik et al. [34] reported that NSAIDs in wastewater reached 0.005-1.59 µgL −1 for diclofenac, 0.05-3.35 µgL −1 for ibuprofen, and 0.03-0.7 µgL −1 for naproxen. Kosjek et al. [35] reported that the diclofenac content in river water was 0.009-0.282 µgL −1 and the naproxen content was 0.021-0.313 µgL −1 . The presented analytical method allows for the determination of the concentration of individual NSAIDs at these levels. The sample was concentrated by adding the appropriate volume of river/lake or sewage water (minimum amount of NSAIDs: 0.2 g in the whole volume) to the column, but not exceeding the total volume of 100 mL. Using a larger volume of water sample applied to the column can reduce the method detection limit by 20 times.

Conclusions
A new analytical technique combining the use of a molecularly imprinted polymer technique and direct FAPA-MS analysis for the determination of ibuprofen, naproxen, and diclofenac was proposed herein. Commercially available AFFINILUTE TM MIP-NSAIDs columns containing an NSAID-selective polymer layer were used. Preconcentration of trace amounts of analytes was performed via the MISPE method. Two types of analyses were performed: the standard HPLC-MS and the new one, direct FAPA-MS. In the standard method, the analysis consisted of five stages. Sample feeding is preceded by conditioning the column, and the sample must be washed before analysis. Approximately 8 mL of various solvents is used in the whole process. In addition, the sample preparation process is time-consuming and the HPLC analysis time is 15 min. In the proposed method, a raw sample is dosed to the column. MIP with the analyte is washed with 1 ml of a solvent, which is the total amount of solvent used in this analytical process. Then, direct and quick FAPA-MS analysis is performed in about 2 min. The use of MISPE concentration only vaguely improved the analytical parameters determined by the HPLC-MS method. However, the combination of MISPE and FAPA-MS not only improved the detection limit of all analytes to 0.2 µg ibuprofen/naproxen/diclofenac applied once per spoon, but also allowed for shortened analysis time, reduced number of analytical steps, and reduced possibility of errors. An additional advantage was the reduction of the amount of solvent used, due to which the proposed method is not only analytically attractive, but also environmentally friendly and compliant with the principles of Green Chemistry.
The presented new analytical method combining MISPE and FAPA-MS can be a successful alternative to MISPE in HPLC-MS. Besides this, the presented method allows us to quickly and easily determine low-molecular-weight organic compounds. By choosing MIPs selective for different particles, we can analyze different groups of compounds, e.g., drugs in wastewater, performance-enhancing drugs in the blood, unauthorized substances in legal highs, and many others. The main limitations of the method are the existence or possibility of the synthesis of relevant MIPs and the ability to detect ions obtained by FAPA by a mass spectrometer.