Cement, Gypsum, Anhydrite, Ettringite Crystals, and PVA; A Discourse
Introduction
There have been many debates on these subjects, and my intention is to, in a reasonably easily understood manner, explain the "why" of it all. I'm not an expert on this, but I have read up on it enough to form an opinion, and I wish to pass on my findings.
This post is based on a pair I did very recently, in a thread pertaining to these matters.
A Brief Summary
If you do not want to read the whole post, here follows a summary of what I'm going to explain.
- PVA is not suitable in any tiling related application.
- Cementious materials and anydrite (or gypsum) materials are not compatible, and must be completely separated by, for example, a primer.
There. Now onto the main part of this discourse.
PVA and Cementious Materials
I hadn't even heard of using PVA for anything tiling related before I came on here. Thus, I read up on it, and here follows my findings:
PVA stands for
polyvinyl acetate, and it is a rubbery synthetic polymer. It is commonly emulsified in water and used as glue. Many know it simply as "wood glue", or "carpenter's glue".
Cementious materials, such as many
tile adhesives and grouts, or other materials which contain
cement, such as concrete, are
alkaline. Simplified, that means they have a high pH.
Alkali
slowly attacks polyvinyl acetate, forming
acetic acid, which has a low pH.
Cement doesn't dry per se; it
cures through
hydration, which means it binds the water you mix it with chemically. This causes the pH of the substance to rise dramatically. Introducing an acid negates that process to some extent, preventing the
cement or conrete from binding all the water it needs to harden properly.
It is hydrolysis which gives
cement and concrete products strength, and holds them together. Without this process, it would merely be the powder you started with.
The acetic acid which is formed when
cement and PVA comes into contact, either through mixing them, or "priming" with PVA, will continually free the water bound in the
cement, and that will weaken the bond and/or integrity of the material. The effect is accelerated if the material is subjected to moisture, which is more or less always the case.
PVA isn't water resistant. It becomes slightly live when exposed to moisture, and this in combination with the exposure to alkali, accelerates the forming of acetic acid. PVA which is marketed as "water resistant" or "exterior grade", has additives which makes them water resistant, but they're not alkali-resistant.
Anhydrite, gypsum, and cement
Anhydrite products are mainly composed of
calcium sulfate, and gypsum products are mainly composed of
calcium sulfate dihydrate. When anhydrite is exposed to water, it forms gypsum. In other words, it hydrates. Essentially, it cures, but not to the same extent as
cement.
Gypsum always has a proportion on anhydrite crystals left in it.
Cement has a proportion of
calcium aluminate. Calcium aluminate
reacts with calcium sulfate, which is the main component of anhydrite products, and which is present in gypsum. The reaction forms
hexacalcium aluminate trisulfate hydration; in other words,
ettringite crystals. These expand, and force away anything which is fixed onto where they form.
As I've previously explained,
cement cures, which means it binds water through hydrolysis. That means water is always present in
cement. If anhydrite is put into direct contact with
cement, there will be a reaction. The reaction won't be as severe with gypsum, as it's already hydrated most of the anhydrite (the dihydrate part), but there is still some present.
Thus, if you want to
tile onto such products, you will need to
separate them entirely. This is best done with a products which seals, and which is also water resistant, such as
acryllic dispersions.
Even if you use water resistant "PVA", the separation will deteriorate with time, due to the chemical reaction between the
cement, which is alkaline, and the polyvinyl acetate. If the bond of the
cement onto the substrate hasn't already been compromised because of that, the formation of ettringite crystals will very likely cause complete debonding.
Rapidly curing cements may have some gypsum added when manufactured. It accelerates the curing, but does not affect the integrity of the product, because it's present in such small quantites, and during the early stages of curing.
Final Notes
PVA is not suitable as a primer, sealer, impregnator, or admix. The uses of PVA may be many, but they do not include anything tiling related. Use proper manufacturer approved primers and additives instead. Using PVA will likely cause liability issues when problems arise, and that is bound to cost alot more than buying proper materials to begin with.
If you want to
tile onto anhydrite or gypsum, make very sure to properly separate the substrate from the adhesive. There will likely be tears otherwise.
Finally, I would like to quote Cliff Anger:
Calcium sulphate and portland
cement are not compatable - whilst a failure is not guaranteed, it is a real possibility. Over time, ettringite crystals may appear, and as they expand, they will cause the tiles to lift and de-bond. It may well take several months to determine how bad the reaction might be.
