PRODUCT DEVELOPMENT LABORATORY, SC)UTHAMPTON FttiE fC" L) tt i Memorandum NQ 77-12-047 Subject Objective Examination of the Taint Problem Associated with du Maurier Cigarettes by Headspace Analysis. 9 TC N. Foster, Esq. arculation Dr. C. 1. Ayres Mr. D. J. Wood Dr. R. G. Hook File 85 Date 13th December, 1977. CONFIDENTIAL -NOT TO BE COPIED OR SHOWN TO UNAUTHORISED PERSOP~g Q.J1J C)N (_M cz::> BATCo document for Legal Services : Health Canada 20 October 1999 JDG/MAB/BMB/85 13th December, 1977. OBJECTIVE EXAMINATION OF THE TAINT PROBLEM ASSOCIATED WITH du MAURIER CIGARETTES BY HEADSPACE ANALYSIS SUMMARY Recently a taint problem arose with du Maurier cigarettes in Canada where subjective testing traced the cause of the problem to odour transfer from packaging to the cigarettes. The,undesirable taint was apparent only on smoking the cigarettes and the taint was significantly reduced on using modified packaging materials. At that time no objective tests were found to be successful in tracing the cause of the taint. This note details the results of applying a novel objective approach to investigating the problem. The approach, known as headspace analysis, uses g.l.c. to analyse the vapour phase of a chamber under equilibrium conditions, which contains a component of the product under investigation, e.g. a cigarette carton. The results of the investigation confirm that measurable quantities of a number of components had migrated to the tainted cigarette. The use of the technique has provided valuable circumstantial evidence to support the subjective findings. It is recommended that the headspace technique should continue to be used, in conjunction with subjective testing techniques currently being established, firstly to investigate taint problems, and secondly as a product development tool to help avoid taint problems in new products. Throughout the following discussion frequent reference is made to CD Uri the chromatograms, or aromagrams, shown in the figures and for convenience~ (-r-J these are listed in Table 1. C-1 0- C-n BATCo document for Legal Services : Health Canada 20 October 1999 - 2 - DISCUSSION The aromagram of a single cigarette from a pack known to be strongly tainted (Fig. 1) has been compared to that of a control untainted cigarette which had not been in contact with packaging card (Fig. 2). Differences are apparent in that the aromagram of the tainted cigarette (Fig. 1) has a number of peaks not present in the control (Fig. 2). The inferences that may be made are :- M the differences in,the aromagrams are directly related to the taint in as much as one or more of the compounds giving rise to the additional peaks are responsible for the taint, or (ii) the differences in the aromagrams are indirectly related to the taint in as much as the taint is apparent when the additional peak-s are present, although the compounds giving rise to the peaks may them- selves not be the cause of the taint. Whichever inference is correct will not affect the followinv argument. For convenience the additional peaks in Figure I compared to Figure 2 will be referred to as "contaminant" peaks. An aromagram, of an empty packet of the strongly tainted cigarettes (Fig. 3) showed a remarkable similarity to that of the contaminated cigarette itself (Fig. 1). Furthermore a sample of the packaging cardboard, printed, varnished and ready to be assembled into packets was judged to be malodourous and gave an aromagram (Fig. 4) which is similar to that of the contaminated cigarette (Fig. 1). These findings focussed attention on the packaging material as being the cause of the cigarette contamination. A sample of the base card, unprinted and unvarnished (ie. card having a foil rendering plus lacquering) which had a faint off-odour, subjectively different from that of the printed and varnished card, gave an aromagram QJ-J 0-1% ON (-n r\,;. BATCo document for Legal Services : Health Canada 20 October 1999 - 3 - (Fig. 5) whose peak profile differed both qualitatively and quantitively from that of the varnished card (Fig. 4). The contributior fr.om the unvarnished card to the cigarette taint would therefore ap~-zar to be minimal and contamination arises from the subsequent print:--g and/or varnishing stages. Two samples of dried varnish (Trade names:- Bowers an-i Colobec) have been examined. The aromagram of the Bowers varnish (Fig. 6) shows distinct similarities to the aromagrams of a) the tainted cigarette (Fig. 1) b) the packet which contained the tainted cigarette-z (Fig. 3) c) the card taken prior to packet assembly (Fig. 4). As the'bontaminant" peaks are evident in the aromagr=- of the Bowers varnish (Fig. 6) it is reasonable to conclude that -:: is this varnish which has been the cause of the "contamination". The aromagram of the Colobec varnish (Fig. 7) shows S=e of the "contaminant" peaks but it is quite different from that of the Bowers (Fig. 6) varnish. A Smoke Panel Test at G.R. & D.C., (No. 1548) was caTried out using tainted and untainted cigarettes in conjunction with the c:rresponding headspace analyses. Only minor differences between the tw:, cigarettes were noted by the panel, however, the corresponding aromag-rams (Figs. 8 and 9) were quite different. The "contaminant" peaks are clearly evident in the sample said to be tainted (Fig. 8) but are essentially absent from the untainted cigarettes (Fig. 9). A second s*_bjective test carried out subsequently on a larger sample of smokers, ap?ears to be in agreement with the results of the objective test detailed above (Note of R. P. Ferris). CrI% C7-N (_TI U- BATCo document for Legal Services: Health Canada 20 October 1999 - 4 - Soon after the taint problem arose it was noted by a subjective test carried out by I.T. Co. (Canada) that the foil between the cigarettes and the packet forms a barrier to the passage of those compounds responsible for the taint. This has been confirmed by headspace analyses of selected samples. The "contaminant" peaks are not evident in the aromagram. of a cigarette taken from a pack with an inner foil (Fig. 9), although the empty packet clearly shows the "contaminant" peaks at an unusually high concentration (Fig. 10). The high concentration is most probably a direct result of the foil preventing migration of the contaminants to the cigarettes. Examination of a pack in which no inner foil was present (ie. in which the cigarettes and packet were in direct contact) shows that the "contaminant" peaks are present in both the cigarette aromagram (Fig. 8) and the packet aromagram (Fig. 11). As a back-up to this work an unrelated cigarette and the card used in the packets, with which no taint problems are apparent, have been examined using the headspace analysis technique. No evidence of the "contaminant" peaks was found in either the aromagram of a cigarette (Fig. 12) or in the aromagram of the card (Fig. 13). Furthermore there appears to be only minor transfer to the cigarette of those volatiles apparent in the headspace of the card. It is suggested that the approach developed during this investigation should be used in conjunction with subjective studies of new or modified cigarette packagings to assess possible taint problems. The detection of contaminants in the headspace of cigarettes, identifiable with a particular component of the packaging would warn of potential taint perception. U1 ON c7N QJ-1 _rz_ BATCo document for Legal Services : Health Canada 20 October 1999 - 5 - In due course, the method might be applicable in a more routine way, to "finger printing" samples from incoming deliveries of materials for established brands. These would be compared with "specified aromagram levels" found by experiment to be safely below the threshold level for taint perception. CONCLUSIONS Headspace analysis provides an objective method of detecting the source of taint problems in favourable cases. In the present examination the "contamination" was traced to the varnish used in the packaging. The results of the headspace analysis are in agreement with subjective testing of corresponding samples. 4- CD LTI LYI BATCo document for Legal Services : Health Canada 20 October 1999 I - 6 - TABLE I Details of the Samples Examined by Headspace Analysis. All figures are in the form of aromagrams:- Fig. 1. A strongly tainted du Maurier cigarette. Fig. 2. An untainted (control) du Maurier cigarette. Fig. 3. Empty cigarette pa cket which contained the strongly tainted samples. Fig. 4. The printed and varnished card from which the du Maurier cigarette packets were made. Fig. 5. The unvarnished base card. Fig. 6. Sample of Bowers varnish (hardened). Fig. 7. Sample of Colobec varnish (hardened). Fig. 8. A tainted du Maurier cigarette of the type examined by the Smoke Panel at G.R. & D.C. Fig. 9. An untainted du Maurier cigarette of the type examined by the Smoke Panel at G.R. & D.C. Fig. 10. The empty packet which had contained the untainted cigarettes of Fig. 9. Fig. 11. The empty packet which had contained the tainted cigarettes of Fig. 8. Fig. 12. Benson & Hedges cigarette used as a comparison. Fig. 13. Card of the type used on Benson & Hedges cigarettes of Fig. 12. L-n LX-j C71- 0-1 BATCo document for Legal Services: Health Canada 20 October 1999 I ;, LI La L2 .... .... .... ... .. ..... .... .... .... C2. 93 : ;* ..... - - - -,:. to ... .. .... .. : - '"" , - '- I I II I I . I . - ! w .... .... ... ... ........ .... F .... .... .. I I I - A STNC-my rA,"Tf.X C44ARET-rf- .. ri IF2. Iti 9 -0 : i - i 9z CD CD EJ _V, 1 1 -571 T, I t:l BATCo document for Legal Services : Health Canada 20 October 1999 N ILJ ~kx I - . . I i 93 =3 so a . 1 10 1 ... .... .... L C3 9'2 4 ; , - i i , . .1 ca: ........... A Ct~w7jtvL cU MAVRIAJZ CICARIETTE: Ij L CCMTAtT W-M 711E 'PACWACLN13. i-3 1 A 40 17 co BATCo document for Legal Services : Health Canada 20 October 1999 I 1 1 In wj CD L2 LJ Lj I . .... .... .... i : '" "' 'I *' , 1 1 . . Kj . . ... ........... : J- I' I Im .... . . .... ... :: 11 ... 4 :. " ": - :, ,, - i . i ma I L2 r2 ca am! a THE Emp-ry -pAC TA'JImigj, C3 1:2 C3 92 cz : : - " r 7 'S le r : I i : I ~ . -J- 1 1 1 1 1 1 1 . i 1 : lu. ~_n U-4 cy- ON f-n 1110 BATCo, document for Legal Services : Health Canada 20 October 1999 71 r3 iri 00: . . ......... an t 4 03 VAKNIS.,tb eAltj, W64 K Er 6, 't"E- Csei^itr_.r, WERF MA~,.E. ILA i 0-N UN ON BATCo document for Legal Services - Health Canada 20 October 1999 THE V04VARMIS.I.C.1 7,A fit C^fta~ SrART OIN ON BATCo document for Legal Services : Health Canada 20 October 1999 . I ri L3 L.1 . . ... .... .. du Li '90MEItSVA21.46j4 Lj. "Aft 3~ZPJF3) sm OW Im CD~ 01, rII-) BATCo.document for Legal Services : Health Canada 20 October 1999 i C2 C3 aj . I ri. ro .... .. .... LI I I M2 I ;r 4- fICUR F- CoLD isz. C I A&ZEAMI) -pr G~% 0- G-N U-Ij BATCo document for Legal Services : Health Canada 20 October 1999 w: I I . Lj I I , '! . . . . CA L21 I ..; . .. : . . I I I I !- , I - ! I CA I - . ., -4 ":- -. - ! I I lu C~ I ; I I I NJ 0. . I- !-[ , ; . ~ I - .. . - Mi t.~j ... . . i A 'rA'N'TE-u d~ MAORJES IM : I - CtCARIT-r& A IM -% EIA14 IOJEb Ci "'ALML A-r 0.9 &.L,4(-. ILI La Tj Url . OIN C7'\ BATCo document for Legal Services : Health Canada 20 October 1999 LW I . I i I I ~ i I ILJ a ]p a i *1 0 4 W ~ r~ t Cj irs 0 1 1 . , I i in C2 CS $ L3 1 92 I L I IL: I I : CI: FICulte R -9 1 AU UJTA.77z:~.j IMAURIEft L~ C%CAIZE77E A, E*A-I-E3~ &~ T"IE 9T Ar 12. U .35. C. C Lj t i BATCo document for Legal Services : Health Canada 20 October 1999 r2 W~j lu jrj : . - -., I ; I : I ar I i - I ILJ . :i - . 1 02 L2: Ila 8:1 ra I . ! I LM I I , I , I F- 10 , i - : ::I TOM EMS"TY TII;C.F.-r ",mw Raj~ M 740- UMTAWTE-3, WA Lj ILI' L! a -rA.ZT SrAIZT un QI-j 0- ON ON BATCo document for Legal Services : Health Canada 20 October 1999 r, I - - . . . : : CM : I ; I I I , 1 1 - I ; W.: 0 ; . . . . - . . ! I . ; . : I-- 1 1 . . , -r HE -0 TAC.r - "JNLCM ;4A-') ~T CO"TPI-CZ THE TA,WE T- o CICAPtE ~rc cF Fi~;. a ON 01% BATCo document for Legal Services : Health Canada 20 October 1999 rj: : La ., . I ; I - I IU I IOU 10. NCR J 10 V3 on' Kj: ILJ 83 so En . .... ... ... VJ11 I : ! Ica. F. ... ... ... ... .A FI C. VRF I Z ---, : I -T. C-4 A SE~S*V A 21 Wfb~&S V"~ A~ A ..i . . . . I - it to ~L 0- O-N CYN 00 BATCo document for Legal Services : Health Canada 20 October 1999, j .......... L2 .... .... .... .... .... . I . i * i ; ILJ .... .... ... I Lj .0 - , , -, I : I - lu ... .... ... C.AftZ I : , . . .. . CW4 ", 4 F- 2.C, a -!6 : Z! 0- 01- %IC BATCo document for Legal Services : Health Canada 20 October 1999