posted by Dave Arnold
I love puffed snacks. Everyone loves puffed snacks. I thought it was time we posted on them.
Why do things puff?
In the snack food industry, most puffed snacks are made from starch passed through an extruder. A screw inside a long barrel mixes, compresses and cooks via friction all at the same time (sometimes extra heat is added, but often friction alone does the trick). By the time the screw reaches the die at the end of the barrel the product mixture is very hot (well over 100°C) and under high pressure. When the product is extruded through the die at the end of the barrel this pressure is released and water boils off very rapidly - puffing and drying at the same time. We don’t have an extruder at school, but that doesn’t keep us from making puffed treats.
The products we puff in the kitchen are technically glasses: amorphous (non crystalline) homogenous solids that are stable and rigid. They also contain water (ideally about 12%). All glasses have what’s called a glass transition temperature at which they go from rigid to rubbery. This temperature is usually a good bit (like 100°C) lower than the melting point where they’d become a liquid. For foods destined to be puffed snacks, the glass transition temperature is ALWAYS higher than the boiling point of water –here’s why: When you heat the glassy pre-snack rapidly above 100°C the water inside wants to boil but can’t because it is trapped inside a glass. When the snack reaches the glass transition temp it suddenly becomes rubbery. The water is able to start boiling and expanding, which it does rapidly. After the water boils away the product sets (because there’s no more water to keep it plastic and flexible) and becomes crunchy.
Points to remember if you want to do your own puffing:
- The product you start with must be amorphous. Crystals won’t puff. As you will see, that is an important point to remember.
- The water content must be correct. Too little water results in burning and less puffing because 1) the temperature of the item will get too high before the glass transition is reached; 2) there is insufficient water to puff adequately; 3) without enough water the snack won’t be flexible enough to expand. In the biz they say water acts as a plasticizer.
- Other substances can also act as plasticizers – such as salt (which also helps taste and might help with heat conduction) and maltodextrin – any relatively small molecule that will decrease the viscosity of your product.
- Heat quickly and evenly. I like frying, but air popping, heat guns and microwaves work too.
- Don’t add things to the snack that will burn before the item is fully puffed –like sugar.
- Don’t add things that will compete with the water in the snack and prevent puffing –like sugar.
Things that will puff fall into two main categories (that I know of): connective tissue and starches.
Connective tissue has lots of collagen. When you cook collagen you break it into to gelatin, which is soft and easy to puff. When you dehydrate the broken-down connective tissue it turns into a glass, which you can heat and puff — a la pork rinds. Wylie Dufresne works similar wonders with beef tendon and cod swim bladders.
Starches are made up of two main molecules: amylopectin and amylose, usually with more of the former. Industrial puffing people care a lot about the amylose to amylopectin ratio. In extrusion applications, about 50/50 gives the highest expansion. The main structure comes from Amylopectin, while the amylose helps to increase the fluidity of the mix and increase expansion. The ratio isn’t so important to us. What is important is that the starch gets cooked thoroughly. Native starch is partially crystalline and won’t puff. The starch granules need to be fully cooked to ensure that the crystal nature is disrupted and the starch can become nice and amorphous. Some chefs (like Wylie) use pre-gelatinized starch to get around this problem. Mix with liquid, dehydrate, and you’re done. If you want to go this rought use Ultra-Sperse from the National Starch Corporation. Don’t bother with Ultra-Tex – it’s hard to hydrate properly.
Almost anything that is mostly starch will puff: rice, corn, pasta, doughs made from tapioca starch (nice and bland), wheat starch, corn starch, yadda yadda.
How do you make a puffed snack?
Making puffs is as simple as: over-cook, dehydrate, and fry.
First: Over-cook your item. In the case of starch this ensures that the starch granules are fully gelatinized and at their lowest viscosity. In connective tissue like pork rinds, overcooking ensures that all the connective tissue is broken down and soft. Cook pork rinds in very salty water for 60-70 minutes, pasta in heavily salted water for 45 minutes.
Second: Dehydrate. If the moisture content of your item is above 15% you won’t get much puffing. If it is below about 10% you won’t get much puffing. Most people don’t have the ability to test moisture content (If you are anal and you know your moisture level is close you can get it just right by sealing your stuff in a container with a saturated solution of ammonium nitrate or sodium nitrite in water. That will maintain 65% relative humidity which will give a moisture content of 12%). In practice determining the moisture content isn’t usually a problem. When the product goes from being flexible to being plastic-y and shrinky-dink like — you’re done. Try puffing a couple and see what happens. Over-dried stuff doesn’t bubble much when fried and burns without puffing. Under-dried stuff starts to puff and boils a lot but has a hard center that doesn’t puff. You’ll quickly get the hang of it. I recommend putting your product in the dehydrator at 135°F for a couple hours and then turning the dehydrator off and leaving the product overnight. In our kitchen this usually works. If you live in a very humid or very dry place your results may vary (sorry Louisiana and Arizona). If you don’t have a dehydrator turn your oven on low and crack the door.
Here are pictures of some stuff puffing in a microwave ( thus the blurry pictures).
Here is the technique for puffing pasta. No, you can’t just puff dried pasta, you have to cook the heck out of it and then dry it. I built a rack to hold the pasta while it was cooking and dehydrating because it was so fragile it wouldn’t hold its shape. The pasta almost breaks under its own weight it’s so soft. You don’t need the rack if you don’t want to keep the holes in the pasta intact. After dehydration the pasta should have almost the texture it had before it was first cooked –just a little more flexible.

Put pasta on the special rack, salt the hell out of some water, boil the hell out of the pasta (like 40 minutes), load the rack into the dehydrator.
Here is the recipe for pork rinds:
Take pork skin and boil it in heavily salted water for 60-75 minutes. Carefully drain and cool in the fridge. After cold it is sturdy enough to handle. Scrape the fat off the fat side; more scraping = better puffing. Cut the pork rinds into pieces and dehydrate until they feel like plastic. Fry in very hot oil.





24 responses so far ↓
Darren teoh // October 28, 2009 at 9:47 am |
Hey man, funny you should mention it. We just made some rice puffs this week for class.
novalis // October 28, 2009 at 10:13 am |
Couldn’t you check moisture content by weight? Weigh the product before hydrating, then weigh periodically during dehydration. When dehydrated weight = prehydrated weight plus 12%, you’re done.
I guess that might not work for pasta, because you probably lose some starch to the cooking liquid, but it would probably work for pork rinds or especially ultrasperse-based puffs, right?
Dave A // October 28, 2009 at 10:33 am |
I was thinking about that. The problem is you’d need to know the initial moisture content of your flour, etc.
novalis // October 28, 2009 at 10:52 am |
Dehydrate the flour first?
Dave A // October 28, 2009 at 12:02 pm |
Maybe.
Jason // October 28, 2009 at 11:03 am |
per the pregelatinized starch bit (ultra-sperse: yes, ultratex: no) what about Wondra flour?
Arguably the name isn’t quite as high-tech (although it does sound magical!), but it’s pregelatinized and designed to hydrate easily.
Chocolatesa // October 28, 2009 at 11:21 am |
Wow that sounds really cool, AND something I could potentially do at home!
Stephen // October 28, 2009 at 4:33 pm |
The rind of cheese puffs up into a light crisp puffy cheesy thing when microwaved, haven’t experimented yet but was wondering if any slice of cheese (hard cheese such as Parmesan) dehydrated would puff, or are there other factors at work creating the rind
Dave A // October 28, 2009 at 4:53 pm |
Interesting. I don’t know. I read a study on puffing imitation cheese, but it only worked on low-fat/high-starch fake cheeses. Keep me posted on puffing real cheese.
Stephen // October 30, 2009 at 2:19 am
Tried an experiment with 3 slices of Parmesan cheese (different thickness, thin to about 1.5mm).
Dehydrated them, then put them in the microwave for 1 minute.
The thinnest one didn’t really puff but but was super crisp.
Next step up puffed up but was quite dense, crispy though.
The thickest puffed up the best, good air bubbles inside, very crisp, and intense flavour.
see pictures here
http://www.flickr.com/photos/30576463@N07/4057066597/
(please excuse quality but had to take them on cell phone as camera is on holiday)
I think next, I’ll try a cheese with a higher initial water content.
Stephen // October 31, 2009 at 12:10 am
sorry new user to flicker, I had accidentally set the photo to private oops! All fixed now.
http://www.flickr.com/photos/30576463@N07/4057066597/
Chocolatesa // November 2, 2009 at 5:53 pm |
Thanks
That looks yummy, I’ll have to try that myself.
chadzilla // October 28, 2009 at 7:30 pm |
Awesome… all the way down to the little fork stand thing.
Funny that we’ve been microwaving pork rinds in Louisiana for years, but never really thought it through further… cajun mind block.
AGP // October 28, 2009 at 8:19 pm |
Haha! Jeremie showed me how to do this several months ago and now we use the pasta for our bar snack! Boil orciette pasta for 45 minutes, dehydrate it overnight in my convection oven, and deep fry it at 375. Wer toss our with smoked tomato powder!! YUM!!!!
Wylie Dufresne // October 29, 2009 at 1:27 am |
if you want to puff real cheese, try microwaving parmesan rind…
jeremiah bullfrog // October 30, 2009 at 1:46 am |
I’m gonna go make chocolate pasta..then puff it up.
Chocolatesa // October 30, 2009 at 9:36 pm |
Stephen, the flikr pic doesn’t work, it tells me it’s a private page…
schinderhannes // November 3, 2009 at 12:37 pm |
Off topic:
it seems impossible to leave a comment on the ABV testing! One always gets to a 404 page.
thus zip comments so far.
Here is mine, you can either hang it there or I´ll restate it later:
I would try to keep as many parameters under control and constant as possible.
The main one seems to be concentration of taste contributing ingredients.
Why don´t you try to make an extra concentrated dink (lots a lime etc) and then dilute always with the same amount of a neutral liquid. Either straight alcohol (200 proof from the pharmacy) or pure water or mixtures thereof.
Thus in a pretty good approximation the only parameter changed is ABV.
The art would be to make these drinks still taste half good. So I would start with a drink normally containing Vodka and replace that by an 80 proof mixture of water and alcohol. Then leave it out to get your stock solution.
Schinderhannes
Danielle // November 5, 2009 at 2:17 pm |
Have you ever seen those Hidden Valley commercials where there’s cauliflower popped like popcorn? It was on TV last night and I was just thinking how awesome it would be if you could have puffed cauliflower with ranch dressing.
Dave A // November 5, 2009 at 5:50 pm |
I haven’t. Wow.
hylan Joseph // November 9, 2009 at 8:57 pm |
try steaming the pasta in a pressure cooker first,than frying it. A company I worked for years ago was trying to develop a pasta snack chip and this yields a pretty awesome puffed product.
nastassiar8se // November 11, 2009 at 6:43 pm |
Great tip. Can’t believe I didn’t think of that.
Jon // December 28, 2009 at 6:16 pm |
So do you recommend using Ultra-Sperse rather than tapioca flour which is what a lot of other chefs use? Are there any percentages for Ultra-Sperse, or is it just by a “when it feels right” kind of way?
Dave A // December 29, 2009 at 3:12 am |
Hi Jon,
Tapioca starch is made into a dough which needs to be fully cooked –usually by steaming. That technique works great. Ultra-Sperse is just made into a dough without cooking and is then dehydrated. Unfortunately, we just do it till it feels right. Janet Carver over at National Starch, the makers of Ultra-Sperse, might have a more accurate recipe.