New and feasible method for total phosphorus and sulfur determination in dietary supplements by ion chromatography

2.1 Instrumentation

A conventional oven (400/2ND, DeLeo, Brazil) was used for sample drying. The sample masses were measured using an analytical balance (AY220, Marte, Brazil), with a resolution of 0.0001 g and a maximum load of 220 g. Digestion procedures were performed using a microwave system (Multiwave 3000, Anton Paar, Austria) equipped with eight high pressure quartz vessels (80 mL). The maximum operational temperature and pressure were set at 280 °C and 80 bar, respectively, which were real-time monitored during the digestion procedures using sensors available in the microwave oven. Conventional laboratory materials were cleaned by immersion in 10% (v/v) HNO₃ (Merck, Germany) for 24 h. The quartz vessels used for sample digestion were cleaned in a microwave oven (Multiwave 3000, Anton Paar) with 6 mL of concentrated HNO₃ (Merck) and the microwave irradiation program was applied: (i) 900 W for 10 min, and (ii) 0 W for 20 min (cooling step). Finally, all materials were washed with ultrapure water, and dried before use in a class 100 laminar flow bench (CSLH-12, Veco, Brazil).

Phosphorus and S were determined as phosphate (PO₄³⁻) and sulfate (SO₄²⁻), respectively, using an ion chromatograph (861 Advanced Compact IC, Metrohm, Switzerland) equipped with a pump (IC liquid handling unit), a conductivity detector and a chemical suppressor module to reduce the conductivity of the mobile phase. To evaluate the digestion efficiency and comparison of results, an inductively coupled plasma optical emission spectrometer (Spectro Ciros CCD, Spectro Analytical Instruments, Germany) with an axial view configuration was also used, determining C (used for DOC determination), P and S. The IC and ICP-OES operational conditions are shown in Table 1.

2.2 Standards and chemicals

All solutions and sample dilutions were prepared using ultrapure water (resistivity of 18.2 MΩ cm) obtained from a purification system (MegaUP, MegaPurity, South Korea). Nitric acid (65%, Merck) was previously distillated below its boiling temperature, using a sub-boiling system (Duopur, Milestone, Italy). Oxygen (99.6%, White Martins, Brazil) was used in vessel pressurization. A standard solution containing 15000 mg L⁻¹ of P and 8000 mg L⁻¹ of S was prepared by dissolving salts of KH₂PO₄ (Merck) and Na₂SO₄ (Merck) and it was used to perform the recovery tests.

Another standard solution (1000 mg L⁻¹) of PO₄³⁻ and SO₄²⁻ was prepared by dissolving KH₂PO₄ (Merck) and Na₂SO₄ (Merck) salts in ultrapure water, and it was used to prepare the calibration curves for IC analysis (1–10 mg L⁻¹ of PO₄³⁻ and SO₄²⁻, in water). The mobile phase used for analytes determination by IC [Na₂CO₃ (3.2 mmol L⁻¹)/NaHCO₃ (1.0 mmol L⁻¹)] was prepared by dissolution of Na₂CO₃ (Merck) and NaHCO₃ (Merck) in water. The mobile phase used in the IC analysis was selected according to the recommendations of the ion chromatograph manufacturer (Metrohm, 2002) as well as in studies reported in the literature (Novo et al., 2018a, b; Pereira et al., 2016). The suppression system was regenerated with 200 mmol L⁻¹ H₂SO₄ (Synth, Brazil) and ultrapure water.

Standard solutions for P and S determination by ICP-OES were prepared before its use by serial dilution of a 10 mg L⁻¹ solution (Spex CertiPrep, Metuchen, USA). The calibration curves were prepared by serial dilution in 5% (v/v) HNO₃ (Merck), and it ranged from 25 to 1000 µg L⁻¹ for P and 250 to 10000 µg L⁻¹ for S. For determination of DOC, the calibration curve (25–100 mg L⁻¹ of C) was prepared by the dissolution of citric acid (Merck) in 5% (v/v) HNO₃ (Merck). In addition, an yttrium solution (1 mg L⁻¹, SpexCertPrep, USA) was used as internal standard for C determination. Argon (99.996%, White Martins) was used in ICP-OES determinations for plasma generation, nebulization, as auxiliary gas, and to purge the digests. The plasma operating conditions, selected emission lines and other parameters used in the determination of C, P and S by ICP-OES were based on previous works (Bizzi et al., 2011; Hartwig et al., 2017; Mesko et al., 2015; Novo et al., 2018a). For analysis of RA, acid-base titration was performed with 0.1 mol L⁻¹ KOH (Merck). All titrations were carried out in triplicate.

2.3 Samples and certified reference material

Five dietary supplements from different brands, including hypercaloric supplements (labeled as samples “A to C”) and energy supplements for gaining muscle mass (labeled as samples “D” and “E”), were purchased in a local market (Pelotas city, RS, Brazil). Powdered samples were used and the values informed by the manufacturers for fat content ranged from 1 to 5%, protein content ranged from 11 to 22%, and carbohydrate content ranged from 68 to 86%. Prior to the sample preparation, samples were dried at 65 °C. Sample “A” was arbitrarily chosen for optimization of the proposed digestion procedure while the other samples were analyzed using the selected optimized conditions. Due to the absence of a CRM with a matrix composition similar to the samples, the accuracy of the proposed method was also evaluated using the CRM BCR 060 (aquatic plant) from Institute for Reference Materials and Measurements (IRMM) mixed with the sample as a type of recovery test. Recovery tests by the addition of standard solutions containing P and S were also performed in the samples prior to the digestion methods.

2.4 Microwave-assisted digestion methods

For sample digestion under oxygen pressure, sample masses ranging from 850 to 1450 mg were transferred to the quartz vessels and 6 mL of HNO₃ solutions (1, 3, 5, 7 or 14.4 mol L⁻¹) were evaluated. After closing and capping the rotor, vessels were pressurized with 5, 10, 15 or 20 bar of oxygen, using a valve especially designed for this purpose. A conventional microwave-assisted digestion method was performed for comparative purposes and the same procedure already described was followed, but without oxygen pressure. The rotor was placed inside the oven, and the following microwave irradiation program was applied: (i) ramp time of 5 min to attain 1000 W, (ii) 1000 W for 10 min, and (iii) 0 W for 20 min (cooling step). After the cooling time, the pressure of each vessel was carefully released and the digests were transferred to volumetric vessels (25 mL) for subsequent P and S determination by IC and ICP-OES. It is important to mention that all the obtained solutions were filtered (a dialysis cell with cellulose triacetate membrane and pore size lower than 0.25 μm) and diluted prior to the analysis by IC and ICP-OES. The efficiency of both digestion methods was evaluated by DOC determination in the digests using the ICP-OES. The sample aliquots were purged with argon (0.1 L min⁻¹) for 2 min prior the determination in order to remove the dissolved CO₂ from the digests. Moreover, the RA in the digests was also evaluated in order to know the amount of acid that was consumed in the oxidation reaction.

All the results were statistically evaluated by Student's t-test (confidence level of 95%, p > 0.05) using GraphPad InStat version 3.00 computer software package (GraphPad, San Diego, USA). The LODs were calculated from the mean of the blank values plus three times the standard deviation of ten replicates. It was also taken into account the sample mass, the final volume of digests, and the dilution factor.

3.1 Conventional microwave-assisted digestion method in closed vessels

Initially, the studies related to dietary supplement were performed using 850 mg of sample to evaluate the minimum HNO₃ concentration for an efficient sample digestion without the need for oxygen pressure. This sample mass was used based on a method recommended by equipment manufacturers for samples mainly composed of protein and carbohydrate (Anton Paar GmbH, 2011). The maximum operational pressure was constantly monitored during the digestion procedures.

Regarding the aspect of the digests, a deep brown color was observed when 1 or 3 mol L⁻¹ HNO₃ was used as digest solution. Under these conditions, the DOC was not determined to prevent damage to the equipment. An improvement in the digestion efficiency was observed using 5 mol L⁻¹ HNO₃. However, the final digests presented a dark-yellow color, and the DOC was around 1300 mg L⁻¹. In turn, the digests presented a clear aspect when 7 or 14.4 mol L⁻¹ HNO₃ were used, and the DOC under these conditions was lower than 320 mg L⁻¹. Thus, 7 mol L⁻¹ HNO₃ was the minimum concentration that enabled an efficient digestion of 850 mg of supplement. Under these conditions, the DOC in digests was 320 ± 42 mg L⁻¹ and the RA was around 2.17 mol L⁻¹ (31%). An additional study was performed to evaluate the maximum sample mass that might be efficiently digested using 7 mol L⁻¹ HNO₃, however, even considering the high digestion efficiency and the relatively high RA previously obtained, an inefficient digestion was observed when increments of 100 mg of sample mass were performed. The DOC in digests was around 1600 mg L⁻¹, and digests presented a dark-yellow color, which was considered non-appropriate for further P and S determination.

After this prior evaluation, it was possible to observe that conventional microwave-assisted digestion (without oxygen pressure) was enough to digest up to 850 mg of dietary supplement with 7 mol L⁻¹ HNO₃ as the digest solution. Despite the high digestion efficiency obtained (DOC = 320 mg L⁻¹), the final solutions presented a high RA (2.17 mol L⁻¹), which should be unsuitable for analysis by IC in view of the high nitrate content in the final solution, and a subsequent dilution step to reduce this interference and prevent damage to the column may be necessary (Kaiser et al., 1999; Slingsby and Kiser, 2001). Therefore, a careful evaluation for P and S determination by IC should be performed, and the oxygen was evaluated as an aid for reducing the acid concentration required for sample digestion.

3.2 Microwave-assisted digestion method under oxygen pressure

Digestion efficiency of different HNO₃ concentrations (1 up to 14.4 mol L⁻¹) under oxygen pressure (5, 10, 15 or 20 bar) was evaluated. Sample masses of 850 mg were also used and the digestion efficiency was performed by determination of DOC and RA in the digests (Fig. 1).

Close

Fig. 1 Dissolved organic carbon determined by ICP-OES in 850 mg of dietary supplements digested by microwave-assisted digestion using 3, 5, 7 or 14.4 mol L⁻¹ HNO₃ under ■ 5, 10, 15 and □ 20 bar of oxygen pressure. Lines represent the obtained residual acidity after digestion under 5 bar (red line with white square) or 20 bar (black line with white circle) of oxygen pressure. Errors bars represent the standard deviation (n = 3).

Export to PPT

An efficient digestion using 1 mol L⁻¹ HNO₃ under different oxygen pressures (from 5 to 20 bar) was not obtained. The final digests presented a dark color with suspended solid particles in the final solution, and the DOC was not determined to prevent damage to the equipment. Under these conditions, the RA in final digests was about 0.02 mol L⁻¹ (2%), corresponding to a consumption of 98% of the acid added, which might explain the low digestion efficiency. Using 3 mol L⁻¹ HNO₃ under 5 bar of oxygen pressure, the digests presented a slightly brown color with solid residues remaining as suspended particles (DOC was around 6847 mg L⁻¹) and the RA was 0.15 mol L⁻¹ (5%). At the same acid concentration, an improvement in the digestion efficiency under 10, 15 and 20 bar of oxygen pressure was observed (DOC lower than 3875 mg L⁻¹). However, a yellow color was observed in the digests and the RA was around 0.22 mol L⁻¹ (7%). Under these conditions, the consumption of the acid ranged from 92 to 95%. Even using high pressure of oxygen as an aid for digestion, it is possible to infer that the amount of acid was not enough for complete oxidation of organic matter.

A clear solution with low DOC (<392 mg L⁻¹) was obtained using 5 mol L⁻¹ HNO₃ under different oxygen pressures (5, 10, 15 or 20 bar). In these conditions, the RA ranged from 0.85 to 0.95 mol L⁻¹ (about 17 to 19% of the HNO₃ added). The suitable digestion efficiency achieved using 5 mol L⁻¹ HNO₃ was not observed for the procedure without oxygen pressure, which is related to the absence of reaction that leads to acid regeneration. As reported in previous studies (Bizzi et al., 2011, 2010), the oxidant action of HNO₃ may be improved if a regenerating process occurs, which depends mainly on the amount of oxygen available in the gas phase during the oxidation of the organic matter. Thus, using 5 mol L⁻¹ HNO₃ under 5 bar of oxygen pressure was enough to digest 850 mg of dietary supplement. Dissolved organic carbon was 350 ± 48 mg L⁻¹ and the RA in the digests was 0.85 mol L⁻¹. Under these conditions, no significant difference in the digestion efficiency was obtained using more concentrated HNO₃. Similarly, no improvements in the digestion efficiency were observed by using higher oxygen pressure. Thus, digestion using 5 mol L⁻¹ HNO₃ under 5 bar of oxygen pressure was selected for further evaluations.

3.3 Evaluation of the maximum sample mass to be digested using diluted HNO₃ and oxygen pressure

An additional study was performed to evaluate the maximum sample mass to be digested by the proposed method using the selected conditions of acid concentration and oxygen pressure (5 mol L⁻¹ HNO₃ under 5 bar). Subsequent sample mass increments of 100 mg were performed, and an efficient digestion was observed up to 1350 mg of dietary supplement. Dissolved organic carbon in the digests was around 560 ± 48 mg L⁻¹ and the RA was 0.25 mol L⁻¹ (5%, corresponding to a consumption of 95% of the acid added). On the other hand, an inefficient digestion was observed using a sample mass of 1450 mg or higher, which presented a dark-yellow color and the DOC was around 1260 mg L⁻¹. Thus, it was possible to observe that when the vessels were pressurized with oxygen, a small amount of acid was sufficient to increase the sample mass digested, whereas in the conventional digestion method it was not possible. In true, it was possible to digest efficiently up to 60% higher sample mass using HNO₃ around 30% more diluted. This improvement in the digestion efficiency occurs due to the HNO₃ regeneration process under oxygen pressure, which remains active while there is sufficient oxygen available in the gas phase during the organic matter oxidation.

An efficient digestion of a high sample mass, such as that achieved by the method proposed (1350 mg of dietary supplement), leads to an improvement in the LODs. The use of diluted acid promotes great compatibility between the digesting solutions and the determination technique, especially in IC analysis. In addition, despite the large sample digested, the maximum pressure did not exceed the maximum limit recommended by the manufacturer of the equipment for safe operational conditions (80 bar). After digestion, it was not possible to observe brown fumes inside the reaction vessel when digestion was performed under oxygen pressure, indicating that all the NO₂ was reabsorbed and contributed to the regeneration process of HNO₃. Considering the phenomenon of NO₂ absorption during sample oxidation, it may have contributed to the mild operational condition even using 1350 mg of sample mass, which was performed without sudden increases in the system pressure.

3.4 Evaluation of the P and S determination by ion chromatography

The presence of a high concentration of DOC, as well as high RA in digests or the presence of some species at a high concentration in the final solutions can interfere in the analytes determination by IC (Kaiser et al., 2001, 1999; Novic et al., 1995; Slingsby and Kiser, 2001; Stewart and Olesik, 1998). Thus, the compatibility between the IC analysis and the final digest solutions obtained after the most suitable condition, using conventional microwave-assisted digestion method (850 mg of supplement using 7 mol L⁻¹ HNO₃) and microwave-assisted digestion method under oxygen pressure (1350 mg of supplement using 5 mol L⁻¹ HNO₃ under 5 bar of oxygen pressure) were evaluated. The chromatograms are shown in Fig. 2.

Close

Fig. 2 Chromatograms obtained after digestion using ( ) 850 mg of dietary supplement using microwave-assisted digestion with 7 mol L⁻¹ HNO₃ without oxygen pressure and (—) 1350 mg of dietary supplement using 5 mol L⁻¹ HNO₃ under 5 bar of oxygen pressure. Dilution factor of 40 times; (1) system peak, (2) chloride, (3) nitrate, (4) phosphate and (5) sulfate.

Export to PPT

Prior to the IC analyses all the samples were diluted to minimize the effects of high nitrate content in the digests from the acid used for the sample preparation methods. The digestion methods without and under oxygen pressure have different behavior during the digestion process in view of the oxygen amount in the system. This reagent will directly impact the RA and nitrate content in the final solution. Thus, dilution factors ranged from 20 to 120 times were evaluated for each digestion method. As observed in Fig. 2,¹ the digest solutions after both digestion methods evaluated presented significant peak of nitrate. In the conventional microwave-assisted digestion method (without oxygen pressure), at least 7 mol L⁻¹ HNO₃ were necessary to digest 850 mg of dietary supplement. Under these conditions, the nitrate peak was extremely large and exceeded the scale of a conductivity detector. The high nitrate content in the digests interferes in the integration of PO₄³⁻ and SO₄²⁻ peaks by IC, even when performing a dilution by a factor of 40 times (Fig. 2, red line). Thus, a dilution factor prior to analysis of 80 times was necessary to minimize the signals of nitrate and to prevent damage to the equipment. The dilution factor and the sample mass digested impair the LODs, and may hinder the quantification of the analytes at low concentrations.

On the other hand, in the microwave-assisted digestion under the oxygen pressure method, a solution of 5 mol L⁻¹ HNO₃ was enough to digest 1350 mg of dietary supplement. This is a condition in which practically all acid was consumed for organic matter oxidation and the RA was only 0.25 mol L⁻¹ (almost 10 times lower than a conventional digestion procedure). This behavior occurs because all HNO₃ remains active during the regeneration process while there is sufficient oxygen available in the gas phase during the organic matter oxidation. Under these conditions, non-observable interferences caused by nitrate peak were observed over PO₄³⁻ and SO₄²⁻ peaks by IC when a dilution factor of 40 times was performed (Fig. 2, black line) in view of the acid consumption during the regeneration process. The lower dilution factor and the higher sample mass digested in the digestion method under oxygen pressure allow to obtain suitable LODs in the determination of PO₄³⁻ and SO₄²⁻ by IC.

3.5 Analytical figures of merit

Accuracy of the proposed method was evaluated by recovery tests using a reference standard solution added on sample prior to the digestion method. The concentration spiked was around 50% of the concentration of the analytes originally present in the sample (100 µL of a standard solution containing 15000 mg L⁻¹ of P and 8000 mg L⁻¹ of S) and the analytes were determined by IC. Recoveries for P and S ranged from 99 to 102%, and the relative standard deviations (RSDs) were below 6%. Additionally, the accuracy of the proposed method was also evaluated by digestion of a mixture of CRM of aquatic plant (BCR 060) with the sample as a recovery test. This study was carried out because it was considered that the added P and S, naturally bound to the CRM matrix, it may behave appropriately for a real evaluation of the proposed procedure. It was considered as a recovery test, but using bound elements instead of only elements present in standard solution. Thus, 200 mg of CRM BCR 060 was mixed with the dietary supplement sample (1150 mg) for further digestion by the proposed method. The CRM BCR 060 was chosen in view of the possibility of adding approximately the same concentration obtained for the sample used in the optimizations (values added: 904 mg kg⁻¹ for P and 448 mg kg⁻¹ for S). Recoveries for P and S ranged from 96 to 98%, and the RSDs were below 10%. The obtained solutions were also analyzed by ICP-OES, and the concentrations of P and S did not present significant differences (Student's t-test, confidence level of 95%, p > 0.05) from those obtained by IC.

To compare the results, the sample used for optimizations of the proposed method was also digested by conventional microwave-assisted digestion using concentrated HNO₃ and subsequent determination of P and S by ICP-OES. It was observed that the obtained results for P (854 ± 17 mg kg⁻¹) and S (511 ± 13 mg kg⁻¹) by ICP-OES determination were in agreement (Student's t-test, confidence level of 95%, p > 0.05) with those obtained by the proposed method (IC determination, P = 828 ± 64 mg kg⁻¹ and S = 500 ± 47 mg kg⁻¹). These results show that the IC determination of P and S was performed with non-observable matrix-related interferences, and practically all the P and S in the final digest were present as PO₄³⁻ and SO₄²⁻, respectively, allowing an accurate indirect determination by IC.

Limits of detection of conventional microwave-assisted digestion associated with the indirect determination of P and S by IC were 1070 mg kg⁻¹ and S: 1400 mg kg⁻¹, respectively. On the other hand, the LODs of microwave-assisted digestion under oxygen pressure for the supplement digest associated with the indirect determination of P and S by IC were 200 and 270 mg kg⁻¹, respectively. It is associated with the higher sample mass digested under oxygen pressure and the great compatibility between the digest solutions and the determination technique. Although the LODs using ICP-OES for P and S (0.53 and 0.07 mg kg⁻¹, respectively) were lower than those obtained by IC, it is important to mention that the proposed method allow to digest a large sample mass of supplement (1350 mg), which makes it possible to obtain suitable LODs for both analytes by IC. This is an important aspect to be considered, since IC has a lower operational cost than ICP-OES. In addition, P and S are normally determined in the vacuum UV (VUV) for ICP-OES, this being another limitation since an optical system specially designed for working in this low emission lines must be available.

3.6 Phosphorus and sulfur determination in supplements

The proposed method was applied for several dietary supplements, and the results are shown in Table 2. For comparison of the results these digests were also analyzed by ICP-OES. No significant differences (Student's t-test, confidence level of 95%, p > 0.05) were observed between the obtained results for P and S in all the dietary supplement samples using IC or ICP-OES. Relative standard deviations obtained by IC determination were always lower than 10%, quite close to RSDs obtained by ICP-OES (RSD = 6%).

It is possible to observe in Table 2 that the concentrations of P and S, determined by the proposed method, in supplements vary in a range of concentrations (from 812 to 2251 mg kg⁻¹ and from 500 to 3039 mg kg⁻¹ for P and S, respectively). Although the variation might be as a result of different manufacturers and the use of different raw materials in their production, it is possible to emphasize that the proposed approach behaves accurately for the entire concentration range. The obtained result for P in sample A (828 ± 64 mg kg⁻¹) was in agreement with the value reported by the manufacturer (780 mg kg⁻¹). On the other hand, the obtained value for P in sample C (2251 ± 94 mg kg⁻¹) was about 3 times higher than that reported by the manufacturer (762 mg kg⁻¹). For sample E, the P concentration found (2230 ± 180 mg kg⁻¹) was about 1.5 times smaller than that reported by the manufacturer (3591 mg kg⁻¹). The concentration of P in the samples D and B was not informed by the manufacturers. With respect to the S content, although it is an essential nutrient for the human organism and it is also associated with disorders that may have adverse effects, there were values for these elements that had not been informed in any of the dietary supplements analyzed. In view of this, and taking into account the wide variation observed for P and S concentration in the samples and the concentrations disagreement with the value informed by manufacturers, an effective quality control should be performed due the impacts to the human health and to ensure food safety.