Investigation of Filter Temperatures and Desorption of Volatiles from Carbon Filters Under Different Smoking Regimes

Investigation of filter temperatures and desorption of volatiles from carbon filters under different smoking regimes.

B. Teillet, X. Cahours and S. Purkis

1 Science & Stewardship 1 Background z Previous work carried out by C. Mueller Poster SSPT21 z Loss of efficiency of carbon filter for the retention of volatiles in the last puffs during Canadian Intense (CI) smoking regime z Higher filter temperature in the last puffs

Î Hypothesis : thermal desorption from the carbon filter might occur during CI and may not in ISO

2 Science & Stewardship 2 From ISO to CI Smoking regimes

z Smoking parameters CI versus ISO

ISO CI Puff volume 35 ml 55 ml

Puff duration 2 sec 2 sec Higher filter load Puff flow rate 1,05 L/min 1,65 L/min Higher flow rate Higher filter temperature Puff frequency 60 sec 30 sec Vent blocking No Yes 100%

Î Experimental design focussed on temperature effects but additional phenomena to be taken into 3 consideration Science & Stewardship 3 ISO and CI Experimental set-up

1. Evaluation of filter temperature

2. Volatiles distribution between filter and vapour phase with the filter temperature effect

3. Volatiles distribution between filter and vapour phase without the filter temperature effect

4. Impact of filter temperature effect on the filter/vapour phase distribution

4 Science & Stewardship 4 Evaluation of filter temperature: experimental

A 7 mg ISO tar commercial cigarette (filter vent 38%) modified by addition of 50 mg of carbon after the filter ventilation holes

Temperature measurement

Experimental conditions Probe in the middle of carbon section (axial and longitudinal) Linear smoking machine equipped with a digital recorder (Sodim) acquisition period 50 msec 3 replicates

5 Science & Stewardship 5 Evaluation of filter temperature: ISO smoking

ISO Smoking regime 60 Slight increase during each puff 50 ure in °C Maximum temperature 27°C in the last puff 40 mperat e t

20 time

6 Science & Stewardship 6 Evaluation of filter temperature: Canadian Intense smoking

50 erature in °C p 40 tem

30 CI Smoking regime High increase during 20 whole smoking period time Maximum temperature:

70°C in the last puffs 7 Science & Stewardship 7 Evaluation of filter temperature: Summary

Differences in filter temperature according to the smoking regimes z ISO z Constant z Slight increase only in the last puff z Maximum temperature Î 27°C z Canadian Intense z High increase during the whole smoking period z Maximum temperature Î 70°C 8 Science & Stewardship 8 Filter/vapour phase volatiles distribution with the filter temperature effect Experimental Filter pad Vapour phase trap 7 mg ISO tar commercial cigarette placed behind the filter pad modified by addition of carbon

Analysis focus on 13 volatiles compounds

VOC : 1,3-butadiene; Isoprene; Acrylonitrile; Benzene. Carbonyls : Acetaldehyde, Propionaldehyde, Acetone, Isobutyraldehyde, Butyraldehyde, Methyl Ethy Ketone (MEK), Crotonaldehyde. Semi-volatiles : Toluene, Styrene.

Method limitation : Acrolein excluded (lack of specificity), Formaldehyde excluded (highly reactive compound), Lower molecular weight compounds excluded (not detectable) 9 Science & Stewardship 9 Filter/vapour phase volatiles distribution with the filter temperature effect Volatiles analysis

Thermal desorption Gas Chromatography Analysis (TDS-GC/MS)

Experimental conditions : TDS-GC/MS Gerstel/Agilent system, RTX VMS column - multi ion monitoring acquisition - Total desorption : 350°C for 5 min - 3 replicates

Acetaldehyde Isoprene Acetone

Propionaldehyde 1,3-butadiene

10 Science & Stewardship 10 Filter/vapour phase volatiles distribution with the filter temperature effect Results expression

Thermal desorption Gas Chromatography Analysis (TDS-GC/MS)

filter vapour phase Results : Area response for each compound 100% 90% from the total desorption of the filter or the vapour phase trap

% 80% e 70% nc Expression : for each compound and each 60% replicate, the relative distribution filter/vapour unda b 50% phase was calculated a e 40% v i t 100% corresponding to the sum of filter + a 30% l e r 20% vapour phase allowing comparison of different 10% smoking regimes. 0% compound

11 Science & Stewardship 11 Filter/vapour phase volatiles distribution with the filter temperature effect ISO smoking regime

FILTER VAPOUR PHASE 100% 90% 80%

e % 70% c n

a 60% d n

u 50% 40% 30%

relative ab 20% 10% 0%

e e e K e e e e n ne de rile E iene re yd o y yde t d p h h M hyd luen a o eh e nzen e o t ldehyd s d de ld e Styren ta I al Acet ral ra B T e on y y Acryloni onald ut t 1,3 Bu Ac pi But ro ob P s Cro 12 Highest I compounds Science & StewardshipLowest 12 Volatility Volatility Filter/vapour phase volatiles distribution with the filter temperature effect CI smoking regime

FILTER VAPOUR PHASE 100% 90% 80%

e % 70% 60%

undanc 50% b

a 40% tive

a 30% l

re 20% 10% 0%

e e e e e e e e e n n d l n de n n yd yd one yde y y e ie h re h t h MEK ze h u re d p e e n e l ty ta ld ld o S u Iso a Ace Be T talde n B e o tyraldeh tyra Acrylonitri ,3 c i u 1 A p b Bu o Crotonald Pro Is 13 Highest compounds Lowest 13 Volatility Science & StewardshipVolatility Filter/vapour phase volatiles distribution without the filter temperature effect Experimental

An 50 mg external carbon Same 7 mg ISO tar commercial filter placed between the cigarette modified by addition of filter pad assembly and the 50 mg of inert semolina in order entrance of the smoking to maintain the pressure drop machine syringe during the smoking (temperature checked : room temperature) vapour phase trap behind the external filter

Filter pad Analysis of volatiles 14 Science & Stewardship 14 Filter/vapour phase volatiles distribution with the filter temperature effect ISO smoking regime

FILTER VAPOUR PHASE 100% 90% 80% 70% 60%

undance % 50% b

a 40% 30%

relative 20% 10% 0%

e e e e e e e e e e n n d d l K n de n ne e yd e yd on y y tri E y e i h r h h h i M ze u re d p e et e n n l ty ta de de o e o u l so ldeh T S I Ac ral B a B eta y cryl n c t A to ,3 pionald o 1 A o bu Butyrald so Cr 15 Pr I compounds Science & Stewardship 15 Filter/vapour phase volatiles distribution without the filter temperature effect CI smoking regime

FILTER VAPOUR PHASE 100% 90% 80% %

e 70% c n 60% 50% 40%

lative abunda 30% e r 20% 10% 0%

e e e e e e e ne ril ne ne en yd on it yde e en hyd re h hyd n MEK r di p et o eh lu y ta lde o lde c de l d St a A al y Benze al To t Is na r Bu Acr on ce io uty 1,3 A op Butyraldehyde rot 16 sob C 16 Pr I compounds Science & Stewardship ISO smoking with/without filter temperature effect: summary

FILTER VAPOUR PHASE 100% 90% 80% % 70% No major difference among nce 60% volatiles distribution between 50%

abunda 40% filter and vapour phase whit or 30%

relative without temperature effect in 20% 10% ISO. 0%

e e e e FILTERe VAPOUR PHASE d ne il K ene e yd on yd yde tr E i hy h t h i M zene Temperature profile showed low 100% d e pr eh n n luene ta ld o ld lde lo Styrene u Ace a a y Be To Is nalde r r naldehyde 90 B% ty ty Acr o 3 io u u temperature during the whole 1, Aceta op b B 80% r o Crot P Is compounds % 70% smoking period. nce 60% 50% bunda a

e 40% 30% lativ

e Î No temperature effect on r 20% 10% carbon filter efficiency in ISO. 0%

e e e e e e d ne de ne d ne e y o yd EK yd e dien hy pr h h nitril M zene h lu a e o cet ehy e lo n ld s A d e To Styrene a I rald ral B y y cry 3 But cet ionalde A tonalde 17 1, A p but o o But ro s C 17 Pr I compounds Science & Stewardship CI smoking with/without filter temperature effect: summary

With FILTER VAPOUR PHASE temperature10 0% Slight differences between the effect 90% maximum filter relative abundance 80% % 70% (Cambridge filter pad in front of the ance 60% external filter : filter load difference) 50% bund 40% ve a 30% Huge differences between the filter lati e r 20% relative abundance for the highest 10% 0% volatility compounds.

e e e e e e e e n d ne d n de il K d ne e y FILTtrER enVAPOyUR PHe ASE en hy re hy to ME e p e eh nz eh tyr Without tadi o ce e 100%u ld ld ldehyde ld ld S ta Is A ra ra ryloni B Tolu Temperature profile showed high temperature e c 3 B iona ty ty A tona 1,90% Ac Bu ro effect op Pr C temperature during the whole 80% Isobu compounds % 70% smoking period. 60% 50% Î Temperature effect on carbon 40% ive abundance t 30% filter efficiency suspected in CI. ela r 20% 10% Further investigations : Desorption 0% estimation by TDS-GC/MS

e e e e ne n de ne de de il ne de ne ie yd e y to y y itr re h e h MEK ze uen y ad eh opr e c eh lon n ehy ol t ld s d A de ld y d T St ta I al al ra r Be 18 n yr y 3 Bu ce o Ac onal , pi ut ut 18 1 A o B rot r ob C P Is compounds Science & Stewardship Filter temperature effect on volatiles distribution: CI experimental

7 mg ISO tar commercial cigarette modified by addition external carbon filter of semolina Vapour phase trap

1 TDS-GC/MS Analysis Analysis in 2 steps

thermal desorption until maximum occurring temperature 60°C/min up to 70°C during 6 sec

full desorption 19 350°C 5 min Science & Stewardship 19 Impact of filter temperature effect on the filter/vapour phase distribution CI desorption simulation

desorption loss FILTER DESORPTION VAPOUR PHASE 100% 90% 80% Negligible 70% <10% <10% 60% ndance % 50% 15% 40% ve abu ti

a 30% 25% 25% rel 20% 10% 45% 0%

e e e e e e le e e e n de ri n ien yd on t yde en hyd h et hy MEK lue yr ad c ehyd nzen eh lde sopre lde ld de yloni d To St ta I A ra al Be al But e na y yr 3 io Acr on 20 1, Ac ut ut ot op ob B 20 s Cr Pr I compounds Science & Stewardship CI smoking with filter temperature effect /TDS simulation : summary

With FILTER VAPOUR PHASE temperature10 0% effect 90% 80% Relative abundance of volatiles in % 70% the filter after desorption

ance 60% 50% simulation in a good agreement bund 40%

ve a with CI including temperature 30% lati e r 20% effect 10% 0%

e e e e e e e e n d ne d n de il K d ne e y tr en y e en hy re hyFILTERto DESORPTIONME VAPOUR PHASE e p e eh nz eh tyr tadi o ce e u ld ld ldehyde ld ld S 100% ta Is A ra ra ryloni B Tolu Hypothesis : e c 3 B iona ty ty A tona 1, Ac op Bu ro 90% C Pr Isobu compounds 80% Thermal desorption of volatiles %

e 70%

nc from the filter to the vapour 60% da 50%

bun phase due to temperature effect a

e 40% v i t a

l 30%

e Higher temperature involved r 20% 10% thermal desorption for the 0% highest volatility compounds

e e e e ne de de ne de il K ne de n e y y o yd tr e y i h rene t h hy i ME z uen e eh e e n eh l yre ad d op d de d lo d o ut s l Ac y St al I a ral ral r Ben al T et n 21 ,3 B c ion ty Ac to 1 p bu Buty o A o Cr 21 Pro Is compounds Science & Stewardship Investigation of filter temperatures and desorption of volatiles from carbon filters under different smoking regimes. Summary

Canadian intense regime involves higher temperatures than the ISO

In comparison with ISO an additional phenomena occurs related to temperature effect. It could drastically reduce carbon filtration power for the highest volatility compounds.

Thermal desorption estimation shows reduction in carbon filter efficiency depending of the compounds from 45% to more than less 10%.

Hypothetically, it might be a combination of both desorption and loss of adsorption due to filter temperature.

Blocking 100% filter ventilation does not allow the smoke cooling effect during the Canadian intense smoking period. However, the Canadian regime may not well represent human smoking conditions. 22 Science & Stewardship 22 Thank you for your attention

23 Science & Stewardship 23