Tobacco Documents Online

, ., ..., fi' . PHILIP MO~RRIS Ul. S. A. 111TER-OFF 1 CE COR'IRESPOHDEHCE RICHMOND, VIRGINIA CONFIDENTIAL To: . Mr. W. F. Kuhn Date: Apri 1 28, 1983 From: • M. A. Jel tema subl'eCt= • Chocolate , Attached are two decoded memos concerning~our use of chocolate in tobacco products.- Our largest use is in the form of cocoa shells Oul.5%). Shells are often used as a cheaper material for extraction of cocoa flavor, since during roasting some of the.flavorants and cocoa butter transfer to~the shells. It is not known if we receive the extracted or unextracted shelils or if the shells provide any other functionality. Bob Ikeda plans to 1) extract the shell~s to see if they contain flavor material~s and 2), research whether tobacco dust could be used instead. 'A cocoa substitute does exist for the shells. This substitute is carob.which is the most common food substitute for chocolate. Block chocolate is used in burley spray. They do have a substitute for this which is most likely based on a commercial chocolate with additives. . The aftercut is an extract of cocoa, probably based from steam. 'distillation. The substitute also has some cocoa. There are probably commercial formulations which could be used. The volatiles have been looked at by G.C. (C. Kro!ustalis). There are approximately 15 cpds. which he saw. Most of these are minor. The major peak (ti85%) is vanillin. With the identificationof the minor components, it is possible that a reformulation could be made. It is not clear that the aftercut is needed, since in our panel data (sniffing aftercuts), the panel could not identify the odd sample until 85-1009 had been altered. Bob Ikeda is taking this further by looking into whether chocolate in any_of the additives is necessary. He plans to evaluate a product containing no cocoa products against a control. As a second step, he plans on, evaluating a product containing no additives (casing, flavors, humectants). If substitutes are used for chocolate, Bob feels these should include only vol.atile cpds. which would distill over intact, avoiding pyrolysis problems of the nonvolatiles. Neither Bob or any other person in Flavor Development has evaluated the vollatiles of chocolate or the effects of cocoa butter. I was told 0 that Daylor di~scourages the identification of compounds used in imitation- F+ formulations to keep the trust of the flavor house. %a .'

Mr. W. F. Kuhn 2 April 28, 1983 Some research into cocoa and cocoa butter is reported iin the literature other than pure-identification of compounds. Leffingwel'1, formerly of R. J. Reynolds has published general organbleptic properties of groups of compounds as well! as specific cpds. He found that cocoa butter acted to smooth and enhance tobacco fl~avor by entrapment of volatiles. No flavor was detected from,cocoa butter itself although they state that other cocoa butter substitutes didn't show the same effects as cocoa butter. In research on individual compounds he found that 2-Phenyl-2-buten-lal had'an enhanced chocolate character in smoke. The aroma of steam distillate compounds was researched by M..Praag. He found the basic fraction containing all the pyrazine cpds. to have a nutty aroma. Others have stated'ithat pyrazines provide nutty, roasted, chocolate aromas. The neutral fraction was found to have a cocoa aroma which disappeWred' on holdinq.,!blggesting highly voliatile cpds. were important. " One compound 5-methyl-2-phenyl-2-hexenal was found to individually possess a strong cocoa aroma. No other single compound' had a cocoa aroma, although the sum did. This is often true of flavor mixtures, which makes the task of.flavor formulation very diffi,cult. There are lists of suggested compounds to be used' in formulations of imitation chocolate, however none of these are thought to individually possess a chocolate aroma. The formulat:ion of one"chocolate will have a different aroma from another formulation based on the favorists perception of chocolate. - The question - what makes chocolate smell like chocolate is the same as what makes burley taste like burley. Once the compounds have been identified it is a difficult task to:determine which few can be combined to produce a given total perception. The optimization of given compounds in a mixture is a much simpler task. . ~. MAJ/bpp ~ ~' cc: Dr. W. A. Farone ,I ~ _ . N tV ~ Q PA Ct