Nicotine delivery from the refill liquid to the aerosol via high-power e-cigarette device

This study deals with the aerosol nicotine delivery by recent high-power ENDS. To the best of our knowledge, this paper proposed for the first time an exhaustive characterization of the airborne nicotine flux including the nicotine content for aerodynamic particle size-fractions from 7 nm to 10 µm. Results perfectly demonstrated that the PG/VG ratio has no significant effect on the frequency distribution of the mass of refill liquid (the airborne carrier) as well as the mass of nicotine (the active product inside this carrier). Whatever the PG/VG ratio, a perfect matching between the frequency mass distribution of the airborne refill liquid and the nicotine was observed. This finding leads to indicate that:

1. (i)

   There is no empty airborne carrier generated by recent high-power ENDS (i.e. droplet of refill liquid without nicotine).

2. (ii)

   The nicotine concentration inside droplets of refill liquid, whatever the aerodynamic size-fractions in the submicron range, is constant for a given PG/VG ratio of refill liquid.

3. (iii)

   The particle size distribution of the airborne refill liquid perfectly fits the particle size distribution of the aerosol nicotine.

Besides, this last point, i.e. the fact that nicotine is equally distributed in droplets regardless of their size, can appear quite obvious. Indeed, the studied system is a homogenous solution of miscible liquids. Consequently, each droplet contains the same proportion of each components (PG, VG and nicotine). Aerosol droplets are formed after condensation of mixed vapor produced by heating of the refill liquid in the ENDS device. This process may change the composition of the condensate comparing to the un-puffed solution (by the principle known from distillation processes, e.g. evaporation and subsequent condensation of alcohol solutions). Nevertheless, the whole condensed liquid has the same composition in term of proportion of PG, VG and nicotine. However, volatile aerosol droplets, and especially in the submicron-sized range such as particles produced by recent high-power ENDS, can lose solvents by evaporation prior to entry into the user’s mouth, or an inertial sizing device such as cascade impactors. As a result, evaporation can cause an increase of nicotine concentration in the droplet and reduction in size. The impact of the evaporation phenomena can be particularly great, for smaller airborne particles compared to larger ones (droplet curvatures have also a strong effect on the evaporation process), when a large volume of dilution air is mixed with the aerosol stream during the sampling step. Nicotine was equally distributed in droplets regardless of their size and the aerosol concentration of nicotine appeared lower than the un-puffed solution. Although the use of a high flowrate for aerosol generation as well as a dilution ratio of 1.5, this finding doubtless proved that no important evaporation bias occurs during the sizing measurement strategy including high flowrate (500 mL.s⁻¹, DLPI set-up, sampling including an initial dilution ration of 1.5).

Few studies were devoted to aerosol features generated by ENDS in term of MMAD measurement. The findings obtained in this study (in the 0.75–0.8 µm range of MMAD) appear in good accordance with cascade impactor data previously published^8,9,10,11. Indeed, the results found in the literature, using various types and brands of ENDS, are often in the 0.5–1 µm range of MMAD when inertial sizing techniques were used: 487 nm to 631 nm for Alderman et al.⁸, from 600 to 650 nm for Bertholon et al.⁹, from 600 to 800 nm for Kane et al.¹⁰ and 1.03 µm for Lerner et al.¹¹). The use of an important flowrate (500 mL.s⁻¹) compared to flowrate (around 20 mL.s⁻¹) used by vapers in real-life practice could have been a drawback for this work. It is possible that high flow rate can affect the particle size distribution. For example, the wick may not have enough time to be saturated with liquid and the residence time should be reduced. Besides, we must keep in mind that we used for this work a newly engineered atomizer including an airflow control ring. As previously showed, no important evaporation bias occurs during the sizing measurement strategy and thus the important flowrate used seems do not strongly affect the MMAD determined.

To more precisely evaluate the bias induced by the high flow rate used, experiments were also performed using the ELPI mode of the DLPI cascade impactors with a puffing behavior similar to real-life practices. Within ELPI, particles are electrically charged via a corona charger and subsequently impacted in one of the 12 size fractions stages of the DLPI impactor. The induced current of impacted particles is measured and related to the aerodynamic cut-off diameter of the impactor stage, providing real-time size-dependent distributions. Results can be expressed by means of number concentrations to calculate the Count Median Aerodynamic Diameter (CMAD). Aerosol sampling was carried out considering 4-s puff and a volume of 60 mL (flow rate of 15 mL.s⁻¹). An in-house interface was designed. Puffs were performed using a 60 mL syringe connected to the ENDS (exactly as described in Fig. 1, replacing the 3L-syringe with a 60mL-syringe). In these experimental conditions, results showed a CMAD of 0.72 ± 0.01 µm (GSD of 2.21) for the 80PG/20VG formulation. We can convert CMAD into MMAD since for a log-normal distribution the known relationship holds that MMAD = CMAD exp[3(log(GSD))²]. Consequently, using this equation we found a MMAD equal to 1.028 µm (conditions: 4-s puff, 60 mL of aerosol, flowrate of 15 mL.s⁻¹, no dilution during the sampling using the 60mL-syringe, CMAD calculation using ELPI and then calculation of MMAD) versus 0.79 µm (Table 1, conditions: 4-s puff, 2 L of aerosol, flowrate of 500 mL.s⁻¹, dilution ratio of 1.5 during the sampling using the 3L-syringe, DLPI size-fractionation and mass measurement using a precision balance). Many parameters vary between these two experimental conditions (and not only the flow rate). As a result, we assume that the possible bias using a high flow rate, if it exists, induces an error of at most 20% on the measurement of the MMAD. All things considered, this study demonstrated that our data on aerosol sizing are quite relevant compared to the literature on MMAD determination.

Another important finding of this study is to highlight a lower aerosol concentration of nicotine compared to un-puffed nicotine concentration of the refill liquid initially introduced inside the tank-type atomizer. A decrease ranging from 24 ± 7% to 38 ± 15% was measured depending on the PG/VG ratio of the refill liquid. No experimental biases could explain these results (i.e. all possible biases, mainly evaporation process of volatile droplets, would have tended to increase aerosol nicotine concentration, not to decrease it). Furthermore, our findings appear highly coherent with results recently published¹². The consistency of our results with the literature data, again confirm that our findings are relevant although the high flow rate used (which seems to play a minor role when the airflow control ring is at the maximum position). Pagano et al. studied the portion of nicotine delivered via aerosolization using ENDS technology dating from early 2014. They demonstrated that, under their experimental conditions and for their given ENDS technology, the portion of aerosol nicotine delivered to filter pads was often less than half that which was available, indicating that most of the nicotine may be left in the ENDS upon depletion. Several assumptions could be hypothesized to explain this phenomenon:

1. (i)

   A partial thermal degradation of the nicotine molecules.

2. (ii)

   A part of the nicotine can be present in the gaseous state (i.e. in a vapor rather than contained in airborne liquid droplets), so it cannot be collected in the cascade impactor.

3. (iii)

   The ENDS process can change the composition of the condensate comparing to the un-puffed solution (the so-called “distillation process”).

Besides, the initial concentration of nicotine in the refill liquid (i.e. the un-puffed liquid) was slightly different from the value indicated by the manufacturer (21.6 ± 0.9 mg/mL and 21.1 ± 0.7 mg/mL vs. 18 mg/mL for the labeled nicotine concentration). Our analyses of un-puffed liquid by LC-MS highlighted that the nicotine content of refill liquid can be considerably different from manufacture’s labeling. There have been relatively few studies on the accuracy of the labeled nicotine concentrations on refill liquid for ENDS^13,14,15. Our findings are perfectly in accordance with a previous work highlighting measured nicotine concentrations higher than the labeled values, with many being over 20% higher¹⁶.

When delivered through the pulmonary route (as with traditional smoking cigarette or with ENDS), nicotine from the aerosol can theoretically be rapidly absorbed into the blood circulation. Crossing the alveolar–capillary barrier, aerosol nicotine reaches the brain within few seconds¹⁷. By contrast, buccal and dermal nicotine absorption (as delivered with usual NRT products) is slower and subject to first-pass metabolism. Therefore, contrary to ENDS, usual NRT products may pose less abuse liability about the risk of initiate or maintain nicotine dependence. However, in perfect accordance with the global objective of smoking reduction or cessation, the development of more effective ENDS technologies appear essential. The main challenge consists in offering an increased and well-controlled nicotine delivery to the vapers. Smokers using ENDS for smoking cessation are already nicotine-dependent persons. But it is possible that technological improvement of ENDS in nicotine delivery efficacy could, as a side effect, make vapers more addicted to nicotine. However, the main issue is that smokers have sometimes difficulties to adequately satisfy their nicotine cravings using ENDS technologies currently on the market. Indeed, ENDS differ remarkably from tobacco cigarettes in terms of systemic nicotine delivery. Although nicotine content in the ENDS aerosol is in the same order of magnitude compared to the mainstream of conventional cigarette, clinical data indicate that the nicotine absorption potential is significantly lower for vapers compared to smokers¹⁸. An improvement in the efficacy of ENDS to deliver aerosol nicotine can be highly beneficial for smokers, with the ability to adjust use patterns for smoking cessation purpose. We expect that new-generation high-power ENDS will deliver well-controlled nicotine at a faster rate and will be more efficient for smokers to satisfy their nicotine need. Consequently, these ENDS technologies could contribute to decrease the number of active smokers and to limit the risk to make vapers more addicted to nicotine. Technological research focusing on airborne nicotine flux into the ENDS aerosol represents an essential step to develop in a next future more effective ENDS technologies, and thus to rise smoking cessation and reduction.

In this frame, two main options can be followed:

1. (i)

   To increase the aerosol nicotine concentration. However, in Europe, the newly revised Tobacco Products Directive proposes to regulate ENDS as a tobacco related product¹⁹. These new rules implemented across the European Union member states applied since May 2016 specified how tobacco related products can be sold, presented and manufactured. Especially, this regulatory decision fixed an upper limit at 20 mg/mL of nicotine concentration in refill liquid. Thus, the increase of nicotine concentration in refill liquid in order to rise the aerosol nicotine concentration becomes impossible. Other ways such as new formulations of refill liquid (e.g. playing on the PG/VG ratio, development of new solvents VG-free, etc.) should be explored.

2. (ii)

   On the other hand, to improve the nicotine absorption level. Longer and deeper puffs increase the nicotine absorption level. Vapers and clinicians have to be aware of this impact of puffing behaviors. ENDS technical features can also be adjusted (mainly power level which can be variable and easy to change using recent high-power ENDS) to modify both the aerosol output and the airborne particle size distribution. These two parameters are critical for the nicotine delivery to lungs.