1

Crystallizers: The crystal-clear answer to trans fat-free margarine productionPalsgaard has examined how crystallizers can step in to fill the production performance gap left by slowly crystallizing, trans fatty acid-free margarines.

By Anders Mølbak Jensen, Product & Application Manager, Lipid & Fine Foods Palsgaard A/S.

PUBLIC ENEMY: TRANS FATSFor the origins of the fight against partially hydrogenated fats and the trans fatty acids they produce, we have to go back to the early nine-ties. That’s when a famed Har-vard Medical School paper was released, reporting the results of a long-term epidemiological study that tracked, among other factors, the dietary habits of almost 80,000 female nurses1. The study was considered by many to be a land-mark in the fight against coronary disease.

The Harvard research team, led by Dr. Walter Willett, arrived at a number of eye-opening conclu-sions, one of which is that trans fats present a significant health risk, with higher risks of cardiovas-cular disease already registered at daily trans fatty acid intake levels of just 5 to 6 grams.

Further research has pointed to an ability of saturated fat and trans fats to raise the level of Low Den-sity Lipoprotein Cholesterol (LDL) in the blood. Excess LDL particles are deposited in the arteries, lead-

ing to hardening of the arteries and heart disease over time2. Si-multaneously, trans fats appear to decrease High Density Lipoprotein Cholesterol (HDL), further increas-ing the risk of heart disease. Other studies have indicated that a high intake of saturated fats and trans fats promotes insulin resistance, too, threatening consumers with Type 2 diabetes.

INSTANT EFFECT ON THE INDUSTRYThe impact of the Harvard study, and follow-up research, on the food

2

industry was powerful and, at least in Europe, instantaneous. Danish consumers, for example, reacted quickly, slashing their purchases of margarine by around 15% almost overnight, and butter grabbed new market share. Legislators moved almost as quickly, and since 2004 in Denmark, for example, oils and fats used in food products have been permitted a maximum of 2% trans fat content (excepting oils and fats of animal origin).

For European margarine manu-facturers, the message was clear: something needed to be done to reduce their recipes’ content of hy-drogenated fats and restore con-sumer confidence.

The US, on the other hand, has been somewhat slower to react, though it introduced mandato-ry labelling for trans fats in 2006. Most recently however, in June 2015, the FDA has finalised a de-termination that trans fats are not generally recognized as safe, and set a three-year time limit for their removal from all processed foods.

Of course, the ban needs to be seen in the light of an 85% re-duction that has been gradually achieved by the industry over the past 10 years, mainly due to the mandatory labelling, but the writing is clearly on the wall for US man-ufacturers: It’s simply not enough.

So, while the industry is pushing to have very low amounts of trans fats allowable in, for example, the sprinkles on cupcakes or cook-ies, manufacturer focus is now on tuning recipes to deliver unaltered product quality in an uncompromis-ing new, trans fat-free landscape.

LIFE WITHOUT TRANS FATSNo trans fats on your new ingredi-ents list? Congratulations: life as a production manager at a marga-rine manufacturer is about to get much more interesting.

Partially hydrogenated oils have long enabled margarine to step up to the plate, forming more stable

products and supporting mouth-feel to deliver an eating experience not far from that of butter. And their positive effect on shelf-life is well documented, too.

Without trans fats, it’s much more difficult to consistently produce high-quality margarine. In fact, ev-ery part of the production process becomes more sensitive to a va-riety of factors that were comfort-ably, even transparently, handled by partially hydrogenated oils in the past.

HIGHER MELTING POINTOne such factor is the higher melt-ing point of other fat types. For op-timum flavour release, it’s best to use fats that melt at approximate-ly mouth temperature: around 35 oC (95 oF). In the old world, trans fats fitted the bill perfectly. In the new one, the only economically feasible, readily available fat type is palm oil, fractions of palm oil, or interestified fat types.

That said, the melting point of palm oil is still a little higher than that of trans fats, and it lacks comparable functionality. Not only that, but in the US, and for reasons that may be difficult to pin down, the use of ‘tropical’ oils has long been given the cold shoulder. So palm oil isn’t a viable option in that part of the world.

The melting point of a fat isn’t just important for the consumer’s eat-ing experience, it also affects the ability of manufacturers to work with the fat during the production process. In a trans fat-free world, you get a mixture of high melting point fractions and liquid oil, giving a higher melting point and a ten-dency toward softer products.

Perhaps the most important phe-nomenon, however, for margarine manufacturers to think about is the slower crystallization speed of trans fat alternatives – and this brings us to the particular focus of this paper.

THE EFFECTS OF SLOWER CRYSTALLIZATIONA lot has been done to address slow crystallization issues, but generally, manufacturers are sim-ply unable to produce as much margarine from the same pro-duction lines as before. Process parameters, including machine settings, need to be adjusted to cope with the slower crystalliza-tion, and most often, investments in new tube chillers are demanded or, for example, the combination two machines where only one was needed before. Whichever route was chosen, final product quality just isn’t the same as that of trans fat-containing recipes.

Without trans fats, for example, margarine begins to crystallize on the tube chiller shaft, reducing production capacity by perhaps 20% from morning to afternoon, as there is gradually less volume to work with. Flushing the tube with heat to restore capacity is one way to fix the problem, but has its own set of drawbacks and can’t be rec-ommended.

Lower production capacity is one effect of crystallization. Another is that, due to the slower crystal-lization speed of alternative fats, crystallization continues to devel-op for longer than the usual 24 or so hours during pre-storage, changing its structure over an ex-

FACT BOX: The effect of trans fatty acid reduction

• Higher melting point

• Slow crystallization of the fat types

• Easier to overwork the fat product

• Post crystallization

• Change structure over time

• Storage: more sensitive to temperature variations

3

tended period of time. So you’re likely to end up with a more brittle product than you’re used to. Stor-age, therefore, has a much greater effect on the final product – and storage temperature variations introduce further sensitivity to re-crystallization.

Pre-storage, therefore, seldom the focus of a margarine production process in the good old trans fat days, has suddenly taken on new significance for the final product’s quality. Now manufacturers need to focus on pre-storage during the first 5 to 7 days following pro-duction, making this aspect part of the overall production process. For example, an attempt might be made to to reduce brittleness and ensure consistency by varying pre-storage temperatures from, for example, 21 degrees for the first 5 days then reducing to 16 degrees thereafter. One implication for many manufacturers is the need to invest in pre-storage facilities that can enable the required tempera-ture control.

Slower crystallization, therefore, is a manufacturer’s nightmare, im-posing both immediate limitations on total production capacity and gradually reduced capacity during production days due to crystalli-zation buildup on chiller shafts. Non-hydrogenated products will have a slower crystallization speed, and interestified products have a tendency to become over-worked and very soft.

IN STEP CRYSTALLIZERS AND EMULSIFIERSWith widespread attention on the ill effects of trans fats, US food manufacturers and their European counterparts now need to come up with trans fat-free recipes that give margarine’s batch-by-batch quality the best possible chance of suc-cess.

According to Palsgaard’s applica-tion center research into the area, and our lengthy experience in the first-to-move European market, that particular ‘something’ is most

likely to be carefully concocted combinations of crystallizers and emulsifiers.

Crystallizers, in particular, can do much to ease the production pro-cess. And we’ll take a thorough look at exactly how shortly. That said, crystallizers will not have significant influence on the lon-ger-term storage phase, during which polymorph structure will make for a firmer margarine over time, an aspect that manufacturers will still need to give special atten-tion to in their trans fat-free strat-egies.

EXPLORING TRANS FAT-FREE PRODUCTION STRATEGIESTo get to the bottom of the case for these ingredients as the toolbox of choice for food manufacturers fac-ing the trans fat ban, we first need-ed to gain a better understanding of crystallization processes in mar-garine oils and fats. We were par-ticularly keen to find out how var-ious process parameters affected the speed and nature of crystalli-zation. Equipped with this knowl-edge, we would then be able to properly examine the role played by emulsifiers in relation to crystal-lization, and to make recommen-dations to margarine manufactur-ers based on our findings. And we set the bar high: our tests would be conducted with one of the most difficult beasts in the business: puff pastry margarine.

WHAT AFFECTS CRYSTALLI-ZATION IN MARGARINE AND SHORTENING?When chilling begins in the marga-rine production process, there are, at first, no crystals. Then the first crystals appear, creating a ‘seat’ for more to build upon, finally ar-riving at a much firmer mass. This firmness needs to be broken down somewhat, restoring plasticity.

To tackle this problem, we used one of our pilot plants to start the seating earlier in the process, al-lowing more time in the machine to reduce post-crystallization time. Longer time in the equipment, however, means greater effect from the pin machine and the fol-lowing tube chillers, resulting in a quite different product.

In our trials, we compared hydro-genated, interesterified fat, and unhydrogenated fat in puff pas-try margarine. Hydrogenated fat, as might be expected, performed best, crystallizing quickly. Far slower to crystallize, unhydroge-nated fats such as palm oil clearly perform worst. In fact, both before and after the pin machine it was hopelessly overworked, and would be impossible to pack or, for that matter, to eat. The interesterified fat medium crystallized well, but again was all too easy to overwork.

Liquid margarine withPalsgaard crystallizer

Liquid margarine without Palsgaard crystallizer

4

CRYSTALLIZERS: THE SOLUTION OF CHOICEAt Palsgaard, we’ve determined the most effective strategy is to use crystallizers as a trans fat-free ‘remedy’.

These can be, for example:• High in behenic acid (C22) and

steric acid (C18) • High in steric acid (C18) and

palmitic acid (C16)• Very high in steric acid (C18)• Mixture of behenic acid (C22),

steric acid (C18) and palmitic acid (C16)

The melting points of various fatty acids, triglycerides and crystalliz-ers can be seen in Table 1. A key aim is to determine a crystallizer whose melting point makes it easy to use in production. This can be achieved with steric acid (18), combined with behneic acids.

Tribehenic and monobehenics have the most extreme melting points for crystallizers. At 180 oF (82.2 oC), they cannot be handled in a normal production environment. If a pipe were to become blocked, for example, things would become very problematic. Despite this, and in the name of thorough research, we did attempt to use this acid in various recipes, but achieved the same or better results with other products. Other triglyceride com-positions, on the other hand, with behenic acids can produce a lower melting point, making them easier to handle in production.

Any of these acids can easily pur-chased, but constructing these triglycerides – using high melting point behenic acids yet ending up with something that has a signifi-cantly lower melting point – is no walk in the park. Locating the be-henic acids in the right place on the chain requires more than a little expertise! With these acid in the right positions, however, the melting point can be brought down to as little as 66 degrees C instead of, for example, 82.

CRYSTALLIZING EFFECT OF CRYSTALLIZER COMPARED WITH TRIBEHENATEUsing Tribehenate Mp, with its 82 oC (180 oF) melting point, speeds up crystallization greatly in com-

parison with our reference (40% palm oil, 40% palm stearin and 20% liquid soya oil without crystal-lizer), as can be seen in Figure 1.

The Mp crystallizer we tested al-most matches Tribehenate Mp’s speed, but with a lower melting point of 61 oC (142 oF). Both solu-tions far outstrip the performance of the reference. The tests also revealed that using crystallizers forms more beta prime crystals, which have a better absorbing ef-fect than other crystal types. The tests also revealed that using crys-tallizers forms more beta prime crystals, which have a better ab-sorbing effect than other crystal types.

TRIVIAL NAME MELTING POINT TRIVIAL NAME MELTING POINTC12:0 Lauric acid 44.4°C (112°F) Trilaurin 46.1°C (115°F)

C14:0 Myristic acid 54.4°C (130°F) Trimyristin 55°C (131°F)

C16:0 Palmitic acids 62.7°C (145°F) Tripalmetin 65°C (149°F)

C18:0 Stearic acids 69.4°C (157°F) Tristearin 72.2°C (162°F)

C22:0 Bebenic acids 80°C (176°F) Tribehenic 82.2°C (180°F)

C18:1 n-9 cis Oleic acid 16.1°C (61°F) Crystallizer : C22 –C18 61.1°C (142°F)

C18,1 trans Elaidic acid 43.8°C (111°F) Crystallizer : C16 –C18 57.2°C (135°F)

C18:2 n-6 cis Linoleic acid -6.6°C (20°F) Crystallizer : C16 –C18-C22 58.8°C (138°F)

C18:3 n-3 cis Linolenic acid -12.7°C (9°F) Crystallizer : C18 72.2°C (162°F)

Table 1:Melting points of fatty acids, triglycerides and crystallizers

820100 340 580 10600

500

100

1500

2000

2500

3000

η in

Cp

Crystallizer Mp 142oFTribehenate Mp 142oF

Reference

Figure 1:Crystallizing effect of crystallizer compared with tribehenate

5

Figure 2:The effect of crystallizer doses on production capacity

0

1

2

3

4

5

6

7

8

9

10

4.4

8.1

0.5

10

bar

0.0% crystallizer 40%

1.5% crystallizer 40%

0.0% crystallizer 70%

1.5% crystallizer 70%

0

1

2

3

4

5

6

7

8

9

10

4.4

8.1

0.5

10

bar

0.0% crystallizer 40%

1.5% crystallizer 40%

0.0% crystallizer 70%

1.5% crystallizer 70%

SFC VALUES WITH AND WITHOUT CRYSTALLIZERMoving on with our exploration, we examined more or less the same puff pastry margarine recipe as in the previous trials, this time adding 1.7 grams of crystallizer. We were keen to determine the Solid Fat Content (SFC) values that resulted from this addition. These are de-picted in Table 2.

The trials demonstrated a some-what surprising result: While 1.7% crystallizer is not a large portion of the recipe, its effect on crystal-lization was strong, quickly getting the seating in place upon which to build further crystals. Judging by the SFC values in Table 2, crys-tallizers have a strong effect on fat crystallization because they acti-vate some of the fats to crystallize more quickly than would otherwise be the case. With other test meth-ods, 24 to 48 hours at 0 oC (142 oF) would have been required to do this.

CRYSTALLIZERS AND PRO-DUCTION CAPACITYNext up, we took a closer look at the effect of crystallizers on pro-duction capacity. Figure 2 shows the results of testing a similar puff pastry margarine recipe to that used in our other trials.

We tested two production capacity levels, each with and without crys-tallizers added. In Figure 2, the

pilot plant’s bar pressure indicates that crystallization is occurring. When producing at low capacity (40%), the fat has time to crystal-lize by itself. But at higher capac-ity (above 70%), low bar pressure shows the fat is more liquid and crystallization is less effective. At low capacity, fast-crystallizing fat types can produce crystalliza-tion, but it requires crystallizers to be added to increase contact and result in sufficient crystallization both at higher and lower capaci-ties.

As a general conclusion, it seems that the more crystallizers you add, the higher the bar pressure and therefore the better you are able

to maintain or increase capacity. Without adding crystallizers, you would need to reduce capacity, al-lowing the blend to remain in the tube chillers for longer. So, while adding crystallizers will increase recipe costs, there’s a worthwhile trade-off in better utilisation of pro-duction equipment.

Still in our experimental corner, we decided to look more closely at the effect of different crystalliz-er dosages on production capaci-ty, moving from zero to 0.5%, 1%, 1.5% and, finally, right up to 2%. To make things more interesting, we simultaneously tested the effect of two different rotation speeds: 400 and 800 RPM.

TRIAL 3 TRIAL 4 TRIAL 5 TRIAL 8 TRIAL 9 TRIAL 10Oil blend: Palsgaard® 6111 1.70% 1.70% 1.70% 0.00% 0.00% 0.00%

Palm stearin 40.00% 42.00% 45.00% 40.00% 42.00% 45.00%

RBD. palm oil 40.00% 42.00% 45.00% 40.00% 42.00% 45.00%

Liquid soya oil 6°C 18.30% 14.30% 8.30% 20.00% 16.00% 10.00%

100% 100% 100% 100% 100% 100%

Melting point: 56.20% 56.90% 56.20% 51.70% 50.80% 52.80%

Avr. SFC % 10°C/50°F 61.90% 63.00% 67.20% 52.00% 55.40% 59.80%

Avr. SFC % 20°C/68°F 50.00% 51.70% 54.80% 40.20% 42.00% 43.90%

Avr. SFC % 30°C/86°F 31.70% 32.20% 34.20% 21.00% 22.40% 24.80%

Avr. SFC % 40°C/104°F 21.00% 21.80% 22.70% 12.40% 13.00% 13.90%

Table 2:SFC values in puff pastry margarine blends with and without crystallizer

6

In Figures 3 and 4, the coloured bars represent the four tube chillers in our pilot lab production setup. It came as no surprise, of course, that the viscosity needed for margarine required more ener-gy toward the end of the produc-tion line. We also confirmed that applying crystallizers does indeed increase viscosity by the end of the process, but the difference they make is nothing extraordinary.

CRYSTALLIZERS AND PRODUCTION CAPACITYMargarine produced for incorpo-ration into puff pastry is very thin when first made, expanding in the oven to arrive at the mouth-feel loved by consumers all over the world. Seen from a puff pastry ex-pansion point of view, trials with Palsgaard’s crystallizers produce a very stable margarine. No matter how hard we work the product, it still performs well in this vital re-gard. But we were curious to learn what might happen if we were to conduct a baking test on the final product itself. So we measured the effect on ten such pastries of in-creasing crystallizer dosages from zero to 2%, switching between three different rotation speeds.

Normally, rotation speed isn’t a pa-rameter most production staff can work with. Often, that’s because most production machinery has a fixed speed – or enables switching only between one or two levels. Now however, in a trans fat-free landscape, the ability to adjust rotation speed in combination with various dosages of crystalliz-ers has become a key factor for achieving the right result.

In Figure 5, rotation speed strong-ly affects expansion results, per-forming best at around 11 RPM. Introducing crystallizers smooths out the effects of rotation speeds, enabling a good expansion result not only with different process parameters but also on different machines – expansion simply be-comes less sensitive to process parameter variations. There is, of course, production machinery that

8.6

9.99.2

11 10.811.2

9.89.3

10.8 10.7

9.69.3

9.710.4 10.7

0

2

4

6

8

10

12

0.0% crystallizer 0.5% crystallizer 1.0% crystallizer 1,5% crystallizer 2,00% crystallizer

Exsp

ansi

on (x

)

400 rpm in all tube chillers 600 rpm in all tube chillers 800 rpm in all tube chillers

Figure 5: The effect of different process parameters on puff pastry expansion

2.4 2.4 2.4 2.42.5

2.6 2.6 2.6 2.6 2.6

2.4 2.4 2.4

2.8 2.8

2.6

2.8 2.8

3.2 3.2

0

0.5

1

1.5

2

2.5

3

3.5

0.0% Crystallizer 0.5% Crystallizer 1.0% Crystallizer 1.5% Crystallizer 2.0% Crystallizer

Am

p

Tube chiller 1 Pin machine 1 Tube chiller 2 Tube chiller 3

Figure 3:The effect of crystallizer doses at 400 RPM

Figure 4:The effect of crystallizer doses at 800 RPM

7

can arrive at comparable results without any addition of crystalliz-ers, but few manufacturers have invested in the up-to-date equip-ment that can do it.

ARE CRYSTALLIZERS AWAYS NECESSARY? Let’s stop for a moment and ask an important question: Have we all too narrowly pursued crystallizers as the only feasible solution to the problem of slower crystallization? Are they, in fact, always neces-sary?

The answer is ‘no’. It is possible to create recipes that perform just as well as, or at least compara-bly to the performance of a trans fat-containing formulation. Cake margarines, for example, that use fast-crystallizing fat types such as palm oil or coconut oil fat, have nothing to be gained by adding crystallizers. But with cheaper fat types, small amounts of crystalliz-ers can make a significant differ-ence.

There is, however, no way to avoid spending some amount of addi-tional funds to cope with a trans fat-free world – and for most, the costs will be high. Manufacturers will need to take a look at their equipment line-up. Most likely, older machinery won’t be enough to maintain current capacity and product quality on the new, trans fat-free playing field – at least, not without applying crystallizer dos-ages as high as 2%.

Upgrading brings new and better technologies with it – enabling, for example, the use of CO2 as a more efficient and effective cool-ing medium. But even after an up-grade to more modern equipment, manufacturers should still expect to incorporate from 0.5 to 1% crys-tallizer content in their recipes.

A TOUGH JOURNEYFor European manufacturers, who have been battling the effects of low or no trans fat recipes for quite some years now, it’s been a tough journey. The only comfort, per-haps, was the fact that everyone

was in the same boat – product quality decreased across the en-tire industry and everyone needed to come up with their own answer to the problems. In the US, things are about to heat up – but there’s much to be learned from the jour-ney already embarked upon by their European competitors.

In a nutshell, to cope with a trans free future, production managers need to realize that the days of pouring in ingredients, setting the machines and retreating to the comfort of the control room are most likely over. Instead, even a minor fluctuation in production re-quires active – and proactive – ef-forts. And most affected are those whose job it is to produce puff pas-try.

WE’RE HERE TO HELPMargarine manufacturers need to develop new production strategies, understanding the new realities described above, experimenting with crystallizers and emulsifiers in their recipes, and preparing for significant capital investments in equipment.

Palsgaard is here to help its cus-tomers work their way through the hurdles, helping to arrive at a roadmap for process, recipe and equipment adaptations. We have all the equipment needed to play with every production parameter, taking samples at many different points and conducting trials that simply wouldn’t be feasible in a live production environment.

Manufacturers have a variety of other avenues to turn to for advice, too, including concentrated efforts being made by oil refineries (de-veloping new oils and fat types), universities, in-house R&D centers and machinery/storage suppliers.

The sooner the industry faces up to the issues discussed in this article, the better. Failure to move quickly enough is likely to result in a roll-ercoaster ride through reduced ca-pacity, zig-zagging product quality and costly re-runs. The sooner the effort is made, therefore, the lower the overall cost of the necessary transition to trans free products – and the smoother the journey for both manufacturers and their cus-tomers.

ABOUT THE AUTHORAnders Mølbak Jensen, is Product & Applica-tion Manager, Lipid & Fine Foods Palsgaard A/S. He can be reached at am@palsgaard.dk or +45 7682 7682. Before joining the com-pany 16 years ago, he held various roles as a laboratory, quality control and R&D manager at a large margarine producer for 10 years. He holds an M.Sc in food science and tech-nology from the University of Copenhagen and a Bachelor of Commerce, IT. His email is am@palsgaard.dk. Get more info at www.palsgaard.com.

1 American Journal of Epidemiology. 2005 Apr 1;161(7):672-9.Dietary fat intake and risk of coronary heart disease in women: 20 years of follow-up of the nurses’ health study.2. Oh K1, Hu FB, Manson JE, Stampfer MJ, Willett WC.