PRODUCT RESEARCH DEPARTMENT
.C:.
C:)
rIIj
-C~*1
(-,4
CD ,
U-1
QN
BATCo document for Province of British Columbia 23
April 1999


INDEX
PAGE CODE TITLE PERSON14EL
I T-7340 Studies of Filter Ventilation: Malaucene
RLR, AP, MPS
Laser-Perforated Tippings and Porous
Plugwraps.
3 Ventilation (and effective area of
plugwrap)
5 Manufacturing variability
6 Smoke deliveries
10 Position of ventilation holes
10 Gap filter
12 Computer Model of Steady-State Flow AP
along a Ventilated Cigarette.
13 Influence of the Cigarette Paper on AP
Carbon Monoxide Deliveries.
T-727S 1.(a) The effect of T102 in paper
14 T-7279 (b) The effect of citrate level in
paper.
15 T-7275 2. Comparison of papers of 'matched'
porosity from two suppliers.
16 Analysis of Filter Plasticizer Level MPS
17 T-5467 Study of Plasticizer Level Variability
MPS
of Filters
19 T-5463 Evaluation of the Eastman Automatic OTE,
MPS
Plasticizer Monitor
20 T-5467 Subjective Effects of Filter Plasticizer
MPS
22 T-7341 Preparation of Alkaline Filters MPS, AB
23 T-7342 Effects of Addition of Nicotine to MPS
Filters
25 T-7338 Development of Carbon Filter Cigarettes
MPS
25 Types of Carbon
26 Filter construction
-Cbl
CZ)
BATCo document for Province of BritiSh Columbia 23
April 1999


INDEX (cont'd)
PAGE CODE TITLE PERSONNEL
27 Distribution of carbon in dal-
matian filters.
27 Effect of intimate contact of
carbon with triacetin.
29 Effect of moisture on carbon off-
taste.
31 Coated MF-3 carbon from GR & DC.
31 Oil/water partition scan and
carbon off-taste.
32 Development of a 9 mg KS carbon
filter cigarette.
34 T-7278 Establishment of Oil/Water Partition MPS
Technique.
36 T-7278 Use of Oil/Water Partition to Compare
MPS, GWB, APS
Flue-cured Grades.
38 Project CORE RLR
40 T-6810 Flue-cured Taste Enhancers HR. NF. MPS,
TAS
41 T-6811 Project GOURMET HR, NF, MPS, TAS
42 T-3175 Project VIKING TAS
44 T-4412 Project LEROY TAS
46 T-4440 Project BEAUTIFY TAS
C=)
BATCo document for Province of BritiSh Columbia 23
April 1999


T-7340 STUDIES OF FILTER VENTILATION:
MALAUCENE LASER-PERFORATED TIPPINGS AND POROUS
PLUGWRAPS
Personnel: R.L. Rice, A. Porter, M. Scherbak
ObJective: To establish technology that could be
usefully employed In
product development, viz:
1) To establish relationships between degrees of
ventilation
and smoke deliveries.
2) To establish the degrees of dilution that can
be obtained
with different Malaucene laser-perforated tippings
in
combination with porous plugwraps.
Status: For our studies of filter ventilation, the
use of Malaucene
laser perforated tippIngs and porous plugwraps was
chosen
as it was anticipated that ventilation achieved by
such a
combination would be less variable than that
achieved by
other techniques.
For the most part, ventilation technology
developed in this
laboratory was obtained through the examination of
two series
of experimental cigarettes. These cigarettes are
described in
Table I under the headings Experiment 1 and
Experiment 2.
The development of new cigarette brands was
carried out
concurrently and in one instance, it preceded the
more
systematic development of technology. It was
hypothesized
that best control of ventilation could be achieved
by using
a very highly porous plugwrap (Schweitzer 260-R)
and employing
the appropriate tipping paper to obtain the
desired degree of
ventilation.
/2
CD
M.)
0 \
BATCo document for Province of BritiSh Columbia 23
April 1999


TABLE 1
DESCRIPTION OF SAMPLES
Experiment 1
Standard M
Tipping Series
Numbers in test
Vent area, range
Plugwrap
Numbers in test
Porosities, range (WTU)
Filters
Numbers in test
Pressure Drop (inches/20 mm
section)
Cigarette Samples
Number of Samples
Ventilation Range
Paper Porosity
7
6-0.5mm2
1
260.000 only
2
5.3 and 3.8
14
31-74%
750 WTU
Experiment 2
Standard M &
Micro-Perforated
3 and 5
3 to O.075mm2
4
4,000-260,000
1
3.0
32
6-53%
440 WTU
/3
C=)
U-4
Un
BATCo document for Province of BritiSh Columbia 23
April 1999


Following this concept, the cigarettes described
under
Experiment I (Table 1) were designed to determine
the
degree of ventilation that could be obtained by
using
the standard M series of Malaucene tipping papers
with
a wide range in vent areas. This wide range in
degree of
ventilation was also desirable so that a
relationship could
be obtained between the degree of ventilation and
the
reduction in smoke delivery over that of the
unvented
cigarette. Changes in smoke deliveries will be
described
under the heading "Smoke Deliveries".
The cigarettes described under Experiment 2 (Table
1) were
designed to gain an appreciation of the effect of
plugwraps
of different porosity on degree of ventilation. In
this
study. both standard M series and micro-perforated
tippings
were examined. This study has been described in
detail at
the BAT Ventilation Conference, (GR & DC Report
No. RD 1613,
paper No. 10).
The degree of ventilation of cigarettes was
measured in this
laboratory as the flow though the vents when
drawing
1050 cc/minute of air from the unlit cigarettes,
using a
flow-meter developed by BAT Hamburg. This has
recently
been complemented by an instrument developed by G.
Levert,
L a modification of the apparatus described by
Tennessee
Eastman at the 32nd TCR Conference.
Ventilation
It can be assumed that in the ventilated
cigarette, the
pressure drop across the vents is the sum of the
pressure
drops of the tipping and the plugwrap.
Xb.
C:)
C,4
CD
(-n
00
BATCo document for Province of BritiSh Columbia 23
April 1999


In this study the following observations were
made:
1) For the standard M series of tippings, the
major part
of the pressure drop of the combined tipping and
plugwrap
was due to the plugwrap.
Example: The pressure drop across the 1 x
M.3,0.0.0
tipping and Schweitzer 260 R plugwrap was 3.9 cm
for
an air flow of 553 cc/minute (53% dilution). The
pressure drop across the plugwrap was 3.8 cm,
while
that across the vents was only 0.1 cm.
With micro-perforated tippings, the pressure drop
of the
tipping had a greater contribution to the total
but in
no case did it exceed 50% of the total pressure
drop.
2) In these ventilated cigarettes, there is a
glue-free
band of tipping stretching either side of the
line(s)
of holes. The width of this band is set by a
skip-tip
roller on the cigarette making machine.
Ventilating
air enters the holes in the tipping, then spreads
laterally between the tipping and plugwrap before
passing through the plugwrap into the filter.
Schweitzer
have coined the term, effective area of pluqwrap,
which
Is the area over which air spreads after entering
a
perforation in the tipping before penetrating the
plugwrap. This effective area of plugwrap is
believed
L to be very small relative to that of the
glue-free band.
Our studies have shown that the effective area of
a plug-
wrap varies with tippings and for a given tipping,
the
effective area varies with different plugwraps.
/5
CD
PQ
C,
BATCO document for Province of BritiSh Columbia 23
April 1999


Manufacturinq Variability
During the development of Matinee Extra Mild
cigarettes
(designed to deliver 4 mg of tar) significant
differences
in the tar delivery were observed for samples made
on
different making machines and in different
operations of the
same machine. Differences in degree of ventilation
(dilution)
accompanied and, in fact, accounted for these
observed
differences in tar delivery. These differences in
dilution
could not be related to either variation in
"up-stream"
pressure drop or variation in tipping or plugwrap.
It
follows that these differences must be attributed
to changes
in the effective area of the plugwrap as a result
of
variability in the tipping operation. A number of
parameters
are being investigated by this laboratory as a
mean of
explaining or better controlling this cigarette
variability
between manufacturing operations. Part of this
work was
described at the BAT Ventilation Conference
(RO.1613, paper
No. 18).
Recently, the Product Development Group have
observed that by
increasing the glue-free band width from 4 mm to 7
= a
reduction in tar is obtained although the reason
for this is
not known. They will be examining the effects of a
10 mm
glue-free band width. They have proposed that the
use of the
appropriate glue-free band width could help
maintain the
tar del-ivery "on spec".
The variation in ventilation of cigarettes from
different
manufacturing operations precludes the development
of
relationships whereby the precise degree of
ventilation can
.../6
CD
(_14
BATCo document for Province of British Columbia 23
April 1999


-6-
be predicted. However, within each of the
experimental
series described earlier, all samples were made in
sequence
on one machine by the same operators. An
intercomparison
of these data should therefore be valid and
permits an
appreciation of the degree of change that can be
obtained
through the use of different tippings and
plugwraps.
Plots of degree of ventilation against "upstream"
pressure
drop have been prepared for most tipping and
plugwrap
combinations that have been examined in this
laboratory.
The flow-pressure drop relationship for the
tipping and
plugwrap was determined and then the graphical
procedure
described by W.A. Selke (Beitr. zur Tabakforsch.
9.4, 190
(1978)) was employed to determine the effect of a
range of
.,upstream" pressure drops on degree of
ventilation.
Smoke Deliveries
Tar, nicotine, water and carbon monoxide have been
determined
for the cigarette samples described in Table 1. In
addition,
hydrogen cyanide and aliphatic aldehyde deliveries
were
measured for five of these samples. The deliveries
of the
ventilated relative to the unventilated cigarettes
are
plotted as a function of unlit ventilation level
in the
figure (the aliphatic aldehyde data have been
omitted for
clarity). (see p. 9).
Tar, to a first approximation, is reduced by
dilution alone
up to a ventilation level of about 50%. At greater
levels
of ventilation the tar is reduced more than would
be expected
by dilution. This is undoubtedly due to the
increase in
filtration efficiency at the lower velocities.
/7
NJ
C)
BATCO document for Province of BritiSh Columbia 23
April 1999


Nicotine Is not reduced significantly by
ventilation up
to the 15% level and this must be largely due to
the
effective change from restricted - to free-smoking
conditions when ventilation is introduced. As the
ventilation level increases the reduction in
nicot1ne
becomes more pronounced. As with tar, this is due
to a
contination of dilution and increased filtration
efficiency.
Carbon monoxide deliveries are reduced linearly
over the
ventilation range covered, and are less than would
be
predicted by dilution alone. This has been
explained by the
increased diffusional loss of carbon monoxide as
the gas
velocity in the rod decreases. One would expect
tht-
diffusion coefficient,which is proportional
to/porosity of
the paperto influence the relative reduction of
carbon
monoxide; i.e. the reduction of carbon monoxide
should be
greater as the paper porosity increases. However,
no
significant difference was found between the two
paper
samples examined, and it must be concluded that at
least for
the porosity range 400-900 WTU the difference in
carbon
monoxide reduction is hidden within the scatter of
the data
points.
Water is reduced drastically by ventilation, as is
apparent
from the plot. However, this apparent reduction
may be an
artifact of the Cambridge filter method of
measuring water.
It has been postulated by some authors that water
may re-
evaporate from the pad more rapidly as the
ventilation level
increases.
Hydrogen cyanide is also reduced markedly by
ventilation.
The mechanism of thig reduction must involve, in
addition
_C~b
CD
0 \
r"i
BATCo document for Province of BritiSh Columbia 23
April 1999


(Xt~ IPz;~
-0
-60
(ZD
N-)
C)
BATCo document for Province of British Columbia 23
April 1999


-8-
to diffusion and filtration, a change in the
combustion
characteristics of the coal. This work was
reported at
the BAT Ventilation Conference, (R0.1613. paper
No. 9).
Empirical Formulae
Empirical formulae for the reduction of tar,
nicotine and
carbon monoxide deliveries have been devised.
These could
be useful if incorporated into a computer program
to predict
smoke component deliveries. Thus:
CO ~ventilated) 0.984 0.0123 0
CO unventilated)
Tar ventilated) exp. 1.1(1 _/100_0
Tar ~unventilated)
icotin: ~ventilated) 00 1.2j]
exp. [0.5(1 100-0
icot n unventilated
L N LO 5
where 0 is the ventilation of the unlit cigarette.
C), -rrl~' 0 0 5 a o
f
V."
JV
A
/9
C:)
rIj
0
LYN
BATCo document for Province of BritiSh Columbia 23
April 1999


L
L
I
L
i
L
1
0.8-
0
P
4 0.6- 20 CN 0 R NICOTINE
of
LU
-j
` 0.4-
0.2-
0
20 40 60 so
DILUTION
... /10
C:)
r\)
BATCo document for Province of British Columbia 23
April 1999


_10-
Position of the Ventilation Holes
In all the research and development work discussed
earlier.
the row(s) of ventilation holes were positioned or
centred
12.5 mm from the mouth-end. This was selected to
give
maximum ventilation for a given vent area, with
minimal
risk of the holes being covered by the consumer's
lips.
Two examples of differences in degree of
ventilation for
different hole positions are shown below, for
cigarettes
having Malaucene I x M3,0.0.0 tipping over
Schweitzer 260R
plugwrap.
Position of Filter Ci t P D % Ventilation
Holes -7. -D. &L
(mm froR -Wouth-end) J~ (Vents open) Found
Predicted
12.5 3.8 3.9 61 56
17.5 3.8 4.8 51 50
12.5 5.3 5.0 63 59
17.5 5.3 6.4 52 51
The predicted % ventilations were derived from the
relationship
between the "upstream" pressure drop and %
ventilation for
the tipping/plugwrap combination on other
cigarette samples, as
described earlier.
"Gap" Filter
Tests of cigarettes with hand-assembled ventilated
filters
have shown that when the ventilating air enters a
5 mm gap
r_j
all
BATCo document for Province of BritiSh Columbia 23
April 1999


_11-
between two filter segments, rather than directly
into the
equivalent position on the equivalent single
filter, tar and
nicotine yields are reduced by about 20S. The gap
presumably
permits the smoke and air to mix in turbulence
before passing
into the second filter section. This avoids the
bull's eye
effect on the end of the filter.
We plan to make such "gap" filters on a Molins
ADPTC machine,
using a stiff porous plugwrap, and to test them on
manufactured
cigarettes as a means either of increasing the
tar-removal
efficiency of a filter at a given level of
ventilation. or of
decreasing the level of ventilation while
achieving the same
tar reduction.
... /12
C:)
C)
Cr%
BATCo document for Province of BritiSh Columbia 23
April 1999


COMPUTER MODEL OF STEADY-STATE FLOW ALONG A
VENTILATED CIGARETTE
Personnel: A. Porter
Objective., To study the effects of various
cigarette parameters on
gas phase concentrations from ventilated
cigarettes,
using computer modelling.
Status: During a visit to GR & DC, a computer
program has been
developed for predicting the pressure drop,
ventilation
level and carbon monoxide concentrations from a
ventilated
cigarette under constant flow conditions Q3R & DC
Report No.
RD.161a. Since the program involves about twenty
cigarette
parameters, the first task now is to construct a
data set
of all the parameters as they relate to ITL
cigarette brands.
Once parameter sets have been established, the
program will
be used to establish the influence of component
variations
on total cigarette ventilation levels. This may be
useful
for interpreting differences in ventilation levels
observed
during cigarette production.
A project for the longer term involves correlating
the
predicted carbon monoxide concentrations with
actual smoking
machine results. This will make the program a b~~
practical
tool in the design of cigarettes with low carbon
monoxide
deliveries.
... /13
C:)
rQ
(:D
CC)
BATCo document for Province of BritiSh Columbia 23
April 1999


-13-
INFLUENCE OF THE CIGARETTE PAPER ON CARBON
MONOXIDE DELIVERIES
Personnel: A. Porter
Objective: 1) To investigate the influences of (a)
titanium dioxide
(b) citrates in cigarette paper on carbon monoxide
deliveries.
2) To compare carbon monoxide deliveries from
cigarettes
made with 'matched' papers from two suppliers.
l(a). T-7275 - THE EFFECT OF TITANIUM DIOXIDE IN
PAPER
Cigarette samples with paper containing different
levels of
titanium dioxide (0, 0.3. 2.4%) have been compared
for
delivery of carbon monoxide. The results showed
that
titanium dioxide in cigarette paper has no
significant
influence on carbon monoxide deliveries. Previous
results which have shown a positive correlation
between
titanium dioxide in cigarette paper and carbon
monoxide
delivery may have been due either to the
methodology used
or to a lack of adequate control samples.
... /14
BATCo document for Province of BritiSh Columbia 23
April 1999


-14-
1 (b) T-7279 - THE EFFECT OF CURATE LEVEL IN PAPER
Two series of Wattens papers (W60 - 430 WTU and Wl
00 -
860 WTU), containing various levels of a 1-1
sodium-
potassium citrate mixture (0-2.5%), have been used
to
make cigarette samples. Both series showed a
reduction in
carbon monoxide per puff with increasing citrate
level,
although it was more pronounced for the W100
series. In
this series an increase in citrate content from 1%
to 2%
would have reduced the CO yield per puff by 6%. In
the
W60 series, this same increase in citrate level
would have
given a much smaller decrease in CO per puff.
However these
decreases compliment the reduced puff numbers of
the papers
with higher citrate levels to further reduce
cigarette CO
yields.
Previously, Dashley (GR & DC Report No. RD.1448)
reported
that a similar increase in level of sodium
citrate, in
papers of porosities similar to W60 and WlOO,
caused increases
of about 20% in CO yields per puff. The choice of
alkali
citrate salt could be the controlling feature
since in
Dashley's work, even a high level of the potassium
salt
gave no increase in CO per puff.
Our 510 paper (440 WTU) from both Schweitzer and
Ecusta has
1 :1 sodium-potassium citrate mixture. Our work
suggests
hat further samples of paper with potassium
citrate alone
:
be obtained for testing, with the aim of achieving
further
reduction of CO yields.
4 b.
CD
_j
CD
BATCo document for Province of BritiSh Columbia 23
April 1999


-15-
2. T-7275 - COMPARISON OF PAPERS OF 'MATCHED'
POROSITY FROM TWO
SUPPLIERS
This study was carried out to see if there was a
significant
difference in carbon monoxide deliveries from
cigarettes
made with papers of equal porosity from our two
main
suppliers, Schweitzer and Ecusta.
Cigarette samples with 510 (440 WTU) and 556 (250
WTU) papers
from each supplier were examined. The measured
paper
porosities were within the range quoted by the
supplier,
but both Schweitzer samples had slightly lower
porosities
than the equivalent Ecusta samples. The carbon
monoxide
delivery was higher (by 10%) for the sample with
Ecusta
510 paper than for that with Schweitzer 510, but
was lower
(by 10%) for the sample with Ecusta 556 than for
that with
Schweitzer 556. Since these differences are within
the
expected variability of the method, we can
conclu&- that these
papers can be used interchangeably as far as
carbonfoonoxide
deliveries are concerned.
... /16
CD
BATCO document for Province of British Columbia 23
April 1999


-16-
ANALYSIS OF FILTER PLASTICIZER LEVELS
Personnel: M.P. Scherbak
Objective: To institute an improve analytical
method for plasticizer
levels on filters.
Status: A modified analytical method has been
instituted for
plasticizer determination, in which filters are
extracted
with 1.2-dichloroethane in an ultrasonic bath. and
the
solutions then analysed by gas chromatography
using a
glass column containing 2% OV-101 on Chromosorb Q
in a
Hewlett-Packard 5840A instrument. Recoveries for
triacetin and TEGDA were 98.6 and 99.1%
respectively,
with coefficients of variation of 3.0 and 3.4%
respectively.
These were determined by spiking 20 mm filter
plugs with
the plasticizers.
... /17
BATCO document for Province of British Columbia 23
April 1999


T-5467 STUDY OF PLASTICIZER LEVEL VARIABILITY OF
FILTERS
Personnel: M.P. Scherbak
Objective: To determine the variability of
plasticizer level in filter
rods.
Status: Two studies have been completed: -
(a) Using Nicotine as a Tracer
Before the improved plasticizer analysis method
was available,
we determined the variability of the level using
nicotine as a
tracer. Initially we showed that acetate filters
spiked with
known amounts of nicotine in triacetin showed no
loss of
nicotine over a period of 2 months. Filter rods
were then
manufactured from two very different tows
(5.0/39000 and
Amcel\ B-45), using a solution of 5% nicotine in
triacetin,
applied to the tow at a target level of 5%. The
variabilities
in plasticizer level were as follows:
Plasticizer Level M
Tow Target Found S.D. C.V.
5.0/39,000 5 4.37 0.32 7.3
Amcell B-45 5 4.77 0.55 11.4
W Usinq the Modified Method for Plasticizer
Analysis
For a study on the subjective effects of
plasticizer level
and type (see later), batches of filter rods were
manufactured
from 3.3/44,000 tow using (a) triacetin at
different target
levels (3%, 6.5% and 10%), (b) TEGDA at a target
level of
6.5% and (c) no plasticizer.
... /18
C:)
0
--J
Q-4
BATCO document for Province of BritiSh Columbia 23
April 1999


-18-
The new gas chromatographic method was used to
determine
the variability of the level of plasticizer on
individual
17 mm filter sections, making 12 determinations
for each
batch. The results werp excellent, as shown in the
following
table:
Plasticizer Triacetin TEGDA
Target % 3.0 6.5 10 6.5
1 Found 3.5 6.1 10.8 7.9
S.D. 0.42 0.38 0.59 0.36
CV% 12.0 6.2 5.5 4.6
... /19
CD
r*j
41.
tA
BATCo document for Province of BritiSh Columbia 23
April 1999


_19-
T-5463 EVALUATION OF THE EASTMAN AUTOMATIC
PLASTICIZER MONITOR
Personnel: D. English, M.P. Scherbak
Objective: To determine the reliability of the
Eastman Automatic
Plasticizer Monitor.
Status: An Eastman Automatic Plasticizer Monitor
was installed on
a XDF-2 filter rod maker in the Quebec Plant.
Manufactured
filter rods were sampled regularly (100 rods per
sample)
over a period of 2 months, noting the monitor
readings at
the time of sampling. The plasticizer contents
were determined
first by the normal wet-dry weighing method and
then by our
improved gas chromatographic method. Statistical
techniques
(ranking and paired-comparison T tests) showed
that the
Eastman monitor gave a more accurate estimation of
the
plasticizer level (using data from the gas
chromatographic
analyses as standard) than did the wet-dry
weighing method
(mean differences from 'standard' 0.12 and 0.45
gm/lOO rods
respectively for the two methods).
.../20
Jib
C:D
BATCO document for Province of British Columbia 23
April 1999


-20-
T-5467 SUBJECTIVE EFFECTS OF FILTER PLASITICIZER
Personnel: M.P. Scherbak
Objective: To determine whether our current levels
of triacetin are
optimum in terms of subjective smoking properties
using
two filter tows and to recheck the relative
subjective
acceptability of using triacetin as compared to an
alternative plasticizer (TEGDA).
Status: For this study, filter rod samples were
manufactured from
3.3/44000 tow using (a) triacetin at different
target
levels (3, 6.5 and 10%), (b) TEGDA at a target
level of
6.5%, and (c) no plasticizer (control). Using
these rods,
modified du Maurier K.S. cigarettes (84/17) were
manufactured.
Smoke Analysis
The cigarettes were smoked to a 30 mm butt, and
smoke yields
were determined for tar, nicotine, total phenols,
total
aliphatic aldehydes (TAA) and HCN. The results in
the
following table showed the expected reduction in
phenols with
increasing triacetin level, but no discernible
trend in tar
or nicotine yield, and no change in TPM pH. Vapour
phase
yields were slightly lower at higher triacetin
levels.
... /21
-0 b.
CD
N-)
C"
BATCo document for Province of BritiSh Columbia 23
April 1999


Plasticizer Level Tar Nicotine Phenols TAA HCN TPM
Puff
% mg/ciqt. mg/cigt. ug/ciqt. mg/cigt. ug/cigt. pH
No.
Control 0 17.2 1.02 70 1.71 289 5.9 9.2
Triacetin 3.5 17.6 1.09 68 1.74 282 5.8 9.3
6.1 17.1 1.06 56 1.65 265 5.8 9.3
10.8 17.4 1.10 46 1.33 259 5.9 9.4
TEGOA 7.9 16.5 1.04 61 1.60 279 5.8 9.3
Subjective Evaluation
The du Maurier K.S. cigarettes with different
levels of
triacetin ranked as follows in descending order of
subjective acceptability:
1. 6.1Z - overall marginally the most acceptable,
mainly on the basis of taste quality.
2. 0% A 10.8% - 10.8% might provide the most
impact
and taste of all the levels tested.
3. 3.5% - marginally the least accepable due to
poor
taste quality.
This confirmed that for this type of filter with a
standard blend, the normal triacetin level (6.5%)
is
optimum. The results of the subjective comparison
of
the du Maurier K.S. cigarettes with triacetin
(6.1%)
and TEGDA (7.9%) indicated that the sample with
the
triacetin plasticizer was subjectively more
acceptable
than that with TEGDA due mainly to less
irritation.
This thus reaffirms earlier findings.
To complete this project, it is planned to
investigate
the subjective effects of triacetin levels of 0%.
3%,
5% and 10% in the Matinde Regular cigarette
configuration.
(less 'tasty' blend, 2.1/52,000 tow, 5% normal
triacetin
level).
... /22 CD
BATCo document for Province of BritiSh Columbia 23
April 1999


-22-
T-7341 PREPARATION OF ALKALINE FILTERS
Personnel: M.P. Scherbak, A. Bishop
ObJective: To find satisfactory means of preparing
alkaline filters.
Status: With flue-cured cigarettes, alkaline
filters are known to
reduce the tar/nicotine ratio of the smoke. Most
previous
work towards alkaline filters has concentrated on
using
dual paper/acetate filters, with the water-soluble
alkaline
material on the paper. Unfortunately paper filters
have an
unattractive effect on smoke taste. In earlier
work at
ITL, attempts were made to apply alkaline
materials to
acetate filters as slurries in triacetin, but the
slurries
were not sufficiently stable.
However stable slurries of 7 alkaline materials in
triacetin
have now been prpared by ball-milling, as
suggested by
Brown & Williamson, at concentrations suitable for
application
to tow by the brush applicator. The materials and
concentrations
are: sodium carbonate (25%), sodium bicarbonate
(40%).
disodium hydrogen phosphate (25%), aluminum
hydroxide (40%),
diammonium phosphate (20%), polyvinyl pyrollidone
(20%) and
magnesium oxide (11%). Stability of the slu"ies
was judged
by their settling rate, viscosity and ease of
resuspension.
These suspensions will be applied to filter rods
(2.1/52,000
tow), and made into dual filters using the ADPTC
machine.
T"A
As a first trial, we plan to determine how these
alkaline
filters change the smoke composition and
subjective smoking
properties of cigarettes giving 9 mg of tar.
/23 J~-
... CD
kA
co
BATCo document for Province of BritiSh Columbia 23
April 1999


T-7342 THE EFFECTS OF THE ADDITION OF NICOTINE TO
FILTERS
Personnel: M.P. Scherbak
Objective: To study the affect of addition to
filterS of nicotine
(or "purified tobacco extract") on the
tar/nicotine
ratio and the subjective acceptance of a higher
level of
nicotine for low tar cigarettes. This would help
cther
studies in the laboratory.
Status: In preparation for this experiment, the
stability of nicotine
on acetate filters was studied by syringe
injection of filter
rods with known amounts of nicotine in triacetin
and deter-
mining nicotine periodically. No significant loss
in
nicotine was observed during 2 months of testing.
Filter
rods were then made with 2.1/52,000 tow using a
solution of
30% nicotine in triacetin and a target plasticizer
level of
5%. In fact the filter rods contained 0.95 mg
nicotine per
10 mm length. Dual filter cigarettes were
hand-assembled,
using 64 mm Matinee tobacco sections, 10 mm
sections of
the 2-1/52,000 tow (with and without added
nicotine), and
5 mm acetate mouthpieces (5.0/39,000 tow). The TPM
and
nicotine yields of these cigarettes, smoked to a
30 mm
butt, are shown in the following table:
/24
jt~.
CD
BATCo document for Province of BritiSh Columbia 23
April 1999


-24-
TPM Nicotine TPM TPM/NIC.
Cigarette mg/cigt. mg/cigt. pH Ratio
Control 13.6 0.61 6.35 22.3
Sample with 12.8 0.84 6.25 15.2
added nicotine-
The addition of nicotine to the filter
significantly
decreased the TPM/nicotine ratio of the smoke
without
increasing its pH. The transfer of the added
nicotine
to the smoke was about 24%, which is much higher
than
the 6-7% transfer found in earlier work when using
PCL
reinforced with nicotine citrate.
The subjective evaluation of the cigarettes
indicated that
the addition of nicotine to the acetate filters
resulted
mainly in increased impact, with a minor increase
in
irritation.
... /25
NJ
CO
CZ)
BATCo document for Province of BritiSh Columbia 23
April 1999


-25-
T-7338 DEVELOPMENT OF CARBON FILTER CIGARETTES
Personnel: M.P. Scherbak
Objective: To find the most acceptable ways of
using carbon filters
in cigarettes with a range of tar yields from 15
mg down
to 5 mg or lower.
Status: We have the machine capability of
manufacturing dalmatian
carbon filter rods (i.e. carbon granules
interspersed in
acetate tow), and for manufacturing the required
dual filters
A by the AiJPTC machine. However while all our
dalmatian rods
have been machine-made, we have not yet started to
used the
ADPTC machine for dual carbon filters, so that to
date, all
our work has been with hand-assembled dual or
triple (chamber)
filters. A number of supporting Investigations
have been
carried out, namely of types of carbon, types of
filter
construction, effects of plasticizer level and
moisture level
and the use of coated carbon from GR & DC.
Types of Carbon
Two major types of carbon are available from
Pittsburgh
Carbon Division of Calgan Corporation, namely PC8
(coconut-
based; virtually Identical to MF3 from GR & DQ and
BPL
(coal-based). Tests of 12 x 30 mesh samples of
each in
hand-assembled triple (chamber) filters have shown
them to
have equal filtration efficiencies for HCN and for
total
aliphatic aldehydes (TAA). A subjective smoking
panel
... /26
NJ
CD
BATCo document for Province of BritiSh Columbia 23
April 1999


-26-
found no signi "cant difference between Cigarettes
with the
2 types of carbon with respect to level of
off-taste and
acceptability of the smokes. However PCB is less
friable
than 8PL, which is an advantage from a
manufacturing stand-
point.
Filter Construction
Dual filters with a dalmatian filter section with
either
PC8 or BPL charcoals, at loadings of 30 and 70 mg,
were
found to be less effective at removing HCN and TAA
than
equivalent triple (chamber) filters having the
same loadings
of carbon in the chamber and matched in total
pressure drop.
The efficiencies for TAA are shown in Table 1.
Table I % Filtration Efficiencies for Aldehydes:
Dalmatian v Chamber
Carbon Type % Filtration Efficiency
Loading As Dalmatian As Granules
Filter in Chamber
PCS - 30 mg is 40
PCB - 70 mg 27 44
SPL - 30 mg 18 40'
BPL - 70 mg 38 51
This difference in efficiency could have been due
to a
number of factors:
a) poor distribution of carbon in the dalmatian
filter
b) intimate contact of carbon in-the dalmatian
filter
with the triacetin plasticizer.
c) greater protection from particulate matter of
the carbon
in the chamber.
toll' . -r
A A 14-1 r ir- e-S .../27
C:)
0
CO
BATCo document for Province of British Columbia 23
April 1999


-27-
a) Distribution of Carbon in Dalmatian Filters
Visual inspection showed that in our dalmatian
rods
and in the carbon filter cigarettes on the
Canadian
market, the carbon granules were concentrated on
one
side of the filter. This gives the smoke chance to
bypass
the carbon as it stream down the inter-fibre
spaces.
This is less marked with partly-filled chamber
filters, kk<"_t1-j
presumably because of turbulence in the space. The
poor
distribution of carbon in the dalmatian rods is
caused by
its tumbling as the tow is funnelled
into-th"arniture.
Following a suggestion by bash Y (iCi DC),
thii-aistriru-
, ly-
tion has since been improveJ-b)rinserting an
additional
bar to support the tow before It enters the
garniture.
This improved distribution resulted in improved
efficiency
for aldehydes, and in fact could account for the
difference
in efficiency In Table 1. viz:
Carbon % Filtration Efficiency
Poor Distribution Improved Distribution
PCB-70 mg 31 44
b) Effect of Intimate Contact of Carbon with
Triacetin
Two tests were made of the effect of direct
contact with
triacetin on carbon filter efficiency for
aldehydes.
(1) Carbon from the dalmatian filters included in
Table I
was retrieved and retested in chamber filters,
giving the
results shown in Table II.
... /28
CD
NJ
(A
CD
00
U-4
BATCO document for Province of British Colurnbia
23 April 1999


-28-
Table II. Fresh v Retrieved Carbon in Chamber
Filters
Carbon Type 'Fresh' Carbon Retrieved
& Loading Carbon from Dalmatian
Filters
PC8 - 30 mg 40 41
PCB - 70 mg 44 44
BPL - 30 mg 40 34
BPL - 70 mg 51 45
Evidently the triacet1n had not deactivated the
PCB carbon
in the dalmatian filter relative to the 'fresh'
carbon
control, although it did reduce the aldehyde
efficiency
of the BPL type, possibly through loss of some
fine
particles of this more friable carbon. This
friability
of BPL may partly explain its higher efficiency in
this
test at the 70 mg loading in both the dalmatian
and chamber
filters.
(ii) Two samples of dalmatian filter rods were
manufactured
using 3.3/37,000 tow and equal loadings (75 mg) of
PC8
carbon, with one having no triacetin and the other
having
5.5% triacetin: Dual - filter cigarettes (10/10,
with a
2.1/52000 acetate mouthpiece) were hand-assembled
just
prior to smoking, using two 64 mm tobacco sections
capable
of giving significantly different smoke yields,
viz. Matinee
blend in 440 WTU citrate paper, and Player's blend
in 140 WTU
phosphate paper. The filtration efficiencies for
aldehydes
are given in Table III.
... /29 CD
CD
CO
BATCO document for Province of British Columbia 23
April 1999


-29-
Table ILI Effect of Tr1acetin in Dalmatian Filter
Tobacco Carbon % Triacetin % Filtration Efficiency
Section Loading (mg) on Dalmatian Filter for
Aldehydes
Matinee 75 0 47
Matinee 75 5.5 46
Player's 75 0 32
Player's 75 5.5 29
These results confirm that close contact with
triacetIn
has little effect on the aldehyde filtration
efficiency
of the carbon.
A subjective smoking test indicated a small
preference for
the samples without plasticizer on the carbon
filter section,
although the smokes were similar in overall
acceptability.
c) Degree of Protection of Carbon from Smoke
Particulate
Matter
It Is well known that the filtration efficiency of
carbon
is lower for the last half of a cigarette than for
the first
half due to deactivation by the smoke particulate
and/or
vapour phases. We have made no test to check the
effect of
protection of a dalmatian filter by a further plug
of
acetate.
Effect of Moisture on Carbon Off-Taste
Earlier work at ITL had indicated that the degree
of carbon
off-taste could be reduced if the carbon was
brought to 10%
... /30 -Ch.
CD
BATCo document for Province of BritiSh Columbia 23
April 1999


-30-
moisture before making filters, relative to
cigarettes
freshly assembled with dry carbon. To recheck this
effect, the subJective smoking panel smoked
freshly made
"Matinee" K.S. cigarettes having 75 mg PCB carbon
in a
dalmatian/dual filter, with the carbon at a series
of
moisture levels (0, 5, 11 and 20%). The carbons
with
11 and 20% moisture gave a less obvious carbon
off-taste,
in contrast to those with 0 or 5% moisture.
In dalmatian filter rods, dry carbon attained 8%
moisture
in sealed Jars, and 19% in a conditioned room (60%
RH)
after 2 weeks.
Comparison of freshly assembled carbon filter
cigarettes
with carbon at 8 and 20% moisture content showed
that the
extra moisture reduced the aldehyde filtration
efficiency
by 14% for cigarettes with the Matinfe tobacco
section,
and 7% for those with the Player's section.
It is suspected that filters containing carbon
brought
to 20% moisture before exposure to tobacco
volatiles may
have a different selectivity towards smoke and
give a
better smoke balance (less off-taste) than filters
brought
to the same moisture content during exposure to
tobacco
volatiles (i.e. equilibrated at 60% RH as
cigarettes). We
plan to test this idea.
... /31
jnbb
CD
Nj
C:)
CO
BATCo document for Province of BritiSh Columbia 23
April 1999


-31-
Coated MF3 Carbon from GR & DC
GR & DC supplied us with development samples of
MF3
coconut-based carbon (equivalent to PCB) in both
coated
(with polyvinyl alcohol) and uncoated control
forms.
Dalmatian filter rods were manufactured with these
carbons
at a loading of 70 mg per 10 mm section. Dual
filter
cigarettes were then hand-assembled for subjective
testing using Matinee tobacco sections, since the
smoking panel has found they could detect the
off-taste
more easily using these tobacco sections.
However after repeated testing, the panel found
off-taste
for each sample, but was unable to find a
significant
difference in level of off-taste between the
samples. This
comparison will be repeated when manufactured dual
filter
cigarettes are available. Coating the carbon
reduced its
filtration efficiency for aldehydes by about 10%.
Oil/Water Partition (O/WP) Scan and Carbon
Off-Taste
GR & DC have established a relationship between
the degree
of carbon off-taste of smoke and data from its
oil/water
partition (O/WP) scan (RD.1330). They obtained a
smooth
curve in plotting panel off-taste scores against
the ratio
of O/WP phenols/pyridines (representing
non-volatiles/semi-
volatiles).
... /32
C=)
rIJ
0
co
BATCo document for Province of BritiSh Columbia 23
April 1999


-32-
We found that this ratio of O/WP phenols/pyridines
was
reduced when carbon was brought to 20% moisture
before
filter manufacture. This is shown in Table IV.
Table IV Comparison of O/WP Data and Off-Taste
(Matinde Tobacco Section, PCB Carbon)
Carbon O/WP Phenols Subjective
Loading (mg)l %
1 Moisture Pyridines Evaluation
75 0 1.98 Obvious off-
taste
75 20 1.59 Less obvious
off-taste
0 (Control) 0 0.74
We intend to use this O/WP ratio as a screening
technique
for development samples of carbon filter
cigarettes, prior
to subJective testing.
Development of a 9 mq KS Carbon-Filter Ciqarette
A single exploratory trial has been made to see
what smoking
experience a 9 mg carbon filter cigarette might
give. Dual
filters (10 mm dalmatian filter, with 75 mg of PCB
at 20%
moisture in 3.3/37,000 towand 10 mm 2.1/52,000
mouthpiece)
were hand-assembled on to 64 mm tobacco sections
(Player's
blend in 140 WTU phosphate paper). Half the
cigarettes were
hand-ventilated using a needle to give a target
tar yield of
9 mg. The TPM and nicotine yields were determined
for the
vented and control cigarettes, and duplicate O/WP
scans
were also run. The results are summarized in Table
V.
... /33
C)
N)
U-4
C)
c0
c0
BATCo document for Province of British Columbia 23
April 1999


-33-
Table V A Ventilated Carbon Filter Cigarette
% Ventilation P.D.(ins) Puff Smoke Yields mg/cigt)
O/WP Phenols
(holes open) No. TPM Nicotine Pyridines
0 5.7 11.0 24.2 1.11 3.2
59 3.1 12.8 9.4 0.54 14.0
The O/WP ratio of phenols to pyridines showed that
the
vented cigarette had a very unbalanced smoke. A
small
panel found that the vented cigarette gave a very
pronounced off-taste, but not of the type normally
associated with carbon. GR & DC have found that
the
combination of ventilation and carbon can produce
smoke
with a high value for this ratio e.g. Viscount No.
1 K.S.,
6.7 (R.D.1604).
We shall pursue this development project when
manufactured
dual filter cigarette samples are available.
... /34
CD
0
00
10
BATCo document for Province of BritiSh Columbia 23
April 1999


-34-
T-7278 ESTABLISHMENT OF OIL/WATER PARTITION
TECHNIQUE
Personnel: M.P. Scherbak
ObJective: To establish the oil/water partition
(O/WP) technique as
a furth:r means of characterizing smoke, to help
in a variety
of proj cts.
Status, The O/WP procedure has been successfully
established, using
a Borgwaldt RM-20 smoking machine and a
Hewlett-Packard
reporting gas chromatograph (Model H/P 5840A).
Tests have been completed to maximize the
performance of this
instrument for O/WP and to improve its
reproducibility,
including the introduction of a modified component
by the
Hewlett-Packard laboratory at Avondale, U.S.A. For
this
work a single grade cigarette has been used as a
standard.
The position and identity of peaks on the O/WP
scan was
ascertained by spiking smoke with the following
pure
compounds (GR DC's numbering of peaks was
adopted):
Peak No. Peak Name
4 Propionaldehyde
9 Acetone
12 Isovaleraldehyde
29 Pyridine
52 Furfural
57 Benzaldehyde
...135
C:)
\b
CD
BATCO document for Province of BritiSh Columbia 23
April 1999


-35-
Peak No. Peak Name
60 5-Methylfurfural
64 Furfuryl alcohol
74 Guaiacol
75 Nicotine
so Phenol
84 m-Cresol
The OjWP technique has been shown to function well
and to give good separation of peaks.
... /36
CD
PQ
BATCo document for Province of BritiSh Columbia 23
April 1999


-36-
T-7278 USE OF OIL/WATER PARTITION TO COMPARE
FLUE-CURED GRADES
Personnel: M.P. Scherbak, G.W. Boswall, A.P.
Schaffer
ObJective: To study the smoke differences among
cigarettes of represent-
ative single-grades of flue-cured tobaccos from
Canada
together with some grades from the United States,
so as to
obtain a better understanding of the scope for
achieving
product development goals through leaf blending.
Status: Duplicate O/WP scans for cigarettes
representative of
single flue-cured grades from the 1976 crop have
been
completed. The data were put on punch cards for
analysis
using the SPSS computer package (Statistical
Package for
Social Sciences). A disc file was created in the
General
Electric computer and our local terminal has been
used for
performing "Discriminant Analysis" of the data.
This
technique seeks combinations of O/WP peaks which
produce the
best possible separation between grades, and
discriminant
functions are given which maximize the variance
between grades
relative to the variance within grades.
Each tobacco sample was identified in two ways:
1) As an individual grade (50 grades were
represented).
2) As belonging to a specific "Grade Type Group"
according
to the ITL leaf classification technique - the
"Stalk
Position vs Maturity-Blemish" grid. Each "Grade
Type Group"
may contain several individual grades (24 "Grade
Type Groups"
were represented).
... /37
CD
rQ
C>
\D
N)
BATCo document for Province of BritiSh Columbia 23
April 1999


-37-
Although we encountered problems which are often
associated
with establishing new systems and procedures in a
laboratory,
a limited number of "Discriminant Analyses" have
been
successfully run on selected O/WP peaks up to peak
44.
Additional analyses are contemplated using other
peaks. With
the maximum number of discriminant functions, it
has been
shown that the computer correctly identified all
the samples
by grade, while the percentage of samples
correctly classified
by the computer into "Grade Type Groups" was only
94.8%.
Reducing the number of discriminant functions to
7, which
accounts for about 95% of the variability
(eigenvalue sum),
the percentage of samples correctly classified by
the
computer into "Grades" totalled 99% (105 samples
out of 106).
Reducing the number of discriminant functions to
3. which
accounts for only 80% of the variability
(eigenvalue sum),
the percentage of samples correctly classified by
the computer
into "Grade Type Groups" was reduced to 76.5%.
Since the discriminant analysis was based on data
obtained from
duplicate O/WP analyses of all grade samples, the
inherent
variability has been artificially reduced and the
results
undoubtedly present an unrealistically good
picture for grade
classification. However, when the grades were
classified
into "Grade Type Groups" according to the ITL
grid. a more
realistic measure of the variability of
like-samples was
obtained as reflected in the reduced ability of
the Discriminant
Analysis to correctly classify the grades into
"Grade Type
Groups". The early availability of the 1977 and
1978 grade
data will enable us to determine If crop years
contribute
significantly to the O/WP profiles of grades.
38
... / C)
C)
BATCo document for Province of BritiSh Columbia 23
April 1999


-38-
PROJECT CORE
Personnel: R.L. Rice
Objective- To develop a cigarette with a central
core and an annulus
made of different tobaccos, and so find a means of
making
more efficient use of our most tasty tobacco
grades.
Status: SAT Hamburg found that if a cigarette
tobacco column has a core
of one tobacco type and an annulus of a different
tobacco, then
the mainstream smoke has the character of that of
the tobacco
in the annulus, while the sidestream smoke has the
character
of the core tobacco.
Sandwich-type cigarettes were hand-made at ITL.
assuming that
this could approximate to the structure of the
core cigarette.
Comparisons were made between a mild Canadian
blend (Matinde)
and burley, CRS and American flue-cured tobaccos.
In the
subjective testing procedure the panel was asked
to smoke a
control cigarette first, made with the test
tobacco at 100% or
at a level at which its smoking characteristics
were readily
identified. Two 'sandwich' cigarettes were then
smoked, with
the test tobacco in the middle or on the outside,
to identify
which cigarette was closer to the control.
The results suggested that for the
cigarettesincorporating burley
or U.S. flue-cured tobacco, significantly more of
the taste
character of these tobaccos was exhibited when
they were in
the centre of the sandwich (i.e. equivalent to
being in the
core of the rod). For cigarettes with CRS, the
position of
the CRS in the rod did not significantly affect
the amount of
CRS taste.
... /40
N)
1-0
BATCo document for Province of BritiSh Columbia 23
April 1999


-39-
Although the affect on smoke was apparently the
opposite
of that found in Hamburg, the modification of a
cigarette
making machine has been considered to obtain more
reproducible
sandwich-type cigarettes, but other priorities
have precluded
further work.
... /49
J:Ib
C)
rIIj
J:-b
110
Un
BATCo document for Province of BritiSh Columbia 23
April 1999


-40-
T-6810 FLUE-CURED TASTE ENHANCERS
Personnel: H. Roubicek, N. Friedmann, M.P.
Scherbak, T.A. Smith
Objective: To evaluate commercial samples of
enhancers of flue-cured
taste.
Status: Flue-cured taste enhancers have been
received from BAT
(PDQ and flavour suppliers. Some of them have
shown
promise on ventilated filter cigarettes (1-7mg
tar) in
preliminary testing based on injected cigarettes,
but their
acceptability varied with tobacco blend and tar
yield
(cigarette design). The following enhancers have
shown
some potential: -
Cigarette
Supplier Description/Applic'n Level Tar Yield
(mg/cigt.)
BAT PDL 8385 (5mg/cigt) 7
Givaudan Tobacco Flavour SPL-2978 (1.5 mg/cigt) 7
Dragoco Artificial Tobacco Flavour (Mint) 3
0/132100 (25,Ajg/cigt)
Dragoco Tobacco Extract 4-046-014 1
Firmenich 303.113 (Modification #1) (0.35 mg/cigt)
I
Firrenich American Flue-Cured Imitation Tobacco 1
53.435
For futher testing of some of these, and for
evaluation
of other samples received, Matinee Extra Mild
cigarettes
(4 mg tar) will be used as 'vehicle'.
... /41
BATCO document for Province of BritiSh Columbia 23
April 1999


-41-
T-6811 PROJECT GOURMET
personnel: H. Roubicek, N. Friedmann, M.P.
Scherbak, T.A. Smith
Objective: To develop a cigarette with a
recognizable flavour, which
will appeal strongly to a small but significant
segment of
Canadian smokers.
Status: Non flue-cured taste has not been
acceptable on our market,
except for menthol and the part-time co-use of
cigarillos.
Currently one brand (Craven Special Mild) seems to
give an
unusual taste. Several internal studies have
indicated
two added flavours to have promise with normal
flue-cured
cigarette smokers in cigarettes yielding 18 mg and
12 mg
tar. The concept of using such flavours in a lower
tar
product is now being tested externally.
CD
CD
I'D
-4
BATCo document for Province of BritiSh Columbia 23
April 1999


-42-
T-3175 PROJECT VIKING
Coordinator: T.A. Smith
Objective: To develop a king-size cigarette giving
a 9 mg tar and a
relatively high nicotine yield (1 mg or higher) so
as to
give sufficient satisfaction for smokers of
mid-tar range
products.
Status: Without using an SCS filter, the design of
a 9 mg king-size
cigarette needs some filter ventilation. As
background work
for VIKING, we have established the
ventilation-smoke reduction
relationships for a ventilation range of 6 to 40%
using a
series of Malaucene m1cro-perforated tippings and
porous
plugwraps (reported under T-7340). This series
used a
common tobacco rod (JPS) in 510 citrate paper (440
WTU) and
a common filter (2.1/45,000; 3" p.d. for 20 mm). A
group of
samples from this series, giving tar yields of
8-10 mg, was
tested subjectively to see if their different
flow/pressure
drop relationships affected their smoking
characteristics.
All samples provided basically good smoking
"mechanics",
(apparent draw resistance, mouthful of smoke,
etc.),
although differences in acceptability were
evident.
We need to identify the best overall smokino
"mechanics" among
cigarette designs yielding 9 mg of tar
(asiftg.different paperS,
filters-and tip ventilation). However, to obtain a
smoke with
a low tar/nicotine ratio, we need to explore the
smoking
properties of some cigarettes with very unusual
blends. some
of which will demand a significantl y different
cigarette design
to give 9 mg tar with good smoking mechanic?. Two
parallel
studies are starting, as follows:
... /43
0
%~D
co
BATCo document for Province of BritiSh Columbia 23
April 1999


-43-
a) To find which design for a 9 mq 'standard' JPS
blend
cigarette is the most acceptable. (This cigarette
will serve
as a control in later 'rounds' of blend
comparisons). A
matrix of cigarette samples is being made using 3
citrate
papers (260, 440 and 875 WTU), 2 filters (3" and
2.5" for
20 mm), each with 2 appropriate microperforated
tippings.
After normal smoke analyses, subjective smoking
comparisons
will be made of those cigarettes which correctly
yield
9 mg of tar.
b) To explore the smoking properties and tar to
nicotine
ratios given by a number of hiqh-nicotine blends
in
cigarettes of a common design. A series of 7
blends (4 with
PCL and CRS, and 3 as strips only) is being made
into cigar-
ettes using 510 paper (440 WTU). a 3" p.d. filter
and selected
microperforated tippings. Their smoke analyses and
subjective
properties will be compared, to be used as
"sighting shots" for
later series of blend testing for smoking
mechanics and balance
of impact, irritation and taste at 9 mg tar.
... /44
C:)
C)
BATCo document for Province of BritiSh Columbia 23
April 1999


-44-
T-4412 PROJECT LEROY
Coordinator: T.A. Smith
Objective: To develop a lower-tar fine-cut tobacco
product.
Status: Marketing Department identified a need for
a perceptably
lighter fine-cut product relative to the Player's
brand.
In anticipation of possible Government or consumer
reaction
following the launch of such a product, a lighter
fine-cut
tobacco was defined as one giving significantly
less tar than
Player's fine-cut tobacco when tested in
manufactured
cigarettes having the Player's Tubes
specifications and con-
taining tobacco coluims of either equal weight or
equal
pressure drop. The new blend was to be as good as
Player's
fine-cut from a physical stand-point, and was to
cost no more
and hopefully be cheaper. The smoking properties
required of
the product by consumers were unknown (e.g.
'light' v 'heavy'
smoke). Hence a consumer product test was planned
to identify
the consumers' needs using a series of blends.
Laboratory trials identified three tobacco blends,
all less
expensive than Player's fine-cut, that give
significantly
lower tar yields by the above definition, and that
give a
variety of good smokes with a range of tastes. The
small-
scale trials also suggested that these blends were
physically
satisfactory.
.../45
C)
BATCo document for Province of BritiSh Columbia 23
April 1999


-45-
Factory trials were carried out on these three
blends
(using Player's fine-cut as control), each using
two
target levels of stem (WTS, 15 and 20%). the lower
levels being included as a 'hedge' against the
possibility
that the preferred 20% level might cause a problem
in the
physical properties of the blend. Large samples
were
prepared to duplicate satisfactorily the products
chosen
1 n the laboratory trials.
Marketing Department conducted some qualitative
external
group research which concluded that the samples
with lighter
smoke character were preferred by consumers. and
two such
samples are now being examined in an external
product test.
... /46-
C)
N)
BATCo document for Province of British Columbia 23
April 1999


-46-
T-4440 PROJECT BEAUTIFY
Coordinator: T.A. Smith
ObJective: To develop a new cigarette tube for use
with fine-cut products,
to provide its users with lower tar yields
relative to Player's
tubes.
Status: Player's tubes are 80 mm long, 27 mm
circumference, with a
13 mm low efficiency acetate filter (5.0/50,000;
0.8 ins P.D.),
a 15 mm tipping patch and a low porosity wetproof
phosphate
paper (110 WTU), with the latter scheduled to be
increased
in porosity to 140 WTU. A matrix of cigarette
samples was
manufactured by making-machines to tube
specifications using
Player's fine-cut tobacco with 2 more efficient
filters
(3-3/44,000 and 2.1/52,000) and 2 more porous
papers (26P -
210 WTU, and 17C - 320 WTU). This afforded a
number of ways
of obtaining the desired reduction in tar yields.
However, the
use of higher porosity paper was later eliminated
because of
anticipated paper breaking problems in tube
manufacturing. The
use of a longer filter and tipping was not
acceptable because
the machine modifications would be too costly.
From tar and nicotine results and cost
considerations, the 3.3/
44,000 tow was selected for use in the new tube.
In external
qualitative group research run by Marketing
Department, the
modified tube could be distinguished by users from
the normal
Player's tube, and it was well received. No
further experimental
work is foreseen.
CD
LA
BATCo document for Province of BritiSh Columbia 23
April 1999