Tobacco Documents Online

/ RJR Interoffice Memorandum TO: Dr. R. L. Suber FROM: T. A. Perfetti DATE: July 9, 1990 SUBJECT: Request for Review and A Additives in STT Filters Background: proval of Tobacco Based Flavors For Use as The Flavor Division of P&AT has undertaken a program to develop and qualify unique flavor systems for cigarette products. Two of the flavor delivery systems that will be fully evaluated and qualified in 1990-91 are: reaction flavor for tobacco products (RFTP) and tobacco based flavors (TBF). Both RFTP and TBF have a common flavor mechanism. That mechanism is the reaction of sugars and an amino acid source to produce Maillard reaction products. In RFTP, the amino acids and sugars are additives, exogenous to tobacco. These amino acids and sugars are reacted and become part of the casing formulation. In the TBF system, tobacco, normally in the form of a spray-dried tobacco extract, which is rich in sugars and free amino acids, becomes the starting material for the Maillard reaction products. In TBF, tobacco is, itself, the endogenous source of the starting materials for the Maillard reaction products. RFTP and TBF materials improve tobacco taste characteristics of smoke, give a more authentic tobacco pack/rod aroma and often reduce harshness and improve the richness of tobacco smoke flavor. The materials for review in this memorandum are a series of seven tobacco based flavors (TBFs). The request is for use of these materials as filter additives in finished cigarette products, up to a level of 25-30% dry weight of filter (which equals to 25-30 mg per dry cigarette weight. We request approval for internal (SED testing) and external (consumer) evaluations of cigarette products, up to a carton quantity. Summary: Tobacco based flavors (TBFs) are prepared by the extraction of ground tobacco with water and formation of a spray-dried powder. The spray-dried powder (175 g in all cases) are combined with an amino acid (50 g in all cases) or reacted alone at 180°C in an aqueous medium under pressure (normally in a Parr-type bomb). The reaction time is normally 30 minutes. The amino acids used to 'date for TBFs are: aspartic acid, glutamic acid, glutamine, asparagine, proline, alanine and phenylalanine. Additionally, the case where no amino acid is used, which is itself a TBF, is being asked for approval.

-2- The TBFs are then applied to a polypropylene web material and made into filter rods. The TBF filter rods are combined with cellulose acetate filter rods to produce the so called STT (smooth tobacco taste) combined filter. The dual filter has a tobacco rod end of polypropylene TBF (15mm in length) and a mouth end segment of cellulose acetate tow. The filter element contains 25-30% TBF in the polypropylene segment. The quantity of TBF in the entire element is 25-30 mg (See Table 1). The filters with TBFs are called TBF/STT filters. A. Ames (Neat) Neat TBF flavors were previously submitted for Ames tests. All TBF flavors were negative and non-toxic, with and without 5% S-9 metabolic-activation. The eight TBF flavors were approved as top dressings under HRRC Proposal No. R9016 (Attachment I). B. Urethane All eight TBFs were previously submitted for urethane analyses. All samples showed levels of urethane at less than a minimum detectable quantity (MDQ) (Attachment II). C. FTC Analvses Nine cigarettes were prepared. Seven employed TBF/STT filters. One was a STT filter with no TBF, an internal control. The ninth product was WINSTON :;S (WKS). The TBF/STT filters are 27mm in length. The WKS employs a 21mm segment of cellulose acetate (CA) 3.3/39,000. The cellulose acetate segment of the TBF/STT filters were made of 3.3/39,000 CA tow. The TBF/STT WKS configuration has 57mm rod and a 27mm filter element. The WKS control has a 63mm rod and a 21mm filter element. A full comparison of all the special smoke analyses were made relative to the internal control STT filter product. Table 2 lists t'r:~ results of all tobacco and smoke analyses completed for the TBF/STT WKS prototypes. All smoke analyses were run on products at zero percent filter ventilation. There were no significant differences in any of the analyses for the test products CT 2057 A-G versus CT 2057 H (the internal control). Changes in the actual deliveries of smoke components were all very small and were within the normal error of each method. Attachment III is a full listing of the Product Specifications for CT 2057 and CT 2056. D. Ames (Smoke) All products were submitted for collection of cigarette smoke condensate (CSC) at zero % filter ventilation. All samples gave Ames activity of CSC non- statistically different from the internal control AN 8200 AH or the WKS control AN 8200 AL. See Attachment IV.

V -3- Summary: Table 3 is a summary of the analytical data compiled for this review. Thank you for reviewing the data. Samples for sensory and consumer evaluation are currently being prepared for testing. I would like to go to the HRRC in as short time as possible. /trr xc: Dr. B. M. Lawrence Mr. L. A. Lyerly Dr. M. E. Stowe