Modelling dynamics of cigarette smoke in denuder

Modelling dynamics of cigarette smoke in denuder tubes
Lukas Pichelstorfer, Werner Hofmann
Department of Materials Research and Physics, University of Salzburg, Hellbrunnerstr. 23, A-5020 Salzburg, Austria
.
Motivation
Results
Cigarette smoke is a very complex aerosol containing several thousand
components that are distributed between the particle and the vapor
phase. Further, particle concentrations in fresh mainstream cigarette
smoke are very high (109 – 1010 cm-3) and particles agglomerate
quickly.[1] As a consequence, direct characterization of it is very
demanding and typically causes changes to the aerosol.
The initial model predictions exceeded the experimentally observed
nicotine deposition (< 0.2% s/cm2).[3] However, there are several
mechanisms that potentially alter the partial nicotine vapor pressure and
therefore evaporation and deposition rates:
Cigarette smoke
PSD
vapor
? Model input ?
composition
In order to determine the particle size
distribution (PSD) and concentration correctly
the aerosol is diluted with clean air to prevent
coagulation artefacts. Thereby the volatiles’
concentrations in the vapor phase are reduced
leading to a loss of these compounds in the
particle phase. Determination of the particle
composition can be carried out by filter
sampling and further analysis. Thus, an aged
aerosol is investigated in this case. Simply
combining the measurement results of vapor
phase, particle phase and particle size
distribution might consequently cause severe
errors.
•Coagulation increases the mean particle diameter thus reducing the
Kelvin term and therefore nicotine vapor pressure at the particle surface.
•Temperature differences between the tube wall and the incoming
aerosol result in a considerable but limited decrease in nicotine
deposition rate in terms of time.
•Water loss to the tube walls boosts the nicotine evaporation rate.
•Diffusion limited transport within the particles slows down the
transport between vapor and condensed phase but requires high
viscosity (100 Pa·s for 40% reduction).
•Finally, nicotine protonation shows the highest potential to explain the
experimental findings. Nicotine salts are non-volatile and smoke pH
measurements suggest most of the nicotine to be in this non-volatile
protonated form (N=35 : pH = 5.36 – 5.88). [4] A 65% to 77% nicotine
protonation at the tube entrance and steady increase to 80% to 92%
within the first third of the tube shift the simulation results into the range
of experimental findings.
Another way to characterize smoke aerosols is by
investigating their interactions with system boundaries. This
indirect measurement does not require any manipulation of
the aerosol. An examples of indirect measurement is the
denuder tube. Freshly generated cigarette smoke is directly
drawn into a vertically arranged glass tube, where the inner
surface of the tube is coated in order to capture nicotine. It
has been found that the nicotine is initially solely present in
the particle phase but deposits as a vapor.[2]
airflow
As the denuder tube experiments are probably the best
defined ones amongst the cigarette smoke experiments, we
chose to use them in order to apply a newly developed
aerosol dynamics model ADiC (Aerosol Dynamics in
Containments).[3]
Fig. 2: Effect of fixed degree of nicotine protonation on nicotine deposition rate in a denuder
tube. The lines represent the model results and the shaded area the range of experimental
data obtained from the literature
Methods
Numerical solution of the aerosols dynamics using a computer model
Representation of particle size distribution:
fixed bins containing moving distributions -> no numerical diffusion
Processes considered:
Coagulation, phase transition, heat transport, vapor transport, mixing/
dilution, deposition
Aerosol altering processes are
calculated consecutively. After
each process, a control parameter
checks whether the changes
within the time step are below a
threshold. Each process has one
or more characteristic control
parameters (e.g.: particle number
concentration
in
case
of
coagulation). If one or more of the
relevant quantities calculated
during one time step exceed
these
limits,
all
changes
calculated within this time step
are undone. Then, the time step is
divided
by
two
and
the
calculations for the new time
period start again. Otherwise,
numerical integration with respect
to time is carried out.
Conclusion
Although a considerable number of publications and data on cigarette
smoke and its behavior exist, it is still impossible to perform reliable
simulations that predict evolution of and interaction with system
boundaries. The reason for this is the complexity of cigarette smoke and
that the largest fraction of the data available focuses on specific aspects
of the cigarette smoke.
Systematic and simulation supported measurement of cigarette smoke.
References
1. Alderman, S.L. and B.J. Ingebrethsen, 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, 2011. 45(12): p. 1409-1421.
2. Lewis, D.A., I. Colbeck, and D.C. Mariner, Diffusion Denuder Method for
Sampling Vapor-Phase Nicotine in Mainstream Tobacco Smoke. Analytical
Chemistry, 1994. 66(20): p. 3525-3527.
3. Pichelstorfer, L. and W. Hofmann, Modeling aerosol dynamics of cigarette
smoke in a denuder tube. Journal of Aerosol Science, 2015. 88: p. 72-89.
4. Lauterbach, J.H., Bao, M., Joza, P.J., and Rickert, W.S., Free-base
nicotine in tobacco products. Part I. Determination of free-base nicotine in the
particulate phase of mainstream cigarette smoke and the relevance of these
findings to product design parameters. Regulatory Toxicology and
Pharmacology, 2010. 58(1): p. 45-63.
Fig. 1: Flow diagram of the ADiC model. C stands for control
parameter, H/V is the abbreviation of Heat/Vapor and t stands for
time
This research was funded in part by British American Tobacco (Investments)
Limited, Southampton, UK (Grant no. A10012363SV).