ArticlePDF Available

Abstract

In electronic nicotine delivery systems (ENDS), coil resistance is an important factor in the generation of heat energy used to change e-liquid into vapor. An accurate and unbiased method for testing coil resistance is vital for understanding its effect on emissions and reporting results that are comparable across different types and brands of ENDS and measured in different laboratories. This study proposes a robust, accurate and unbiased method for measuring coil resistance. An apparatus is used which mimics the geometric configuration and assembly of ENDS reservoirs, coils and power control units. The method is demonstrated on two commonly used ENDS devices—the ALTO by Vuse and JUUL. Analysis shows that the proposed method is stable and reliable. The two-wire configuration introduced a positive measurement bias of 0.086 (Ω), which is a significant error for sub-ohm coil designs. The four-wire configuration is far less prone to bias error and is recommended for universal adoption. We observed a significant difference in the coil resistance of 0.593 (Ω) (p < 0.001) between the two products tested. The mean resistance and standard deviation of the reservoir/coil assemblies was shown to be 1.031 (0.067) (Ω) for ALTO and 1.624 (0.033) (Ω) for JUUL. The variation in coil resistance between products and within products can have significant impacts on aerosol emissions.
International Journal of
Environmental Research
and Public Health
Article
Method for Quantifying Variation in the Resistance of
Electronic Cigarette Coils
Qutaiba M. Saleh 1, Edward C. Hensel 2, * and Risa J. Robinson 2
1Department of Computer Engineering, Rochester Institute of Technology, Rochester, NY 14623, USA;
qms7252@rit.edu
2Department of Mechanical Engineering, Rochester Institute of Technology, Rochester, NY 14623, USA;
rjreme@rit.edu
*Correspondence: echeme@rit.edu
Received: 10 September 2020; Accepted: 22 October 2020; Published: 24 October 2020


Abstract:
In electronic nicotine delivery systems (ENDS), coil resistance is an important factor
in the generation of heat energy used to change e-liquid into vapor. An accurate and unbiased
method for testing coil resistance is vital for understanding its eect on emissions and reporting
results that are comparable across dierent types and brands of ENDS and measured in dierent
laboratories. This study proposes a robust, accurate and unbiased method for measuring coil
resistance. An apparatus is used which mimics the geometric configuration and assembly of ENDS
reservoirs, coils and power control units. The method is demonstrated on two commonly used ENDS
devices—the ALTO by Vuse and JUUL. Analysis shows that the proposed method is stable and
reliable. The two-wire configuration introduced a positive measurement bias of 0.086 (
), which is a
significant error for sub-ohm coil designs. The four-wire configuration is far less prone to bias error
and is recommended for universal adoption. We observed a significant dierence in the coil resistance
of 0.593 (
) (p<0.001) between the two products tested. The mean resistance and standard deviation
of the reservoir/coil assemblies was shown to be 1.031 (0.067) (
) for ALTO and 1.624 (0.033) (
) for
JUUL. The variation in coil resistance between products and within products can have significant
impacts on aerosol emissions.
Keywords: coil resistance; e-cigarette; electronic nicotine delivery system; pod style; atomizer
1. Introduction
1.1. Theoretical Foundation
Typical pod-style electronic nicotine delivery systems (ENDS) consist of three subsystems:
the reservoir, power control unit (PCU) and lithium battery. The reservoir includes a mouthpiece,
e-liquid storage compartment, a heating element herein called the coil, sometimes a wick, and an
aerosol generation chamber, sometimes referred to as the atomizer. Reservoirs designed for re-use,
permitting users to refill the reservoir with e-liquids, are sometimes called “open systems.” Reservoirs
designed to be disposable, not intended to permit e-liquid refills by the user, are sometimes called
“closed systems.” Most pod-style ENDS reservoirs are designed with the heating element, or “coil”,
fully integrated with a wick to deliver e-liquid from the reservoir to the heating element such that
thermally generated aerosol mixes with inhaled air for delivery through the mouthpiece to the user.
Other ENDS reservoir designs may permit the user to replace the coil, wick, or adjust the flow path
through the reservoir. The PCU contains electronics which manage the user interface, controls energy
delivered to the pod, and withdraw or replace (recharges) energy to the battery. The lithium battery is
the energy storage unit and provides energy to the PCU and the pod. The three subsystems interact
Int. J. Environ. Res. Public Health 2020,17, 7779; doi:10.3390/ijerph17217779 www.mdpi.com/journal/ijerph
Int. J. Environ. Res. Public Health 2020,17, 7779 2 of 16
with one another in an integrated manner. Variation in one component, such as the coil in the reservoir,
may have dierent impacts on aerosol generation depending upon its interaction with the PCU and
the battery.
The aerosol generation performance of an ENDS is jointly dependent upon the physical
characteristics of the ENDS pod, PCU and battery [
1
6
], characteristics of e-liquid in the
reservoir [
2
,
3
,
5
,
7
], and user behaviors of puflow rate and puduration [
1
,
8
12
]. Understanding the
theory of ENDS operation elucidates potential eects of variation in resistance arising from interactions
with the PCU and battery. All ENDS are fundamentally heat and mass transfer devices, which are
commonly studied in engineering disciplines [
13
] for medical, industrial, residential and commercial
products. The purpose of the heating coil is to convert electrical power discharge from the battery
into thermal power dissipated inside the coil. The resulting thermal power (W) is distributed via
heat conduction to the surface area of the coil as a heat flux (W/m
2
), which is then transferred to the
surrounding air/e-liquid via heat convection. As the e-liquid solvent temperature reaches its saturation
temperature (the eective boiling point of the e-liquid mixture), mass transfers from the e-liquid
reservoir into the air stream to form an aerosol. The combination of surface heat flux, coil surface
temperature, heat and mass transfer coecients and e-liquid composition jointly aect the rate of
aerosol generation. The aerosol generated at the interface between the coil and e-liquid experiences
further changes as it progresses through the flow channel of the ENDS toward the user. While the
eectiveness of aerosol generation is impacted by many factors, it is dominated by the electrical
energy from the battery dissipated as thermal energy in the coil. The amount of heat energy, E,
in Joules (J), dissipated in the coil is defined as the product of the instantaneous direct current power,
P, supplied to the coil, in watts (1 W
1 J/Sec), and time duration over which the power is supplied
(Sec). The power is a function of the electrical current, I
coil
(A), flowing through the coil, the voltage,
V
coil
(V), applied across the terminals of the coil, and resistance, R
coil
(
) of the coil itself as shown in
Equation (1).
P=Vcoil Icoil =V2coil
Rcoil =I2coil Rcoil (1)
The coil resistance is an inherent physical characteristic of the coil which depends primarily
upon the composition and purity of the coil and its geometry. In this study, we include the internal
electrical connections between the coil and the pod housing to quantify the eective resistance of the
coil assembly. The voltage, V
coil
, and current, I
coil
, passing through the coil are related to one another
using the classical definition of Ohm’s law, V
coil
=I
coil ×
R
coil
, from physics [
14
]. The PCU controls
the duration over which power is supplied to the coil. The PCUs employed in early ENDS designs
simply shorted the voltage available from the battery across the coil for an interval of time. As the
battery discharged over time, its available voltage decreased and hence the power delivered to the coil
decreased. All modern ENDS PCU designs control the time duration of power delivery, while some
PCU designs actively control the current, I
coil
, flowing through the coil and other designs actively
control the voltage, V
coil
, applied across the terminals of the coil. Fully understanding the eects of
variation in coil resistance, R
coil
, on aerosol emissions cannot be accomplished without understanding
the logic implemented in the ENDS PCU.
Equation (1) shows that the desired power of an ENDS can be achieved by a specific ratio of coil
resistance and applied voltage. In the lowest-cost ENDS designs, the applied voltage is limited by
physical constraints of the most common lithium batteries, which peaks at approximately 3.7 (V) and
decreases as the battery discharges. Higher-cost ENDS designs may actively control the output voltage
using a “boost” converter [
15
], at the penalty cost of reduced operating time between recharging.
ENDS manufacturers, over time, have sought to increase the power dissipation in the coil in order to
increase the rate of e-liquid aerosolization. Given the physical limitations of low-cost rechargeable
lithium batteries, ENDS designers choose to reduce the coil resistance as the most appealing parameter
to increase power. This is illustrated by an example. Consider an ENDS designer who specified a power
level of 12 (W) and a lithium battery operating at nominally 3.7 (V). Equation (1) dictates that a coil
resistance of 1.14 (
) should be used. If power of more than 12 (W) is desired, even smaller resistance is
Int. J. Environ. Res. Public Health 2020,17, 7779 3 of 16
required. Otherwise, a stack of two or more batteries connected in series or a DC-to-DC boost converter
can be used to step up the applied voltage [
15
]. Using two batteries is not a desirable solution, because it
increases the cost, weight and volume of the ENDS. Boost converters are increasingly common in
modern ENDS but appear to be primarily used as a means to extend operating lifetime and maintain
steady power while the battery discharges.
The desire for high-power ENDS pushes designers to increasingly use coils with low resistance
values, which led to the introduction of the sub-ohm devices which use coils that have a resistance
of <1 (
) [
16
,
17
]. Reducing the coil resistance to 0.068 (
) permits instantaneous power as high as
200 W. The sub-ohm devices are reported to satisfy several features desired by users such as intense
flavor, warm vape and big clouds, which is associated with high airflow that is suitable for the
direct-to-lung inhale style [
18
,
19
]. Sub-ohm coils are mostly available in the box-mod ENDS style;
however, some pod-style ENDS started to use coils with resistances of less than 1 (
) such as SMOK2
pod (0.8 (
) and NORD2 (0.3, 0.4, 0.6 (
)) [
20
], TARGET PM80 (0.2, 0.3, 0.6, 0.8 (
) [
21
], and Z-BIIP
(0.48 ()) [22].
Several studies investigated the eects of power values on the performance of ENDS devices
while others focused on coil resistance values. As demonstrated with Equation (1), power levels can
be controlled in real time by manipulating the applied voltage and current, or by installing a coil
with a dierent resistance. The power level of an ENDS may be increased by increasing the applied
voltage (which has the eect of increasing the current) or by decreasing coil resistance values while
keeping the applied voltage constant. The same approach can be followed for decreasing power levels.
For this reason, most of the results achieved from the studies focusing on power and voltage values
can be generalized to coil resistance values with appropriate adjustments and vice versa. However,
it is essential that studies investigating the eect of power on emissions document both the power
dissipation and coil resistance in order to make results generalizable to other products.
1.2. Context of Prior Work
Previous studies have shown that coil resistance aects both the amount of vapor generated and
constituents. Cirillo et al. 2019 [
23
] showed that reducing coil resistance leads to higher concentrations
of some carbonyls and reactive oxygen species (ROS). Their research also reported that vapor generated
by coils with lower resistances has a higher negative impact on cell viability. The same research
group [
24
] also showed in a separate work in 2019 that the production of selected aldehydes increased
as coil resistance decreased from 1.5 (
) to 0.25 (
). The eects of the aerosols generated by the
two coils on Sprague–Dawley rats was studied. The rat group exposed to the 0.25 (
) vape showed
disorganization of alveolar and bronchial epithelium. The same group also showed higher perturbation
of the antioxidant and phase II enzymes compared to the 1.5 (
) groups. Gillman et al. 2016 [
25
] studied
the eects of changing power on the total yield mass and the formation of aldehyde. Their results
showed that power has significant impact on the concentration of aldehyde in the vapor. Although their
main focus was power, they used several coils with various resistance values to control the power.
This indicates that in this study, the physically important factor is resistance values, as already
demonstrated in Equation (1). Chausse et al., in 2015 [
26
], suggested coil resistance could be the key
to lung toxicity. Their analysis showed that the combination of certain voltage and coil resistance
values may lead to a high impact on human health. Hiler et al. 2019 [
27
] investigated the eects
of changing heating coil resistance on nicotine delivery, pung topography, subjective eects and
liquid consumption. They used o-the-shelf coils with two resistance values of 0.5 (
) and 1.5 (
)
which are supposed to consume a power of 40.5 (W) and 13.5 (W), respectively. Hiler reported that
lower-resistance coils were found to deliver higher nicotine and have higher liquid consumption.
Several prior studies did not document the method used to measure coil resistance or assess variation
in this key parameter. Cirillo et al. 2019 [
23
,
24
], Sleiman et al. 2016 [
28
], Ogunwale et al. 2017 [
29
]
and Soulet et al. 2018 [
30
] reported coil resistance as a study parameter without documenting the
method used to measure resistance. Researchers may be tempted to select o-the-shelf coils with
Int. J. Environ. Res. Public Health 2020,17, 7779 4 of 16
resistance values that suit their study. This approach assumes that the coil resistances reported by
the manufacturer are accurate, neglects manufacturing variation between coils, and may limit study
designs to ENDS brands that oer coils with dierent resistance values.
Conversely, other studies reported the measurement methods employed. Gillman et al.
2016 [
25
] used a measurement instrument that is specialized in milliohm resistance measurement
(Extech milliohm meter, 380560) which claims high precision and low error rates. Hiler et al. [
27
] used
an o-the-shelf ohm meter “Coil Master 521 TAB v2”, advertised to test homemade coil resistance
and ensure proper operation before yse. The Coil Master [
31
] is reported to have a 510 threaded
connector, compatible with many box-mod ENDS, and is equipped with a fire button which applies
voltage to the coil during test. The manufacturer reports that readings have an error of approximately
+/0.05 ()
, which corresponds to a reading error of +/
10% when the measured coil has a resistance
of 0.5 (
). The high error margin of this meter suggests that it is generally not appropriate for use in
scientific studies.
In summary, many studies have been performed to assess the impact of power and coil resistance
on emissions and toxicity, but no studies have been performed to evaluate the accuracy of the methods
used to measure coil resistance. Many modern commercially available box-mod ENDs are equipped
with a resistance meter to report coil resistance to the user, but no studies could be found that validate
the reliability of these values.
With the current trend to reduce coil resistance values to sub-ohm levels [
16
,
17
], manufacturing
variations might have greater eects on the actual resistance of the coils, which in turn may give rise to
variation in the composition of emissions from the ENDS. To date, we have found no reports that quantify
such variation nor how it might aect the expected performance of the ENDS. Previous inconsistencies
in reporting coil resistance may be caused by the absence of a robust standard method.
1.3. Study Objectives
This study focuses on demonstrating accurate and robust methods for quantifying: (1) the
eective resistance of the heating coil in a fully assembled pod, (2) variation in eective coil resistance
across pods from a single manufacturer, (3) the repeatability of the test method and apparatus,
and (4) dierences in eective coil resistance between manufacturing designs. The proposed methods
are demonstrated on two popular [
32
34
] commercially available pod-style ENDS: ALTO by Vuse [
35
]
and JUUL [
36
]. The test fixtures described herein are supported by detailed published protocols [
37
],
enabling researchers to use commercially available test equipment and instructions to fabricate custom
test fixtures for measuring the eective coil resistance of integrated pods from various manufacturers.
2. Materials and Methods
2.1. Constant Current Resistance Measurement Method
The constant current method is commonly used for resistance measurement in conjunction with a
digital multimeter (DMM). Modern DMMs for laboratory use incorporate an integrated voltmeter and
constant current source. The DMM constant current source, I
Source
, is used to supply current to the
device under test (DUT) while measuring the voltage, V
Voltmeter
, across the DUT. The resistance of the
DUT, R
Measured
, is determined as the ratio of measured voltage over the applied current, consistent with
Ohm’s law:
RMeasured =VVoltmeter
ISource
(2)
Two configurations are commonly used with the constant current method. The most common
two-wire configuration is appropriate for general purposes with large DUT resistances on the order of
kilo-ohms and mega-ohms. The four-wire configuration is more appropriate for accurate measurement
of sub-ohm coil resistances. The eect of measurement configuration is assessed herein by using both
Int. J. Environ. Res. Public Health 2020,17, 7779 5 of 16
configurations on a sample of pods from two manufacturers and conducting a repeated-measures
dierence between configurations.
2.1.1. Two-Wire Configuration
The schematic shown in Figure 1is known as the two-wire configuration [
38
]. Two terminals
(denoted as the +and
terminals) are supplied with a constant current, I
Source
, from the DMM.
two-wire leads are connected from the terminals of the DMM to the opposing ends of the DUT coil
being tested. Kircho’s voltage law states that the voltage drop across three components in series is
the summation of the voltage drop across each component in the series. Thus, the voltage, V
Voltmeter
,
measured by the DMM is the summation of the voltage drops across the DUT and the two-wire leads.
Kircho’s current law asserts that the steady current flowing through three resistive components in
series are equal. Consequently, the measured current through the DUT will be identical to the current
through the wire leads. Equation (3) shows the calculations in detail, which can be simplified to the
second form if we assume that the two-wire leads are of the same composition, diameter and length.
RMeasured =VVoltmeter
ISource =VLead+VDut +VLead
ISource =RDUT +2×RLead (3)
Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 5 of 16
2.1.1. Two-Wire Configuration
The schematic shown in Figure 1 is known as the two-wire configuration [38]. Two terminals
(denoted as the + and terminals) are supplied with a constant current, ISource, from the DMM. two-
wire leads are connected from the terminals of the DMM to the opposing ends of the DUT coil being
tested. Kirchoff’s voltage law states that the voltage drop across three components in series is the
summation of the voltage drop across each component in the series. Thus, the voltage, VVoltmeter,
measured by the DMM is the summation of the voltage drops across the DUT and the two-wire leads.
Kirchoff’s current law asserts that the steady current flowing through three resistive components in
series are equal. Consequently, the measured current through the DUT will be identical to the current
through the wire leads. Equation (3) shows the calculations in detail, which can be simplified to the
second form if we assume that the two-wire leads are of the same composition, diameter and length.
𝑅 =𝑉

𝐼
=𝑉
 +𝑉
 +𝑉

𝐼
=𝑅
 +2×𝑅
 (3)
The resistance of wire leads can vary widely between research laboratories and test benches,
easily between 0.010 (Ω) and 1.000 (Ω) depending mainly on their materials and lengths. The
resistance of the wire leads, RLead, introduces a significant bias error between the DMM observed,
RMeasured, and the actual coil, RDUT, when the DUT has low resistance. For example, when measuring
the resistance of an ENDS coil with a true resistance of 1 (Ω) in this configuration with two-wire leads
that have a resistance of 0.050 (Ω) each, the measured value would be approximately 1.1 (Ω), a 10%
bias error in measurement. If we use the same apparatus and wire leads to measure sub-ohm coils,
the percent of bias error would be correspondingly higher. The wire leads are the main source of
error in this configuration. The bias error can be reduced by using very short wire leads with high
electrical conductivity (e.g., gold instead of copper). However, short leads are often difficult to
manipulate in the lab, and this approach does not completely remove the bias introduced by the wire
leads. The four-wire configuration offers a practical and robust approach to removing the bias error
associated with the two-write configuration.
Figure 1. Two-wire resistance measurement schematic. Dotted box labeled digital multimeter (DMM)
is a simplified version of internal schematic of the digital multimeter.
2.1.2. Four-Wire Configuration
Figure 1.
Two-wire resistance measurement schematic. Dotted box labeled digital multimeter (DMM)
is a simplified version of internal schematic of the digital multimeter.
The resistance of wire leads can vary widely between research laboratories and test benches,
easily between 0.010 (
) and 1.000 (
) depending mainly on their materials and lengths. The resistance
of the wire leads, R
Lead
, introduces a significant bias error between the DMM observed, R
Measured
,
and the actual coil, R
DUT
, when the DUT has low resistance. For example, when measuring the
resistance of an ENDS coil with a true resistance of 1 (
) in this configuration with two-wire leads
that have a resistance of 0.050 (
) each, the measured value would be approximately 1.1 (
), a 10%
bias error in measurement. If we use the same apparatus and wire leads to measure sub-ohm coils,
the percent of bias error would be correspondingly higher. The wire leads are the main source of error
in this configuration. The bias error can be reduced by using very short wire leads with high electrical
conductivity (e.g., gold instead of copper). However, short leads are often dicult to manipulate in the
lab, and this approach does not completely remove the bias introduced by the wire leads. The four-wire
Int. J. Environ. Res. Public Health 2020,17, 7779 6 of 16
configuration oers a practical and robust approach to removing the bias error associated with the
two-write configuration.
2.1.2. Four-Wire Configuration
The four-wire configuration employs two current wire leads (force +and force
) to supply
current through the DUT and two separate sensor wire leads (sense +and sense
) to measure the
voltage across the DUT. The four-wire configuration eliminates the bias eect of lead resistances
described in the two-wire resistance measurement configuration [
38
]. Figure 2shows the schematic
of this configuration. The voltmeter has very high resistance (on the order of megaohms) and thus
very low current (on order of picoohms) flows through the sense +/
wire leads. Thus, the voltage
drop across the sense +/
wire leads is negligible and the voltage measured by the voltmeter is the
same as the voltage cross the DUT. The resistance of the wire lead is totally insignificant, and the
measured resistance is unbiased compared to the two-wire configuration. Random errors such as
those associated with analog-to-digital conversion in the DMM remain in both configurations but are
negligible in comparison to the quantities typically required for characterization of ENDS coils.
Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 6 of 16
The four-wire configuration employs two current wire leads (force + and force ) to supply
current through the DUT and two separate sensor wire leads (sense + and sense ) to measure the
voltage across the DUT. The four-wire configuration eliminates the bias effect of lead resistances
described in the two-wire resistance measurement configuration [38]. Figure 2 shows the schematic
of this configuration. The voltmeter has very high resistance (on the order of megaohms) and thus
very low current (on order of picoohms) flows through the sense +/ wire leads. Thus, the voltage
drop across the sense +/ wire leads is negligible and the voltage measured by the voltmeter is the
same as the voltage cross the DUT. The resistance of the wire lead is totally insignificant, and the
measured resistance is unbiased compared to the two-wire configuration. Random errors such as
those associated with analog-to-digital conversion in the DMM remain in both configurations but are
negligible in comparison to the quantities typically required for characterization of ENDS coils.
Figure 2. Four-wire resistance measurement schematic. Dotted box labeled DMM is a simplified
version of internal schematic of the digital multimeter.
2.2. ENDS Product-Specific Test Fixture
An apparatus was assembled to conduct constant current method resistance measurements in
both the two-wire and four-wire configuration. The DMM used for all resistance measurements in
this study was a Model 34465A from KEYSIGHT™ [39] and supports both the two-wire and four-
wire resistance configurations with several user-selectable ranges (100 (Ω) to 1,000 (MΩ)). The DMM
was set to the “auto scale” option for all observations conducted in this study, resulting in every
observation being measured on the 100 (Ω) range. The DMM was connected to a desktop computer
running the Microsoft Windows™ 10 operating system (Microsoft, Redmond, WA, USA) with a USB-
2 cable. A custom MATLAB™ script (MathWorks, Inc., Naatick, MA, USA) was used to trigger
measurements and collect readings from the DMM and save the data to comma separated value
(CSV) test files for later analysis.
While commercial off-the-shelf (COTS) four-wire leads, commonly called “Kelvin leads”, may
be used to connect the DUT to the DMM, we elected to fabricate custom four-wire leads which were
permanently soldered to each DUT fixture. We determined that COTS Kelvin leads might not be the
best option for measurement of ENDS coil resistance. In some types of ENDS, the coils are separable
from the reservoir and wick. However, during and after usage, the coil is in contact or submerged
with the e-liquid. In other types of ENDS, such as the pod-style device studied here, the coil is
permanently attached within the pod and is submerged in the e-liquid reservoir. In removable pod
Figure 2.
Four-wire resistance measurement schematic. Dotted box labeled DMM is a simplified
version of internal schematic of the digital multimeter.
2.2. ENDS Product-Specific Test Fixture
An apparatus was assembled to conduct constant current method resistance measurements in
both the two-wire and four-wire configuration. The DMM used for all resistance measurements in
this study was a Model 34465A from KEYSIGHT
[
39
] and supports both the two-wire and four-wire
resistance configurations with several user-selectable ranges (100 (
) to 1000 (M
)). The DMM was set
to the “auto scale” option for all observations conducted in this study, resulting in every observation
being measured on the 100 (
) range. The DMM was connected to a desktop computer running
the Microsoft Windows
10 operating system (Microsoft, Redmond, WA, USA) with a USB-2 cable.
A custom MATLAB
script (MathWorks, Inc., Naatick, MA, USA) was used to trigger measurements
and collect readings from the DMM and save the data to comma separated value (CSV) test files for
later analysis.
While commercial o-the-shelf (COTS) four-wire leads, commonly called “Kelvin leads”, may be
used to connect the DUT to the DMM, we elected to fabricate custom four-wire leads which were
Int. J. Environ. Res. Public Health 2020,17, 7779 7 of 16
permanently soldered to each DUT fixture. We determined that COTS Kelvin leads might not be the
best option for measurement of ENDS coil resistance. In some types of ENDS, the coils are separable
from the reservoir and wick. However, during and after usage, the coil is in contact or submerged
with the e-liquid. In other types of ENDS, such as the pod-style device studied here, the coil is
permanently attached within the pod and is submerged in the e-liquid reservoir. In removable pod
ENDS devices, the pods most often make electrical contact with the ENDS PCU using two dierent
style connectors. Several pod-style ENDS employ two spring-loaded connectors, also known as pogo
pin connectors, for the +and
terminals on the PCU side while mating with two corresponding
flat connectors on the pod side. The connectors come in various sizes and shapes across ENDS
designs and, most importantly, they have dierent distances between the +/
terminals. Additionally,
the mechanical means of retaining the pod in the ENDS is dierent from one manufacturer’s design to
another. Some manufacturers use a friction fit, some use a magnetic clasp, and others may use detents.
This variation between ENDS designs means that the pod-style coil terminals are not directly exposed
and cannot directly connect to COTS Kelvin lead clips. Further, we desired a test fixture which would
allow us to assess the impact of the pod-retaining mechanisms employed in various ENDS designs
on the repeatability of eective coil resistance measurement. Electrical contact resistance caused by
the connectors is a potential confounder in natural environment operations of ENDS products, and is
worth investigation.
Accordingly, this study incorporates a special holding fixture which is unique to each ENDS
product and retains the pod with its integrated coil in position while measurements are made with
the DMM. Each holding fixture uses a scavenged ENDS PCU housing and original manufacturer’s
connectors to mimic the housing, connectors and pod retention employed in the original ENDS to
create a more realistic setup and produce accurate resistance measurements. This study employs
connectors and PCU housing of the same design as the ENDS device under test. Figure 3shows the
holding fixture built for an ALTO-style ENDS with a snap pod retainer as part of its internal structure.
The Model 34465A KEYSIGHT
DMM [
39
] is used in conjunction with each unique ENDS holding
fixture. Each test fixture is secured vertically using a table-top vise with the pod housing on the top
end while the four wires are to the bottom. This setup made it easy to position the test fixture and
switch between pods during the experiment.
The process has currently been demonstrated on two ENDS designs. Detailed tutorials for building
resistance measurement holding fixtures are available at [
37
]. The process of building the pod holding
fixture includes multiple steps which are briefly summarized here:
1. Discharge the battery of the ENDS prior to opening the device.
2.
Open the ENDS PCU to access its internal structure. Use standard electrical safety precautions
when working in the presence of possible charge carrying components such as capacitors.
Avoid shorting any electrical leads during disassembly.
3. Remove the battery and the readily accessible PCU electronic components.
4.
Locate the internal side of the connectors. These spring connectors are used to connect the PCU
to the pod or tank section of the ENDS. The spring connectors may be directly soldered to a
printed circuit board (PCB) as in the JUUL, or indirectly connected to the PCB via thin wires as in
the ALTO.
5.
Solder four lead wires, force +/
and sense +/
, to the ENDS PCU +/
connectors, respectively.
Care must be taken not to damage the connectors or gaskets which lie between the ENDS PCU
PCB and the END PCU pod receiver. Details of these connections are important in developing
an accurate and robust fixture and are discussed in detail and with photographic guides in [
37
].
In the case of the device built for ALTO, the four lead wires are soldered to the cut end of the
manufacturer’s thin wires linking the connectors to the PCB, taking care to protect the connector
and surrounding plastics case from soldering heat. These two thin wires will be added to the
measured coil resistance in addition to the resistance of the connectors themselves. These extra
resistances can be measured and subtracted from coil resistance or can be simply neglected if they
Int. J. Environ. Res. Public Health 2020,17, 7779 8 of 16
appeared to be very small. This point is discussed in detail in the Section 3. These thin wires are
inherently present in the ENDS circuitry to supply power to the coil. Their resistance is added to
the coil resistance contribute to the total resistance seen by the ENDS PCU. Some designs of ENDS
PCU may rely on these thin wires to dynamically sense coil resistance. Thus, including the thin
wires in the testing apparatus circuitry represents an accurate measure of the eective resistance
of the pod/coil assembly.
6.
Make a small groove at the end of the ENDS PCU housing to make room to pass the four wires
out from the fixture to the DMM.
7.
Connect color-coded banana plugs to the free end of the four lead wires for inserting into
the DMM.
Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 7 of 16
ENDS devices, the pods most often make electrical contact with the ENDS PCU using two different
style connectors. Several pod-style ENDS employ two spring-loaded connectors, also known as pogo
pin connectors, for the + and terminals on the PCU side while mating with two corresponding flat
connectors on the pod side. The connectors come in various sizes and shapes across ENDS designs
and, most importantly, they have different distances between the +/ terminals. Additionally, the
mechanical means of retaining the pod in the ENDS is different from one manufacturer’s design to
another. Some manufacturers use a friction fit, some use a magnetic clasp, and others may use
detents. This variation between ENDS designs means that the pod-style coil terminals are not directly
exposed and cannot directly connect to COTS Kelvin lead clips. Further, we desired a test fixture
which would allow us to assess the impact of the pod-retaining mechanisms employed in various
ENDS designs on the repeatability of effective coil resistance measurement. Electrical contact
resistance caused by the connectors is a potential confounder in natural environment operations of
ENDS products, and is worth investigation.
Accordingly, this study incorporates a special holding fixture which is unique to each ENDS
product and retains the pod with its integrated coil in position while measurements are made with
the DMM. Each holding fixture uses a scavenged ENDS PCU housing and original manufacturer’s
connectors to mimic the housing, connectors and pod retention employed in the original ENDS to
create a more realistic setup and produce accurate resistance measurements. This study employs
connectors and PCU housing of the same design as the ENDS device under test. Figure 3 shows the
holding fixture built for an ALTO-style ENDS with a snap pod retainer as part of its internal structure.
The Model 34465A KEYSIGHT™ DMM [39] is used in conjunction with each unique ENDS holding
fixture. Each test fixture is secured vertically using a table-top vise with the pod housing on the top
end while the four wires are to the bottom. This setup made it easy to position the test fixture and
switch between pods during the experiment.
Figure 3. Custom electronic nicotine delivery systems (ENDS) test fixture for Vuse/ALTO pod-style
ENDS coil resistance measurement in conjunction with a Keysight Model 34465A digital multimeter.
The exploded view shows an inside view of the wire connections.
Figure 3.
Custom electronic nicotine delivery systems (ENDS) test fixture for Vuse/ALTO pod-style
ENDS coil resistance measurement in conjunction with a Keysight Model 34465A digital multimeter.
The exploded view shows an inside view of the wire connections.
2.3. Data Sampling Procedure
Each pod to be tested is marked with a unique identification number. Prior to making a resistance
measurement, this ID number is entered into the data logging script. The operator places the pod
to be tested in the holding fixture and presses a button on the computer to initiate data sampling.
The script has the ability to read the resistance using the two-wire and four-wire modes without
additional operator intervention. The tested pod remains inserted between the two-wire and four-wire
measurements. The script can read and report a single observation of coil resistance or can read
and record 120 sequential readings of the same pod taken at one-second intervals, to evaluate the
performance of the fixture and validate stability. Each data reading is tagged with the pod ID, two-wire
vs. four-wire configuration, trial number (for test/re-test repeatability), a time-date stamp (to identify
120 sequential readings), and the numerical value of the resistance reported by the DMM. All readings
Int. J. Environ. Res. Public Health 2020,17, 7779 9 of 16
are made in the same lab under the same environments including the same storage condition and
room temperature.
2.4. Test Specimens
Two popular pod-style ENDS devices were chosen for this study. The Vuse ALTO [
35
] and
JUUL [
36
] ENDS are two of the most popularly used e-cigarettes especially among teenagers [
32
34
].
We purchased N=22 Vuse ALTO pods and N=16 JUUL pods filled with nicotine flavor e-liquid with
a manufacturer-reported 5% nicotine concentration. The manufacturers report that ALTO pods are
filled with ~1.8 mL of e-liquid [
35
], while JUUL pods are filled with ~0.7 mL of e-liquid [
36
]. All pods
were new and in the manufacturer’s original sealed packaging until opened for this test. Pods were
purchased from local retail brick-and-mortar establishments and national online vendors. We observed
that the Vuse ALTO pods were identified with N=5 unique manufacturing lots, and the JUUL pods
were identified with N=2 manufacturing lots.
2.5. Statistical Analyses
Descriptive statistics were computed for each ENDS design including the mean, median,
interquartile range and outlier analysis. Standard statistical tests were used to assess all data
collected. Assessment of bias evident between the two-wire and four-wire configuration of the constant
current resistance test method was conducted with a repeated-measure (single-sample) t-test for each
ENDS design studied. Assessment of the one-to-one intra-class correlation coecient, ICC
1:1
[
40
],
was conducted in the same lab using the same testing apparatus to assess the repeatability of the results
when the same sample of ALTO and JUUL pods were individually tested, and then re-tested after a
5 months interval of time. A total of 120 independent readings using the four-wire configuration of
each uniquely identified pod were taken on two dierent days, separated by five months between.
The test/re-test correlation coecient, r
test/re-test
[
40
], was conducted to test the consistency or reliability
of the measurement. Ten readings using the four-wire configuration were taken for each pod and
the pod was removed from the test fixture and inserted back between consecutive readings with a
2–5 s interval. The one-to-one intra-class correlation coecient, ICC
1:1
[
40
], and test/re-test correlation
coecient, r
test/re-test
[
40
], were computed and reported for each ENDS studied. Assessment of
dierences in mean eective coil resistance between ENDS designs was conducted using a two-sample
t-test under the assumptions of a normal distribution, unequal sample sizes and unequal variances.
The assumption of normally distributed samples was evaluated using a quantile–quantile (Q–Q) plot
of data observations vs. theoretical normal distribution. Assessment of manufacturing variation was
conducted using a normal distribution with point estimates for the mean and standard deviation to
predict the ±6σrange of coil resistances anticipated for mass-produced ENDS pods.
3. Results
Coil resistance test fixtures are built for two pod-style products: ALTO and JUUL. The constant
current resistance measurement method was used to measure N=22 ALTO pods and N=16 JUUL
pods with both the two- and four-wire lead configurations. A sampling distribution of the mean
was conducted for each test specimen, consisting of 120 repeated-measure resistance readings with
the digital multimeter at one-second intervals. Variation between repeated readings exhibited a 95%
confidence interval of less than 0.0002 (
) for every set of 120 repeated observations per pod and test
configuration (2 wire vs. 4 wire), indicating excellent stability of the test fixture.
The mean resistance value of the N=22 ALTO pods had a range of 1.018–1.304 (
) for the
two-wire configuration and 0.933–1.214 (
) for the four-wire configuration. The mean resistance of the
N=16 JUUL pods had a range of 1.631–1.744 (
) for the two-wire configuration and 1.544–1.659 (
)
for the four-wire configuration.
Figure 4shows a box plot of four groups of data: the two-wire and four-wire readings for
ALTO pods and JUUL pods. The mean resistance and standard deviation of N=22 ALTO pods was
Int. J. Environ. Res. Public Health 2020,17, 7779 10 of 16
observed to be 1.118 (0.053) (
) using the two-wire lead configuration and 1.031 (0.052) (
) using the
four-wire lead configuration. The ALTO data exhibits a slight positive skew and a paired t-test between
the two-wire lead and four-wire lead configurations exhibits a dierence in means of
δ
=0.087 (
)
(
p<0.001
). The mean resistance (and standard deviation) of N=16 JUUL pods was observed to be
1.710 (0.032) () using the two-wire lead configuration and 1.624 (0.033) () using the four-wire lead
configuration. The JUUL data exhibits a negative skew and the paired t-test between the two- and
four-wire configurations exhibits a dierence in means of
δ
=0.086 (
) (p<0.001). Results demonstrate
that the two-wire lead configuration exhibits a positive bias of
0.087 (
), as would be expected due
to the additional resistance (2
×
R
Lead
) present in the two-wire lead test configuration illustrated in
Figure 1. For this reason, only four-wire configuration data is taken to the next analysis step.
Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 10 of 16
0.087 (Ω) (p < 0.001). The mean resistance (and standard deviation) of N = 16 JUUL pods was observed
to be 1.710 (0.032) (Ω) using the two-wire lead configuration and 1.624 (0.033) (Ω) using the four-wire
lead configuration. The JUUL data exhibits a negative skew and the paired t-test between the two-
and four-wire configurations exhibits a difference in means of δ = 0.086 (Ω) (p < 0.001). Results
demonstrate that the two-wire lead configuration exhibits a positive bias of 0.087 (Ω), as would be
expected due to the additional resistance (2 × R
Lead
) present in the two-wire lead test configuration
illustrated in Figure 1. For this reason, only four-wire configuration data is taken to the next analysis
step.
Figure 4. Box plot and t-test results of ALTO (N = 22) and JUUL (N = 16) coil resistance readings for
the two-wire and four-wire configurations using the constant current method. The differences
between the two-wire and four-wire configurations are 0.087 (Ω) (p < 0.001) and 0.086 (Ω) (p < 0.001)
for ALTO and JUUL, respectively. The magenta difference bars are pointing at the group means while
the red line in the box plot refers to the group medians. The box notches illustrate the 95% confidence
interval on the median.
The test/re-test correlation coefficient, r
test/re-test
, and the one-to-one intra-class correlation
coefficient, ICC
1:1
, were computed to assess the repeatability of the testing apparatus. The one-to-one
intra-class correlation coefficient is conducted on two sets of readings. The first set is the means of
the 120 readings reported for in the previous paragraph. The second set consists of repeated
measurements for the same pods five months later. This comparison will show the stability of the test
fixture and pods over a period of five moths. The ICC
1:1
for ALTO (N = 13) is 0.9997 (13) p < 0.001 and
for JUUL (N = 16) is 0.9960 (16) p < 0.001. The test/re-test correlation coefficient is used to show the
consistency or reliability among measurements. A set of 10 consecutive readings with a 2–5 second
interval is taken for each pod. The test/re-test correlation coefficient for ALTO (N = 17) is 0.9997 (144)
p < 0.001 and for JUUL (N = 16) is 0.9873 (135) p < 0.001.
A quantile–quantile (Q–Q) plot of each data sample obtained using the unbiased four-wire lead
configuration is shown in Figure 5. The data was evaluated against several standard distributions,
and the normal distribution was found to be the closest fit to both ALTO and JUUL data, with a slight
positive and negative skew, respectively, consistent with the box plots in Figure 4.
Figure 4.
Box plot and t-test results of ALTO (N=22) and JUUL (N=16) coil resistance readings for
the two-wire and four-wire configurations using the constant current method. The dierences between
the two-wire and four-wire configurations are 0.087 (
) (p<0.001) and 0.086 (
) (p<0.001) for ALTO
and JUUL, respectively. The magenta dierence bars are pointing at the group means while the red line
in the box plot refers to the group medians. The box notches illustrate the 95% confidence interval on
the median.
The test/re-test correlation coecient, r
test/re-test
, and the one-to-one intra-class correlation
coecient, ICC
1:1
, were computed to assess the repeatability of the testing apparatus. The one-to-one
intra-class correlation coecient is conducted on two sets of readings. The first set is the means of the
120 readings reported for in the previous paragraph. The second set consists of repeated measurements
for the same pods five months later. This comparison will show the stability of the test fixture and
pods over a period of five moths. The ICC
1:1
for ALTO (N=13) is 0.9997 (13) p<0.001 and for JUUL
(N=16) is 0.9960 (16) p<0.001. The test/re-test correlation coecient is used to show the consistency
or reliability among measurements. A set of 10 consecutive readings with a 2–5 s interval is taken for
each pod. The test/re-test correlation coecient for ALTO (N=17) is 0.9997 (144) p<0.001 and for
JUUL (N=16) is 0.9873 (135) p<0.001.
Int. J. Environ. Res. Public Health 2020,17, 7779 11 of 16
A quantile–quantile (Q–Q) plot of each data sample obtained using the unbiased four-wire lead
configuration is shown in Figure 5. The data was evaluated against several standard distributions,
and the normal distribution was found to be the closest fit to both ALTO and JUUL data, with a slight
positive and negative skew, respectively, consistent with the box plots in Figure 4.
Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 11 of 16
Figure 5. Q–Q plot of the four-wire coil resistance readings of ALTO (N = 22) and JUUL (N = 16) for a
standard normal distribution. The data was tested against several standard random distributions and
normal distribution was found to be the closest fit.
Figure 6 shows a normalized histogram and fitted pdf for the four-wire resistance readings of
the ALTO and JUUL. The plot shows that both pod products exhibit manufacturing variations in coil
resistance. The mean and standard deviation of the sample distributions from the unbiased four-wire
lead configuration are 1.031 (0.067) (Ω) for the ALTO and 1.624 (0.033) (Ω) for the JUUL. A two-
sample t-test was conducted to assess the effective mean resistance between the two pod products,
demonstrating a result of δ 0.593 (Ω) (p < 0.001).
Figure 5.
Q–Q plot of the four-wire coil resistance readings of ALTO (N=22) and JUUL (N=16) for a
standard normal distribution. The data was tested against several standard random distributions and
normal distribution was found to be the closest fit.
Figure 6shows a normalized histogram and fitted pdf for the four-wire resistance readings of
the ALTO and JUUL. The plot shows that both pod products exhibit manufacturing variations in
coil resistance. The mean and standard deviation of the sample distributions from the unbiased
four-wire lead configuration are 1.031 (0.067) (
) for the ALTO and 1.624 (0.033) (
) for the JUUL.
A two-sample t-test was conducted to assess the eective mean resistance between the two pod
products, demonstrating a result of δ0.593 () (p<0.001).
Int. J. Environ. Res. Public Health 2020,17, 7779 12 of 16
Int. J. Environ. Res. Public Health 2020, 17, x FOR PEER REVIEW 12 of 16
Figure 6. Normalized histogram and fitted Probability Density Function (PDF) for four-wire coil
resistances of ALTO (N = 22) and JUUL (N = 16). The group mean resistance of ALTO pods is 0.593
(Ω) (p < 0.001) less than that of JUUL pods. The sample of ALTO pods tested exhibits more
manufacturing variability than the sample of JUUL pods.
4. Discussion
The four-wire constant current method is the preferred technique for quantifying electronic
cigarette coil resistance. Between-product comparisons may not be reliable when resistance
measurements are taken using the popular two-wire method. Thus, the popular two-wire method
should be avoided in future studies. For example, in our controlled setting, the two-wire method
introduced a bias of 0.086 (Ω), which was 15% of the observed difference between product means of
0.593 (Ω). If care is not taken with the wire lead resistance, the bias can be even larger than 15% and
obfuscate important variation between products. Similarly, if two-wire comparisons are made
between data collected by two different labs with two different apparatus, we may inadvertently
conclude that there is no significant difference between product characteristics or ascribe a difference
to the product which could actually be a result of the test apparatus. While two-wire resistance
measurement methods are common in many laboratory settings, they may not be sufficiently
accurate for measuring the resistance of electronic cigarette coils with a resistance of 3 (Ω) or lower.
Results demonstrated that a sample of coils from a single manufacturer procured across
manufacturing lots exhibit variation, which is a significant fraction of the nominal coil resistance.
Such manufacturing variation is expected. The amount of manufacturing variation in coil resistance
associated with a particular product may have significant implications on the emissions resulting
from the use of the product. The coefficient of variation (standard deviation over the mean) was
observed to be 5.1% for ALTO and 2.0% for JUUL. Using the sampling distribution of the mean, it
is reasonable to infer that the ±3σ coil resistance of ALTO and JUUL pods vary by as much as ±15%
and ±6% for ALTO and JUUL, respectively, when considering the true population of large production
lots typical of a national and global distribution channel.
Differences in coil resistance, whether associated with bias error from the two-wire
configuration, variation between products or variation within a single product, may yield important
variations in the emissions of total particulate matter (TPM) and presence of hazardous and
Figure 6.
Normalized histogram and fitted Probability Density Function (PDF) for four-wire coil
resistances of ALTO (N=22) and JUUL (N=16). The group mean resistance of ALTO pods is 0.593 (
)
(p<0.001) less than that of JUUL pods. The sample of ALTO pods tested exhibits more manufacturing
variability than the sample of JUUL pods.
4. Discussion
The four-wire constant current method is the preferred technique for quantifying electronic
cigarette coil resistance. Between-product comparisons may not be reliable when resistance
measurements are taken using the popular two-wire method. Thus, the popular two-wire method
should be avoided in future studies. For example, in our controlled setting, the two-wire method
introduced a bias of 0.086 (
), which was 15% of the observed dierence between product means of
0.593 (
). If care is not taken with the wire lead resistance, the bias can be even larger than 15% and
obfuscate important variation between products. Similarly, if two-wire comparisons are made between
data collected by two dierent labs with two dierent apparatus, we may inadvertently conclude
that there is no significant dierence between product characteristics or ascribe a dierence to the
product which could actually be a result of the test apparatus. While two-wire resistance measurement
methods are common in many laboratory settings, they may not be suciently accurate for measuring
the resistance of electronic cigarette coils with a resistance of 3 () or lower.
Results demonstrated that a sample of coils from a single manufacturer procured across
manufacturing lots exhibit variation, which is a significant fraction of the nominal coil resistance.
Such manufacturing variation is expected. The amount of manufacturing variation in coil resistance
associated with a particular product may have significant implications on the emissions resulting from
the use of the product. The coecient of variation (standard deviation over the mean) was observed to
be 5.1% for ALTO and 2.0% for JUUL. Using the sampling distribution of the mean, it is reasonable
to infer that the
±
3
σ
coil resistance of ALTO and JUUL pods vary by as much as
±
15% and
±
6% for
ALTO and JUUL, respectively, when considering the true population of large production lots typical of
a national and global distribution channel.
Int. J. Environ. Res. Public Health 2020,17, 7779 13 of 16
Dierences in coil resistance, whether associated with bias error from the two-wire configuration,
variation between products or variation within a single product, may yield important variations in
the emissions of total particulate matter (TPM) and presence of hazardous and potentially hazardous
constituents (HPHCs) in the emissions. Furthermore, the impact of variability in coil resistance in the
pod (or e-cigarette reservoir) on emissions is closely related to the electronics used in the power control
unit (PCU) of the e-cigarette. The PCUs of modern e-cigarettes are far more sophisticated than a simple
battery. However, examining the eect of variation in coil resistance on power dissipated in the coil by
a simple direct current (DC) circuit of a battery across a coil provides insight into the joint impacts of
both the coil and the PCU on emissions.
Consider an electronic cigarette with a nominal coil resistance of 1 (
) (measured using four-wire
configuration) and powered by a battery with a fully charged voltage of 3.7 (V). To illustrate the point,
consider the PCU as being a simple “on/o” switch with no active voltage or current control and no
supporting pulse width modulation. The nominal power dissipated will be
P=3.72
1.0 =
13.69
(W)
,
as given by Equation (1) and the nominal current flowing through the coil will be I
coil
=V
coil
/R
coil
=
3.7/1.0 =3.7 (A), as determined by Ohm’s Law, Equation (2). It is well known from DC circuit analysis
that the total energy delivered from the coil to the liquid is limited by the product of the nominal
power and duration of activation. Similarly, it is well known from heat and mass transfer analysis
that the rate of mass transfer from the liquid to the aerosol stream is aected by the surface area,
flow path, and flow rate. As the power, P, and energy increase, we can anticipate more emission of
TPM. As the current, I
coil
, increases, the temperature of the coil itself will increase through a well-known
phenomenon known as “Ohmic heating.” We might anticipate, then, that increases in current flowing
through the coil might give rise to increased production of HPHCs in the emissions resulting from
thermal decomposition of e-liquid constituents. Thus, understanding the influence of variation in the
coil resistance of e-cigarettes is essential to understanding emissions.
We demonstrated that the two-wire configuration introduced a positive bias, overestimating the
true value of the coil resistance by 0.086 (
). If we use this biased estimate of coil resistance, we would
underestimate the power,
P=3.72
1.086 =
12.6
(W)
, and current, I
coil
=3.6 (A), in the coil. These biased
estimates are not conservative, and could very well give rise to apparent inconsistencies observed
between the emissions produced when comparing products to one another.
The same concern holds true when we assess the eect of manufacturing variation between coils
of the same product design. A manufacturing variation of +/
15% in coil resistance with a simple PCU
would result in variations in coil power and current dissipation of
13%/+18%. As the manufacturing
variation in coils increases, the potential adverse consequences of changes in HPHC emissions also
increases. It has been reported that coils with lower resistance values may have higher negative health
impacts [
23
25
,
27
]. Therefore, it is essential to develop a full understanding of the manufacturing
variation in coil resistance associated with electronic cigarettes. This high dierence in the expected
instantaneous power has the potential to drastically change the performance of the device, constituents
of the aerosol produced, and the e-liquid consumption rate.
Electronic cigarette power control units (PCUs) which employ active voltage control may mitigate
the adverse impact of variation in coil resistance. Both the ALTO and JUUL are equipped with PCUs
exhibiting active control logic. The sophistication of PCUs varies widely between product designs.
The potential of the PCU to mitigate manufacturing variation in coil resistance deserves further
attention. Little is known about the performance of various PCUs and their limitations.
The robust method introduced herein provides a foundation for investigation of several research
questions, including the following. What resistance variation is exhibited between coils of dierent
products? What variation in resistance is exhibited within a product design? What are the eects of
resistance variation on the aerosol composition and generation rate? How does the resistance of a coil
change over the course of its operating life? How does the resistance of a coil change during a vaping
session? How do these variations aect the performance, lifetime and safety of the lithium battery?
What is the relation between coil resistance, power, and the performance of the PCU? To what extent
Int. J. Environ. Res. Public Health 2020,17, 7779 14 of 16
can the PCU overcome variations in coil resistance? What is the eect of using interchangeable coils
(such as the now common 510 threaded reservoir) in conjunction with PCUs on emissions? How does
this framework inform potential adverse health consequences of product misuse and product hacking?
5. Conclusions
The preferred constant current resistance measurement method using four-wire leads is
demonstrated to provide stable, accurate, repeatable and unbiased observations of the resistance of
pod-style electronic cigarette coil assemblies. The commonly employed two-wire lead configuration is
demonstrated to introduce a positive bias, the magnitude of which is dependent upon the laboratory
testing apparatus and impedes reproducing results between independent laboratories. The constant
current four-wire lead method is recommended as the standard method for measuring the resistance of
electronic cigarette coil assemblies. The coil resistance measurement method, which was demonstrated
using two brands of pod-style electronic cigarettes, is broadly applicable to other brands and styles of
electronic cigarettes.
A quantile–quantile assessment demonstrated that the manufacturing variation in the eective
coil resistance of pod assemblies for two brands of popular electronic cigarettes is normally distributed.
The mean coil resistance of pods varies significantly between the brands of electronic cigarettes tested,
demonstrating that the sample mean resistance of the ALTO pods was 0.593 (
) lower than the mean
resistance of JUUL pods (p<0.001). Furthermore, the sample of ALTO (N=22) pods tested exhibited a
mean and 99% confidence interval of 1.031
±
0.0405 (
), while that of the JUUL (N=16) pods exhibited
a mean and 99% confidence interval of 1.624
±
0.0243 (
). The resistance measurement method is
thus valuable for assessing variations between brands of electronic cigarettes and for quantifying
manufacturing variation which may be anticipated within a single product brand. Products exhibiting
larger manufacturing variation in coil resistance may result in the wider variability of emissions
generated from those products. As a result, it is recommended that the distribution of manufacturing
coil resistance (using the constant current four-wire lead method) be reported in future comprehensive
emissions studies and new product applications for coils and coil assemblies.
Author Contributions:
Conceptualization, Q.M.S. and E.C.H.; methodology, Q.M.S. and E.C.H.; software, Q.M.S.
and E.C.H.; validation, Q.M.S., E.C.H. and R.J.R.; formal analysis, Q.M.S. and E.C.H.; investigation, Q.M.S., E.C.H.
and R.J.R.; resources, E.C.H. and R.J.R.; data curation, Q.M.S. and E.C.H.; writing—original draft preparation,
Q.M.S. and E.C.H.; writing—review and editing, R.J.R.; visualization, Q.M.S. and E.C.H.; supervision, E.C.H.
and R.J.R.; project administration, E.C.H. and R.J.R.; funding acquisition, E.C.H. All authors have read and agreed
to the published version of the manuscript.
Funding:
Research reported in this publication was supported by the National Institute of Environmental Health
Sciences (NIEHS) of the National Institutes of Health and FDA Center for Tobacco Products (CTP) under award
number R21ES029984. The content is solely the responsibility of the authors and does not necessarily represent
the ocial views of the NIH or the Food and Drug Administration.
Acknowledgments:
The authors would like to acknowledge Gary DiFrancesco who assessed, with laboratory
hardware, the initial ENDS test apparatus build; Shehan Jayasekera for his help with software script for reading
from the DMM; Samantha Emma Sarles for her help with initial ENDS dissection.
Conflicts of Interest: The authors declare no conflict of interest.
References
1.
Robinson, R.J.; Eddingsaas, N.C.; DiFrancesco, A.G.; Jayasekera, S.; Hensel, E.C., Jr. A framework to
investigate the impact of topography and product characteristics on electronic cigarette emissions. PLoS ONE
2018,13, e0206341. [CrossRef] [PubMed]
2.
Bitzer, Z.T.; Goel, R.; Reilly, S.M.; Foulds, J.; Muscat, J.; Elias, R.J.; Richie, J.P., Jr. Eects of solvent and
temperature on free radical formation in electronic cigarette aerosols. Chem. Res. Toxicol.
2018
,31, 4–12.
[CrossRef] [PubMed]
3.
Behar, R.Z.; Luo, W.; McWhirter, K.J.; Pankow, J.F.; Talbot, P. Analytical and toxicological evaluation of flavor
chemicals in electronic cigarette refill fluids. Sci. Rep. 2018,8, 1–11. [CrossRef] [PubMed]
Int. J. Environ. Res. Public Health 2020,17, 7779 15 of 16
4.
Kosmider, L.; Sobczak, A.; Fik, M.; Knysak, J.; Zaciera, M.; Kurek, J.; Goniewicz, M.L. Carbonyl compounds
in electronic cigarette vapors: Eects of nicotine solvent and battery output voltage. Nicotine Tob. Res.
2014
,
16, 1319–1326. [CrossRef] [PubMed]
5.
Farsalinos, K.E.; Voudris, V. Do flavouring compounds contribute to aldehyde emissions in e-cigarettes?
Food Chem. Toxicol. 2018,115, 212–217. [CrossRef]
6.
Farsalinos, K.E.; Spyrou, A.; Tsimopoulou, K.; Stefopoulos, C.; Romagna, G.; Voudris, V. Nicotine absorption
from electronic cigarette use: Comparison between first and new-generation devices. Sci. Rep.
2014
,4, 4133.
[CrossRef]
7.
Eddingsaas, N.; Pagano, T.; Cummings, C.; Rahman, I.; Robinson, R.; Hensel, E. Qualitative analysis of
e-liquid emissions as a function of flavor additives using two aerosol capture methods. Int. J. Environ. Res.
Public Health 2018,15, 323. [CrossRef]
8.
Hensel, E.C.; Eddingsaas, N.C.; DiFrancesco, A.G.; Jayasekera, S.; O’Dea, S.; Robinson, R.J. Framework to
estimate total particulate mass and nicotine delivered to E-cig users from natural environment monitoring
data. Sci. Rep. 2019,9, 1–9. [CrossRef]
9.
Robinson, R.J.; Hensel, E.C. Behavior-based yield for electronic cigarette users of dierent strength eliquids
based on natural environment topography. Inhal. Toxicol. 2019,31, 484–491. [CrossRef]
10.
Kim, K.-H. Mass change tracking approach as collection guidelines for aerosol and vapor samples released
during e-cigarette smoking. Anal. Methods 2016,8, 2305–2311. [CrossRef]
11.
Korzun, T.; Lazurko, M.; Munhenzva, I.; Barsanti, K.C.; Huang, Y.; Jensen, R.P.; Escobedo, J.O.; Luo, W.;
Peyton, D.H.; Strongin, R.M. E-cigarette airflow rate modulates toxicant profiles and can lead to concerning
levels of solvent consumption. ACS Omega 2018,3, 30–36. [CrossRef] [PubMed]
12.
Robinson, R.; Hensel, E.; Al-Olayan, A.; Nonnemaker, J.; Lee, Y. Eect of e-liquid flavor on electronic cigarette
topography and consumption behavior in a 2-week natural environment switching study. PloS ONE
2018
,
13, e0196640. [CrossRef]
13.
Cengel, Y.; Ghajar, A. Heat and Mass Transfer: Fundamentals and Applications; McGraw-Hill Higher Education:
New York, NY, USA, 2014.
14. Tipler, P.A.; Mosca, G. Physics for Scientists and Engineers; W.H. Freeman: New York, NY, USA, 2003.
15.
Wens, M.; Steyaert, M. Design and Implementation of Fully-Integrated Inductive DC-DC Converters in Standard
CMOS; Springer Science & Business Media: New York, NY, USA, 2011.
16.
Wagener, T.L.; Floyd, E.L.; Stepanov, I.; Driskill, L.M.; Frank, S.G.; Meier, E.; Leavens, E.L.; Tackett, A.P.;
Molina, N.; Queimado, L. Have combustible cigarettes met their match? The nicotine delivery profiles
and harmful constituent exposures of second-generation and third-generation electronic cigarette users.
Tob. Control 2017,26, e23–e28. [CrossRef] [PubMed]
17.
Talih, S.; Salman, R.; Karaoghlanian, N.; El-Hellani, A.; Saliba, N.; Eissenberg, T.; Shihadeh, A. “Juice
Monsters”: Sub-ohm vaping and toxic volatile aldehyde emissions. Chem. Res. Toxicol.
2017
,30, 1791–1793.
[CrossRef] [PubMed]
18.
Soulet, S.; Duquesne, M.; Toutain, J.; Pairaud, C.; Mercury, M. Impact of vaping regimens on electronic
cigarette eciency. Int. J. Environ. Res. Public Health 2019,16, 4753. [CrossRef]
19.
McAdam, K.; Warrington, A.; Hughes, A.; Adams, D.; Margham, J.; Vas, C.; Davis, P.; Costigan, S.; Proctor, C.
Use of social media to establish vapers pung behaviour: Findings and implications for laboratory evaluation
of e-cigarette emissions. Regul. Toxicol. Pharmacol. 2019,107, 104423. [CrossRef]
20. SMOK Website. Available online: https://www.smoktech.com/(accessed on 10 April 2020).
21.
Vaporasso Website, TARGET PM80. Available online: https://www.vaporesso.com/vape-kits/target-pm80
(accessed on 24 April 2020).
22.
Innokin Website, Z-BIIP. Available online: https://www.innokin.com/vaporizers/z-biip/(accessed on
24 April 2020).
23.
Cirillo, S.; Urena, J.F.; Lambert, J.D.; Vivarelli, F.; Canistro, D.; Paolini, M.; Cardenia, V.;
Rodriguez-Estrada, M.T.; Richie, J.P.; Elias, R.J. Impact of electronic cigarette heating coil resistance on the
production of reactive carbonyls, reactive oxygen species and induction of cytotoxicity in human lung cancer
cells in vitro. Regul. Toxicol. Pharmacol. RTP 2019,109, 104500. [CrossRef]
Int. J. Environ. Res. Public Health 2020,17, 7779 16 of 16
24.
Cirillo, S.; Vivarelli, F.; Turrini, E.; Fimognari, C.; Burattini, S.; Falcieri, E.; Rocchi, M.B.L.; Cardenia, V.;
Rodriguez-Estrada, M.T.; Paolini, M. The customizable e-cigarette resistanceinfluences toxicological outcomes:
Lung degeneration, inflammation, and oxidative stress-induced in a rat model. Toxicol. Sci.
2019
,172, 132–145.
[CrossRef]
25.
Gillman, I.G.; Kistler, K.A.; Stewart, E.W.; Paolantonio, A.R. Eect of variable power levels on the yield of
total aerosol mass and formation of aldehydes in e-cigarette aerosols. Regul. Toxicol. Pharm.
2016
,75, 58–65.
[CrossRef]
26. Chausse, P.; Naughton, G.; Dutheil, F. Electronic cigarettes. Chest 2015,148, e29–e30. [CrossRef]
27.
Hiler, M.; Karaoghlanian, N.; Talih, S.; Maloney, S.; Breland, A.; Shihadeh, A.; Eissenberg, T. Eects of
electronic cigarette heating coil resistance and liquid nicotine concentration on user nicotine delivery,
heart rate, subjective eects, putopography, and liquid consumption. Exp. Clin. Psychopharmacol.
2019
.
[CrossRef] [PubMed]
28.
Sleiman, M.; Logue, J.M.; Montesinos, V.N.; Russell, M.L.; Litter, M.I.; Gundel, L.A.; Destaillats, H.
Emissions from electronic cigarettes: Key parameters aecting the release of harmful chemicals.
Environ. Sci. Technol. 2016,50, 9644–9651. [CrossRef]
29.
Ogunwale, M.A.; Li, M.; Ramakrishnam Raju, M.V.; Chen, Y.; Nantz, M.H.; Conklin, D.J.; Fu, X.-A.
Aldehyde detection in electronic cigarette aerosols. ACS Omega 2017,2, 1207–1214. [CrossRef]
30.
Soulet, S.; Duquesne, M.; Toutain, J.; Pairaud, C.; Lalo, H. Influence of coil power ranges on the e-liquid
consumption in vaping devices. Int. J. Environ. Res. Public Health 2018,15, 1853. [CrossRef]
31.
Coil Master Website. Available online: https://www.coil-master.net/product/coil-master-521-mini-v2
(accessed on 24 April 2020).
32.
Cullen, K.A.; Gentzke, A.S.; Sawdey, M.D.; Chang, J.T.; Anic, G.M.; Wang, T.W.; Creamer, M.R.; Jamal, A.;
Ambrose, B.K.; King, B.A. E-Cigarette use among youth in the United States, 2019. JAMA
2019
,322, 2095–2103.
[CrossRef]
33.
Tan, A.S.; Soneji, S.S.; Choi, K.; Moran, M.B. Prevalence of using pod-based vaping devices by brand among
youth and young adults. Tob. Control 2019. [CrossRef]
34.
Kavuluru, R.; Han, S.; Hahn, E.J. On the popularity of the USB flash drive-shaped electronic cigarette JUUL.
Tob. Control 2019,28, 110–112. [CrossRef]
35. VUSE Website. Available online: https://vusevapor.com/alto-complete-kit (accessed on 10 April 2020).
36. JUUL Webstie. Available online: https://www.juul.com/(accessed on 10 April 2020).
37.
Saleh, Q.; Hensel, E.; Robinson, R. Coil resistance testing apparatus for VUSE ALTO. Available online: https:
//protocols.io/view/coil-resistance-testing-apparatus-for-vuse-alto-bibnkame (accessed on 23 October 2020).
38.
Janesch, J. Two-wire vs. Four-wire Resistance Measurements: Which Configuration Makes Sense for Your
Application. Available online: https://doc.xdevs.com/doc/Keithley/Appnotes/2Wire_4Wire%20Resistance%
20Article.pdf (accessed on 23 October 2020).
39.
Digital Multimeters 34460A, 34461A, 34465A (6
1
2
digit), 34470A (7
1
2
digit). Available online: https:
//www.keysight.com/us/en/assets/7018-03846/data-sheets/5991-1983.pdf (accessed on 10 April 2020).
40.
McGraw, K.O.; Wong, S.P. Forming inferences about some intraclass correlation coecients. Psychol. Methods
1996,1, 30. [CrossRef]
Publisher’s Note:
MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
aliations.
©
2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).
... Furthermore, individuals have their own unique vaping/puff topographies which contributes to the complexity of coil lifetime. The short life expectancy of the TFV16 Mesh 0.17 Ω coil [28], as well as variations in the resistance of the coils resulting from the manufacturing process [29,30], could account for the variations in the electrical parameters we noted over time (i.e., 300 consecutive puffs) within a channel and between channels. Regardless, discovering imperfections associated with coil production is not the aim of this study. ...
... High atomizer temperatures can be an issue, particularly when puffing continuously (4 puffs/minute for 75 min or 300 puffs). High temperatures make it uncomfortable to handle the atomizer and, more importantly, are known to generate harmful chemicals which are released with the ECIG-generated aerosol [29][30][31][32][33]. For this reason, the target wattage of the ECIG generator was set at ≈120 W or 60 W for each atomizer in parallel, despite the company's recommendation of 120 W for each atomizer. ...
... In comparison, plateau temperatures between channel 1 and channel 2 displayed significant variation (average difference of~4.0 • C between channel 1 and channel 2 after the second and third sets of 100 puffs). It is unclear why this temperature difference exists other than to say the two atomizer coils are inherently different [28], including manufacturing variations that could impact the coil's resistance [29,30]. Below, a comparison is made between the current study and three other studies [31][32][33] found in the literature. ...
Article
Full-text available
Vaping (inhalation of electronic cigarette-generated aerosol) is a public health concern. Due to recent spikes in adolescent use of electronic cigarettes (ECIGs) and vaping-induced illnesses, demand for scientific inquiry into the physiological effects of electronic cigarette (ECIG) aerosol has increased. For such studies, standardized and consistent aerosol production is required. Many labs generate aerosol by manually activating peristaltic pumps and ECIG devices simultaneously in a predefined manner. The tedium involved with this process (large puff number over time) and risk of error in keeping with puff topography (puff number, duration, interval) are less than optimal. Furthermore, excess puffing on an ECIG device results in battery depletion, reducing aerosol production, and ultimately, its chemical and physical nature. While commercial vaping machines are available, the cost of these machines is prohibitive to many labs. For these reasons, an economical and programmable ECIG aerosol generator, capable of generating aerosol from two atomizers simultaneously, was fabricated, and subsequently validated. Validation determinants include measurements of atomizer temperatures (inside and outside), electrical parameters (current, resistance and power) of the circuitry, aerosol particle distribution (particle counts and mass concentrations) and aerosol delivery (indexed by nicotine recovery), all during stressed conditions of four puffs/minute for 75 min (i.e., 300 puffs). Validation results indicate that the ECIG aerosol generator is better suited for experiments involving ≤ 100 puffs. Over 100 puffs, the amount of variation in the parameters measured tends to increase. Variations between channels are generally higher than variations within a channel. Despite significant variations in temperatures, electrical parameters, and aerosol particle distributions, both within and between channels, aerosol delivery remains remarkably stable for up to 300 puffs, yielding over 25% nicotine recovery for both channels. In conclusion, this programmable, dual-channel ECIG aerosol generator is not only affordable, but also allows the user to control puff topography and eliminate battery drain of ECIG devices. Consequently, this aerosol generator is valid, reliable, economical, capable of using a variety of E-liquids and amenable for use in a vast number of studies investigating the effects of ECIG-generated aerosol while utilizing a multitude of puffing regimens in a standardized manner.
... Less attention has been given to the manufacturing variability of coil resistance and the relative impact on variations in aerosol yield. Our previous work [17] introduced a robust method to measure coil resistance of e-cigarettes and documented manufacturing variation in coil resistance of two popular pod-style ENDS: Vuse ALTO and JUUL. Pod 2 of 11 units included in the test showed variation in coil resistance of~30% and~7.4% for ALTO and JUUL, respectively. ...
... The test fixture presented in [17,29] was used to measure coil resistance, built by repurposing the housing of the PCU (power control unit) of the targeted ENDS, mimicking the geometrical and electrical conditions of the original ENDS. The test fixture provides measurement of the effective coil resistance which accurately represents the resistance seen by the PCU during operation. ...
... The coil resistance test fixture [17,29] was held vertically using a tabletop vise to ensure consistency in the measurement and minimize error resulting from motion. The test fixture was connected to the digital multimeter and communicated with the PES-1 personal computer via USB serial connection. ...
Article
Full-text available
This work investigated the effects of manufacturing variations, including coil resistance and initial pod mass, on coil lifetime and aerosol generation of Vuse ALTO pods. Random samples of pods were used until failure (where e-liquid was consumed, and coil resistance increased to high value indicating a coil break). Initial coil resistance, initial pod mass, and e-liquid net mass ranged between 0.89 to 1.14 [Ω], 6.48 to 6.61 [g], and 1.88 to 2.00 [g] respectively. Coil lifetime was µ (mean) = 158, σ (standard deviation) = 21.5 puffs. Total mass of e-liquid consumed until coil failure was µ = 1.93, σ = 0.035 [g]. TPM yield per puff of all test pods for the first session (brand new pods) was µ = 0.0123, σ = 0.0003 [g]. Coil lifetime and TPM yield per puff were not correlated with either variation in initial coil resistance or variation in initial pod mass. The absence of e-liquid in the pod is an important factor in causing coil failure. Small bits of the degraded coil could be potentially introduced to the aerosol. This work suggests that further work is required to investigate the effect of e-liquid composition on coil lifetime and TPM yield per puff.
... Each trial consisted of nominally 50 homogenous puffs, at the conclusion of which the mass decrease of the ENDS and the mass increase of TPM collected on a Cambridge filter pad was measured gravimetrically. The coil resistance of each product was measured using an unbiased four-lead method [4,5] before and after the emissions trials. The propylene glycol to glycerin ratio (PG:GL) of the un-puffed e-liquid was measured using NMR and the nicotine mass ratio of both the un-puffed e-liquid, fNic,ELiq, and aerosol collected on each pad, fNic, was determined using GC-MS. ...
... The constituent mass ratio of nicotine, or any other aerosol constituent of interest, may be dependent upon operating parameters such as coil power, temperature, and e-liquid composition in addition to user topography as shown in Equation (4). Investigating the mass ratio of constituents as a function of flow conditions enables research laboratories to leverage the work of one another. ...
Preprint
This study introduces and demonstrates a comprehensive, accurate, unbiased approach to robust quantitative comparison of Electronic Nicotine Delivery Systems (ENDS) appropriate for establishing substantial equivalence (or lack thereof) between tobacco products. The approach is demonstrated across a family of thirteen pen- and pod-style ENDS products. Methods employed consist of formulating a robust emissions surface regression model, quantifying the empirical accuracy of the model as applied to each product, evaluating relationships between product design characteristics and maximum emissions characteristics, and presenting results in formats useful to researchers, regulators, and consumers. Results provide a response surface to characterize emissions (total particulate matter and constituents thereof) from each ENDS appropriate for use in a computer model and for conducting quantitative exposure comparisons between products. Results demonstrate that emissions vary as a function of puff duration, flow rate, E-Liquid composition, and device operating power. Further, results indicate that regulating design characteristics of ENDS devices and consumables may not achieve desired public health outcomes; it is more effective to regulate maximum permissible emissions directly. Three emissions outcome measures (yield per puff, mass concentration and constituent mass ratio) are recommended for adoption as standard quantities for reporting by manufacturers and research laboratories. The approach provides a means of (a) quantifying and comparing maximal emissions from ENDS products spanning their entire operating envelope, (b) comparative evaluation of ENDS devices and consumable design characteristics, and (c) establishing comparative equivalence of maximal emissions from ENDS. A consumer-oriented product emissions dashboard is proposed for comparative evaluation of ENDS exposure potential. Maximum achievable power dissipated in the coil of ENDS is identified as a potentially effective regulatory parameter.
... The operator noted if any bubbles appeared to be generated in the ENDS Reservoir, if droplets were evident in the connection between the exit plane of the ENDS device and the surface of the filter pad, or if discrete droplets or gravity distribution of deposition pattern were evident on the filter pad. Likewise, the operator noted if there was a significant increase in the coil resistance between the "before" and "after" resistance measurements when using the fixture 4 wire resistance measurement method (39,40), and if so, would retire that reservoir from further testing to decrease the likelihood of using a failed coil in further trials. The operator noted whether each puff-activated ENDS appeared to consistently activate for every puff in the multi-puff sequence, or if the ENDS device operated unreliably. ...
... The primary outcome measures for the operating envelope (MinAF, MaxAF, MinAD, MaxAd) are shown in the upper portion of the table. The mean, median and standard deviation effective coil resistance using the four wire resistance measurement method (40), emissions nicotine mass ratio, unpuffed E-Liquid nicotine mass ratio, and solvent Propylene Glycol to Glycerin composition are reported for each product. ...
Article
Full-text available
Many Electronic Nicotine Delivery Systems (ENDS) employ integrated sensors to detect user puffing behavior and activate the heating coil to initiate aerosol generation. The minimum puff flow rate and duration at which the ENDS device begins to generate aerosol are important parameters in quantifying the viable operating envelope of the device and are essential to formulating a design of experiments for comprehensive emissions characterization. An accurate and unbiased method for quantifying the flow condition operating envelope of ENDS is needed to quantify product characteristics across research laboratories. This study reports an accurate, unbiased method for measuring the minimum and maximum aerosolization puff flow rate and duration of seven pod-style, four pen-style and two disposable ENDS. The minimum aerosolization flow rate ranged from 2.5 to 23 (mL/s) and the minimum aerosolization duration ranged from 0.5 to 1.0 (s) across the ENDS studied. The maximum aerosolization flow rate was defined to be when the onset of liquid aspiration was evident, at flow rates ranging from 50 to 88 (mL/s). Results are presented which provide preliminary estimates for the effective maximum aerosolization flow rate and duration envelope of each ENDS. The variation in operating envelope observed between ENDS products of differing design by various manufacturers has implications for development of standardized emissions testing protocols and data reporting required for regulatory approval of new products.
... The test fixture presented in [25,32] was used to measure coil resistance, built by repurposing the housing of the PCU (power control unit) of the targeted ENDS, mimicking the geometrical and electrical conditions of the original ENDS. The test fixture provides measurement of the effective coil resistance which accurately represents the resistance seen by the PCU during operation. ...
... The coil resistance test fixture [25,32] was held vertically using a table top vise, to ensure consistency in the measurement and minimize error resulting from motion. The test fixture was connected to the digital multimeter and communicates with the PES-1 personal computer via USB serial connection. ...
Preprint
Full-text available
This work investigated the effects of manufacturing variations including coil resistance, initial pod mass, and e-liquid color on coil lifetime and aerosol generation of Vuse ALTO pods. Random samples of pods were used until failure (where e-liquid was consumed, and coil resistance increased to high value indicating a coil break). Initial coil resistance, initial pod mass, and e-liquid net mass ranged between 0.89 to 1.14 [], 6.48 to 6.61 [g], and 1.88 to 2.00 [g] respectively. Coil lifetime with light color e-liquid was (mean) = 149, (standard deviation) = 10.7 puffs while pods with dark color e-liquid was = 185, = 22.7 puffs with a difference of ~36 puffs (p <0.001). Total mass of e-liquid consumed until coil failure was = 1.93, = 0.035 [g]. TPM yield per puff of all test pods for the first session (brand new pods) was = 0.0123, = 0.0003 [g]. During usage, TPM yield per puff of pods with light color e-liquid was relatively steady while it was continuously decreasing for pods with dark e-liquid. Coil lifetime and TPM yield per puff were not correlated with either variation in initial coil resistance or variation in initial pod mass. The absence of e-liquid in the pod is an important factor in causing coil failure. Small bits of the degraded coil could be potentially introduced to the aerosol. There is a potential correlation of e-liquid color with both coil lifetime and TPM yield per puff. Change of e-liquid color might have been a result of oxidation which changed some nicotine into nicotyrine.
... It may also be assumed that the delivery of the e-liquid to the heating zone occurs faster with a high PG content. In general, the characteristics of coils and their dependence on devices' parameters have been examined in a number of studies (Saleh et al., 2020;Zhao et al., 2016) but further research is still needed, especially regarding the influence of the composition of the liquid on the coil temperature. ...
Article
Full-text available
Electronic nicotine delivery systems (ENDS) generate an aerosol by vaporising e-liquids that usually consist of propylene glycol (PG), vegetable glycerine (VG), and other ingredients (water, nicotine, and flavours). The chemical and physical properties of these components have a significant effect on aerosol formation and must be identified in order to improve product attractiveness and assess the degree of health risks. The aim of this article is to provide a description of the composition of the e-liquid base and its impact on the physical properties of the liquid used and the behaviour of the aerosol generated and particles separately. For this purpose, 46 articles were selected using a series of keywords. Englishlanguage publications were chosen. The impact of the PG/VG ratio on the physical properties of the e-liquid (boiling point, viscosity, volatility, hygroscopicity), aerosol emission characteristics (refractive index, light scattering coefficient, particle size distribution, concentration, emission of harmful compounds), vape attractiveness (taste, “throat-hit”, “cloud effect”), nicotine flux, coil temperature, and puff topography is presented. The PG/VG ratio is strongly correlated with the emission of carbonyls, which has adverse health effects and should be optimised. Furthermore, PG and VG also affect the other important characteristics of the aerosol generated by ENDS, which impact on both attractiveness and the consumption of harmful compounds. These findings could be considered for further research with the aim of improving electronic nicotine delivery systems as this can reduce levels of toxicants. This can be achieved by optimising the geometry of the components with respect to heating power and e-liquid.
... Each trial was conducted with the ENDS oriented at an inclination angle of 30 degrees to ensure that e-liquid could not be gravity fed from the reservoir into the emissions collection system. The coil resistance of each product was measured using an unbiased four-lead method [4,5] before and after the emissions trials. The propylene glycol to glycerin ratio (PG:GL) of the unpuffed e-liquid was measured using NMR and the nicotine mass ratio of both the unpuffed e-liquid, f Nic, Unpuffed , and aerosol collected on each pad, f Nic , was determined using GC-MS. ...
Article
Full-text available
This study introduces and demonstrates a comprehensive, accurate, unbiased approach to robust quantitative comparison of electronic nicotine delivery systems (ENDS) appropriate for establishing substantial equivalence (or lack thereof) between inhaled nicotine products. The approach is demonstrated across a family of thirteen pen- and pod-style ENDS products. Methods employed consist of formulating a robust emissions surface regression model, quantifying the empirical accuracy of the model as applied to each product, evaluating relationships between product design characteristics and maximum emissions characteristics, and presenting results in formats useful to researchers, regulators, and consumers. Results provide a response surface to characterize emissions (total particulate matter and constituents thereof) from each ENDS appropriate for use in a computer model and for conducting quantitative exposure comparisons between products. Results demonstrate that emissions vary as a function of puff duration, flow rate, e-liquid composition, and device operating power. Further, results indicate that regulating design characteristics of ENDS devices and consumables may not achieve desired public health outcomes; it is more effective to regulate maximum permissible emissions directly. Three emissions outcome measures (yield per puff, mass concentration, and constituent mass ratio) are recommended for adoption as standard quantities for reporting by manufacturers and research laboratories. The approach provides a means of: (a) quantifying and comparing maximal emissions from ENDS products spanning their entire operating envelope, (b) comparative evaluation of ENDS devices and consumable design characteristics, and (c) establishing comparative equivalence of maximal emissions from ENDS. A consumer-oriented product emissions dashboard is proposed for comparative evaluation of ENDS exposure potential. Maximum achievable power dissipated in the coil of ENDS is identified as a potentially effective regulatory parameter.
... Furthermore, coil temperature may also be influenced by e-liquid composition, which changes the viscosity and heat capacity or by air flow rates in the device, as faster air flow rates have higher cooling effects. 43,48,49 Thus, a single vaping device may produce different temperature ranges for the same voltage input upon minor alterations in operational scenarios. 50 In addition, the aerosol emissions will change as a result of the users' puffing regimen. ...
Article
E-cigarette aerosol is a complex mixture of gases and particles with a composition that is dependent on the e-liquid formulation, puffing regimen, and device operational parameters. This work investigated mainstream aerosols from a third generation device, as a function of coil temperature (315-510 °F, or 157-266 °C), puff duration (2-4 s), and the ratio of propylene glycol (PG) to vegetable glycerin (VG) in e-liquid (100:0-0:100). Targeted and untargeted analyses using liquid chromatography high-resolution mass spectrometry, gas chromatography, in situ chemical ionization mass spectrometry, and gravimetry were used for chemical characterizations. PG and VG were found to be the major constituents (>99%) in both phases of the aerosol. Most e-cigarette components were observed to be volatile or semivolatile under the conditions tested. PG was found almost entirely in the gas phase, while VG had a sizable particle component. Nicotine was only observed in the particle phase. The production of aerosol mass and carbonyl degradation products dramatically increased with higher coil temperature and puff duration, but decreased with increasing VG fraction in the e-liquid. An exception is acrolein, which increased with increasing VG. The formation of carbonyls was dominated by the heat-induced dehydration mechanism in the temperature range studied, yet radical reactions also played an important role. The findings from this study identified open questions regarding both pathways. The vaping process consumed PG significantly faster than VG under all tested conditions, suggesting that e-liquids become more enriched in VG and the exposure to acrolein significantly increases as vaping continues. It can be estimated that a 30:70 initial ratio of PG:VG in the e-liquid becomes almost entirely VG when 60-70% of e-liquid remains during the vaping process at 375 °F (191 °C). This work underscores the need for further research on the puffing lifecycle of e-cigarettes.
Article
Electronic nicotine delivery systems (ENDS) have been associated with a dramatic increase in youth becoming addicted to nicotine following decades-long decline in cigarette smoking uptake. The United States Food and Drug Administration, Center for Tobacco Products (FDA/CTP) is responsible for regulating devices and consumable materials associated with ENDS. State and federal regulations regarding flavoring compounds in ENDS liquids (e-liquids) may be circumvented when vendors market refillable reservoirs side-by-side with noncompliant e-liquids. This study investigated the effect of third-party refillable versus manufacturer-supplied single-use reservoirs on total particulate matter (TPM) and nicotine emissions. The maximum TPM yield per puff was 5.6 times higher for the third-party (Blankz) reservoir (12.4 mg/puff) in comparison with the manufacturer’s (JUUL) reservoir (2.2 mg/puff), whereas the maximum TPM concentration was over 7 times higher for third party (0.200 mg/ml) versus manufacturer (0.028 mg/ml) pod. The third-party pod was tested with nicotine concentrations ranging from 0% to 4%. The mass ratio of nicotine present in the aerosol (mg Nic/mg TPM) was found to be approximately the same as the mass ratio of the e-liquid (mg Nic/mg e-liquid) for both pods and all 3 nicotine laden e-liquids tested. Toxicant exposure may increase when consumers use third-party pods with ENDS devices. Refillable reservoirs are a significant barrier to regulatory restrictions on potentially toxic additives to e-liquids. It is recommended FDA/CTP require emissions characterization of third-party reservoirs used with each ENDS they are compatible with and should be required to demonstrate no increased potential toxicant exposure in comparison with manufacturer-provided reservoirs.
Preprint
Full-text available
Measuring coil resistance of Electronic Nicotine Delivery Systems (ENDS) accurately is critical in any research studying the characteristics of electronic cigarettes and their effects on the performance of these devices. It has been shown in several papers that changing coil resistance has the potential to change the Hazardous and Potentially Hazardous Constituents (HPHC) of emissions and consequently health effects on users. This protocol describes how to build a test apparatus for coil resistance measurement for ENDS. This apparatus mimics the geometrical and electrical characteristics of the ENDS and thus provides accurate measurements of the effective coil resistance. The steps shown in this protocol are illustrated for creating a VUSE ALTO test apparatus, but the general idea can be applied to other devices.
Article
Full-text available
Introduction: The extent to which electronic cigarette users will compensate for lower nicotine eliquids has implications on the risk associated with regulating eliquid composition. This article elucidates topography as a compensatory mechanism by investigating the impact of nicotine strength on total particulate matter (TPM) and nicotine consumed per puff. Methods: Thirty-three experienced vape pen users were assigned an NJOY™ VapePen and AVAIL™ brand Tobacco Row eliquid with their usual nicotine strength (L = 6 mg/mL, M = 12 mg/mL, H = 18 mg/mL) and vaped through RIT’s wPUMTM vape pen monitor to record every puff during 1 week. Nicotine and TPM yield per puff was determined accounting for the impact of topography characteristics on emissions and used to compute participant-specific mean yield per puff. Results: Nicotine yields ranged from 0.01 to 0.05 mg/puff and varied widely within each group (L, M, and H nicotine strength). Group-wise mean flow rate was lower for L compared to M (p = 0.2) and duration was higher compared to M (p = 0.09). Larger TPM was consumed per puff for L compared to M (p = 0.07), yet nicotine per puff for L was less than M (p = 0.3). H users took smaller volumes than L (p = 0.1) or M (p = 0.17), and there was little difference between L and M (p = 0.47). Conclusions: Evidence was provided for topography as a compensatory mechanism. Use of low nicotine strength eliquids can increase TPM, which can lead to increase in HPHC. Regulatory review of new products should consider natural use topography and realistic use exposures to nicotine, TPM and HPHCs.
Article
Full-text available
Electronic cigarette (ECIG) nicotine delivery and other effects can be influenced by device and/or liquid characteristics and user puffing behavior. One class of ECIGs includes "sub-ohm" devices that incorporate heating coils with resistance less than 1 ohm (Ω), lower than that observed in conventional devices (e.g., ≥1.5 Ω). Relative to conventional ECIGs that operate at ≤10 W, low-resistance coils can be used to increase device power (e.g., 40-300 W). However, little is known about the individual and combined effects of ECIG power, manipulated by coil resistance, and liquid nicotine concentration on ECIG acute effects. Experienced ECIG users (N = 32) completed 4 sessions that differed by ECIG power and coil resistance (40.5 W, 0.5 Ω or 13.5 W, 1.5 Ω) and liquid nicotine concentration (3 or 8 mg/ml). In each session, participants used a 4.5-V Kanger SUBOX ECIG in a 10-puff directed and 60-min ad libitum bout. Nicotine delivery, heart rate, subjective effects, puff topography, and liquid consumption were measured. Nicotine delivery was greatest in the 8mg/ml+0.5Ω condition and lowest in the 3mg/ml+1.5Ω condition. The greatest reduction in abstinence symptoms were observed in the 8mg/ml+0.5Ω condition, although the highest ratings for satisfaction and liking were reported in the 3mg/ml+0.5Ω condition. Use of ECIGs containing 3 mg/ml nicotine liquid resulted in longer and larger puffs and increased puff frequency, though high-power/low-resistance ECIGs resulted in greater consumption of ECIG liquid. ECIG device and liquid characteristics and user puff topography should be considered simultaneously when making regulatory decisions aimed at protecting public health. (PsycINFO Database Record (c) 2019 APA, all rights reserved).
Article
Full-text available
Most recent studies on electronic cigarettes (e-cigs) have been carried out using vaping regimens consistent with mouth-to-lung inhalation (MTL) and not with direct-to-lung (DTL) inhalation. This paper aimed to characterizing the influence of inhalation properties (puff duration, puff volume, airflow rate) on the mass of vaporized e-liquid (MVE). Because the literature on DTL is non-existent, an intense vaping regimen consistent with DTL inhalation (i.e., puff volume = 500 mL) was defined. The use of a low or standard (ISO/DIS 20768) regimen and the proposed intense vaping regimen were first compared using the Cubis 1 Ω atomizer on a large power range, and then by using two atomizers below 1 Ω and two others above 1 Ω on their respective power ranges. An analysis of the e-cig efficiency on the e-liquid vaporization was proposed and calculated for each MVE. The intense vaping regimen allowed a broader power range in optimal heating conditions. MVE linearly increased with the supplied power, up to over-heating conditions at higher powers. Moreover, the e-cigs’ efficiencies were higher when low-resistance atomizers were tested at high powers. All these results highlighted that the generated vapor might be better evacuated when an intense vaping regimen is used, and illustrate the obvious need to define a suitable standardized vaping regimen consistent with DTL inhalation.
Article
Full-text available
The recent growth in e-cigarette use has presented many challenges to Public Health research, including understanding the potential for e-cigarettes to generate toxic aerosol constituents during use. Recent research has established that the way e-cigarettes are puffed influences the magnitude of emissions from these devices, with puff duration the dominant driving force. Standardised puffing machine methods are being developed to harmonise testing approaches across laboratories, but critical to their success is the degree with which they accurately reflect vapers real-world puffing behaviours (topography). Relatively limited data is available examining the way vapers puff, with significant inconsistencies between studies. Here we report the creation and analysis of a large database of public-domain vaping videos to establish e-cigarettes puffing behaviour in near natural settings. Over 300 videos containing 1200 puffing events from 252 vapers were obtained from social media sources, divided approximately equally amongst cigalike, Ego and Advanced Personal Vapouriser (“APV”, also referred to as “boxmod”) types of e-cigarettes. Analysis showed that similar mean puff durations were found for all three categories of vaping devices. This includes direct-to-lung as well as mouth-to-lung puffing behaviours. A 3 s puff duration, as used in the recently published ISO puffing standard ISO 20,768:2018, appears appropriate for average behaviours. However, the wide diversity of puffing durations observed amongst vapers means it may be challenging to identify a simple yet comprehensively representative single machine-puffing regimen for laboratory studies. A puff duration of around 5.6 s appears to represent 95th percentile puffing behaviours amongst vapers and may be an appropriate choice for scientists and regulators seeking an additional more intense puffing regime. A range of new behavioural patterns have been identified whose impact on aerosol exposure need to be considered. Public-domain video records of vapers provides valuable and accessible insights into real-world use behaviours. It is freely available, and constantly updated with new material, and therefore provides a valuable resource for scientists seeking to understand real-world vaping behaviours.
Article
Full-text available
A framework describing the joint effect of user topography behavior and product characteristics of one exemplar device on the total particulate mass (TPM) and aerosol constituent yield delivered to a user is presented and validated against seven user-specific ‘playback’ emissions observations. A pen-style e-cig was used to collect emissions across puff flow rates and durations spanning the range observed in the natural environment. Emissions were analyzed with GC-MS and used to construct empirical correlations for TPM concentration and nicotine mass ratio. TPM concentration was demonstrated to depend upon both puff flow rate and duration, while nicotine mass ratio was not observed to be flow-dependent under the conditions presented. The empirical model for TPM and nicotine yield demonstrated agreement with experimental observations, with Pearson correlation coefficients of r = 0.79 and r = 0.86 respectively. The mass of TPM and nicotine delivered to the mouth of an e-cig user are dependent upon the puffing behavior of the user. Product-specific empirical models of emissions may be used in conjunction with participant-specific topography observations to accurately quantify the mass of TPM and nicotine delivered to a user.
Article
Importance The prevalence of e-cigarette use among US youth increased from 2011 to 2018. Continued monitoring of the prevalence of e-cigarette and other tobacco product use among youth is important to inform public health policy, planning, and regulatory efforts. Objective To estimate the prevalence of e-cigarette use among US high school and middle school students in 2019 including frequency of use, brands used, and use of flavored products. Design, Setting, and Participants Cross-sectional analyses of a school-based nationally representative sample of 19 018 US students in grades 6 to 12 participating in the 2019 National Youth Tobacco Survey. The survey was conducted from February 15, 2019, to May 24, 2019. Main Outcomes and Measures Self-reported current (past 30-day) e-cigarette use estimates among high school and middle school students; frequent use (≥20 days in the past 30 days) and usual e-cigarette brand among current e-cigarette users; and use of flavored e-cigarettes and flavor types among current exclusive e-cigarette users (no use of other tobacco products) by school level and usual brand. Prevalence estimates were weighted to account for the complex sampling design. Results The survey included 10 097 high school students (mean [SD] age, 16.1 [3.0] years; 47.5% female) and 8837 middle school students (mean [SD] age, 12.7 [2.8] years; 48.7% female). The response rate was 66.3%. An estimated 27.5% (95% CI, 25.3%-29.7%) of high school students and 10.5% (95% CI, 9.4%-11.8%) of middle school students reported current e-cigarette use. Among current e-cigarette users, an estimated 34.2% (95% CI, 31.2%-37.3%) of high school students and 18.0% (95% CI, 15.2%-21.2%) of middle school students reported frequent use, and an estimated 63.6% (95% CI, 59.3%-67.8%) of high school students and 65.4% (95% CI, 60.6%-69.9%) of middle school students reported exclusive use of e-cigarettes. Among current e-cigarette users, an estimated 59.1% (95% CI, 54.8%-63.2%) of high school students and 54.1% (95% CI, 49.1%-59.0%) of middle school students reported JUUL as their usual e-cigarette brand in the past 30 days; among current e-cigarette users, 13.8% (95% CI, 12.0%-15.9%) of high school students and 16.8% (95% CI, 13.6%-20.7%) of middle school students reported not having a usual e-cigarette brand. Among current exclusive e-cigarette users, an estimated 72.2% (95% CI, 69.1%-75.1%) of high school students and 59.2% (95% CI, 54.8%-63.4%) of middle school students used flavored e-cigarettes, with fruit, menthol or mint, and candy, desserts, or other sweets being the most commonly reported flavors. Conclusions and Relevance In 2019, the prevalence of self-reported e-cigarette use was high among high school and middle school students, with many current e-cigarette users reporting frequent use and most of the exclusive e-cigarette users reporting use of flavored e-cigarettes.
Article
Electronic cigarette (e-cigarette; e-cig) use has grown exponentially in recent years despite their unknown health effects. E-cig aerosols are now known to contain hazardous chemical compounds, including carbonyls and reactive oxygen species (ROS), and these compounds are directly inhaled by consumers during e-cig use. Both carbonyls and ROS are formed when the liquid comes into contact with a heating element that is housed within an e-cig's atomizer. In the present study, the effect of coil resistance (1.5 Ω and 0.25 Ω coils, to obtain a total wattage of 8 ± 2 W and 40 ± 5 W, respectively) on the generation of carbonyls (formaldehyde, acetaldehyde, acrolein) and ROS was investigated. The effect of the aerosols generated by different coils on the viability of H1299 human lung carcinoma cells was also evaluated. Our results show a significant (p < 0.05) correlation between the low resistance coils and the generation of higher concentrations of the selected carbonyls and ROS in e-cig aerosols. Moreover, exposure to e-cig vapor reduced the viability of H1299 cells by up to 45.8%, and this effect was inversely related to coil resistance. Although further studies are needed to better elucidate the potential toxicity of e-cig emissions, our results suggest that these devices may expose users to hazardous compounds which, in turn, may promote chronic respiratory diseases.
Article
Despite the knowledge gap regarding the risk-benefit ratio of the electronic cigarette (e-cig), its use has grown exponentially, even in teenagers. E-cig vapour contains carcinogenic compounds (e.g., formaldehyde, acetaldehyde and acrolein) and free radicals, especially reactive oxygen species (ROS) that cause toxicological effects, including DNA damage. The role of e-cig voltage customization on molecule generation has been reported, but the effects of the resistance on e-cig emissions and toxicity are unknown. Here we show that the manipulation of e-cig resistance influences the carbonyls production from non-nicotine vapour and the oxidative and inflammatory status in a rat model. Fixing the voltage at the conventional 3.5 V, we observed that the amount of the selected aldehydes increased as the resistance decreased from 1.5 to 0.25 Ω. Under these conditions, we exposed Sprague Dawley rats to e-cig aerosol for 28-days, and we studied the pulmonary inflammation, oxidative stress, tissue damage and blood homeostasis. We found a perturbation of the antioxidant and phase-II enzymes, probably related to the increased ROS levels due to the enhanced xanthine oxidase and P450-linked monooxygenases. Furthermore, frames from scanning electron microscope showed a disorganization of alveolar and bronchial epithelium in 0.25 Ω group. Overall, various toxicological outcomes, widely recognized as smoke-related injuries, can potentially occur in e-cig consumers who use low-voltage and resistance device. Our study suggests that certain "tips for vaping safety" cannot be established, and encourages further independent investigations to help public health agencies in regulating the e-cig use.
Article
Background An estimated 10%–44% of youth and young adults have ever used JUUL, the leading e-cigarette brand in the USA,1 while 8%–9% reported past 30-day use of JUUL.2–5 Although there is growing attention on the prevalence of JUUL use, prevalence of using other brands of pod-based vaping devices is unknown. This information is important to assess whether newer brands are gaining popularity among young people and to complement sales data which do not track online purchases and sales through non-participating retailers or provide information about characteristics of users.1 This study assesses the prevalence of current use of JUUL, Suorin and Vuse across demographic and tobacco use characteristics among US youth and young adults. Methods Study population and procedure Data came from 2000 US youth and young adults recruited through the SSRS national online opt-in survey panel6: 15–17 years old (500 ever smokers, 500 never smokers) or 18–24 years old (500 who smoked within the past 30 days and had smoked >100 lifetime cigarettes, 500 who did not smoke in the past 30 days or had smoked <100 lifetime cigarettes). Participants completed an online survey from December 2018 to January 2019. Measures Participants were asked if they have ever used JUUL, Suorin and Vuse. The question for Vuse referred to using the brand of e-cigarettes in general and was not limited to the pod-based sub-brand (ie, Vuse Alto). Response options were past 30-day use, ever use but not in the past 30 days, aware of the product but never used it, or not aware of the product. We computed a derived variable indicating past 30-day use of any of the three brands (yes/no). Analyses We estimated the prevalence of using JUUL, Suorin and Vuse, stratified by youth and young adults and by individual characteristics. Data were weighted to reflect the distribution of the US population of these age groups by demographics and smoking status.7 8 A final adjustment was made so that the youth and young adult groups are representative of the nationwide age distribution.9 Results Of the three brands, JUUL use was most prevalent among youth (12.9%) and young adults (18.1%), followed by Suorin (4.4% and 6.7%, respectively), and Vuse (2.1% and 6.2%, respectively; figure 1). Overall, 15.0% of youth and 22.0% of young adults reported using any of these three brands in the past 30 days. Prevalence of using JUUL and Vuse were significantly higher among young adults than youth. Over three in four youth (76.4%) and two in three young adults (68.8%) are aware of JUUL. In comparison, less than half the youth are aware of Suorin (36.7%) and Vuse (42.2%). Among young adults, 37.0% are aware of Suorin and 51.9% are aware of Vuse. Among youth, ever and past 30-day smokers and ever tried non-cigarette tobacco products were associated with JUUL, Suorin and Vuse use (table 1). Additionally, non-Hispanic white youth were more likely to use JUUL (compared with other races/ethnicities), while youth in 11th or 12th grades were more likely to use Suorin (compared with 10th grade or below). Among young adults, being male and ever tried non-cigarette tobacco products was associated with JUUL, Suorin and Vuse use. Furthermore, ever and past 30-day smoking was associated with JUUL use, while past 30-day smoking was associated with Suorin use. Figure 1