A Look at Studies and Experimental Brew with Tested Results
Fermentables in wort are about 70% maltose, glucose, fructose, sucrose, and maltotriose the other 30% are mostly maltotetraose (6%) and dextrins (22%). Dextrins are unfermentable extract that remains in wort from starch that is not completely broken down to the main fermentable sugars (glucose, maltose, and maltotriose) from amylases during the mashing process. In general, dextrins are said to potentially contribute to mouthfeel, but will not add any flavor.1
A 1989 paper looking at the potential for dextrins influencing mouthfeel didn’t have high hopes going into the study when stating in the opening paragraph “whether dextrins contribute significantly to any other beer character than its caloric value is doubtful.” This statement comes despite the science at the time of the study saying that dextrins do increase the viscosity of beer (1962).2 Again in 1962, dextrins were said to have a thick and pasty effect on mouthfeel3 In 1977, dextrins were said to be tasteless but still agreed that they do have an effect on viscosity and add to the mouthfeel of beer.4
Back to the 1989 study, here a trained panel was given commercial light beer (with no dextrins) dosed with dextrins ranging from 20 to 70 g/liter and asked which samples were thicker or more viscous. The authors found that in order for there to be a detectable increase in viscosity or mouthfeel, there needed to be more than 50 g/liter of dextrins added to the beer. They disagreed with most of the previous research at this point in time saying dextrins likely contribute less to mouthfeel and it has more to do with beta-glucans, ethanol, glycerol, glycoproteins, and melanoidins because most beers (at this point) had between 10 and 50 g/liter of dextrins. Interestingly, the ratings for sweetness were closely matched to viscosity when added at concentrations of 60 g/liter or more to beer. Although not surprising, the authors also found that when the samples were uncarbonated, they had a higher rating for viscosity, which they explain is why stouts with less carbonation than perceived as heavier, creamier, and more viscous that ales or lagers that are more heavily carbonated.5
A later study (1991) measured chemical and physical properties of beer which might contribute to mouthfeel and then correlated these with sensory attributes. Thirty different samples were used in the study, including different beer types and styles ranging from non-alcoholic lagers to barleywines and served to a trained tasting panel who were asked to rate the intensity of the samples for the different descriptors. One of the measurements taken in the study with HPLC was dextrins (separated out into groups DP4-DP9 and >DP10) and they found a correlation at the 0.1% level for the terms viscosity and thickness. Other parameters that correlated well to density and viscosity were polyphenols, fermentable sugars, chloride, and glycerol. This is another example of chloride linked to improved mouthfeel, which previously wrote about. Glycerol was actually correlated higher to viscosity than beta-glucans, which again makes me interested in having English strains used in NEIPAs tested for total glycerol production (my test results below actually do include glycerol for WLP007). The paper does conclude that brewing trials need to be conducted to confirm which parameters cause these mouthfeel sensations, however.6
A 1993 paper looking at multiple studies on dextrins, like this article, ultimately concluded that dextrins have little or no effect on the mouthfeel attributes related to fullness. One study cited did a series of mashes on worts with different dextrin content and found that the beers with higher dextrins lacked palate fullness.7 Another study added enzymes to beer to hydrolyze the dextrins and found that this did not diminish the fullness of the beer (no sensory data to back up these claims was included).8 However, the paper also cited a study that suggested that a high wort dextrin level in beer results in low attenuation and gives beer a thick and pasty mouthfeel.9 Although the papers seem to contradict each other, I feel most confident in the paper finding that you need levels over 50 g/liter to achieve a mouthfeel difference, which is where I focus most of my discussion on my experimental brew test results.
Diastatic power is the term used to know how much starch-converting enzymes a specific malt might have, which break down starches into fermentable sugars. These enzymes consist of beta-amylase, alpha-amylase, limit dextrinase, and alpha-glucosidase. Only limit dextrinase (hence the name) has the ability to break down dextrins (by cleaving alpha-1,6-linkages in amylopectin and limit dextrins) into fermentable sugars. Considering that most beers have about 20% of small branched dextrins that are carried into the finished beer,10 it’s important to understand what factors might encourage or discourage limit dextrinase activity to either leave these dextrins alone or to break them up. Basically, the more active limit dextrinase in the wort, the more fermentable sugars11 and higher alcohol content and less total dextrins.
All of the starch degrading enzymes are created during the germination phase of malting (unmalted grains, therefore, wouldn’t have any limit dextrinase activity). It appears difficult to know the limit dextrinase potential in specific malts, it was significantly correlated in a study of more than 70 different barley varieties to extract protein content, but noted that 80% of the variation is explained by “other factor or factors,” which to me suggests for now, that it’s fairly random. However, there does appear to be some studies that show what types of mash conditions might favor limit dextrinase activity. We do know that lower malt kilning temperatures retain more of the limit dextrinase enzyme, specifically, the enzyme is stable at 120°F to 149°F, but dramatically declines at temperatures increase to 179°F.12
A study looking at different mashing environments found that at mash temperatures of 135°F for 15-minutes, approximately two-thirds (60-70%) of the total level of limit dextrinase activity was inactivated. The more the mash increased above 135°F, the more inactive the limit dextrinase. At normal mash temperatures starting around 149°F, the activity was almost non-existent. However, at 125°F, most of the starting limit dextrinase remained13, which suggest that if you conduct a protein rest (113-138°F) you would likely have a good temperature range for limit dextrinase activity. So a protein rest might do more than just break down large proteins into smaller ones, but also break down dextrins, which may be one reason why John Palmer states in How to Brew that “using a protein rest on fully modified malts tends to remove most of the body of a beer, leaving it thin and watery.”
Another study looking at mash temperatures also found that limit dextrinase activity was fairly stable at 113°F-131°F and declined as the temperatures in the mash increased. After 30 minutes at 158°F limit dextrinase activity was undetectable (which happens to be the temperature I mash most of my NEIPAs, likely encouraging dextrins into the final beer).14 So, it seems we can think of the limit dextrinase enzyme similarly as we do with beta-amylase, which is that they are effective at lower temperatures.
Saliva and Dextrins
An interesting study looked at how dextrins might potentially contribute to the aftertaste of beer through salivary α-amylase. In other words, they studied whether our mouth’s saliva might have the ability to break down dextrins into sugars and have an impact on the taste of beer by asking participants were asked to put beer in the mouths for periods of time ranging from 5 seconds to 5 minutes and then the saliva was analyzed.
The authors actually found that the conditions in our mouths are adequate for the hydrolysis of beer dextrins. When saliva was added to 2.5 ml of beer, they detected large amounts of maltose and maltotriose were produced as well as small amounts of glucose. The study says the reaction was rapid, however, they tested the solution after 1-minute of contact time. I certainly don’t hold beer in my mouth for that long before swallowing! Ultimately, the study concluded that because of the period of time the sugars take to produce in the mouth, it may not cause a detectable sweet taste, but it could potentially partially mask the bitter aftertaste in beer.15
If I’m really trying to taste and describe a beer, the contact time in my mouth is around 4-7 seconds. Looking at the charts in the study, a number of sugars produced in this time period are so low, I doubt it can have much of an impact. The increase in sugar production really seems to start climbing at 30 seconds of contact time. So I’m a little skeptical of the idea that dextrins can break down to sugars quickly enough in our mouth to have much of an impact on the overall taste of beer or even partially mask bitterness.
Lessons from Scotch Whisky
For brewers who sometimes don’t boil their wort, I found a study on scotch production and limit dextrinase activity to be relevant. Unboiled wort has been found to produce greater alcohol yields than boiled worts,16which is a technique used in producing scotch. During scotch production worts are often made in batch style mashes and never boiled. This is what led researchers to look at limit dextrinase activity during fermentation, because if limit dextrinase survives mashing and goes into the fermenter without being deactivated by boiling or pasteurization temperatures, then maybe the enzyme could continue to break down dextrins as the scotch ferments. In the study, mashing of malted barley was done at 149℉ for 60 minutes and run off into a fermenter and sparging was done in four different batches at different temperatures, the later runoffs are done hotter bringing the wort to around 185℉. The first two runoffs are pulled off into the fermenter without experiencing hotter temperatures, where they discovered that most of the limit dextrinase survived and was mostly extracted from the wort prior to the warmer water batch sparges.
The authors found that limit dextrinase not only survived the 60-minute mash but remained relatively high during the early stages of fermentation and even increased around the 10-15 hour mark. This means that with limit dextrinase available during the early stages of fermentation, it’s available to break down dextrins into fermentable sugars with the potential to increase the alcohol yield (beneficial in scotch production) and potentially reduce the mouthfeel because of the reduction in dextrins.17 This is likely only relevant to beer brewers who don’t boil their wort and also don’t bring their beer up to pasteurization temperatures prior to cooling. Essentially, if you do a mash on the lower end of the temperature spectrum and then runoff directly into a fermenter, you could see the continual breakdown of dextrins during fermentation leading to an increase in alcohol production.
In the Oxford Companion to Beer, Carapils is described as a highly modified pale malt kilned at 230°F and never above 320°F, which at the higher temperatures the sugars caramelize and become glassy and unfermentable. The book also makes mention that in some styles such as bocks, carapils can be as much as 40% of the grist for fuller body and mouthfeel. Briess Carapils is a true glassy caramel/crystal malt, albeit one that isn’t roasted enough to develop the color or flavor associated with darker caramel malts. According to the book Malt: A Practical Guide from Field to Brewhouse, specialty glassy malts are made with low temperature and high moisture to make pale malts with a glassy endosperm. How exactly Briess is making their Carapils is uncertain as it’s a proprietary process, however. Weyermann Carafoam (Carapils outside the US) is different than Briess Carapils and is akin to chit malt, high in protein and under-modified. It is mealy/starchy so it too is converted into fermentable sugars when mashed, but would be unsuitable for steeping. Weyermann suggests it can be used as up to 40% of the grist.18 For the purposes of this post, I focus on and use Briess Carapils.
To test out the research on dextrins, I brewed an experimental NEIPA with a grist consisting of 50% Briess Carapils and 50% 2-row (not counting a small addition acidulated malt added for the mash pH adjustment). Combining the heavy dose of Carapils (Briess recommends only 1-5% of the grist) with a mash temperature at 160°F where limit dextrinase activity is low, should be enough to create a beer with substantial dextrins. I sent a sample of the experimental beer to White Labs to get the actual dextrin concentration tested (results below) to go along with sensory descriptions.
As for the recipe for the beer, I was anticipating (or hoping for) a big mouthfeel beer so I chose to use creamy caramel-like hops in an attempt to create a shake-like beer. To do this, I used the Fantasia blend of hops in the whirlpool from the Barth-Haas Group, which is advertised to add a “silky touch of cream and caramel” to beer.19 For the dry hopping, I doubled down on the rich aroma hops by first using Experimental Green Machine hops, which are supposed to have an orange and vanilla characteristics.20 I then added Otto Supreme hops to the serving keg to try and add a fruitier element to the beer (peach, pineapple orange, guava21).
Water: 100% Reverse Osmosis treated at .20 grams/gallon gypsum and .60 grams/gallon chloride
Estimated IBU: 44.1
Mash temperature: 160°F
Batch Size: 5 Gallons
50% Briess Carapils
(Also 8 ounces of Acidulated Malt for mash pH adjustment)
15 grams Styrian Golding First Wort
30 grams Styrian Golding @ 30 minutes
40 grams Styrian Golding 25 minute whirlpool
100 grams Fantasia Blend 25 minute whirlpool
3 ounces Experimental Green Machine on day 3 of fermentation.
1 ounce of Experimental Green Machine on day 9 of fermentation (purging headspace as I added the addition)
2 ounces of Otto Supreme in the serving keg, left at room temperature for 24 hours before going into the fridge.
Fermentation: WLP007 at 68°F
I wasn’t sure what kind of efficiency I was going to get with 50% Carapils, so I overestimated my grains a touch assuming I might see a short drop, but I actually got a little better efficiency than I normally do. In terms of fermentation, I didn’t notice a difference from other beers fermented with WLP007. On day 10 of fermentation, it already reached its final gravity of 1.014, which is also fairly typical for that high of a mash temperature with the strain. I reached out to Briess to see if they had a comment on the seemingly normal fermentation I experienced with the heavy dose of Carapils, which is advertised to add non-fermentables to wort and below is the response:
“For a multi-step mash with sufficient enzymes we typically see about a 10% reduction in apparent attenuation from the Carapils portion of the wort, i.e. 50/50 brew of pale (80% apparent) and Carapils (72%) would expect to yield an apparent attenuation of 76%. If enzymes are deficient or Beta Amylase rest is shortened or omitted you would expect to see much lower attenuation rates from Carapils, maybe down to about 30%.”
This doesn’t really explain the normal fermentation in my case since beta amylase activity is optimal between 140-150ºF 22 and I mashed immediately at 160ºF. Skipping the beta amylase step, I had 76.7% apparent attenuation, well above a potential low of 30%. I’ve recently had three brewing friends brew beers high in Carapils as well and none of them experienced poorer than expected attenuation.
Looking at the White Lab test results shown below, it’s interesting to see that even with 50% Carapils, the beer tested at 41.3 g/liter, which is well below the threshold of 50 g/liter of dextrins needed to notice an increase in viscosity and mouthfeel as the research above found. A representative at Briess was nice enough to respond on this topic and slightly disagree with the study, responding “as far as dextrin perception is concerned, it seems like 50g/L (~5 Plato, or a .020 increase in specific gravity) is quite high, and the detector is not very sensitive.”
After getting the results back, I’m a little confused what exactly Carapils is doing. Advertised as a dextrin-style malt that enhances mouthfeel by adding non-fermentables doesn’t seem to match up with my results. With 50% Carapils in the grist (well over the 1-5% recommendation) skipping a beta amylase step, I still had normal fermentation and was under the 50 g/liter dextrin threshold cited in the literature. I asked if White Labs would share a more traditional grist test result to compare how the dextrin content varied, which would have shed some light on the total amount of dextrins Carapils was adding, but wasn’t able to obtain one without getting another beer tested. Going back to my exchanges with Briess, they did mention that increased body could also be coming from the elevated levels of beta glucans from Carapils, but since the process is proprietary, I’m not really sure how this happens or what the beta glucan levels are.
I speculate that if an increase in beta glucans is desired oats would do a better job. A mash consisting of 100% malted barley was tested with a beta-glucan content of only 20 mg/L, adding just 10% unmalted oats increased the beta-glucan content to 393 mg/L. Replacing 40% of the malted barley with unmalted oats increased the beta-glucan content to 1,949 mg/L, which is a 97-fold increase.23 It’s not clear however how much beta glucans are being added by the Carapils.
A great experiment done by Brulosophy tested the differences between a beer made with Carapils and another beer made without. The experiment was setup so that a blond ale brewed with 10% Carapils was compared to an identical beer except that the Carapils addition was replaced with just base malt. They ultimately found that only 2 of 11 people were able to identify the beer made with Carapils in the triangle test. Although a small sample size, using well over the recommended range of using 1-5% of the total grist, Carapils didn’t seem to make much of a difference.
In terms of the sensory side of things, this particular experimental NEIPA does have a nice soft mouthfeel, but I wouldn’t describe it super thick or chewy. To my memory, it’s not a whole lot different than NEIPAs, I’ve made without Carapils (however oats were used).To be fair though, I didn’t have an identical beer to taste side-by-side only swapping the Carapils with something like 2-row. The one thing I did notice is that the head retention was better than I typically see, which was mainly noticeable in the sticky lacing around an empty glass. Perhaps this has something to do with the protein content of Carapils, but this information doesn’t seem to be available either.
Changing gears and focusing on the strange mix of hops I used, the aroma and taste are certainly different than I’m used to when using typical fruitier varieties. There is a strong caramel aroma with a hint of vanilla, reminding me of caramel square candies. This is likely a combination of both the Fantasia whirlpool and Experimental Green Machine hops. There isn’t a whole lot of fruitiness to the beer, I should have upped the Otto Supreme addition and lowered the Green Machine to achieve this. There is a noticeable hop spice in the beer, similar to the traditional kettle hop taste you can get with hot-side hopping (which this beer had plenty of).
It’s a different beer for sure, reminding me more of a caramel shake than a fruity NEIPA! It’s strange to get so much crystal malt like aroma and tastes from the hops. Some of that I guess could be coming from the high amount of Carapils, but it’s advertised as not influencing color or aroma. Fantasia and Experimental Green Machine might be interesting to play with in a milkshake NEIPA!
- After doing the research and test results back on my NEIPA, I’m leaning more towards dextrins not really having that much of an impact, which is certainly something I didn’t see coming. Mouthfeel may have more to do with proteins, beta-glucans, ethanol, glycerol, and melanoidins than with total dextrin content (or the combination of all of them).
- After doing this exercise, I just don’t see how such small additions (1-5%) can really have much of an impact on mouthfeel.
- Lighter carbonation may contribute to a fuller more viscous mouthfeel and may be beneficial in a NEIPA.
- Glycerol was correlated higher to viscosity than beta-glucans. Because my test results showed that in this beer fermented with WLP007 produced 1.8 g/L of glycerol, I’m now even more interested to get additional NEIPA strains tested.
- Only limit dextrinase (hence the name) has the ability to break down dextrins. Limit dextrinase activity is greatly reduced as the temperature of the mash increases. The temperatures of a protein rest (113-138°F) is likely the best environment to encourage limit dextrinase activity. Hot mashes on the other hand held around 158°F will likely end all limit dextrinase activity fairly quickly.
- Limit dextrinase activity can continue during fermentation if the wort is not brought up to pasteurization temperatures, which is only relevant to brewers to don’t boil their wort and runoff directly into a fermenter.
- Otter, G. E., Popplewell, J. A. & Taylor, L. J. European Brewery Convention Proceedings of the 12th Congress, Interlaken, 1969, 481
- De Clerck, J. Cours de Brasserie, Universite de Louvain Institut Agronomique, Section de Brasserie, Heverlee-Louvain, 1962
- Vermeylen, J. Traite de la Fabrication du Malt et de la Biere, Association Royale des Anciens Eleves de Institute Superieur des Fermentations, Gand. 1962
- Bradee, L. H. “The Practical Brewer”, Masters Brewers Association of the Americas, Madison, WI, 1977.
- Ragot, F., Guinard, J., Shoemaker, C. F., & Lewis, M. J. (1989). The Contribution Of Dextrins To Beer Sensory Properties Part I. Mouthfeel. Journal of the Institute of Brewing, 95(6), 427-430. doi:10.1002/j.2050-0416.1989.tb04650.x
- Langstaff, S. A., Guinard, J., & Lewis, M. J. (1991). Instrumental Evaluation Of The Mouthfeel Of Beer And Correlation With Sensory Evaluation. Journal of the Institute of Brewing, 97(6), 427-433. doi:10.1002/j.2050-0416.1991.tb01081.x
- De Clerck, J. A Textbook of Brewing, Vol. 1. Chapman & Hall, London, 1957
- Scriban, R. Petit Journal du Brasseur, 1971, 79, 537
- Vermeylen, J. Traite de la Fabrication du Malt et de la Biere. Association Royale desAnciens Eleves de I’Institut Superieur des Fermentations, Gand. 1963
- Envoldsen, B. S., and Bathgate, G. N. Structural analysis of wort dextrins by means of Beta-amylases and the debranching enzymes pullanse. J. Inst. Brew. 75:433-443, 1969.
- MacGregor, A. W., Bazin, S. L., Maori, L. J., and Babb, J. C. Modeling the contribution of alpha-amylase, beta-amylase and limit dextrinase to starch degradation during mashing. J. Cereal Sci. 29:161-169,1999
- Lee, W., & Pyler, R. (1984). Barley Malt Limit Dextrinase: Varietal, Environmental, and Malting Effects. Journal of the American Society of Brewing Chemists, 42. doi:10.1094/asbj-42-0011
- The Impact of the Thermostability of alpha-Amylase, beta-Amylase, and Limit Dextrinase on Potential Wort Fermentability. (2003). Journal of the American Society of Brewing Chemists. doi:10.1094/asbcj-61-0210
- Lee, W., & Pyler, R. (1984). Barley Malt Limit Dextrinase: Varietal, Environmental, and Malting Effects. Journal of the American Society of Brewing Chemists, 42. doi:10.1094/asbj-42-0011
- Bréfort, H., Guinard, J., Buhlert, J. E., & Lewis, M. J. (1989). The Contribution Of Dextrins Of Beer Sensory Properties Part Ii. Aftertaste. Journal of the Institute of Brewing, 95(6), 431-435. doi:10.1002/j.2050-0416.1989.tb04651.x
- Hopkins, R. H., and Wiener, S. J. Inst. Brew. 61:488,1955.
- Walker, J. W., Bringhurst, T. A., Broadhead, A. L., Brosnan, J. M., & Pearson, S. Y. (2001). The Survival of Limit Dextrinase during Fermentation in the Production of Scotch Whisky. Journal of the Institute of Brewing, 107(2), 99-106. doi:10.1002/j.2050-0416.2001.tb00082.x
- Schnitzenbaumer, B., Kerpes, R., Titze, J., Jacob, F., and Arendt, E. K. (2012) Impact of various levels of unmalted oats (Avena sativa.) on the quality and processability of mashes, worts and beers, J. Am. Soc. Brew. Chem.