Tobacco Documents Online

CONFIDENTIAL PROJECT 2305 FLAVOR DEVELOPMENT PERIOD COVERED BY REPORT: August 1967 - August 1968 DATE OF REPORT: August 28, 1968 PROJECT LEADER: F. L. Daylor, Jr. The objective of the Flavor Development Group is to supply the flavor and fragrance support for the consumer product development activities. The group's activities include formula- tion, application, and evaluation. Only a relatively small part of the group's work is charged to 2305. The major part of the work is directeditoward cigaret and other development projects. The objective for Project 2305 specifically is to explore, develop, and evaluate new flavors for tobacco, confectionary, and toiletries products. The following is a report of the Flavor Development Group's Work. 1. CIGARET FLAVORS A. Parliament with Plastic Dilution Tip No off flavor due to the plastic had been found and work had been in progress to develop flavors for the 100mm Parliament with the plastic dilution tip. Models delivering 14mg tar and 18mg tar had been developed. The obvious flavor defects were the thin body and the hot andidry effects. The hot and'dry effects had also been noted in earlier work with other air dilution candidates. The earlier work hadpresented several leads on desirable flavor additives for this type cigaret. Many flavors were screened by injection into cigarets made with regular Parliament filler with the normal flavors and the appropriate filter dilution system~. The promising candidates were studied further by overspraying the additives on Parliament filler and making cigarets from the treated filler. After a series of panel evaluations and other evaluations

K by Projiect and Manufacturing Department personnel, flavors 38'86-89 + 90 and 3728-9'6-A were selected as the most promising candidates for the 100mm Parliament with the plastic dilution tip delivering 18mg tar. Both~flavors were applied to 100 pound quantities of filler and cigarets were made in Tobacco Processing. A control cigaret was also made and it delivered the target tar, unfortunately both flavored sampl'es delivered slightly less than the target level of tar. Booth testing prior to the mailout test was favorable and the tests were released. Both flavor candidates were tested against the control and the flavor candidates were tested~against each other. The report of the test results stated that the panelists could~not distinguish between the cigarets on any of the three compari.sons. Prior to the completion ofthe mailout test, the 100mm fluted-diluted Parliaments were introduced into test markets in Albany and Providence. At the same time, 80 and 85mmifluted- diluted Parliaments were also introduced into test markets in Albany and San Antonio replacing the regular Parliament product. The 80 and 85mm fluted-diluted products were soon withdrawn~and replaced by the regular product. The 100mm fluted-diluted product continued inthe Albany and Providence markets. The early test marketing of the 80 and 85mm products left the flavor formulation work for these products incomplete. However, flavors were developed and evaluated through POL booth testing for the following products: 85mm Parliament 80mm Parliament 14mg Tar 14mg Tar 12mg Tar 11mg Tar N Flavors were also developed for the 100mmifluted-diluted Parlia- ment delivering 14mg tar. These flavor candidates were also evaluated through booth testing at POL. x Gn CO GT N 2

B. North Pole For about five years, menthol filters have been~the fastest growing segment of the cigaret market. However, Philip Morris has not captured the same proportionate share of the menthol market as they have with regular filters; this is the reason for the North Pole project. The objective for the flavor development work for the North Pole cigaret was to develop a mentholated cigaret with flavor characteristics similar to Kool. C. I. data showed higher nicotine delivery and lower menthol delivery for Kool than Philip Morris menthol brands. Jointly with Manufacturing Department personnel and Mr. Mason, the Marlboro Green blend. was modified to increase the nicotine delivery of the cigaret. On the first attempt, the nicotine delivery was increased to a level beyond that of Kool, however, the menthol level was far too low. At the same time, flavor formulation studies were in progress. The studies identified some flavor additives besides the higher nicotine delivery blends that gave similar effects in smoke to tne flavor of Kool. A series of five samples of 100mm mentholated cigarets were made. The samples included an equilibrated~ menthol cigaret and flavors suggested by both R & D and Manuufac- turing Department personnel. From this series of five, two (3909 and 3912) were selected for large scale evaluation. 3909 and 3912 were tested against each other in a POL mailout test. The results of the test were not clear cut, however, a slight lead for 39'12 over 3909 was indicated. Both 3909' and 3912 (100mm cigarets) were tested against 85mm Kools in separate HTI tests. The results were very favorable for both North~Pole candidates, so much~so that the results were attributed more to the 100mm length rather than~to the flavors. The results of this series Un Cz C11 . CJ 3

of tests involving 100mm North Pole candidates indicated that 3912 enjoyed a very slight preference over 3909 and.was more similar to Kool than 3909. Hence, 3912 was remade as an 85mm cigaret for testing against Kool in another HTI test. For 3912, a Benson and Hedges type blend is used with a modified Marlboro Green flavor along with flavor 3871-1-4. While the HTI test of the 85=3912 was in progress, several new higher alkaloid~blends were evaluated and none of them showed any major advantage over the 3912. At the same time, additional flavor candidates were considered on both the Benson and Hedges type filler and the high alkaloid blends; none of these showed any major advantage over the 3912. The 85mm 3912 was a definite winner over Kool on the HTI test. The 85mm model delivered about 19mg tar. The 100mm models had delivered about 21mg tar. There was interest in a: diluted 100t=North~Pole that would deliver about 15mg tar. Samples of the rods from each of the five 100mm candidates were hand assembled with an appropriate filter dilution system to reduce the delivery to 15mg tar. From this experiment, sample 3911 was selected as the most promising flavor candidate of the five readily available. Flavor formulation work did not quickly produce a better candidate. Hence, a diluted 3911 which is a Benson and Hedges type rod with modified'Marlboro Green flavors with a fig additive is considered the best candidate for this purpose. i By this time the name Lido had been~selected for the brand. It was recognized that 3912 was a "mainstream" menthol filter cigaret. The question arose whether the Lido could.be given a mark of distinction. There were several novel flavors under N development. The one of these that had drawn most favorable 0 comment was the orange flavor. Hence, an orange modification of 3912 was also developed. The orange modification of 3912 is being evaluatedin an HTI test against 3912. Gn U G1 -.'~ 4

C. Virginia Slims The objective of the flavor development work in the develop- ment of the Slim~cigaret was to be evaluation. The first samples of Slims to come to our attention were made with Benson and Hedges filler and were found to be harsher than Benson and Hedges 100's by the Descriptive Panel. With the product aimed for the female market, the rationale was that the product should be mild and certainly no harsher than the Benson and Hedges 100. It would facilitate matters if the filler selected were one of the regular blends rather than a:special blend. Consequently, the Commander blend was evaluated. Subjective response to the cigaret was favorable but the delivery was higher. The sweetness of the Commander was the desirable flavor feature. It was recognized that the Slim would have to compete with Silva Thins both in flavor and in subjective response. To more fully evaluate the available blends, a series of samples were made in the factory using Marlboro, Benson and Hedges, and Command'er filler at two RTD levels. Flavor evaluations of this series indicated that Marlboro filler gave similar response to the Benson and Hedges filler and.the choice to be made was between these two and the Commander. Evaluations with Manufac- turing Department personnel showed that at the 5.0" RTD level the Commander rod gave the smoother smoke. Both the Commander and the Benson and'Hedges Slims were made at the 5.0 and the 5.5 RTD levels. At the lower RTD the sweetness of the Commander was - preferable but at the higher RTD the Bensonand Hedges and Commander rods delivered comparable flavors. At the higher RTD the Benson and Hedges Slims delivered about 17.0mg tar and the Commander delivered 19.5mg tar. Considering the flavor and the tar, the Benson and Hedges filler was selected. 5

The Benson and Hedges Slims were compared with Silva Thins in booth testing. The Slim cigarets were found to be a more accept- able cigaret by a slight margin. The Silva Thins evaluated was a recently revised product. The former Silva Thin contained a higher percentage of reconstituted tobacco plus inert filler. The former product was not available for our panel testing. There was interest in a mentholated companion for the VirginiaSlim. The mentholated product was specified to be a cigaret similar to the regular Slim with a menthol impact equivalent to Benson and Hedges 100 menthol. Menthol levels were adj,usted both on the basis of smoking evaluations and'puff by puff menthol delivery data. HTI tests were scheduled for the non-menthol product against Benson and Hedges 100''s and for the Slim menthol against the Benson and Hedgea100 menthol. The non-menthol Slim was booth tested against Silva Thins and~the higher tar delivery of the Slim resulted in greater response to the Slim than to the Silva Thins. The same Slim sample was also booth tested against Benson and Hedges 100's and'the Slim performed very well. The Slim menthol also performed very well in booth testing against the Benson and Hedges 100 menthol. These favorable performances were particularly impressive when considering that the Benson and Hedges products deliver about 2 more mg of tar and'the obvious difference in circumference. Flavor evaluations of the samples for the HTI tests with Manufacturing Department personnel suggested that perhaps the flavor level on the Slims was to high. New sampLes were generated. The samples with the reduced flavor level were a little low in tar delivery. However, the change in flavor level did not indicate any advantage nor did it show that the normal level was a problem. 01 O oo~o ao . c~ - 6 - ~'

HTI test results are not available yet. Observations in the field indicate a potential problem with the apparent fast burn. The combination ofthe fast burn and the small circumference might lead the smoker to the conclusion that he is being "short changed." Slower burning papers are being evaluated to determine whether they can slow down the apparent burn and not excessively increase the delivery nor detract from the flavor. Samples are under preparationin the Manufacturing Department. D. Air Cured Cigarets Interest in an air cured cigaret arose from two separate sources as follows: (1) Animal and statistical data indicated lower incidence of respiratory disease fromithe smoking of air cured cigarets as compared with the smoking of flue cured or blended cigarets. (2) A market research concept test indicated' that the name "'Air Cured" had very favorable connotations to consumers. By definition, Turkish tobacco is consideredto be air cured, however, in animal testing the results with Turkish tobacco were comparable to flue cured. Consequently, the advisability of using,Turkish was questioned, although it would be permissable by definition. Maryland was acceptable from both~ points of view. An all air cured RCB would also be required. It was concluded that Turkish~should be excluded from the product. The RCB and stem usage should be either all burley or burley and Maryland. All of the previous work on burley cigarets was reviewed. The Manufacturing Department produced an air cured RCB. Using the Parliament burley blend a series of samples were produced in Tobacco Processing that would evaluate the contribution of the Maryland tobacco and determine whether Turkish would have been desirable from the flavor point of view and whether the recently developed Turkish casing would be desirable. Several ur) . C.'? C1 - 7 - ~

acids had been shown to have important effect on the smoke flavor of all burley cigarets. Re-evaluation of the effects of these acids was also included in this series. Smoking evaluations of this series by Manufacturing Department and project personnel gave the following indications: (1) Up to 20% usage of Turkish in a burley cigaret was not a major flavor advantage. (2) Turkish casing makes a modest smoke flavor improvement. (3) Maryland does not hurt. (4) Lactic acid results in a major smoke flavor improvement. This fourth point confirms earlier work done in Chemical Research. Two air cured blends were suggested by the Leaf Department, one using 50% of the air cured RCB and the other 30%0 of that RCB. Evaluation of the two showed the 30% blend to be more desirable. Flavor formulation work resulted in several menthol and non-menthol flavors for the air cured samples. A side- stream odor problem was noted as a fact of life with air cured tobacco. Additives that masked the sidestreamiwere easily detected in mainstream smoke flavor and were considered to be an equal problem. Evaluations by Manufacturing Department and project personnel indicated that the non-menthol sample with the Turkish casing was preferred over other available candidates. .These cigarets were 85mm filtered~cigarets with the 30% RCB level and'Marlboro type after cut flavors. The special casings used include lactic acid. The menthol sample selected as the best available was also~an 85mm filter cigaret using the 30% M-~ RCB blend, Marlboro menthol type after cut flavors with flavor Gin 4165-40 were used. These cigarets delivered about 16mg tar. The menthol delivery (0 .6mg) . was a little high. (,l x 8

The non-menthol sample was booth tested against Marlboro and the mentholated was testediagainst Lido. The Marlboro and the Lido were the better performers as was expected. It had been anticipated that the obvious taste characteristics of the air cured cigarets would be detected as a departure from the smoke flavor that is normally expected. However, the work to date was encouraging and the interest in the air cured project had not diminished. A new blendlusing 20% RCB was evaluated and was recognized, as a further improvement. Flavor 4097-76 was judgedito be help- ful for both mainstream smoke flavor and sidestream odor. Filler rods with this flavor, the new blend (including the lactic acid) were combined with flush fluted, charcoal filters. The resulting product delivered about llmg tar. The same rod combined with a Philip Morris Multifilter filter system delivered about 14mg tar. Both of these samples will be evaluated further in booth testing against Tareyton andwill very probably be HTI tested. Meanwhile, work is continuing on the development of flavors for an air cured mentholated cigaret. E. Novel Flavors The continuing work to find distinctive flavors for cigarets that will enjoy consumer acceptance and, hopefully, compensate for some of the deficiencies in low delivery cigarets has been given greater emphasis in the past year. Certainly menthol as used is distinctive andenjoys growing acceptance. The anethole flavored Saratoga has had a limited degree of acceptance. With Un this in mind work was begun to seek flavors that could go along W with either menthol or possibly the anethole. A long series of col flavors was generated, some in response to specific requests. Current popularity of lime flavors and fragrances in other consumer products made it an item of interest. Lime, lime-menthol and lime-anise flavors were applied to cigarets. Lime and lime-menthol were shown t&be more promising in SEF evaluati.ons. Extended smoking evaluations were also encouraging,. Stability studies 9

indicated no major problems with rancidity, however, flavor retention on storage appeared to be a definite consideration. Cost of the lime flavor is also a consideration. The evalua- tion of the market potential for lime flavor cigarets was a definite problem. A test was devised that is intended to give some insight into the market potential. The test attempts to measure the acceptability of the lime concept against a standard mentholated cigaret, and~an anethole flavored cigaret, all three at about a 20mg TPM level. A 10mg lime-menthol model is also tested~against a 10mg anethole flavor and~against a standard delivery Parliament. Hopefully this procedure will give some insight into the acceptability of the lime-menthol flavor, per se, and the role that distinctive flavors can play in improving the flavor of low delivery cigarets. An orange flavor modification of the Saratoga drew very favorable comments. However, retention was a major problem with orange flavor. Flavor oxidation studies showed no problem with rancidity to date. These studies are continuing. Encapsulationof the orange flavor was evaluatedland rejected as a solution to the retention problem~. The problem stems from the volatility of the oil and the high terpene content. The high terpene content makes orange and other citrus oils weak flavors. Concentration of the oils to reduce the terpene content and increase the oxygenated:(flavorful) components results in a - more retentive flavor, and a much more expensive flavor on a per pound basis. Even the folded citrus flavors require higher usage levels than might be expected from experience with other flavor materials. Finally, anorange flavor with satisfactory retention and no obvious rancidity problems was developed. Shelf life studies are a continuing part of this work. This flavor has been filler and filter evaluated. - 10 -

At the suggestion~of Manufacturing Department personnel, a clove modification of the orange flavor was developed. This flavor was evaluated on the air cured cigaret, and as a flavor modification for the Parliament Menthol. An orange modification of the Lido cigaret is under evaluation. A long list of distinctive flavor candidates have been developed. Further work on these has been held in abeyance while pursuing the lime and orange flavors. F. Turkish Casing The objective for the Turkish casing work was to develop a casing that will impart the smoke flavor effects of Turkish~ tobacco to the smoke of a blended cigaret that does not contain any Turkish. While the flavor part of the Turkish casing work was beginning, a program of synthesizing compounds related to Turkish~tobacco was also begun. The first part of the flavor work included evaluation of flavors available from suppliers. This was not fruitful. Compounds suggested by Chemical Research~ were not of immediate interest. However, a number of the acids present in Turkish tobacco were found to be important contributors to Turkish character. First formulations of Turkish casing produced the desired effect, however, the effect was not retained on storage. 3649-85-C based mthe mixed glycerol esters of the acids of interest gave the desired~effect andialso had improved retention. Continuing flavor formulation work resulted in samples 4021-5 and'4021-18. These flavors were evaluated by project personnel and evaluations were made by Turkish Leaf personnel. 4021-5 was selected for extensive evaluation. It was applied to Turkish-free Marlboro~blend(Turkish being,replaced by increased bright usage). Panel testing showed no~significant differences, while panel testing did show differences when the F, - 11 -

control Marlboro was tested against the Turkish free cigaret but without the Turkish casing. In the course of this work, the point of additionof the casing was also considered. Happily, the casing can be added as a component of the after-cut flavor and still retain the desired~effect. The cost of the Turkish casing is of minor concern as the casing added considerably less than one cent to the cost of a thousandcigarets. The Turkish casing is under evaluation in a two part mail- out test. One part of the test is a comparison of a Marlboro control against a Turkish-free Marlboro with the Turkish casing, again with the Turkish replaced by bright. The other part of the test is a comparison of the Marlboro control against the Turkish~-free Marlboro with increased bright and no Turkish casing,. The results of these tests should be available shortly. While this test is in progres&to deternine whether a satisfactory Turkish~substitute has been developed, the Turkish casing is being applied'to and evaluatedlon tobaccos received for this purpose from Argentina and Gautemala. To date the effort on the Turkish casing has been profitable. The Turkish casing,has been the best of the most promising flavors for the fluted'-diluted Parliament. It is the flavor being sent to Argentina in quantity for the Big Ben brand. It has potential as a flavor for reconstituted materials. Samples based on Turkish casing have been evaluated as flavors in Guatemala, Nigeria and Switzerland. G. Flavors for Carbon Filters The subjective response that has beencalled"Carbon Taste" has remained a problem. This is a flavor effect noted when an GO active carbon filter is used. The intensity of the effect is related to the weight of carboniin the filter. This has been - 12 -

confirmed by both Manufacturing Department and project personnel. When the Tareyton 100 product was introdiuced', our evaluations indicated that perhaps the American Tobacco Company was also aware of the problem and were trying to correct the situation. Apparently added flavors were part of their solution. This is not to say that Tareyton 1001has no carbon taste, but rather that it is reduced'. Flavor formulation work on this problem resulted iniflavor 4223-74 which was judged by project personnel to mask carbon taste. At this time, Filter Development personnel were developing a flush-fluted product with~a high gas phase efficiency carbon filter. The flavor was applied'to that product. The product performed very well in booth testing at POL against both Tareyton and Lark. The flavor is currently under evaluation on Philip Morris Filters and further evaluations are planned on the flush- fluted product. H. Flavors for Low Delivery Cigarets The search for flavors for low delivery cigdrets remains an important and continuing part of our work. This work is complementary to the distinctive flavor work. Here the objective is to increase the body and fullness of the naturally thin smoke of a:10mg cigaret by adding blended (not distinct) flavors. This work resulted~ in flavor 4223-68-C which was applied to a diluted~cigaret delivering 11.9mg tar. The flavor was applied to Marlboro filler and the prod'uct delivered 0.29mg menthol. This product was booth tested at POL against both~True Menthol (12mg tar) and Montclair (9.lmg tar) with very favorable results. This flavor is now under evaluation on the Philip Morris Filter Menthol developed'by the Filter Development Group. - 13 -~

Under other projects, the flavor contributions of bright, burley and Turkish rodreflavor section&(tobacco filters) were determined. Descriptive Panel data showed that the reflavor sections modified the smoke to a slight degree. However, the main smoke flavor character came from the combustible rod portion. When a reconstituted rod was used the main flavor was that of the reconstituted rod, regardless of the reflavor section. To date work with reflavor sections has not resulted in major improvements in smoke flavor. However, one additional experiment will be completed. The Marlboro blend will be rearranged, putting the low delivery tobaccos in the combustible section and the higher delivery tobaccos in the reflavor section. This work is being conducted jointly with Tobacco Material Division personnel. II. CIGARET FLAVOR DEVELOPMENT FOR INTERNATIONAL For the past year the requirements for flavor services to International have increased considerably. Approximately one full man year has been devoted to the requests for services that have come to Flavor Development through the International Liaison Officer. The following is a summary of the flavor development activities for International: . A. Argentina Argentina planned to introduce an 85mm filter cigaret to compete with L & M and Viceroy. The cigaret used Argentina tobacco. After considering the flavor of the L & M brand, flavor 3893-90 was developed for inclusion with the standard after-cut flavor used. This flavor is now a routine component of the Argentine Philip Morris Filter brand. Argentina also planned to introduce a 100=cigaret composed of 100% Argentine tobacco. It is intended to be similar to the domestic B & H 100. Several flavor were developedand evaluated.here. Flavor 4021-5-A was selected - 14 -

as the most promising and samples were sent to Argentina for their evaluation. 4021-5-A is based on the Turkish casing work. The flavor was ordered in sufficient quantity to satisfy the original productiion requi_rements for the Big Ben brand and the flavor is now a routine part of the after-cut flavor for this brand. To date flavors have been supplied for 320,000 pound&of tobacco. The flavor has been re-ordered' for 800,000 pounds of tobacco. Work is in progress to develop a flavor for an 85 mentholated cigaret. 3912 type flavor and flavor 4135-90 have beerr evaluate.d here and' have been sent to Argentina for their evaluati.on. B. Canada The first objective for the flavor development work for Canada was to deve?.op flavoLs for an 8.0mg tar vErsion of the Canadian Belvedere. This is an all bri:ght cigaret diluted to the 8mg tar le-,;,el. Several fla;;~-ors were considered and evaluatedhere by project and Manufacturing Department personnel. As a resul.t of this evaluatio-a, four flavors were sent to Canada on prototype cigarets. Of these, Canadian personnel selected sample D7BGI with flavor 4021-67 as an item~ of further inte.rest. Sufficient 4021-67 for 1,000 pounds of tobacco was sent, promptly. Hcwever, the F.a flavor was not incorporated' into the Viscount brand on its introduction. Samples of the R& D made mcdel were tested N a ainst the Viscount in POL booth testing. The g preference x was for the R & D model. The Viscount was also tested against its competition9 Richmond, and.the preference was for the Richmond cigaret. The Richmond ci.garet has the L1 Strickman filter. The R & D flavor :roodel was also preferred over the Richmond. A series of flavors for mentholated - 15 -

versions of the Viscount were also developed and sent to Canada. No response was forthcoming on these samples. In working with cigarets in this very low tar delivery range, the flavor level usage is very critical. The flavors do contribute to the tar delivery. This is particularly noticeable with the mentholated version. Also, with the low tar delivery the usual levels of flavor become considerably more obvious. C. Guatemala The first priority of work for Guatemala was to supply flavors for an 85mm Guatemalan tobacco filter cigaret. This item is intended to compete with the Belmont brand by B.A.T. Several flavors were suggested and sent to Guatemala for their evaluation. Flavor 4021-18 was selected for incorporation in the Diplomat and is included in the after-cut flavor for that brand. The flavor had been supplied in sufficient quantity to treat 266,000 pounds of tobacco. Other work for Guatemala has included attempts at modifying the hot spicy character of the Regios brand. Flavor has been sent to Guatemala for their evaluation. The B.A.T. brand Windsor is of interest. Smokers note a certain sweetness of the lighting puffs of that cigaret and there is interest in duplicating that effect. Scans of the volatiles from Windsor, Derby (another B.A.T. brand) and Belmont were obtained but no meaningful differences were noted. Flavor formulation work resulted in flavors that approach the effect but that work has not been finalized. Gr CL - 16 - ~

D. Mexico The first objective for the work for Mexico is to develop flavors for a brand to compete with the B.A.T. Fiesta. A series of flavors was fornulated and evaluated here on the available Mexican tobaccos. Four flavors were made into cigarets in Tobacco Processing and sent to Mexico for their evaluation. Menthol flavors have also been formulated for Mexican cigarets. Evaluation of these is also incomplete. E'. Nigeria The objective for the wcrk done for N',igeria has been to develop a flavor to duplicate the flavor effects of the Gold Leaf brand. Several flavors were evaluated on the Nigerian tobaccos and flavor 4021-18 was selectedias the most promising. Samples were sent to Nigeria. Their reaction was favorable. Flavor in sufficient quantity to treat 300 pounds of tobacco has been sent. To date no commentahave been received~on the results of their experiments. III. CHEWING'GUM FLAVORS The first priority for the chewing gum flavor work is to supply flavors for the lemon, orange, grape, strawberry, and cherry gums with real fruit. Teaberry Replacement flavors have previously been evaluatedthrough market testing,. However, with the regulation on the continued use of methyl salicylate still pending, provision has been made to work on Teaberry Replacement flavors "as necessary'''. Second priority was assigned to flavor development work on the line of Quasi-pharmaceutical gums. Recently, flavor development has assumed responsibility for the evaluation of the products under development. The following is a summary of the gum flavor work: G0 Ur _ 17 -

A. Fruit Smash Gum An orange flavor for fruit smash gum was developed. Orange juice sacks were obtained as the real fruit component for the orange gum. The orange flavor and mint and vanilla variations of the orange flavor were evaluated and considered to have limited market potential. The high terpene content of the orange flavor also resulted inproblems with sloppy chewing properties. On this basis, work on the orange gum has been held in abeyance and interest had developed' in a raspberry gum. A flavor for raspberry gum with real fruit, which was~considered to be a good' raspberry, was the subject of a POL mailout test. Results of the test showedilimited potential for this product and further work on raspberry has been dropped. Earlier work with lemon flavor for gum had resulted in a flavor that had greater duration than the flavor for Sour Lemon. This flavor incorporated citral propylene glycol acetal, which at the time was not approved as a food additive. Prior to the clearance of that chemical, emphasis on Sour Lemon had'faded and'the work was discontinued. With the chemical being cleared and new interest in lemon flavor for g=, this flavor was evaluated for its application in the Fruit Smash Lemon gum. Lemon juice sacks were also obtained in quantity as the real fruit component for the gum. The new lemon flavor with real fruit was tested against Sour Lemon in a POL mailout test. The results of the test showed that Sour Lemon was preferred. The preference for the Sour Lemon was in part attributed to two factors: (1) the appearance of the real fruit componen.t and (2) the higher acid N level. Flavors were selected'for the Fruit Smash Grape Gum. After a series of reasonably encouraging internal panel tests, the x - 18 -

Fruit Smash Grape gum was compared with~Adams Grape Gum in a POL mailout test. The test results were significantly in favor of the Adam's product. Cherry flavors were also developed'for the Fruit Smash line. These, too, were tested internally and finally in a POL mailout test against the Adam's product. Results were similar to those of the Grape Gum testing. The Fruit Smash Strawberry gum was also the subject of a POL mailout test against the Adam's Sour Strawberry product. Preference was for the Adami's product. With this painful experience behind, the internal panel testing procedures were reviewed for the purpose of determining whether better guidance could be obtained in boothitesting. If nothing else, longer chewing times shou~ld improve the internal panel testing. It appears that the fruit bits in the Fruit Smash gum are not pleasing in appearance and this has inpart contributed to the lack of acceptance of the Fruit Smash line. Reviewing the overall Fruit Smash development scheme with Gum Project personnel, a program has been devised to fully determine the optimum level of acid, the acceptability of the flavor, and the influence of the fruit bits on flavor and the influence of their appearance on acceptability. Currently, interest in Fruit Smash gum is concentrated on the cherry and strawberry products. B. Quasi-Pharmaceutical Gums Flavor Development work for the line of Q.P.'s included~ flavors for the following: Anti-Oral Odor, Throat Soothing, Stimulant, and Full Stomachigums. A series of mouthwashitype - 19 -

flavors were developed for the Anti-Oral Odor (AOO) gum:. The Listerine type flavor of this group was of the greatest interest to the Gum Project personnel. The Listerine flavor was the subject of a U. S. Testing Company test against other oral od.or masking agents. The Listerine flavor was effective against oral odor. However, Doublemint gum was also effective. Typical cough drop type flavors have been developed for the throat soothing gum. These include honey, honey-lemon, wild cherry, cocillana and menthol-eucalyptol. Of these, the Gum Project personnel chose honey-lemon for further work. For information purposes, the honey-lemon gum was booth tested against Wrigley's Doublemint and the Doublemint was found to be far more acceptable as a confectionary product. Coffee flavors for the stimulant gum containing caffeine have beendeweloped. Freeze dried~coffee has been used as both a source of flavor and the caffeine. A new supplier of freeze dried coffee has been identified. His freeze dried coffee appears to be more acceptable than the present supplier. Flavor supplements for the real coffee have been incorporated in the product. The Coffee Gum was also booth tested against Wrigley's Doublemint and the Doublemint was overwhelmingly preferred as a confection, as expected. Peppermint flavors have been supplied for the full stomach gum (Afta-Mint). To date, no panel testing has!been completed. However, a distinct flavor effect due to the bicarbonate has been observed. Work is continuing on the Q. P. line. - 20 -

C. Fruit Punch With Real Fruit There is interest in using the real fruit concept in the Fruit Punch Gum. The stated objective for this work is to make a Fruit Punch Gum with real fruit that will have the same flavor effect as the standard Fruit Punch. Work to date with the real fruit components has not been too encouraging. The real fruit components make a definite flavor contribution and result in a departure from the standard Fruit Punch flavor. Results to date indicate that the addition of real fruit to Fruit Punch lessens the acceptability of the product. It is not clear whether this is a result of the flavor change or due to the appearance or to combination of the two. D. Viokase Peppermint flavor 3059-88 is still the flavor of choice for the anti-calculus gum containing Viokase. Artificial sweeteners are being evaluated. A series of panel tests is in progress to select the best formulation for the product. E. International Recently, there has been opportunity to offer flavors for use in Guatemala. A series of cherry, grape, raspberry, straw- berry, lemon, lime, orange, and peppermint flavors has been sent to Guatemala. They have responded favorably to the cherry and grape flavors. Flavors have been sent in sufficient quantity for factory trials. Bubble gum flavor is under development. Other gum flavor activities include review of the Teaberry Replacement flavor 1677-84 which had been successfully market tested. At this time we are prepared for a regulation that would limit the use of methyl salicylate to the proper tolerance of 3,300 ppm. Formulation work to develop other replacement flavors at higher levels of methyl salicylate are in progress but are currently enjoying low priority. X J - 21 - I'J

Large inventories of both peppermint and spearmint oils had been acquired over the past few years. This was a requirement. It is generally standard practice to have 18 months - 2 years supply of both oils on hand. Changes in the outlook for Clark Peppermint and Spearmint brands left little immediate use for the large inventories of the oils. Flavor Development was of assistance to the Purchasing Department in the disposing of the inventory. IV. CANDY FLAVORS The flavor work in support of the Fruit Smash hard candy line has been the first priority for candy flavor work. Life Saver flavors were established as the goal for this work. Lemon, lime, orange, raspberry, strawberry, and cherry flavors were sought for the line. Interest in grape and'blueberry flavors is no longer active. Flavors for each of the above types have been developed and selected for factory trials at Charms. The products of the factory trials have been evaluated using the internal panels. The results of these evaluations indicate further modifications. The results suggest modifying the flavor levels, acid levels and perhaps some of the factory procedures. Stability studies on citrus flavors have been carried'~ out. The results of the studies indicate that proper storage + conditions for oils are extremely critical. Use of anti- oxidants prolong the shelf life of citrus oils. Observations on gum and~candies indicate that once the flavor is in the product it is fairly well protected from d'eterioration. V. TOILETRIES The emphasis for the flavor development work for the toiletries line was planned to be flavors for Speak Easy. - 22 -

Several new flavors were developed for the aerosol mouth refresher. Among these were several mouthwash type flavors (Micrin, Lavoris, etc.), Spearmint, Teaberry and Cinnamint flavors were adapted for the mouth refresher. A Creme de Menthe flavor developed especially for the mouth refresher was especially well regarded by Product Development personnel. The use of one key ingredient, Wormwood Oil, was ques- tionable. Before the question on the use of Wormwood was resolved, a series of HTI tests of the Spearmint Mouthwash flavor, Teaberry and Cinnamint flavors were scheduled. The development flavors were to be tested against the Colgate 100 product and Whisper. The performance of Teaberry, Cinnamint and Spearmint flavors in this series of tests showed these flavors are quite acceptable and shouldbe considered for line expansion. However, at this time there are no plans for expanding the Speak Easy line. Having had some exposure to peppermint oils in other product areas, Flavor Development evaluated the oil purchased' for Speak Easy. Instrumental and subjective analyses were made. Instrumentally the oil was noticeably different from any oil that had previously been considered. The Analytical Division confirmed that the reason for the differences between this and other peppermint oil was the fact that this oil was adulterated. Their further work indicated that Arvensis Oil was the adulterant. There is also evidence that NOPOL is also an adulterant in the oil. A new source of flavor for the Speak Easy product is being used in current production. This is being done on an interim~ basis pending thorough determination of the acceptability of the oil from the new source. X -23'-

VI. PLANS Flavor Development work will continue in tobacco, confectionery and general products work with the major emphasis on tobacco flavoring. Tobacco flavoring will account for more than half of project personnel time. This effort will almost be exclusively cigaret flavoring with only a minor effort devoted' to smoking tobacco. Flavors for low delivery cigarets and distinctively flavored products will continue to enjoy priority. Specifically, the following will be pursued: 1. Completion of the Lido flavor work including the Lid_- with orange flavor. 2. Evaluate the flavor effects of changes in the burning rate of Virginia Slims. 3. Complete the development of regular and mentholated~ flavors for air cured cigarets. 4. Complete stability studies on the Parliament menthol with orange clove flavors. 5. Develop and evaluate flavors for lcw delivery and carbon filtered cigarets. 6. Develop flavors for specifipd brands for both Mexico and'Guatemala. 7. Adapt the Itizrkish~ casing for use in International brands. 8. Flavor development will participate in the MP filter (,r work by assistingin evaluation of their prototypes. At this time no flavor formulation work is plannedi. The evaluation work should be good preparation for any flavor formulation that may be requiredlat a latter time. - 24 -

9. Flavor Development will also assist in the evaluation~ of the Controlled Profile cigaret. No formulation work is planned immediately. 10. Continue to evaluate compounds suggested for flavor use by Chemical Research. Continue to seek new flavor materials fr=outside sources. 11. Determine the flavor characteristics of all the leading brands. . 12. Initiate studies in Flavor Transfer. Flavor development work for the general product develop- ment line will be expanded to include flavors for the snack food development work. The chewing gum product development work will continue to establish the fruit smash gum line, the line of Q. P.'s, and anti-caries, and the anti-tartar gums. Efforts will be made available for work on Teaberry Replacement flavors as necessary. The search for new and long lasting flavors will continue. Hopefully, more frequent opportunities will arise to offer chewing gum flavor services tolInternational. The first priority for candy flavor work will be the establishment of the Fruit Smash hard candy line. Included in this will be the supplies of the real fruit citrus materials for both candy and gum. For Speak Easy, the work will be d'irectedtoward obtaining the most acceptable peppermint flavor for this product. /ad' - 25 -