Introduction

High biological activity of formic acid generates the interest in monitoring of its sources for humans. The elevated level of the acid occurring in the human body can cause significant fatal effects such as serious damage to optical nerve, respiratory failure, hepatic and renal failure, and coma. Blood concentrations above 10 mmol/L (0.5 mg/mL) can induce deep metabolic acidosis and lead to death (Kinoshita et al. 1998; Tanaka et al. 1991). Formic acid can enter the human body through diet, production by intestinal microflora, and, as a result, methanol ingestion or inhalation of its vapors. Methanol ingestion is the most important source of high concentrations of formates in body fluids and tissues. Methanol is rapidly, within 2–24 h after ingestion, metabolized (dehydrogenated) by alcohol dehydrogenase (ADH) to formaldehyde and further to formic acid of substantially higher than methanol toxicity (Hovda et al. 2005; Kinoshita et al. 1998; Kruse 1992; Wallage and Watterson 2008). The content of formate in blood is a marker of methanol intoxication. Its rapid screening in the human body is of great significance for clinical diagnosis. High concentration of formic acid, corresponding to those related to fatal methanol poisonings, can be found in putrefied postmortem physiological fluids, presumably due to bacterial action and decomposition of lipids, carbohydrates, and proteins (Viinamäki et al. 2011).

Small amounts of methanol can occur in commercial alcoholic drinks as a by-product of fermentation and distillation steps (5.2–85.6 mg/L in vodkas (Arbuzov and Savchuk 2002) and 3.1–4.6 mg/100 mL ethanol in rums (Lachenmeier et al. 2003)). Wide consumption of alcoholic beverages can make a source of formic acid for humans (Kapur et al. 2007). Until now, data on the content of formates in such products are scarce. Analytical control of the sources of formic acid for humans is of great value for the general food industry. Organic acids, including formic acid, can affect sensory properties of food items as well as their stability.

Gas chromatography (Lee et al. 2012; Lee et al. 2013; Ligor et al. 2008; Ryhl-Svendsen and Glastrup 2002; Sokoro et al. 2007), capillary electrophoresis (Klampfl et al. 2000; Kubáñ and Boček 2013; Kubáň et al. 2013; Kubáň et al. 2014; Pantučkova et al. 2013; Suárez-Luque et al. 2006; Travassos Lemos et al. 2015), and ion chromatography (Brudin et al. 2010; Käkölä et al. 2008; Kuo 1998; Lodi and Rossin 1995; Magne et al. 1998; Rantakokko et al. 2004; Walford 2002; Wojtczak et al. 2013) are techniques most often used for the determination of low-molecular-mass carboxylic acids in various materials, including food products. Complications arising from the major sample components, e.g., proteins, lipids, and salts, in the examination of the samples of biological matrices, can generate the necessity of coupling of the technique with a suitable sample preparation step for their elimination.

Ion chromatography (IC) is a technique widely applied in the investigations of ionic components of various materials (Weiss 2004). Fast, simple, and selective determination of inorganic and small organic ionic species in one run makes the technique competitive to the other instrumental techniques. Little sample preparation steps and small samples required for analysis are the advantages of IC analytical procedures. The possibility of the use of ion chromatography for rapid identification and quantification of small organic acids, among them formic acid, in various food and beverages samples was documented (Lodi and Rossin 1995; Magne et al. 1998; Walford 2002; Brudin et al. 2010; Wojtczak et al. 2013). The content of organic acids in foods can serve as a marker of degradation and fermentation processes occurring during production and storage steps. The IC technique is particularly suitable for direct examination of the occurrence of small organic acids in drinking waters, among them those being by-products of ozonation processes (Kuo 1998; Rantakokko et al. 2004). Such compounds can be consumed by microorganisms stimulating their regrowth. Simultaneous inorganic (cations and anions) multicomponent analytical results can make the basis of the differentiation in the originality of alcoholic products (Arbuzov and Savchuk 2002; Lachenmeier et al. 2003). Reliable determination of formates and acetates in the presence of inorganic anions requires careful optimization of the IC analytical procedure for effective baseline separation of formate, acetate, and fluoride signals exhibiting close retention times (Kontozova-Deutsch et al. 2008; Kontozova-Deutsch et al. 2011; Janiszewska and Balcerzak 2013).

In the present work, a rapid method for the determination of formates in vodka and rum products, differing in production processes, by the use of suppressed anionic IC with conductometric detection was developed. Vodka is produced by rectifying ethyl alcohol, the product of fermentation of grains (wheat, rye, corn), potatoes, sugar beets, and fruits. The distillation products (of 96–97% alcohol content) are diluted with water usually to alcohol level below 40% prior to bottling. Rum is a distilled alcoholic beverage made from sugarcane by-products (molasses or syrup) or sugarcane juice. The distilled spirits are diluted with water to 60–70% of alcohol and aged (usually for a year) in oak barrels. During aging, the rum acquires golden color.

Materials and Methods

Apparatus

The 761 Compact IC System from Metrohm AG (Herisau, Switzerland) with a MSM II (Metrohm Suppressor Module) suppressor and a conductometric detector was used. Anion separations were investigated with the use of two different chromatographic columns, Metrosep A Supp 7 (250 mm × 4 mm) from Metrohm, Switzerland, and IonPac AS9-HC (250 mm × 4 mm) from Dionex, USA (the volume of a sample injection loop was 20 μL). Data acquisition and evaluation of chromatograms were carried out with the IC Net 2.3 Metrodata (Metrohm) software.

An infrared (IR) radiation lamp was used for sample evaporation.

Filters of 0.45 μm pore size (MCE, Fisherbrand) were used for sample filtration throughout experiments.

Reagents

All reagents used were of analytical grade. Water purified in a Millipore Elix3/Simplicity UV system (a specific resistance >18.2 MΏ cm) was used in all experiments.

Standard solutions of formate and acetate were prepared from high-purity acids (from Fluka) dedicated to IC and containing 1000 ± 4 mg/L of each anion. Standard solutions of inorganic anions (fluoride, chloride, bromide, nitrate(V), phosphate, and sulfate) were prepared from high-purity sodium salts (from Fluka) containing 1000 ± 4 mg/L of each anion and dedicated to IC.

The eluent of 3.6 mmol/L of sodium carbonate prepared by dissolution of 381.6 mg of Na2CO3 (from Fluka) in 1 L of deionized water was used.

Alcoholic Drinks Examined

Two different Polish industrial alcoholic products, pure vodka (Absolwent) and rum (Golden Rum), were bought on the local market. The samples taken from commercial bottles were directly analyzed by the analytical procedure described below.

Analytical Procedure

Alcohol samples of 5–50 mL were transformed into a 50-mL beaker and evaporated to ca. 1 mL under IR lamp (20 cm distance between the lamp and the sample surface). Each evaporated sample was diluted to 5.0 mL with deionized water and submitted to chromatographic analysis after filtration through a 0.45-μm filter.

Results and Discussion

Formate is a small organic anion of ion chromatographic signal registered under the conditions commonly used for the determination of inorganic anions. The ion chromatogram of the mixture of formate and some inorganic anions (F−, Cl−, NO2 −, Br−, NO3 −, HPO4 2−, and SO4 2−) registered when using a very popular anion-exchange A Supp 5 Metrohm column, Na2CO3 + NaHCO3 eluent, and conductometric detection is given in Fig. 1. Acetate and oxalate are organic acids which signals can also be registered under such conditions. Close retention times of formate, acetate, and fluoride ions can provide problems with quantification of particular anions in the samples containing all of them.

Fig. 1
figure 1

The ion chromatogram of synthetic mixture of formate, acetate, and common inorganic anions registered with Metrohm Metrosep A Supp 5 chromatographic column (3.2 mmol/L Na2CO3 + 1.0 mmol/L NaHCO3 eluent)

Full size image

A kind of chromatographic column and an eluent applied limit the effectiveness of the separation of formate ion signal from those of acetate and inorganic anions in case of their occurrence in the examined sample. In this work, the effectiveness of two anion-exchange columns, Metrosep A Supp 7 from Metrohm and IonPac AS9-HC from Dionex, in the separation of F−, CH3COO−, and HCOO− ions and quantification of formates was examined. Both columns are dedicated to the determination of inorganic and small organic anions occurring in the examined solutions. The experiments carried out have shown that complete resolution of fluoride, acetate, and formate signals is possible with both columns and Na2CO3 as an eluent. Figure 2 presents the ion chromatograms of the mixture of the ions obtained with the use of the examined columns and the same (3.6 mmol/L Na2CO3) eluent. Both columns exhibited similar effectiveness in the separation of F−, CH3COO−, and HCOO− ions. The Metrosep A Supp 7 column was chosen for further studies owing to shorter analysis time of the samples containing the other inorganic anions (Cl−, NO2 −, Br−, NO3 −, HPO4 2−, SO4 2−) usually occurring in food products and giving IC signals under the conditions used. The separation of all ions was possible in shorter, ca. 35 min, time on the Metrosep A Supp 7 column as compared with ca. 60 min when using the IonPac AS9-HC column. The analysis time of the mixture of all ions on the IonPac AS9-HC column decreases in case of more concentrated eluent (ca. 30 min for 9 mmol/L Na2CO3). The isolation of formate signal from the acetate is, however, worse under such conditions.

Fig. 2
figure 2

The chromatograms of synthetic mixture of formate (10 mg/L), acetate (10 mg/L), and fluoride (2 mg/L) ions, and the values of asymmetry factors and resolutions obtained with the use of a Metrohm Metrosep A Supp 7 and b Dionex IonPac AS9-HC (3.6 mmol/L Na2CO3 eluent, 0.7 mL/min)

Full size image

Five standard solutions containing the mixtures of formates with acetates and fluorides were used for the calibration procedure (Table 1). Each of the calibration solution was run triplicate. Two calibration curves were generated by plotting (1) the peak areas and (2) the peak heights against the concentrations of formate in the injected standards (Fig. 3). Both parameters, peak areas and peak heights, can make the basis of the quantitation procedure. Statistical data of formate results obtained using peak areas for quantitation are presented in Table 1. Detection limit (DL) for formate ions was evaluated in the analysis of synthetic mixtures containing their minimum concentrations being detected: 0.25 mg/L (HCOO−) in the mixture with 0.25 mg/L (CH3COO−) and 0.05 mg/L (F−) ions and calculated as a triple standard deviation (SD) of the results (n = 10). DL and quantitation limit (QL) were evaluated as 0.014 and 0.042 mg/L, respectively. Calibration range corresponded to 0.04–10.0 mg/L of formate ions.

Table 1 Statistical data for formate IC results (n = 3) in mixtures with acetates and fluorides
Full size table
Fig. 3
figure 3

Calibration curves for formate ions using 1 peak areas and 2 peak heights (Metrosep A Supp 7 column, 3.6 mmol/L Na2CO3 eluent, 0.7 mL/min)

Full size image

Analysis of Alcoholic Drink Samples

Two typical Polish alcoholic products, vodka (Absolwent) and rum (Golden Rum), available on the local market were analyzed for the confirmation of the applicability of the described chromatographic procedure for the determination of formates in such kind of samples. The samples of 5–50 mL taken from the commercial bottles were preliminary evaporated to ca. 1 mL to minimize the amount of alcoholic matrix. The evaporation was carried out under IR lamp (located ca. 20 cm above the surface of the sample), and the residue was diluted with deionized water to 5.0 mL of the final volume prior to submitting to chromatographic analysis. The experiments have shown that such approach improves baseline separation of formate signal. Figure 4 shows the IC chromatograms of the commercial rum sample directly submitted to IC analysis and after preliminary evaporation of ethanolic matrix. Lower amount of alcoholic matrix provides better adjustment of formate signal to the baseline. Changing the matrix from pure water to ethanol-water (1:4) does not practically affect quantification results. The regression curves for formate ions in pure aqueous and ethanolic-aqueous matrices calculated on the basis of peak areas are presented in Fig. 5.

Fig. 4
figure 4

The IC chromatograms of commercial Golden Rum. a Directly submitted to IC analysis. b After preliminary evaporation of ethanolic matrix (Metrosep A Supp 7 column, 3.6 mmol/L Na2CO3 eluent, 0.7 mL/min)

Full size image
Fig. 5
figure 5

Calibration curves for formate ions in 1 aqueous and 2 ethanolic-aqueous (1:4) solutions

Full size image

The ion chromatograms of the examined spirit products under the conditions used are presented in Fig. 6. The evaluated contents of formate in both examined spirit drinks are presented in Table 2. Higher amounts of formates were determined in rum (3.11 ± 0.04 mg/L) as compared with vodka (0.21 ± 0.03 mg/L) samples. Spike recovery of formates (standard solution added to the commercial sample prior to the evaporation) was in the range of 96.0–107.0%. Figure 7 presents the chromatograms of the vodka sample spiked with formate (0.2 mg/L), fluoride (0.04 mg/L), and acetate (0.60 mg/L) ions.

Fig. 6
figure 6

The IC chromatograms of commercial. a Rum and b vodka samples (A Supp 7 column, 3.6 mmol/L Na2CO3 eluent, 0.7 mL/min) (?—unidentified peak)

Full size image
Table 2 The results for formate in the examined commercial spirit drinks (n = 3)
Full size table
Fig. 7
figure 7

The IC chromatograms of the original and spiked (formate (0.2 mg/L), fluoride (0.04 mg/L), and acetate (0.60 mg/L)) Absolwent vodka sample (?—unidentified peak)

Full size image

The developed method offers the possibility of rapid evaluation of the multi-anionic profile of alcoholic drinks. Figure 8 presents the whole anionic chromatogram of the examined Golden Rum sample under optimum conditions evaluated in this work (Metrosep A Supp 7 column, 3.6 mmol/L Na2CO3 (0.7 mL/min) eluent). The anionic profile can serve as a criterion for distinguishing among spirit drinks, including their adulteration and authentication.

Fig. 8
figure 8

The multi-anionic profile of Golden Rum sample (A Supp 7 column, 3.6 mmol/L Na2CO3 eluent, 0.7 mL/min)

Full size image

Conclusions

Ion chromatography provides the possibility of rapid evaluation of the multi-anionic (small organic acids and common inorganic anions) profile of alcoholic drinks. Two analytical columns examined, Metrosep A Supp 7 from Metrohm and IonPac AS9-HC from Dionex, allow effective isolation of formate signals from such kind of samples. The analysis time of the mixture of the examined organic (HCOO−, CH3COO−, and C2O4 2−) and common inorganic (F−, Cl−, NO2 −, Br−, NO3 −, HPO4 2−, and SO4 2−) ions is shorter (ca. 35 min) when using the Metrosep A Supp 7 column as compared with the IonPac AS9-HC column (ca. 60 min) (3.6 mol/L Na2CO3 eluent in both cases). The described analytical procedure is applicable to direct analysis of alcoholic commercial products for the content of formates. Minimizing of the amount of alcoholic matrix by preliminary careful evaporation of the sample under IR light source improves baseline separation of formate ion signal. The evaluated amounts of formate ions in the examined rum samples were about 15 times higher (ca. 3 mg/L) than in vodka samples (ca. 0.2 mg/L) examined.