Louis Pasteur did his ground-breaking work on pasteurization in the 1860's, showing how heating wine to high temperatures could inactivate spoilage organisms. While this is certainly one of the greatest achievement in the annals of food science history, he forgot to mention the effect on the flavor. Pasteurization generally does not do us any favors in that arena. Modern flavor chemists still haven't found all the answers to the challenge of flavoring pasteurized products, but it certainly has come a long way since the early days.

The heat is on

Classical pasteurization refers to any high heat process designed to inactivate pathogens and spoilage organisms, including retorting. But the food industry generally uses the term to refer to a controlled time and temperature treatment applied to milk and other fluid products that may not necessarily permit ambient or long-term storage without spoilage. For fluid milk, that means at least 145°F for thirty minutes. However, modern milk pasteurization usually means a high temperature short time process (HTST) of 161°F for not less than 15 seconds or ultra-high temperature (UHT) treatments which can practically run from 191°F for 1 second to up to 212°F for 0.01 second.

Juices follow the same general relationships, but enzyme inactivation is also a priority. Orange juice for example is commonly pasteurized for 16 seconds at 165°F or 1 second at 185°F.

When subjecting fluid products to these temperature conditions, some deterioration of the flavor is inevitable. Several major types of reactions contribute to this: volatilization of low boiling point flavor compounds, chemical chances to desirable flavor compounds, physical entrapment and the generation of off-flavors from chemical reactions in the product matrix. In addition, formulation, processing, and non-inert packaging- materials all can influence the finished product flavor.

"It's pretty straightforward," maintains Mike Tyrrell, senior flavorist at Givaudan-Roure, Clifton, NJ. "Generally, the longer a flavor is exposed to high temperatures, the greater the chances for its degradation through chemical interactions with the product base. If the pasteurizing equipment is not a closed system, flavor components are lost via evaporation and the profile can become unbalanced."

What's going on?

Volatility

If the boiling point of a flavor compound is below that of the heat treatment, it becomes a gas and enters the atmosphere. For example, ethyl butyrate, a component of banana flavor, and acetaldehyde in orange juice are very volatile and can be lost readily through heating. Whether these volatiles are actually lost depends on the process.

"Some pasteurization methods give better results than others," explains Joe Dono, manager of Givaudan-Roure's beverage application group. "A plate exchanger is a closed system and that's obviously better than a batch system. You don't drive the volatiles out into the atmosphere. In most cases, you can reintroduce them into the product, especially at normal pasteurization temperatures. If you start going higher than 190 to 195°F, then that becomes more difficult. Below that, you're generally okay. As long as the product doesn't remain hot for an extended period, probably in the range of an hour, you're going to have good results."

Chemical changes
As in most food products, there are innumerable chemicals making up both the flavor system – either naturally occurring or added – and the product matrix. And with the application of heat, some of these become quite reactive. These reactions frequently result in a tangled web of lost flavors and newly generated flavors – usually undesirable ones.

"The more simple the product, the fewer problems you're going to have," says Dono. "With dairy products, you've got fat, proteins, etc. Combining these with high temperatures for extended periods of time increases the number of reactions you can have. You don't get caramelized notes in a simple sugar and water mixture."

"Very often, the prolonged exposure to high temperatures can cause sugars and proteins to form a Maillard reaction," notes Dick Stypula, manager of dairy applications, Givaudan-Roure. "These types of reactions, particularly in dairy products, contribute to off-notes and loss of the original flavor profile. Volatile loss in HTST processes tends to be minimal – it's a closed system. You are concerned with maintaining a minimal exposure time to protect the flavor components from reactions with resident proteins. Unsaturated aldehydes and many ketones will react and produce compounds that are responsible for off-notes or which actually may mask a product's normal flavor profile. Many of these off-notes and reduced flavor profiles can be found in no-fat or low-fat vanilla frozen desserts where the protein and vanillin have reacted."

"Sugars will react, especially fruit juice, to form something like furfural or hydroxy methyl furfural, which are considered detrimental to the flavor in orange juice," says Dave Bowen, director of creative development, Quest International, Owings Mills, MD. "These occur when you heat sugars. They give a caramelized cooked flavor. There are secondary reactions that form furanones, or, if there are proteins present, you'll form pyrazines. While something like pyrazine is pleasant in cooked chicken, it's not in milk – it gives it a scorched character."

Flavor entrapment

Sometimes the loss of flavor in pasteurized foods is due to entrapment of the flavor molecule by some substance in the matrix, usually a high-molecular weight compound such as a protein or a polysaccharide hydrocolloid. The exact mechanism is not fully understood, but in essence, these molecules can physically trap flavor molecules. While these entrapped molecules remain chemically unchanged and can be picked up analytically, they are shielded from any taste receptors.

Other reactions and conditions
Sometimes the problem is not so much a change in the flavor itself as it is a change in the flavor perception. This happens in non-fat foods. Fat is going to provide a mouth coating effect and change the way the flavor comes across by extending the duration. But to compound the problem, many of the ingredients that act as fat mimetics change the viscosity of the product. In pasteurization, that often translates into more exposure to heat as the rate of heat transfer slows.

"Product matrices do make a significant difference on flavor performance," says Carol McBride, manager, flavor applications, Dragoco, Totowa, NJ. "Ingredients such as starch, hydrocolloids and dairy powders, as well as product parameters like pH and viscosity, will influence the flavor."

Because of the effects of the product matrix, flavors that perform well in one pasteurized system will not perform the same in a different one.

"If you have a flavor containing aldehydes and ketones, they will attach to protein molecules and actually disappear from your perception," explains Bowen. "That is pH-dependent and varies with the type of protein as well as the specific flavor component."

"Acid association will affect the flavor perception threshold," Bowen continues. "The threshold of acetic acid is about 54 ppm in water, but in oil it is zero. On the other end, the threshold of dodecanoic acid in water is zero, but is 750 ppm in oil. Depending on the fat content, these sorts of considerations are very important to the level of each of the acids you use. The more lipophilic acids in a full-fat product will tend to disappear."

The sweetener is another ingredient that affects the perception of many flavors. Fructose, for example, gives a different perception of sweetness than sugar or high-fructose corn sweeteners. The sweetness peaks quickly then quickly dissipates.

"Many of the reduced-calorie products are requiring us to redefine flavors," observes Paulette Lanzoff, associate technical director of applications and flavor creation, Universal Flavors, Kearny, NJ. "When you start using artificial sweeteners, for example, a lot of them will give you an entirely different flavor perception than when you use sugar."

Packaging interactions with the product are more likely to create problems. Either off flavors from insufficiently cured packaging material will transfer to the product, or flavor volatiles will be absorbed by the package.

"You have to look at the packaging material, even if you're not doing aseptic processing, unless you are using glass," cautions James LaMarta, Ph.D., applied research, Givaudan-Roure. "If you are using some form of PET, you have to be careful and work with the packaging companies to make sure that there isn't any leaching. You'll always get some monomers in your product, but you need to know that the level is acceptable and will not adversely affect the flavor or the product."

Flavor fixers

All marketing claims aside, we are not yet at the stage were we can exactly replicate all of the flavor attributes of a fresh product once it has undergone heat processing. However, a number of techniques can bring it closer to the fresh ideal.

Obviously the first step is to optimize the process so that the severity of the heat treatment and its effect on the flavor is minimized.

"A modem HTST or UHT process is far kinder to a flavor system than the old 145°F for thirty minutes," maintains Bowen.

But since designing the product for the process is a way of life and product designers are frequently called in to provide answers, here are several. But they only apply to added flavors. For those working with pure, unflavored systems, go back to the first option.

Flavor systems can sometimes be designed with less reactive flavor compounds. It may be a product with a higher boiling point or one that is more resistant to a particular undesirable reaction.

"Diacetyl provides a buttery note in dairy products and is itself quite reactive with certain portions of protein molecules," states Tyrrell. "But by using a flavor compound further along the homologous series -- such as acetyl propionyl -- there is a good possibility it will be less reactive yet provide some of that diacetyl character."

In general, the larger the molecule or the longer the chain, the more stable the flavor is to heat within a homologous series.

"You may not get a lot, especially in the case of esters, but you do get a little," notes Brian Mullin, senior flavorist, Sanofi Bio-Industries, Trevose, PA. "It may be enough to get there. A lot of these processes are rapid heat and quick cool down in a scaled system."

Often, designers of food products believe that artificial flavors are more stable than natural flavors. That may be true if the artificial contains more stable compounds, but if the compounds are the same, there will be no difference. And while the natural components available in the past were possibly more volatile than the ones that could be developed through artificial means, modern flavor processes now derive many of these more stable flavor ingredients naturally.

"Physical parameters like boiling points and flash points are the same whether they are natural or artificial as long as they are the same purity," remarks Mullin. "You won't see a difference between natural and artificial ethyl butyrate. Sometimes with naturals, as a general category, you can lose some different components that distort their character and so taste less authentic than artificials where you can provide things that are not lost as quickly. Also with natural flavors, many customers prefer ethanol as a diluent which has a low boiling point."

Another way to formulate a more resistant flavor system is by looking at the solvent. Solvents such as propylene glycol have relatively low boiling points. As they boil off they co-distill the more volatile portion of the flavoring with them. A solvent that suppresses the volatility will retain more flavor.

"One of the ways we create higher boiling point flavors is to change the diluent," says Mullin. "Incorporation of something like ethanol into your flavor drops the boiling point of the entire flavor. Using propylene glycol or vegetable oil can help preserve some of the flavor."

A third option would involve using special flavor formats.

Encapsulation, spray drying and emulsions all show improved resistance to degradation than a liquid format.

"Some gums help to fix a flavor into a system," claims Lanzoff. "If you use the right kind of gum to develop a flavor emulsion, it can help to some degree."

True encapsulation technology where a complete capsule is formed may be of limited practical application at this time. Because of the sophisticated process needed, the flavors are much more expensive.

Also, pasteurization involves high heat, high shear and high moisture. That kind of environment is a challenge to encapsulation technology.

It's difficult to find a compound that will stand up to all that abuse and still release the flavor once it hits the mouth.

"With some of the more novel encapsulation techniques, you actually encase the flavor in a high-temperature melting wax," says Mullin. "The problem with that is, unfortunately, the waxes don't go away. You're left with pieces of encapsulate you can sense in your mouth."

Spray drying, although a form of encapsulation, in this case relies more on the flavor components than the encapsulating material for stability. During the spray dry operation, the flavors undergo a high-heat process. This eliminates any extremely volatile components, leaving those that are stable at process temperatures. This concept holds true for most flavors generated through high heat and high pressure processes.

"Pyrazines are a family of compounds that are very stable -- nuts, chocolate, coffee," advises LaMarta. "So is vanillin. We also have quite a few reaction chocolate flavors manufactured by a heat and pressure process, which is a good example of how to produce a heat stable flavor whether that was the intention or not."

Flavor emulsions also provide a measure of stability against heat. Fat can work as a protectant against heat.

If none of these techniques give the desired result, another way to achieve a balanced profile after processing is to use an unbalanced one going in. If the flavor is overloaded with the notes that change or dissipate during processing, a sufficient amount may survive the process. Merely increasing the level of the entire flavor system may achieve this end. This however could become a costly alternative. Another way to look at the problem would be to use similar compounds that change to the desired profile.

"In essence you're talking about a type of reaction flavor," notes Stypula, "It's similar to using a precursor for a particular profile that has been compounded in such a way that the pasteurization finishes the flavor's processing and yields the desired profile. In the screening to create such flavors, you determine the starting point by working back from the final needed profiles. The sensory profile of the flavor prior to processing may bear little similarity to what you have in the finished product."

Behind the mask

Since the heat process generates flavors -- especially the Maillard reaction products – the finished product may not just lose flavors but actually gain some that are overpowering or otherwise objectionable. Of course, this also depends on the exact nature of the finished product.

"Aseptically packed milk is generally considered to have an off-flavor relative to fresh milk," points out Stypula, "but it's a flavor note you usually need to make a good-tasting vanilla ice cream – some of the caramelized/cooked character of condensed milk. Most of us drink fluid milk packaged in the traditional manner and this off-note isn't a concern."

If you are looking at covering up objectionable flavors that heating generates, masking flavors may help. These can physiologically dull the senses to off notes. But these can be difficult to work with as they may affect the good as well as the bad.

"With mask notes, the masking affects all the flavors; some more than others," LaMarta discloses. "In some cases, you can target. But in general it's hard to mask just one note. Someone once had masked the off notes so well that the product tasted like water."

"There are two approaches here," Tyrrell adds. "Masking is rarely totally effective where you have an objectionable flavor sticking way out, but there are ways of minimizing the problem by blocking or overpowering that flavor note.

Hopefully, after that you can build your flavor, although the masking agents do tend to never totally eradicate that strong note. The other approach is to develop a flavor where that nasty note is actually a part of the flavor. You use that as your vehicle since you're not masking the off-note and build a flavor that doesn't dull the senses and is perceived much better."

Go with the current

Last, it helps to use flavors that go with the base and any changes in it that pasteurization creates.

"With aseptic milks, for example, there are certain flavors that just won't work because you can't cover up those artifacts giving it a non-fresh taste," notes Tyrrell. "A fresh, green fruity character is difficult to produce. If it's a chocolate, a vanilla, a coffee or even a jammy/processed strawberry you can get away with it. There are limitations on the types of flavors you can successfully develop in those products."

"Banana goes well with milk," Dono adds. "It's a strong flavor with a lot of different notes. In fact banana cream pie is a good example of how it marries well with the brown flavors that show up in cooked milk."

The product itself can also be a source of a "reaction" flavor and that particular note can then be left out so the balance is right in the finished product.

"If you are working with a 'jammy' strawberry and you know that some of the brown notes will be generated by the matrix during heating, leave them out of the compound," advises Lanzoff.

Current trends indicate that more and more products will be undergoing pasteurization. While the explosive growth in aseptic juices may have slowed, more and more pasteurized products are appearing on the shelf.

"New age beverages are here to stay," notes Dono. "Companies are making capital investments in heat processing equipment which will go against conventional bottling plants. These types of products taste better that those that are chemically preserved. The label that says 'no preservatives' is very appealing to the consumer."

Until some alternate method of eliminating heat to ensure the safety of foods, we are stuck with the results. But until that day we have to live with the flavor consequences of heat treatments like pasteurization. If we look at how far we've come, I think Louis would be proud of us.