The of Changes Caused by Processing

Marlene Stauffer Blommer Chocolate Company

he purpose of this paper is to discuss Formulations using a single bean vari- Tand demonstrate flavor differences ety versus a blend of beans may make dra- among milk that are formulated matic flavor differences in a chocolate for- with a constant recipe, altering only the mula. To date, much of the understanding processing parameters. A product devel- of chocolate flavor has been based on gas opment team worked together to develop chromatographic and mass spectral data a protocol.The ground rules were set.The of the volatile aroma fraction of dried first step was to establish the formula, then and/or roasted cocoa beans and the result- determine processing techniques, focus on ing formulated chocolate. More than 400 what processing techniques to alter, ana- compounds have been found in cocoa fol- lyze the product and finally to evaluate the lowing fermentation, drying, roasting and Stauffer is the sensory effects of the process changes. . Some flavor elements in choco- director of quality The goal was to maintain the formula in late are highly volatile compounds con- assurance and new Figure 1 throughout the test process. tributing sour notes; moderately volatile product develop- Certain assumptions that would affect compounds contributing roasted or flow- ment at Blommer flavor were made prior to any samples actu- ery notes; and low volatile compounds con- Chocolate. She has ally being manufactured. Roasting, the type tributing milky or notes. The been with Blommer and amount of raw materials, the refining ensemble of all these flavor components since 1982. process and the conching process are all produces a unique chocolate flavor. Spe- variables that play a part in determining cific beans are chosen for their known con- the final flavor outcome. tribution to chocolate flavor. For our test, we have chosen to keep the roasting para- ROASTING meters constant using a low roast West Roasting parameters could alter the flavor African-type and to focus by the selection of bean type, degree of on the chocolate manufacturing process. roast (whether it is a low, medium or high roast chocolate liquor) and also the amount MILK COMPONENT of chocolate liquor in the formula. Milk as a raw material in a Marlene Stauffer

54th PMCA Production Conference, 2000 145 The Flavor of Milk Chocolate

Functional formulation is critical. Functional perfor- Deodorized was chosen for performance, mance, sensory properties, cost and stor- the test samples so that the flavor of the sensory properties, age life are all factors determining the value cocoa butter would not interfere with the of milk in a formulation. chocolate intensity of the samples and to cost and storage The type of milk selected is also a fac- reduce the variability in flavor which can life are all factors tor. Whole milk powder is the most com- sometimes be found in prime press cocoa determining the mon milk ingredient in a milk chocolate. butter of similar blends. value of milk in a Roller-dried whole milk is often used for its formulation. high free- usage and more yellow color. FLAVORINGS Spray-dried whole milk is more readily Flavorings are added to further enhance available; however, the process used will or modify already existing notes in a choco- determine the amount of free fat available late.The most common flavor additive used to the chocolate manufacturer. It is typi- in chocolate is , either in the form cally more white than a roller-dried prod- of natural vanilla or the now more com- uct.The amount of fat in a whole milk pow- monly used artificial . Salt is the sec- der may also vary from 26 to 28.5 percent. ond most common additive. These addi- A combination of nonfat dry milk and tions are not used to mask the chocolate anhydrous milkfat is sometimes used in flavor, but to intensify it.Vanilla and vanillin chocolate formulations for a more buttery help to add creamy notes, whereas salt flavor and a high free-fat system. This will accents clean, crisp notes and helps to obviously alter the rheological properties of reduce bitterness. a particular chocolate formulation and is Vanillin, the most common additive to sometimes used when cost is an issue. milk chocolate, was the flavoring added to Milk crumb is also another common the test samples manufactured. milk product used to manufacture special formula milk chocolates. The milk crumb REFINING helps provide caramel . Maillard Various types of refining equipment exist in browning reactions occur between free the chocolate industry today. Two-stage amino groups and reducing . These refining consisting of a prerefiner and five- reactions result in colors and roll refining is the most common system caramel flavors.The degree of carameliza- and was used for the manufacturing of the tion can be precisely controlled in crumb test samples. Controls of this equipment manufacture. set the stage for preconditioning a choco- For this project, roller-dried whole milk late mass to achieve an optimal conching powder with a fat content of 28.5 percent process. It is important to envelope the par- was chosen as the milk component. ticles with fat so that undesirable absorp- tion of water during refining does not occur. COCOA BUTTER If refining is poorly adjusted, it will be Chocolate manufacturers have a number impossible to make up for this during the of requirements with respect to the quality conching cycle. of cocoa butter. Prime press cocoa butter The refining process has certain para- Formula from a blend of chocolate liquors might be meters that can be changed that might alter ...... 50% Milk ...... 19% used to further enhance chocolate flavor the flavor. Refining will determine the size Cocoa Butter . . . . 16% or a deodorized cocoa butter could be used reduction of a chocolate mass as it is being Chocolate Liquor . 15% to give a more mild chocolate flavor. The manufactured. Whether a product is fine, Soya . . . . 0.3% choice depends on the intensity of the fla- medium or coarse ground will determine Vanillin ...... 0.01% vor desired in the end product as well as the palette’s flavor perception.The particles Figure 1 the melting and solidification properties. will be coated with fat, which is the flavor

146 54th PMCA Production Conference, 2000 The Flavor of Milk Chocolate carrier. When these particles enter the ufacturing and the conche temperature. The function of the mouth, the melt, sweetness and With greater conching intensity, derivatives conche is to all will influence how the product .A have been found indicating sugar break- remove unwanted product with a high fineness (coarse) will down. The most prominent in the process flavors but at the also appear darker in color compared to a is the caramelization of lactose, which same time retain low fineness (fine) chocolate. A low fine- would be a source for the malty and ness product will have more surface area, intensely caramel notes. the more desirable requiring a greater amount of fat to coat A goal in conching is also to obtain the ones. each particle, and will sometimes give a optimum viscosity (flow properties) at the sticky mouthfeel due to the lowest practical fat content. Key elements required to smooth the particles across the in this goal would be moisture reduction, tongue. A coarse product will have more input of energy, control of superfines and free fat due to less cocoa butter needed to amount of free fat available. coat the solid particles. This will create a There are several different types of more free-flowing (thinner) viscosity conches in the industry today—from dry requiring less energy to coat the tongue conching to wet conching to continuous and giving a different flavor perception. conching.Test samples were manufactured from a wet-conching pilot system. CONCHING It is well known that adjustments to conch- PARAMETERS ing parameters can modify chocolate fla- The parameters we chose to focus on to vor. Conching can be manipulated by time, examine the flavor profile of the exact same temperature, moisture content and shear. paste formula follow: Introducing energy to the product mass in • Conching time the form of shearing and increased tem- • Conching temperature perature over a period of time will create • Particle size certain flavor developments. The function These parameters were chosen based on the of the conche is to remove unwanted fla- above previous assumptions. We felt these vors but at the same time retain the more particular sample types would demonstrate desirable ones. The amount of conching is differences by process and by examples. important. Typically, conched chocolate is considered more mellow and blended as PROCEDURES compared to unconched chocolate. The paste was manufactured using granu- Acidity and bitterness are reduced dur- lated sugar, roller-dried whole milk pow- ing conching. Ziegleder states that 30–40 der, deodorized cocoa butter, low roast percent of highly volatile elements are West African-type chocolate liquor, soya removed through the conching process and lecithin (an emulsifier) and vanillin. The Schneider feels that reduction of acid in a raw materials remained constant through- chocolate mass can succeed much easier out the test process. in a dry substance rather than in a pasty The products were mixed in a Hobart substance.Temperature adjustment will aid mixer, measuring each ingredient very pre- in the degree of caramelization based on cisely, then sent through the refining the Maillard browning reaction, where the process and into a high shear Frisse pilot basic reaction is between free amino groups conche. Time, temperature and amperage and reducing sugars.The were monitored throughout the cycle.The may well depend heavily upon the recipe; product was then standardized to the the prior treatment of raw materials such as agreed-upon formula and analyzed. Sam- liquor treatment; milk process; crumb man- ples were hand-tempered, cooled and eval-

54th PMCA Production Conference, 2000 147 The Flavor of Milk Chocolate

The chocolate uated from a sensory perspective. The ini- more dairy notes with about the same level data shows tial samples were prescreened by the prod- of sweetness as the unconched sample.The that physical uct development group before sending unconched sample was found to have mod- samples to the full panel.The sensory erate nutty notes not present in the characteristics are evaluation was conducted by a team of 15 conched sample. It also had off-notes definitely different trained panelists using independent grouped as raw and unconched at a slight between conched descriptive analysis. A 0–7 intensity scale level of detection. and unconched was utilized, with 0 indicating the absence By not conching a milk chocolate, the chocolate. of an attribute and 7 indicating a very product has flavor spikes and is unrounded, strong and overpowering attribute. The with hints of raw notes and disjointed fla- process continued for each set of samples vors. The conched sample is much more until all parameters were adjusted. The rounded, with smooth, chocolatey notes. samples were evaluated at an average age The milk, sugar and chocolate flavors are of one month. well blended together and balanced. According to Ziegleder, completely EFFECTS OF CONCHING TIME conched masses will be perceived to give a Physical Characteristics more fine taste compared to unconched Initial paste formulas (the actual refined masses due to the penetrating sweetness paste) were exactly the same.The variables of the sugar. The roundness occurs by the were a long conching time process and a coalescing of the surfaces and flavor ele- short conching time process with a con- ments becoming more available with the stant conching temperature. The conched help of the fatty phase. There is still the sample was conched for 12 hours at 64°C question of whether chemical reactions and the unconched sample was conched occur in conching as they do in roasting. for 1 hour at 64°C. Comparing the two samples, conching does When it came time to standardize the help to round flavor, smooth particles and products, the unconched sample required reduce viscosity (which has an added ben- additional cocoa butter to bring it to a efit of reducing cost). workable viscosity to make the test sam- ples.The fat content was increased and vis- EFFECTS OF CONCHING TEMPERATURE cosity altered. The chocolate data in Fig- The second set of samples looked at altering ure 2 shows that physical characteristics the conching temperature.The low temper- are definitely different between conched ature conching was at 64°C for 12 hours and and unconched chocolate. the high temperature conching was at 74°C Sensory techniques used for the evalua- for 12 hours with constant conching time. tion of test samples were followed. Each Samples were manufactured using the sample used all five senses to capture the previously established protocol.The conch- full mouth experience and the palette was ing times remained the same while alter- cleared between each taste session with ing only the conching temperature. Typi- crackers and water. cally, it is assumed caramelization occurs around 72°C. The data in Figure 4 shows Sensory Evaluation the major difference between the two sam- Panelists focused on degree of chocolate, ples was flowability. But what about the caramel notes, sweetness, dairy and any flavor of the samples? other perceivable flavor notes. The spider For this set of samples, the panelists were diagram in Figure 3 indicates that the sen- focusing on caramel notes as well as choco- sory panel found the conched sample to late, dairy and sweetness. have more chocolate, more caramel and The spider diagram in Figure 5 illustrates

148 54th PMCA Production Conference, 2000 The Flavor of Milk Chocolate

Changing Conching Time Changing Conching Temperature Completely conched masses Time Fineness Viscosity Percent Temp. Viscosity Percent Sample (hrs.) (microns) (NCA) Fat Sample (°C) Fineness (NCA) Fat will be perceived Long 12 15 155 29.3 Low 64° 15 155 29.3 to give a more Short 1 15 too high 29.3 High 74° 15 100 29.3 Short+ 1 15 145 34.2 fine taste Figure 2 Figure 4 compared to unconched masses Conching Time Flavor Profiles Conching Temperature Flavor due to the Profiles caramel caramel penetrating sweetness of the raw, sugar. unconched chocolatey

dairy chocolatey

nutty sweet

dairy sweet

Sample A: unconched Sample A: lower temperature (64˚C) Sample B: conched Sample B: higher temperature (74˚C) Figure 3 Figure 5 that the higher-temperature sample had a but also with the viscosity at the same level more pronounced caramel note with slightly of fat. The physical characteristics of the less perceivable chocolate notes.The higher- samples are obviously different, but how temperature sample is perceived sweeter did the flavor change? with less dairy apparent.The flavor percep- tion of a low-temperature, long-time conch- Results for Fine versus Coarse Chocolate ing cycle shows a well-rounded sample, with Panelists concentrated on mouthfeel, milky, chocolatey notes, very even.The high- sweetness, caramel, chocolate and dairy fla- temperature, long-time conching cycle shows vor notes. slightly less chocolate notes, with caramel Even though the only parameter changed notes perceivable. with this set of samples was the particle size, you can see from the spider diagram in Fig- EFFECTS OF PARTICLE SIZE ure 7 that perceptions other than the mouth- The third set of samples demonstrated feel are also affected. The most notable varying the particle size of a chocolate change is the increased perception of sweet- mass.Will we see a change in flavor or only ness in the more coarse sample.This sweet- in flavor perception? ness seems to mask the intensity of the Samples were manufactured with 15 caramel, dairy and chocolate notes. and 35 microns by adjustments to the The taste and the flow characteristics of refining process of the mass.All conching a chocolate mass are developed not only times and temperatures were held the during the conching cycle, but also during same at 64°C for 12 hours. The chocolate size reduction and homogenization in the data in Figure 6 reflects the differences in refining process. the physical characteristics, showing dif- Chocolate refined to a very high degree ferences not only with the particle size, has a more balanced flavor, while the more

54th PMCA Production Conference, 2000 149 The Flavor of Milk Chocolate

The taste Changing Particle Size Effects of Change and the flow Fineness Viscosity Percent Conching Time characteristics of a Sample (microns) (NCA) Fat Short time (1 hour) Fine 15 155 29.3 less chocolate; raw, unconched; nutty chocolate mass Long time (12 hours) Coarse 35 60 29.4 more chocolate, more caramel, high dairy are developed not Figure 6 Conching Temperature only during the Low Temp (64°C) chocolate, round; caramel; dairy conching cycle, but Changing Particle Size Flavor Profiles High Temp (74°C) also during size caramel more sweet, more caramel, less dairy reduction and Effects of Particle Size homogenization in 15 microns chocolate, round; caramel; dairy the refining 35 microns process. more sweet, less caramel, less dairy dairy chocolatey Effects of particle size demonstrate not only flavor but texture differences between fine and coarse milk chocolates using the same recipe. A coarse product gives the flavor sweet perception that it is more sweet with less Sample A: fine (15microns) overall flavor.The particle size selected for Sample B: coarse (35 microns) a milk chocolate depends to a great degree Figure 7 on final product application requirements. coarse chocolate is perceived as more Chocolate flavor and its development is sweet.The chocolate recipe and the size of a complex process. True examples indi- the individual ingredient particles largely cating the various flavors can be produced determine the extent to which product taste from one single formulation using con- can be controlled. Fracturing of particles stant raw materials. It is wonderful that in the refining process opens the particles chocolate making, with all the scientific up to be simultaneously enveloped by the reactions that occur in the process, is also fat, which will carry the flavor and be fur- an art. The magic and variation of all the ther enhanced in the conching process. flavors available lend themselves to sat- CONCLUSIONS isfying all of our cravings. Effects of conching time definitely give a more rounded chocolate flavor. Unconched REFERENCES samples exhibit raw, unconched nutty notes Beckett, S.T. Industrial Chocolate Manufacture and Use. 3rd Edition. with less overall chocolate flavor. Each spe- Kuster, Werner. Two-Stage Refining— An cific chocolate recipe will require experi- Important Step for Optimizing the Conching mentation to determine the optimum conch- Process. Buhler AG, . Minifie, B.N. Chocolate, Cocoa and Confec- ing time required to meet a specific flavor. tionery. 3rd Edition. Effects of conching temperature show dif- Rozin, P.The Flavor-Fusion Illusion. PMCA Proceedings, 1996. ferences in the degree of caramelization Schneider, F. Is Dry Conching Still Needed?, perceived. A low-temperature conching 1996. cycle will round all flavors together.A high- Seguine, E.S., Factors Influencing the Taste, Selection and Specification of Chocolate. temperature conching cycle will promote PMCA Proceedings, 1998. caramelization. The temperature chosen Zak, D.L.,The Development of Chocolate Fla- for a particular recipe will also depend on vor. Manufacturing Confectioner, November 1988. the treatment and type of raw materials Ziegleder, G. Aroma Development in Conch- and the final desired flavor. ing. Susswaren 1997.

150 54th PMCA Production Conference, 2000