Introduction

Nicotinic acid N-methylbetaine (NAMB), also known as trigonelline, is an alkaloid with the chemical formula C7H7NO2. It is a niacin metabolic product excreted in mammalian urine [1]. When the nitrogen atom of niacin is methylated, a zwitterion is formed (vitamin B3). It is reported to have antitumor, antibacterial, antiviral, neuroprotective, anti-migraine, sedative, hypolipidemic, and hypoglycemic properties. This could also improve memory, impede the aggregation of platelets, and reduce the invasive activity of cancer cells [2, 3]. Since NAMB affects a wide range of human diseases and disorders, many researchers are interested in measuring its concentration in biological fluids [4, 5]. Biological fluids such as plasma and urine are complex matrices which often consider analyte extraction as a fundamental step in the analytical procedure [6,7,8]. Cleaning up the sample matrix would be very critical for preconcentrating the analyte from a dilute sample solution and achieving low detection. Different methods of sample preparation, consisting of liquid–liquid extraction (LLE) and solid-phase extraction (SPE) are reported for the study of various types of analytes. These procedures are commonly lengthy and time-consuming, and they disclose significant consumption of toxic organic solvents [9]. In recent years, considering some advantages arising from the intrinsic properties of magnetic particles, substantial attention has been paid to use of magnetic particles for solid-phase extraction [10]. MSPE technique overwhelms difficulties like packing of the column and separation of the phases that could be accomplished simply by utilizing field of an external magnet [11]. Magnetic nanoparticles themselves actually have very little ability to remove analytes from matrices, especially complex matrices such as biological fluids, and therefore many magnetic sorbents have been developed as a composite of magnetic nanoparticles and a wide variety of sorbents [12]. Magnetic nanocomposites are one of the efficient approaches with numerous promising applications including biomedical applications [13, 14] and water remediation [15,16,17,18]. MNPs such as Fe3O4 are good candidates for magnetic carrier technology due to high sorption capacity, strong magnetic properties, and low toxicity [19]. Owing to the special physicochemical characteristics, including thermal and mechanical stability, high cation exchange efficiency, large chemically active surfaces, and unusual interlamellar surfaces, the clay mineral may be used as a matrix of magnetic nanoparticles and other suitable molecules [20,21,22,23,24]. Smectites, kaolinite, and micas are the three main clay groups. If montmorillonite or a part of other smectites is the notable clay mineral in the clay, it is termed bentonite (Ben) with the basic formula of (Na)0.7(Al3.3Mg0.7) Si8O20(OH)4. nH2O [25]. Ben contains 70–95% montmorillonite, dioctahedral specie of the smectite mineral group. Montmorillonite is a 2:1 type hydrous aluminosilicate with the octahedral sheet “sandwiched” between two tetrahedral sheets. Ben is a 2:1 sheet of rock with one octagonal layer made up of magnesium, aluminum or iron oxide, and two tetrahedral layers of silicon oxide (SiO4)4, forming a plate in which van der Waals forces hold layers together [26]. Negative facial charges of the particles are due to isomorphic substitutions of Al3+for Si4+ in tetrahedral sites and Mg2+ for Al3+ in octahedral sites [25, 27]. Because of its swelling, rheological and colloidal properties with specific features, such as high capacity for cation exchange, high specific surface area, small price, and widespread availability, are easily accessible for use in various industrial fields [28]. In many cases, the use of adsorbent is not adequate to isolate the analytes more selectively, and the use of a selector medium in the extracting composite is required. A broad variety of selector materials has been used, from molecularly imprinted polymers to compounds that form inclusion-occlusion complexes such as cyclodextrins. Cyclic oligosaccharides known as cyclodextrins (CDs) (α, β or γ) can form complexes of inclusion in aqueous solutions with a wide range of polar and non-polar compounds like monoaromatic and polyaromatic hydrocarbons through 1:1, 1:2 or even 2:2 stoichiometry [29]. β-CD, perhaps the most commonly used CDs in the biomedical science and various chemistry fields [30, 31], is a cyclic oligosaccharide consisting of seven D-glucopyranose covalently joined by α-1,4-glycosidic bonds obtained from common starch enzymatic hydrolysis, that has an inner truncated cone shape of almost 8 Å depth and 6,0–6,4 Å internal diameter. The cavity has a relatively low polarity, and then it can host inside guest molecules, which may improve the solubility, dispersing properties, physical and chemical stability of some drugs via the inclusion complexes. CDs can be coupled with INPs, which have physicochemical interesting multifunctional properties for biomedical applications. The complexation of β-CD with aromatic compounds [32], drugs [33, 34], cholesterol [35], and Ben/INP composites [36], with some chemicals like methylene blue, [37] lead cations [38], phosphorus [39], and for chemical sensing [40] has been studied. In the current method, for the first time, a novel strategy for extraction and determination of NAMB as a supplement in plasma samples by applying magnetic Ben-βCD /Fe3O4 NPs was developed and validated.

Materials and methods

Chemicals and reagents

Reagents and chemical compounds utilized in this study have been of analytical quality, and all samples were made using deionized high purity water purchased from Sabalan Nano Co. (Iran). Methanol of HPLC grade, octane sulfonic acid, FeCl2 0.4 H2O, NaBH4, ethanol, and trichloroacetic acid was provided by Merck Co. (Darmstadt, Germany). NAMB and β-CD have been obtained from Sigma-Aldrich, Germany. Bentonite was purchased from Kerman bentonite mine, Kerman, Iran. This was pulverized and sieved to a particle diameter less than 75 μm and dried in the oven at 110 °C for 24 h until it was used. In a 100 mL volumetric bottle, a NAMB stock standard solution of 100 mg/L has been prepared by dissolving 10.0 mg of NAMB in about 10 mL of ethyl alcohol and diluting to volume with deionized water.

Instrumentation

Knauer Platinum Blue V6900 UPLC instrument used to perform liquid chromatographic analysis (Berlin, Germany). The UPLC equipped with online degasser K-5020, UV detector K-2501, a 10 μL loop six-port/three-channel injection valve, and a p-1pump. The chromatographic data were processed using a desktop computer with a Eurochrom® 2000 LC software. All separations were made on a VDSPher® 100 C18 guard column and ODS-H C18 (5 μm in particle size, 4.6 mm in internal diameter and 250 mm in length). A combination of methanol and 0.05 M ammonium acetate of pH 3.1 (90:10, v/v) was used in isocratic elution as the mobile phase at a flow rate of 0.5 mL/min. Analytes were quantified by a UV–Vis detector set at maximum wavelength of NAMB (264 nm). It should be mentioned, of course, that a UV–Vis spectrophotometer (lambda 25, PerkinElmer, USA) was used to measure NAMB absorbances throughout the optimization phase. A Digital ion analyzer made by Metrohm Company (Switzerland) has been used to measure the pH of solutions at room temperature. Infrared spectra were obtained by the FTIR spectrometer (Bruker, USA). TESCAN vega3 scanning electron microscope (Czech Republic) was used to obtain SEM micrographs of the delicately dispersed powdered sample on the stainless steel inserts. XRD was performed on an X-ray diffractometer (Philips-pw12c, Amsterdam), with a source of radiation of Cu Kα. The magnetic properties of the samples were investigated using vibrating sample magnetometer (MKK-LBKFB, Iran). Atomic force microscopy (AFM) (Ara 0101) was used to examine the surface of the matrix.

Preparation of the Ben/INP

The nanoparticles of Ben/INP have been obtained by reducing Fe2 + to Fe by borohydride. Fe-to-bentonite mass ratio was 1:5. In summary, a solution of Fe was prepared by dissolving 1.07 g FeCl2·4H2O equal to 0.3 g Fe and was added to the ethanol: water (4:1, v/v) mixture, subsequently to this solution, 1.50 g bentonite was incorporated and the mixture was allowed to stand for 30 min in an ultrasonic shaker [25]. In the meantime, sodium borohydride solution was made by dissolving 0.61 g NaBH4 in 100.0 mL of deionized water. The borohydride solution was then added to the aqueous Ben/Fe2+ dispersion drop-wise. The mixture was stirred for an extra 10 min since introducing the borohydride solution. The reaction represents the reduction in Fe ions by borohydride ions is:

 > Fe2+and Fe2+(aq) + 2BH4− + 6H2O →  > Feo(s) + Feo(s) + 2B(OH)3 (aq) + 7H2 (g).

 > Fe2+ refers to Fe ions bound to the bentonite surface, Fe2+ refers to aqueous Fe ions, > Fe° (s) refers to Fe nanoparticles on the bentonite surface, and Fe°(s) refers to nanoparticles of unadsorbed Fe. Vacuum filtration was used to remove the solid material from the liquid phase. Absolute ethanol (25 mL portions) was applied to clean the solid particles at least three times, subsequently dried overnight at 50 °C [25].

Preparation of the nano-sorbent

10 mL of 4 mg/mL Ben/INP suspension was combined with 10 mL of 4 mg/mL β-CD aqueous solution for the linking of nano-hybrid Ben/INP with β-CD. The vial was placed in a water bath at 60 °C for 3.5 h, upon being vigorously stirred. The resulting substance was precipitated, and to eliminate excess β-CD, it was isolated by a magnetic field over many cycles [41].

Adsorption and desorption experiments

In a small glass beaker, 50 mg of the nano-sorbent was loaded. After which, 10 mL of the standard or test solution that contains 10 mg/L NAMB was added to the flask. The beaker was put in a shaker to distribute the nano-sorbent homogenously throughout the solution. NAMB adsorption on the substrate has been conducted at room temperature by mixing the combination under sustained mechanical stirring for 10 min at 800 rpm. After that, the sorbent was collected by the external magnetic force across one bottom of the bottle, and the supernatant solution was immediately removed. The isolated sorbent had been eluted with 2 mL of NaCl 1 M to desorb the analyte. The desorption process was greatly enhanced by swirling on a mixer over 20 min. With the aid of a magnet, the above stage had also been finished. The eluent clear solution containing NAMB was injected into the HPLC. Working solutions were made to use the same concentration of NAMB to conduct the absorbance readings. For more than eight hours, all standard solutions were required to enter equilibrium with Ben/INP/β-CD, insulated from light. Then, the absorbance of each solution was determined by a 264 nm wavelength spectrophotometer. Almost all tests were done at ambient temperature.

To find important variables for adsorption in the screening step, Design Expert software version 10 was used to design the experiments. Experimental variables including time, pH, and amount of adsorbent were selected as major factors may affect the adsorption. In this section, using the method of two-level factorial design with three center points, 27 runs were designed, which is presented in the table S1(in supplementary information). In order to optimize the variables identified at the screening phase, experiments consisting of 11 runs were designed using the software by the response surface methodology (RSM) and the central composite design (CCD) as shown in Table S2. From such a stage, the optimal conditions for adsorption were being assessed to verify the model (p = 0.05). All methods described for adsorption, were conducted in the same manner, for desorption utilizing experimental design approach based on the variables including temperature, salt, time, and volume in three stages of screening, optimization, and validation to achieve an acceptable mathematical model. The screening and optimization steps, built with the aid of the software, involved 18 and 11 runs which are given in table S3 and table S4, respectively.

Real sample analysis

The proposed method was applied to the analysis of blank plasma samples spiked with NAMB. Collected plasma samples were used to be spiked with the prepared sample solutions consisting of 20–10,000 ng/mL of NAMB. The 950 µL plasma samples spiked with 50 µL standard NAMB solutions were thoroughly mixed and then diluted with 200 μL perchloric acid 10% and centrifuged at 4000 rpm for 10 min to precipitate and remove interfering proteins. In a 1000 μL microtube, 0.9 mL de-proteinized plasma and 10 mg of nano-sorbent was added. The microtube was put inside an ultrasound bath for 1 min, in order to distribute the nano-sorbent uniformly across the entire solution. NAMB adsorption on the sorbent was accomplished by shaking within 5 min at ambient temperature with steady stirring (500–800 rpm) of the sample. Eventually, the sorbent was collected through an external magnetic field on one side of the beaker. To desorb the analyte, the separated sorbent was eluted at 100 °C with 0.5 mL of NaCl 0.8 M. The process of desorption has been promoted by mixing on a shaker. The bright, filtered liquid was used to be applied in the HPLC system for NAMB determination. Precision and recovery rate of the method were performed with three repetitive measurements at least in three concentrations of low, medium, and high in the concentration range of the calibration curve. The LOD was also calculated by multiple measurement of the blank response using Eq. (1):

$$LOD = \overline{X} + 3SD$$
(1)

The LOQ was also calculated using the formula (2):

$$LOQ = \overline{X} + 10SD$$
(2)

Pharmacokinetic assay

The method was fully used for the assessment of NAMB in human plasma in a PK study. The PK research project was authorized by the Ethics Board of Kerman University of Medical Sciences (No. IR.KMU.REC.1396.2511). Three healthy male participants aged between 23 and 27 years (average 25) and an average body weight of 68 kg were recruited. Before joining the research, the participants received a complete description of the goals and purpose of the project, the safety of the medications under study, and were able to withdraw from the study at any point without restriction, and they signed an informed consent and gave it to the authorized person. Recruiters were required to avoid taking any medicine for ten days prior to enrolment and must not smoke or drink alcohol. They were checked for any detrimental effects during and 48 h after the trial. Study participants were given a 250 mg dose of NAMB on an empty stomach with 250 ml of water. After dosing, plasma samples collected from of the scalp vein at 0.00, 0.17, 0.33, 0.50, 0.75, 1.00, 1.50, 2.00, 3.00, 4.00, 6.00, and 12.00 h. Blood samples were gathered into EDTA test tubes, and the plasma was separated by centrifugation for five minutes at 3000 rpm.

Results and discussion

Characterization of bentonite/Fe nanoparticles /β-CD nano-hybrid

The buildup of Ben-βCD-INP nano-hybrid was examined by FTIR spectroscopy. Figure 1(a–f) shows the FTIR spectra of β-cyclodextrin, iron nanoparticles, Ben-βCD, Ben, Ben-βCD-INP, and Ben-INP, respectively. The spectrum of raw Ben exhibits bands at 3423 and 3621 cm−1. The sharp peak at 3621 cm−1 is assigned to OH stretching vibrations of the structural hydroxyl group. The bands about 1640 cm−1 and 3440 cm−1 seem to be due to the bending vibration of O\\H and H\OH stretch vibrations in coordinated water molecules.

Fig. 1
figure 1

FTIR spectra of a β-cyclodextrin, b iron nanoparticles, (c)bentonite/β-cyclodextrin, d bentonite, e bentonite/iron nanoparticles/β-CD and f bentonite/iron nanoparticles

Full size image

The bending presented characteristic (Si, Al)\OH stretching bands (3622 cm−1) and stretching Si\O\\Si (1035 cm−1). Among Ben-INP, relatively similar bands appeared in about 2926 cm−1, 1390 cm−1, and 595 cm−1. It was also observed that the C–O–C stretching bands were at 1010–1060 cm−1 for the α (1–4) glycosidic linking. A complex 1038 cm−1 band is attributed to the stretching vibrations of Si–O groups, whereas the 518 and 466 cm−1 related bands are linked to Al–O– Si and Si–O–Si bending vibrations, respectively. At 630 cm−1, the band was allocated to out-of-plane vibrations coupled with Al–O and Si–O.

The X-ray diffraction of a powder (XRD) is a highly useful tool for characterizing material structure and to show the production of Fe nanoparticles on the surface of Ben and the structure of crystals. Figure 2 displays Fe nanoparticles and Ben-βCD-INP XRD pattern. The characteristic Ben peaks (2θ = 20.9°, 26.7°, and 27.8°) are lowered when Fe nanoparticles are increasingly loaded. A variety of small peaks in Ben-INP XRD patterns originate from the intrinsic structure of Ben. In XRD patterns, characteristic peaks were observed at 2θ = 32.7°, 43.4°, 53.4°, 57.3°, 63.2°, and 74.7° (Fig. 2 e) which matched with the standard iron nanoparticle sample [42]. The reflection peak positions of Fe nanoparticles and Ben-βCD were in well agreement with the reported XRD patterns [43] and indicated that the crystalline structure of Fe3O4 was preserved after the polymerization and CD conjugation processes.

Fig. 2
figure 2

XRD patterns of a β-cyclodextrin, b bentonite/ iron nanoparticles c bentonite/ iron nanoparticles/β-cyclodextrin, d bentonie/β-cyclodextrin, e iron nanoparticles and f bentonite/ iron nanoparticles/sodium Borohydrid

Full size image

SEM has examined the surface morphology of the as-prepared products. Figure 3a–c shows the SEM image of Ben/INP, Ben, and Ben-βCD, respectively. From the SEM micrographs, one can see that the Ben/INP displayed more depleted and porous surface morphology. Consequently, there were a number of INP on the surface whose size was 49.6, 57, 57.1, and 59.4 nm, as shown in Fig. 3a. Then during the reaction, the magnetic INP was placed appropriately on the Ben surface. The TEM images of INPs and Ben-βCD-INP are shown in Fig. 4a and b. TEM images revealed all forms and also illustrated the INPs core–shell structure. The diameter of the particle was within 10–60 nm, with a coating thickness of 3–4 nm. In the TEM images, the absence of the crystal structure periphery suggested the coating to be amorphous. In the scattered nanoparticles, the thickness of the shell and chain-like morphology were identical. TEM images for samples of Ben/INP showed that over time the nanoparticles tended to maintain their dispersion over the Ben particles.

Fig. 3
figure 3

SEM images of a bentonite/ iron nanoparticles, b bentonite and c bentonie/β-cyclodextrin

Full size image
Fig. 4
figure 4

TEM images of a iron nanoparticles and b bentonite/ iron nanoparticles/ β- CD

Full size image

VSM could be applied to evaluate the magnetic properties of the samples at ambient conditions. Analysis of magnetic hysteresis loops, dispersion, and agglomeration of the (a) INPs and (b) Ben/β-CD-INP is shown in Fig. 5a and b. In either curve, there is no obvious hysteresis indicating super-paramagnetic. It is found that the profile of the two curves is very similar, although the saturation magnetization value of Ben/β-CD-INP (56 emu. g−1) is lower than that of the INP (29 emu. g−1) showing that INPs have been coated by Ben/β-CD. Figure 6, shows three-dimensional (3D) atomic force microscopy of (a) bentonite/iron nanoparticles, (b) bentonite/ iron nanoparticles/β-cyclodextrin, and (c) bentonie/β-cyclodextrin. BET analysis of (a) β-cyclodextrin, (b) bentonite/iron nanoparticles (c) bentonite/iron nanoparticles/β-cyclodextrin, (d) bentonie/β-cyclodextrin, (e) iron nanoparticles and (f) bentonite/iron nanoparticles/sodium Borohydrid is shown in Fig. 7 and Table 1.

Fig. 5
figure 5

Magnetization curves of a iron nanoparticles and b iron nanoparticles/bentonite/β-cyclodextrin

Full size image
Fig. 6
figure 6

Atomic force microscopy (AFM) of a bentonite/ iron nanoparticles, b bentonite/ iron nanoparticles/β-cyclodextrin, and c bentonie/β-cyclodextrin

Full size image
Fig. 7
figure 7

BET analysis of (a) β-cyclodextrin, b bentonite/ iron nanoparticles c bentonite/ iron nanoparticles/β-cyclodextrin, d bentonie/β-cyclodextrin, e iron nanoparticles and f bentonite/ iron nanoparticles/sodium Borohydrid

Full size image
Table 1 BET analysis results of the as-prepared nanocomposites
Full size table

Pure nanoparticles as an adsorbent are widely known to be unsuitable in many applications for a variety of reasons. The most important issue is that these materials have a high tendency to agglomerate due to strong magnetic attraction among particles, van der Waals forces, and a high energy surface, resulting in severe reactivity loss and extremely high pressure drops in conventional treatment systems [44,45,46]. In the absence of any surface coating material, magnetic iron oxide particles have hydrophobic surfaces with a significant surface-area-to-volume ratio [47]. The usage of iron nanoparticles in combination with clay minerals could help the problem such nanoparticle aggregation in aqueous solutions and iron nanoparticles' comparatively high zero point of charge. High adsorption capabilities for cations, medicines, and colors have been demonstrated for β-cyclodextrin (46). Cyclodextrins are hydrophobic inside and hydrophilic outside cyclic oligosaccharides that can form complexes with hydrophobic compounds [48]. Huang et al., 2014 showed that the cyclodextrin is also capable of increasing the stability of magnetic nanoparticles in water, improving its dispersion in the aqueous medium for a longer period. CDs are used for controlled delivery of organic, inorganic, biological, and pharmaceutical molecules due to their ability to form inclusion complexes with diverse guest molecules by encapsulating the non-polar part of the guest into its hydrophobic cavity and stabilizing the polar part by the polar rims [49]. Efficiency of the current extraction method for plasma samples was compared with other reported methods. In previous studies, liquid–liquid extraction technique [50, 51], protein precipitation pretreatment [52] and isotope dilution assay [53] had been reported to extract NAMB in biological samples. The present method possessed critical advantages such as specific affinity to the analyte, using nontoxic and economical solvents rather than expensive organic solvents and acceptable recovery values.

Optimization of the experimental design by design expert

The model prediction for maximum NAMB extraction was compared to the experimental result at optimal operating conditions. A good agreement between the model prediction and experimental results confirms the soundness of the developed model. Model F-value of 24.31 (Table S5) means that the model is meaningful. There is just a 0.02 percent risk that such a high F-value will happen because of noise. Values less than 0.05 for “p value” indicate that the terms of the model are significant and greater than 0.10 indicate the model terms are not important. The 18.38 “Lack of Fit F-value” (Table S5) means that there is a 5.24 percent probability that this high “Lack of Fit F-value” will occur because of noise and indicate that it is not significant. In the next step, the model was optimized to find the optimum amount of the variables (table S2). Data (table S6) showed that the model was significant, and the lack of fit was insignificant. Among the variables, pH showed a significant effect. To find the best variables and the range of the variables affecting the desorption of NAMB adsorbed on the adsorbent, the experimental design was used (table S3). Table S3 shows eighteen experiments designed by design expert to assess the effects of four variables including time, salt amount, temperature, and volume of the eluent. Results showed two variables of time and volume are important (p < 0.05) that the ranges should be optimized. Table S7 shows that the linear model based on two variables is significant and Fig. S1 shows the 3D diagram for the changes of response at various time and volumes. Summary of experimental design results for NAMB by RSM and CCD is depicted in Fig. 8.

Fig. 8
figure 8

Summary of experimental design results for NAMB by RSM and CCD

Full size image

Real sample analysis

HPLC method figures of merit and validation parameters are summarized in Tables 2 and 3. NAMB determination in spiked samples showed that the recovery of the method is between 91 to 115%. The concentration–time profile of NAMB (Fig. 9) showed a multi-exponential behavior as a result of the absorption, distribution, and elimination phases of NAMB. The terminal phase was characterized by an apparent terminal half-life of about 10 h.

Table 2 Figures of merit of the presented method for determination of NAMB
Full size table
Table 3 Precision and accuracy of the MSPE/ HPLC assay for the quantification of NAMB in human plasma (n = 6)
Full size table
Fig. 9
figure 9

Concentration–time profile of NAMB after oral administration of 250 mg of NAMB to healthy volunteers (mean ± SD, n = 3)

Full size image

Conclusion

In this study, Bentonite-β-cyclodextrin-iron nanoparticles (Ben-βCD-INP) composite was synthesized, characterized, and utilized as an innovative magnetic solid-phase extraction (MSPE) adsorbent to separate and preconcentrate NAMB. The present extraction method had major advantages such as specific affinity to the analyte, nontoxicity, use of cost-effective solvents, and acceptable recovery values. The method showed high efficiency in selectivity, sensitivity, and linearity. It was successfully used for PK analysis of NAMB in human plasma samples.