Rakı is a double distilled, traditional aniseed spirit, one of the most consumed alcoholic beverages produced in Turkey (Darici et al., 2019). Rakı was accorded Protected Designation of Origin (PDO) by the Turkish Patent and Trademark Office in 2009 (TPI, 2009). In accordance with this PDO document and the Turkish Food Codex Communiqué on Distilled Alcoholic Beverage (TFC, 2017), Rakı is produced by distilling Suma (grape distillate) with aniseed (only Pimpinella anisum) in a traditional copper pot still. Suma is a distillate of grapes/raisins, which are distilled to up to 94.5 % v/v alcohol by column still distillation with the purpose of retaining the flavour and smell of grapes. Rakı is also one of the most adulterated alcoholic beverages in Turkey. It is often mixed with ethyl alcohol, methanol, water, sugar, and aniseed aroma to increase the profit margin. Adulteration with methanol specifically can cause severe health issues, such as blindness, and can even result in death (Ghadirzadeh et al., 2019). In 2020, 33 people lost their lives in Turkey due to consumption of counterfeited Rakı adulterated with methanol. While the addition of methanol to alcoholic beverages can cause many health problems, other types of chemical and metal contamination can be carcinogenic to humans, and their high consumption can cause the same or worse problems as methanol does (Lachenmeier et al., 2011).

Ethyl carbamate (EC) (C₂H₅OCONH₂) has been classified by the International Agency for Research on Cancer (IARC) as a Group 2A agent, being carcinogenic to animals and possibly to humans (IARC, 2007). Ethyl carbamate can be formed through multiple pathways, and it forms naturally in fermented products including distilled spirits (Figure 1). The most common pathway for EC formation is the reaction between ethanol and N-carbamyl compounds, such as urea produced from arginine metabolism by yeast or lactic acid bacteria (Choi et al., 2018). EC can also be formed during distillation and 48 hours after the distillation is complete (Bruno et al., 2007). Because of its carcinogenicity, EC has attracted attention in many countries, and its concentrations in alcoholic beverages have been limited (Table 1). However, the Turkish regulatory agency has not yet established the limits for ethyl carbamate, and no data is available for Turkish Rakı. Rakı is one of the most consumed distilled spirits in Turkey and the most prone to adulteration; therefore, the determination of EC levels in Rakı is needed. EC is commonly determined by GC-MS, UPLC-MS and HPLC with a fluorescence detector (Riachi et al., 2014). To our best knowledge, the present study contains the first reported data for EC levels in Rakı.

Table

Table 4 shows the levels of ethyl carbamate (EC) in commercial Rakı. None of the Rakı samples contained EC. This may be due to the double distillation of Rakı, which can reduce EC levels by up to 80 % (Correa et al., 2014). For instance, in whisky, only 1 % of EC can be formed during the second distillation (Riffkin et al., 1989). Moreover, high concentrations of EC are generally found in stone-fruit alcoholic beverages produced from, for example, sour cherry and plum, because the process of breaking stones can initiate the formation of amygdalin-derived ethyl carbamate (Deak et al., 2010). Brazilian distilled spirit Cachaça produced from sugar cane was reported to have a high level of EC: as much as 5589 µg/kg, which is above the permitted limit of 150 µg/kg (Andrade-Sobrinho et al., 2002). A distilled spirit in China, Luzhou-flavour, had EC concentrations of between 17 and 151 µg/kg (Fang et al., 2018).

Table

titre du tableau
Samples	Ethyl carbamate (µg/kg)	Furfural(mg/L)	Ethanol(%)	Methanol(mg/100 ml)
R1	ND	ND	45.18 ± 0.12	34.17 ± 1.22
R2	ND	ND	45.16 ± 0.06	30.73 ± 0.56
R3	ND	ND	49.90 ± 0.24	24.69 ± 0.35
R4	ND	ND	42.88 ± 0.13	23.86 ± 0.11
R5	ND	ND	45.05 ± 0.22	38.12 ± 0.54
R6	ND	ND	47.15 ± 0.22	39.96 ± 0.35
R7	ND	ND	43.14 ± 0.15	37.54 ± 0.39
R8	ND	ND	45.07 ± 0.23	37.73 ± 0.44
R9	ND	ND	45.31 ± 0.25	39.31 ± 0.36
R10	ND	ND	44.96 ± 0.29	38.60 ± 0.87
R11	ND	ND	50.15 ± 0.14	25.52 ± 0.36
R12	ND	ND	45.03 ± 0.23	43.20 ± 0.14
R13	ND	ND	44.93 ± 0.21	23.87 ± 0.36
R14	ND	ND	45.11 ± 0.20	29.45 ± 0.55
R15	ND	ND	41.89 ± 0.33	31.90 ± 0.33
R16	ND	ND	40.18 ± 0.12	25.68 ± 0.21
R17	ND	ND	47.72 ± 0.30	42.82 ± 0.39
R18	ND	ND	45.54 ± 0.18	34.23 ± 0.11
R19	ND	ND	45.24 ± 0.20	40.49 ± 0.55
R20	ND	ND	45.10 ± 0.12	35.76 ± 0.29
R21	ND	ND	45.16 ± 0.16	52.63 ± 0.68
R22	ND	ND	44.95 ± 0.24	28.78 ± 0.23
R23	ND	ND	45.04 ± 0.08	45.38 ± 0.74
R24	ND	ND	45.22 ± 0.06	16.52 ± 0.19
R25	ND	ND	42.25 ± 0.08	16.35 ± 0.21
R26	ND	ND	45.02 ± 0.06	36.47 ± 0.54
R27	ND	ND	43.25 ± 0.19	14.55 ± 0.25
R28	ND	ND	47.14 ± 0.09	35.47 ± 0.39
R29	ND	ND	45.13 ± 0.24	12.36 ± 0.87
R30	ND	ND	45.10 ± 0.18	13.40 ± 0.65
R31	ND	ND	43.22 ± 0.21	43.85 ± 1.33
R32	ND	ND	43.06 ± 0.27	16.46 ± 0.78
R33	ND	ND	45.32 ± 0.11	46.22 ± 1.28
R34	ND	ND	45.12 ± 0.17	36.51 ± 0.56
R35	ND	3.97 ± 0.02	40.46 ± 0.11	ND
R36	ND	3.79 ± 0.02	56.91 ± 0.25	27.38 ± 0.36
R37	ND	2.06 ± 0.09	54.28 ± 0.32	48.85 ± 1.55

Furfural (FR) levels in commercial Rakı are shown in Table 4. The range of FR levels was between undetectable and 3.97 mg/L, which is below the generally permitted limit (50 mg/L). Furfural contamination is found when the distillation continued after the tail part of the wine is reached. Therefore, carrying out the production steps carefully may help to reduce contamination. Commercial Rakı samples did not contain furfural, but three counterfeited Rakı samples had low amounts. In literature, many distilled spirits have been evaluated for their furfural concentrations; for example, sugarcane spirits from Brazil and Mozambique were found to contain 1.6–17 mg/L (Bortoletto and Alcarde, 2013) and 2.8–270 mg/L (Tabua et al., 2018) respectively, and brandy and cognac were found to contain 0.5–82.5 mg/L (Tsakiris et al., 2016) and 6.63 mg/L (Awad et al., 2017) respectively. Our results show that commercial Rakı do not contain furfural, but adulterated ones have a low amount of furfural contamination. The ethanol levels of the Rakı samples are shown in Table 4. Commercial Rakı samples complied with regulations and the label. However, two Rakı samples, R36 and R37, which were not legally produced, had very high ethanol concentrations. Similarly, the methanol levels of the commercial Rakı complied with regulations. Only sample R35, a counterfeit Rakı, did not contain methanol since it was produced using an agricultural ethanol purified from some other alcohol.

Table 5 shows the metal contamination of Rakı samples. None of the samples contained arsenic, except for samples R34 and R35, which contained 0.018 mg/L and 0.022 mg/L respectively; these arsenic concentrations can be considered negligible. The absence of arsenic may be due to the arsenic-free copper alembics that was used during the production process. While arsenic is not common in distilled spirits, copper is problematic; it can, for example, catalyse the formation of ethyl carbamate, and the consumption of a high concentration of copper can produce toxic effects, such as diarrhea, excessive vomiting and liver damage (Silva et al., 2020). In the present study, the concentrations of copper in the Rakı samples were low: between 0.046 and 0.414 mg/L, which is also below the limits for distilled spirits (i.e., 5 mg/L). Navarro (2007) reported copper concentrations in whisky, gin, rum, liquor, brandy, wine, and beer as being 1.01 mg/L, 0.1 mg/L, 2.34 mg/L, 0.59 mg/L, 8.01 mg/L, 0.39 mg/L and 0.39 mg/L respectively. Another toxic metal, Pb, was found at ppb levels in the commercial Rakı, most likely due to lead-free copper alembics. The permissible limit of Pb in alcoholic spirits in some European countries is in the range of 0.2–0.3 mg/L (Kostic et al., 2010). Zinc concentrations of the Rakı was in the range of 0.01–0.39 mg/L. The permissible limit for zinc is 5 mg/L, thus Zn concentrations in the commercial Rakı were below the limit. The absence or a very low amount of zinc in the copper alembics could explain these results.

Table

titre du tableau
Samples	As (mg/L)	Cu (mg/L)	Pb (mg/L)	Zn (mg/L)
R1	<0.004	0.069	0.002	0.018
R2	<0.004	0.077	0.004	0.025
R3	<0.004	0.093	0.004	0.035
R4	<0.004	0.086	0.005	0.041
R5	<0.004	0.076	0.003	0.030
R6	<0.004	0.130	0.008	0.131
R7	<0.004	0.101	0.006	0.042
R8	<0.004	0.118	0.005	0.114
R9	<0.004	0.059	0.006	0.023
R10	<0.004	0.173	0.008	0.045
R11	<0.004	0.076	<0.002	0.020
R12	<0.004	0.161	0.008	0.024
R13	<0.004	0.101	0.004	0.033
R14	<0.004	0.074	0.003	0.022
R15	<0.004	0.066	0.004	0.024
R16	<0.004	0.073	0.003	0.045
R17	<0.004	0.079	0.006	0.046
R18	<0.004	0.138	0.003	0.032
R19	<0.004	0.199	0.003	0.032
R20	<0.004	0.130	0.006	0.041
R21	<0.004	0.414	0.005	0.058
R22	<0.004	0.253	0.026	0.392
R23	<0.004	0.310	0.007	0.157
R24	<0.004	0.060	0.003	0.020
R25	<0.004	0.153	0.010	0.076
R26	<0.004	0.242	0.005	0.029
R27	<0.004	0.050	0.003	0.019
R28	<0.004	0.071	0.003	0.011
R29	<0.004	0.077	0.004	0.026
R30	<0.004	0.046	<0.002	0.031
R31	<0.004	0.183	0.002	0.030
R32	<0.004	0.102	0.006	0.053
R33	<0.004	0.169	0.005	0.030
R34	0.0181	0.115	0.004	0.051
R35	0.0217	0.075	0.004	0.022
R36	<0.004	0.064	0.004	0.062
R37	<0.004	0.075	0.004	0.045

This study aimed to determine the chemical and metal contamination of commercial Rakı. Our results showed that neither ethyl carbamate nor furfural were detected in the commercial Rakı. The arsenic, copper, lead and zinc levels of the commercial Rakı were either below the limits or undetectable. In conclusion, the chemical and metal contamination of commercial Rakı is very low.

The authors would like to thank Prof. Luis E. Rodriguez-Saona (The Ohio State University, Department of Food Science and Technology) for his technical support during this study. The authors would also like to thank M.Sc. Koray Ozcan for providing Rakı samples.