White pigments | Double Hombre swirl (quantitative) | Am I too stupid for glycerin rivers?

Soapmaking Forum

Help Support Soapmaking Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.


Notorious Lyear
Jan 14, 2021
Reaction score
  1. Decide for a unit weight X (all further quantities are multiples of this).
  2. Make 7X soap batter, divide into one pure portion (P) of 4.17X and two (Z, T) of 1.42X
  3. Stir your favourite colours into the Z and T batters
  4. Cast 1X of Z into your mould
  5. Add 0.90X of P into Z, carefully cast 1X of this blend on top of the first layer
  6. Add 0.69X of P into it, cast 1X (it is used up now)
  7. Cast 1X of P
  8. Add 0.18X of T into P, cast 1X
  9. Add 0.24X of T into this blend, cast 1X (it is used up now)
  10. Cast the remaining 1X of T.
Am I overthinking this? Maybe. Anyway, this gives me an hombre-like double gradient with reproducible contents of colourant, with a factor of 1:√10 in between each, i. e. the next but one dilution has a tenth of the colour strength of the first one.

Now, imagine that you chose for the suggestive batter names the colourant zinc oxide (ZnO) for Z and titanium dioxide (TiO₂) for T, two white pigments of which you want to estimate the brightness and opacity they lend to soap. How would that look like?


A few days ago, I made this soap dedicated to pigment dilution (recipe based on abyssinia oil to exploit its impressive natural transparency). Still in some planing mood, I grated it into thin sections of about 1.8 mm thickness each. Now, a few days into cure, have largely reached their final moisture/opacity.

One reason for this is that I found it surprisingly difficult to get reliable mass-based numbers for dosage of pigments. Everyone except me seems satisfied with these stupid tsp ppo numbers, that will vary between manufacturers, top or bottom of the bag, air humidity, electrostatic charge, height above sea level, moon phase, house number, etc.
But this approach gave me a solid conception what to expect from (my) zinc and titanium oxides as white pigments (percentages are relative to total oils).
  • There is a reason why TiO₂ is so popular. It's just a powerful pigment. It adds a pure white in reflected light (like when you look at the soap under normal lighting conditions), and at the same time increases opacity (makes the soap less translucent). Noticeable at 0.1%ppo.
  • Zinc oxide increases the whiteness too, albeit by a somewhat lower amount (you'll need more for a similar brightness. However, it doesn't increase opacity as much. If you aim to just “lighten up” a soap (e. g. counterbalance natural tan of oils or FO discolouration) but you don't want to make the soap look bright white, ZnO is not a bad choice (if you have it at hand). My initial concerns regarding the lye solubility of zinc were not confirmed – obviously the pigment stays intact to a large degree. ZnO has barely a (visual) effect at dosages below some 0.3%ppo. Note that I did not care about the medicinal/toxicological implication of zinc in soap (neither about its potential to create zinc soaps, an exotic type of soap scum).

A different objective of this batch was another attempt of mine to obtain glycerin rivers. I did everything I could to provoke them. I added quite some TiO₂, ramped down the lye concentration to as low as 29%. I made the lengthiest CPOP of my life: put the mould into a water bath at 60°C and put it into the oven over night – it was still at over 30°C by the next noon. But no, still no glycerin rivers for me. :(
To those who have regular glycerin river issues: Which factors are you usually avoiding, that I could try to provoke them?

And yes, I abused my laptop screen as a backlight. The bottom right graphic is the crystal structure of rutile (natural TiO₂ mineral).
Glycerin rivers are just evil like that. And who know, maybe there’s more to them than water, heat, and particles? Speed of saponification and cooling?

You’re probably not the only one who cares about mass and particle dispersion in soap, but you’re the only one I know of so far! I enjoyed your process. I don’t have the patience or knowledge to deal with it except to stick with td from one supplier and learn which of my recipes needs what amount...and eyeball it. NSS seems the densest in my experience so I use less than I want in the batter because after saponification it lightens even more.
Yes, Auntie Clara's is always worth a read. Her series of articles was my reason to use a very diluted 1:2.45 lye, and tried to cool down as slowly as possible.

It turns out that Kevin Dunn has something to say there too (she links that article). However, my lye concentation is already “comfortably on the high-water side of this threshold”, and the heat treatment/slow cooling definitely exceeded 4 hours as well. One pointer is added glycerol (I didn't, and maybe that's part of the secret, since my high-abyssinia no-lauric blend doesn't produce quite as much glycerol by its own as a more conventional recipe).

Oof. Suppliers, that's another hot potato. I didn't mention that my pigments are not from a cosmetic supplier, but artist grade. Can't even tell where they are from, I got dozens of pigments for free from liquidation of a painter's atelier. So much fun to play with them (and I promise I'm putting only the less toxic ones into soap).

In the end, eyeballing sounds quite like the order of the day. ;)
Update: DOS attack!

I'm really surprised how, about one month in, these soap sheets are under so heavy DOS attack! I wouldn't have guessed recipe-wise, but apparently crambe/abyssinia oil isn't quite rancidity resistant (at least not as resistant as I thought when I decided to not put ROE in them :().
Since all layers appear affected at some point or another, I can exclude an unfortunate recipe deviation due to the complicated batter mixing procedure. Interstingly, the specks appear greasy, like drops of brownish oil put on top of the soap. I even could sweep away some of the smaller patches.

To be honest to myself, abyssinia oil is IV-rated somewhere between HO safflower (which I'd deem quite DOS safe) and rice bran oil (which I'd recommend to take DOS precautions to myself and others and/or not use exceedingly). Those two drops of dilutd ROE definitely wouldn't have hurt. The 10 days older bar of abyssinia castile (with ROE added) still shows not the faintest trace of DOS.