Formate Emission in the Mainstream Aerosols of Heated Tobacco Products Distributed in Japan

Abstract

Heated tobacco products (HTPs) are newly developed nicotine delivery systems via the inhalation of mainstream aerosols generated during the heating of tobacco leaf materials. Previous studies have shown that the amount of chemicals generated is much lower than that generated by conventional combustible cigarettes. However, little attention has been paid to formate, a conjugated base of formic acid with potentially toxic effects on human health. This study aims to understand the actual emission levels and behaviour of formate in mainstream aerosols produced by commercially available HTP devices in Japan. Aerosols were generated from four types of devices with regular and menthol-type flavours using a vaping machine following the CRM 81 puffing protocol. Formate was tapped in 5 mM sodium carbonate solution and subsequently analysed using ion chromatography. The results showed that the total emission amount of formate ranged from 0.0027 ± 0.0031 to 0.27 ± 0.055 mg L⁻¹, varying with heating temperature and flavour type. Moreover, the majority of formate existed in a particulate form due to the weak-basic property of the aerosol, and the formate emission level was much greater than the workplace exposure limit for the direct inhalation of mainstream aerosols. The formate in the mainstream aerosol can be considered a health concern, when using “high-temperature type” HTPs over a long period.

1. Introduction

The adverse health effects of cigarette smoking are well recognised. Most smoking-related diseases are not related to nicotine [1] but are associated with other toxic chemicals generated by the combustion of tobacco leaves. Tobacco harm reduction is becoming a public health approach to reduce the impact of cigarette smoking; switching smokers to less harmful tobacco products could reduce some of the harm associated with smoking cigarettes [2]. The possible alternative includes oral nicotine products and electronic nicotine delivery systems (ENDS) such as electronic cigarettes (e-cigarettes) and heated tobacco products (HTPs) [2,3]. Oral nicotine products deliver nicotine mainly by absorption through the user’s oral mucosa [2]. Meanwhile, ENDS are electrically operated devices that deliver nicotine in a manner similar to conventional cigarettes, potentially offering a non-combustible alternative for smokers [2,3]. However, in Japan, e-cigarette liquids which contain nicotine itself are prohibited to use by law, so the HTPs are being used as an acceptable alternative for smokers [4]. The HTP, also referred to as a heat-not-burn tobacco product, is a type of ENDS that delivers nicotine into the body by generating mainstream aerosols during the heating of processed tobacco-leaf materials [5]. Recent comprehensive chemical analyses have shown that HTPs also generate numerous types of chemicals during heating, but the amount of generated chemicals is much less than that of conventional combustible cigarettes [6,7,8,9], mainly because of the lower heating temperature without combusting tobacco. Thus, HTPs are promoted as less hazardous or harmless products by tobacco industries. Although there is little evidence that the short-to-medium term use of HTPs causes major harm to users, the effects of long-term use are still uncertain. Therefore, it is necessary to carefully investigate the aerosol components, including unknown and unidentified substances.

Formate (IUPAC name: methanoate) is a conjugate base of formic acid—the simplest carboxylic acid—with a potentially toxic effect on human health. Formic acid inhalation causes eye and nose irritation, sore throat, cough, chest tightness, headache, and confusion [10]. It is also an intermediate in methanol poisoning. Methanol has low toxicity, and in the human body, it is first metabolised to formaldehyde by alcohol dehydrogenase, and then converted to formate [11]. Formate accumulates in the body and causes visual damage, optic nerve injury, abdominal problems, nausea, and headaches [11]. Tobacco leaves contain several organic acids, including formic acid, which contribute to leaf quality wherein the taste and aroma of tobacco products are closely linked to their organic acids [12,13]. Therefore, these organic acids are found in mainstream aerosols of combustible cigarettes [14]. However, despite their potential significance, formate generated by HTPs has received little attention. The mainstream aerosol generated by the HTPs consists primarily of “water droplets”, which contain glycerine and/or propylene glycol that aid in aerosol forming [15,16]. Because formic acid is a water-soluble weak acid, it is presumed to be easily soluble in water droplets to form formate with counter cations such as protonated nicotine and ammonium ions.

Bentley et al. [9] conducted a comprehensive untargeted chemical analysis of aerosols generated from the Tobacco Heating System 2.2 (commercial name: IQOS) using two-dimensional gas chromatography with time-of-flight mass spectrometry (GC×GC-TOFMS) and liquid chromatography with high-resolution accurate mass spectrometry (LC-HRAM-MS), and demonstrated the emission amounts of over 500 chemicals including formic acid. While this information on formate emissions is valuable, it pertains to only one product among HTPs. It has been noted that the emission levels of trace hazardous chemicals such as carbonyl compounds, tobacco-specific nitrosamines, and polycyclic aromatic hydrocarbons, in the mainstream aerosols can vary significantly depending on the device used [17]. At present, Japan is the largest HTP-consumer country, with a variety of devices which employ different heating temperatures ranging from 40 to 350 °C and a broad range of consumable tobacco sticks with numerous flavours [17]. Given that users select their preferred product from a diverse range of HTPs, the range of formate emission levels should be determined to evaluate the public health risk associated with exposure to the chemical in mainstream aerosols. In this study, we aimed to grasp the actual emission levels and behaviour of formate in mainstream aerosols produced by commercially available HTP devices in Japan by comparing them with those of acetate (IUPAC name: ethanoate), which is one of the most abundant chemicals found in mainstream HTP aerosols [9]. The results showed that formate in the mainstream aerosol can be considered a health concern when using ‘high-temperature type’ HTPs over a long period.

2. Materials and Methods

2.1. Heated Tobacco Products

Mainstream aerosols were collected from four HTP types: A–D. All materials were purchased from a retail tobacco store in Tokyo, Japan.

Table 1. Heated tobacco products used in this study and amount of puffs for the collection of formate in mainstream aerosols.

HTP	Heating Temp. (°C)	Flavour Type	Abbreviation	Puff Number
A	350	Regular	AR	10
		Menthol	AM
B	350 (maximum)	Regular	BR	10
		Menthol	BM
C	200	Regular	CR	10
		Menthol	CM
D	40	Regular	DR	50
		Menthol	DM

summarises the devices, their heating temperatures, flavour types, and the corresponding abbreviations used here. The heating temperature was obtained from a commercial catalogue or technical information provided by each manufacturer. Devices A–C were categorised as a “high-temperature type” and device D as a “low-temperature type”.

Device A consists of a charger, holder, and consumable tobacco stick. After charging, the tobacco stick is inserted into the holder which has a flat heating blade. The blade-heated tobacco leaves are impregnated with aerosol formers in the tobacco stick. The aerosol is generated during the direct heating of the tobacco leaves with the blade at ~350 °C. The configuration of Device B was the same as that of Device A; however, the heating method is different. The tobacco stick has a metal heating element placed in its core. The aerosol is generated by the direct heating of the tobacco leaves from inside to premiere by the inductively heated metal element at temperatures up to ~ 350 °C. Device C comprises a rechargeable battery, a heating furnace in the battery body, and a tobacco stick. The direct heating of tobacco leaves inserted in the furnace produces the mainstream aerosol at ~200 °C. Device D consists of a rechargeable battery, cartridge, and tobacco capsule. The aerosol is generated by heating the liquid in the cartridge containing aerosol formers and passing this through the tailor-made tobacco capsule at ~40 °C. Each device has a variety of flavour sticks or capsules, with regular and menthol (IUPAC name: 2-isopropyl-5-methylcyclohexanol) used here. These devices and tobacco sticks used served for the test promptly after purchase.

2.2. Measurement of Formate and Acetate

The measurement of formate and acetate was conducted in the chemical laboratory of Shonan campus, Tokai University in Kanagawa, Japan, from June to October 2022. As shown in Figure 1, the mainstream aerosols were generated from all devices using an LM4E Linear Vaping Machine for E-cigarettes (Borgwaldt KC, GmbH, Hamburg, Germany), following the puffing regime specified in the Cooperation Centre for Scientific Research Relative to Tobacco (CORESTA) Recommended Method 81 (CRM81) [18]—55 mL puff volume, 3 s puff duration, 30 s puff interval, and a “Rectangle” type puffing profile. The vaping machine was operated in a separated booth with a local exhaust system at room temperature (~24 °C) and 34% relative humidity.

For the formate and acetate measurements in the aerosols, two glass impingers (G-1 type, Sibata Scientific Technology, Tokyo, Japan) filled with trapping solution (15 mL) were connected in series to the outlet of the vaping machine with a silicon tube, with or without a Cambridge filter pad (44 mm diameter, Whatman, Buckinghamshire, UK), which was used for the separation of the gas and particle phases [18,19]. As a trapping solution, 5 mM aqueous sodium carbonate (Na₂CO₃) solution—the ion chromatography effluent—was used to reduce the water dip in the ion chromatograms. Ten puffs were each collected from devices A, B, and C. In the case of the “low-temperature type” device D, the amount of formate collected at 10 puffs was below the detection limit in the preliminary test. Thus, the puffing number was set at 50 for this device only (Table 1).

The formate and acetate contents were subsequently determined using ion chromatography after filling up to 15 mL with 5 mM Na₂CO₃. Ion chromatography was performed using a Thermo Fisher Scientific Dionex Aquion IC system with a chemical suppressor. The following condition was used: separation column, 4.0 mmφ × 250 mm, Dionex AS-9-HC Analytical (Thermo Fisher Scientific, Waltham, MA, USA); guard column, Dionex AS-9-HC Guard (Thermo Fisher Scientific, Waltham, MA, USA); eluent, 5 mM Na₂CO₃ at 1.0 mL min⁻¹ (isocratic); and regenerant, 1.5 mM sulfuric acid (H₂SO₄). Dilution series of formate and acetate ranging from 0.0 to 1.0 mg L⁻¹ in Milli-Q ultrapure type1 water (18.2 MΩ·cm at 298 K, Simplicity^® water purification system, Merck KGaA, Darmstadt, Germany) were prepared from their sodium salts (HCOONa and CH₃COONa) and used for calibration (r > 0.99 for each concentration versus peak area). All the reagents were purchased from Kanto Chemicals (Tokyo, Japan).

Three repeated measurements were conducted for all the runs. Blank samples were collected simultaneously without connecting HTPs. After subtracting the mean blank reading of the storage from sample readings, the collection amounts of formate and acetate (mg) were converted to emission amounts in a puff volume E (mg L⁻¹) using the following equation:

E = W/V

(1)

where W is the collection amount of formate or acetate (µg) and V is the total puff volume (0.55 L at 10 puffs and 2.75 L at 50 puffs). The limit of determination was defined as 10-fold the standard deviation of the blank reading and resulted in 3.9 and 0.78 µg L⁻¹ at 10 and 50 puffs, respectively, for formate, and 5.1 and 1.0 µg L⁻¹ at 10 and 50 puffs, respectively, for acetate.

The percentage of the particulate form of acetate and formate, P_p (%), was quantified using the following equation:

P_p = (E_total − E_g)/E_g × 100

(2)

where E_total is the total emission amount of analytes collected without the Cambridge filter pad and E_g is the emission amount of gaseous forms of analytes collected with the Cambridge filter pad (the particulate form was removed by the filter pad). Accordingly, the percentage of the gaseous form of acetate and formate, P_g (%), was obtained using Equation (3).

P_g = 100 − P_p

(3)

2.3. Statistical Analysis

Experimental data on the emission amount of acetate and formate were presented as the arithmetic mean and standard deviation. Statistical analyses were performed on Microsoft^® Excel 2021 MSO 64 bit for windows. Differences in the emission amounts of formate and acetate between regular and menthol type flavour types were analysed using Welch’s t-test (unequal variances t-test) for each device. The degrees of freedom and the t value were calculated for each data set, and the p value was obtained from the t-distribution table. Statistical significance was accepted at p < 0.05, with 0.05 ≤ p ≤ 0.1 interpreted according to guidelines by Curran-Everett and Benos [20].

3. Results

3.1. Emission Amounts of Formate and Acetate

Acidic substances from the mainstream aerosols of HTPs were collected in two serially connected impingers without installing a Cambridge filter pad. Significant peaks of acetate and formate—with retention times of 4.9 and 5.4 min, respectively—were observed in all HTP samples, whereas the peaks of chloride, nitrate, and sulphate ions were almost absent (blank levels). Acetate and formate were also found in the second impinger, resulting in collection efficiencies of 84 ± 1.9 and 81 ± 4.7% for acetate and formate, respectively, in all runs (n = 12). The emission amounts of both organic acid ions were then determined for all samples.

The arithmetic means of the emission amounts of acetate collected from all devices are shown in Figure 2. Error bars indicate the standard deviation of triplicate runs. The amount of acetate emitted varies with the device and flavour type. Bentley et al. [9] conducted a comprehensive chemical analysis of aerosols from the Tobacco Heating System 2.2 and reported that the emission amount of acetic acid was 944 µg item⁻¹, which corresponds to 1.5 mg L⁻¹. Even though the devices may not be exactly the same and storage conditions during distribution are not controlled in this study, the “high-temperature type” devices showed equivalent levels of acetate emission—AR: 1.1 ± 0.20 mg L⁻¹, BR: 1.6 ± 0.11 mg L⁻¹, and CR: 1.5 ± 0.084 mg L⁻¹. The acetate emission from the “low-temperature type” device, DR, was much less (0.13 ± 0.040 mg L⁻¹). Meanwhile, except for AR and AM, significant differences in acetate emissions were found between regular and menthol flavour types with p values less than 0.05 (Welch’s t-test), probably because of the differences in the ingredients of the tobacco leaf materials.

The arithmetic means of the emission amounts of formate collected from all devices are shown in Figure 3. Error bars indicate the standard deviation of triplicate runs. The emission amount of formate increased with an increase in the heating temperature—AR: 0.27 ± 0.055 mg L⁻¹; BR: 0.19 ± 0.071 mg L⁻¹; CR: 0.048 ± 0.20 mg L⁻¹; and DR: 0.0027 ± 0.0031 mg L⁻¹. The levels were lower than those of acetate. The emission levels of the regular type were significantly greater than those of the menthol type for devices A and C (p < 0.05), and a greater emission of the regular type was suggested for device B (p = 0.090). No significant difference was observed between DR and DM, mainly because of their lower emission levels. This means that the use of HTPs with a higher heating temperature and regular flavour results in the inhalation of more formate into the body.

3.2. Gas to Particle Distribution of Formate and Acetate

Owing to the acid-base equilibrium of nicotine, the basic environment of aerosols is favourable for nicotine absorption into the body [21,22]. Actually, the “water droplets” generated from HTPs were weak-basic [19], so that acetate and formate are likely to form salts and exist as particulate matter. The acetate and formate emissions from the selected devices BR, BM, and DM were measured using a Cambridge filter pad which trapped the particulate species emitted from these devices, and the percentages of the particulate form, P_p, and those of gaseous form, P_g, were obtained for acetate and formate using Equations (2) and (3). Figure 4 shows the results in pie charts. Gaseous acetate was found in the samples of BR and BM, but not in that of DR; thus, the percentage of particulate acetate was 95, 92, and 100% for BR, BM, and DR, respectively. Similarly, gaseous formate was found in the samples of BR and BM, but not in that of DR. The percentage of particulate formate resulted in 99, 96, and 100% for BR, BM, and DR, respectively. These results indicate that the formic acid in the mainstream aerosol of HTPs was distributed mainly in the particle phase, as was predicted.

4. Discussion

We detected formate in the mainstream aerosols of HTPs distributed in Japan and found that their emission amounts depended on the heating temperature and varied depending on the type of flavour. Here, we need to pay attention to the validity of formate emission levels shown in Figure 3, because Bentley et al. [9] reported the emission amount of formic acid as 0.233 µg item⁻¹, corresponding to 0.00035 mg L⁻¹, which is much less than the values obtained here. The difference in trapping and/or analytical methodologies is a possible explanation. Thus, we also quantified the acetate and formate in the trapping solution of AR using a reverse-phase HPLC method with UV detection (detection wavelength: 220 nm). The results of the HPLC method showed equivalent levels of both compounds: 1.0 ± 0.13 and 0.23 ± 0.13 mg L⁻¹ for acetate and formate, respectively (n = 3). Therefore, the notable differences in results between the previous study [9] and this work may lie in the analyte trapping methodology.

The inhalation of formic acid vapour is known to severely irritate the mucous membranes of the nose and mouth, leading to inflammation. The inhalation of aerosolised formic acid also causes inhalation injuries [23]. Even though the formate mostly existed as a particulate form in HTP aerosols, the emission amounts from “high-temperature type” devices ranging from 0.028 ± 0.0039 mg L⁻¹ (CM) to 0.27 ± 0.055 mg L⁻¹ (AR) were much greater than the workplace exposure limit of 9 mg m⁻³ (or 0.009 mg L⁻¹) recommended by the US National Institute for Occupational Safety and Health (NIOSH) [10] and other organisations. Thus, the inhalation of formate from HTPs is a health concern, especially with long-term use. Similarly, we should note that the formate emitted from the exhaled breath of a user can be a source of air pollution. Utilising the “low-temperature type” devices or regulating the heating temperature to lower levels can effectively suppress the formate emission in mainstream aerosols, thereby mitigating the direct exposure to users and the impact of second-hand smoke from exhaled air. However, HTP users tend to favour high-temperature devices, which enable higher nicotine consumption and a richer flavour experience. Consequently, it is anticipated that the low-temperature heating devices will be phased out from the market in Japan. To address this potential health risk of formate emissions from “high-temperature type” HTPs, it is necessary to elucidate formate emission mechanisms for further improvement of products.

Formate and other organic acids are originally present in tobacco leaves [12]. Therefore, the transition from tobacco leaf material to water droplets during heating could be an emission route for formate from HTP products. However, despite the fact that formate and acetate have similar chemical properties, the ratio of formate and acetate emissions was not constant between the HTP devices, and the ratio tended to increase with the heating temperature, as shown in Figure 5. This suggests the existence of another formate emission route.

Glycerin and/or propylene glycol are used as HTP aerosol formers due to their hygroscopic property [15,16]. However, they are not necessarily stable, and the thermal decomposition of these ingredients is known to initiate the formation of formaldehyde and other aldehydes [24] (Figure 6). Formaldehyde is labile for further oxidation and thus produces formic acid. Formic acid can subsequently dissolve in water and is emitted mainly as formate in the particulate phase. Methanol can also be a precursor of formaldehyde in aerosols because a certain amount of methanol (0.32 mg L⁻¹) was found in the aerosol of the IQOS [9].

A limitation of this study is that the HTPs used were purchased from distribution, and the storage conditions from manufacturing and transportation to purchase were not controlled. Particularly for imported products, the transportation period may be long, and it is necessary to investigate the influences of temperature and humidity on emission levels in association with the condition of tobacco leaf materials during storage. In addition, formate was deduced to be a by-product of aerosol formers and other ingredients. However, because the composition of the raw materials was not disclosed, it was not possible to fully consider the formate precursors. Formate is a highly hazardous chemical that can be considered a health concern, considering the emission amounts found in this study. In the future, we aim to ask tobacco manufacturers to cooperate and conduct tests that consider storage conditions and ingredients in tobacco leaf materials.

5. Conclusions

Today, Japan is the largest HTP consumer country, and there are some different HTP devices and a broad range of consumable tobacco sticks. Our objective was to understand the actual emission levels and behaviour of formate in mainstream aerosols produced by commercially available HTPs in Japan, in order to evaluate the health risk associated with exposure to this potentially harmful chemical. The results showed that the total amount of formate emitted increased with increasing heating temperature and varied with flavour type. The majority of formate exists in particulate form owing to the weak basic properties of the aerosol. It should be noted that formate emission levels were much greater than the workplace exposure limit when mainstream aerosols were directly inhaled. Therefore, formate in mainstream aerosols is a concern for human health when using the “high-temperature type” HTPs over a long period. Further studies are required to elucidate formate emission mechanisms for the safe use of HTPs.