zz
Z-7
.~V b !.% . -, - "C',
-,6.7
.JRI. -19 A
R
AL
Aj
ZOUZ660


RESTRICTED'
GRADC RESEARCH PROGRAMME
STATUS REVIEW NOTES
covering the period
November 1984 April 1985
DISTRIBUTION:
Mr. A.L. Heard
Dr. R. Binns
Mr. R.G. Nicholls
Dr. C.J.P. de Siqueira
Mr. P.J. Dunn
Dr. S.R. Massey
Herr E. Rittershaus
Dr. F. Seehofer
co
1985 British American Tobacco Co. Ltd. This must
riot be copied or shown to unauthorised pe~
BATCo document for Province of British Columbia 11
November 1999


RESTRICTED
GR&DC RESEARCH PROGRAMME
STATUS REVIEW NOTES
covering the period
November 1984 - April 1985
These Status Review Notes are based on the Work
Areas described in
the 'GR&DC Research Programme and Resource
Allocation, 1984 1986".
Contents:
Page
Work Area 01 Biological Research 1
02 Smoke Research 5
03 Smoker Behaviour 20
04 Tobacco Treatment 27
05 Flavour and Biotechnology 35
06 Process Research
(A) Process Technology 41
(B) Physics of Tobacco 46
(C) Novel Cigarette Technology 50
(D) Measurement Technology 59
07 Analysis and Test Method Development 60
-Ij
co
1985 Bridsh Amerien Toba= Co. LA& TWs a" W be =pW
or *xwn to umthodud pasom
BATCo document for Province of British Columbia 11
November 1999


WORK AREA 01
BIOLOGICAL RESEARCH
U-4
c0
C:)
1985 BriUsh Amrien ToWwo Co. UA TWs mu" wt be copW
or sbom to wmawdsW pasom Ln
BATCo document for Province of British Columbia 11
November 1999


STATUS REVIEW NOTE
Work Area: 01 Biological Research
Author: G. Smith
Period Covered: November 1984 - April 1985
OKECTIVE:
To assess, in quantitative terms, the biological
activity
(?s defined by in-vitro models and in-vivo animal
inhalation
st;dies) of cigarette mainstream and ambient smoke
and be in a
pc,sition to evaluate the significance of product
modifications.
AREAS OF RESEARCH:
W Parameters affecting the biological activity of
smoke: It
is essential to realise how variations in the
physical and
chemical characteristics of a cigarette influence
the
biological activity of smoke.
(2" Short-term tests: It is unsound to rely on any
one assay.
The aim is to have knowledge of a number of tests
which,
when used together, provide a relevant base for
evaluating
new products.
PKGRESS:
W Parameters Affecting the Biological Activity of
Smoke
The commercial cigarettes supplied by the CAC
countries as part
of Project RIO have been evaluated using the Ames
bacterial
muttagenicity test and their relative activity
determined. In
thr--e of the four countries, a BAT product was
the most active.
The samples from Canada were generally the least
active compared
to those from Australia, Brazil and Germany. This
is probably
due to the absence of Burley tobacco, which is
noted as having
a vajor influence on the activity of condensate.
Filter ventil-
at~on also affects activity such that low delivery
cigarettes
tend to have condensate with a relatively high
specific activ-
ity. Naturally, this must be balanced against the
reduced
delivery of smoke.
Several parameters known to influence the Ames
mutagenicity did
not stand out (e.g. stem) as levels did not vary
greatly between
products.
cc
LM Bnush Amencan Tobacco Co. LuL Thu must not be
copigd or show to unaudkmized pcoons.
CP\
BATCo document for Province of British Columbia 11
November 1999


Observed differences in bacterial mutagenicity
between products
could, at times, not b-. explained as there were
too many design
variables between the cigarettes.
Level of nicotine did not correlate with activity.
Inhalation of whole smoke from the commercial
cigarettes of
Project RIO induced drug metabolising enzymes in
the rat lung.
The level of induction was similar for all
products.
Cigarette circumference has been shown to
influence the carcino-
genicity of condensate as measured by mouse skin
painting.
Examining cigarettes over a broader range (13 mm -
29 mm) has
verified that circumference has a significant
effect on mutagen-
icity as measured using the Ames test. Condensate
from ultra-
slim cigarettes is 2.5 times less active than
condensate from
fat cigarettes. As the slim ~_roduct has a lower
delivery, the
per cigarette activity is a great deal less. Slim
cigarettes
are also less likely to induce the drug
metabolising enzymes.
The inhalation toxicity of whole smoke from the
same series of
cigarettes did not show any significant
differences between
products.
The observed differences noted with cigarettes of
different
circumference may be due to the effect of changing
combustion
characteristics. This possibility was pursued by
examining the
influence of puff volume. In general, increased
puff volume
reduces the specific Ames mutagenicity of the
condensate. In
contrast, there is an increased response in the
rat respiratory
tract following inhalation of whole smoke
generated by the
higher puff volumes. It is noteworthy that, when
condensate
only is examined, as in the Ames test, increased
linear flow
rate, i.e. slim cigarettes and increased puff
volume, reduces
activity. With inhalation toxicity of whole smoke,
circumf-
erence has little effect whilst puff volume is
positively
related to activity. In the former case, the
vapour phase to
particulate phase ratio does not alter; In the
latter, at
higher puff volumes there is an increase in
certain vapour
phase components for a constant level of
particulates.
The interpuff frequency of cigarettes has been
found to have
little effect on the specific Ames mutagenicity of
condensate.
The Japanese Toray company has presented further
data indicating
a reduction in Ames activity using their ion
exchange resin.
In a joint experiment, using cigarettes
manufactured by the
Toray scientists, we were unable to confirm this
observation.
This has been reported to the Japanese company,
who have yet
to reply.
co
C:>
1995 British American Tobacco Co. LUL This must
not be copied or shown to unautim%ised penom _1J
BATCo document for Province of British Columbia 11
November 1999


(2) Short-Term Tests
(a) Yeast and mammalian cell mutagenicity
Following early difficulties, the yeast-based
system is now
established and we are in a position to compare
products.
This parallels work being carried out under
contract
(Microtest Ltd.) where a mammalian cell
mutagenicity assay
is being established. Toxicity problems have been
overcome
and condensates from standard products are now
being
evaluated. Both of the above tests are noted in
the
legislature as a natural progression from the Ames
test.
(b) Biochemistry
The development of the drug metabolising enzyme
package
is now complete. It involves the measurement of
enzymes
in rats thought by some experts to produce
potentially
harmful (viz. P448) or harmless (viz. P450)
metabolites.
Enzymes capable of handling both types of
metabolite will
also be measured. Using this system it has been
shown
that nicotine does not induce P448.
Research being sponsored at Surrey University is
examining
the relationship between enzyme specificity and
the shape
of the molecule. This should provide fundamental
informa-
tion on how a compound is altered by P448 and
P450.
FUTURE WORK:
Unless further requests are received from the CAC
countries,
the commercial aspect of Project RIO is now
complete. The
final report is in preparation.
Smoke inhalation studies in progress in rats
include the exam-
ination of fresh sidestream smoke and the
influence of heat
treatment during primary processing on the
activity of main-
stream smoke. The influence of magnesium oxide and
potassium
citrate on sidestream activity is being examined
using the
Ames test .
The influence of the combustion characteristics on
biological
activity is being investigated, in particular the
significance
of changes in the part i cul ate/vapour phase
ratio. A wide
spectrum of flow rates is to be examined by
smoking cigarettes
of variable circumference using a range of puff
volumes.
Annular cigarettes from Novel Cigarette Technology
(Work Area
06 M) may also be examined. The eventual aim is to
be able
to explore the relationship between combustion
characteristics
of a product and its biological activity.
(J-i
co
0 1985 British American Tobacco Co. Ltd. This must
not be copied or shovin go unnutborised persons.
CC)
BATCo document for Province of British Columbia 11
November 1999


WORK AREA 02
SMOKE RESEARCH
c0
C:>
0 1985 Bhtish A-fic- Tobcco Co. La& Thi. .t ~ be
copied or shown to o_dwdxCd pgraom 10
BATCo document for Province of British Columbia 11
November 1999


- 4 -
Due to problems in manufacture, the examination of
tobacco
blends will be later than originally planned.
However, it
should be started within the next few months.
Work will continue to further refine the yeast and
mammalian
cell assays.
Human studies
The company is financing a 3-year project at
Southampton Univer-
sity examining the influence of smoking on food
utilisation.
The project, now at the end of its first year, has
started to
produce results. From the observations to date, it
appears
that female smokers maintain a lower body fat
content and yet
have a higher energy intake when compared with
matched non-
smokers. The female smoker has a higher resting
metabolic rate
and a greater thermic response to a standard meal
than her non-
smoking counterpart, i.e. tends to "burn off" the
food rather
than deposit body fat.
None of these findings is applicable to the males
studies to
date, possibly due to the poor matching of these
subjects.
cc
1985 British American Tobacco Ca, Lid. This must
not be copied or shown to unauthoris4d persons.
<--
BATCo document for Province of British Columbia 11
November 1999


- 5 -
STATUS REVIEW NOTE
Work Area: 02 Smoke Research
Author: R.R. Baker
Period Covered: November 1984 - April 1985
OBJECTIVE:
To quantify the physical and chemical processes
which contribute
to the levels of substances in mainstream,
sidestream and ambi-
ent smoke, as a basis for designing and developing
new products.
AREAS OF RESEARCH:
(1) Sidestream Smoke: To study how cigarette
parameters and
construction affect sidestream emission, how they
may be
utilised to reduce sidestream visibility,
irritation and
aroma, and the build-up of ambient smoke in the
GR&DC
environmental room as a function of room
conditions.
(2) Enhanced Sensory Filters: To apply principles
of smoke
filtration and gas/smoke flow to the development
of new
filter concepts which have, inter alia, enhanced
sensory
characteristics.
(3) Smoke Formation: To quantify the key physical
and chemical
processes occurring in a burning cigarette, their
interac-
tions and effects on smoke formation and
deliveries.
(4) Nitrosamines and Chemical Reaction Mechanisms:
To provide
basic chemical information related to tobacco
curing,
aging, processing and burning, relevant to the
fonnation
of modified levels of specific components in
smoke,
including nitrosamines.
PROGRESS:
(1) Sidestream Smoke
(a) Chemical and physical modification of
cigarette papers
The previously described study relating to
chemicals that
reduce visible sidestream smoke has been
completed. Over
ninety chemicals were added to cigarette paper and
seventy
of them reduced the visible sidestream plume by
more than
40%, as measured using the optical density
technique. The
cc
1995 British American Tobcco C. Lai This must am
be copied or shown to unauthorised persons.
BATCo document for Province of British Columbia 11
November 1999


- 6 -
thermal decomposition characteristics of the
chemicals have
also been determined using both derivative
thermogravimetry
(DTG) and differential scanning calorimetry (DSC).
The bulk of the chemicals appear to reduce the
sidestream
visibility in line with an earlier hypothesis that
water
from the chemical released at 300-500*C is
important.
However, there are some exceptions to this and, as
yet,
no alternative explanation has been developed for
these
exceptions.
The DTG and DSC studies have also shown that, when
the
chemicals are applied to the cigarette paper, both
the
paper decomposition temperature and the heat given
out by
the paper decomposition are reduced. Furthermore,
the
lower the quantity of heat given out by the
decomposing
paper, the lower the static burn rate of the
resultant
cigarette.
The visibility of the sidestream plume in the
above work
has been measured using an optical density system
developed
previously. Three further optical density
instruments are
currently being built - one for B&W, one for
Central Analy-
tical Services (Work Area 07) and one for
Biological
Research (Work Area 01).
Previous work has correlated sidestream nicotine
with side-
stream aroma, irritation and annoyance. The
cigarette
paper pH has been modified by the application of
inorganic
and organic acids (15 in total) in an attempt to
reduce
the sidestream nicotine yields. Initial results
have
shown that the pH of conventional papers
containing chalk
or titanium dioxide as fillers can be reduced by
the appli-
cation of acid. This also extends to papers
containing
tri-potassium citrate as a burn additive. However,
it
is virtually impossible to reduce the pH of
cigarette
papers containing magnesium oxide/hydroxide when
acids
are applied.
The application of citric acid to conventional and
low
visibility sidestream cigarettes containing
potassium
citrate has produced reductions in sidestream
nicotine
yields of the order of 10-15%. This has manifested
itself
in small reductions in sidestream aroma,
irritation and
annoyance. Certain of these subjective parameters
have
been reduced to a statistically significant
extent. There
has been no apparent change in the gross objective
or
subjective results relating to the mainstream
smoke charac-
teristics of these cigarettes when compared to
control
cigarettes made from the same cigarette papers
without
acid treatment. Further work is being carried out
with
this approach in order to bring about greater
reductions
inisidestream aroma, irritation and annoyance.
Currently, C=)
a series of double wrapped cigarettes using acid
treated
papers is being evaluated.
CO
190 British American Tobacco Co. Lt& This must
itot be copied or shown to unauthorised persons.
NJ
BATCo document for Province of British Columbia 11
November 1999


(b) Other subjective studies on sidestream smoke
Subjective studies have been conducted aimed at
determining
the effect of further alteration of the
equilibrium ratio
between sidestream particulate and vapour phases.
The
total sidestream smoke from both conventional and
low
visibility cigarettes has been subjectively
compared to
the sidestream smoke to which quantities of
sidestream
vapour phase have been added. This was
accomplished by
comparing:
M whole sidestream smoke obtained when 660 mg of
the cigarette smouldered
to (ii) (whole sidestream smoke from 440 mg
tobacco)
(sidestream vapour phase from 220 mg tobacco).
No significant changes were observed with either
conve-
ntional or low visibility cigarettes.
It is therefore tentatively concluded that further
reduc-
tion in sidestream visibility, beyond those that
can be
currently achieved (about 45%) with compensating
increases
in vapour phase, will have little effect on
sidestream
aroma, irritation and annoyance. This conclusion
applies
to existing combinations of tri-potassium citrate
and
magnesium hydroxide.
(c) Ambient smoke
All the analytical instruments in the GR&DC
environmental
rooms have now been interfaced to the Sirius
minicomputer.
A computer program has also been written for
transferring
data from the Sirius to the GR&DC Cyber mainframe
computer.
Preliminary work using this complete system has
indicated
that the reproducibility of the rise and fall of
ambient
levels of carbon monoxide and particulate matter
is good
(Coefficient of Variation <6%). Furthermore, the
half-
life time for carbon monoxide decay in the room
varied from
5 to 155 minutes as the air circulation rate and
percentage
fresh air varied from 30 changes per hour, 100%
fresh air
to 1 change per hour, 0% fresh air. Room
temperature
(15-30*0 and relative humidity (50-90%) had
virtually no
effect on the half-life time.
An instrument for the continuous monitoring of
ambient
nicotine, known as an Airborne Vapour Monitor
(AVM) has
been developed by the Analysis and Test Method
Development
Group (Work Area 07). This has been tested in the
environ-
mental rooms and calibrated against the
time-averaged
absorption tube method for ambient nicotine.
Preliminary
experimental results with the AVM have shown that
ambient
nicotine appears to decay more rapidly than
ambient parti-
culate matter, carbon monoxide or hydrocarbon.
co
I M British American Tobacco co, Lit This must no
be copied or shown to unauthcrised persons.
BATCo document for Province of British Columbia 11
November 1999


- 8 -
Systematic studies on the build-up and decay of
ambient
smoke components as a function of the following
environ-
mental conditions is under way:
temperature 15 - 30%
relative humidity 50 - 90%
air change rage I - 30 changes/hour
% fresh air 0 - 100%
number of cigarettes 2 - 8
In addition, the build-up and decay of the number
concen-
tration of different aerosol particle sizes in the
ambient
smoke is also being monitored for some of the room
condi-
tions, using the LAS-X Laser Aerosol Spectrometer.
Preli-
minary results have shown that, for particles
smaller than
0.25 pm diameter, the ambient particle number
concentration
starts to decrease at the end of smoking, but then
starts
to increase again about 1000 seconds after the
smoking
has finished. The number concentration of larger
particles
increases continuously for about 500 seconds after
the
smoking has finished and then gradually decreases.
(2) Enhanced Sensory
(a) Filtration mechanisms
A report describing the effects of both puff
velocity and
puff volume on the filtration efficiency of
nicotine by
highly ventilated (80%) cellulose acetate filters
has been
completed. It was shown that the dead volume
effects in
both the filter and tobacco rod led to a higher
efficiency
in highly ventilated filters than in unventilated
and mod-
erately (40%) ventilated filters. When these dead
volumes
were taken into account, good agreement between
observed
and theoretical filtration efficiency was
obtained.
Measurement of filtration efficiencies in three
cellulose
acetate filters (same tow but different packing
fractions
of 0.106, 0.097 and 0.089) using a monodisperse
aerosol
has been extended to lower flow rates than those
reported
in the last Status Review Note. Use of a flow
splittin?
device has made measurements at 0.5 and 1.05 f
min-
possible, giving a complete range of 0.5 to 6 f
min-1.
Results are in line with the smoke filtration
efficiencies
obtained for the same three filters.
A report of the aerosol contract work carried out
at
Harwell has been received, covering the
measurement of
filtration efficiencies using a polydisperse
aerosol.
Detailed calculations to compare the experimental
results
with predicted efficiencies have yet to be
-carried out.
No more work is planned in this area in the
foreseeable C>
future. 110
CO
1985 British American Tobacco Co. Ltd. This must
not be copied or shown to ummthorised persons.
BATCo document for Province of British Columbia 11
November 1999


- 9 -
Work for Project SHIP has been completed in both
the
ventilated and unventilated filtered cigarettes.
Resul ts
indicated that tobacco rods containing high levels
of stem
(100 and 50%) gave rise to higher nicotine
filtration
efficiencies in both the ventilated and
unventilated
filters. However, when ventilation was present,
the effi-
ciency in these filters was suppressed but was
still higher
than in the other cigarettes. It was suggested
that changes
in smoke pH were the cause of this higher
efficiency since
both the measurement of PMWNF filtration
efficiencies and
smoke particle size indicated that all tobacco
rods gave
rise to similar particle size distributions.
(b) Novel filters
Further studies into filters that generate a swirl
of smoke
have taken place. Some manufactured deep slot
filters from
Filtrona (Germany) have been investigated in the
labora-
tory. A swirl of smoke was obtained when
ventilation in the
range 50 to 85% was introduced over the groove. At
the high-
est level of ventilation studied, the ratio of
deliveries
at 70 and 35 cm3 was 5.6, indicating a positive
response to
changes in puff volume at the 1.5 mg delivery
level. At
higher delivery levels, the effect was not as
significant.
Suitable manufactured samples are awaited before
carrying
out any large-scale sensory testing of such
filters.
One further filter design has been patented for
producing
a swirl of smoke. The Hauni laboratory laser was
used to
put a single hole through a filter containing a
central
tube. When these filters were attached to tobacco
rods,
a swirl of smoke was observed at all levels of
filter
ventilation greater than 40%.
In an attempt to quantify swirl, a small infra-red
sensing
device has been placed into the swirl chamber.
Although it
was possible to calibrate the device down to 1 mg
of smoke
(conventional filter), no reading was obtained
with a 1 mg
swirl product. Thus the amount of smoke in the
vicinity of
the sensor was decreased but was insufficient for
detec-
tion.
The FIDES computer program has been extended to
include
the calculation of smoke deliveries and is being
tested
with sample data. Both a filter and tobacco rod
data base
have been set up for use with the program.
(3) Smoke Formation
(a) Computer modelling of burning cigarettes:
MULTICELL Model
Computer modelling studies aimed at representing
the beha-
viour of burning cigarettes has concentrated on
application
and development of the puff-by-puff delivery model
and the
MULTICELL model for complex geometries.
1995 British American Tobacco Co. Ltd. This must
not be copied or shown to unauthorized persom
c0
BATCo document for Province of British Columbia I
I November 1999


- 10 -
The MULTICELL model , which represents the
internal charac-
teristics of tobacco rods and filters as
influenced by the
geometric features of the cigarettes, has been
extended to
include predictions of smoke component
concentrations. The
concentration prediction includes the influence of
air
dilution through permeable wrappers, such as
cigarette
papers or ventilation zones, diffusion through
cigarette
and tipping papers and filtration- within the
tobacco rod
and filter. Preliminary calculations have
indicated that
the program is executing in a satisfactory manner
and the
graphical interpretation of the predictions has
been exten-
ded to include contour distribution of smoke
concentration
in the cross-section of the cigarette. Detailed
applica-
tion tests involving delivery predictions have yet
to be
undertaken.
The MULTICELL pressure/flow model, without the
inclusion
of smoke component concentration. has been used to
investi-
gate the effect of ventilation zone blocking on
the level
of ventilation in filters. These predictions are
comple-
mentary to experimental measurements carried out
on the
blocking of filters with naturally porous and
laser perfor-
ated ventilation zones. The measurements indicated
-that
the reduction in ventilation is strongly
influenced by the
ventilation level of the unblocked filter. This
behaviour
was reflected by the corresponding MULTICELL
predictions
and calculations carried out with an alternative
electrical
network representation of the filter. Typically,
with the
simple resistor model, blocking 50% of the
ventilation
zone on an 80% ventilated filter caused only a 5%
reduction
in the unblocked ventilation level. Conversely,
50% block-
ing causes a 45% reduction in the unblocked
ventilation
on a 20% ventilated filter.
For all predictions with the MULTICELL model the
resistance
to flow, within tobacco rods and filters, has been
assumed
to be isotropic. That is, the viscous and inertial
imped-
ances for flow across the cigarette in the radial
direction
are specified as equal to those for flow along the
cigar-
ette, in the axial direction. The impedances in
the
latter case are routinely determined from
pressure/flow
measurements on bound cigarettes. To test this
assumption
of isotropy, measurements have been made, in
collaboration
with Process Research: Physics of Tobacco (Work
Area 06
(B)), of the pressure/flow relationships both
along and
across tobacco beds. The impedance values obtained
from
the pressure/flow data indicate that tobacco rods
are
anisotropic. Both the viscous and inertial
Impedances
are a factor of two greater in the radial
direction than
in the axial direction. The influence of this
impedance
anisotropy has yet to be evaluated with the
MULTICELL
model.
1985 British Aaneri= Toba=o Co. LuL This mug gwt
be c*ed or sbown to =authorised persom CN
BATCo document for Province of British Columbia 11
November 1999


- 11 -
(b) Puff-b -puff model of cigarette deliveries:
DELIV9 model
Refinements to the puff-by-puff delivery model
(DELIVER)
to accommodate the influence of different
cigarette paper
additives has been completed. In addition to
agreement
with puff-by-puff experimental data on cigarettes
with
phosphate additive, agreement between the model
and exper-
imental data for citrate and chemical-free papers
is now
achieved.
In order to accurately detail the correct puff
numbers for
the wide range of cigarettes investigated with the
puff-by-
puff delivery model, adjustments had to be made to
the
static burn rate (SBR) as measured on the
cigarettes.
Implicit in the original DELIVER model was the
assumption
that, between puffs, the cigarette smouldered at
the SBR.
In all cases the SBR had to be reduced to give an
appro-
priate interpuff burn rate (18R). This reduced IBR
takes
account of the apparent delay in movement of the
burnline
immediately following the puff. A functionally
simple
expression for evaluating the IBR has been
developed and
included in the DELIVER model. This expression
includes
the influence of interpuff period which becomes
increas-
ingly important as the interpuff period is
reduced.
Typically, with an inter puff period of 28
seconds, the
IBR can be reduced to 30% of the SBR.
(c) Amalgamation of DELIVER and FIDES models
Refinements to the puff-by-puff DELIVER model have
con-
centrated on the characteristics of the tobacco
rod. No
changes have been made to the original filter
representa-
tion, which is essentially restricted to
conventional
unventilated and ventilated filters. In an attempt
to
increase the flexibility of the puff-by-puff
model, a
joint project has been initiated with the Hamburg
Physics
Group in which the DELIVER model is being
amalgamated
with the Hamburg filter design model FIDES. The
initial
feasibility study in the development of this
model, DELFI,
has now been completed. This study has indicated
that
the detailed two-dimensional flow calculation in
the
puff-by-puff model can satisfactorily be connected
to the
one-dimensional flow calculation in the filter
resistor
model.
M Combustion characteristics and biological
activity
Work has started on a range of different
circumference
cigarettes on which the association between the
cigarette
thermal characteristics and biological activity,
as defined
by tests listed in Work Area 01, is being
investigated.
In addition to routine analyses of cigarette
deliveries,
measurements have been made of: CD
110
CO
1995 British Ameri"n Toba=* OL U& 71kis must not
be copied or shown to uasudwrisW psmms.
BATCo document for Province of British Columbia 11
November 1999


- 12 -
Smoulder and puff temperatures as indicated by the
infra-red thermovision system. The peak smoulder
temperature on the coal surface reduces as the
cigar-
ette circumference increases whereas the peak puff
temperature increases.
Particle size distributions - the particle size
systematically increases as the cigarette circum-
ference Increases. Typically, for the 13 mm cir-
cumference cigarette the mass median diameter is
0.26 pm and this increases to 0.60 pm for the 29
mm
cigarette.
A wide range of other measurements is in hand,
including
puff-by-puff deliveries, internal combustion zone
tempera-
tures and detailed smoke chemistry.
(e) Project SHIP combustion studies
In an attempt to characterise the burning
behaviour of
reconstituted sheet material in comparison with
that of
stem material, experimental investigations have
been con-
ducted in three areas.
(I) Thermal behaviour
Length burnt measurements, on 100% lamina, stem
and
sheet samples and samples containing different
propor-
tions of the three materials, have indicated that
static burn rates are linearly additive functions
of the stem/sheet composition. A linearly additive
relationship was also identified for nicotine
transfer
efficiency in the puff, the nicotine transfer
effi-
ciency in the lamina (38%) being significantly
higher
than that of the sheet material (22%) and WTS
(9%).
(ii) Puff-by-puff analysis
From analysis of puff-by-puff delivery data under
standard smoking conditions, the Ellis product and
filtration parameter have been evaluated. The fil-
tration within the rod for both lamina and sheet
material for nicotine and TPM is significantly
lower
than that for the stem material. Typically for
nicotine the filtration coefficient is 0.09 cm-11
0.09 cm-1 and 0.12 cm-1 for the lamina, sheet and
stem material respectively. Furthermore, the
nicotine
product efficiency of the stem material is less
than
that for the lamina and sheet.
(III) Tobacco pyrolysis and thermal analysis
Pyrolysis of 100% samples of lamina, sheet and
stem C=)
material was carried out in argon at a linear
heating 110
rate of 1.4*C/s. These pyrolysis experiments
indicated %~O
CO
1985 British American Tobacco Co. Ltd This m=t am
be copied or shown to unaudmised persons. CO
BATCo document for Province of BritiSh Columbia 11
November 1999


- 13 -
that both the sheet and stem material
exothermically
decomposed at a temperature in the region of
310'C.
This exothermic behaviour was tentatively
associated
with the higher nitrate content of the sheet and
stem
material.
Thermal analysis of the samples by both DTG and
DSC
has indicated that the sheet' material has
character-
istics more like lamina than stem. The exothermic
behaviour of stem and sheet, seen during
pyrolysis,
was observed, but was much less pronounced in the
DSC investigation. The reasons for this are
unclear,
but could be due to the much smaller quantities of
materials used in the DSC experiments.
(f) Project SHIP aerosol formation studies
Experiments have been completed on the
characterisation
and formation temperatures of aerosol particles
from the
Project SHIP samples of stem, reconstituted sheet
and
lamina. The samples were pyrolysed at a heating
rate of
3*C s-1 in a thermal analysis system in either an
air or
nitrogen atmosphere. In general, particles are
formed in
two temperature regions: about 200 to 600*C and
about
600 to 900%.
Particle formation is, in general , enhanced in
the presence
of an inert atmosphere, the effect on lamina
material
being considerably greater than stem and
reconstituted
sheet. For the first major zone of particle
production
(i.e. between 200* and 6000C), the median particle
diameter
is smaller in air than in nitrogen. The possible
explan-
ation may be that the components which normally
re-condense
into aerosol particles after being distilled out
of the
tobacco in an inert atmosphere react with oxygen
in the
presence of air and remain in the vapour phase.
In the second major production zone (i.e. above
700%).
the LAS-X Laser Aerosol Spectrometer (with 0.09 pm
as cut-
off diameter) readings suggest enhanced generation
in the
presence of air. Median particle size, as well as
number
concentrations, increase remarkably for lamina and
sheet
material, whereas particles produced from stem are
the
least affected by the generation atmosphere. It is
probable
that the combustion products released above 700%
readily
form particles by homogeneous condensation (the
result of
high supersaturation due to rapid cooling).
Alternatively,
if insoluble solid tobacco materials are released
at these
high temperatures, they could act as condensation
nuclei
for some of the combustion products and increase
the final
particle size. For stem, however, aerosol
formation mech-
anisms for particles above 0.09 pm in diameter
stay the
same in nitrogen and air. C=)
110
U-4
CO
1985 British American Tobacco Co. LAd. This must
not be copied or shown to ~tborisad persons.
BATCo document for Province of British Columbia 11
November 1999


- 14 -
(g) Effect of tobacco heating rate on aerosol
formation
100% lamina tobacco has taen heated at four
different
heating rates (0.3, 1. 2 and 3% s-1) in both air
and
nitrogen atmospheres. As the heating rate was
increased,
the zones of tobacco weight decay and aerosol
particle
formation shifted towards higher temperatures, as
expected
from thermal analysis theory. The effect of
atmosphere on
the mass concentration of the aerosols generated
in the
200-500% zone is greatest at the highest heating
rate.
The largest median particle sizes (0.20 pn) are
also ob-
served at the highest heating rate. The observed
influences
of increasing heating rate on aerosol formation
can be
explained as follows:
(i) Faster decomposition creates higher
supersatur-
ation, enhancing the nucleation/condensation
process;
(ii) More insoluble nuclei are emitted from
tobacco
at high heating rates which enhance heterogeneous
nucleation; .
(iii) A sudden outburst of particles of different
sizes
which are a mixture of insoluble nuclei and homo-
geneously nucleated particles interact by some
sort of coagulation, resulting in larger median
diameter aerosol.
However, as coagulation should result in a
decrease in
number concentration, unlike the observations made
during
the experiments, (i) and (ii) above are more
likely to be
the dominating effects.
(h) Synthetic aerosol generation
A project is being planned in which the sensory
perception
of aerosol particles of different sizes and
characteristics
will be assessed. To facilitate this project, the
synthetic
aerosol generator developed previously to produce
a fairly
monodisperse, stable, multicomponent particle in
the size
range typical of cigarette smoke has been. further
improved.
A six jet nozzle-collision-nebuliser has been
incorporated
so that the generator can produce mass
concentrations close
to the delivery of TPM from an ultra-low delivery
(1 mg)
cigarette. The generator has been re-calibrated
using
di-(2-ethylhexyl) sebacate (DEHS). Two-component
particles
containing DEHS and amyl salicylate (a common
tobacco
flavour material) will be generated in the near
future for
the sensory perception work. Analysis of the
aerosol
phase by gas chromatography is planned.
M Secondary aerosol formation from smoke vapour
phase
A short study has been completed in which
cigarette smoke
vapour phase was held in a 1 litre sampling bag
and the
CO
N.)
IM Bdtish American Tobamo Co. LuL This must not be
copitd or shown to unsodmsed persom CD
BATCo document for Province of British Columbia 11
November 1999


- 15 -
possibJlity of particle formation from the vapour
phase as
a function of time was examined using the LAS-X
laser
aerosol spectrometer. After passing none standard
puffs
through a Cambridge filter into the bag, taking 8
minutes,
approximately 104 particles/cm3 in the smallest
size range
detectable (0.09 - 0.11 pm) were observed.
Consequently
the secondary formation of particles does occur.
To determine the formation as a function of time,
the bag
was filled with smoke vapour phase by taking a
puff every
3 seconds - 30 puffs were collected in 90 seconds.
Second-
ary particles started to appear 140-160 seconds
later.
A maximum concentration of about 2 x 104
particles/cm3 was
recorded 300 seconds after completion of vapour
collection,
when the median particle diameter was 0.23 pm.
After this
time the number concentration decreased while the
median
diameter increased up to 0.4 pm, due to
coagulation.
This secondary aerosol formation is insignificant
compared
to the total particle concentration in mainstream
smoke
(about 109 particles/cm3). Furthermore, typical
maximum
particle concentrations in the GRUC environmental
roams,
after smoking 2 to 6 cigarettes, are 2.5 x 104
particles/
cm3. Consequently, in such large volumes,
secondary
particle formation will also be negligible.
(2) Nitrosamines and Chemical Reaction Mechanisms
(a) Nitrosamine formation during processing and
curing
A small laboratory experiment has been carried out
to
measure the amounts of nitrosamines transferred to
the
atmosphere when Burley tobacco is heated to 180"C.
It was
found that 6.1 pg of tobacco-specific nitrosamines
(TSN)
were transferred under an air flow of 2 litres
min-1 from
10 q of tobacco.
Samples of Virginia tobacco that had been taken at
one day
intervals during flue-curing have been analysed
for nitros-
amine content. The total period of time for the
flue-
curing was 9 days and it was found that
nitrosamine forma-
tion did not start until after 6 days, and
measurable
amounts were not in evidence until the 8th and 9th
days.
In the second part of this work, samples of Burley
tobacco
will be analysed which have been removed during
air-curing.
(b) Nitrosamine formation in model chemical
studies
Model studies on nitrosamine formation have
continued. It
has been shown that the reaction of nornicotine
and
anatabine with nitrite proceeds rapidly, at a pH
range 2-8,
to produce N'-nitrosonornicotine (NNN) and
N'-nitroso-
anatabine (NUB). When these reactions are allowed
to CD
110
CO
1995 Bdtish Asmuican Tobacco Co. LuL This must am
be copied or sho" to mautborind peraouL
BATCo document for Province of British Columbia 11
November 1999


- 16 -
continue for an extended period of time (> 24
hours) the
maximum yield of NNN (at pH 2.5) is 65%, while I
he yield
of NAtB is quantitative. The reaction of nitrite
with
anatabine is particularly rapid (t1/2 < 1 hour).
Kinetic
calculations indicate that the nitrosation of
nornicotine
is first order with respect to nitrite. This was a
surprising result as normally (in.the
concentration range
studied) nitrosation of secondary amines is second
order
with respect to nitrite. Further work is in
progress to
verify this result. This is important as it will
provide
valuable information on the mechanisms of
nitrosation of
tobacco alkaloids.
Problems with artifactual formation of
nitrosamines, have
been encountered with the analysis when nitrite
ion is
present. Consequently, a small study was
undertaken to
ascertain the effect of added nitrite during the
initial
work-up (soxhlet extraction) of tobacco. It was
found
that nitrite had little effect on TSN levels until
in
excess of 150 p9 was added, after which the
artifactual
formation of TSN rises rapidly with further
addition of
nitrite. This result means that great care will
need to
be taken in the model studies to avoid false
positive
results, especially when investigating nitrosation
on
cellulose surfaces. Perhaps cold extraction with
dichloro-
methane will be preferred to soxhlet extraction.
(c) Reactions of carbonyls and amines
The reaction of two common smoke carbonyl
compounds, acet-
aldehyde and acetone, at temperatures up to 900*C,
has
been studied using the pyrolysis/mass spectrometer
system.
No information was available in the literature
with regard
to high temperature reactions and it was
anticipated that
this work would provide information to assist in
the inter-
pretation of some of the findings in Project SHIP.
Contrary
to expectations, it was found that neither
compound reacted
with ammonia.
(d) Effect of pectin on nicotine transfer
As a follow-up to some Project SHIP work, the
effect of
pectin on the release temperature of nicotine has
been
studied. This was done by applying a mixture of
pectin
and nicotine to cellulose rods, pyrolysing the
system and
analysing the evolved nicotine by mass
spectrometry.
When a 1:1 mixture of nicotine and pectin was
examined,
the maximum release temperature was 165*C (as for
100%
nicotine) but, additionally, there was a further
release
at higher temperatures (300 and 364*C). On
examination
of nicotine with an excess of pectin (10:1),
release of
C=
1 ~0
CO
NJ
1995 BriLish American Tobacco Co. LA& This must
not be copied or showm to mauthorised perums.
BATCo document for Province of British Columbia 11
November 1999


- 17 -
nicotine occurred only at the higher temperature
(maximum
367%) and none at all at 165*C, clearly
demonstrating
that pectin increases the temperature at which
nicotine
transfers.
Pectin contains polygalacturanic acid and hence is
an
acidic material containing carboxylic acid groups.
One
possibility is that pectin and nicotine combine to
form a
nicotine salt which releases nicotine at higher
tempera-
tures. To check this, the hydrochloride and
citrate salts
of nicotine were examined; these have been studied
prev-
iously and liberate nicotine at conventional
temperatures
(180 - 200*C). It was found, for both salts, that
the
release temperature of nicotine was unaffected by
mixing
with pectin. This would suggest that only free
base
nicotine can interact with pectin to increase the
release
temperature.
To test whether the pectin effect is a function of
molecu-
lar weight, nicotine salts of five fatty acids
(Clo - C18)
were prepared. No increase In nicotine release
temperature
was found on pyrolysis of these salts on
cellulose,
confirming that the pectin effect is not related
merely to
the high molecular weight of pectin.
A pectin-nicotine mixture (10:1) was heated in an
aluminium
dish instead of on paper. Surprisingly, nicotine
was
released at the lower temperature. This would
indicate
there is some interaction between nicotine,
pectinate and
cellulose.
These findings have also been confirmed using pure
poly-
galacturanic acid, three other commercially
available pec-
tins and de-nicotinated tobacco instead of
cellulose.
In order to ascertain the effect of pectin on
mainstream
and sidestream deliveries, de-nicotinated tobacco
rods were
loaded with nicotine and with a 10:1
pectin-nicotine mix-
ture (so that the amount of added nicotine was
identical).
It was found that mainstream and sidestream
deliveries of
nicotine from these cigarettes were similar. Thus,
although
pectin can dramatically increase the temperature
at which
nicotine is released, it appears to have no effect
on main-
stream and sidestream deliveries of nicotine.
FUTURE WORK:
(1) Sidestream Smoke
The effect of using reconstituted tobacco sheet
material which
has been treated with an odour modifier (supplied
by De Mauduit)
on sidestream emissions and subjective assessment
will be
examined.
19115 British American ToWcco Co. Ltd. This must
a" be copied or shown to unautbori persons.
L.^J
c0
NO
U_J
BATCo document for Province of British Columbia 11
November 1999


- 18 -
Samples of fruit extracts have been obtained from
Brazil
and their effect on improving or masking butt
aroma will be
assessed.
A systematic study on the effect of design
parameters of conven-
tional cigarettes (diameter, packing density, cuts
per inch,
paper permeability) on sidestream generation has
been largely
completed experimentally and the results will be
assessed.
The design potential of co-axial cigarettes for
reducing side-
stream emissions will be assessed in conjunction
with Novel
Cigarette Technology (Work Area 06 (C)).
Systematic work on the build-up and decay of
ambient smoke
components as a function of environmental
conditions will con-
tinue, including monitoring aerosol number
concentrations and
size variation. Techniques for the collection of
ambient smoke
for subsequent nitrosamine analysis will also be
developed.
A consultant has been appointed who will examine
the role of
building design and ventilation in
reducing/eliminating the
social problems associated with passive smoking.
(2) Enhanced Sensory Filters
At present no more practical work is planned in
the filtration
mechanism area. However, the effect of flow
velocity on filtra-
tion efficiency will be measured for specific
filters in order
to obtain filtration parameters for use in the
FIDES computer
model.
Further subjective and objective testing of tube
and groove
filters will be carried out when suitably
manufactured samples
become available.
In order to assess further the physical
characteristics of
smoke swirl and the distribution of smoke within
the mouth,
work has commenced an determination of smoke
particle size and
carbon monoxide concentration within the
artificial mouth. It
is intended to extend these measurements to other
vapour phase
constituents and physical parameters.
An assessment of the FIDES computer program using
test data for
tobacco rods and filters will take place.
Determination of
parameters for the tobacco rod and filter will be
made based on
comnercially available and sample cigarettes to
build up the
data base available to FIDES.
A range of Filtrona filters has been used to
manufacture cigar-
ettes to investigate the properties of such
filters. Both
ventilated and unventilated filter cigarettes have
been prepared
and smoke analysis will be undertaken.
CO
M-)
MS B3ritish Americ- Tobacm Co. Ltd. ThLs must not
be copied or shmm to maudworised perso- 4~-
BATCo document for Province of British Columbia 11
November 1999


- 19 -
(3) Smoke Formation
Computer modelling work will concentrate on
application of the
MULTICELL model and development of the
puff-by-puff delivery
model in collaboration with Hamburg. The former
will include
an assessment of the effects of impedance
anistropy and the
representation of species concentration. Both
these topics
will be supported by appropriate experimental
work. The next
stage of the puff-by-puff project will involve
restructuring
the DELIVER model to accommodate the FORTRAN
version of the
filter model FIDES.
In the combustion/biological activity
investigations, measure-
ments will be carried out on internal coal
temperatures, coal
impedances and smoke chemistry using the mass
spectrometer.
The mass spectrometer will also be used to further
investigate
the mechanism of reduced CO production on shredded
stem samples.
The work is in support of Process Technology (Work
Area 06 (A)).
A second synthetic aerosol generator will be built
and cali-
brated, to be used in the sensory assessment of
particles of
different sizes and characteristics.
Initial probe studies will be undertaken to
determine particle
size characteristics as the smoke travels through
the cigarette.
(4) Nitrosamines and Chemical Formation Mechanisms
Tobacco samples obtained from different stages
during the curing
of Burley tobacco will be analysed for nitrosamine
content.
Work on nitrosamine model studies will continue.
The effect of
potential catalysts and inhibitors of nitrosamine
formation
will be investigated. The results will then be
used to attempt
the control of the nitrosamine content of
tobaccos.
Further samples of tobacco will be treated in the
Tobacco
Processing Cylinder and nitrosamine levels
measured.
Twenty-four cigarette brands from different
companies will be
analysed for nitrosamines in tobacco, mainstream
and side-
stream smoke. It is anticipated that this will
shed some
light on the various processing conditions and
tobacco types
used by the different companies.
C:>
1 ~0
NO
-1.4
L^J
c0
1985 British American Tobacco Co. Ltd. This must
no be copicd or shown to unandmised persona. N.)
BATCo document for Province of BritiSh Columbia 11
November 1999


- 20 -
STATUS REVIEW NOTE
Work Area: 03 Smoker Behaviour
Author: G.A. Read
Period Covered: November 1984 - April 1985
OBJECTIVE:
To provide the means to identify and characterise
the fundamen-
tal nature of the smoking process. The research
will provide
the opportunity to develop products with improved
quality and
satisfaction across the tar delivery range.
AREAS OF RESEARCH:
Experimental studies in five broad areas are in
progress.
(1) Smoking Behaviour - Technological Developments
A series of technological developments has been
achieved that
enables laboratory-based consumer recording and
duplication to
be conducted, where appropriate, on a group-wide
basis. In
addition, a facility is available for estimating
puff deliveries
on a real time basis, considerably increasing the
opportunity
for subjective and objective assessment in
relation to product
development. Research is in progress to facilitate
the trans-
ition between laboratory and remote monitoring of
smoking
behaviour.
(2) Smoking Behaviour - Consumer and Psychological
Studies
A range of consumer studies has been conducted
during this
period. Projects have been designed to identify
the relation-
ship between smoking style and consumer behaviour.
product
design and consumer response and the physical
characteristics
of the product and smoker behaviour. A study has
been designed
and implemented to examine the relationship
between smoker
behaviour and electrocortical activity.
(3) Respiratory Research
A series of studies is in progress to examine and
characterise
both gas and particle behaviour within the
respiratory system
in an attempt to provide a fundamental basis for
determining
smoke aerosol interaction in the lung. This
research will
provide a basis for modifying the smoke aerosol to
maximise
smoke sensations from a minimum of smoke dose.
co
10 1995 British Amerk= Tobacm Co. La This mum am
b. cWW or sho" to unautharised perwas. NJ
BATCo document for Province of BritiSh Columbia 11
November 1999


- 21 -
(4) Aerosol Retention Studies
An extensive range of studies has been conducted
to identify
both the physiological and cigarette design
parameters that
influence particulate retention. Studies are
continuing that
attempt to identify and further characterise the
mechanisms
involved in particulate retention to provide the
means to spec-
ifically modify lung retention. Research to
quantify ambient
smoke retention using radio-active techniques is
continuing to
make good progress.
(5) Pharmacological Studies
The experimental work to characterise nicotine
pharmacokinetics
and dynamics in man has been completed. Studies to
evaluate the
pharmacological properties of eugenol have been
completed and
reported. Considerable progress has been achieved
in the devel-
opment of techniques to evaluate the relationship
between saliv-
ation and the dry mouth property of certain
products. The
nicotine receptor studies have been extended to
characterise
the functional significance of previously reported
changes in
the kinetic properties of the receptor following
nicotine
exposure.
PROGRESS:
(1) Smoking Behaviour - Technological Developments
Arising from the evaluation of the~previously
reported in-house
constructed human smoking behaviour recording and
duplication
equipment, the system has been modified to allow
duplication to
be driven from the pressure signal generated
during puffing.
This additional facility enables the potential for
the orifice-
type cigarette holder to be replaced by a more
conventional
holder system as well as allowing inter-product
duplication
based on smoker effort rather than on volume
considerations.
A reflectance device has been constructed which is
built in to
the puff recording system and is undergoing
evaluation. This
facility allows an estimate of smoke deliveries to
be determined
in real time. Following a recent publication from
the Oak Ridge
National Laboratory, in which the reflectance
device was used
to estimate smoke deliveries, the GR&DC device has
been modified
in terms of its software to allow puff flow and
reflectance to
be compared on a continuous basis. This
modification will allow
both per cigarette and puff-by-puff delivery
estimates of parti-
culate delivery to be achieved with even greater
precision.
The combined device, the portable puff recordi ng
and reflectance
system, will be an invaluable tool for product
development and
assessment, providing a real time estimate of
puffing behaviour,
dose estimation and subjective assessment. Studies
of this type
will enable the relationships between smoker
effort and reward
in terms of puff volume, puff flow, smoke dose and
qualitative
U_J
c0
NJ
1985 Brftish Amedem Tobewo C& EA& TMs mum imt be
cop6d or shin to -thoriud penom
BATCo document for Province of British Columbia 11
November 1999


- 22 -
assessment to be elucidated. These factors, and
their inter-
relationship in terms of influencing product
acceptability
and satisfaction, are still poorly understood, yet
are essential
in product development.
A study involving consumers, in which their
smoking behaviour
was recorded using both a conventional
orifice-type holder and
video techniques, has been completed. Within this
study, an
attempt is being made to test the hypothesis that
puff duration
is related to puff volume. If this can be
validated on an
individual smoker basis, video techniques may
provide a viable
alternative to conventional smoking behaviour
monitoring and
allow subjects to be monitored remotely.
(2) Smoking Behaviour - Consumer and Psychological
Studies
A collaborative study between Smoker Behaviour and
Psychol-
ogy groups has been conducted in which subjects
from Project
NARINERS have been monitored in terms of their
smoking behaviour
whilst smoking and subjectively assessing a range
of cigarettes.
All consumers had been assessed in terms of their
motivational
and attitudinal behaviour to cigarette smoking
over an extended
time course. Subjects were selected for the study
in terms of
their smoking histories, the groups consisting of
smokers in
different tar categories that were either stable
or unstable
in terms of that tar group. The major objective of
this exercise
is to identify smoking behaviour parameters that
are associated
with these defined groups whilst smoking own brand
and switching
to alternative products both within and outside
their tar
category. The study will be reported during the
next review
period.
The first phase of a smoker behaviour consumer
study, designed
to test smokers' reaction to a UK Virginia
ultra-slim product
(17 mm circumference, containing Du Maurier
blend), has been
completed. Subjects were compared in terms of
smoking behaviour
and product assessment whilst smoking their own
brand, a
conventional Du Maurier product (low-middle tar)
and the ultra-
slim product (low-middle tar). The ultra-slim
product differs
from the conventional Du Maurier product in having
a consider-
ably higher pressure-drop (approximately x2). In
addition to the
direct product comparison, this study is allowing
a comparison
of smoking style, draw resistance and
effort/reward relation-
ships to be further elucidated in combination with
subjective
assessment. A second phase of this study is in
progress to
assess the response of these consumers to a
product of equiv-
alent delivery to the Du Maurier product but with
considerably
reduced pressure drop characteristics.
A detailed analysis of the findings obtained in a
collaborative
study between GRADC and a TAC funded project with
Dr. N. Pride
has been completed and reported. Although the
study was prim-
arily concerned with investigating the
relationship between .110
puffing and inhalation pattern and the development
of respira- 110
tory disease, an interesting observation arose
from this study. _1J
L.P4
CO
N)
CO
1985 British Americam Tobacco Co. L& This mum not
be copW or sho%~ to unautborised persom
BATCO document for Province of BritiSh Columbia 11
November 1999


- 23 -
These consumers had an average age of
approximately 55 years
with long frequently life-time, smoking histories,
yet the
puffing af,d inhalation patterns are comparable
with other
consumer studies conducted with considerably
younger popula-
tions. These findings indicate that smoking style
is consis-
tent over time.
A series of psychological and behavi6ural studies
has been
initiated with Professor A. Gale, University of
Southampton.
The first study is in progress and has been
designed to invest-
igate the relationship between electro-cortical
activity and the
objective measures of smoking behaviour. This
study is an essen-
tial pre-requisite to subsequent joint studies
with GROC and
BAT Hamburg, planned to identify the quantitative
significance
and interaction between nicotine, product and
psychological
state.
In addition. this research is specifically aimed
at identifying
the physiological and psychological factors that
initiate or
prompt a smoking event.
(3) Respiratory Research
The work reported in this section primarily
involves external
research activity with Dr. T. Higenbottam of
Addenbrooke's
Hospital, Cambridge, and covers three main areas.
A series of studies has been initiated to
characterise the
behaviour of gases within the respiratory system
as an integral
part of a research programme aimed at defining the
detailed
functioning of the lung. These studies will
substantially ad-
vance our understanding of the lung in terms of
smoke component
distribution, deposition and uptake. The studies
are designed
to characterise the lung in terms of volume of
distribution for
smoke components, the lung surface fluid volume
available to
soluable smoke gases (and particulate components)
and the nature
of blood flow through the lung and hence rate and
site of
component removal. These studies are particularly
relevant for
evaluating the significance and site at which
smoke is capable
of eliciting sensory effects in the lung.
There is considerable interest currently from
medical and phar-
maceutical groups in terms of characterising the
distribution,
deposition and clearance of aerosols within the
lung. It has
been suggested that cigarette smoke particulate
can be cleared
via three possible routes - by macrophage
scavenging followed
by either clearance on the mucociliary escalator
or by resorp-
tion of the macrophage into the lung walls and in
changes in
alveolar wall permeability. Studies are in
progress that use
suitable radiolabelled particles of differing
sizes to specifi-
cally "plot" the pathways of particle clearance.
These studies
will provide a basis for evaluating the effect of
cigarette
smoke on lung clearance mechanisms and the fate of
smoke part- C:)
icles within the human respiratory system.
LON
CO
r1 J
1995 Britih Americ- Tobsoco Co. Ltd. This must not
be copied at sho"m to unsuduxised persom 110
BATCo document for Province of British Columbia 11
November 1999


- 24 -
The final area of study is concerned with specific
radio-
labelling of cigarette smoke with a short-lived
but highly
active isotope for ise in characterising smoke
particulate
distribution, depostion and clearance.
(4) Aerosol and Retention Studies
A series of studies has been completed that
examines the effect
of respiratory pattern, smoke dose, tobacco blend,
the phsyical
properties of the smoke aerosol and other
cigarette design
criteria on particulate retention in the lung. A
report has
been prepared that documents the detailed
findings. From the
findings it can be concluded that smoke dose has
little or no
effect on the percentage of particulate material
retained by
the lung, whilst depth inhalation and repiratory
hold does
significantly increase retention. Examination of
retention
during smoking of UK Virginia, Oriental, ai
r-cured and fermented
air-cured products revealed that the fermented
air-cured product
had the lowest retention characteristics. In
addition, it was
also observed that this product had the lowest
particle size
characteristics of the products investigated.
In studies in which humectants were added to
standard tobacco
rods, the level of particulate retention was
substantially
increased. The increase in retention was accounted
for in part
by the quantitative uptake of humectant from the
smoke and also
by increased retention of the residual smoke
presumed to arise
from increased deposition. These findings are
consistent with
at least two mechanisms influencing retention -
particle stripp-
ing of volatile and semi-volatile material and
particle deposi-
tion probably arising from the hydroscopic nature
of smoke.
The fundamental nature of these mechanisms and
their influence
on particulate retention is being further
elucidated in studies
characterising particulate retention based on
particle numbers
rather than mass. In addition, the retention of
particulate
smoke following re-breathing of the smoke aerosol
is providing
an insight into the significance of specific
components on
particle behaviour and component retention within
the lung.
The external study being conducted at AERE Harwell
to char-
acterise the retention of ambient smoke by
non-smokers is
progressing well. Following the successful
development and
characterisation of a sidestream marker
(radiolabelled iodohexa-
decane, report in progress) and the construction
of a suitable
environmental room for smoke exposure, a
preliminary study
involving human volunteers was initiated and is in
progress.
The purpose of this study is to determine the
exposure condi-
tions and the level of radiolabelling necessary
for the main
study. These studies will enable an estimate of
ambient parti-
culate smoke deposition to be determined from a
knowledge of
smoke particulate concentration and duration of
exposure in
natural environments.
CD
ILA
cc
1985 Bitih A..A. T.I.- Co. Ltd. Thi. must im be
copied or x1bow to unmtbarisod persouL LP4
BATCo document for Province of BritiSh Columbia 11
November 1999


- 25 -
(5) Pharmacological Studies
The experimental work relating ,o the
pharmacokinetic studies
with Professor C. George, University of
Southampton, has been
completed. The study has investigated the effect
of administer-
ing equivalent nicotine doses via four different
routes - from
cigarette smoking, continuous infusion, pulsed
infusion and from
an aqueous nicotine aerosol. The physiological
measures taken
during this experiment are currently under
evaluation and a
report is in progress. The biological samples
continue to
be analysed for nicotine and cotinine and hence
the pharmaco-
dynamics and kinetics will be reported later.
A series of studies has been conducted to monitor
the rate of
salivation in relation to smoke exposure. A
technique has been
developed and evaluated as a means of collecting
saliva and
assessing salivary flow during smoking. These
procedures are
being used to collect "uncontaminated" saliva as a
means of
obtaining a biological fluid sample essentially
non-invasively.
The technique is also being used to investigate
the relationship
between salivation and subjective assessment of
products with
reported dry mouth characteristics. A report is in
preparation
that documents-the progress in this area.
A series of internal and external studies has been
completed
that investigated the potential pharmacological
properties of
eugenol. The observed findings indicate that in
the systems
investigated, eugenol has no gross pharmacological
properties
at typical doses calculated from human smoke
doses. A report
documenting the detailed findings has been written
and will be
issued shortly.
Following the previously reported observation of
changes in
nicotine receptor numbers following smoke and
nicotine exposure.
additional studies have been conducted to
establish the funct-
ional significance of this change. This has been
achieved by
developing an experimental technique that can be
used to monitor
receptor response through the release of
(pre-loaded) radio-
labelled dopamine. Using this technique, it has
been possible
to demonstrate significant modifications to the
receptor follow-
ing smoke and nicotine treatment. These findings
may be related
to the development of nicotine tolerance, i.e. a
reduced res-
ponse to nicotine at equivalent. concentrations.
These studies
will be reported during the next review period.
FUTURE STUDIES:
It is considered to be of primary importance to
continue to
develop, evaluate and implement the means to
assess smoking
behaviour both remotely and ultimately
surreptitiously to deter-
mine how products are smoked under naturalistic
conditions.
Considerable research effort is being directed
towards achieving
these objectives.
CX:)
1985 British American Tobacco Co. EAd. This most
am be copied or shown to unauthwind person& L^J
BATCo document for Province of British Columbia 11
November 1999


- 26 -
Following the development and availability of
relatively simple
and rapid techniques for assessing smol?
deliveries, the
detailed relationship between smoking style,
product delivery
and subjective assessment will be examined
extensively to
clarify our understanding in this area of product
assessment.
The identification of the parameters that
influence smoke
retention and, more importantly, that 'can
specifically modify
particulate deposition within the lung, are a
major objective
and considerable research effort is being directed
to elucidate
these parameters.
The salivation technique will be used during the
next review
period to characterise the dry mouth properties
associated with
specific products. In addition, research is in
progress to
investigate the fundamental interaction of smoke
and components
within the mouth in terms of salivary flow and
smoke aftertaste
assessment.
CO
1985 British American Tobacco Co. UA This must not
be copied or shown to unauthori--d persons. t~-J
NJ
BATCo document for Province of British Columbia 11
November 1999


WORK AREA 04
TOBACCO TREATMENT
CD
110
%10
-,j
U-4
co
0 1995 British American Tobacco Co. LJA This -t va
be copied or sho-n to unauOorised Persom LN
L.-j
BATCo document for Province of British Columbia 11
November 1999


- 27 -
STATUS REVIEW NOTE
Work Area: 04 Tobacco Treatment
Author: R.A. Crellin
Period Covered: November 1984 - April 1985
OBJECTIVE:
To identify, investigate and implement the
technology of smoke
sensory enhancement/modi f i cation through a
programme involving
preferred chemical changes induced through
processing of
tobacco.
AREAS OF RESEARCH:
The work area divides into two major areas of
activity:
(1) Flavour Generation during Processing
The chemical nature and influence on sensory
properties of
reactions which occur or can be induced during
tobacco pro-
cessing is investigated. In particular, amino
acid/sugar
reactions and reactions generating
sulphur-containing flavours
are studied. Novel and/or modified existing
processes are
then examined to determine how these reactions may
be enabled
and optimised.
(2) Analysis and Mapping of Products and Processes
Spectroscopic, chromatographic and chemical
techniques which
aid the identification and interpretation of
tobacco and smoke
components relating to sensory properties are
developed. The
techniques are used either
(a) in analytical mode via subfractionation,
enabling the
identification of components induced by process or
intro-
duced by competition
or
(b) in mapping mode via statistical techniques, so
that the
overall effects of blends, processes and additives
on
sensory properties can be depicted.
Both approaches are applied to brands and
processes from a
number of markets worldwide.
c0
1985 Bri6sh American Tobacco Co. Lt& This must not
be copied or shown to unsutho~ persom LAJ
4 ~h-
13 ATCo document for Province of British Columbia
11 November 1999


- 28 -
PROGRESS:
(1) Flavour Generation during Processing
(a) Novel Tobacco Processing
The Tobacco Processing Cylinder (TPC) has been
used to
evaluate novel cased Burley redrying treatments. A
Burley
sub-blend ex-B&W has been redried from inlet
moistures of
36% and 28% to exit moistures of 6% and 3% via
three
different treatment cycles. Thus, overall, twelve
combina-
tions have been carried out. The three treatments
involved
(i) hot air drying,
(ii) air/steam drying,
(iii) hot air drying, preceded by a "stewing"
cycle
during which tobacco temperature was raised but
without drying.
The optimisation of the redrying step is now a
major goal
for B&W. The treated tobaccos- are being
subj-ected to
very detailed chemical analyses, including pectin
analysis
in Hamburg. Partly to meet the analytical needs of
the
project, a new method for simultaneous and rapid
quantifi-
cation of theabromine and glycyrrhizin marker
components
for cocoa and liquorice respectively has been
developed.
The Processing Research programme carried out by
Souza
Cruz has resulted in the devising of a process
scheme for
building pyrazine precursors into tobacco. The
scheme
involves post-GLT adjustment of tobacco pH and TPC
treat-
ment subsequent to ageing. Investigation of this
approach
has been discussed with ITL (Canada) and is being
implemen-
ted for two grades available in large quantity,
with the
aim of quality improvement so that greater
quantities could
be successfully utilised. The tobaccos are being
additive-
treated in GR&DC and returned to Canada for
ageing.
In a quite separate ITL project, nine different
GLT drying
modes are being investigated involving different
humidities
and tobacco bed depths/apron speeds. The study has
been
completed for three different grades. GR&DC is
conducting
a full tobacco analysis of all samples, plus smoke
analysis
of cigarettes made in Canada. Sensory testing of
these
samples will also be conducted, although
preference testing
will rightly be carried out by ITL.
TPC processing of flue-cured and air-cured lamina
has
shown that tobacco-specific nitrosamines (TSN)
decrease
with increasing process severity for flue-cured
lamina,
but increase for air-cured lamina.
CD
110
c0
1985 British American Tobacco Co. Ud. This must
not be copied or sbown to unauthorised persons.
(__J
r.-M
BATCO document for Province of British Columbia 11
November 1999


- 29 -
Improvements to the subjective smoke character of
stem is
a major aim. Burley and flue-cured CRS have been
treated
under a variety of sets of conditions in the TPC.
Cigar-
ettes are being made incorporating these
TPC-treated stems
for chemical and sensory study. TPC-treated
flue-cured
lamina is being similarly evaluated. As an adjunct
to
the CRS work, TPC-treated CRS and shredded stem
were
separately classified using the Hambro Fluid Bed
Dryer.
The small size fraction of shredded stem contains
three
times the lignin level of the normal size
fraction.
Normal and small size fractions of CRS show no
difference.
This is in accord with the view that stem
shredding causes
shear along lines of weakness parallel to
cellulose fibres
and may have implications for the carbon monoxide
reduction
work reported elsewhere.
(b) Sulphur Compounds
A joint project with the Flavour House PFW has
begun with
the aim of assessing the importance of sulphur
chemistry In
tobacco and thereby ultimately the formulation of
tobacco-
compatible flavours based on relevant sulphur
compounds.
GROC provide the tobacco-related expertise
including the
choice of tobacco types and qualities; PFW provide
sulphur-
related expertise in the areas of analytical
chemistry and
sensory evaluation. Burley tobacco has been
extracted
using the Schultz technique and the extract
subfractionated
by HPLC into "cuts" of differing sensory
character. PFW
now have these samples for analysis.
(2) Analysis and Mapping of Products and Processes
(a) Market Mapping
Project SHIP (SAW) was completed before the
year-end and
8 reports are now in preparation. providing a full
account
of the many conclusions drawn from this major
study.
Some of the main technical conclusions are as
follows:
(I) Fourier Transform Infra-Red (FTIR) and X-Ray
Fluorescence (XRF) mapping of brands and experim-
ental products demonstrates that a close approach
to the Philip Morris region is achieved by two
key product development samples, Amelia and
Stetson Lights.
(ii) Transport of stem from the WTS line to the
RTS
line results in higher nicotine transfer effic-
iency (NTE) to mainstream.
(III) Scavenging of lamina nicotine by RTS pectin
does
not increase NTE to mainstream but does increase
free-base nicotine levels in smoke; this may be
1985 British American Tobacco Co. La Tlas must act
be copied or sho" to unauthorised pasces.
CO
L.1-4
ON
BATCo document for Province of BritiSh Columbia 11
November 1999


- 30 -
a consequence to the substantially higher nico-
tine release temperature from nicotine/pectin
complexes.
(iv) Deliveries of ammonia and nicotine are a
sensi-
tive function of sugar levels, whether introduced
by casing or by Virginia tobaccos.
(v) Samsun and UAB Oriental tobaccos contain the
highest levels of tobacco labdanoids (believed
part-responsible for Oriental character); their
use causes irritation masking.
(vi) Preferred reaction flavour systems contain
high
concentrations of aminated glucose; these
systems smooth the smoke and enhance smoke body.
A further exercise, termed SHIP II, is nearly
complete.
This comprises an evaluation of 'new-style" B&W
leaf
purchases (relevant to Amelia/Stetson) with
"old-style"
purchases in terms of leaf/smoke chemistry and
product
mapping. The new-style grades provide elevated
smoke free-
base nicotine levels and the new style Burley
contains
substantially increased levels of tobacco
cembranoids
(believed part- responsi bl e for tobacco flavour
character).
A Virginia SHIP exercise, termed SHIP III, has
been agreed
wi th Wi 11 s, Austral ia. Caref ul considerati on
of the appl i -
cability of the SHIP lessons to a Virginia market,
coupled
with awareness of the constraints and
opportunities offered
by the Australian market, led to the
identification of
the following three project areas:
(i) Market mapping - by linking together consumer
data and O/WP, FTIR and XRF data through chemo-
metric (chemistry plus statistics) methods.
(ii) Competitor product matching - by computer-
blending domestic and off-shore grades to the
FTIR/XRF profiles of key competitor brands.
(iii) Stem additive evaluation - by examination of
the
shifts in mapping position of blends containing
WTS or shredded stem treated with ammonium salts
(e.g. diammonium phosphate) and lamina waste
solubles.
The major Virginia and Blended brands in the
Malaysian
market are presently being mapped (SHIP IV) by
FTIR and
XRF. Subsequent phases of the Malaysian project
will
involve the mapping of lamina grades and -other
blend
components by FTIR and XRF plus comparison of the
spectroscopic data with sensory data provided by
MTC.
00
1985 Bridsh Amedcan Tobacco Co. UA This must not
be copied or sb- to -horised pe-a.
BATCo document for Province of British Columbia 11
November 1999


- 31 -
BAT Germany have recently presented 32 masked and
coded
domestic brands in white packs to large numbers of
consu-
mers. Smokers are asked to rate the masked
products on an
array of sensory parameters. GRADC will gather
FTIR and
XRF data on all the brands (SHIP V) so that
sensory and
spectroscopic links may be forged.
Cocoa hulls have been approved for use in West
Germany
following successful application by Philip Morris.
Hulls
have a similar chemistry to tobacco blend
components,
rendering their introduction into Marlboro
difficult to
prove/disprove. FTIR will be used to see if the
technique
can index cocoa hull presence and thus determine
if and
when the material is introduced into German
Marlboro.
(b) Analytical and Mapping Techniques
The fleet of SHIP projects has necessitated
further work
in the streamlining of statistical analysis
through arti-
ficial intelligence techniques. This has been
progressed
through the availability of sensory and NIR data
collated
by B&W for a large number of products on the USA
market.
From this data, a pathway to the optimisation of
product
design has been proposed and reported. Firstly,
sensory
and NIR maps are 'locked" together to produce the
same
map, allowing assessment of blends. Secondly,
sensory and
cigarette design data maps based upon the
optimised blend
are also locked, allowing assessment of cigarette
design.
With the availability of sensory cigarette design,
FTIR
and XRF from project SHIP, the approach has been
shown to
be a powerful route to product development and
will be
exploited in SHIPs III, IV and V.
A puff-by-puff whole smoke FTIR scan has been
developed.
The whole smoke spectrum is highly discriminating
between
Virginia and Blended products and, within the
US-Blend
type, spectral differences between manufacturers
have been
demonstrated. As in the case of tobacco
characterisation,
the method is low resolution and hence can be
performed on
a low cost spectrometer.
A variety of tobacco and cigarette smoke
extraction
procedures have been developed. The principal
objectives
have been
(i) the discovery of natural tobacco aroma and
flavour components;
(ii) the ability to analyse competitor
developments.
110
110
__~J
L__4
CO
1995 British American Tobacco Co. LA This must am
be copied or shown to unwtharised persons.
CO
BATCo document for Province of British Columbia 11
November 1999


- 32 -
The Schultz technique has now been evaluated; this
allows
larger quantities of tobacco to be extracted than
does the
Sandra technique and can be operated at reduced
pressure
if particularly mild conditions are required. The
larger
Schultz extracts have been fractionated by HPLC
and the
sub-fractions with interesting organoleptic
properties
analysed by gas chromatography (GC) and mass
spectrometry
(MS). Experience of the combined
Schultz/HPLC/GC/MS system
indicates that information relating to competitor
additives
is forthcoming. This claim is made on the basis of
results
obtained from wet snuff products; with these
techniques,
the time cannot be far off when routine analysis
of
competitor products for trace additives, such as
those in
top dressing flavours, is possible.
Microencapsulated menthol supplied by PFW results
in a
significant stabilisation of menthol delivery to
smoke;
over a ten week ageing period, delivery is
constant.
Conventionally mentholated cigarettes show a
considerable
menthol delivery reduction over the same ageing
period due
to menthol migration to the filter and the
packaging.
Flavour release agents have also been examined.
One of
these is designed to release tetramethylpyrazine
(TMP) and
benzaldehyde into smoke on pyrolysis. The compound
works
as intended, providing a 12% transfer of TMP to
mainstream
smoke. Again, this value will be stable over long
ageing
periods; conventionally-applied TMP transfer
values
dwindle to 1% on ageing.
Such developments can also provide puff-by-puff
control of
flavour delivery. The relevance of this has been
demonstra-
ted by the determination of puff-by-puff
deliveries of
conventionally applied flavours. Volatile flayours
follow
the puff-by-puff pattern exhibited by TPM, whilst
less
volatile flavours. have profiles which rise
rapidly, flatten
during middle puffs and decrease in delivery at
the end of
the cigarette.
Several flavours have been evaluated for chemical
integrity
on transfer to smoke. Radiolabelled anisole,
isoamyl
isovalerate, benzaldehyde, vanillin, anisaldehyde
and
methyl cinnamate were selected by BAW. All
flavours except
anisaldehyde transfer intact to mainstream smoke.
In the
case of anisaldehyde 10% of the added material
transferring
to the mainstream degraded to an unidentified
decomposition
product. In the sidestream, decomposition was
detected
for methyl cinnamate, vanillin and anisaldehyde. '
The
tentative implication is that less volatile
flavourants
are more likely to suffer degradation.
co
LA
198S Bridsh American ToW=0 QL Ltd. TI& must am be
copied or shown to unautborised persom 110
BATCO document for Province of British Columbia 11
November 1999


- 33 -
FLr-URE WORK:
(I- Flavour Generation during Processing
(a) Novel Tobacco Processing
The TPC-treated cased Burley sub-blends have been
returned
to B&W for manufacture into US-Blended cigarettes
and
sensory testing by B&W. Samples of these
cigarettes will be
supplied to GR&DC for analyses similar to those
conducted
in Project SHIP. As well as analysis of potential
irritants
this could also include Sandra smoke scan and
smoke FTIR.
Mapping by tobacco spectroscopy, i.e. FTIR and
XRF, would
logically be conducted on the existing Burley
sub-blends.
The GLT drying mode experiment conducted by ITL
(Canada)
will also provide a consignment of cool air-dried
*green'
lamina for TPC drying studies In GR&DC, to
supplement the
drying modes evaluated by Canada. Subsequent to
this work,
the effects of ageing will be explored by
re-analysis of
the same tobaccos after twelve months' storage in
the
Aylmer warehouse.
A third area of process collaboration with ITL is
concerned
with low quality grade improvement through
sugarlamino
sugar/ amino acid modification to match high
quality
grades. Initially this will be tackled by
gathering data
on high and low quality and high and low sugar
Canadian
grades at various stages between GLT and
manufacture.
Methods are under development within GR&DC to
improve
ability to determine aminosugar compounds. In this
way, a
clear picture of the amino acid/sugar
deficiency/imbalance
of low quality grades will be built up, allowing
the iden-
tification of potential post-GLT correction
procedures.
The SHIP III (Wills, Australia) stem modification
project
(treatment with ammonia salts and lamina waste
extracts)
will commence also in the near future.
(b) Sulphur Compounds
In order to follow the progress of the PFW
collaborative
project "in-house", a flame photometric detector
will be
installed, assessed and optimised. It may prove
possible
to use this system to evaluate sulphur component
profiles
of the TPC processed Burley sub-blends.
(2) Analysis and Mapping of Products and Processes
(a) Market Mapping
The map locking technique already described for
linking
sensory and spectroscopic data together - thus
providing
cc
61
.985 British American Tobacco Co. Ltd. This must
am be copied or shown to unauthorised
CD
BATCo document for Province of British Columbia 11
November 1999


WORK AREA 05
FLAVOUR AND BIOTECHNOLOGY
co
X%
0 1915 British American Tobacco Co. U& This man
not be copied or sho-n to amwthorised persons.
BATCo document for Province of British Columbia 11
November 1999


- 34 -
a firm basis for sensory description and
prediction - will
be applied to SHTP III (Australia), SHIP IV
(Malaysia) and
SHIP V (West Germany) .
The tobacco FTIR and XRF techniques will also be
used to
study the effects of DIET on single flue-cured
grades for
BAT (UK&E), and map the shifts induced.
The next stage of the whole smoke FTIR scan
development
will determine whether the FTIR-based technique
has as much
discriminating power as O/WP using a selection of
Austral-
ian brands from SHIP III. If FTIR is as
discriminating as
O/WP, then O/WP will be redundant for product
development
with a far more powerful tool taking its place,
advantages
being
(i) puff-by-puff information;
(ii) speed of analysis.
If this is successful, the technique can be
incorporated
into the map locking method for optimisation of
product
design through chemometrics.
(b) Analytical and Mapping Techniques
The SHIP activity has led to the purchase of a
second
FTIR spectrometer of low resolution for tobacco
character-
isation. This will free the existing high
resolution
spectrometer for more fundamental studies,
particularly
leaf scanning and leaf depth profiling, techniques
of
potential interest and assistance to blenders.
Further exploitation of the Schultz/HPLC/GC/MS
could
involve any of a wide array of SHIP and TPC
processing
studies. The most appropriate application remains
to be
identified, but the Burley sub-blends ex TPC
treatment may
be the right candidate in terms of filling major
knowledge
gaps.
Finally, with an eye to the further future, a
report is
being assembled in which techniques for
fingerprinting
free radicals in smoke are reviewed, this with a
view to
applying chemometrics in this area.
CO
_M'.
1985 British Azncricm Tobacco Co, Lui This must
not be copied or shown to unauthorised persom r"i
BATCo document for Province of BritiSh Columbia 11
November 1999


- 35 -
STATUS REVIEW NOTE
Work Area: 05 Flavour and Biotechnology
Author: T.G. Mitchell
Period Covered: November 1984 - March 1985
OBJECTIVE:
To evolve new flavour compounds for use by the
Group's BAT
flavourists and identify viable routes based on
chemical syn-
thesis or biosynthesis for the supply of new or
existing flavour
ingredients.
AREAS OF RESEARCH:
The main divisions of the work are:
(1) Flavour Chemistry: development,
characterisation and under-
standing of materials for use as additives to
promote smoke
flavour in the broad sense.
(2) Leaf and Biotechnology: to investigate and
develop routes
to useful additives or improved products using
biochemical
or biological processes.
PROGRESS:
(1) Flavour Chemistry
(a) Svenska Tobaks - BAT Project
The final stages of evaluation of a series of
fractions
are almost complete, with the examination of the
following
on reference cigarettes using flue-cured, US
blended,
Oriental and flue-cured + expanded tobaccos:
Cembranoid.fractions : S2 - polar fraction of
McNair 944
. gum, singlet oxygen treated.
S41 - sub-fraction McNair 944.
S53 - polar fraction McNair 944
gum, permanganate/periodate
treated.
Labdanoid fractions S34 - abienol, singlet oxygen
trea-
ted.
S50 - polar fraction Variety PB,
singlet oxygen treated.
1985 British American Tobacco Co. Ud. This roust
not be copied or sbwn to unautborised persorLL
C:)
I'D
I'D
_J
(_P4
co
-C-
C-74
BATCo document for Province of British Columbia 11
November 1999


- 36 -
Test levels were 1 and 2 mg per cigarette for
cembranoid
fractions and 0.1 mg for the labdanoidr, injected
into the
cigarettes with an additional evaluation of S2
sprayed
onto the tobacco.
In view of the unfavourable economics of fractions
based
on large-scale green tobacco extraction, the
recommendation
will be made at the next joint meeting with
Svenska Tobaks
(May 1985) to terminate the project. A research
link
between the groups may be maintained and we are
continuing
the evaluation of flavourants derived from abienol
using
Canada Balsam as source (see following).
W Abienol Oxidation
Oxidation of abienol produces a smoke flavour
material of
high potency which appears to have merit as an
Oriental
enhancer. Production starting from Canada Balsam,
in
which abienol is present at about 10-12%, offers a
simpler
and more direct route than growing specific
tobaccos for
extraction. The current programme is aimed at
identifying
the basic materials costs for a useful fraction
and at
further evaluations of laboratory-produced samples
of oxi-
dised abienol in potential areas of application
identified
by Product Applications Group.
(c) Pectin-Nicotine Studies
This area of work, which is being undertaken
jointly with
Smoke Research (Work Area 02), is intended to
examine the
prospects for controlling or improving the release
of
components into mainstream smoke, in the belief
that
complexing between volatile components such as
nicotine
or flavour materials and cell wall constituents is
a
major factor. Pectin-nicotine interaction is the
first
area being examined.
Liquid NMR studies have demonstrated that
polygalacturonic
acid (as a model for pectin) forms a salt with
nicotine in
a manner completely analogous to simpler acids
(acetic,
oxalic, citric and hydrochloric). There is one
small diff-
erence in the 13C spectrum of nicotine
polygalacturonate,
by comparison with the other salts, which suggests
a
secondary interaction. such as hydrogen bonding,
as well
as the primary one of salt formation.
The pyrolysis-MS examinations have shown
differences in
the temperature of release of nicotine from
complexes
with pectin, depending on it being heated in a
metal boat
(200-260*0 or on a cellulose support (about
390*C). The
type of pectin used led to variation of release
temperature
from 260-410*C when added to denicotinised
tobacco. Solid
state NMR will be used to Investigate the apparent
stabil-
isation of the nicotine-pectin complex by
cellulose.
LN
co
0 1985 British American Tobacco Co. UAL This must
am be copied or shown to unwhorised persons. 41-
4 t..
BATCo document for Province of British Columbia 11
November 1999


- 37 -
(2) Leaf and Biotechnology
(a) Megastigmatrienones
Whilst transformation of megastigmatriene had been
obtained
with microbial cultures, the instability of the
compound
in aqueous media suggested that this route to
megastigma-
trienones had poor prospects for success. At the
same time
an alternative route has been identified using
isophorone
as a relatively cheap and readily available
starting point.
Screening a range of fungi identified two of
twelve which
transformed the isophorone - Aspergillus flavus
and
A.niger. Hydroxy isophorone was produced in 15%
yield in
shake flask culture, with a semi-defined medium
giving
best results and improved recovery of product
using ethyl
acetate in place of trichloroethane for
extraction.
Pilot studies have been made on the further
chemical
conversion of the isophorone, using the latter in
place of
hydroxyisophorone to Identify a procedure for
protection of
the keto group. This was successfully achieved
with ethane
dithiol whereas ethylene glycol and triethyl
orthoformate
did not react.
(b) Microbial Transformations
Only limited effort apart from the
megastigmatrienone
studies has been undertaken in this area in the
review
period. Future requirements, given a successful
synthesis
and product, would be a route allowing high yields
and
ease of separation for the biotechnology step.
Approaches
to this include immobilised and 2-phase systems
and initial
studies of support media (reticulated foam) have
been made.
A culture of Cladosporium resinae has been
obtained as
potentially having high tolerance to appropriate
solvents
in a 2-phase system.
(c) Plant Cell Cultures
The contract at Southampton University for them to
maintain
tobacco cell lines and produce quantities at
regular
intervals has been continued, with the emphasis
remaining
on biotransformations using Izmir and Palotina
cells.
A standard procedure has been followed with
freeze-drying
of contents at the end of the study to obtain
material for
sensory evaluation. This led to identification of
differ-
ences in aroma of cells of the two different lines
being
noted in the absence of any added substrate. To
examine
this aspect further, cell lines of four further
tobacco
varieties - Kentucky 14, NC 2326, Bahia and Basma
Drama -
are being developed at the University using cells
supplied
by us.
L-N
CC)
1995 British American Tobacco Co~ Ltd. This must
not be copied or shown to unautboriud persons.
BATCo document for Province of British Columbia 11
November 1999


- 38 -
Effect of growth phase on abienol transformation
has been
examined by following changes after addition at 0,
3, 7
and 14 days after subculture. Growth curves based
on cell
number, packed cell volume, wet and dry weight,
have also
been conducted. Addition at 7 days (late log
phase)
appeared optimal for abienol transformation.
In studies of effects of plant cells on
megastigmatriene,
unusual effects had been noted earlier when BHT
was added
as an antioxidant. A bright yellow-orange pigment
was
formed, subsequently shown to be the dimeric form
of
butylated hydroxytoluene (BHT). Enzymes with
action
on phenolics seem to be present in Increased yield
in
plant cell cultures as shown by peroxidase assayed
on
pyrogallol: Palotina leaf - 22 units/gram,
Palotina cell
culture - 262 units/gram.
Horse-radish peroxidase was . not active in
causing the
dimeric form of BHT to be produced, but extracts
of tobacco
cells have still to be examined. A cause for
concern is
that high enzyme activities of this type may
interfere in
attempts to cause accumulation of active compounds
in
transformation reactions. The inhibitory actions
of meth-
anol and ascorbic acid on tobacco cell peroxidase
have
been-confirmed, but it is still to be established
whether
they can be used with growing cells.
(d) Leaf studies
The experiment involving spraying
cembranoid-containing
gum onto leaves pre-curing led to a general
preference
for the leaf to which the gum had been added in
cigarette
evaluations. A repeat study was initiated in 1984,
but
is still to be evaluated.
The CASE-award student at Nottingham University
has carried
out a study of cembratriene diol (CBD) breakdown
in curing,
using radiolabelled material. Only 10% of the
starting
material remained unaccounted for after
flue-curing using
a CBD and 17% using o CBD. 70% of the a form was
present
as degradation products whilst 0 CBD was more
stable, with
38% remaining unchanged. A report on this work has
been
prepared. A new. shorter route to synthesis of
labelled
cembrene has been developed and is almost
complete: it
offers the choice of tritium or carbon 14
labelling.
The work at Nottingham has led to a valuable
insight and
experience over the prospects for obtaining
labelled
tobacco components in order to study their fate In
cigar-
ettes.
The study at Birminghan University on use of
pollen
irradiation was completed earlier. The technique
appears
to offer no advantage over other methods for
selective
CO
t985 British American Tobacco Co. Lt& This must
not be copied or sh- to unauthorized persons. _rb.
CN
BATCo document for Province of British Columbia 11
November 1999


- 39 -
gene transfer. By contrast, the marker systems
devised
for its evaluation remain of interest, both for
their own
potential and as simple models. A report on this
work has
been prepared.
An examination of the prospects for feeding
compounds to
tobacco leaves has been conducted, to see whether
this
offered an approach to enhancement in the tobacco
for
subsequent product advantage. Using nicotine, the
concen-
tration in the leaves was generally doubled up to
5 or 6%
dry weight using solutions of either free base or
salts,
with the leaf maintained under conditions
favouring trans-
piration. Nicotine malate gave the best results
overall.
With menthol, whilst some success was achieved
considerable
problems were experienced due to the need for an
organic
solvent and a tendency for the leaf tissue to be
killed
before adequate uptake could be achieved. This
work will
be reported in the next period.
A trial of the Cylproc process has been initiated
in
Venezuela to evaluate its applicability to Burley
with
particular reference to nitrate levels.
Discussions on needs for trials using tobacco in
support of
research projects have been held and areas for
examination
in 1985 identified. These are:
M Nitrosamines - confirmation of 1984 flue-curing
findings (USA);
(ii) Expansion - use of agronomic variables to
control
physical properties and burn of leaf for expan-
sion (USA);
(iii) Alkaloids - production of small quantities
of
alkaloids other than nicotine for pharmacology
(pilot study - GRADC).
(e) Supporting Activities
As part of a Process Research (Work Area 06(B))
project
on the influence of sugar on stem filling power.
yeasts
are being used to prepare samples with reduced
sugars and
minimal additional change.
FUTURE WORK:
With the anticipated completion of the Svenska
Tobaks project,
work on flayourants in the short term will
concentrate on
finalising the position on products derived from
abienol and
megastigmatrienone production.
1935 British American Tobacco Co. LrA This must
ont be cvied or shwu to unautboised persons. CO
BATCo document for Province of BritiSh Columbia 11
November 1999


- 40 -
The main chemical work will be to concentrate on
elucidation of
the role of cell wall components in respect of
smoke transfer
of nicotine. It is proposed to examine a range of
nicotine
complexes with polysaccharides using the solids
state WR probe
which is expected shortly. The possibility of
using a larger
liquid state probe is being examined which it is
believed would
have particular value in allowing the rapid
characterisation of
complex materials such as reaction flavours as
well as other
nuclei, e.g. in relation to fate of nitrate and
nicotine.
Biotechnology activities will include examining
use of enzymes
to aid analysis of cell wall components and the
further screen-
ing of plant cell transformations.
%-0
CO
1995 British American Tobacco Co. Lt& This must
zwt be copW or shown to =authorized persons. _Mh.
CO
BATCo document for Province of British Columbia 11
November 1999


WORK AREA 06
PROCESS RESEARCH (A)
PROCESS TECHNOLOGY
c::>
1985 British American Tobacco Cn~ LUL This must
not be copied or sbovro to unauthorised persons.
c0
BATCo document for Province of British Columbia 11
November 1999


- 41 -
STATUS REVIEW NOTE
Work Area: 06 Process Research (A)
Process Technology
Author: R.W. Hedge
Period Covered: November 1984 - April 1985
OBJECTIVE:
To establish the principles of tobacco expansion
for a second
generation high-order lamina expansion process and
to test the
commercial prospects of recent discoveries.
To obtain basic data on the capability of new
process/processing
equipment.
AREAS OF RESEARCH:
(1) Fundamental Studies
(2) Feasibility Studies
EXPANSION - PROGRESS:
(1) Fundamental Studies
A number of experiments using the glass linear
tube expander
have been carried out to finalise the "mixed agent
synergism"
patent. Additionally, studies on the optimisation
of process
conditions using a pentane/acetone expansion agent
have con-
tinued so as to provide details for the design and
cost studies
being undertaken by external contractors. The work
areas
covered include:
(a) Further work with different solvent mixtures
to confirm
the synergistic effect found with pentane/acetone
has con-
tinued. Marked synergism occurs with a mixture of
agents,
one of which is a non-polar hydrocarbon such as a
C3 to
C6 alkane, alkene, aliphatic or cyclic, the other
of which
is a miscible, C1 to C4. more polar,
oxygen-containing
compound such as an aldehyde, ketone, ether or
ester.
Examples of synergistic pairs so far identified
are
pentane/acetone, hexane/acetone, hexene/methyl
formate,
cyclo-pentane/acetone.
c0
1985 British American Tobacco Co. Ltd. This mm not
be copied or shown to unautharised person&.
BATCo document for Province of British Columbia 11
November 1999


- 42 -
(b) Optimisation of the running conditions in the
Linear Tube
Expander using pentane/acetone as the expansion
agent have
been completed. A low residence time Hambro
separator has
replaced the original cyclone separator used
previously.
It allows much greater control of running
conditions, which
have been set to give a fully expanded product at
a final
tobacco moisture of 14% thus avoiding the need to
re-order.
The initial moisture of the tobacco should ideally
be 30%
and a tube flow rate and temperature of 50 m/sec
and 350*C
are used. Residence time from the start to the end
of the
process is less than 1 second.
(c) A study to compare different levels of DIET
and APEX (GROC
expansion process using acetonelpentane expansion
agent)
in manufactured flue-cured cigarettes has been
completed.
Inclusion levels used were 10, 20, 50 and 100%.
Weight
savings at all levels were significantly greater
with APEX
tobacco and chemical losses were far less than
those in
the DIET samples. Full physical and chemical
analyses,
as well as sensory panel testing, have been
completed.
(d) An additional comparative study between APEX
and DIET
using 6 single tobacco grades has been completed.
Again
the expansion and weight saving achieved with the
APEX
process far outstripped that of DIET. Another
benefit
shown by this study was that the APEX treatment
ameliorated
and improved the taste and flavour of several
poor, unripe
grades evaluated. The full physical and chemical
analyses
and sensory panel testing will be completed by May
1985.
(e) A method for determining pentane and acetone
residues
remaining in APEX expanded tobacco has been set
up.
Routine monitoring of residues showed that little
or no
acetone remains after expansion but significant
levels of
pentane are present. The pentane residues do not
appear
to cause any problems in the product, but ways of
reducing
the residues and further monitoring studies are
being
undertaken, as warehousing fire problems could
arise.
(f) The I i near tube expander has been used to
carry out a range
of preliminary High Humidity Drying (HHD) studies.
Tobacco
subjected to different temperatures and flow rates
has been
processed and the resulting products are currently
being
made into cigarettes for analysis and sensory
testing.
(g) Initial photographic studies on the glass
linear tube have
been carried out and show the flow paths and
distribution
of tobacco particles in the expander tube. This
work was
undertaken as a "sighting exercise' for the larger
study
to be run in conjunction with the SPS group from
Harwell.
(h) The first major expansion report has been
completed and
issued. The second part will be completed Summer
1985. The
examples for the synergy expansion patent are
complete; the C=>
final patent application will be submitted in June
1985. 110
I'D
_1 J
CIP4
CO
1985 British American Tobacco C~ Ltd. This must
not be copied or shown to unauthorived pgrsonL
BATCo document for Province of BritiSh Columbia 11
November 1999


- 43 -
(I) Work on the Vapour Pressure Release Apparatus
(VPRA) has
continued. The synergistic expansion effects found
in the
linear tube using agent pairs such as
pentane/acetone have
been confirmed in the VPRA.
(j) An experiment to determine whether expansion
agent vapour
can be successfully applied to tobacco inside the
VPRA has
been carried out using a second pressure vessel
connected
to the VPRA. The results show this to be a very
effective
method of agent application which considerably
reduces
process cycle time.
(2) Feasibility Studies
Consideration has been given to two areas:
(a) Safety
A "coarse" HAZOP study was carried out In
conjunction with
an engineering consultancy. It concluded that
"assuming
the action points raised will be incorporated,
there is no
reason to expect that the proposed process could
not be
operated in a safe and efficient way". The study
excluded
consideration of a solvent recovery system, which
is known
technology.
(b) Economics
A consultant, Chemical and Thermal Engineering
(CTE), found
that physical data on pentane-acetone-water
systems was
available for the design of a recovery system.
However,
an outline recovery proposal by CTE indicated no
obvious
simplification and cost benefit relative to Freon
recovery
in the G-13 process. Thus this was not a key area
of the
new process. It was decided to carry out a fuller
economic
evaluation by undertaking a design and costing
study for
both the linear tube expansion process (APEX-C)
and the
pressure release process (APEX-B). This would
allow, in
principle, the construction of a pilot plant
incorporating
all relevant commercial features for either or
both. It is
planned to select a design contractor from three
invited
to tender for such a study - to be completed by
the end of
June 1985.
EXPANSION - FUTURE WORK:
(1) Fundamental Studies
Future work will continue on both the Linear Tube
Expander and
Vapour Pressure Release Apparatus. Feasibility
studies into
both process options will be undertaken.
L-4
co
0 1935 Brifish American Tobacco Co. LA This must
not be cvkd or shown to unaudwdsad persons. U-1
rQ
BATCo document for Province of BritiSh Columbia 11
November 1999


- 44 -
(a) Linear Tube Expander
Emphasis will be on three areas of research:
(i) To continue to provide supporting data for
commercial design studies using an azeotropic
expansion mixture of pent a ne/ acetone. The work
will include further optimisation of operating
conditions and the impregnation stage, as well
as the production of samples for cigarette
manufacture and product assessment.
0 i) Further work on low order water-only
expansion
(HHD) will be carried out. The purpose of this
work will be to obtain maximum lamina expansion
with no detectable chemical or sensory attribute
changes.
(iii) The photographic studies will continue in
con-
junction with Harwell. They will be used to
locate expansion within the linear tube and to
determine the expansion profile (i.e. speed of
expansion, whether collapse occurs, etc.).
Additionally, the path the tobacco takes in the
gas/air steam will be photographed to observe
the mass flow characteristics within the tube.
This will aid the definition of the type of
equipment best suited to the new agent or HHD
systems.
(b) Vapour Pressure Release Apparatus
Work with the VPRA will include further testing of
in situ
vapour impregnation using synergistic mixtures.
Optimisa-
tion of temperatures, pressures and cycle times
will also
be carried out using the existing apparatus. A
new,
larger VPRA will be built to allow larger tobacco
samples
to be expanded. It will be used to assess the
scale-up
possibilities of the process to pilot plant size.
The new
VPRA will be completed by May 1985 ready for
evaluation.
(2) Feasibility Studies
It is anticipated that a design tender proposal
will be selected
early in April. There will be close liaison with
the contracor
during the period of tender preparation. If
approval for the
construction of a pilot plant facility is
obtained, a plant and
building contract will be arranged and progressed.
The total
exercise includes not only the provision of
process machinery,
but also site selection, a process building and
all services.
U-4
I M British American Tobacco Cm Ltd. This must not
be copied or shown to unwthodsed persons, CO
Ln
LF4
BATCo document for Province of British Columbia 11
November 1999


- 45 -
PROCESS STUDIES - PROGRESS:
(a) No work has been initiated on fundamental
studies using
the Tobacco Processing Cylinder, following
agreement on
studies undertaken and reported within Work Area
04.
(b) Report RD.1983 has been issued in which it was
concluded
that the carbonaceous reduction Of C02 to CO
occurs more
readily at the surface of WTS or CRS than for
shredded
stem. This led to comparisons of the effect for
extracted
and unextracted stems to try to elucidate the
cause. as
described in the last Status Report.
Studies of the surface of CRS and shredded stem
from
unextracted stem, using X-ray photon spectroscopy,
have
been undertaken at Harwell. The results suggest
that,
while the bulk compositions of shredded stem and
CRS are
very similar, their surfaces are markedly
different both
In respect of the amounts of sulphur, chlorine,
potassium
and calcium and in the ratios of these elements.
A draft report describing the properties of the
six stem
samples (pairs of shredded stem and CRS at three
solubles
levels) and 100% stem cigarettes made for these
samples Is
in preparation. This will include determinations
(currently
in progress in Combustion Studies Group) of the
relative
abilities of the six samples to reduce C02 to CO-
PROCESS STUDIES - FUTURE WORK:
(1) CO Reduction
In view of the surface differences of shredded vs.
rolled, cut
stems further work is planned in order to explore
the surface
components which influence the C02 - CO reduction
step with a
view to achieving lower CO deliveries from lamina
as well as
stem.
cc
(J1
1995 British American Tobacco Co. LuL This must
not be copied or shown to unauthodsed persong,
_Mh.
BATCo document for Province of BritiSh Columbia 11
November 1999


WORK AREA 06
PROCESS RESEARCH (B)
PHYSICS OF TOBACCO
1985 British American Tobacco Co. U4. This must w
be copied or shown to wtauthorised perwnL c0
Ln
LM
BATCo document for Province of British Columbia 11
November 1999


- 46 -
STATUS REVIEW NOTE
Work Area: 06 Process Research (8)
Physics of Tobacco
Author: R.G. Hook
Period Covered: November 1984 - April 1985
OBJECTIVES:
To obtain basic scientific data on the effect of
process condi-
tions on the physical properties of cut tobacco
and cigarettes,
with the aim of reducing cigarette densities,
maintaining cigar-
ette quality, maximising process yields and
reducing tobacco
degradation.
AREAS OF RESEARCH:
(1) Removing-barriers to density reduction.
(2) Process improvements.
PROGRESS:
(1) Removing Barriers to Density Reduction
A survey has been carried out among the CAC
companies to define
the problems associated with ends quality and to
collate the
existing know-how on possible solutions. The
answers confirm
the importance of poor ends quality as a factor
limiting consu-
mer acceptability, as a cause of wastage, and as a
barrier to
density reduction.
It is generally agreed that tobacco particle size
is a key
factor but there are apparent conflicts and
inconsistencies in
the evidence from different sources. One of the
aims of the
work in GR&DC is to specify the relationship in a
form which is
generally applicable and to resolve the apparent
anomalies. A
first step has been the development of a
simulation model which
can be used to predict the effects of different
particle size
distributions. The main predictions of the model
are being
tested experimentally.
The model also deals with the problem of
variability in ends
phenomena. This variability shows itself in large
differences
between repeat measurements of ends fall-out, in
the unrelia-
bility of the relationship between packer rejects
and ends
fall-out, and in the common experience of several
companies
that ends quality problems arise and then go away
without any C=>
clearly Identified cause. The theoretical model,
backed by
1985 British Amerim Tobacco Co. Ltd. This must not
be copied or sho" to wmthcrised pc~ co
Ln
0 N
BATCo document for Province of British Columbia 11
November 1999


- 47 -
experiments on large numbers of individual
cigarettes, makes
it clear that the variability is an intrinsic and
unavoidable
aspect of the phenomenon. Although ends fall-out,
firmness
and pressure drop all depend on the same set of
basic factors,
there are good reasons why ends fall-out should be
more variable
than the other properties.
The practical implication is that more 'effective
control over
ends fall-out would probably be achieved by
monitoring and
responding to changes in particle size
distribution than by
attempting to respond directly to fluctuations in
the fall-out
measurement itself.
A study of the effects of circumference on
firmness has been
completed. It was found that firmness measured by
the Borgwaldt
method increased with increasing circumference at
constant
density. This Is in line with theoretical
predictions for any
firmness method applying a fixed load to one or
more levers.
The same theoretical model predicts that firmness
measured by
methods involving radial compression by a fluid
should be
independent of circumference.
A review of hot collapse was circulated to
operating companies
for comment. There was no clear concensus in the
replies to
indicate what further work is necessary in this
area. The
study of the effects of temperature and moisture
content on
firmness of unlit cigarettes is being extended to
include
cigarettes of varying circumference and with
different papers
(e.g. low sidestream papers).
(2) Process Improvements
The expansion of stern particles has been studied
in helium, air
and superheated steam at 200% and atmospheric
pressure. The
rate and the final level of expansion in helium
was intermediate
between those in air and steam. The heating rate
was higher in
helium than in air because of the differences in
thermal
cDnductivity. The heating rate in steam was even
higher because
of condensation and/or inhibition of evaporation.
The qualita-
tive behaviour of the sample in steam was
different from that
in the other gases. A short video film has been
prepared
showing these effects. These results support the
accepted
view that steam has unique properties as a drying
medium for
tobacco, over and above differences in thermal
conductivity
and thermal capacity.
An investigation of the reasons for an inverse.
correlation
between the expandability of stem and sugar
content has begun
(Work Area 05). Several experiments have been
carried out
involving the addition and removal of sugar from
stems followed
by expansion under different conditions. Although
no consistent
pattern has yet emerged it is quite clear that,
whatever the
effect of sugar per se, other factors which
correlate with
sugars (e.g. maturity and stalk position) must
also play some
part.
IM British American Tobacco Co. Ltd. This must not
be copied or shown to unauthori3ed persons.
C:)
%10
CO
BATCo document for Province of British Columbia 11
November 1999


- 48 -
Methods are being explored to obtain combinations
of high
initial heating rate and moderate maximum tobacco
temperature
which are outside the range of existing drying
systems. The
potential applications, if successful, would be to
a low order
expansion without agents or to an alternative
heating stage
for the DIET process. The heating methods include
combinations
of microwave. radio frequency and hot gases acting
on a bed
constrained relative to the gas stream.. Although
the aim is to
determine the effects on physical and chemical
properties,
experiments to date have concentrated on
measurement of the
temperatures and heating rates.
Pilot plant experiments are being carried out to
optimise stern
cutting conditions. Complete hand-stripped
Brazilian flue-
cured stems were used without rolling and were cut
in three
ways: randomly, as birds-eyes and as long slivers,
the latter
only being followed by a second random cut at 14
cuts per inch
(cpi). In each mode, four nominal cuts per Inch
settings (60,
100, 200) were used on a Molins "windmill" type
cutter.
Each cutting variation was combined with three
different wetting
and drying treatments. Cylinder filling value
results show
that the optimum setting of cuts per inch depends
on the mode
of cutting and the subsequent treatment. The
optimum is
generally between 100 and 160 cpi but for the
double cut slivers
the maximum filling value was obtained using 200
cpi in the
first cut. Measurements of the actual thickness of
cut con-
firmed that this was generally greater than would
be expected
from the nominal setting. Using a different type
of cutter,
BAT Germany have found the actual thickness to be
less than
indicated by the machine setting. It is clear that
a lot
remains to be learned about the optimisation of
stem cutting
and subsequent treatment.
In connection with the work on
being carried out to identify
secondary processes where there
degradation of tobacco.
ends stability, experiments are
the parts of the primary and
is scope for reducing mechanical
Cut flue-cured lamina at 10% moisture content has
been subjected
to tumbling action for up to 40 minutes with only
minimal
changes in particle size distributions. Similar
tests are
planned with expanded tobacco and WTS and at
different moisture
contents.
Although it is well known that tobacco is
mechanically degraded
during cigarette manufacture and that there is a
loss of filling
value as measured by cylinder tests, there is
conflicting evid-
ence on where the degradation occurs, whether it
is avoidable
and what effect it has an cigarette properties.
Experiments
are in hand to clarify some of these issues.
00
1995 Bridsh American Tobacco Co. Lt& TW mw no be
oo*d or iW~ to unwawrisw pe~ Lr
CO
BATCO document for Province of British Columbia 11
November 1999


WORK AREA 06
PROCESS RESEARCH (C)
NOVEL CIGARETTE TECHNOLOGY
c0
IM Mush Amed= Tobacco Co. Ls& TWs must not be
copied or shown to unauthonsed persogm L-n
110
BATCo document for Province of British Columbia I
I November 1999


- 49 -
Laboratory measurements have been made of the
mechanical press-
ures required to compress small samples (ca. 0.5
9) of cut
tobacco In the range of densit'es and compression
rates which
apply in the garniture of a cigarette making
machine. As
expected, the pressures are much greater than
those needed to
compress larger samples slowly to similar
densities in a filling
value cylinder. Also, the variability is much
greater because
of the small sample size. The pressure required to
compress
tobacco to a fixed height under the tongue can
fluctuate
considerably above the average level. It has been
postulated
that these very high stress rates could result in
particle
breakage. Preliminary experiments in the
laboratory and on a
cigarette maker suggest that, even at these high
stress rates,
the degree of degradation in the garniture region
is small at
normal making moisture and temperature. However,
variability
in either moisture content or temperature could
cause some of
the tobacco to become brittle and lead to a sharp
increase in
degradation.
FUTURE WORK:
Cigarette making trials to define the
relat-ionship between ends
stability at constant firmness and particle size
distribution
and to identify other significant factors.
Experiments to identify ways of reducing tobacco
degradation
and controlling particle size distribution.
Measurements of firmness at different temperatures
and moisture
contents on cigarettes of varying circumference
and wrapped in
different cigarette papers (including low
sidestrearn papers).
Evaluation of properties of cigarettes made from
stern cutting
experiments.
Further work on the effects of sugars on stem
expansion.
Continuing development of drying methods giving
high ini-
tial heating rates combined with moderate maximum
tobacco
temperature.
1995 Mtish Amenican Tobocco Co. Ltd. This must not
be copied or shown to unisuakwised pe ca
BATCO document for Province of British Columbia 11
November 1999


- 50 -
STATUS REVIEW NOTE
Work Area: 06 Process Research (C)
Novel Cigarette Technology
Author: J.A. Luke
Period Covered: November 1984 - April 1985
OBJECTIVE:
To conceive and develop cigarette making
technology aimed at the
production of novel cigarettes with novel
materials.
AREAS OF RESEARCH:
These are similar to those discussed in the
previous review.
However, since inception the Novel Cigarette
Technology group has
been asked to undertake a number of short-term,
high priority
product concept exercises for CAC companies, which
has impacted
on the group's main research programme. In
recognition of this
trend in emphasis, the 1985 programme has been
re-styled to show
30-40% of effort In support of specific novel
product/technolog-
ies for CAC companies and to assist the MPDC/PAG
in ventures such
as KILT/SMITH, etc. The programme is now
summarised as follows:
(1) Novel Structured Cigarettes/Materials
(a) Improved Smoker Reaction
(b) Low Weight - Low Cost
(2) Exploitation of Selected Novel
Products/Innovative
lechnologies, to support
(a) CAC Companies
(b) MPDC/PAG Projects
PROGRESS:
(1) Novel Structured Cigarettes /Materials
(a) Improved Smoker Reaction
(I) Annular cigarettes
The first phase of examination of annular
cigarettes
containing wrapped cores is approaching
completion.
Three series of samples were produced with cores
con-
taining 30% V/V of the total tobacco rod. Series I
included low sidestream papers for both outer and
core
S%0
LJQ
c0
1985 British American Tobacco Co. LuL This must
not be copied or shown to unauthorized persons.
BATCo document for Province of British Columbia 11
November 1999


- 51 -
wrappers, using a single tobacco blend; Series II
included Cytrel in the core or annulus, with
rontrols,
to examine effect on sidestream smoke, etc.;
Series
III included WTS in the core or annulus, with con-
trols, to test the well known hypothesis of the
annular region providing most of the mainstream
smoke character. Most objective measurement data
is now available. Results of panel tests on Series
III samples are also available. The panel tests
indicate some directional trends in support of the
mainstream character hypothesis but the effects
were
not marked for the Series II cigarettes.
Cigarettes of each series were sent to CAC
countries
for examination. A report is being prepared.
A second phase of examination is to commence
shortly,
including the effect of. density changes in core
and
annulus, nicotine siting and transfer, the
introduc-
tion of shredded stem, and further work on the
"main-
stream" hypothesis using zones with more
significant
flavour differences.
Evolution of an annular cigarette with an
unwrapped
core is well advanced. The basis of the method is
the extensive modification of a cigarette machine
hopper to embody the principle of using a wide
suction
band which is shaped into a C profile prior to the
introduction of a core tobacco. A number of ways
of
introducing the core are being studied including
pneumatic feeding, a second narrow suction band or
a
second narrow chimney. It is anticipated that
sample
cigarettes will be available in May. The sample
series has been provisionally specified.
(ii) Variable Composition or Density
Producing cigarettes of variable density by the
shaped
ecreteur method is mechanically satisfactory, but
smoke tests on such designs have given
disappointing
results. It is considered, for example, from
various
studies. that a density difference of 2:1 is
required
along a tobacco rod to achieve effects such as the
levelling of the puff profile. It has proved
almost
impossible to achieve this in practice, simply
because
the range of density from (say) 320 mg/cc to 160
mg/cc
produces difficulties at both extremes in terms of
the ability to produce rods. Tobacco rods that
have
been produced contained a much reduced range from
320
mg1cc to 220 mg/cc, and it is assumed that this
has
masked any significant effect on, for example, the
puff profile.
CO
1985 British American Tobecco Co. LuL This must
not be copied or shod to mautborised persons. C7*,
BATCo document for Province of British Columbia 11
November 1999


- 52 -
A similar principle of construction was used for
the
"Codevac" two component variable composition prod-
ucts. This gave an asymmetrically structured
tobacco
rod which led to various objections including
uneven
burning and ash appearance.
A symmetrical design of this type requires a
conical
shaped region of one of the tobacco components,
with
the cone base either at the lit end or the filter
end. A "cone" can be formed by the shaped ecreteur
method acting on the core of an annular cigarette
with phased similar treatment of the annulus
tobacco.
For this reason the experimental hopper described
above is designed to produce both annular and
variable
composition tobacco rods. First examples are
expected
in May/June.
Variable composition products with more than two
components require a different form of
construction.
The feasibility of controlling and mixing several
streams of tobacco particles at cycling rates
equiv-
alent to tobacco rod lengths is being studied.
(iii) Thermally modified cigarettes
The role of carbon in cigarette combustion is
being
comprehensively explored as a fuel, a conductor
and
a selective retention medium. One key aspect is
obtaining suitable forms of carbon from
appropriate
suppliers.
During this period flexible graphite sheet has
been
obtained from Morgan Crucible to extend the work
on
coal cooling by thermal conduction, in an attempt
to
match the CO reduction in both sidestream and
main-
stream smoke recorded for cigarettes containing
metal
conductors. This novel material has a high
thermal'
conductivity and low density. The graphite sheet
was
easily shredded into particles of similar size to
tobacco particles, and blended well. No
difficulties
were experienced in making up cigarettes with 5%
of
shredded graphite in the blend, unlike the
situation
when 5% Asgard carbon fibre was blended with
tobacco.
Analysis showed that, for the former, graphite
distri-
bution cigarette-to-cigarette was good, whereas
greater variability was noted for the carbon fibre
with much carbon fibre remaining entangled on
hopper
comb needles, etc. Cigarettes were submitted for
objective testing and ad-hoc panel smoking. In
addi-
tion, the Aga Thermovision camera was used to
measure
coal surface temperatures. One significant finding
was that, whereas a reference cigarette displayed
a
U-4
co
0 1985 British American Tobacco Co. LAdL This must
not be copied or shown to =wdlaised personL
BATCO document for Province of BritiSh Colurnbia
11 November 1999


- 53 -
marked peak temperature of 850-900*C followed by a
sharp fall, the test cigarette containing graphite
displayed a temperature profile with no distinct
peak
whilst rising only to 650-700*C. Initial
measurements
of mainstream deliveries showed only a small
reduction
in CO (6%), whilst sidestream measurements are not
yet
available. The "ad-hoc" panel has so far been
unable
to confirm an earlier observation of '
'smoothness" in
cigarettes containing carbon fibre, but it is
recog-
nised that the degree of stabilisation in the
carbon
fibre is likely to be important in this respect.
Further work is in hand on both studies.
A range of experimental cigarettes has been
produced
in which carbon dust was spray-coated onto tobacco
as
a suspension in invert sugar casing. Cigarettes
await
analysis. This appears to be rather more reliable
as
a method than the introduction of carbon granules
into
tobacco blends.
M Low Weight - Low Cost
M Reduced circumference tobacco rods
The smoke characteristics of tobacco rods with
circum-
ference in the range 13-19 own have been
extensively
investigated. A report on this subject will be
issued shortly.' covering objective and subjective
performance. Specific variants wrapped in the low-
additive paper 60NC have been submitted as part of
the
circumference-oriented "research series". Variants
have also been used for sidestream studies and
smoker
behaviour.
(it) Shaped cigarettes - ovals
The use of shape as a means of reducing tobacco
weight
has not been tackled seriously in the past, yet it
is
well known that a cylinder is an efficient shape
for
containing a given volume (sic. weight) of
material
such as tobacco. It was discussed earl ter that a
1 own
reduction of one axis of the cross section of an 8
mm
diameter cigarette would yield a 12.5% weight
saving,
yet the resulting shape may not be detected in use
when compared with a handled product from a
soft-cup
pack. However, an equally important aspect of
shape
is how it might be used generally to disguise the
real
volume of a cigarette. For example a circular
product
of 5 mm diameter has the same tobacco weight as a
6.5 mm x 3.8 mm oval product, yet visually the
latter
looks larger. An oval product offers also the
potent-
ial for the use of a proportionately greater
amount of
wrapper for the weight of tobacco contained, which
is
an advantage in further reducing visible
sidestream. CD
CO
19&5 Britisb American Tab- Co. LUL This must rat
be copied or 2hown to unaudwised persom C711
Jt:..
BATCo document for Provin