Evaluation of a Resorcinarene-Based Sorbent as a Solid-Phase Extraction Material for the Enrichment of L-Carnitine from Aqueous Solutions

: We present ﬁve new sorbents for the evaluation of the pre-concentration of L-carnitine. The sorbents were obtained from copolymerization between butylmethacrylate (BuMA) and ethylene dimethacrylate (EDMA), which were physically modiﬁed on their surface by each one of the ﬁve synthetized resorcinarenes of variable chain length studied, with long-chain resorcinarenes exhibiting better ﬁxation on the copolymer. The characterization of the synthesized resorcinarenes was done using FTIR-ATR, 1 H NMR, 13 C NMR, and mass spectrometry, while characterization of the sorbents was done using FTIR-ATR and scanning electron micrography (SEM). The molecular interaction between L-carnitine and the synthesized resorcinarenes was studied in DMSO via 1 H-NMR spectroscopy and, in the gas phase, via electrospray ionization-mass spectrometry (ESI-MS). The results showed that the short-chain resorcinarenes underwent a stable interaction with the neurotransmitter. Once the sorption of resorcinarenes on the copolymer was accomplished, the best parameters for the evaluation of the L-carnitine preconcentration were established. The solution tests were carried out through LC/MS analysis, obtaining better results for L-carnitine absorption with the short-chain resorcinarenes such as tetra(ethyl)calix[4]resorcinarene and tetra(propyl)calix[4]resorcinarene


Introduction
Some amino acids and compounds derived from them have the ability to participate in metabolic or cellular communication processes as neurotransmitters that chemically regulate neuronal synapses by selectively acting on receptor proteins located in the cell membrane. In general, these chemical transmitters can be classified as high molecular weight neuroactive polypeptides, as well-known amino acids and their derivatives, or as organic compounds (such as hormones) of low molecular weight [1,2], whose bioactive stage can generate any type of response, inhibition, or synaptic excitation. L-carnitine is a trimethylammonio carboxilate synthesized from methionine and lysine that is found in the human diet and plays an important role in fatty acid oxidation; consequently, quantification of carnitine levels in the human body can be used to diagnose disorders of fatty acid oxidation [1][2][3]. Several methods for the detection of carnitine are found in the literature, given its great importance in the human diet [3][4][5]. Among the various methods that have been described for detecting this organic cation, the following should be highlighted: analysis with biosensors [6], radioactive markers [6][7][8], chromatographic methods [9][10][11][12], capillary electrophoresis [13], and fluorescent methods [14][15][16]. Additionally, over the last few decades, advances have been reported in the detection of carnitine through supramolecular chemistry [17,18].
Of these methods, HPLC-MS is currently the most widely used for the analysis of carnitine; however, one of the great difficulties is the matrix effect, which prevents adequate

Materials and Methods
All chemical reagents were acquired from commercial suppliers and used without further purification. IR spectra were recorded on a Thermo Fisher Scientific Nicolet iS10 FT-IR spectrometer with the monolithic diamond ATR accessory and absorption in cm −1 . 1 H-NMR spectra were recorded in DMSO-d 6 at 400 MHz on a Bruker Avance 400 instrument. Chemical shifts are reported in ppm, using the residual solvent signal as a reference. For liquid chromatography-mass spectrometry (LC/MS), the samples were centrifuged at 15,000 rpm for 3 min at room temperature, the supernatant was diluted 1000 times, and 2 µL was analyzed in a Bruker Impact II LC/MS Q-TOF equipped with electrospray ionization (ESI) in positive mode. The chromatographic conditions were as follows: an intensity C18 column (2.1 × 100 mm, 1.8 µm) (Bruker Daltonik, Billerica, MA, USA) at a temperature of 40 • C and a flow rate of 0.250 mL min −1 . Mobile phase water (A) and acetonitrile (B), each containing 0.1% formic acid, were used. Isocratic elution B 5% was used for 2 min. The ESI source conditions were as follows: end plate offset 500 V, capillary 4500 V, nebulizer 1.8 bar, dry gas nitrogen 8.0 L/min, and dry temperature 220 • C. The morphological characterization of the copolymer and the physical modification of the resorcinarenes was done using a QUANTA 200 (FEI) scanning electron microscope with secondary electron and backscattered electron detectors and an EDAX probe. The samples, Processes 2023, 11, 1705 3 of 13 being non-conductive, were pretreated with a gold metallic coating and carbon evaporation equipment. For particle reduction of the synthesized compounds, a 0.8" × 2" high ASTM E11 analytical sieve with an aperture of 106 µm with 140 mesh from Endecotts was used. Melting points were taken on an Electrothermal 9100 SMP11 melting point meter with a heating rate of 1 to 10 • C per min from Stuart-Equipment.

Synthesis of calix[4]resorcinarenes
The procedure for obtaining the macrocycles was adapted from previous research [24,26,38]. Hydrochloric acid (3 mL) was carefully added to a solution of resorcinol (10 mmol) and the respective aldehyde was slowly added (10 mmol), in a 1:1 ethanol/water mixture (20 mL) for short-chain-length resorcinarenes (1-3), while pure ethanol was used for longchain-length resorcinarenes (4,5) due to aldehydes' insolubility. The reaction mixture was reacted at reflux with constant stirring for 8-12 h and subsequently cooled in an ice bath down to 0 • C, then filtered and washed with water. Finally, the filtrate was dried under a vacuum and was characterized via FTIR-ATR, 1 H-NMR, 13 C-NMR, and ESI-MS. In this way, five resorcinarenes were obtained: tetra(ethyl)calix [4]resorcinarene (1), tetra(propyl)calix [4]resorcinarene (2), tetra(hexyl)calix [4]resorcinarene (3), tetra(nonyl)calix resorcinarene (4), and tetra(undecyl)calix [4]resorcinarene (5). The spectroscopic data is shown in Table 1. The binding ability of macrocycles 1, 2, and 3 with L-carnitine was studied in DMSO-d6 via 1 H-NMR. A solution with a concentration of 1.2-2 mg/mL of the macrocycle host was treated with various volume amounts of L-carnitine, according to different molar ratios set between 1:0,1 and 1:2. After each addition, 1 H-NMR was recorded. It should be clarified that the initial concentrations of carnitine varied between 13 and 68 mg/mL, depending on the resorcinarene to be studied, and that this analysis for compounds 4 and 5 was not carried out due to their insolubility in the deuterated solvents used.

ESI-MS Studies
As mentioned above, the gaseous phase experiments were performed on a Bruker Impact II LC/MS Q-TOF equipped with electrospray ionization (ESI) in positive mode. Resorcinarenes 1-5 were each injected into the ion source as an ethanol solution with L-carnitine using 1:1, 2:1, and 1:2 molar ratios.
The mixture was transferred to an airtight container and heated in a water bath at 57 • C for 19-40 h, cooled to room temperature, filtered, washed with EtOH, and dried under a vacuum at 55 • C to constant mass. Finally, the polymer was crushed and sieved to a particle size of 106 µm. Chemical characterization was performed using FTIR-ATR and morphological characterization with SEM.

Physical Modification of Polymers with Resorcinarenes
Resorcinarene (0.6 mmol) was dissolved in 20 mL CHCl 3 (in the case of sample 1, CO(CH 3 ) 2 was used because of insolubility in CHCl 3 ). Then, the sieved polymer (1 g) was added. The mixture was magnetically stirred for 36 h at room temperature, and the resulting solid was filtered under a vacuum, washed with EtOH to neutral pH, and dried at 57 • C until it reached a constant mass, which was compared with the initial mass of the polymer added to determine how much resorcinarene was adsorbed on its surface. Finally, the polymeric material was crushed and sieved to a particle size of 106 µm. The procedure was carried out with a sieved polymer without resorcinarene additions.

SPE Protocol
For the optimization of the SPE procedure, the influences of various parameters on the extraction performance, including time, volume of the sample solution, elution solvent, pH, and concentration of the analyte were studied. Once the ideal extraction conditions were determined, quantitative tests were carried out using a standard solution of carnitine at 0.5 M, which was diluted in water for the preparation of an HPLC-MS calibration curve (six solutions 0.1-0.6 mg/mL) and the preparation of the sample solution 0.02 M to be eluted through a SPE cartridge to quantify the carnitine remaining in solution.
In each quantitative test performed, 200 mg of the sorbent resulting from impregnation was packed into the SPE cartridge using frits and filter paper to avoid loss. Then, the cartridge was preconditioned with methanol and water, and finally the sample solution was passed through the SPE cartridge and was collected in a volumetric flask (10 mL) to be analyzed using HPLC-MS.
The desorption process was done immediately after each sample elution. In order to carry this out, 10 mL of a solution of CH 3 COONH 4 was eluted through the SPE cartridge and was collected in a volumetric flask (10 mL) to be analyzed using HPLC-MS.

Synthesis of Resorcinarenes
The synthesis of tetra(ethyl)calix [4]resorcinarene (1), tetra(propyl)calix [4] resorcinarene (2), tetra(hexyl)calix [4]resorcinarene (3), tetra(nonyl)calix [4]resorcinarene (4), and tetra (undecyl)calix [4]resorcinarene (5) was performed using the methodology described in the literature [24][25][26] (Scheme 1). In this way, the obtaining of compounds 1-5 was done through the acid-catalyzed cyclocondensation of the respective aldehyde with resorcinol in ethyl alcohol at reflux condition. The specific conditions are summarized in Scheme 1. In all cases, a solid was formed that was filtered, washed with water, and characterized using the spectroscopic techniques mentioned. As shown in Table 1, the characterization of the compounds was performed using ESI-TOF/MS, FT-IR, 1 H-NMR, and 13 C-NMR spectroscopy ( Figures S1-S4). Compounds 1-5 had been previously synthesized [39][40][41], and our spectroscopic data agreed with those reported (Figures S5-S18); however, the following aspects should be noted: the signals of the macrocyclic compounds are described and they exhibited similar behavior. The characteristic band for OH was seen in the FT-IR spectrum, and the characteristic bands for aromatic and alkyl substituents were resolved. In 1 H-NMR, the macrocycles showed similar profiles, which contained the characteristic signal of the methylene bridge between 4.09 and 4.32 ppm. For the protons corresponding to the hydroxyls, the signals were resolved between 8.93 and 8.87 ppm. Additionally, the signals corresponding to the aromatic system were observed between 6.15 and 7.23 ppm. The signals at 6.96 ppm for compound 1 and 7.35 ppm for compound 2, which are singlets, showed that the majority conformation was crown in DMSO. Finally, the 13 C-NMR spectra of compounds 1 and 2 showed seven and nine signals, respectively, confirming the obtention of the products.

Synthesis of Resorcinarenes
The synthesis of tetra(ethyl)calix [4]resorcinarene (1), tetra(propyl)calix [4] resorcinarene (2), tetra(hexyl)calix [4]resorcinarene (3), tetra(nonyl)calix [4]resorcinarene (4), and tetra(undecyl)calix [4]resorcinarene (5) was performed using the methodology described in the literature [24][25][26] (Scheme 1). In this way, the obtaining of compounds 1-5 was done through the acid-catalyzed cyclocondensation of the respective aldehyde with resorcinol in ethyl alcohol at reflux condition. The specific conditions are summarized in Scheme 1. In all cases, a solid was formed that was filtered, washed with water, and characterized using the spectroscopic techniques mentioned. As shown in Table 1, the characterization of the compounds was performed using ESI-TOF/MS, FT-IR, 1 H-NMR, and 13 C-NMR spectroscopy ( Figures S1-S4). Compounds 1-5 had been previously synthesized [39][40][41], and our spectroscopic data agreed with those reported (Figures S5-S18); however, the following aspects should be noted: the signals of the macrocyclic compounds are described and they exhibited similar behavior. The characteristic band for OH was seen in the FT-IR spectrum, and the characteristic bands for aromatic and alkyl substituents were resolved. In 1 H-NMR, the macrocycles showed similar profiles, which contained the characteristic signal of the methylene bridge between 4.09 and 4.32 ppm. For the protons corresponding to the hydroxyls, the signals were resolved between 8.93 and 8.87 ppm. Additionally, the signals corresponding to the aromatic system were observed between 6.15 and 7.23 ppm. The signals at 6.96 ppm for compound 1 and 7.35 ppm for compound 2, which are singlets, showed that the majority conformation was crown in DMSO. Finally, the 13 C-NMR spectra of compounds 1 and 2 showed seven and nine signals, respectively, confirming the obtention of the products.

Molecular Interaction Studies between Resorcinarenes and L-Carnitine
Once the resorcinarenes were synthesized, molecular interaction experiments evaluating resorcinarenes 1 to 5 against carnitine were carried out in solution via 1 H-NMR and, in the gas phase, by means of ESI-MS (Figures S20, S22 and S24). It was found that the complexes of compounds 1, 2, and 3 with carnitine are stable in solution and in the gas phase. For example, as shown in Figure 1, the variation in the 1 H-NMR chemical shift of a single proton resonance in carnitine as a function of resorcinarene 2 (for compounds 1 or 3 it is the same behavior) in a molar ratio of 1:1 to carnitine in DMSO-d6 shows great molecular interaction in solution (Figures S19, S21 and S23), while the interaction of L-carnitine with compounds 4 and 5 was considerably lower.

Molecular Interaction Studies between Resorcinarenes and L-Carnitine
Once the resorcinarenes were synthesized, molecular interaction experiments evaluating resorcinarenes 1 to 5 against carnitine were carried out in solution via 1 H-NMR and, in the gas phase, by means of ESI-MS (Figures S20, S22 and S24). It was found that the complexes of compounds 1, 2, and 3 with carnitine are stable in solution and in the gas phase. For example, as shown in Figure 1, the variation in the 1 H-NMR chemical shift of a single proton resonance in carnitine as a function of resorcinarene 2 (for compounds 1 or 3 it is the same behavior) in a molar ratio of 1:1 to carnitine in DMSO-d 6 shows great molecular interaction in solution (Figures S19, S21 and S23), while the interaction of L-carnitine with compounds 4 and 5 was considerably lower.
The stoichiometric relationship was confirmed via electrospray ionization-mass spectrometry, which is a soft ionization technique that allows evaluating this type of interaction, as previously studied with various macrocyclic systems [42][43][44]. The experiments were carried out with different host:guest stoichiometries, producing the same signal for the 1:1 complex in all cases. Using this method, it was possible to establish that the m/z of the complex formed between carnitine and compound 1 was 762.5 ( Figure S20), while the complexes formed between carnitine and compounds 2 and 3 were m/z = 821.4 ( Figure S22) and m/z = 986.6 ( Figure S24), respectively ( Figure 2). The formation of these complexes is not surprising, since the structural arrangement of resorcinarenes in crown conformation and the tetraalkyl ammonium group of carnitine coordination is possibly dictated by πinteractions in the resorcinarene cavity. However, as observed by 1 H-NMR, the interaction  The stoichiometric relationship was confirmed via electrospray ionization-m trometry, which is a soft ionization technique that allows evaluating this type o tion, as previously studied with various macrocyclic systems [42][43][44]. The exp were carried out with different host:guest stoichiometries, producing the same s the 1:1 complex in all cases. Using this method, it was possible to establish that th the complex formed between carnitine and compound 1 was 762.5 ( Figure S20), w complexes formed between carnitine and compounds 2 and 3 were m/z = 821.4 The observed results for the behavior of resorcinarenes 4 and 5 in solution were difficult to evaluate due to their low solubility in the solution of the deuterated solvents used for the study via 1 H-NMR. On the other hand, in the gas phase, the interaction was not evident, despite obtaining homogeneous solutions, in the mass spectra of the signals of carnitine and resorcinarenes observed separately. These results suggest a low interaction of these resorcinarenes with L-carnitine. Despite the fact that the resorcinarenes on their upper rim are structurally equivalent, the effect of the interaction of the alkyl chains on the lower rim affects the cavity size, this effect being greater with long-chain resorcinarenes, as has previously been observed in other studies [45,46].

Obtention and Characterization of Sorbents
As shown in Scheme 2, the copolymer poly(BuMA-co-EDMA) was obtained from the reaction between EDMA and BuMA in cyclohexanol in the presence of 1,1-azobis (cyclohexanecarbonitrile), in accordance with previous reports [23,24]. The chemical characterization of poly (BuMA-co-EDMA) was performed via FT-IR, and the morphological characterization was performed via SEM. This copolymer had been previously synthesized, and our experimental data agreed with those reported [24,25]. The next step in the process was the physical modification of poly(BuMA-co-EDMA) with resorcinarenes 1-5. In this way, impregnation is a useful technique for the adsorption of molecules on the surface of a polymeric matrix, which enables more adsorption sites, rapid mass transfer, fast binding kinetics, and high selectivity copolymer modification due to lipophilic interactions between alkyl chains on the lower rim of the resorcinarenes with butyl chains in the copolymer. The observed results for the behavior of resorcinarenes 4 and 5 in solution were difficult to evaluate due to their low solubility in the solution of the deuterated solvents used for the study via 1 H-NMR. On the other hand, in the gas phase, the interaction was not evident, despite obtaining homogeneous solutions, in the mass spectra of the signals of carnitine and resorcinarenes observed separately. These results suggest a low interaction of these resorcinarenes with L-carnitine. Despite the fact that the resorcinarenes on their upper rim are structurally equivalent, the effect of the interaction of the alkyl chains on the lower rim affects the cavity size, this effect being greater with long-chain resorcinarenes, as has previously been observed in other studies [45,46].

Obtention and Characterization of Sorbents
As shown in Scheme 2, the copolymer poly(BuMA-co-EDMA) was obtained from the reaction between EDMA and BuMA in cyclohexanol in the presence of 1,1-azobis (cyclohexanecarbonitrile), in accordance with previous reports [23,24]. The chemical characterization of poly (BuMA-co-EDMA) was performed via FT-IR, and the morphological characterization was performed via SEM. This copolymer had been previously synthesized, and our experimental data agreed with those reported [24,25]. The next step in the process was the physical modification of poly(BuMA-co-EDMA) with resorcinarenes 1-5. In this way, impregnation is a useful technique for the adsorption of molecules on the surface of a polymeric matrix, which enables more adsorption sites, rapid mass transfer, fast binding kinetics, and high selectivity copolymer modification due to lipophilic interactions between alkyl chains on the lower rim of the resorcinarenes with butyl chains in the copolymer.  The ATR-FTIR spectra of the obtained sorbents show characteristic signals. For example, poly(BuMA-co-EDMA)-1 and poly(BuMA-co-EDMA)-2 ( Figure 3d) show signals corresponding to a C=O bond at 1714 cm −1 . Other characteristic peaks were found at 3070 cm −1 , assigned to the C-H stretching vibration of aromatic rings, 2910 cm −1 due to the C-H stretching vibration alkyl chains, 1455 cm −1 due to the C-H bending vibration, and 1155 cm −1 due to the C-O stretching vibration of the ester group. The adsorption bands at 3330 cm −1 of the copolymer modified with compounds 1 or 2 can be ascribed to the characteristic adsorption peaks of the OH stretching vibrations of resorcinarene rings. Furthermore, typical bands of benzene rings increased, as observed at 1400-1500 cm −1 . The morphologies of the obtained sorbents were analyzed via scanning microscopy images (Figure 3a-c). First, it was observed that the morphology of the copolymer is quite uniform, with continuous pores. The uniform and rich open-pore structure is suitable for mass transfer efficiency in the impregnation process with resorcinarenes 1, 2, or 3. Secondly, the morphologies of the impregnated copolymers poly(BuMA-co-EDMA)-1 poly(BuMAco-EDMA)-2 or poly(BuMA-co-EDMA)-3 maintain their porosity characteristics, which are suitable for mass transfer efficiency as well as for carnitine adsorption.
The amount of fixed resorcinarenes 1-5 on the polymeric surface was determined via gravimetric analysis; these results show that the polymer poly(EDMA-co-GMA) has a high degree of substitution of the studied macrocycles on the surface (Figure 4). The best substitution was found with long-chain resorcinarenes (4 and 5), which was 352.2 µmol/g for poly(BuMA-co-EDMA)-4 and 367.5 µmol/g for poly(BuMA-co-EDMA)-5. The gravimetric data indicated that resorcinenarenes 1-3 were attached to the copolymer, but in a smaller proportion; however, it can be considered that the impregnation process was efficient. These results are in agreement with other investigations of the physical modification of poly(GMA-co-EDMA) [23,24].

Optimization of the SPE Procedure
In order to optimize the extraction conditions, the best modified polymer was chosen for these tests, in this case poly(BuMA-co-EDMA)-1, because resorcinarerene 1 exhibited the best interaction with carnitine in aqueous solution. In this way, the effect of the contact time in the extraction process for different times (30-120 min) was tested. As was evident from the results, the extraction reached a maximum at 60 min for the sorbents tested; after this time, no appreciable changes were observed. The flow rate of the sample solution was constant for all the experiments. To obtain high desorption efficiency, the selection of a suitable elution solvent was necessary. Five solvents were tested: water, methanol, acetonitrile, ethanol, and acetone. Preliminary experiments via HPLC showed that there were interfering peaks in the chromatograms when ethanol, acetone, or acetonitrile was used as the eluent, possibly because part of the macrocycle dissolves in the eluent. Between the other two solvents, using methanol as the elution solvent provided the highest extraction recoveries; however, with water the results were comparatively good. For this reason, water was chosen as the solvent in the process. The appropriate volume of water was further investigated using volumes of 1.0-4.0 mL. The extraction recoveries increased with the increasing water volume from 1.0 mL to 3.0 mL and remained nearly constant at higher volumes. Therefore, water with a volume of 3.0 mL was selected as the eluent.
The concentration of analyte played a significant role in the SPE procedure. Different concentrations of the analyte, from 0.005 to 0.1 M, were tested. The results showed that the extraction efficiency increased with increasing concentration up to 0.02 M; above this value, the results tended to be constant. When the concentration of L-carnitine was too small, most of the analyte was not extracted because there were too many unoccupied active sites, resulting in low extraction recovery. Therefore, a concentration of 0.02 M was

Optimization of the SPE Procedure
In order to optimize the extraction conditions, the best modified polymer was chosen for these tests, in this case poly(BuMA-co-EDMA)-1, because resorcinarerene 1 exhibited the best interaction with carnitine in aqueous solution. In this way, the effect of the contact time in the extraction process for different times (30-120 min) was tested. As was evident from the results, the extraction reached a maximum at 60 min for the sorbents tested; after this time, no appreciable changes were observed. The flow rate of the sample solution was constant for all the experiments. To obtain high desorption efficiency, the selection of a suitable elution solvent was necessary. Five solvents were tested: water, methanol, acetonitrile, ethanol, and acetone. Preliminary experiments via HPLC showed that there were interfering peaks in the chromatograms when ethanol, acetone, or acetonitrile was used as the eluent, possibly because part of the macrocycle dissolves in the eluent. Between the other two solvents, using methanol as the elution solvent provided the highest extraction recoveries; however, with water the results were comparatively good. For this reason, water was chosen as the solvent in the process. The appropriate volume of water was further investigated using volumes of 1.0-4.0 mL. The extraction recoveries increased with the increasing water volume from 1.0 mL to 3.0 mL and remained nearly constant at higher volumes. Therefore, water with a volume of 3.0 mL was selected as the eluent.
The concentration of analyte played a significant role in the SPE procedure. Different concentrations of the analyte, from 0.005 to 0.1 M, were tested. The results showed that the extraction efficiency increased with increasing concentration up to 0.02 M; above this value, the results tended to be constant. When the concentration of L-carnitine was too small, most of the analyte was not extracted because there were too many unoccupied active sites, resulting in low extraction recovery. Therefore, a concentration of 0.02 M was selected for use in the subsequent experiments. Finally, the pH of the sample solution will have an influence on both L-carnitine and the sorbent; nevertheless, the structures of resrocinarenes grafted onto a copolymer play a crucial role in enriching and extracting the neurotransmitter in the proposed SPE procedure. The hydroxyl groups in resorcinarene provide hydrogen bonding and dipole-induced dipole interactions, which aid in extracting L-carnitine from the aqueous solution. Among these groups, the ionic form of the hydroxyl group is influenced by the pH of the sample solution. Therefore, very high pH values were not considered to be appropriate because they would affect the stability of the modified sorbent. To investigate the effect of the pH of the neurotransmitter solution on the extraction process, different pH values (2)(3)(4)(5)(6)(7)(8) were tested. An increase in the extraction efficiency for L-Carnitine was observed when the pH of the aqueous solution was changed from pH 2 to 6. At low pH, the hydroxyl groups interacted weakly with the tetralkylammonium group of L-carnitine, while at pH 6, some of the deprotonated hydroxyls of the macrocyclic resulted in stronger interactions with L-carnitine.

Analyses of the Extraction of L-Carnitine with the Modified Sorbents
Once the best conditions had been established for the carnitine extraction process with the modified sorbents, the final step consisted of determining the amount of carnitine recovered in an extraction-desorption cycle. For this purpose, it was decided to evaluate the process by means of a carnitine calibration curve obtained with the HPLC-MS technique, as described in the experimental section. As shown in Figure 5, the calibration curve of L-carnitine showed good linearity in the range of 0.1-0.6 mg/mL, with a correlation coefficient (R2) of 0.9984. According to the results established for the extraction with the three sorbents, it was observed that evidently poly(BuMA-co-EDMA)-1 provided higher extraction efficiency for L-carnitine (67%) than poly(BuMA-co-EDMA)-2 and poly(BuMAco-EDMA)-3, which had efficiencies close to 35%. Finally, the stability of the sorbents was evaluated by performing adsorption-desorption cycles. The sorbents were recycled at least four times without any obvious loss of extraction recovery, indicating that the modified sorbents were stable enough to be recycled in the extraction procedure. tetralkylammonium group of L-carnitine, while at pH 6, some of the deprotonated hydroxyls of the macrocyclic resulted in stronger interactions with L-carnitine.

Analyses of the Extraction of L-Carnitine with the Modified Sorbents
Once the best conditions had been established for the carnitine extraction process with the modified sorbents, the final step consisted of determining the amount of carnitine recovered in an extraction-desorption cycle. For this purpose, it was decided to evaluate the process by means of a carnitine calibration curve obtained with the HPLC-MS technique, as described in the experimental section. As shown in Figure 5, the calibration curve of L-carnitine showed good linearity in the range of 0.1-0.6 mg/mL, with a correlation coefficient (R2) of 0.9984. According to the results established for the extraction with the three sorbents, it was observed that evidently poly(BuMA-co-EDMA)-1 provided higher extraction efficiency for L-carnitine (67%) than poly(BuMA-co-EDMA)-2 and poly(BuMAco-EDMA)-3, which had efficiencies close to 35%. Finally, the stability of the sorbents was evaluated by performing adsorption-desorption cycles. The sorbents were recycled at least four times without any obvious loss of extraction recovery, indicating that the modified sorbents were stable enough to be recycled in the extraction procedure.

Conclusions
We have presented a copolymer modified with resorcineranes for the pre-concentration of carnitine with an acceptable efficiency in the analyte recovery process. These macrocycles were highly hydrophobic, which is essential for the modification and stability of the sorbent. The host-guest interaction between resorcinarenes and L-carnitine was confirmed via 1 H-NMR and mass spectrometry, showing good affinity in both cases. Among the two macrocycles evaluated as candidates for the pre-concentration process, tetra(ethyl)calix [4]resorcinarene exhibited the highest affinity in the adsorption process

Conclusions
We have presented a copolymer modified with resorcineranes for the pre-concentration of carnitine with an acceptable efficiency in the analyte recovery process. These macrocycles were highly hydrophobic, which is essential for the modification and stability of the sorbent. The host-guest interaction between resorcinarenes and L-carnitine was confirmed via 1 H-NMR and mass spectrometry, showing good affinity in both cases. Among the two macrocycles evaluated as candidates for the pre-concentration process, tetra(ethyl)calix [4]resorcinarene exhibited the highest affinity in the adsorption process on the modified sorbent. The sorbents were successfully applied to the pre-concentration of L-carnitine in a synthetic solution. This macrocycle-copolymer provides a new method for the pre-concentration of the neurotransmitter L-carnitine. While this investigation aimed to propose a new lipophilic sorbent for the pre-concentration of L-carnitine and focused more on fundamental characterization, the obtained sorbents could be useful in biological samples and for pre-concentration processes above micromolar range. However, we believe there is still room for technical improvements based on the proposed copolymer, including the fabrication of sorbents for the pre-concentration of other neurotransmitters.

Data Availability Statement:
The data that support the findings of this study are not openly available due to the fact that the data form part of an ongoing study but are available from the corresponding author upon reasonable request.