The pre-packaged ice cream market isn’t very friendly to low carb dieters. They offer us low fat frozen desserts that are full of sugar to compensate for the lack of butterfat. Oh sure, you can find some “no sugar added” products, but they are still high enough in carbohydrates to be considered verboten. The few that are truly sugar free are loaded with sugar alcohols that make your stomach do flip-flops and raise your insulin levels. Why the heck can’t there be at least one brand that’s high in fat yet contains no sugar? Unfortunately, I don’t see this getting better anytime soon. The industry is pushing the boundaries so far in the low fat/low cost direction that their products can’t legally be called ice cream anymore. I guess if I want something done right, I gotta do it myself…again.
Choosing a recipe
I decided the best way to start was to use a known good baseline recipe and then tweak it. The “Ben & Jerry's Homemade Ice Cream & Dessert Book” published in 1987 was the most credible source that I found. They use fresh ingredients and have simple recipes I can follow. It was this book that introduced me to the science (and art) of ice cream making. They use raw eggs in their recipes because the lecithin in the yolks acts as an emulsifier. The risk for salmonella is low, but using pasteurized (in the shell) eggs should alleviate that concern.
It’s usually a good idea to use multiple sweeteners in any recipe. They tend to work synergistically with each other, so less of each one need to be used. This helps avoid any aftertaste most artificial sweeteners have when used in large quantities.
Over the counter powdered sucralose (e.g., Spenda) contains dextrose and maltodextrin for bulking. One little packet contains an insignificant amount in carbohydrates. However, a cup of Splenda yields 24 grams of carbohydrate. This is why I like to use liquid sucralose. It’s truly 0 carbs and 0 calories.
Truvia, which is mostly erythritol, is my other usual choice. Erythritol is an excellent sugar alcohol because it doesn’t cause stomach distress and has a glycemic index of zero.
I don’t care for any of the other sugar alcohols except for maybe small amounts of xylitol. It’s used in some prepackaged low carb products like sugar free honey. Most of the other sugar alcohols usually cause gas and bloating and/or have significant glycemic effects. A particularly pervasive one is maltitol. It’s realistically no better than sucrose due to its relatively high glycemic response and low sweetness when compared to sucrose.
Freezing Point Depression
My first batch of ice cream was a partial success. I made French Vanilla with liquid sucralose instead of sugar. The taste was great, but scooping it was nearly impossible. It was hard as a rock!
My background is not in food science, but I am intrigued by the chemistry involved in making ice cream. I was able to get a second-hand copy of “Ice Cream” by Goff and Hartel and I read it cover to cover. The amount of information for sugar free ice cream making is relatively scant, but on page 441, I discovered an interesting study:
Whelan et al. ( 2008 ) examined a number of polyol sweeteners in low glycemic index formulations. Once the freezing curves were matched, other physicochemical properties also were found to match.Ah! If I can achieve the same freezing properties as sugar, then my ice cream will be a success! The researchers used a combination of sugar alcohols, polydextrose, and inulin. While they did achieve an acceptable taste and texture, the resulting mixes described in the paper wouldn’t be to my satisfaction. The potential glycemic response and intestinal distress weren’t worth it. Looking back at my first batch of ice cream, using erythritol would have resulted in a much more scoopable product at low temperatures. Erythritol is 2.8x as effective at lowering the freezing point compared to sugar. The ratio of sugar to Truvia is 8:1, so the freezing point depression problem can’t be solved by using it alone.
I created a spreadsheet to calculate the freezing point depression of a recipe. (I’ll explain how it works in a future post, but feel free to muck around with it.) If I can match the freezing curve of the regular recipe, then I’d have a shot at making a viable sugar free ice cream with the same mouthfeel and melting characteristics. What I needed were soluble ingredients that have a very low molecular weight to depress the freezing point without affecting flavor or insulin response. One obvious additive is salt. It’s almost six times as effective as sugar. Two recipes, butter pecan and chocolate, both benefit from ½ tsp of sea salt. I suspect that a little added salt won't negatively impact flavor in any of the mixes that were written as salt free. (I’ve eliminated whole milk completely from the recipes, so the salts from non-milk fat solids are much lower anyway.) The Ben and Jerry’s book mentions adding alcohol for flavor along with a warning not to add too much. (I later calculated that it’s 7.4 times as effective as sucrose.) It’s edible antifreeze! Vanilla extract is approximately 34% alcohol, so a little will certainly help. However, I still needed something else to fill the gap. Eventually, I stumbled upon glycerin.
Glycerin (a.k.a. glycerol) is a controversial additive in the low carb universe. Recent research has shown that glycerin does not significantly elevate blood insulin levels and only minimally elevates blood sugar levels. I’m looking into this further and plan to make a more in depth post on glycerin. For now, let’s assume the results from that research are true.
What makes glycerin so attractive is its low molecular weight. It’s about 3.7 times as effective as sucrose in lowering the freezing point. Therefore, it would only take a modest amount (1 tbsp. / 18.9g) to provide equivalent melting characteristics to a sugar free ice cream mix made with Truvia. I was able to find pure vegetable glycerin on Amazon.com.
Stabilizers are important because they prevent large ice crystal from forming. They also allow the ice cream to keep its shape when melting instead of turning into a puddle of flavored milk.
I have been using xanthan gum as a stabilizer, but according to page 80 of the Goff and Hartel book, it’s not ideal. I found Cremodan 30, but that requires cooking (and I believe there needs to be sugar in the mix). It contains carrageenan, which is great for making ice cream, but not so great as a food additive. I'd be very interested in trying unflavored gelatin as a stabilizer, but there's very little information available on the Internet on how much to use and how to prepare it. I tried Fibersol 2, which is a non-digestible dextrin. It's completely soluble in liquid in high concentrations, has a glycemic index of approximately 5, has a good laxation threshold (70g), and doesn't alter taste. There’s even some preliminary evidence that it may provide a lot of the same benefits of resistant starch that’s so fashionable nowadays. It does perform well as a stabilizer at 50g per 1,000g (i.e., 5%) of mix, but it caused too much gas and bloating for me to continue using it.
Quite frankly, ¼ tsp of xanthan gum works pretty darn well and is easy to find in the supermarket. I plan on trying a mix of gums to see if they will work better. Maybe one day, I’ll give unflavored gelatin a shot.
The next few posts on this topic will discuss into the equipment, include a detailed recipe for Butter Pecan, a breakdown of my freezing point depression calculator spreadsheet and more ice cream science!