It was a bit late, but I wanted to get done with the soapmaking so I could start concentrating on posting this story. Only three oils remained. The first was jojoba:
Jojoba oil is weird. It’s a liquid, and not a particularly viscous one (I think this has to do with the degree of saturation), but the molecules are ridiculously long–about forty carbons each. For perspective, most oils are somewhere in the mid-teens; forty is more typical of a wax than an oil.
The saponification number is also a tiny 0.069, which stands to reason: if the molecules are so huge, then there will be fewer of them to saponify, right? Blah blah blah molar blah blah blah concentration blah blah blah molecular mass blah blah blah stochiometry. BRILLIANT! Oh, man, it is a great mystery of life why I do not have a Nobel and a Pulitzer on my wall.
I only used 80 grams of jojoba oil, because I wanted to have some left over. So the recipe was:
80 g jojoba oil
30 g water
5.5 g of lye
The soap had hardly thickened at all. I poured it into the trays, hoping for the best.
My next victim was Crisco.
Crisco is odd because it’s really a mixture of oils:
The webpage I saw recommended using a factor of .136, so that’s what I’m using:
100 g Crisco
38 g water
13.6 g lye
The Crisco soap was watery when I poured it, but I’d heard it made a fairly hard bar. We’ll see.
Now it was time for the oil that got me into this mess in the first place: Trader Joe’s Lavender Body Oil.
This is another mixture of oils:
As MOTD surmised, this isn’t really lavender oil–"lavender-scented oil" would be more accurate, as the (non-saponifiable) compounds from the lavender are dissolved in a much larger quantity of so-called "carrier oils." Happily, the first two carrier oils on the list (safflower and sweet almond) have identical saponification values of 0.136, and the third (sesame) is only fractionally lower at 0.133.
100 g oil
38 g water
13.6 g lye.
Generally, fragrances and colorants are added to soap at the end of the process (when they trace) because the lye tends to react with them and destroy whatever property you hoped they would lend. I didn’t notice that here, although I think the smell was subtly different.
The Crisco was starting to set up nicely. I poured the lavender into the molds, and doffed my goggles.
It was time for a soap family portrait:
(If you click on the photo, you can see the annotated version at Flickr.)
The jojoba looked very discouraging–to be honest, I didn’t have particularly high hopes for it, which is another reason I didn’t do as much of it. Perhaps this was a self-fulfilling prophecy, but soap and students work in different ways. I think I was just too close to the margin of error on my scale when I was making it. It’s probably only in the saponification tables to let people know they need to take it into account when they add 5 or 10% of it to a relatively large batch.
I decided to fix the blender with epoxy resin. Epoxy is really interesting stuff, almost like a plastic analog of concrete (which is my nominee for Most Underappreciated Material, but that’s another story). This page has a nice explanation of how they work.
I mixed the resins together and glopped it onto the housing. I knew I’d have to sand it a bit so the blades could rotate again, but that wouldn’t be too difficult.
J-B Weld takes about 24 hours to set up. I didn’t feel like waiting that long, and I didn’t want to do the Shake-N-Bake thing I did last night, so I looked around my apartment for something else I could use. Cookwares came to my rescue yet again–I had a one-pint Pyrex measuring cup that was just the thing.
But first I wanted to get out of my apartment for a while to let the epoxy set, because the vapors are kinda nasty. I went to Progressive Grounds, in the Mission, and found material for a blog entry on the way.
By the time I got back, I was kinda tired, and I didn’t want to deal with making soap again. I had the next day off, and I figured I’d be able to give my new soapmaking apparatus a workout then, along with finishing up my blender repair. So, the next day, I made sesame-oil soap.
I used TJ’s Toasted Sesame Oil. It had a strong, robust, nutty smell.
The saponification factor was 0.133. Using the hand mixer was harder than the blender, but I think the soap thickened. We’ll see.
The cells in the blue ice cube trays I had moved to were much smaller than those in the white ones I was using before. That’s the sesame oil in the middle; the four rightmost ones have the tail end of the flax soap.
A friend of mine suggested that I try butter, and warned me of the foul smell of butyric acid. Butyric is one of the shorter fatty acids, having only four carbon atoms. Dylan "Just Like Bacon" Stiles has a nice little graph of how the alkanoic acids smell, of which I think Edward Tufte would mostly approve. Anyway, butyric acid is called that because it’s found in butter–it’s from the Greek βουτυ�?ος, meaning butter.
Curiously, while butyric acid itself smells vile, butyric acid esters smell pretty nice–in the case I linked to, like Juicy Fruit gum. (Such nice scents are fairly typical of esters; they’re a very large and versatile part of the artificial-flavoring artist’s arsenal.)
As long as I’m on a food-science tear, let me recommend a book: Harold McGee’sOn Food and Cooking. McGee has the heart of a gourmet, the mind of a scientist, the knowledge of a pedant, the curiosity of a toddler, and the expository power of a popular-science writer of the first water. If you dig food or science, you should get a copy; if you dig food and science, you must get a copy. :-)
Actually, he also looks a great deal like my Contracts professor, now that I think about it. Hmm. Anyway, he also has a blog with nice little food-science tidbits–thanks to my little bro’ Jase (who’s quite a good writer himself, come to that) for pointing it out.
What was this series of blog entries about again? Oh, right! Soap! Made from butter!
So I decided I’d use plain butter, and then try clarified butter, or ghee. (When I mentioned this to Dinah, she said "Ghee soap! That sounds interesting!")
Butter is the first soapmaking material I used that wasn’t almost pure fat. There are other significant components, chiefly water and protein.
14 grams of butter had 11 grams of fat in it, so for 100 grams of fat, I’d need to use 100 * 14 / 11 = 127 grams of butter. The remainder would be mostly water, with some proteins (USDA rounding rules let them say zero grams on the label, I guess), cholesterol, and whatnot. (Do you think the label should say things like "WhatnotNg"? I do.) Anyway, the melted butter looked like this:
, the process of mixing the soap looked like this:
, and the final soap looked like this:
By the way, I got a tripod at Goodwill this weekend, so I’ll be able to make videos of soapmaking–I’ve got a notion in my head about making a short, three-minute instructional piece in iMovie. I haven’t used iMovie before, but hey, gotta learn somehow. Anyway, ghee.
Ghee is clarified butter; butter without those other components I mentioned above. You can buy jars of it at many stores, especially Indian grocery stores, but making it yourself from any old butter is easy. First you melt the butter (driving off the water in the process):
Quite simple, really. The advantage of using ghee is that it has a much higher smoke point (no milk solids or proteins to burn), which means you can fry foods at much hotter temperatures without burning your frying medium. Grapeseed oil is also known for having a high smoke point.
Happily, I had made almost exactly 100 g of ghee. I didn’t record the starting mass, but I used one whole stick and most of another, which would be just under 200 g. I don’t know whether this yield is typical; it seems a little low to me, but then again, I hadn’t made ghee before. Skimming off the solids was the hardest part; I lost a lot of fat there. The best technique turned out to be "herding" the solids to one side of the container so I could scoop up more of it at a time. Here’s the finished product:
Unsurprisingly, the ghee soap looked and poured a lot like the butter did. I put some ghee soap in the upper left of the top tray, with the butter soap in the right part. In the blue tray, from left to right, it’s ghee, sesame, and flax (remember flax?).
The next night I took some of the soaps out of the trays. The lard soap was nice and hard:
whereas the canola oil was doughy:
The peanut oil was much softer underneath the surface; it had developed a sort of crust:
The grapeseed oil had an even more pronounced crust:
Oh, and there’s the bit of the blender I lost last time!
The flax oil was just kinda weird and grainy:
I also checked in on the other soaps. The olive oil soap was surprisingly hard at this point. The walnut was softer, but it had set up appreciably.
Nothing too crazy to report with respect to everything else. The most–let’s say "distinctive"–one is the sesame oil soap. It reminds me a bit of a lab we had in high school where we dissected squid. As part of this, the teachers prepared calamari over a Bunsen burner. On these days, you’d walk into the science wing and think, "Oh, that smells good." And then about ten minutes later you’d think, "Ok, I don’t want to smell that for a while now." It’s much the same phenomenon with the soap.
The butter and ghee soaps have a slightly buttery smell, and I think it’s a little less pronounced in the ghee soap. We’ll see what happens as they set up.
This peanut oil has vitamin E added to it. No idea how that’s going to change things; I seem to recall something about it acting as a preservative or perhaps accelerating curing. Recipe:
100 g peanut oil
38 g water
13.6 g lye.
By the way, the 0.38 factor I’m using for the water comes from this page on the Walton Feed soapmaking site. People who are more interested in the chemistry of soap should look at this page on that site by James Hershberger, a soapmaker and chemical engineer. Anyway, speaking of that 0.38 factor . . .
I have now twice hit 38 grams on the first try when pouring a glug of water out of my pitcher. This is a little weird.
The soap had a weird sort of greenish cast to it. It was taking forever to firm up . . .
Oh, no! Some kind of weird black goop was seeping out the bottom of the pitcher!
I poured the soap into the tray, hoping it was still good, or at least that I might be able to salvage it later.
It was time for a thorough cleaning of the blender. This is actually the most time-consuming part of the experiment.
I’m not sure what the goop was from. I’m leaning towards soap that somehow leaked out of the blade assembly.
I wanted to get at least one more soap in before I went to bed. I understand that soaps made with grapeseed oil set up faster. A 100% grapeseed oil soap . . . well, we will see.
100 g grapeseed oil
38 g water
12.65 g lye
The grapeseed oil had an aroma between nutty and musty. I’d just bought it a couple days ago, though, so it couldn’t have gone rancid.
The soap did thicken fairly soon for a soap made with such a light oil. It was an encouraging thing to see.
Here it is, at the lower right of the tray. On the left is the lard soap, and then the canola soap, and then the peanut soap:
Substantially less encouraging, though, was more of that black goo:
Most soapmakers who use blenders don’t use the pitcher-based variety I have. They use what they call a "stick blender," which is basically my blender turned upside-down. In retrospect, this makes more sense: you can use the blender in whatever sort of vessel pleases you, and needn’t worry whatever you’re blending befouling the bearings and stuff. "Blinkin’ barnacles! Th’ bloody bearings ‘re befouled wi’ blended beetles!"
I thought I saw a lump of something make its way into the grapeseed soap as I was pouring it into the mold. As I was disassembled the blender pitcher, I saw the source of the lump, and probably, of the goo:
Well, shoot. Maybe I’ll try to do something about that tomorrow. For now, I’d try something different–shaking my soap in a Nalgene bottle. This has something of a modern-hippie aspect to it, but hey, I have something of a modern-hippie aspect myself on good days. And I had a modern-hippie-style oil: flax!
100 g flax oil
38 g water
13.57 g lye
I went a couple grams long on the flax oil, because it said it had a few extra ingredients in it. Also, if you look at one of the larger versions of the picture on Flickr, you can see that they explicitly name-check a bunch of fatty acids: stearic, lauric, palmitic, alpha-linolenic (an omega-3 fatty acid), and linoleic (omega-6), as well as the class of omega-9 fatty acids. You might recognize these roots in the names of some of the salts mentioned on MotD.
Here’s a video of me shaking the Nalgene bottle:
Yes, fully handmade. I even used a bottle from Powells Books.
I had a little extra soap, so I poured it into another ice-cube tray after filling up the old one. It was very, very thin, possibly even worse than the peanut.
I believe the idea behind putting bits of flaxseed in the oil was to implement something analogous to "flavor crystals" by introducing an anisotropy into a usually very isotropic product. Some people like their soap slightly anisotropic (bits of herbs and the like floating around in there–the "Additive Type" mentioned in this page on the types of soapers), but as I understand, you usually add those things after the soap has gotten a bit thicker, and you never end up with them sitting around in the jar after you’ve shaken the soap. Then again, if you’re blending your soap by shaking a Nalgene bottle, you’re not exactly an exemplar of the state of the soapmaking art, no matter how cool your blog is:
The next day, the grapeseed soap had developed beads of some sort of clear fluid, which I wasn’t expecting. None of the other soaps did this, and I have no explanation. If you do, I’d be interested:
Katie was in town from Sacramento, so I went to Café du Soleil with her. Afterwards, we went to 826 Valencia and a nifty bookstore in the Mission.
When I returned, I took a closer look at the blade assembly, and hatched a plan for fixing it. More on that soon . . .
We haven’t made an animal-fat soap yet. Let’s try lard.
When the Black Table kids made soap, they started from bacon, and rendered the fat themselves, but cautioned readers that the procedure "smell[s] like a short-order cook after a triple shift." It is uncertain whether bacon bits have the same problem. Anyway, I decided to skip the whole bacon thing and use lard.
If you want to get lard in San Francisco, you have three options:
Safeway or another supermarket (BORING!)
A Mexican grocery store (the Spanish word is "manteca,") (Better, but not quite as cool as–)
I melted the lard in my microwave. There were some impurities in it:
While these are probably valued among pirates, I removed them. This probably shows my middle-class suburban bourgeois sensibilities, but unidentified lumps of stuff in a substance which is supposed to make me clean doesn’t really fit in my world-view. If you want to make a postmodern soap, you could use unfiltered lard and make some kind of deconstructionist argument about how the very "cleansing of the cleaner" before it is made cheapens the transformative aspect of this folk-art undertaking. If you’ve actually read Derrida, feel free to comment with further contributions to postmodern critical soap theory.
Anyway. Fortunately, I know about filtration from experience. And I had all the appropriate apparatus to hand:
The lard had a faintly musty odor. Perhaps it had absorbed some of the smell of the wooden lard vat.
It took surprisingly long for the lard to strain through the coffee filters. I think it might have been freezing on the filter paper, slowing things down.
Finally, my wait was rewarded with clean lard:
The finished lard soap looked like this:
Next up: canola oil.
Canola oil is one of the newer additions to the human cupboard, but it comes from a long-domesticated plant: rapeseed. It was originally grown for lamp oil, and was used as a lubricant for steam equipment in the Industrial Revolution. Navies required a great deal of rapeseed oil for this purpose during World War II. Canadian farmers started growing it then, but they needed to find other uses for it when the war ended.
The problem with regular rapeseed oil is that it contains erucic acid, which is believed to cause cancer. Also, it has a lot of glucosinolates, strong-tasting compounds that make the oil unsuitable for either animal or human food.
Through selective breeding, Canadian crop scientists developed strains of rapeseed that were low in both erucic acid and glucosinolates. They called it canola, from CANada Oil, Low Acid. (Note the country-of-origin information on the label above.)
While I usually disapprove of coining new words for old or only marginally-new things, I really can’t hold it against the makers of canola oil. How would you like to be responsible for selling thousands and thousands of bottles of cooking oil, all of which are prominently labeled "RAPE"? "Hey, I’ve got a lot of ‘Nonconsensual and Unlawful Carnal Knowledge Oil’ up ins; ya wanna buy any of it?" Doesn’t work.
Incidentally, the plant’s name comes from the Latin word for turnip, rapum. The turnip, rapeseed, rutabaga, cabbage, cauliflower, broccoli, and mustard plants are all members of the genus Brassica. There’s also an interesting story behind the genetics of the plants, but this is already getting pretty far afield.
I used 150 g of canola oil, because I wanted to see how a larger batch would turn out, and because I had a whole liter of the stuff, which was even cheaper than low-end olive oil.
Canola oil’s saponification factor is 0.134:
150 g canola oil
57 g water
20.1 g lye.
Mustard is the extrovert of genus Brassica. Most of its other members are rather bland, and canola oil carries on the tradition. It’s almost eerily scentless–it smells slightly oily, but that’s it.
Nothing too spectacular happened, but it did get harder than I was expecting it to. Having learned from my early olive-oil experiment, I ran the blender for a longer time. The soapmaking term for when the blend thickens is "trace," because the stirring spoon would leave traces in its path in the soap vat. Soaps without solid fats of some kind (tallow, lard, palm oil, coconut oil, or the like) take a long time to trace. Using a blender instead of a spoon makes things much faster, as does adding a bit of solid fat. Actually, now that I think about that, it also could be the case that I didn’t wash out all the lard from the prior batch. Oh, well.
The lard soap was already starting to set up a bit. This is pretty consistent with what I know of lard soap; it makes a fairly hard bar. I tried to get video of me pouring the canola oil soap into the tray, but it didn’t come out too well–pouring soap glop, steadying the tray, and trying to keep the soap glop in the tray and off your body doesn’t make for the best cinematography. Maybe I can bug Joe into coming over and helping out.
I thought I was just going to make single-oil soaps, but coconut and olive oil soaps have interestingly complementary properties, so perhaps half-and-half . . .
50 g olive oil
50 g coconut oil
38 g water
16.2 g lye (the arithmetic mean of the numbers used in the last entry)
The mixed oil looked like this:
,
and the lye looked like this:
.
For completeness, the plain lye looked like this:
Earlier, Lauren asked whether you just dumped everything in the blender and turned it on–and yep, that’s about all there is to it. There are only a couple caveats:
When you mix the lye into the water, it will heat up. Let it cool down a bit before you mix it into the oil. About 90 or 100 degrees Fahrenheit is good. This doesn’t matter too much, as long as you don’t dump it into the oil immediately after adding the lye.
Add the lye solution to the oil, and not the other way around.
The lye solution gives off nasty vapors. They don’t have a smell as such, just a stinging sensation. Have a window open, don’t lean over the lye solution, and you should be fine.
You must wear goggles and look like an Intepid Soap Warrior from an anime cartoon, comme ça:
As for this soap, it was about halfway between the two oils, in the amount of time it took to thicken, the ease of pouring it into the tray, and in color and smell. As for the others, the olive-oil soap was a little bit stiffer, while the coconut-oil one was already quite hard.
When it was in the blender, it looked like this:
And, after I poured it, like this:
Well, so much for the first few classic soapmaking oils. Part of the fun here was going to be using really strange things, so let’s start with . . . walnut!
I just noticed that this can is labeled “ROASTED WALNUT OiL”. The funky capitalization makes me think of “BiL”, the fannish abbreviation for Minneapolitan folk-rockers Boiled in Lead. Anyway.
The can gives it a sort of wood-shop feel. Unsurprisingly, walnut oil has a nutty smell.
Walnut oil’s saponification value is a fairly standard 0.135:
100 g walnut oil
38 g water
13.5 g lye.
I was expecting the walnut oil to be darker in color, for some reason. Maybe it was the woodshop metaphor.
Joe and Jane came over shortly before I made the walnut oil soap. Once it was in the tray, we went out to dinner and to meet up with Robin, who was in town for a family reunion.
After I got back, I cracked off a bit of the coconut-oil soap that was hanging over the edge of the tray and washed my hands with it. It worked! I made soap! This is really cool!
The next day, the mustardy color of the olive-oil soap had abated somewhat, and it was a bit harder. Here’s a picture of the tray:
I took the soaps out of the ice-cube tray. The coconut-oil and coconut-olive soaps popped right out with a little persuasion. I didn’t pop the others out as much as I scooped them, with a knife.
The olive-oil soap was the consistency of soft peanut butter. The walnut-oil soap was harder, but not by much.
I spread these two out on a plate in hopes that I could dry them out a bit. (By the way, my drying rack is also made by Ingvar Kamprad Scientific Supply.)
I cleaned up shop and went on to to the next day’s soapmaking.
I decided to start with olive oil, because I had a bunch of it, it was relatively cheap (especially for the lower grades–you get no benefit from extra-virgin in soap; I just used it because I had it around), and it makes a classic soap.
You’ll frequently see people advertising “Castile” soap. This comes from the days when Castilian soapmakers had a lot more olive oil than they did tallow (the standard soapmaking fat in those days), and figured out that they could make some very mild, high-quality soap with it. These days, people add all sorts of other stuff, but all castile soaps are based on olive oil.
Olive oil’s saponification number is 0.134, so the recipe will be:
100 g olive oil
38 g water
13.4 g lye.
My scale only reads down to 1 gram precision, so I stuck with 13 grams to avoid having too much lye in the soap. I measured out ingredients in a beaker from Ingvar Kamprad Scientific Supply.
After a few minutes, I decided to call it. It looked a little thicker, and that’s about all you can hope for with 100% olive.
I poured it into the ice-cube tray I had waiting. The soap was the color of mustard, which shouldn’t have been all that surprising, but the resemblance was striking.
My next victim was the anti-olive oil: coconut.
Olive oil
Coconut oil
Makes very soft soap
Makes very hard soap
Soap is skin-friendly
Soap dries out skin
Soap lathers very little
Soap is very sudsy
14% saturated fat
86% saturated fat (!)
Supposed to be healthy for you (Mediterranean diet, antioxidants, etc.)
Will leap out of jar and jam your aorta like the 101 during rush hour if provoked
Liquid at room temperature
Solid at room temperature, and actually forms largish crystals to boot
Here’s a picture of the end of the jar, showing the crystals:
And yes, it was much harder. It was almost a slurry when I poured it out of the blender, and I didn’t run it for nearly as long. Oh, ingredients:
100 g coconut oil
38 g water
19 g lye (saponification number 0.190)
Here’s what the liquid coconut oil looked like after I stuck it in the microwave for a while:
And here’s the final product. The olive-oil soap was a little darker, but not quite as dark as it looks in this picture. The coconut-oil soap looks like it’s at least firm in this shot, and it is. It sets up very hard, and very fast:
There was one kind of berry, a dark red.
I tried it, one by one, and hours apart.
Sub-acid, and not bad, no ill effects;
and so I made home-brew. . . .
Home-made, home-made! But aren’t we all?
I felt a deep affection for
the smallest of my island industries.
(You should actually read the whole poem, as it’s very good. I love how smoothly Bishop’s poetry reads–is that the sort of thing English majors say? In any event, I’m digressing before I start.)
Some time ago, a friend of mine pointed me towards Trader Joe’s Lavender Body Oil. I don’t remember exactly how I came to this thought, but I eventually decided it would be interesting to make soap from this oil, and possibly from others . . .
Eventually I accumulated quite a few “others”:
And so here we go! I’m going to make soap from all the oils in this picture. It’ll be an adventure; my approach will be much closer to Cockeyed.comScience Club than to Angewandte Chemie. (Frankly, I think Rob has much more fun than the Ang. Chem. guys, but don’t tell them that.)
First, some background. I’m going to radically oversimplify the science here, and will probably get some things wrong, but this is essentially what happens.
What we call “fat” is, chemically speaking, almost all triglycerides–molecules formed of a glycerine backbone with three fatty acids attached. If you add water and a base to fat, the fat undergoes a process called hydrolysis, where the glycerine separates out and the fatty acids bond with the base, giving fatty acid salts. This is called saponification. (Check out that last link; it’s from Molecule of the Day, which is made by a guy who actually knows this stuff, and gives a great overview of the saponification process.)
Anyway, saponification is extremely useful, especially for those fatty-acid salts we get out of them. Those salts have split personalities: one end is strongly attracted to water, and the other is strongly attracted to oil. (I’m oversimplifying, as I mentioned.) This means that these molecules are very good at lifting oils and other non-water-like things out of whatever they’re in and into the water, so you can wash them away. These bundles of molecules are called micelles. (Also MOTD, and also worth a look.)
The practical upshot of all this is 1) you can make soap if you add a base to a fat and 2) you can clean things with said soap. The only real gotcha here is that the best bases for making soap, like lye, are not terribly friendly to the skin, as anyone who has seen Fight Club can testify. (The first time I saw Fight Club, I was in an airplane flying over the Atlantic Ocean. Anyway.)
So you need to know how much lye you need to react with whatever amount of oil you’re using. Fortunately, this is easy to find–it’s called the saponification number. There are all kinds of tables of these numbers on the Internet, as well as calculators for determining how much of all the ingredients you need. Also, this site explains all this with a cute story involving bunnies, wolves, border collies, and sheep.
I was at Safeway tonight, waiting in line behind a guy in a Whole Foods shirt (I found this amusing), when he addressed me:
“Excuse me; can I ask you something?”
“Of course.”
“What do you use coconut oil for? What sort of things do you cook with it?”
“Actually, I’m not going to cook with it at all–I’m going to make soap with it.”
“Soap?”
“Yes. That’s what I’m using the Crisco and the peanut oil for as well. I’m getting a bunch of different kinds of fat, and then I’m going to make soap from them and make a web page about it.”
“Oh, I see.”
“Kinda a silly thing, but it should be fun.”
I imagine this guy thought I was completely weird, but considering the fact that I have a shelf that looks like this, he might be right:
So, inspired by the Cockeyed.com Science Club and the MOTD guy’s experiments, I will soon be posting pictures and experiences from my adventures in the world of soapmaking. Stick around; this will be fun!
