Determination of vitamins B1 and B6 in infant formula and food supplement samples using magnetic layered double hydroxide nanoadsorbent before liquid chromatography-tandem mass spectrometry

Abstract A magnetic Mg/Fe layered double hydroxide (magnetic LDH) nanocomposite was successfully prepared and used as a suitable adsorbent to simultaneously extract vitamins B1 and B6 from complex infant formula and food supplements. The important parameters affecting the extraction were optimized and the final parameters are as follows; the amount of adsorbent (20 mg), extraction time (6 min), elution time (8 min), and elution solvent ethanol in 750 µL. As-prepared nanoadsorbent was successfully in magnetic solid phase extraction set-up with no hazardous solvents and extraction performed in 6 min. The structure of nanoadsorbnet is lamellar, providing a wide surface to extract the analytes. The quantification of the analytes was accomplished using the liquid chromatography-tandem mass (LC/MS-MS) technique. Under optimum extraction conditions, the linearity ranged from 4 to 1000 ng/mL (vitamin B1) and 20 to 1000 ng/mL (vitamin B6), and the correlation of coefficients (R2) was obtained better than 0.99. The intra-day (n = 3) and inter-day precisions (n = 3 working days) calculated in the form of percent relative standard deviations (%RSDs) were obtained below 10.20%. The proposed method was successfully practiced for analyzing vitamins B1 and B6 in several brands of infant formulas and food supplements. Graphical Abstract


Introduction
Water-soluble vitamins primarily belong to the vitamin B complex group, essential food nutrients, especially in infant formulas. These vitamins have been engaged in metabolic pathways and show significant health benefits when added to our daily diet. Thiamin (vitamin B1) was structurally determined to be the first water-soluble vitamin and exists as unphosphorylated and as phosphorylated thiamin derivatives; thiamin-monophosphate (TMP), thiamin-diphosphate (TDP), and thiamin-triphosphate (TTP) [1] . Our body cannot synthesize this vitamin, and due to its solubility in water, it is easily and quickly excreted from the kidneys and its storage time in the body is very short. Therefore, according to these conditions, people should get thiamin regularly through their daily diet [2] . Furthermore, vitamin B1 deficiency in infancy affects health function and balance skills in childhood [3] . A new study shows that thiamin deficiency in infants severely affects the health of preschool children who were breastfed in the first year of life. Conclusions were based on a retrospective study of children who received a brand of Remedia formula, utterly devoid of vitamin B1, in 2004 [4] . Having enough of this vitamin in infancy is essential for brain development.
Pyridoxine (vitamin B6), a water-soluble vitamin, is an essential vitamin in metabolism and is added to foods and supplements. Vitamin B6, like vitamin B1, involves cellular metabolism, such as the metabolism of sugars, proteins, and fats. The formation of hemoglobin severely needs the use of vitamin B6. Vitamin B6 in food emerges as pyridoxamine, pyridoxal and pyridoxine, either free or bound to phosphate, proteins, or amino acids, is known as pyridoxine [5] . This vitamin plays a functional role in the nervous and the immune systems as a neurotransmitter responsible for understanding pain or depression. Vitamin B6 deficiency in infants causes growth retardation, including weight loss and neurological problems. The U.S. Infant Formula Act of 1980 recommends that infant formula be fortified with pyridoxine hydrochloride at a minimum of 35 g/100 kilocalories of the product (8.6 g/100 kJ) by the European Union (EU) minimum requirement of 9 g/100 kJ.
According to the Food and Drug Administration (FDA) and food and supplement manufacturers, there are daily nutritional requirements for these vitamins, so quantitatively assaying the vitamin amount in such products is necessary. However, simultaneous measurement of these two vitamins in food is complicated due to the complexity of the sample environment and the small concentration of vitamins [6,7] .
Numerous advanced devices have been used to measure vitamins B1 and B6 in food, dietary supplements, and pharmaceuticals, such as an HPLC device coupling with different detection systems [7][8][9][10][11][12] . For example, in using the HPLC-UV method, the method's sensitivity is not acceptable due to disturbing substances [13] . Moreover, HPLC-UV methods need ion-pairing reagents or a pre-column set-up. However, liquid chromatography-mass spectrometry methods seem good to accommodate vitamin B analytes' evaluation; unfortunately, it is still impossible to inject samples with complex matrices despite all these advanced and powerful devices. So an analyst should look for sample purification and analyte concentration. All of these practices are called sample preparation. Therefore, implementing an efficient extraction method coupled with advanced devices for material analysis can be one of the researchers' goals.
The preparation method for analysis of B group vitamins is dispersive liquid liquid microextarction (DLLME) in some reports [14,15] . The mentioned approach presents multiple drawbacks when practiced with complex food samples. Furthermore, extracting solvent requires additional processing of the solvent evaporation before introduction to analytical systems. Emulsion's difficulty, which hinders the full recovery of the extract, is another limitation observed in DLLME. For conventional solid phase extraction (SPE) methods used in a published work [16] , the cartridges should be conditioned with large volumes of the solvent before the extraction procedure. Several extraction steps with consumption of hazardous organic solvents were used to obtain optimum output, in which the recycling cost of these dangerous solvents was extremely high. Besides, the affecting factors, including the type, the volume, and the flow rate of the solvent, should be optimized, which was time-consuming and expensive. An appropriate adsorbent is synthesized instead of cartridges or disks used in conventional SPE methods. Despite the advantages of solid phase microextraction (SPME), none of its versions is practiced to extract the vitamin B group components.
As a promising sample pretreatment technique, magnetic solid phase extraction (MSPE) has recently gained considerable attention because it avoids the main limitations of other conventional pretreatment methods, such as consuming organic solvents [16] . The proposed method has the benefits of (i) fast equilibrium, (ii) low cost, (iii) a simple set-up handling, (iv) a reusable adsorbent and (v) an environmentfriendly sample preparation step. These characteristics present it as the first choice for the typical laboratories interested in the analysis of vitamin B group of vitamins. Using the magnetic separation process, the filtration and centrifugation steps will be eliminated and therefore the extraction process will be faster. Adsorbents used in magnetic separation must have superparamagnetic nature, high surface area, high dispersion, and small diffusion resistance. Recently, significant research has been done on the synthesis of new magnetic nanoparticles or the introduction of the properties of magnetic nano hybrids and nanocomposites [17][18][19] .
Nowadays, the interest in using nanostructured materials has increased a lot, and this is due to the unique properties of these materials, especially their high surface-to-volume issue, cost-effectiveness, and straightforward synthesis. One of these new nano adsorbents is layered double hydroxides (LDH) with the chemical formula [(M II ) 1Àx (M III ) x (OH) 2 ] xþ (A mÀ x/m )ÁnH 2 O], which M is representative of metal and A is an anion with m charge. LDHs are composed of octagonal layers of brucite stacked on top of each other, and the space between the layers is filled with water and anions where the charge of the structure is balanced with positively charged brucite-like layers [20] . The network of lamellar LDHs provides a high surface-to-volume ratio, allowing strong interactions with molecules. The interlayer area is very flexible and can host several kinds of compounds, so this property explains the interest in these materials as an extractor. In addition, the exchangeable interlayer anions, ease of synthesis, cost-effectiveness, and composite flexibility, makes LDHs a good candidate for use as an adsorbent in the extraction process [21,22] . By adding the magnetic or magnetizable material into the structure of LDH, the extraction performance does not need extra filtration or centrifuge set-up.
In this study, we synthesize LDH and magnetize it with Fe 3 O 4 to obtain magnetic LDH to extract vitamins B1 and B6 simultaneously. We selected the ideal conditions by optimizing all the effective parameters in the extraction. After preparing the sample, we injected it into the HPLCtandem MS system and observed the required response. In addition, examination of actual samples of infant formula and food supplements was also tested.

Instrumentation
For analysis of samples, a 2695 Waters HPLC instrument (Waters Milford, MA) coupled with a triple quadrupole tandem mass spectrometer (Waters, USA) was used. The target analytes were detected by adjusting in a positive multiple reaction monitoring (MRM) mode. The instrumental parameters were set as follows; spray voltage þ 3 kV, source temperature 300 C, desolvation temperature 100 C, capillary voltage, þ2 kV, cone voltage, þ25 V; extractor, þ3 V; and desolvation flow, 600 L/h. The MRM transitions are m/z 265.2 > 144.0 and m/z 265.2 > 122.0 for vitamin B1 and m/z 170 > 152 and m/z 170 > 134 for vitamin B6 [23] . In the case of vitamin B1, the precursor ion of 265.2 is related to M þ ion, as the vitamin has an intrinsic positive ion in its structure.
The auto-sampler tray and HPLC column temperatures are 4 C and room temperature, respectively. The samples were separated on a C18 column (100.0 Â 4.6 mm; particles size of 3.5 mm) (Agilent, USA) by injecting 20 mL of them with a column temperature of 30 C. The mobile phase consisted of water and methanol gradient. A gradient elution utilizing 0.1% FA in water as solvent A and 0.1% FA in methanol as solvent B was performed, having a varying flow rate, nonlinear gradient steps. The gradient was as follows: 0 min (97%A and 3%B, 0.50 mL/min, curve initial), 2 min (70%A and 30%B, 0.50 mL/min, curve 6), 4 min (3%A and 97%B, 0.80 mL/min, curve 11), 6 min (97%A and 3%B, 0.80 mL/min, curve 11) and 7 min (97%A and 3%B, 0.50 mL/min, curve 11).
FT-IR spectra were recorded on a Vector 22 (Bruker, Germany) Fourier transform infrared spectrometer in the 4000À400 cm À1 wavenumber range. XRD measurements were carried out on a Bruker D8 Advance (model AXS, Germany) X-ray powder diffractometer generated at 40 kV and 35 mA.

Chemicals and solutions
Standards of vitamins B1 and B6 were received as a gift from Zahravi pharmaceutical company (Tabriz, Iran All reagents used were of analytical reagent grade and all solutions were prepared with high quality and deionized water (Shahid Ghazi Co., Tabriz, Iran). For all solvents used as mobile phase, 0.1% formic acid was added. To protect against samples instability and degradation, all stock and working standards were kept in the refrigerator before being used; all prepared samples were analyzed within four hours. All operational standards and samples were filtered through 0.22-mm nylon filters.
Preparation of 1000 mg/mL stock solutions of vitamins B1 and B6 was completed in a 250 mL volumetric flask separately and shaken well until the standards were completely dissolved in methanol. Then diluent was added to fill the flask to the marked line. The level-6 calibrants were prepared serially by diluting the working standard. Calibrations were made in concentrations of 4-20-100-400-2000-10000 ng/mL. The details of calibration equations are in Table 1S.

Preparation of infant formula and food supplements
Five infant formula powders (labeled IF1 to IF5) were prepared by weighing out each powder's necessary amount: 1 g infant formula was dissolved in 5 mL water (20% w/v), and then 5 mL of ACN was added to precipitate the proteins. After centrifugation, the supernatant (6 mL) was collected, and the final volume was adjusted to 20 mL by adding distilled water with pH 4.5. The acidified media was used for protein crash/precipitation, which allows the analyst to eliminate the fatty/solid material via centrifugation [24] . Two mL of the supernatant was then poured carefully and functioned to magnetic SPME. Food supplements in the form of tablets were powdered thoroughly before being weighed and dissolved in methanol. Then the sample solution was centrifuged to separate the solid residue, and the upper phase was ready to apply magnetic SPME.

Synthesis of magnetic Mg-Fe-Cl LDH
Synthesis of Mg-Fe-Cl LDH Firstly, 10 mL of a solution containing 3 mmol MgCl 2 .6H 2 O and 1 mmol FeCl 3 .6H 2 O were added rapidly to a 40 mL of NaOH 0.15 M solution while the reaction mixture stirred vigorously almost for 10 min. The resultant mixture was centrifuged, and the sediment phase was washed with deionized water. This solid phase was dispersed into 40 mL deionized water and transferred into 45 mL Teflon-lined autoclave, where synthesis was performed hydrothermally for 16 h at 100 C [25] .

Preparation of magnetic LDH
To magnetize the as-prepared LDH, 2.36 g of FeCl 3 .6H 2 O and 0.86 g FeCl 2 .4H 2 O were put into a 2-neck round bottom glass flask and dissolved with 40 mL of deionized water ultrasonically. Then the reaction mixture was placed in an oil bath at 80 C for 30 min while stirring continuously. After that 5 mL ammonia solution (25%) was added to this mixture and was stirred for another 30 min. To prepare the magnetic LDH, 10 mL of these magnetic nanoparticles and 10 mL of as-prepared LDH were individually shaken using ultrasonic for 1 h. Then these two solutions were mixed and ultrasonicated for another 2 h. The sediment phase was discarded using an external magnetic field and dried overnight at room temperature under vacuum conditions [25] .

Magnetic solid phase microextration
The composition of supplements consists of different formulas and both lipophilic and hydrophilic compounds, therefore, the extraction of vitamins B1 and B6 by a selective solvent is impossible. In this project, a magnetic solid phase microextration (SPME) was applied to extract vitamins B1 and B6 from infant formula and food supplements, and the results of extraction were evaluated in the matter of recovery and matrix effect. Lipid and lipid-like co-existing materials were not bound to the magnetic nanoadsorbent, and pure and clean extraction was achieved. The experiments are as follows; first, 2.0 mL of each sample (prepared according to the methodology of Section "Preparation of infant formula and food supplements") was placed in a microtube. Next, 20 mg of the adsorbent was weighed and added to this microtube. The sample vial is placed in the ultrasound device for 6 minutes. After that, the magnetic LDH containing the analytes was separated using a magnetic field. Next, the 750 mL of desorption solvent (ethanol) is added to the solid phase. The sample is placed in a sonicator for another 8 minutes and then the upper phase is discarded using a magnetic field. The solution was injected into the HPLCtandem MS system.

Characterization of the synthesized nanoadsorbent
The Fourier-transform infrared spectroscopy (FT-IR) spectra of the bare Fe 3 O 4 , LDH, and Fe 3 O 4 /LDH hybrid are shown in Figure 1(a-c). IR spectrum of Fe 3 O 4 nanoparticles in Figure 1(a) shows that a sharp peak at 565 cm À1 and a broad peak at 3500 cm À1 are stretching vibration modes of Fe-O and Fe-O-H, respectively. The broad band appears around 3440 cm À1 in both spectra of LDH and magnetic LDH is related to OH stretching bond present in LDH structure and its interlayer water molecules. Peaks at 1634 and 1637 cm À1 are related to LDH and hybrid LDH, respectively, confirming the OH bending mode. The low-frequency region at 400-900 cm À1 is ascribed to M-O stretching and M-OH (M ¼ Mg and Fe) bending in LDH and hybrid LDH [26] .
X-ray powder diffraction (XRD) confirmed the identification of synthesized materials. The several sharp peaks that appeared in the XRD pattern of LDH (Figure 2a) are a sign of the crystalline nature of this material. Several peaks appear in Fe3O4 XRD pattern (Figure 2c) appeared at 2h region of 30-70˚that are assigned to the crystal plate at (220), (311), (400), (422), (511) and (440), respectively. Reflection peaks at (003) and (006) in 2h values of 11.14 and 22.28 , respectively, stressed the synthesis of LDH and are in agreement with the literature [27,28] . Sharp and narrow peaks in the lower 2h regions are prominent features of high crystalline lamellar compositions [29] .
The morphology of the hybrid structure was revealed using a scanning electron microscopy (SEM) graph. A regular hexagonal-shaped and nano-sized LDH image is shown in Figure 3. The plate-like and smooth surface of the prepared nano adsorbent is in accordance with the literature [30,31] .

Optimization of magnetic solid phase microextraction
According to our experiences, 6 factors affect analytes' adsorption (vitamins B1 and B6) by magnetic LDH, namely  the amount of adsorbent, time of extraction and elution, pH, kind and volume of extraction solvent. The experiments were done one by one while keeping the other parameters instant. The peak area of the analytes was used to estimate the optimum value of the parameter. Each sample was injected into LC/tandem-MS and determined in triplicate.
One of the important advantages of the solid phase microextraction method is the possibility of using the minimum amount of adsorbent (mg) because these nanoparticles' size is very small and provides a high surface-tovolume ratio for reaction and bonding. In this step, we considered the range of 3 to 25 mg and according to the results (Figure 4a), 20 mg of adsorbent is adequate. In dispersive solid phase extraction, the interaction between the analytes and the adsorbent is very important. The better the adsorbent dispersibility, the better it can load the analytes. As the amount of adsorbent increases up to 25 mg, this dispersion decreases and the adsorbent particles aggregate, and the extraction power decrease. After mixing the adsorbent and the sample solution, we must give it enough time to adsorb the analytes on the adsorbent well. In the proposed method, we use ultrasonic to improve this contact. Different time intervals, i.e., 2, 4, 6, 8, and 10 min, were tried. We observed that by increasing the extraction time to more than 6 minutes, the structure of nanoadsorbent is disturbed, and the absorption rate decreases, so we chose a time of 6 min as extraction time (Figure 4b).
The contact time of the elution solvent is also an important parameter in the extraction. In our selected intervals (Figure 4c), 8 min is the best time to wash the vitamins B1 and B6 from the adsorbent surface with acetonitrile. An elution solvent with sufficient power is needed to clean the analytes from adsorbent without any structural changes.
According to the articles and the structure of vitamins B1 and B6, we selected several solvents for testing, i.e., acetone, acetonitrile, ethanol, and methanol and among them, the best solvent was ethanol (Figure 4d).
The amount of desorption solvent should be large enough to collect all the adsorbed vitamins B1 and B6 from the adsorbent surface, so this is one of the most important optimizations. After desorption by four volumes of 250, 500 750 and 1000 mL of elution solvent, we conclude that the best result is obtained with 750 mL of ethanol (Figure 4e).
The pH of the solution determines the ionic state of the analytes in the sample solution. It also affects the ionization balance of some molecules and thus controls the solubility and extraction ability of analytes. The values of acid constants are pKa1 ¼ 4.8; pKa2 ¼ 9.2 (vitamin B1) and pKa1 ¼ 5.6; pKa2 ¼ 9.4 (vitamin B6) [32,33] . Depending on the pH, vitamin B1 could be a monovalent or divalent cation.  To achieve the appropriate pH to continue this step, the sample solution was adjusted in pHs at 4.5, 7.5, and 9.0 by 0.1 M HCL and 0.1 M NaOH. Solutions with pH lower than 4 were not practiced due to the probable desolvation of adsorbent in this pH. After extraction, the analytical signal was maximum when the pH of the sample solution was set at 4.5.

Statistical (regression) analysis
The validity of proposed magnetic solid phase microextraction coupled with LC-MS/MS was thoroughly investigated by linearity, recovery, the lower limit of quantification (LLOQ), as well as the inter-day and intra-day precision according to the United States Food and Drug Administration (FDA) guidelines [34] . LLOQ is the lower concentration at the calibration curve plotted by preparing the calibrators using proposed magnetic solid phase microextraction. Recovery was calculated as the ratio of the peak area of extracted analytes over the peak of samples in water. We carry out statistical (regression) analysis on the data to obtain the calibration function. The horizontal axis is defined as the x-axis and the vertical axis as the y-axis. When plotting data from a calibration experiment, the convention is to plot the instrument response data on the y-axis and the values for the standards on the x-axis. This is because the response is dependent to standards (predictive variables). The coefficient of determination (R 2 ) is one of the statistics commonly used in analytical measurement. R 2 is a measure of goodness of fit and often used to describe the fraction of the total variance in the data which is contributed by the line that has been fitted. Ideally, if there is a good linear relation, the majority of variability can be accounted for by the fitted line. R 2 should therefore be close to 1.
After plotting the data and fitting the regression model, the calibration data are judged to be satisfactory the calibration equation can be used to estimate the concentration of the analytes in real samples.
Precision and accuracy were evaluated by three concentrations (100, 400, and 1000 ng/mL) of analytes assayed on one day (inter-day) and three different days (intra-day) for both vitamin B1 and B6. The chromatogram of vitamins B1 and B6 is reported in Figure 1S.
The linearity of the method was assessed by plotting the peak areas versus concentrations curve in the range of 4-1000 ng/mL for vitamin B1 and 20-1000 ng/mL for vitamin B6 With coefficients of determination (R 2 value) better than 0.99. The LLOQ value extracted from the calibration curve was 4 and 20 ng/mL for vitamins B1 and B6, respectively.
The accuracy and precision (intra-day and inter-day) were studied by spiking vitamin B1 and B6 at three different levels into food supplements. The spiked samples were prepared according to the sample preparation and extracted under magnet solid phase microextraction method. All analyses were done in five times. The RSD and accuracy values were 7.2% and 10.2% for inter-and intra-day evaluations, respectively. The details are shown in Table 1.

The magnetic solid phase microextraction mechanism
The adsorption mechanism of vitamin B1 and B6 will be beneficial for understanding the extraction mechanism and   developing new technologies for food treatment. Vitamin B1 consists of an aminopyrimidine and a thiazolium ring connected by a methylene bridge. Vitamin B1 pKa of 9.2 is related to the quaternary N on the thiazole ring molecule and remains cationic over a wide pH range. Another pKa ($4.8) is due to the protonated pyrimidine [32] .
Vitamin B6 consists of 2-methyl-4-aminopyrimidine group attached via a methylene group to a thiazole ring functionalized with a methyl group in the position 4 and a hydroxyethyl group in position 5. Vitamin B6 has pKa$5.6 value for pyridinium N and pKa$9.4 value for its phenolic group [33] . On the other hand, vitamin B6 is an amphoteric molecule (pI ¼ 7.5) and can exist in cationic, anionic, and neutral forms. Vitamins B1 and B6 are both stable in acidic media. As extraction performance occurred at pH 4.5, vitamin B6 exists mainly in the cationic form at pH values lower than its isoelectric point. In pH ranges (we have illustrated the pKa ± 2 pH to determine the charge state) 7.6 to 3.6 and 7.4 to 11.4, the positive and negative net charges gradually increased, respectively. At very acidic or very basic pHs, the vitamins are not stable.
Removal of water molecules from the interlayer of LDH or the adsorption of analytes on the surface of LDH through hydrogen bonding are the mechanisms imagined for the adsorption capacity of magnetic hybrid LDH. Also, the lamellar nature of the prepared adsorbent has an important role in the analyte adsorption and provides adequate space for the small-sized vitamins B1 and B6 molecule's precipitation. The surface zeta potentials of magnetic LDH is negative and it becomes more negative with increasing the pH. Therefore more vitamins B1 and B6 adsorb through electrostatic interaction. When the repulsion forces between adsorbents are maximum, the aggregation and sedimentation are at their minimum level and the stacking of adsorbents is limited because of the coulombic repulsion [35] . Therefore, a decreased inter-particle electrostatic repulsion is a reason for the maximum adsorption. Moreover, Surface functional groups such as hydroxyl and carboxyl groups provide hydrogen bonding interaction with surface hydroxyl and amine groups of vitamins B1 and vitamin B6.

Real samples
The proposed method was evaluated on five infant formula brands available. Each sample was prepared as detailed in the experimental section and injected in triplicate. Table 2 displays the recovery results of each vitamin observed in the infant formula samples prepared. The three sample profiles appear to be similar, which was to be expected, bearing in mind that the label claims are usually identical for this type of product. We used the ±20% accuracy level interval for accepting the label amount. In sample IF#1, the label amount of both vitamin B1 and B6 are not matched with the label amount. In sample IF#2, the amount of vitamin B1 is several times more than the label amount; however, the vitamin B6 amount is confirmed. In all cases of tablet food supplements, vitamins B1 and B6 label amounts are at acceptable levels. The proposed analysis method is valid and can be performed for various samples with different matrices. For example, the complexity of the matrix of a sample should not limit the implementation of the method or affect the method's response, as reflected in recovery calculations. We calculated the relative recovery and repeated each real sample with two different spike concentrations.

Comparison of the proposed method with the reported ones
As the structure of these water-soluble vitamins is different, different approaches have been reported to measure them (Table 3). Due to the effects of the infant formula matrix and the low concentration of the analytes, cleaning the matrix, concentrating the analyte, and connecting it with simple quantification techniques is necessary. The traditional methods of sample preparation have several limitations. For example, the solid phase is fixed in the cartridge in the SPE method [40] , which is a very time-consuming and expensive method. After several uses, it is blocked or destroyed and needs to be replaced. Also, a large amount of solvent is used in liquid phase extraction, such as dispersive liquid-liquid extraction (DLLME) or cloud point extraction, separation of two phases is very difficult. Like SPE, the amount of hazardous solvent used in conventional sample preparation methods is high [14,15,39] . Moreover, using detection systems such as spectrometry hampered the accurate detection. In our proposed method, nano-adsorbents are used in only 20 mg. Due to their excellent dispersion properties and the surface-to-high volume ratio of adsorbent particles, separation, purification, and concentration of the sample are done in the best way; thus, we can minimize the side effects of the matrix. Furthermore, safe and nontoxic substances are used to synthesize the adsorbent free from any organic solvents, and the extraction process is performed in less than 20 min. The calibration curves were wide enough to cover the analytical requirements, and the instrumental device was accurate and precise LC/tandem MS. Disadvantages of the present method may be the lack of high concentration factors, which is not considered a disadvantage compared to similar works. Because in complex matrices, most of the reported methods have a low concentration factor and have not been reported. Due to the absence of vitamins studied in the present method being very low concentrations and therefore no need to concentrate analytes, a high concentration factor may not be considered an advantage. And in such complex matrices the ability to purify the reported extraction method can be an important advantage.

Conclusion
Due to the critical and essential role of vitamins, B1 and B6 in our body and the inability of the body's natural system to synthesize them, children, infants and the elderly should regularly meet our body's needs through nutrition or dietary supplements. It is more important to discuss the amount of these vitamins in certain groups of food supplements. Older adults should always have enough thiamin and pyridoxine to increase their immune system and many other things, and infants should always have these vitamins to develop their nervous system and motor system. One of the sources of vitamins B1 and B6 supply for infants can be their powdered milk, which according to the manufacturer, most of them are enriched with these vitamins. To confirm the claim of food supplement manufacturers, we have developed a new and efficient method for simultaneously measuring vitamin B1 and B6 with magnetic SPME coupled with LC/MS-MS. LDHs are used as extractants due to the ease of synthesis and having a high surface area for extraction. Considering the surface charge of LDHs and their OH groups, the combination of LDHs with other nanomaterials is feasible. To facilitate the extraction process without needing centrifuge or filtration, we magnetize the LDHs using Fe 3 O 4 . Positively charged LDHs and negatively charged Fe 3 O 4 form the stable nanocomposite via electrostatic interaction. Our proposed method can now be used in food control laboratories and regulatory bodies to measure vitamins B1 and B6 in food and pharmaceutical products.