HPLC‐Orbitrap‐MS for the Determination of B‐Vitamins in Fruit Juices and Food Supplements

TL;DR

This paper introduces a new method using HPLC-Orbitrap-MS to accurately measure B-vitamins in fruit juices and supplements with high sensitivity and precision.

Contribution

The study presents a validated HPLC-Orbitrap-MS method for simultaneous determination of seven B-vitamins in complex food matrices.

Findings

The method showed excellent linearity (R² > 0.998) and good accuracy (96%-112%) for supplements and near 100% for juices.

Matrix effects were mitigated using matrix calibration, and detection limits indicated high sensitivity.

Freshly squeezed juices had higher B-vitamin concentrations than commercial ones.

Abstract

Accurate assay of vitamins in foods is a considerable analytical challenge due to the chemical complexity of matrices and molecular structures. Orbitrap MS technology coupled with liquid chromatography through electrospray ionization source (HPLC‐ESI‐MS) was applied for the simultaneous determination of seven B‐vitamins (thiamin, riboflavin, niacin, pantothenic acid, pyridoxine, biotin, and folic acid) in fruit juices and dietary supplements. The method employed an easy sample treatment procedure, involving direct dilution for juices and a fast solvent extraction for supplements. Chromatographic separation was achieved by a reversed‐phase column with a gradient elution of water and acetonitrile. Mass spectrometry detection was performed in full‐scan mode and using both positive and negative ionization to maximize sensitivity. The method was validated, demonstrating excellent linearity (R
2 > 0.998), acceptable accuracy (96%–112% for supplements, near 100% for juices with matrix‐matched calibration), and good precision (RSD < 15%). The matrix effect was investigated and mitigated using a matrix calibration curve for fruit juices. Limits of detection and quantification were determined, indicating good sensitivity. The validated approach was successfully applied to the analysis of real samples, showing good agreement with label claims for supplements and confirming the expected vitamin levels in fruit juices, with freshly squeezed juices exhibiting higher concentrations than commercial ones.

Figures

Chemical structure of analytes.

LC‐HRMS chromatograms of B‐vitamins: extracted ion current of thiamin [M + H]+ 265.1118, rt 1.12 min; niacin [M + H]+ 124.0393, rt 1.60 min; pyridoxine [M + H]+ 170.0812, rt 2.73 min; pantothenic acid [M‐H]− 218.1034, rt 8.07 min; biotin [M‐H]− 243.0809, rt 10.34 min; folic acid [M‐H]− 440.1324, rt 11.34 min; riboflavin [M + H]+ 377.1456, rt 12.35 min.

Tables

TABLE 1. Exact mass of B‐vitamins analyzed.

Analyte	Ionization mode	Exact mass
Thiamin	Positive	265.1118
Pantothenic acid	Negative	218.1034
Folic acid	Negative	440.1324
Biotin	Negative	243.0809
Niacin	Positive	124.0393
Riboflavin	Positive	377.1456
Pyridoxine	Positive	170.0812
Labelled riboflavin (IS)	Positive	383.1530
Labelled thiamine (IS)	Positive	268.1218
Caffeine (IS)	Positive	195.0877

TABLE 2. Analytical parameters of accuracy, LOD, LOQ, and reproducibility (RSD*) for dietary supplements calculated in solution.

Dietary supplement
	Spiked (S1) 40 μg/L	Accuracy (%)	Spiked (S2) 150 μg/L	Accuracy (%)	Spiked (S3) 300 μg/L	Accuracy (%)	RSD % a	LOD (μg/L)	LOQ (μg/L)
Thiamin	41.7 ± 1.3	104	148.5 ± 2.3	99	292.3 ± 16.7	97	13.1	3.0	6.0
Pantothenic acid	38.4 ± 1.0	96	150.6 ± 2.9	100	325.3 ± 21.0	108	8.4	6.0	12.0
Folic acid	44.1 ± 0.1	110	158.8 ± 3.6	106	314.7 ± 15.8	105	9.3	6.0	12.5
Biotin	37.9 ± 2.1	95	152.7 ± 7.2	102	307.3 ± 11.2	102	12.6	12.0	20.4
Niacin	42.2 ± 3.5	106	148.6 ± 6.2	99	311.8 ± 14.3	104	8.4	3.0	6.2
Riboflavin	38.3 ± 1.1	96	161.7 ± 0.3	108	335.8 ± 5.5	112	11.9	12.0	15.0
Pyridoxine	43.5 ± 3.1	109	161.5 ± 7.2	108	290.8 ± 13.4	97	7.4	3.0	7.0

TABLE 3. Analytical parameters of accuracy, LOD, LOQ, and reproducibility (RSD%*) for fruit juices.

Fruit juices
	Spiked (JS1) 150 μg/L	Accuracy (%)	RSD % a	Spiked (JS2) 300 μg/L	Accuracy (%)	RSD % a	LOD (μg/L)	LOQ (μg/L)
Thiamin	144.2 ± 3.9	96	9.4	301.2 ± 8.5	100	8.2	12.0	20.0
Pantothenic acid	144.6 ± 6.2	96	7.6	302.6 ± 8.5	101	4.3	6.0	9.8
Folic acid	153.4 ± 7.3	102	11.4	296.3 ± 6.7	99	5.7	12.0	20.1
Biotin	143.3 ± 9.5	96	10.9	301.5 ± 7.8	100	9.8	6.0	10.2
Niacin	141.0 ± 2.2	94	5.8	303.2 ± 10.5	101	3.5	6.0	11.0
Riboflavin	154.2 ± 4.7	103	9.4	298.93 ± 9.2	100	4.6	12.0	19.8
Pyridoxine	136.4 ± 6.2	91	4.5	280.6 ± 7.3	94	2.6	12.0	20.3

TABLE 4. Amount of B‐vitamins found in real samples.

Sample	Thiamin		Pantothenic acid		Biotin		Niacin		Riboflavin		Pyridoxin		Folic acid
Dietary supplements	Amount (mg/g)	RSD%	Amount (mg/g)	RSD%	Amount (mg/g)	RSD%	Amount (mg/g)	RSD%	Amount (mg/g)	RSD%	Amount (mg/g)	RSD%	Amount (mg/g)	RSD%
DS 1	1.32 ± 0.02	1.5	—	—	0.21 ± 0.01	4.7	18.3 ± 2.2	12.0	24.7 ± 1.8	7.3	3.26 ± 0.25	7.7	0.10 ± 0.01	10
DS 2	0.40 ± 0.01	2.5	1.74 ± 0.07	4.0	—	—	—	—	2.4 ± 0.2	8.3	0.54 ± 0.02	3.7	—	—
DS 3	0.83 ± 0.01	1.2	4.10 ± 0.05	1.2	0.042 ± 0.007	16	8.8 ± 1.2	13.6	—	—	5.58 ± 0.60	10.7	0.041 ± 0.001	2.4

Full text

Introduction

1

Currently, a rising number of people are embracing nutrient‐rich diets to promote health and prevent disease, recognizing the vital importance of nutrition for general well‐being. Among the essential nutrients, the intake of vitamins is particularly crucial for human body. Those of the B group are water‐soluble compounds and contribute to the healthy and vital maintenance of the human organism, playing a key role in energy, lipid, glucose, and protein metabolism, and especially support the immunity system [1, 2, 3]. B‐vitamins have different chemical structures and include the following molecules: thiamin (B1), riboflavin (B2), niacin (B3), pantothenic acid (B5), pyridoxine (B6), biotin (B7), and folic acid (B9), and cyanocobalamin (B12), which, although they have specific functions, perform interconnected biological activities (Figure 1). The deficiency of these micronutrients can be manifested by asthenia, nervous, and dermatological disorders [4, 5, 6]. They are not synthesized by the human body, so they must be taken daily through food. A balanced and varied diet, rich in wholegrains, legumes, fruit, and vegetables, is generally sufficient to prevent their deficiency and promote optimal functioning of the body, but in some cases, it may be necessary to resort to dietary supplements for their more correct intake. The latter have become increasingly popular, and many people use them to improve their diet.

Chemical structure of analytes.

Food and beverage manufacturers are increasingly interested in emphasizing the nutritional value of their products, particularly their nutraceuticals content, as this information on packaging attracts health‐conscious consumers. As with many other nutrients, the health properties of vitamins are also confirmed by the European Union's Regulation 432/2012 [7], which authorizes the labelling of products with various health claims only if they contain a specific amount of the vitamins in question. It is essential to emphasize that nutrition and health claims are subject to strict rules designed to prevent misleading information for consumers; regarding vitamins, European Regulations 1169/2011 and 90/496/EEC establish the criteria for labelling, defining the minimum levels and presentation modality [8, 9].

In this context, analytical checking of food and supplement labels is essential for protecting consumer health. For B‐vitamins, accurate and reliable analytical methods are crucial due to their chemical stability and the complexity of food matrices. This approach is the only guarantee for precise nutritional information and fraud prevention. Consequently, the exact quantification of B‐vitamins in complex matrices (food supplements, fruit juices) presents several analytical challenges.

Generally, the assay of B‐vitamins is commonly performed using liquid chromatography with UV/diode array, fluorescence, electrochemical, and mass spectrometry detectors [10, 11, 12, 13, 14, 15, 16]. Mass spectrometry, due to its high specificity and selectivity, certainly represents the ideal analytical tool for the identification, structural characterization, and quantitative determination of a wide range of compounds in complex matrices such as food and derivatives [17, 18, 19].

Among all the different mass spectrometry‐based working modes, tandem mass spectrometry (MS/MS) experiments, especially those performed using a triple quadrupole in multiple reaction monitoring (MRM) mode, are the most widely employed for the accurate quantification of small molecules in food matrices [20, 21, 22, 23, 24]. The high selectivity of this technique allows for precise results even at low concentrations, minimizing interference. However, when analyzing complex matrices, high‐resolution mass spectrometry becomes essential. In fact, while MS/MS with MRM mode excels in quantification, high‐resolution techniques are particularly valuable for identifying and understanding matrix effects caused by interferents, as well as for providing a more detailed characterization of sample composition [25, 26, 27, 28].

Our work focused on the investigation of a highly accurate analytical method for the simultaneous assay of seven B‐vitamins (B1, B2, B3, B5, B6, B7, and B9) in fruit juices and dietary supplements. This approach takes advantage of high‐performance liquid chromatography (HPLC) coupled with high‐resolution Orbitrap mass spectrometry, enabling more accurate and reliable quantification of these essential vitamins.

Materials and Methods

2

Chemicals

2.1

Pure analytes of vitamins, thiamine, pantothenic acid, biotin, riboflavin, folic acid, pyridoxine, niacin, and caffeine, (−)‐riboflavin‐^13^C_4_,^15^N_2_, and thiamin‐^13^C_3_ were purchased from Sigma‐Aldrich (St. Louis, Missouri, United States). HPLC‐grade solvents and reagents, including methanol, ammonium acetate, formic acid, tetrahydrofuran (THF), and acetic acid, were also obtained from Sigma‐Aldrich.

Standard Solutions

2.2

Stock solutions of 200 mg/L were prepared for each standard compound in ultrapure water. Only biotin (B7) and folic acid (B9) were dissolved in a 1% ammonium acetate solution, while riboflavin (B2) was solubilized in a 5% aqueous solution of acetonitrile. Each standard solution was diluted to 20 mg/L using ultrapure water, and these were used as working solutions for the preparation of the calibration curves.

Real Samples

2.3

Eight real samples, representing products commonly available on the market, were examined. In detail, three multivitamin supplements and five fruit juices were selected, two of which were cold‐pressed and three purchased from commercial distributors. All samples came from local supermarkets.

Sample Preparation

2.4

Dietary supplement samples, previously pulverized, homogenized, and stored at 4 °C, were subjected to an extraction procedure, according to a slightly modified method already published [29]. Briefly, 100 mg of each sample was treated with 5 mL of a 5% aqueous tetrahydrofuran (THF) solution for 10 min using a vortex mixer. Eight milliliters of a 1% aqueous acetic acid solution was then added, and the mixture was stirred for a further 10 min. After centrifugation at 6000 rpm for 10 min, the supernatant was filtered through a PTFE 0.45 μm filter, the internal standards were added, and the sample was diluted 10 times with a solution composed of acidic water (0.1% formic acid) and methanol 70:30 (v/v).

All juice samples were subjected to centrifugation at 12000 rpm for 4 min to separate the liquid from the solid components. Subsequently, they were filtered through 0.45 μm PTFE syringe filters to remove any residual particles. The filtered samples were diluted with a mixture of acidified water and methanol (70:30 v/v) at varying ratios.

HPLC‐Orbitrap‐MS Analysis

2.5

The detection and quantification of vitamins were performed with an Orbitrap Exploris 120 (Thermo Fisher Scientific, San José, California, United States). The instrument was composed of a heated electrospray ionization (H‐ESI II) probe. The mass spectrometer was coupled with a Vanquish system consisting of HPLC pump and autosampler (Thermo Fisher Scientific, San José, California, United States). The chromatographic separation was achieved employing a C_18_ reversed‐phase column (Hypersil GOLD, 2.6 μm, 100 × 2.1 mm, Thermo Fisher Scientific, San José, California, United States). The sample injection volume was 2 μL, while the flow rate was set at 0.3 mL/min, using as elution solvents 0.1% HCOOH in water (solvent A) and acetonitrile (solvent B) under gradient conditions. The gradients steps were the following: 5% B in isocratic for 1.5 min, from 5% to 40% B (1.5–4.0 min), 40% B in isocratic for 2 min, from 40% to 90% B (4.0–8.0 min), 90% B in isocratic for 4 min, from 90% to 5% B (12.0–15.0 min), and then an isocratic flow for 5 min to equilibrate the system before starting a new analysis. The total run time was 20 min. The MS system using the following conditions: spray voltage: 3.5 kV, (−4.0 kV in negative mode); sheath gas, aux gas, and sweep gas 20, 10, 1 a.u., respectively; ion transfer tube temperature 300 °C; vaporizer temperature 280 °C; the scan range was set in the range 100–1000 m/z, while the RF lens was set to 70% of the maximum value and the orbitrap resolution was set 60 000. MS experiments were performed in full scan modality monitoring the ion current of the exact masses of protonated and deprotonated molecules.

Matrix Effect

2.6

The matrix effect was assessed by using the following equation: ME (%) = (B/A) * 100. In the equation, A is the peak area of the analyte obtained from the analysis of a standard solution, and B is the analyte peak in a blank matrix [30].

Results and Discussion

3

LC‐HRMS Experiment and Sample Preparation

3.1

The analytical method proposed in this study was developed specifically to quantify B‐vitamins naturally found in fruits and in significant amounts in dietary supplements. Given the complexity of the analyzed matrices, the concentrations and the chemical stability of these vitamins, we opted to use high‐resolution mass spectrometry coupled with liquid chromatography (HPLC‐HRMS). This instrumental combination ensures high sensitivity and specificity in the analysis, allowing us to accurately identify and quantify these molecules. Sample preparation is found to be very simple, especially regarding fruit juices. In fact, in this case, the vitamins were quantified directly in the juices without subjecting samples to extraction and/or enrichment procedures, while in the case of supplements, they were subjected to a simple solvent extraction procedure in accordance with a method already available in the literature which reports quantitative recovery of the investigated analytes [30].

Method development has involved a sequential optimization process. Initially, the most appropriate ionization mode for each vitamin was selected in order to obtain the highest intensity signal; then, the chromatographic conditions were optimized to achieve good separation of the compounds.

The MS experiments were carried out in both positive and negative modes to maximize the ionization efficiency of the vitamins based on their chemical structure, and therefore to obtain high sensibility. Table 1 shows the ions monitored during the analysis.

The analytical determinations were performed in full scan mode, monitoring the extracted ion current (EIC) for each vitamin under investigation (Figure 2). For the quantitative assay performed using the internal standard calibration method, labelled riboflavin, labelled thiamine, and caffeine were used. In detail, the first standard was useful for determining riboflavin, pantothenic acid, and folic acid, the second for biotin and thiamine determination, and caffeine was used for pyridoxine and niacin quantification. Chromatographic separation was necessary to minimize the matrix effect and help achieve good specificity values. The separation was achieved using a C18 reversed‐phase and a gradient elution of water and acetonitrile.

LC‐HRMS chromatograms of B‐vitamins: extracted ion current of thiamin [M + H]+ 265.1118, rt 1.12 min; niacin [M + H]+ 124.0393, rt 1.60 min; pyridoxine [M + H]+ 170.0812, rt 2.73 min; pantothenic acid [M‐H]− 218.1034, rt 8.07 min; biotin [M‐H]− 243.0809, rt 10.34 min; folic acid [M‐H]− 440.1324, rt 11.34 min; riboflavin [M + H]+ 377.1456, rt 12.35 min.

Method Validation

3.2

A full validation of the developed analytical approach was performed, encompassing linearity, accuracy, matrix effects, limits of detection (LODs), limits of quantification (LOQs), and both intra‐ and inter‐day precision.

Linearity, Accuracy, and Matrix Effect

3.3

The linearity of the instrumental response was evaluated by generating calibration curves for each analyte. Five standard solutions, with concentrations ranging from 25 to 400 μg/L, were analyzed using both external and internal calibration methods. As already specified, for internal calibration curves, three different internal standards were used at a fixed concentration of 100 μg/L. Excellent linearity was observed for both approaches, as demonstrated by correlation coefficients greater than 0.998 in all cases.

After the linearity evaluation, analytical parameters were determined to assess the method's reliability using spiked samples for both matrices, fruit juice and dietary supplement.

Accuracy was evaluated by preparing three fortified blank samples for each matrix at the following concentrations for all vitamins: 40 μg/L (S1), 150 μg/L (S2), and 300 μg/L (S3), in order to cover the whole calibration range. Spiked fruit juice samples were prepared using a cola beverage, while those for food supplements were prepared using a mineral‐salt‐only supplement.

For food supplements, the calculated accuracy was acceptable (96%–112%), indicating negligible matrix effects (Table 2). Therefore, standard calibration curves prepared in solution with internal standards were considered suitable for the analysis of vitamins in supplements.

On the other hand, the quantification of vitamins in the fortified cola samples revealed a significant overestimation for some vitamins and an underestimation for others, with accuracy values above 120% in the former and below 80% in the latter. This phenomenon is attributable to a strong matrix effect. In view of the latter results, the matrix effect was calculated for the juice samples as specified in the experimental section, and a matrix‐matched calibration curve was implemented using an external calibration. Cola drink diluted 10‐fold in water was used as a blank matrix and calibration curves were constructed by adding known amounts of analytes, in the range of 25 to 400 μg/L, without the use of internal standards. This last choice was made because linearity was found to be quite good even in the case of external calibration, and this can be an added value for the methodology in terms of saving time and reagents. The instrumental response was linear also using matrix solutions, obtaining a correlation coefficient (R ^2^) around 0.98 for each analyte. In order to evaluate the accuracy, which in this case matches with the reproducibility of the calibration curves, two fortified samples were prepared at 150 μg/L (JS1) and 300 μg/L (JS2). The evaluation of the concentration in these samples was correct as the accuracy was near 100% for all the investigated vitamins. The results are shown in Table 3.

Limits of Detection (LODs) and Limits of Quantification (LOQs)

3.4

To assess the sensitivity of the analytical method, the limit of quantification (LOQ) and the limit of detection (LOD) were determined. The LOQ represents the lowest concentration of analyte that can be quantified with an accuracy value between 80% and 120%, while the LOD indicates the lowest detectable concentration but not quantifiable with precision. For these determinations, fortified samples (supplements and juices) were prepared at decreasing concentrations and analyzed using the calibration curves described above. The spiked samples were prepared at the following concentrations: 20, 12, 6, 3, 1.5, 1 μg/L. The LOQ and LOD values for fruit juices, calculated in the matrix, are reported in Table 3, while those for supplements are in Table 2. In both cases, the LOQ values are lower than 25 μg/L, which is the minimum of the calibration curves used, confirming the good sensitivity of the proposed method.

Precision (Repeatability and Reproducibility)

3.5

The precision of the methodology was assessed in terms of repeatability and reproducibility of the analyses. The repeatability was calculated as relative standard deviation by subjecting each real sample and each fortified sample to six measurements (intra‐day precision). Specifically, the results are presented in Table 4 for real samples, and in all cases, the RSD% values were lower than 15%, revealing a good repeatability of analyses. The reproducibility was also expressed as relative standard deviation (RSD%*) and obtained by analyzing each fortified sample six times over a period of 1 week (inter‐day precision). Once again, the results were below 15% and are shown in Tables 2 and 3.

Method Application to Real Samples

3.6

Finally, after evaluating the analytical parameters, real samples of dietary supplements and fruit juices were analyzed, and the results are shown in Table 4.

The measured B‐vitamin content in dietary supplements closely matched label claims, further validating the analytical method. Regarding fruit juices, those obtained by squeezing are an excellent source of vitamins, with significantly higher levels compared to commercial juices. Analyses have shown high concentrations of micronutrients studied, particularly in squeezed bergamot and citron juices, where the total amount of vitamins reaches almost 2 mg/L in the former case and almost 4 mg/L in the latter. Consistent with existing literature, folic acid was not detected in any of the citrus juices analyzed.

Conclusion

4

A novel LC‐HRMS method has been developed for the quantitative determination of B‐vitamins. High resolution and chromatographic separation ensured method specificity. Key analytical parameters, including accuracy, reproducibility, limits of quantification (LOQs), and limits of detection (LODs), have demonstrated satisfactory performance for most analytes, confirming the trueness of the approach. Matrix effects observed in fruit juice analysis were effectively mitigated using matrix‐matched calibration. Notably, dietary supplement analysis was free from matrix effects, and internal standards were employed to ensure accuracy and precision. This versatile method enables rapid screening of water‐soluble vitamins in several matrices.

Author Contributions

Conceptualization: L.B., F.M., L.D.D.; methodology: L.B., F.M., L.D.D.; validation: L.B. and F.M.; formal analysis: L.B. and I.S.; investigation: L.B.; data curation: L.B. and F.M.; writing—original draft preparation: L.B. and F.M.; writing—review and editing, L.B. and F.M.; supervision: L.B., F.M., L.D.D.; project administration: L.B. and L.D.D.; funding acquisition: L.B. and L.D.D. All authors have read and agreed to the published version of the manuscript.

Conflicts of Interest

The authors declare no conflicts of interest.

Funding

The authors have nothing to report.