Characterization of E-cigarette coil temperature and toxic metal analysis by Infrared Temperature Sensing and Scanning Electron Microscopy- Energy Dispersive X-Ray

Abstract

Introduction:

Electronic cigarettes (e-cigarettes) have rapidly evolved since their introduction to the U.S. market. The rebuildable atomizer (RBA) offers user-driven modification to the heating element (coil) and wicking systems. Different coil materials can be chosen, based on user needs and preferences. However, the heating element of an e-cigarette is believed to be one-source for toxic metal exposure.

Methods:

E-cigarette coils from Kanthal and nichrome wires were constructed in a contact and non-contact configuration and heated at four voltages. The maximum temperatures of the coils were measured by infrared temperature sensing when dry and saturated with 100% vegetable glycerin or 100% propylene glycol. The metal composition of each coil was analyzed with Scanning Electron Microscopy-Energy Dispersive X-Ray (SEM-EDX) when new, and subsequently after 1, 50, and 150 heat cycles when dry.

Results:

The coils reached temperatures above 1000 °C when dry, but were below 300 °C in both liquid-saturated mediums. Metal analysis showed a decrease of 9–19% chromium and 39–58% iron in Kanthal wire and a decrease of 12–14% iron and 39–43% nickel in nichrome wire after 150 heat cycles. Significant metal loss was observed after one heat cycle for both coil alloys and configurations.

Conclusions:

The loss of metals from these heat cycles further suggests that the metals from the coils are potentially entering the aerosol of the e-cigarette, which can be inhaled by the user.

Introduction

Electronic cigarettes (e-cigarettes) have rapidly evolved since their introduction to the U.S. market in 2007. All e-cigarettes operate under the same, basic principle. The device consists of two main components: a battery power supply and an atomizer that contains a coil and wicking system (Breland et al. 2016). The coil is typically a metal filament wrapped around a wicking material of silica or cotton, saturated in a refill formulation liquid (e-liquid). The e-liquid consists of a ratio of propylene glycol and vegetable glycerin, nicotine, and usually, a flavoring agent (Peace et al. 2016). When the battery is activated, it delivers an electric charge to the coil, which heats and aerosolizes the e-liquid. The generated aerosol is then inhaled by the user (Breland et al. 2016; Peace et al. 2016).

The first e-cigarette, designed in China in 2004, resembled a traditional cigarette and was known as the “ciga-like”. Future generations of e-cigarettes consisted of the cartomizer (a cartridge atomizer) and clearomizer (a transparent tank cartomizer), which contained replaceable tank-and-atomizer systems (Breland et al. 2016). A popular generation of e-cigarette is known as the rebuildable atomizer (RBA), which allows for complete user-driven customizability. The intent of the RBA is to enable the user to modify the device to create a nicotine delivery system that an e-cigarette consumer would find satisfying (Grana et al. 2013). Users take advantage of adjustable power supplies and coil configurations to impact heat production with the desire to maximize drug delivery.

The use of rebuildable atomizers caused polarizing discussions about what components will create the best “build” and therefore, the best vaping experience. Early e-cigarette models used nichrome 80:20 wire for coil builds in commercial atomizers (William et al. 2013). The metal is comprised of nickel and chromium and has a maximum operating temperature of 1200 °C (Consolidated… 2020). Kanthal A-1 is a resistance wire that grew popular with RBA builders in the vaping community. The metal is comprised of iron, chromium, and aluminum alloys and has a maximum operating temperature of 1400 °C (Kanthal Datasheet… 2018). With wire composition preferences came choices in wire configurations. Generally, the coils in the e-cigarette can be wrapped into two configurations: contact or non-contact. The contact build has the coil wraps touching together and the non-contact build has the coil wraps separated in space. Despite the numerous ways to create an e-cigarette coil, there is no definitive answer on which type of coil build provides the best user experience and most efficient drug delivery.

Potential toxic metal exposure from inhaling e-cigarette aerosols is of concern with the use of metallic filaments as the heating element. Chromium and nickel, two major metal alloys present in nichrome wire are listed as carcinogens and are linked with chronic lung infections and other respiratory illnesses such as reduced lung infections and bronchitis (ATSDR 2005, 2013). Chromium is also present in Kanthal wire. Some studies for iron oxides also suggest that inhalation over time can lead to lung cancer (Kornberg et al. 2017). Early studies assessed e-cigarette aerosol for metals and identified tin, silver, iron, nickel, aluminum, and cadmium present in the aerosol (Williams et al. 2013; Hess et al. 2017). Several studies highlighted a (Hess et al. 2017; Williams et al. 2019; Halstead et al. 2019; Zervas et al. 2020) variability in metal concentrations in aerosols between e-cigarette brands, potentially resulting from the different heating units within the devices, device age, and access of the liquid to metal. One study assessed the metal content present in e-liquids directly from the manufacturer bottle verses the tank and the resulting aerosol. The study determined that there was a higher metal content present in the e-liquid from the tank and the aerosol, suggesting that the coils had an effect on metal concentration (Olmedo et al. 2018). A later study found that factors such as e-liquid composition and boiling temperature of the e-liquids had an effect on the transfer of metal from two commercial coils to the e-liquid aerosolized in the device (Zervas et al. 2020).

The purpose of this study was to characterize two types of common metal wires used to construct coils for electronic cigarettes with typical user modifications. The temperature output of the two coils when in a contact and non-contact configuration under dry conditions and wet conditions when saturated with 100% vegetable glycerin, 100% propylene glycol e-liquids, and a 50:50 mixture was determined using an infrared (IR) temperature sensor. The metal profile of the two coils before and after heating was determined using scanning electron microscopy energy dispersive X-ray (SEM-EDX).

Experimental

Reagents and Supplies

For all studies, a Kayfun Lite Styled Rebuildable Atomizer (RBA) was purchased from FastTech (Hong Kong, China). The Pure-Atomist Kanthal A-1 wire of 30, 32, and 34 American Wire Gauge (AWG) and ceramic tweezers were purchased from Lightning Vapes (Brandenton, FL). The consolidated chromium wire of 30, 32, and 34 AWG were purchased from McMaster Carr (Elmhurst, IL). The 2 mm high quality silica wick and Master 5-in-1 coiling kit were purchased from Amazon (Seattle, WA). The 100% pure vegetable glycerin (VG) and 100% pure propylene glycol (PG) were purchased from Wizard Labs (Almonte Springs, FL). A thermoMETER M# and M1/M2 temperature sensors ranging from 100–600 °C and 650–1800 °C were purchased from Micro-Epsilon (Raleigh, NC). The Sorensen power supply was from Amtek (San Diego, CA) and a multimeter was purchased from Fluke (Everett, WA).

Coil Building

Kanthal A-1 and nickel chromium (nichrome) wires of 30 AWG (254.6 μm), 32 AWG (201.9 μm), and 34 AWG (160.1 μm) were used to wrap all coils in two configurations: contact and non-contact (Figure 1). The coils in each gauge and configuration were tested under two types of conditions: (i) “dry medium” containing no silica wick and no e-liquid and (ii) “wet medium”, which contained the 2mm silica wick and e-liquid. Each coil was constructed using the 1.5 mm diameter rod from the Coil Master 5-in-1 coiling kit and were wrapped to 1.8 Ω. For non-contact configuration, the coils were pulled apart so that the wraps were no longer in contact. The coils were housed in a Kayfun Lite Styled RBA and a Fluke multimeter was used to ensure that every coil type, configuration, and gauge was wrapped to 1.8 Ω. In the contact configuration, the coils were wrapped and placed in the atomizer and then annealed to ensure even heating across the coil before the resistance was tested. For all “wet medium” conditions in the temperature studies, a 2 mm silica wick was threaded through the coil and the atomizer was filled with the e-liquid. E-liquid refill formulations used for the temperature studies were 100% VG, 100% PG, and a 50:50 PG:VG made in-house with no nicotine.

Figure 1:

Electronic cigarette coil wrapped in a contact (A) and non-contact (B) configuration.

Coil Temperature Studies

The atomizers were heated using a Sorensen power supply, which controlled the voltage output. The coils were heated at increasing voltages of 3.5, 3.7, 4.2, and 5.0 V (6.8, 7.6, 9.8, 13.9 W). The trials were completed in triplicate, resulting in three trials for each coil under each voltage, configuration, and dry and wet (100% PG and 100% VG). Both coil materials were then tested under contact and non-contact configurations at 32 gauge in a 50:50 PG:VG solution. Two dual laser infrared temperature sensors by Micro-Epsilon were used to measure the temperature outputs of the coils. A thermoMETER CT Laser M3 with a measuring range of 100–650 °C and spectral range of 2.3 μm was used for all “wet medium” studies. A thermoMETER CT Laser M1/M2 with a measuring range of 650–1800 °C and a spectral range of 1/1.6 μm was used for all “dry medium” studies. The coils were heated for eight seconds, and the maximum temperature at both ends near the pole and at the center of the coil was recorded. The temperature at the three positions was measured ten times. Compact Connect version 1.9.8.6 was used to record the temperatures of the coils as they were heated.

Coil SEM Analysis

The coils analyzed for SEM were heated from the Sorensen power supply at their maximum operating voltage for ten seconds under dry conditions. The coils were heated for one, 50, and 150 times at their maximum operating temperatures and stored in a trace metal free microcentrifuge tubes. The max operating temperatures were determined by the voltage and coil combination that could withstand a ten second burn without melting the coil. The analysis of the coil composition pre and post-heating was conducted using a Hitachi SU-70 Scanning Electron Microscope coupled with Energy Dispersive X-Ray Spectroscopy (SEM-EDX). Samples were mounted on an aluminum disc coated with carbon tape, with the ends taped down and the wraps pulled apart and pressed into the carbon tape. An EDX spectrum was obtained simultaneously with the SEM images. The SEM-EDX was used with an accelerating voltage of 20 keV and the center of the coil was identified and a measurement was made in that center and to the immediate left and right, totaling 12 measurements for each gauge at each configuration at 2000x magnification. The EDX spectrum and weight percent of each metal present in the coil were also obtained. SEM-EDX data for coils at 32 gauge for both coil types and configurations were recorded at the center of the coil for new, one, and50 heat cycles in the presence of a wet media of 50:50 PG to VG.

The Brown-Forsythe test was used to test for equality of variances. A significance level of 0.05 was used. The data that was determined to exhibit equal variance was analyzed using ANOVA. Data with unequal variance was analyzed using Tukey-HSD. Data was analyzed using JMP pro 12 statistical software.

Results

Coil Temperature Studies

In the dry heat studies across the length of the coil, the center of the coils was visibly brighter than the ends in both configurations suggesting that the coils produced the most heat in the center of the coil and was confirmed with 5–10% higher temperature measurements in the center. Kanthal in a contact configuration with the lowest gauge (thickest wire) had the most significant increase in temperature output, from 1134 °C to 1436 °C, with increasing voltage. Wire gauge only significantly impacted temperature at the lower voltages in contact configuration, with the thicker wire exhibiting the lowest starting temperature (1134 °C) at the lowest voltage (3.5 V) (Figure 2). The thinnest wire, 34 gauge, in contact configuration was unable to withstand the heat produced at 5.0 V and burnt out/melted before the maximum temperature could be recorded. Kanthal in non-contact configuration had lower operational temperatures compared to the contact configuration with increasing voltages and wire gauges. Initial temperatures of 30 and 32 gauge wires at 3.5 V, non-contact, and dry were 975 °C and 1016 °C versus 1134 and 1354 °C in contact, respectively. Temperatures at 4.2 and 5.0 V of 34 gauge, non-contact could not be recorded. For nichrome wire with dry heat in a contact configuration, the operational temperature range of 1051–1234 °C was lower compared to Kanthal across the voltages and wire gauges (Figure 2). In a contact configuration, the temperature profile of 34 gauge nichrome could not be recorded even at the lowest voltage due to burnout. In a non-contact configuration, the wire had a temperature range of 911–1234 °C across the voltages and wire gauges. Similar to non-contact Kanthal, the non-contact nichrome wire at 34 gauge could not be recorded at 4.2 and 5.0 V.

Figure 2:

Maximum temperature outputs of Kanthal in contact (C) and non-contact (NC) configurations in three mediums; Dry (A), 100% VG (B), and 100% PG (C). Maximum temperature outputs of nichrome in contact and non-contact configurations in three mediums: Dry (E), 100% VG (F), and 100% PG (G).

For the wet heat studies with 100% vegetable glycerin (VG), the temperature output was greatly reduced compared to the dry heats. Kanthal wire in a contact configuration had a temperature range of 266–298 °C with increasing voltage and wire gauge (Figure 2). In a non-contact configuration, the Kanthal wire had a temperature range of 241–275 °C. Statistical analysis of the two coil configurations determined that the temperature output in a contact verses a non-contact configuration for Kanthal wire was not statistically different (P = 0.453). For nichrome wire in a contact configuration, the temperature of the coils ranged from 253–283 °C across all voltages and gauges. In a non-contact configuration, the nichrome wire had a temperature range of 245–293 °C. Statistical analysis of the two coil configurations determined that the temperature output in a contact vs non-contact configurations for nichrome wire was not statistically different (P =0 447). Comparing the two coil types and the temperature output in Kanthal versus nichrome wire was determined to not be statistically different with varying wire gauge and voltage (P = 0.198).

For the wet heat studies with 100% propylene glycol (PG), the coil temperature output was also greatly reduced compared to the dry heats. Kanthal wire in a contact configuration had a temperature range of 182–386 °C with increasing voltage and wire gauge (Figure 2). In a non-contact configuration, the Kanthal wire had a temperature range of 148–186 °C across all voltages and gauges. Statistical analysis demonstrated that the differences in temperature output of Kanthal in contact and non-contact configurations was statistically significant (P < 0.001). For nichrome wire in a contact configuration, the temperature of the coils ranged from 171–241 °C across all voltages and wire gauges. In a non-contact configuration, the the nichrome wire had a temperature output range of 160–184 °C. Statistical analysis demonstrated that the difference in temperature output of nichrome coils in contact and non-contact configurations was not statistically significant (P = 0.138). The temperature output between coil configurations combined with coil metal was statistically significantly different. Comparing the two coil configurations and the temperature output in Kanthal versus nichrome wire was statistically significant when varying wire gauge and voltage (P = 0.001).

In a 50:50 PG:VG solution, the 32-gauge Kanthal in a contact configuration had a temperature of 197–240 °C, and, in a non-contact configuration the range was 190–227 °C (Figure 3). In a 50:50 PG:VG medium, the Kanthal wire was significantly different from coils heated in 100% vegetable glycerin (P = 0.001), but not significantly different from coils heated in 100% propylene glycol (P > 0.05). For nichrome wire in a contact configuration, the temperatures ranged from 204–240 °C and for a non-contact configuration, 212–240 °C. For both coil types, the temperatures of the nichrome coils heated in a 50:50 PG:VG medium was significantly different than those heated in either 100% VG or 100% PG (P < 0.05).

Figure 3:

Maximum temperature outputs of 32 G Kanthal and nichrome wire in contact (C) and non-contact (NC) configurations in three mediums; 100% VG (A), 50:50 PG:VG (B), and 100% PG (C).

Coil SEM Metal Analysis

When heated dry, the metals that showed the most appreciable loss following heating were iron and chromium for Kanthal A-1 wire and iron and nickel for nichrome wire. The oxygen content on each coil increased with each burn cycle. Metal analysis performed on new unheated coils yielded Kanthal baseline compositions to be 59.29–66.26% iron, 19.72–21.03% chromium, and 3.60–4.79% oxygen. The baseline composition of the nichrome wire was 18.81–19.69% iron, 51.50–54.34% chromium, and 3.44–4.78% oxygen for the three gauges.

With dry heating, chromium composition in the Kanthal wire in contact configuration decreased from 19.72–21.03% to 13.47–15.53% after one heating cycle and to 5.76–8.54% after 50 burn cycles across the three gauges (Figure 4). After 150 heating cycles, the final chromium composition was 4.55%, 2.54%, and 1.86% for 30, 32, and 34-gauge wire, respectively. The iron composition decreased from 59.59–66.62% to 33.52–43.17% after 1 heating cycle and to 14.85–23.50% after 50 heating cycles. After 150 heating cycles, the metal content for nickel was 23.50%, 14.85%, and 22.15% for the three gauges, respectively (Figure 4). The total loss of metal in a contact configuration was between 15.27–19.17% for chromium and 52.48–58.26% for iron. For each metal, a significant difference in metal composition at each heating cycle increment was observed across the three gauges (P < 0.05). When comparing the total metal loss between each wire gauge, for both chromium and iron, there was a significant difference between 30-gauge and the 32 and 34-gauge wires (P < 0.05). There was not a significant difference in 32 and 34-gauge wire for total metal loss for metal for chromium and iron (P = 0.4468, 0.5003).

Figure 4:

Metal loss of Kanthal (A, B, and C) and nichrome wire (D, E, and F) in contact (C) and non-contact (NC) configurations after one, 50, and 150 burns in a dry medium.

When heated dry, Kanthal wire in a non-contact configuration, after 1 heat the metal content for chromium dropped from 19.72–21.03% to 15.86–17.00% across the three gauges and after 50 heats, to 8.97–11.45% (Figure 4). After 150 heats, the final metal content for chromium was 10.34%, 6.98%, and 5.54% for 30, 32, and 34-gauge wire, respectively (Figure 3). For iron, the metal content dropped from 59.59–66.62% to 38.64–48.33% after 1 heat and to 22.98–32.11% after 50 heats. After 150 heats, the metal content for nickel was 26.66%, 18.02%, and 16.64% for the three gauges, respectively (Figure 4). The total loss of metal in a non-contact configuration was between 9.38–15.49% for chromium and 39.96–46.10% for iron. For each metal, there was a significant difference in metal composition at each burn stage across the three gauges (P < 0.05). When comparing the total metal loss between each wire gauge, for both chromium and iron, there was a significant difference between 30 and 32 gauge and 30 and 34 gauge (P < 0.05). There was not a significant difference in 32 and 34-gauge wire for total metal loss for metal for chromium and iron (P = 0.3923, 0.8797). For Kanthal wire, there was a significant difference between the two configurations (contact and non-contact) at each gauge for both chromium and iron (P < 0.005).

Oxygen composition in a contact configuration for the Kanthal wire increased from 3.60–4.74% to 20.15–24.04% after one heating cycle and to 30.64–32.79% after 50 heat cycles across the three gauges (Figure 4). After 150 heating cycles, the final oxygen composition was 34.46%, 32.79%, and 31.29% for 30, 32, and 34 G respectively. The total gain of oxygen in the contact configuration was 30.85–35.09%. Metal composition increased significantly at each heat stage across the three gauges (P < 0.05). When comparing the total oxygen gain between each wire gauge, there was no significant difference (P = 0.6261). In a non-contact configuration, oxygen content increased to 16.39–23.03% after one heat cycle and to 25.30–27.79% after 50 heat cycles. After 150 heat cycles, the final oxygen composition was 26.17%, 28.53%, and 31.06% for 30, 32, and 34 G respectively. The total increase of oxygen in the non-contact configuration was 22.57–26.32%. With all three gauges, there was not a significant difference in the oxygen levels from 50 to 150 heat cycles (P = 0.9393, 0.0819, and 0.2824 for 30, 32, and 34 G respectively). For 34 gauge there was also no significant difference between the oxygen levels from 1 burn to 50 heat cycles (P = 0.0573). When comparing the total oxygen increase between each wire gauge, there was no significant difference (P = 0.2848). For Kanthal wire, there was a significant difference (P < 0.05) between the two configurations (contact and non-contact) at each gauge at 150 heat cycles. For 30 and 32 gauge, there was also a significant difference in oxygen accumulation at 50 heat cycles.

For nichrome wire in a contact configuration, after 1 dry heat, the metal content for iron dropped from 18.81–19.09% to 11.22–16.91% across the three gauges and after 50 burns, to 5.32–8.19% (Figure 4). After 150 dry heats, the final metal content for chromium was 5.57%, 7.01%, and 5.75% for 30, 32, and 34-gauge wire, respectively (Figure 3). For nickel, the metal content dropped from 51.50–54.30% to 21.03–37.41% after 1 heat and to 9.16–14.41% after 50 heats. After 150 heats, the metal content for nickel was 10.84%, 11.51%, and 9.88% for the three gauges, respectively (Figure 4). The total loss of metal in a contact configuration was between 12.05–14.12% for iron and 40.77–43.50% for nickel. For each metal, there was a significant difference in metal composition after the first and fiftieth heat cycles across the three gauges (P < 0.05). However, there was not a significant difference in the loss of metal for nickel and iron from 50 to 150 heat cycles. When comparing the total metal loss between each wire gauge, for both iron and nickel, there was no significant difference between the three gauges (P > 0.05).

For nichrome wire in a non-contact configuration, after 1 heat, the metal content for iron dropped from 18.81–19.09% to 13.04–15.77% across the three gauges and after 50 burns, to 5.70–8.37% (Figure 2). After 150 heats, the final metal content for chromium was 12.13%, 12.87%, and 14.11% for 30, 32, and 34-gauge wire, respectively (Figure 4). For nickel, the metal content dropped from 51.50–54.30% to 21.03–36.99% after 1 heat and to 17.55–9.16% after 50 burns (Figure 4). After 150 heats, the metal content for nickel was 14.76%, 7.59%, and 9.88% for the three gauges, respectively (Figure 3). The total loss of metal in a non-contact configuration was between 12.13–14.11% for iron and 39.58–44.69% for nickel. For each metal, there was a significant difference in metal composition after the first and fiftieth heat cycle across the three gauges (P < 0.05). However, a significant difference in the loss of metal for nickel and iron from 50 to 150 heats was not observed. When comparing the total metal loss between each wire gauge, for nickel there was not a significant difference between 30 and 32-gauge wire (P = 0.5632), but a significant difference between 30 and 34-gauge and 32 and 34-gauge was observed (P < 0.05). For iron, there was a significant difference when comparing 30 gauge wire to 32 and 34-gauge wire (P < 0.05), but not between 32 and 34-gauge wire (P = 0.8105). When comparing contact and non-contact configurations of nichrome wire, a significant difference in metal loss between the two configurations for both iron and nickel was observed. However, for 32 and 34-gauge wire, there was a significant difference for iron loss (P < 0.05), but there was no significant difference for nickel metal loss (P > 0.05).

Oxygen composition in a contact configuration for the nichrome wire increased from 3.44–4.78% to 18.04–27.01% after one heat cycle and to 29.12–30.14% after 50 heat cycles (Figure 4). After 150 heating cycles, the final oxygen composition was 29.99%, 27.36%, and 27.95% at 30, 32, and 34 gauges respectively. The total gain of oxygen in a contact configuration was 23.41–25.21%. With 30 and 32 gauge, there was a significant difference in oxygen levels between new, 1 heat cycle, and 50 heat cycles. (P < 0.05). For 34 gauge contact, there was only a significant difference in oxygen levels between new and 1 heat cycle. There was not a significant difference in the amount of oxygen accumulated between the three gauges (P = 0.5682). For a non-contact configuration, oxygen levels increased to 20.91–25.60% after one heat cycle and 24.12–34.03% after 50 heat cycles. After 150 heat cycles, the oxygen levels were 22.19%, 24.48%, and 24.06% for 30, 32, and 34 gauge respectively. The total increase of oxygen in a non-contact configuration was 17.41–20.62%. For 30 and 34 gauges, there was no significant difference in oxygen levels after one, 50 and 150 heat cycles (P > 0.05). For 32 gauge wire, there was a significant difference between new, one, and fifty heat cycles, but no significant difference between oxygen levels after one heat cycle and 150 heat cycles (P = 0.6427). There was not a significant difference in the amount of oxygen accumulated between the three gauges (P = 0.2968). For nichrome, 30 and 34 gauge wires did not have a significant difference between the two configurations (Contact and non-contact), but 32 gauge had a significant difference in oxygen levels between the two configurations at one heat cycle and 50 heat cycles (P < 0.05).

The significance of the metal loss with increasing heat cycles could be seen visually via SEM analysis. Figure 5 shows both 30-gauge Kanthal and nichrome wire at new/pre-heated and after 1, 50, and then 150 heats. Both wires exhibit a smooth surface prior to burning and even after the first burn, the metal begins to show signs of decay and fissures. By the 150^th heat cycle, both wires are pitted and have rough surfaces. In some instances, charring could be seen visually prior to SEM analysis. When all coils were heated while with a 50:50 PG:VG e-liquid, no statistical difference was observed between non-contact and contact configurations of either Kanthal or nichrome wires at either 1 or 50 heat cycles (P > 0.05). Additionally, no statistical difference was observed between 1 and 50 heat cycles (P > 0.05).

Figure 5:

Coil surface morphology - SEM images of Kanthal wire coils new (A & B), and after 1 (C), 50 (D), and 150 (E) dry heating cycles (2000x magnification) and nichrome wire coils new (F & G), and after 1 (H), 50 (I), and 150 (J) dry heating cycles.

Under a wet media of 50:50 PG:VG, the coils exhibited less visible deterioration under SEM when compared with the dry heat cycles. Statistical comparison of new, one, and 50 heat cycles were not statistically different in a wet media. However, consecutive loops within coils displayed varying degrees of deterioration (Figure 6). The difference in the two medias (wet vs. dry) was significant after 50 heat cycles (P < 0.05).

Figure 6:

Coil surface morphology - SEM images of contact and non-contact configurations of Kanthal wire coils after one (A) 50 (B) wet heating cycles and nichrome wire coils after one (C) and 50 (C) wet heating cycles.

Discussion

Coil Temperature Studies

Online forums for e-cigarette consumers who use rebuildable atomizers (RBA) have created numerous posts discussing whether Kanthal A-1 and nichrome creates the best experience. Most participants on those forums will claim that the choice between wire type is a personal one. However, it has been highlighted by consumers that Kanthal wire can withstand higher temperatures and thus, lasts longer than nichrome (Reddit… 2015). The other coil topic frequently discussed in these forums is the configuration of the coil. Most RBA users will say that contact configuration coils will burn hotter than a non-contact coil, and that the non-contact coil will create a more even burn across the wick and coil (Contact… 2014). Consumers also say that vaping with a contact coil will lead to a “dry burn” or heating the wick to the point where there is no e-liquid left to aerosolize. Further, assessing the coil temperature outputs while dry is critical given that a model called the rebuildable dripper atomizer (RDA) enables users to drip an e-liquid directly onto a coil or coil/wick system (Poklis et al. 2017; Harrell and Eissenberg 2018).

When assessing the coil temperatures heated in the absence of a liquid, Kanthal wire reached higher temperatures than nichrome and could withstand greater voltages. Heating coils in a dry medium and with increasing voltage and wire gauge showed that contact coils would reach a temperature where they would melt or “burn out” and cease to work above a certain voltage (Figure 1). For Kanthal wire, 34-gauge was the only wire in a contact configuration that could not withstand the highest voltage, 5 V. For nichrome, only 30-gauge wire, the thickest wire, could withstand the temperatures produced from all four voltages. However, in a wet medium, both coils were able to withstand the same amount of voltage applied in both 100% PG and 100% VG e-liquids.

In 100% VG, no significant difference between the coil type and configuration in regards to their observed temperature outputs was observed (P = 0.198). However, in 100% PG, the two wire types had a statistically significant difference in temperature outputs. Kanthal also showed a difference in temperature output between the two coil configurations. Propylene glycol is less viscous than vegetable glycerin. As it was previously mentioned, Kanthal reaches higher temperatures in a dry medium when compared to nichrome wire and a higher temperature in its contact configuration over non-contact. The less viscous properties of propylene glycol allow Kanthal coils to reach a higher temperature than nichrome during the eight second heating period used in this study. At these higher temperatures, a greater risk of “dry burning” the coil in 100% PG e-liquids exits. Some consumers who have recently transitioned to e-cigarettes from traditional cigarette usage prefer 100% PG based e-liquids, citing that the vapor produced emulates a traditional smoking experience (Legostar 2013).

The 50:50 PG:VG ratio data for Kanthal wire showed a similar trend in comparison to 100% VG vs 100% PG (Figure 2). No significant difference in the temperature output of the coils when compared to 100% VG was observed, but when compared to the 100% PG medium, Kanthal temperature outputs were statistically different. Similar 100% VG and 100% PG, the difference in media viscosity influences the temperature outputs of the coils and their ability to conduct heat.

Coil SEM Analysis

SEM data gathered in this study showed a significant metal loss for iron, chromium, and nickel in the coils after being heated while dry. The metal loss from the coils may be present in aerosol inhaled by the e-cigarette user. The metals that exhibited composition changes in this study were consistent with metals identified and quantified in previous studies (Williams et al. 2013; Goniewicz et al 2014; Hess et al. 2017; Olmedo et al. 2018). When reporting studies about metals in e-cigarette aerosols, it is important to include the coil specifications. The coil type and configuration will have an effect on which metals are present and the amount of metal present in the aerosol. Previous studies have highlighted the dangers of nickel and chromium being inhaled via e-cigarette aerosol. Nickel and chromium are both carcinogenic and have been linked to illnesses such as chronic bronchitis and lung cancer (Hess et al. 2017).

Dry heating a coil is a common technique used with rebuildable atomizers. It allows the user to shape the coil and ensure even heating across it when building. It is also used to “clean” the coil of e-liquid build-up when changing or refilling the tank. Online forums have reported that the build-up of e-liquids primarily happens on coils that are built in a contact configuration. This study determined that there was a significant metal loss from newly constructed coils in both contact and non-contact configurations after just one dry heat cycle. Further, a significant difference in the amount of metal loss between the contact and non-contact configurations was observed. Figure 3 shows that contact configurations had a greater metal loss overtime when compared to non-contact configurations. This metal loss and differences between wire type and configuration is supported by the temperature differences in the coil temperatures when dry (Figure 2). The dry burning method for coils has caused members of the scientific community to caution e-cigarette users from pursuing this method, citing that the possible metal loss from the intense heat of the dry burn could be inhaled by the users (Farasalinos 2015).

Changes to the coil from dry heating over time showed not only metal loss, but visual changes to the coils. SEM images showed increased surface pitting and the formation of a white solid, indicating oxidation. SEM images of the coils showed that pieces of the wire were missing, possibly due to disintegration. Oxygen levels increased with each heat cycle and an increase of aluminum in nichrome and Kanthal, respectively, was also observed. Kanthal and nichrome wire in a contact configuration both showed significant oxygen changes after the first heat cycle. Nichrome wire, however, showed that by the 50 and 150 heat cycles, oxygen levels did not change significantly, which was also seen in both wire types non-contact configurations. The apparent loss of oxygen seen in nichrome wire between 50 and 150 heat cycles may be due to the metal degrading over time.

The increase in oxygen content is possibly due to the oxidation of the coils as they are heated. Nichrome wire will form a chromium oxide on the outer side of the coil as it is heated and Kanthal will form an aluminum oxide coating (Kanthal… 2018). The formation of these oxides are intended to protect the coil from further oxidation, however, the metal oxides can be inhaled into deep lung tissue, where injury is likely to occur. A particle size analysis of e-cigarette aerosol shows that in the three e-liquid compositions presented in this study (100% VG, PG, and 50:50 PG:VG), aerosol particles were less than 5 μm in size, demonstrating that the aerosol deposits in the deep lung tissue (Mulder et al. 2019). Further, other studies that assessed metal aerosol content from e-cigarettes using Kanthal and nichrome wire also observed aluminum and chromium, suggesting that oxidized metals can still be inhaled by the e-cigarette user (Williams et al. 2019; Olmedo et al. 2018)..

The SEM-EDX data from wet heat cycles demonstrated no significant difference in metal deterioration between coil types and configurations. This supports the temperature data which also did not change significantly between the coil types and configuration. In the end, user preferences are likely a reflection of subconscious behavioral effects.

Conclusions

Despite enthusiastic conversations about e-cigarette coil types and configurations, in the presence of e-liquids, there is no significant difference between the two types of wire and configurations on temperature output or as observed by SEM-EDX. Dry burning has been shown to cause significant metal loss from the coil as early as one heat cycle and can potentially be inhaled by the e-cigarette consumer in the process of building or cleaning their coil. The metals identified on the coils are known to cause negative health effects when inhaled over time. It is also important that in future studies that report the presence and concentrations of metal in e-cigarette aerosol include all specifications about the atomizer, including the heating element. This is because the metal compositions will affect the presence and abundance of certain metals in the aerosol.