1. Introduction

Biodiesel has significant potential as an alternative diesel due to its biodegradability, renewable energy source, better lubricity, combustion efficiency, and low toxicity compared to other fuels (Tan et al., 2015). Glycerol is produced as a by-product, representing around 10% of the total volume of biodiesel produced (Bilck et al., 2015; Chi et al., 2007).

Studies show that the conversion of glycerol to other chemicals with high added value, such as lactic acid, makes the glycerol industry more profitable (Arcanjo et al., 2017; Manfro and Souza, 2014), making biodiesel production a more suitable option to replace fossil fuels (Huang et al., 2012).

Glycerol can be dehydrogenated to form dihydroxyacetone (Santos et al., 2018). This intermediate went through dehydration to pyruvaldehyde on solid acid catalysts and rearranged via Cannizzaro reaction to lactic acid. Therefore, from the literature, pyruvaldehyde is a crucial intermediate on glycerol valorization in lactic acid (Lari et al., 2016; Santos et al., 2018).

Lactic acid is a raw material to benefit from synthesizing several products, applying in the textile industry and cosmetic industry (Drumright et al., 2000). In the food industry, its main applications are as a pH regulator and food additive (Bruno et al., 2020; Drumright et al. 2000). It has been used as a precursor of solvents in green chemistry, such as ethyl lactate, and used in the synthesis of polylactic acid (PLA), applied in the production of biodegradable packaging (Bruno et al., 2020; Drumright et al., 2000).

Lactic acid is commonly sold as an 88% solution, varying its price according to its application in the market. In general, lactic acid price follows the cost of starch and sugar raw materials used for fermentation (Biddy et al., 2016). With a market growth of 16.2% per year, it is expected that this demand will reach 960.1 kt in 2025, which should represent U$ 9.8 billion in the global market. As with lactic acid, PLA demand is also expected to grow, reaching $ 6.5 billion worldwide in 2025. One reason for this would be the Asia-Pacific PLA market growth rate of 22.4% per year due to various beverages and foods and the cosmetics manufacturing units (Oliveira et al., 2018).

The hydrothermal conversion of glycerol to lactic acid has been widely studied recently as an alternative to the fermentation route, with the advantage of using a low-cost raw material. Furthermore, for environmental reasons, there is an increase in demand for lactic acid, as it is the raw material for producing a biodegradable polymer (PLA) (Bruno et al., 2018; Shen et al., 2009).

Some methods described in the literature involve the determination of different organic acids by high-performance liquid chromatography (HPLC). However, these methods cannot efficiently separate glycerol and lactic acid, observing the overlap of peaks. In this context, the present work describes a methodology for separating the lactic acid produced by hydrothermal conversion of glycerol using HPLC.

Materials and methods

2.1 Standard samples of glycerol and lactic acid

Glycerol and lactic acid HPLC grade were obtained from Sigma-Aldrich (St. Louis, U.S.A.) and Proquímios (Rio de Janeiro, Brazil). Two calibration curves were obtained using the following concentrations 0.6, 1.8, 3.0, 4.2, 5.4 and 6.6 g L^–1 for the composition of the curve points. Standards were diluted in Milli-Q water to 10.0% vol. (Millipore, Bedford, USA) for method analysis. Diluted solutions acted as standards for linearity assessment.

2.2 Design of Experiments

A central composite design (CCD) was proposed to study the influence of experimental variables on glycerolysis reaction. The effect of two variable parameters, time and temperature, requiring nine experiments using face-centered stair points was investigated. The low and high values of the variables were chosen based on preliminary investigations. For this purpose, sodium hydroxide and glycerol concentrations, catalyst content, time, and temperature were used as independent variables, and lactic acid concentration was used as the dependent variable, as presented in Tab. 1. The experimental design and optimization were conducted with Statistica 12.0 software.

2.3 Reaction Procedure

A 100 mL 4590 series reactor (Parr Instrument Company, Moline, U.S.A.) was used for hydrothermal conversion of glycerol. The operating conditions were modulated by a 4848 reactor controller (Parr Instrument Company), and the reagents were fed into the reactor in the desired mass ratio of glycerol: NaOH (1:1) and desired temperatures (220, 250 and 280 °C), with the use of 5% of a copper catalyst supported on silica. Then, the reactor contents were stirred at 1000 rpm at a vacuum pressure of 800 psi. The total reaction time was 3, 4.5 and 6 h.

2.4 High-performance liquid chromatography method

Analyses were performed employing the Nexera series UHPLC system (Shimadzu, Japan) equipped with an LC-40 XR binary pump, a RID-20A detector, and a SIL-40XR autosampler. A Hi-Plex H column (Agilent Technologies, Santa Clara, U.S.A.), 300 mm long by 7.7 mm internal diameter and 8 µm particle size was used to separate the products. The mobile phase used was H₂SO₄ 0.001 mol L^–1 + 10% ACN (organic modifier), using ultrasound previously to remove bubbles. The conditions used for this method were the temperature of detector and oven at 60 °C, 20 min of running time, and flow rate of 0.6 mL min^–1. The flow rate of 0.6 mL min^–1, column oven and detector temperature at 60 °C and 20 min for the total execution time. Samples were diluted to 10% vol in ultrapure water (Milli-Q) as a diluent and injected in triplicate. The results were analyzed using the LabSolutions software (Shimadzu).

2.5 Method validation

The developed method was validated according to ICH (2005) guidelines in terms of the following analytical parameters: linearity, limit of detection (LOD), limit of quantification (LOQ), precision, and accuracy (Dias et al., 2020).

2.6 Linearity

Linearity was given by injection in triplicate of the standards of each point used to build the calibration curve (Novaes et al., 2018). Two distinct curves (glycerol and lactic acid) were drawn in the interval of 0.6–6.6 g L–1 (0.6; 1.8; 3.0; 4.2; 5.4 and 6.6 g L^–1). Homogeneity was assessed by the Cochran test (Shen et al., 2009), through Eq. 1.

where s2max is the highest value variance and Eik = 1s2i is the total of all samples variances.

2.7 Limit of detection and limit of quantification

The LOD and LOQ were calculated using Eqs. 2 and 3, respectively

where σ_LC is the standard deviation of the minor number concentration (LC) tested and Importar imagen is the slope of the curve obtained.

2.8 Precision

Precision was calculated by repeatability, using three different concentrations (lowest, average, and highest) in triplicate (Rocha and Bacelar Júnior, 2018). Concentrations were 0.6, 3.0 and 6.6 g L^–1 for both curves, and the result was expressed as relative standard deviation (RSD, in %), as in Eq. 4.

where σ represents the deviation from the lowest standard concentration and Importar imagen is the mean concentration.

2.9 Accuracy

This parameter was represented by the recovery (R, in %) by calculating different concentrations (lowest, average, and highest) in triplicate for the different curves. Concentrations used were 0.6, 3.0 and 6.6 g L^–1 for both curves. Accuracy was calculated with Eq. 5.

where C_m and C_e are the obtained and theoretical concentrations, respectively.

3. Results and discussion

3.1 Reactions

Table 2 shows the results of conversion and selectivity for lactic acid under the reaction conditions of the experimental design.

According to Tab. 2, it can be observed that the method proposed by HPLC was efficient in separating the two compounds, glycerol and lactic acid, allowing the quantification of each one of these compounds and enabling the calculation of the conversion and the selectivity for lactic acid. It is essential to mention that pyruvaldehyde was obtained as a by-product, in smaller amounts, in all reactions, which is expected for catalysis reactions in the primary medium due to the greater predisposition of formation of this intermediate by the Cannizzaro reaction (Lari et al., 2016; Santos et al., 2018).

It can also be noticed that, according to Fig. 1, when comparing the temperatures, increasing this parameter favors the formation of secondary compounds, such as pyruvaldehyde. When the reaction time increases, the appearance of lactic acid is preferred, consequently increasing its selectivity. Therefore, temperature increase combined with the increase in reaction time is favorable to the rise of the conversion and the rise in the selectivity of the lactic acid, which corroborates the results expected in the literature (Evans et al., 2020).

3.2 Analysis by HPLC

All standards and reaction products were examined by HPLC and showed similar chromatographic appearance, which can be seen in Fig. 2. The overlapping peaks indicate coelution of interference that shares the same transition, with consequences such as inadequate quantification of the compound of interest or even the absence of identification of the combination of interest (Welch et al., 2009). Diluted sulfuric acid as a mobile phase showed this coelution of desired compounds, generating an incorrect quantification.

Beltrán-Pietro et al. (2013) validated a chromatographic method using an ionic column at a temperature of 70 °C with a refractive index detector (RID) detector connected to a UV detector, with a phase composed of 3 mmol L^–1 of sulfuric acid (H₂SO₄, pH 2.00). with a flow rate of 0.5 mL min^–1 in an analysis time of 30 min. With this method, it is possible to identify compounds such as glyceraldehyde, oxalic acid, dihydroxyacetone, among others; however, it presents coelution of other compounds, such as glycerol and lactic acid.

Although some methods in the literature report the separation of glycerol and lactic acid in the chromatographic profile, it is observed that, in fact, there is a co-elution of these compounds, not allowing an efficient quantification of these products, requiring the development of a new method.

As an alternative, using the mobile phase with the organic modifier was essential for separating glycerol and lactic acid compounds, allowing a better chromatographic profile to visualize both peaks. The choice of acetonitrile as an organic modifier was due to its low viscosity (0.4 cP at 25 °C), which generates less pressure and allows a better retention time response (Leite, 2008; Sadek, 1996). Some tests were performed with 5, 10 and 20% vol of acetonitrile to decide the best proportion of acetonitrile in the mobile phase. The best chromatographic separation occurred with the 10% vol of acetonitrile.

3.3. Method validation

3.3.1 Linearity

Six standard samples of glycerol and lactic acid were analyzed in triplicate by HPLC for each curve constructed, as seen in Fig. 3. This parameter was evaluated in the interval of 0.6–6.6 g L^–1. The determination coefficient (R²) was above the acceptance rule values (0.990) (Leite, 2008; Ribani et al., 2004; Sadek, 1996), indicating excellent linearity.

The Cochran test was used with a 95% confidence level using the number of samples (n = 6) in triplicate. The calculated values of C were 0.661 for the glycerol curve and 0.130 for the lactic acid curve, being lower values when compared to the tabulated (0.616), indicating homoscedastic variances (Novaes et al., 2018).

3.3.2 Limit of detection and limit of quantitation

For the glycerol analytical curve, LOD and LOQ values were 0.076 and 0.385 g L^-1, respectively, while for lactic acid, the LOD and LOQ values were 0.221 and 1.166 g L^–1. The results showed the possibility of differentiation and quantification of the different samples of each curve. The responsible standards require values below 0.5 g L^–1, and the present method is applicable (Cassini et al., 2013; Souza and Junqueira, 2005), except only for lactic acid that showed a value above that specified by the standard.

3.3.3 Precision

Through Tab. 3, it is possible to conclude that the values are accurate about repeatability since they are lower than the 5% determined by the standards (Novaes et al., 2018; Ribani et al., 2004; Sadek, 1996).

3.3.4 Accuracy

The values obtained for this parameter are between 98.33 and 100%. Regarding the glycerol curve, the mean recovery was 99.44 ± 0.03%, while for lactic acid, it was 100.00 ± 0.02%, indicating an amicable agreement between experimental and theoretical data. Furthermore, these results are within what is expected by regulatory standards (70–120%) (Goutal et al., 2016; Lanças, 2004).

4. Conclusions

The methodology described in this article allowed the separation of lactic acid and glycerol by HPLC in a quick and straightforward run of 20 min. In addition, it was possible to quantify each one, allowing the calculation of conversion and selectivity for lactic acid for the reactions mentioned in the experimental design.

Authors’ contribution

Conceptualization: Cruz, A. F. A.

Data curation: Cruz, A. F. A.

Formal Analysis: Cruz, A. F. A.; Dias, G. P.; Bittencourt, F. R.

Funding acquisition: Not applicable

Investigation: Cruz, A. F. A.

Methodology: Cruz, A. F. A.; Dias, G. P.; Bittencourt, F. R.

Project administration: Aranda, D. A. G.

Resources: Aranda, D. A. G.

Software: Cruz, A. F. A.; Dias, G. P.; Bittencourt, F. R.

Supervision: Aranda, D. A. G.

Validation: Cruz, A. F. A.

Visualization: Cruz, A. F. A.; Aranda, D. A. G.

Writing – original draft: Cruz, A. F. A.; Aranda, D. A. G.

Writing – review & editing: Cruz, A. F. A.; Aranda, D. A. G.

Data availability statement

All data sets were generated or analyzed in the current study.

Funding

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq). Grant No: 167934/2019-8.

Acknowledgments

LIPCAT – Laboratório de Intensificação de Processos e Catálise

References

Arcanjo, M. R. A.; Silva Júnior, I. J.; Rodríguez-Castellón, E.; Infantes-Molina, A.; Vieira, R. S. Conversion of glycerol into lactic acid using Pd or Pt supported on carbon as catalyst. Catal. Today.2017, 279 (Part 2). https://doi.org/10.1016/j.cattod.2016.02.015

Beltrán-Prieto, J. C.; Pecha, J.; Kašpárková, V.; Kolomazník, K. Development of an HPLC method for the determination of glycerol oxidation products. J. Liq. Chromatogr. Relat. Technol.2013, 36 (19), 2758–2773. https://doi.org/10.1080/10826076.2012.725695

Biddy, M. J.; Scarlata, C.; Kinchin, C. Chemicals from Biomass: a Market assessment of bioproducts with near-term potential. NREL/TP-5100-65509; National Renewable Energy Laboratory: Golden, CO, 2016. https://www.nrel.gov/docs/fy16osti/65509.pdf (Accessed 2021-08-19).

Bilck, A. P.; Müller, C. M. O.; Olivato, J. B.; Mali, S.; Grossmann, M. V. E.; Yamashita, F. Using glycerol produced from biodiesel as a plasticiser in extruded biodegradable films. Polímeros, 2015, 25 (4), 331–335. https://doi.org/10.1590/0104-1428.1803

Bruno, A. M.; Chagas, C. A.; Souza, M. V. M.; Manfro, R. L. Lactic acid production from glycerol in alkaline medium using Pt-based catalysts in continuous flow reaction system. Renew. Energ. 2018, 118, 160–171. https://doi.org/10.1016/j.renene.2017.11.014

Bruno, A. M.; Simões, T. D. R.; Souza, M. M. V. M.; Manfro, R. L. Cu catalysts supported on CaO/MgO for glycerol conversion to lactic acid in alkaline medium employing a continuous flow reaction system. RSC Adv.2020, 10 (52), 31123–31138. https://doi.org/10.1039/D0RA06547A

Cassini, S. T.; Antunes, P. W. P.; Keller, R. Validação de método analítico livre de acetonitrila para análise de microcistinas por cromatografia líquida de alta eficiência. Quim. Nova.2013, 36 (8), 1208–1213. https://doi.org/10.1590/S0100-40422013000800019

Chi, Z.; Pyle, D.; Wen, Z.; Frear. C.; Chen, S. A laboratory study of producing docosahexaenoic acid from biodiesel-waste glycerol by microalgal fermentation. Process Biochem.2007, 42 (11), 1537–1545. https://doi.org/10.1016/j.procbio.2007.08.008

Dias, G. P.; Santos, R. C.; Carvalho, R. C.; Souza, C. G.; Santos, A. P. F.; Andrade, D. F.; D’Ávila, L. A. Determination of Methanol in Gasoline and Ethanol Fuels by High-Performance Liquid Chromatography. J. Braz. Chem. Soc.2020, 31 (5), 1055–1063. https://doi.org/10.21577/0103-5053.20190272

Drumright, R. E.; Gruber, P. R.; Henton, D. E. Polylactic Acid Technology. Adv. Mater.2000, 12 (23), 1841–1846. https://doi.org/10.1002/1521-4095(200012)12:23%3C1841::AID-ADMA1841%3E3.0.CO;2-E

Evans, C. D.; Douthwaite, M.; Carter, J. H.; Pattisson, S.; Kondrat S. A.; Bethell, D.; Knight, D. W.; Taylor, S. H.; Hutchings, G. J. Enhancing the understanding of the glycerol to lactic acid reaction mechanism over AuPt/TiO₂ under alkaline conditions. J. Chem. Phys.2020, 152 (13), 134705. https://doi.org/10.1063/1.5128595

Goutal, S.; Auvity, S.; Legrand, T.; Hauquier, F.; Cisternino, S.; Chapy, H.; Saba, W.; Tournier, N. Validation of a simple HPLC-UV method for rifampicin determination in plasma: Application to the study of rifampicin arteriovenous concentration gradient. J. Pharm Biomed. Anal. 2016, 123 (1), 173–178. https://doi.org/10.1016/j.jpba.2016.02.013

Huang, D.; Zhou, H.; Lin, L. Biodiesel: an Alternative to Conventional Fuel. Energy Procedia. 2012, 16 (Part C), 1874–1885. https://doi.org/10.1016/j.egypro.2012.01.287

ICH International Conference on Harmonization. Validation of Analytical Procedures: Text and Methodology – Q2 (R1); Geneva, Switzerland, 2005. https://www.ema.europa.eu/en/documents/scientific-guideline/ich-q-2-r1-validation-analytical-procedures-text-methodology-step-5_en.pdf (Accessed 2021-08-19).

Lanças, F. M. Validação de Métodos Cromatográficos de Análise; Editora RiMA, 2004.

Lari, G. M.; García-Muelas, R.; Mondelli, C.; López, N.; Pérez-Ramírez, J. Glycerol oxidehydration to pyruvaldehyde over silver-based catalysts for improved lactic acid production. Green Chem.2016, 18 (17), 4682–4692. https://doi.org/10.1039/C6GC00894A

Leite, F. Validação em Análise Química; Átomo, 2008.

Manfro, R. L.; Souza, M. M. V. M. Production of Renewable Hydrogen by Glycerol Steam Reforming Using Ni–Cu–Mg–Al Mixed Oxides Obtained from Hydrotalcite-like Compounds. Catal. Lett. 2014, 144 (5), 867–877. https://doi.org/10.1007/s10562-014-1196-6

Novaes, C. G.; Yamaki, R. T.; Paula, V. F. Nascimento Júnior, B. B.; Barreto, J. A.; Valasques, G. S.; Bezerra, M. A. Otimização de Métodos Analíticos Usando Metodologia de Superfícies de Resposta – Parte I: Variáveis de Processo. Rev. Virtual Quim.2018, 9 (3), 1184–1215. https://doi.org/10.21577/1984-6835.20170070

Oliveira, R. A.; Komesu, A.; Rossell C. E. V.; Maciel Filho, R. Challenges and opportunities in lactic acid bioprocess design—From economic to production aspects. Biochem. Eng. J.2018, 133, 219–239. https://doi.org/10.1016/j.bej.2018.03.003

Ribani, M.; Bottoli, C. B. G.; Collins, C. H.; Jardim, I. C. S. F.; Melo, L. F. C. Validação em métodos cromatográficos e eletroforéticos. Quim. Nova.2004, 27 (5), 771–780. https://doi.org/10.1590/S0100-40422004000500017

Rocha, K. R.; Bacelar Júnior, A. J. ANOVA Medidas repetidas e seus pressupostos: Análise passo a passo de um experimento. Perspect. Cienc. Tec.2018, 10, 29–51. https://doi.org/10.22407/1984-5693.2018.v10.p.29-51

Sadek, P. C. The HPLC Solvent Guide; Wiley, 1996.

Santos, K. M. A.; Albuquerque, E. M.; Borges, L. E. P.; Fraga, M. A. Discussing Lewis and Brønsted acidity on continuous pyruvaldehyde Cannizzaro reaction to lactic acid over solid catalysts. Mol. Catal.2018, 458 (Part B), 198–205. https://doi.org/10.1016/j.mcat.2017.12.010

Shen, Z.; Jin, F.; Zhang, Y.; Wu, B.; Kishita, A.; Tohjj, K.; Kishida, H. Effect of Alkaline Catalysts on the Hydrothermal Conversion of Glycerin into Lactic Acid. Ind. Eng. Chem. Res.2009, 48 (19), 8920–8925. https://doi.org/10.1021/ie900937d

Souza, S. V. C.; Junqueira, R. G. A procedure to assess linearity by ordinary least squares method. Anal. Chim. Acta.2005, 552 (1–2), 25–35. https://doi.org/10.1016/j.aca.2005.07.043

Tan, Y. H.; Abdullah, M.O.; Nolasco-Hipolito, C. The potential of waste cooking oil-based biodiesel using heterogeneous catalyst derived from various calcined eggshells coupled with an emulsification technique: A review on the emission reduction and engine performance. Renew. Sust. Energ. Rev.2015, 47, 589–603. https://doi.org/10.1016/j.rser.2015.03.048

Welch, C. J.; Brkovic, T.; Schafer, W.; Gong, X. Performance to burn? Re-evaluating the choice of acetonitrile as the platform solvent for analytical HPLC. Green Chem.2009, 11 (8), 1232–1238. https://doi.org/10.1039/B906215G

Notas de autor

andrezafalvesc@gmail.com