Scholarly article on topic 'Transient measurement of the effective particle density of cigarette smoke'

Transient measurement of the effective particle density of cigarette smoke Academic research paper on "Chemical sciences"

CC BY
0
0
Share paper
Academic journal
Journal of Aerosol Science
OECD Field of science
Keywords
{"Cigarette smoke" / Tobacco / "Effective density" / CPMA}

Abstract of research paper on Chemical sciences, author of scientific article — Tyler J. Johnson, Jason S. Olfert, Ross Cabot, Conor Treacy, Caner U. Yurteri, et al.

Abstract The real-time effective particle density of cigarette smoke was determined using a Centrifugal Particle Mass Analyzer (CPMA) and Differential Mobility Spectrometer (DMS). A Puff Inhale Exhale (PIE) simulator was used to produce the smoke from various research and commercial cigarettes following the International Standard Organization (ISO) puffing parameters (35ml puff of 2s duration, every 60s) or the Health Canada Intense (HCI) puffing parameters (55ml puff of 2s duration, every 30s). The impact of modifying parameters, such as smoke mass, cigarette format, filter type, inhalation volume and mouth hold period, on the effective particle density was also investigated. All of the effective density functions were found to be independent of particle size within the bias uncertainty of the measurement system, indicating that the cigarette smoke particles likely had a spherical morphology. Trends in the average effective particle densities were observed for the different cigarettes and puffing parameters. While all of these shifts were within the bias uncertainty of the CPMA–DMS system, two-sample t-tests and the Tukey method were used to identify where the shifts were statistically probable. However due to the complexity of cigarette smoke, the aerosol mechanisms behind most of these shifts were unknown and require further investigation. For all of the tested cases the average effective particle density, considering puffs 3–6, varied from 1090 to 1518kg/m3, with a majority (9 out of 16 cases) falling within 1300 to 1394kg/m3. The Tukey method identified no statistical change in the effective particle density over the duration of an ISO puff, but it did identify significant differences between effective densities produced by different cigarettes.

Academic research paper on topic "Transient measurement of the effective particle density of cigarette smoke"

Contents lists available at ScienceDirect

Journal of Aerosol Science

journal homepage: www.elsevier.com/locate/jaerosci

Transient measurement of the effective particle density of cigarette smoke

Tyler J. Johnson a, Jason S. Olferta'*, Ross Cabotb, Conor Treacyb,

Caner U. Yurterib, Colin Dickens b, John McAugheyb, Jonathan P.R. Symonds'

a Department of Mechanical Engineering, University of Alberta, Edmonton, Alberta, Canada T6G 2G8 b British American Tobacco, Group Research & Development, Southampton SO15 8TL, UK c Cambustion Ltd., J6 The Paddocks, 347 Cherry Hinton Road, Cambridge CB1 8DH, UK

CrossMark

ARTICLE INFO

Article history: Received 16 February 2015 Received in revised form 8 May 2015 Accepted 13 May 2015 Available online 21 May 2015

Keywords: Cigarette smoke Tobacco

Effective density CPMA

ABSTRACT

The real-time effective particle density of cigarette smoke was determined using a Centrifugal Particle Mass Analyzer (CPMA) and Differential Mobility Spectrometer (DMS). A Puff Inhale Exhale (PIE) simulator was used to produce the smoke from various research and commercial cigarettes following the International Standard Organization (ISO) puffing parameters (35 ml puff of 2 s duration, every 60 s) or the Health Canada Intense (HCI) puffing parameters (55 ml puff of 2 s duration, every 30 s). The impact of modifying parameters, such as smoke mass, cigarette format, filter type, inhalation volume and mouth hold period, on the effective particle density was also investigated. All of the effective density functions were found to be independent of particle size within the bias uncertainty of the measurement system, indicating that the cigarette smoke particles likely had a spherical morphology. Trends in the average effective particle densities were observed for the different cigarettes and puffing parameters. While all of these shifts were within the bias uncertainty of the CPMA-DMS system, two-sample t-tests and the Tukey method were used to identify where the shifts were statistically probable. However due to the complexity of cigarette smoke, the aerosol mechanisms behind most of these shifts were unknown and require further investigation. For all of the tested cases the average effective particle density, considering puffs 3-6, varied from 1090 to 1518 kg/m3, with a majority (9 out of 16 cases) falling within 1300 to 1394 kg/m3. The Tukey method identified no statistical change in the effective particle density over the duration of an ISO puff, but it did identify significant differences between effective densities produced by different cigarettes.

© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC

BY license (http://creativecommons.org/licenses/by/4.0/).

1. Introduction

Tobacco smoking is a recognized risk to health that is dose-related (Doll, Peto, Boreham, & Sutherland, 2004; International Agency for Research on Cancer, 2004; U.S. Department of Health and Human Services, 2010). These risks can be assumed to result from chronic exposure to a range of tobacco smoke toxicants (Fowles & Dybing, 2003; U.S.

* Corresponding author. Tel.: +1 780 492 2341; fax: +1 780 492 2200. E-mail address: jolfert@ualberta.ca (J.S. Olfert).

http://dx.doi.org/10.1016/joaerosci.2015.05.006

0021-8502/© 2015 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).

Department of Health and Human Services, 2012) via mechanisms of inflammation or oxidative stress, or to exposure to individual or classes of toxicants that exert toxic effects through more specific mechanisms (Stratton, Shetty, Wallace, & Bondurant, 2001).

Tobacco smoke is a complex dynamic mixture partitioned between vapour and particle phases, of more than 6000 individual substances identified to date (Rodgman & Perfetti, 2013). Smoke has a complex composition as a consequence of its formation process, that is a combination of combustion, pyrolysis and distillation of the cut tobacco leaf (Baker, 1999). These combined processes at temperatures ranging from 600 °C in the smoulder to 950 °C whilst a puff is drawn initially generate a vapour cloud, which cools rapidly behind the cigarette coal from 600 °C to near ambient in a few milliseconds (Baker, 1975). Thus, supersaturation of the vapour cloud will occur. It is known that non-volatile substances such as metals also transfer to smoke, probably as a consequence of cellular eruption under thermal stress. These non-volatiles are a likely source of seed nuclei for subsequent condensation processes (Morie, 1977; Stöber, 1982) and promote droplet formation and condensation in the air stream rather than condensation onto the residual tobacco rod (although such condensation does occur). It is estimated that early aerosol concentrations are of the order of 1010 particles cm"3, but this will decrease by an order of magnitude over 2-3 s via coagulation (Ingebrethsen, 1986; Robinson & Yu, 1999). As these early physical behaviours are dominated by coagulation and condensation, there is an expectation that chemical composition should be relatively uniform and the droplets spherical, the latter having been demonstrated by Johnson et al. (2014).

Local and systemic doses of smoke and its individual constituents are an important part of the toxicological assessment, and deposition and retention mechanisms will vary with the physical and chemical form of each compound (Baker & Dixon, 2006; St. Charles, McAughey, & Shepperd, 2013). The deposition of particle (or droplet) bound species will be driven by inhalation behaviour and properties such as particle size (Baker & Dixon, 2006; International Commission for Radiological Protection, 1994; St. Charles et al., 2013). In this context the particle density of the smoke droplets may be a significant parameter in modelling particle behaviour on inhalation.

Differences in smoke composition and therefore potentially smoke effective density are more likely to arise from differences in tobacco blend types (US blend, Virginia, etc), cigarette construction (Regular, King Size, Super Slim, etc.) or puffing parameters (ISO versus HCI).

Effective particle density (pef) is an important aerosol characteristic and relates a particle's mobility diameter (dm) to its mass (mp) (McMurry, Wang, Park, & Ehara, 2002) by 6 mp

Peff = n d3p 0)

This property can be used to convert mobility size distributions to mass distributions (Adam, McAughey, McGrath, Mocker, & Zimmermann, 2009; Alderman & Ingebrethsen, 2011) or to model impaction and settling losses, such as in human lung deposition (Pichelstorfer, Winkler-Heil, & Hofmann, 2013; Saber & Bazargan, 2011; Zhang, Kleinstreuer, & Hyun, 2012).

Johnson et al. (2014) measured a steady-state average effective particle density of 1180 +113 kg/m3 from smoke produced using a smoking machine from a University of Kentucky 3R4F research cigarette using a Differential Mobility Analyzer (DMA) and CPMA system. Johnson et al. (2014) determined the size-resolved effective density was independent of particle mobility-size (i.e. the particles were spherical with constant density). Using a Millikan Cell with a one hour holding time and 1000 times dilution, Lipowicz (1988) measured an effective particle density of 1120 kg/m3 for particles larger than 1 mm from a University of Kentucky 1R3F research cigarette. Chen, Namenyi, Yeh, Mauderly, and Cuddihy (1990) also measured a mass-weighted average effective particle density of 1120 kg/m3 from a University of Kentucky 2R1F research cigarette using a cascade impactor and Scanning Electron Microscope (SEM) in parallel. However all of these studies measured the steady-state effective particle density of cigarette smoke where significant smoke aging occurred.

This research determines the real-time mass-resolved effective particle density of cigarette smoke using a CPMA-DMS system. The impact of different cigarettes and puffing parameters on the effective particle density produced is also investigated. By comparing these measurements to steady-state effective densities measured in previous studies, the effects of smoke aging on the effective particle density are determined. In the future these measurements can also be used to produce more accurate lung deposition models (Yurteri et al., 2014) or to calculate real-time cigarette smoke mass distributions from measured mobility-size distributions (i.e. from DMS measurements).

2. Experimental set-up

The experimental set-up used to generate and measure the real-time effective particle density of cigarette smoke is shown in Fig. 1. A Puff Inhale Exhale (PIE) simulator (Warrington, McGrath, Perkins, and Reavell, 2011) was used to generate the cigarette smoke as it mimics the human smoking process with a puffing phase, followed by a mouth hold period. A University of Kentucky 3R4F research cigarette smoked following the International Standard Organization (ISO) puffing parameters (ISO3308:2012, 2012) and completing the mouth holding period in a mouth cast, was the base case for this research. All of the other trials only changed one experimental variable from this case, such as cigarette type or puffing parameter. The different cigarettes smoked following the ISO puffing parameters are summarized in Table 1, and include research cigarettes from the University of Kentucky (3R4F, 1R5F) and CORESTA (CM6) and a series of commercial cigarettes used previously in a study of mouth level exposure (Ashley, Sisodiya, McEwan, McAughey, & Prasad, 2011) and a study of regional lung deposition (Yurteri et al., 2014). The different puffing parameters tested with the University of Kentucky 3R4F

Table 1

Cigarettes used with experimental set-up following ISO puffing parameters.

Cigarettes Circumference (mm) Length (mm) 'Tar'(NFDPM) (mg/cig) Nicotine (mg/cig) Charcoal filter % Tip ventilation Source

3R4F 24.8 84 9.4 0.73 No 30 A

CM6 24.5 83 14.0 1.4 No 0 B

1R5F 24.9 83.9 1.7 0.16 No 69 A

KSSS1 17 83 1.6 0.15 Yes 88 C

KSSS4 17 83 4.1 0.38 Yes 64 C

KSSS7 17 83 7.5 0.66 Yes 33 C

KSNC4 25 83 4.7 0.35 No 56 C

KSNC7 25 83 7.2 0.50 No 39 C

KSC1 25 83 0.5 0.05 Yes 82 C

KSC4 25 83 4.0 0.35 Yes 56 C

KSC7 25 83 7.6 0.63 Yes 43 C

A: University of Kentucky (2013); B: CORESTA (2013); C: Ashley et al. (2011).

research cigarette are summarized in Table 2. The research cigarettes chosen were commercially available products, which addressed a number of variables including tar yield under regulatory smoking regimes and naming convention. The naming convention represents KS (King Size standard 25 mm circumference) versus KSSS (King Size Super Slim 17 mm circumference), C (cellulose acetate filter containing charcoal) versus NC (non-charcoal cellulose acetate filter) and 1, 4 and 7 representing notional 'tar' deliveries. Measured nicotine and 'tar' (as Nicotine Free Dry Particulate Matter) are reported in Table 1 (ISO3308:2012, 2012).

For each case the real-time effective particle density of the cigarette smoke produced by the PIE simulator was measured using Cambustion's (Cambridge, United Kingdom) Centrifugal Particle Mass Analyzer (CPMA) and Differential Mobility Spectrometer (DMS500). A Kr-85 radioactive neutralizer was first used to charge the cigarette smoke particles, allowing CPMA classification to occur. A CPMA classifies particles by their mass-to-charge ratio by applying opposing electrostatic and centrifugal forces on the particles (Olfert & Collings, 2005). The centrifugal and electrostatic forces are generated by passing the particles between two spinning concentric cylinders with a voltage potential placed between them. The mobility size distribution of the mass-to-charge classified particles was then measured using a DMS500. A DMS measures the transient particle mobility size distribution by passing the particles between two concentric cylinders with a potential difference placed between them (Biskos, Reavell, & Collings, 2005; Reavell, Hands, & Collings, 2002). Twenty-two electrometer rings, stacked along the inside of the outer DMS classifier cylinder, measure the current generated from the particles discharging at impact. To increase the current measured on each ring, the particles are further charged using a Corona unipolar charger before entering the DMS classifier. The path of the particles from the DMS classifier inlet to the electrometer rings is dependent on their electric mobility diameter. An inversion matrix via a de-convolution algorithm is used to convert the measured ring currents to an electrical mobility size distribution. Therefore a CPMA-DMS system can measure the transient effective particle density at one CPMA setpoint (Johnson, Symonds, & Olfert, 2013).

The radioactive neutralizer used in this experiment applies a distribution of elementary charge as a function of the particle size. The multiply-charged particles (particles with greater than one elementary charge) from this distribution complicate the results obtained from a CPMA-DMS system. Since the CPMA classifies particles by their mass-to-charge ratio, a particle with n charges and n times the mass of the singly-charged setpoint (where n is a positive integer) will also be classified through the CPMA. As a result the DMS will also measure the mobility size of particles with masses that are a multiple of the singly-charged mass-to-charge setpoint of the CPMA. Uncharged particles are also able to pass through the CPMA if they are smaller than the cut-off mass determined by the CPMA rotational speed (Symonds, Reavell, & Olfert, 2013). These uncharged particles are then charged by the DMS corona charger and measured in the DMS classifier. Therefore the effects of uncharged and multiply-charged particles must be removed to accurately measure effective particle density using a CPMA-DMS system and this was done using the method described by Johnson et al. (2013).

The effects of uncharged particles were limited by lowering the cut-off mass of the CPMA. This was accomplished by operating the CPMA at higher rotational speeds by selecting a higher CPMA resolution of 10. At the largest mass-to-charge

Table 2

Puffing parameters used with the PIE simulator and University of Kentucky 3R4F research cigarettes.

Routine Number of Puff volume Inhalation volume Inhalation time Mouth hold Puff interval Ventilation blocking Mouth

puffs (ml) (ml) (s) (s) (s) (%) cast

ISO 8 35 500 5 2 60 0 Yes

ISO-1s 8 35 500 5 1 60 0 Yes

ISO-8s 8 35 500 5 8 60 0 Yes

ISO-In 8 35 1000 5 2 60 0 Yes

HCI 12 55 125 4 2 30 100 Yes

Mouth 8 35 500 5 2 60 0 No

setpoints selected during the campaign, the CPMA resolution was limited by the maximum voltage potential of the classifier (i.e. 1000 V). For these few cases the highest possible CPMA resolution was used. The effects of the remaining uncharged particles that passed through the CPMA were manually removed by setting the DMS rings currents generated by these particles to zero. Uncharged particles were identified as the source of these ring currents as the DMS particle mobility size distribution was bimodal with a clear definition between the peaks as described by Johnson et al. (2013).

The effects of multiply-charged particles were removed through a correction process developed by Johnson et al. (2013). In brief, the particle mobility size distribution of the unclassified cigarette smoke was measured directly from the PIE simulator using a DMS. The real-time size distribution for each experimental case was measured twice and averaged to account for small variations between each cigarette. The aerosol charge distribution generated by the radioactive neutralizer was approximated using Wiedensohler's (1988) bipolar charging model. The CPMA transfer function was then applied to this distribution to determine the size and concentration of charged particles that passed through the CPMA. The aerosol charge distribution generated by the DMS Corona charger was approximated by Fuchs' (1963) unipolar charging model. The DMS transfer function was then applied to determine the theoretical DMS ring currents generated by the mass-to-charge classified aerosol. Three scaling factors within the model were used to minimize the differences between the theoretical and experimental ring currents. After constrained minimization, the fraction of current generated by multiply-charged particles, as determined by the model, was removed from the DMS experimental ring currents. These corrected ring currents were reinverted to a particle mobility size distribution using the DMS inversion matrix. This mobility size distribution was fitted with a lognormal curve through chi-squared minimization and determined the Count Median Diameter (CMD) of the distribution. The CMD was then corrected using the DMS size calibration curve that was determined experimentally with known sizes of atomized polystyrene latex (PSL) particles. Using the CPMA mass-to-charge setpoint and the size calibrated CMD from the mobility size distribution of singly-charged particles the effective particle density was determined.

3. Results and discussion

Table 3 summarizes effective densities measured during this study for the range of experimental parameters studied, including type of cigarette, temporal changes in effective density, puffing parameters, and the effect of particle dielectric constant on the measurement system.

3.1. Measurement uncertainty

Johnson et al. (2013) determined the bias uncertainty of the CPMA-DMS system is 10% in terms of particle mobility size and 30.1% in terms of effective particle density. These uncertainties apply to all of the results shown herein. However since the same experimental set-up was used for all of the cases, approximately the same bias (or systematic) uncertainty was introduced to each measurement. Therefore, two sample t-tests or the Tukey method was used to identify statistically significant changes in the effective density between the different test cases.

These uncertainties do not account for specific sources of error, such as the volatile and semi-volatile components of the cigarette smoke particles, which can evaporate over time (Baker & Proctor, 1990; Chen et al., 1990; Ingebrethsen & Sears, 1989; Johnson, Olfert, Yurteri, Cabot, & McAughey, 2015) and thus influence the effective particle density. However since the CPMA-DMS system measured the real-time effective particle density, the only appreciable smoke aging time occurred during the mouth hold period which was controlled by the PIE simulator. Rapid particle evaporation could be a concern due to the significant sample dilution and pressure changes that occur within the sampling system and the DMS classifier. The DMS dilutes the CPMA-classified sample by a 6:1 ratio with a rotating disc dilutor and the sample is further diluted when mixed with the sheath flows in the DMS charger and classifier. The DMS classifier drops the pressure of the sample from approximately 1 bar to 0.25 bar. Evaporation would cause the DMS to measure a smaller CMD and thus overestimate the effective particle density, however the maximum sample residence time in the instrument is just 2 s making this unlikely. Ingebrethsen, Alderman, and Ademe (2011) stated that evaporation effects of cigarette smoke particles in the DMS were "relatively small" in terms of particle diameter, and furthermore the size calibration data taken by Johnson et al. (2013) with a liquid Di-Ethyl-Hexyl-Sebacate (DEHS) aerosol do not indicate any significant systematic bias.

Table 3

Summary of statistical analysis results with respect to cigarettes and measurement parameters.

Test case Variable # of Samples Mean (kg/m3) Standard deviation (kg/m3) Tukey grouping or t-test results

Research and commercial cigarettes KSNC7 17 1518 199 A

KSSS7 13 1513 198 A

KSC4 23 1429 213 A B

KSC7 21 1404 177 A B

KSSS1 18 1394 192 A B

KSSS4 17 1377 138 A B C

1R5F 15 1346 185 A BC

KSNC4 18 1307 133 BC

KSC1 22 1300 151 BC

CM6 18 1221 176 CD

3R4F 56 1090 79 D

Puff profile -2 s 36 1101 101 A

0s 56 1090 79 A

+ 2 s 51 1059 73 A

Mouth hold 8 s 22 1388 156 A

1s 18 1377 255 A

2s 56 1090 79 B

Dielectric constant 4.03 52 1090 73 A

8.72 56 1090 79 A

13.5 55 1077 78 A

Inhaled volume 500 ml 56 1090 79 95% CI for difference ( - 311.6, -130.8)

1000 ml 17 1311 172 t(18) — -5.14 p-Value — 0.000

Mouth cast Yes 56 1090 79 95% CI for difference (131.2,302.2)

No 16 1306 156 t(17)— 5.35 p-Value—0.000

Puff regime ISO 56 1090 79 95% CI for difference (20.8,139.9)

HCI 21 1170 123 t(26)— :2.78 p-Value—0.010

3.2. Base case

Since the CPMA-DMS system measures in real-time, it generates a large amount of data over a very short period. The sampling period of the CPMA-DMS system was limited by the time resolution of the DMS which was set at 0.1 s. However the offline algorithm described by Johnson et al. (2013) takes significant time to correct the data points for multiply-charged particles. In general, the values shown herein are the effective particle density determined over a narrow range of time within each puff. This time period was selected to be during the maximum particle number concentration of each puff. The averaging period of the measured values at the maximum concentration was where the unclassified mobility-size distribution was approximately constant (in terms of the CMD and geometric standard deviation), which was typically a 1-2 s period (or 10-20 DMS samples).

The particle masses and mobilities measured at the maximum particle number concentration from a University of Kentucky 3R4F research cigarette for the fourth ISO puff are shown in Fig. 2. The effective particle density can be calculated using Eq. (1) for each particle size. It was found that the effective particle density was independent of particle mobility size, within the bias uncertainty of the CPMA-DMS system, indicating that cigarette smoke particles likely have a spherical morphology. Therefore the remaining data is represented as the average effective density measured across the mobility size distribution for each individual puff. The average effective particle density for ISO puffs 3 to 6 measured using the CPMA-DMS system was 1090 + 21 kg/m3 with precision error only and 1090 + 329 kg/m3 with precision and bias uncertainties combined. The latter agrees with the average effective density determined by Johnson et al. (2014) of 1180 + 113 kg/m3 as determined using a two sample t-test1 (t[47] = 1.62, p=0.112).2

Johnson et al. (2014) measured the steady-state effective particle density using a DMA-CPMA system with University of Kentucky 3R4F research cigarettes following ISO puffing parameters. In their experiment, Johnson et al. (2014) collected smoke from University of Kentucky 3R4F research cigarettes in a Tedlar® bag filled with 9 L of dilution air over a period of several minutes. Despite the drastically different timescales (minutes versus seconds) between Johnson et al.'s (2014) steady-state measurements and this transient study, the volatility of the cigarette smoke had no statistical impact on the average effective particle density. Due to this agreement it is unlikely significant evaporation occurred in the sampling

1 The different bias uncertainties between the CPMA-DMS system and DMA-CPMA-CPC system were accounted for in the statistical analysis by calculating the standard error using the root mean square of the bias and precision uncertainty from each system.

2 Where t[DOF] is a statistical value used to quantify if significant difference between two groups exist at the calculated degree of freedoms (DOF) and p represents the probability of the two groups being statistically similar.

700 650 600 550 500 450 400 350 300

----Johnson et al., 2014 (Constant pef( of 1180 kg/m3)

-Boundaries of Johnson et al., 2014 Uncertainty (±113 kg/m3)

O ISO Puff 4 (Average p „ of 1105 kg/m3)

6 8 10

30 40 50 60

Particle Mass, m (fg)

Fig. 2. Mass and mobility of particles produced by ISO puff 4 from a University of Kentucky 3R4F research cigarette and compared to the mass-mobility measurements completed by Johnson et al. (2014), where the shaded region outlines the measurement uncertainty of that study.

Fig. 3. Average effective particle density of each ISO puff from various research cigarettes with the calculated standard deviation, where the mean effective densities that do not share a letter are identified as significantly different using the Tukey method with a 95% confidence.

system or DMS classifier as this would have caused the DMS to underestimate the CMD and the effective particle density to be overestimated by the CPMA-DMS system.

3.3. Tobacco product influence

The average effective particle densities at the maximum particle number concentration of each ISO puff from three different research cigarettes are shown in Fig. 3. The effective particle densities from puffs 1 and 2 were found to vary significantly based on how the cigarette lit. If the electric lighter was held too close to the cigarette an external flame would be generated, while too far away and the cigarette would go out half way through the first inhalation period. The effective particle densities from puffs 7 and 8 were also found to vary significantly as the flame might reach the cigarette filter by this point in the smoking cycle causing it to burn. This was especially common for the slim or narrow cigarettes, which burned faster. These variations in puffs are consistent with other studies. The cigarette puff chemistry data collected by Adam, Baker, and Ralf (2007) showed that the chemical composition of the particles produced by ISO puffs 1 and 2 were different than puffs 3-8, which had an approximately constant chemical composition. Therefore the average effective particle density (shown on the right end of the plot with the standard deviation) for each research cigarette and all of the results herein only considers puffs 3-6. The Tukey method (at a 95% confidence) identified the average effective densities produced by the 3R4F

Fig. 4. Average effective particle density of each ISO puff from various commercial cigarettes with the calculated standard deviation, where KSSS is a King Size Super Slim cigarette, KSNC is a King Size cigarette with No Charcoal filter and KSC is a King Size cigarette with a Charcoal filter. The mean effective densities that do not share a letter are identified as significantly different using the Tukey method with a 95% confidence.

œ 2 >

? 1 <13 Q. ¡t=

to <u <15 Q

1750 . 1500 -1250 -1000 -, 750 -, 500 -250 -0

Fig. 5. Average effective particle density at three different points of each ISO puff subsequently sampled from the mouth cast, from a University of Kentucky 3R4F research cigarette with the calculated standard deviation, where the mean effective densities that do not share a letter are identified as significantly different using the Tukey method with a 95% confidence.

and 1R5F research cigarettes were significantly different, while both were not significantly different from the average effective density produced by the CM6 research cigarette.

The effective particle densities from various commercial cigarettes smoked following the ISO puffing parameters were also determined as shown in Fig. 4. The number in each commercial cigarette identifier indicates the pack tar in mg per cigarette. All of the commercial cigarettes were 83 mm long, with the Super Slims having a 17 mm circumference and the King Sizes a 25 mm circumference. In most cases, the average effective particle density was found to increase slightly as cigarette tar content increased. However this trend was not identified by the Tukey method (at a 95% confidence) as significantly different. There was no observable trend in the change in the average effective particle density between the three categories of commercial cigarettes. The Tukey method did identify the higher effective densities produced by KSNC7 and KSSS7 cigarettes are significantly different than other cigarettes, it also identified the lower effective densities produced by KSC1 and KSNC4 cigarettes were significantly different than other cigarettes.

The effective particle densities produced by the commercial cigarettes were 19-39% higher than the base case (University of Kentucky 3R4F research cigarette). The reason for this difference is unknown, but likely statistically significant as the Tukey method (at a 95% confidence) identified significant differences between samples from the two groups as summarized in Table 3.

■c CO

ffi '(/>

1750 1500

1250 1000

750 500

1500 «

Fig. 6. Effect of the mouth hold period on the average effective particle density of each ISO puff from a University of Kentucky 3R4F research cigarette with the calculated standard deviation, where the mean effective densities that do not share a letter are identified as significantly different using the Tukey method with a 95% confidence.

3.4. Evolution within a puff

The change in effective particle density over the duration of an ISO puff for the base case (University of Kentucky 3R4F research cigarette smoked following the ISO puffing parameters) was also determined as shown in Fig. 5. The same averaging period as described above was used, but shifted two seconds before and after the maximum number concentration of each puff. The average effective particle density was found to decrease slightly over the duration of the puff. However this trend was found to not be significantly different by the Tukey method at a 95% confidence.

3.5. Puffing parameter influence

The impact of puffing parameters on the effective particle density was also investigated. The effects of halving and quadrupling the mouth hold period of the ISO puff parameters on the effective particle density from a University of Kentucky 3R4F research cigarette are shown in Fig. 6. These results were unexpected as only particle evaporation and coagulation were significant factors within the mouth hold period. Coagulation would not have affected the effective particle density as the particles are predominantly liquid (Johnson et al., 2014). Furthermore, evaporation is preferential to lighter or less dense components. Therefore evaporation of lighter components would have caused the effective particle density to increase with longer mouth hold periods rather than the parabolic trend observed. Further investigation is required to determine why mouth hold period has such an effect on the effective particle density.

The influence of the ISO inhalation volume on the effective particle density from a University of Kentucky 3R4F research cigarette is shown in Fig. 7. The larger inhalation volume produced a higher effective particle density than the base case and the values are significantly different as determined by a two sample t-test (t[18] = - 5.14, p=0). This shift could be caused by the lighter components of the particles evaporating within the mouth cast due to increased sample dilution.

The influence of passing the sample through a mouth cast on the effective particle density from a University of Kentucky 3R4F research cigarette is shown in Fig. 8. Completing the mouth hold period in a 3/4 in. tube rather than the mouth cast produced a higher effective particle density. The values are significantly different as determined by a two sample t-test (t [17] = 5.35, p=0) and the mechanisms behind this change are unknown and should be the focus of future work.

The effective particle densities produced from a University of Kentucky 3R4F research cigarette following the Health Canada Intense (HCI) puffing parameters (Canada, 2014) are shown in Fig. 9. The effective particle density produced by the HCI puffing parameters, averaged over HCI puffs 3-6, was slightly higher than the ISO puffing parameters. The similarity in effective particle density between the two cases is interesting (as determined by a two sample t-test; t[26] = 2.78, p=0.010), especially considering that changes in puffing parameters can generate different mobility size distributions. Adam et al. (2009) found that doubling the puff volume from 35 ml to 70 ml caused the CMD and particle number concentration to change from 226 + 6 nm to 183 + 3 nm and 6.63 x 1010 + 0.46 x 1010 cm- 3 to 7.50 x 1010 + 0.17 x 1010 cm- 3 respectively. This is similar to the trends observed by Ingebrethsen, Cole, and Alderman (2012) who found the CMD and number concentration produced from a 3R4F Kentucky reference cigarette changed from 217 + 8 nm to 185 + 8 nm and 1.21 x 109 + 0.15 x 109cm-3 to 3.14 x 109 + 0.36 x 109cm-3 respectively when the puff volume was increased from 35 cm3 to 55 cm3 both over a 2 s duration. These magnitudes are similar to the results measured by Fuoco, Buonanno,

Fig. 7. Effect of the inhalation volume on the average effective particle density of each ISO puff from a University of Kentucky 3R4F research cigarette with the calculated standard deviation.

Fig. 8. Effect of a mouth cast on the average effective particle density of each ISO puff from a University of Kentucky 3R4F research cigarette with the calculated standard deviation.

Stabile, and Vigo (2014) of a 165 nm CMD and total particle concentration of 3.14 x 109 + 0.61 x 109 cm"3 from a conventional cigarette with a 0.8 mg of nicotine per cigarette concentration.

3.6. Dielectric constant

The unipolar charging model developed by Fuchs (1963), used to approximate the charge distribution produced by the DMS corona charger in the multiply-charged particle correction process, requires the dielectric constant of the particles to be known. The dielectric constant of the cigarette smoke particles was calculated by assuming only its major components (tar, water and nicotine) were significant. The particle composition produced by either the ISO or HCI puffing parameters was approximated from the data collected by Counts, Morton, Laffoon, Cox, and Lipowicz (2005). By applying the dielectric constant of nicotine (Haynes, 2013), coal tar (Workman, 2001) and water (Haynes, 2013) to these compositions, the dielectric constant of the cigarette smoke particles was determined. Depending on the tobacco product smoked, the dielectric constant for an ISO puff varied between 4.03 and 13.5, while the HCI puff varied between 21.7 and 30.5. Choosing products of a similar nature to the ones tested in this study, dielectric constants of 8.72 for the ISO puff and 28.2 for the HCI puff were determined and used in the correction model. Using the above limits, the sensitivity of the base case effective

2500 „

Fig. 10. Sensitivity of the effective particle density results with the calculated standard deviation of the base case (University of Kentucky 3R4F research cigarette smoked following the ISO puffing parameters) to the dielectric constant used within the correction process, where the mean effective densities that do not share a letter are identified as significantly different using the Tukey method with a 95% confidence.

particle density to the dielectric constant was found to be small and the Tukey method at a 95% confidence did not identify any significant differences between the average effective densities calculated using the different dielectric constants (Fig. 10).

4. Conclusions and summary

The real-time effective density of cigarette smoke was found to be independent of the particle mobility size within the bias uncertainty of the CPMA-DMS system for all of the cases tested. This constant effective density function indicates that the cigarette smoke particles likely had a spherical morphology. The commercial cigarettes and modified ISO puffing parameters produced higher effective particle densities than the base case (a University of Kentucky 3R4F research cigarette following standard ISO puffing parameters). The reason for these shifts is unknown and requires further investigation. The ISO and HCI puffing parameters produced similar effective densities despite having drastically different particle mobility-size distributions. The change in effective density over the duration of each ISO puff was also found to be small. All of the measurements agreed within the bias uncertainty of the CPMA-DMS system. Due to this high bias uncertainty, the transient

ISO 3R4F case measured also agreed with the steady-state effective particle density determined in previous studies based on a two sample t-test (t[47] = 1.62, p=0.112).

Acknowledgements

This research was funded by British American Tobacco (Investments) Ltd. References

Adam, T., Baker, R.R., & Ralf, Z. (2007). Characterization of puff-by-puff resolved cigarette mainstream smoke by single photon ionization-time-of-flight mass spectrometry and principal component analysis. Journal of Agricultural and Food Chemistry, 55(6), 2055-2061.

Adam, T., McAughey, J., McGrath, C., Mocker, C., & Zimmermann, R. (2009). Simultaneous on-line size and chemical analysis of gas phase and particulate phase of cigarette mainstream smoke. Analytical and Bioanalytical Chemistry, 394(4), 1193-1203.

Alderman, S.L., & Ingebrethsen, B.J. (2011). Characterization of mainstream cigarette smoke particle size distributions from commercial cigarettes using a DMS500 fast particulate spectrometer and smoking cycle simulator. Aerosol Science and Technology, 45(12), 1409-1421.

Ashley, M., Sisodiya, A., McEwan, M., McAughey, J., & Prasad, K. (2011). Comparison of mouth level exposure to 'Tar' and nicotine in smokers of normal and superslim king size cigarettes in Romania*. Contributions to Tobacco Research, 24(6), 277-288.

Baker, R.R. (1975). Temperature variation within a cigarette combustion coal during the smoking cycle. High Temperature Science, 7(3), 236-247.

Baker, R.R. (1999). Smoke chemistry. In L.D. Davis, & M.T. Nielson (Eds.), Tobacco: Production, Chemistry and Technology. Blackwell: London, pp. 308-439.

Baker, R.R., & Dixon, M. (2006). The retention of tobacco smoke constituents in the human respiratory tract. Inhalation Toxicology, 18(4), 255-294.

Baker, R.R., & Proctor, C.J. (1990). The origins and properties of environmental tobacco smoke. Environment International, 16(3), 231-245.

Biskos, G., Reavell, K., & Collings, N. (2005). Description and theoretical analysis of a differential mobility spectrometer. Aerosol Science and Technology, 39 (6), 527-541.

Canada. (2014, December 8). Canada Government Tobacco Act: Tobacco Reporting Regulations, SOR/2000-273. Part 3: Emissions from designated tobacco products. Health Canada.

Chen, B.T., Namenyi, J., Yeh, H.C., Mauderly, J.L., & Cuddihy, R.G. (1990). Physical characterization of cigarette smoke aerosol generated from a Walton smoke machine. Aerosol Science and Technology, 12(2), 364-375.

CORESTA. (2013). Cooperation Centre for Scientific Research Relative to Tobacco-Reference Products. (http://www.coresta.org/Guides/Guide-No08-Monitor-Production_Apr09.pdf) Retrieved 05.12.13.

Counts, M., Morton, M., Laffoon, S., Cox, R., & Lipowicz, P. (2005). Smoke composition and predicting relationships for international commercial cigarettes smoked with three machine-smoking conditions. Regulatory Toxicology and Pharmacology, 41(3), 185-227.

Doll, R., Peto, R., Boreham, J., & Sutherland, I. (2004). Mortality in relation to smoking: 50 years observations on male British doctors. British Medical Journal, 328(7455), 1519.

Fowles, J., & Dybing, E. (2003). Application of toxicological risk assessment principles to the chemical constituents of cigarette smoke. Tobacco Control, 12(4), 424-430.

Fuchs, N. (1963). On the stationary charge distribution on aerosol particles in a bipolar ionic atmosphere. Geofisica pura e applicata, 56(1), 185-193.

Fuoco, F., Buonanno, G., Stabile, L., & Vigo, P. (2014). Influential parameters on particle concentration and size distribution in the mainstream of e-cigarettes. Environmental Pollution, 184(0), 523-529.

Haynes, W.M. (Ed.). (2013). CRC handbook of chemistry and physics (93rd ed.). CRC Press/Taylor and Francis: Boca Raton, FL.

Ingebrethsen, B.J. (1986). Aerosol studies of cigarette smoke. Recent Advances in Tobacco Science, 12, 54-142.

Ingebrethsen, B.J., Alderman, S.L., & Ademe, B. (2011). Coagulation of mainstream cigarette smoke in the mouth during puffing and inhalation. Aerosol Science and Technology, 45(12), 1422-1428.

Ingebrethsen, B.J., Cole, S.K., & Alderman, S.L. (2012). Electronic cigarette aerosol particle size distribution measurements. Inhalation Toxicology, 24(14), 976-984.

Ingebrethsen, B.J., & Sears, S.B. (1989). Particle evaporation of sidestream tobacco smoke in a stirred tank. Journal of Colloid and Interface Science, 131(2), 526-536.

International Agency for Research on Cancer (2004). Volume 83: Tobacco smoke and involuntary smoking. IARC: Lyon.

International Commission for Radiological Protection (1994). ICRP publication 66. Human respiratory tract model for radiological protection. Annals of the ICRP, 24, 1-3.

ISO3308:2012 (2012). Routine analytical cigarette-smoking machine - Definitions and standard. Routine analytical cigarette-smoking machine - Definitions and standard conditions. International Standard Organization: Geneva.

Johnson, T.J., Olfert, J.S., Cabot, R., Treacy, C., Yurteri, C.U., & Dickens, C., et al. (2014). Steady-state measurement of the effective particle density of cigarette smoke. Journal of Aerosol Science, 75, 9-16.

Johnson, T.J., Olfert, J.S., Yurteri, C.U., Cabot, R., & McAughey, J. (2015). Hygroscopic effects on the mobility and mass of cigarette smoke particles. Journal of Aerosol Science, 86, 69-78.

Johnson, T.J., Symonds, J.P.R., & Olfert, J.S. (2013). Mass-mobility measurements using a centrifugal particle mass analyzer and differential mobility spectrometer. Aerosol Science and Technology, 47(11), 1215-1225.

Lipowicz, P.J. (1988). Determination of cigarette smoke particle density from mass and mobility measurements in a millikan cell. Journal of Aerosol Science, 19(5), 587-589.

McMurry, P., Wang, X., Park, K., & Ehara, K. (2002). The relationship between mass and mobility for atmospheric particles: A new technique for measuring particle density. Aerosol Science and Technology, 36(2), 227-238.

Morie, G.P. (1977). Selective filtration of tobacco smoke components: A review. Recent Advances in the Chemical Composition of Tobacco and Tobacco Smoke: New Orleans, Louisiana.

Olfert, J., & Collings, N. (2005). New method for particle mass classification - The Couette centrifugal particle mass analyzer. Journal of Aerosol Science, 36 (11), 1338-1352.

Pichelstorfer, L., Winkler-Heil, R., & Hofmann, W. (2013). Lagrangian/Eulerian model of coagulation and deposition of inhaled particles in the human lung. Journal of Aerosol Science, 64(0), 125-142.

Reavell, K., Hands, T., & Collings, N. (2002). A fast response particulate spectrometer for combustion aerosols. SAE International: Warrendale 2002-01-2714.

Robinson, R., & Yu, C. (1999). Coagulation of cigarette smoke particles. Journal of Aerosol Science, 30(4), 533-548.

Rodgman, A., & Perfetti, T.A. (2013). The chemical components of tobacco and tobacco smoke second ed.). CRC Press: Boca Raton.

Saber, E.M., & Bazargan, M. (2011). Dynamic behavior modeling of cigarette smoke particles inside the car cabin with different ventilation scenarios. International Journal of Environmental Science & Technology (IJEST), 8(4), 747-764.

St. Charles, F.K., McAughey, J., & Shepperd, C.J. (2013). Methodologies for the quantitative estimation of toxicant dose to cigarette smokers using physical, chemical and bioanalytical data. Inhalation Toxicology, 25(7), 383-397.

Stober, W. (1982). Generation, size distribution and composition of tobacco aerosols. In E. F. Litzinger, C. F. Mattina, & L. P. Bush (Eds.), Recent advances in tobacco science (Vol. 8, pp. 3-41).

Stratton, K., Shetty, P., Wallace, R., & Bondurant, S. (Eds.). (2001). Clearing the Smoke: Assessing the science base for tobacco harm reduction. National Academies Press: Washington, DC.

Symonds, J.P.R., Reavell, K.S., & Olfert, J.S. (2013). The CPMA-Electrometer System - A Suspended Particle Mass Concentration Standard. Aerosol Science and Technology, 47(8), i-iv.

U.S. Department of Health and Human Services. (2010). How tobacco smoke causes disease: The biology and behavioral basis for smoking-attributable disease: A report of the surgeon general. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health.

U.S. Department of Health and Human Services. (2012). [Docket No. FDA-2012-N-0143] Harmful and potentially harmful constituents in tobacco products and tobacco smoke; Established list. [Docket No. FDA-2012-N-0143] Harmful and potentially harmful constituents in tobacco products and tobacco smoke; Established list. (F. a. Administration, Ed.) Federal Register. 77, No. 64. Food and Drug Administration.

University of Kentucky. (2013). Reference Cigarette Program. <http://www2.ca.uky.edu/refcig/> Retrieved 12.11.13.

Warrington, P., McGrath, C., Perkins, J., Reavell, KDickens, C., &. (2011). An instrument to simulate puffing, inhaling and exhaling cigarette smoke. Manchester, UK: European Aerosol Conference.

Wiedensohler, A. (1988). An approximation of the bipolar charge distribution for particles in the submicron size range. Journal of Aerosol Science, 19, 387-389.

Workman, J.J. (2001). 26 - Dielectric constants of materials (in alphabetical order). In J. Workman (Ed.), The handbook of organic compounds. Academic Press: Burlington, pp. 275-292.

Yurteri, C. U., Dickens, C. J., McGrath, C. M., Cabot, R. A., Hawke, J., Perkins, J. W., et al. (2014). Estimating the regional deposition of tobacco smoke in the human respiratory system. October 24-25, 2014, Beckman Center Auditorium, Irvine, California: Advances in Aerosol Dosimetry Research.

Zhang, Z., Kleinstreuer, C., & Hyun, S. (2012). Size-change and deposition of conventional and composite cigarette smoke particles during inhalation in a subject-specific airway model. Journal of Aerosol Science, 46(0), 34-52.