Geopolymer – some formulas, methods, and links

While living in Florida a couple of years back, I’d played around with making some geopolymer and cement like things. It didn’t work all that well. Largely as I was testing first using high CaCO3 content (rather like cements). I did that as a first step, intending to then move on to silicate sands and ‘waterglass’ or liquid sodium silicate.

I got as far as buying the materials, then had to move everything and everyone back to California as my contract ended (and Disney decided to outsource the whole group to India… so chances of getting hired dropped to zero… an action that landed them in the news, in court, and in PR trouble – so some of the building was not fired after all; but new hires were still not on the cards. But I digress…)

So somewhere in my boxes of stuff that were carted back here is a small can of waterglass. I scattered the sand in the yard in Florida since it doesn’t make much sense to drag a couple of $ of sand coast to coast. I’ve also got a small jug of lye in the box.

Well, prior to starting experiments again, I figured I’d do another “quick check” to see if there was more published now about how to DIY, and maybe more shot in the dark experimentation was not needed. Turns out there’s a lot more published now about “how to” and processes. So this article will mostly be a collection of links to them, a few quotes, and some comments about how I assess them.

My general intuition that lye and waterglass would be important is confirmed in the links. Yet there do seem to be other processes and formulae possible. There’s still room for the experimenter.

Sidebar on Sodium Silicate

As a reminder, waterglass is sodium silicate and can be made by many, in some cases fairly primitive, methods. Well inside the skill of ancient Egyptians:


Soluble silicates of alkali metals (sodium or potassium) were observed by European alchemists already in the 1500s. Giambattista della Porta observed in 1567 that tartari salis (cream of tartar, potassium hydrogen tartrate) caused powdered crystallum (quartz) to melt at a lower temperature. Other possible early references to alkali silicates were made by Basil Valentine in 1520, and by Agricola in 1550. Around 1640, Jean Baptist van Helmont reported the formation of alkali silicates as a soluble substance made by melting sand with excess alkali, and observed that the silica could be precipitated quantitatively by adding acid to the solution.

In 1646. Glauber made potassium silicate, that he termed liquor silicum by melting potassium carbonate (obtained by calcinating cream of tartar) and sand in a crucible, and keeping it molten until it ceased to bubble (due to the release of carbon dioxide). The mixture was allowed to cool and then was ground to a fine powder. When the powder was exposed to moist air, it gradually formed a viscous liquid, which Glauber called “Oleum oder Liquor Silicum, Arenæ, vel Crystallorum” (i.e., oil or solution of silica, sand or quartz crystal).

However, it was later claimed that the substances prepared by those alchemists were not waterglass as it is understood today. That would have been prepared in 1818 by Johann Nepomuk von Fuchs, by treating silicic acid with an alkali; the result being soluble in water, “but not affected by atmospheric changes”.

The terms “water glass” and “soluble glass” were used by Leopold Wolff in 1846,, by Émile Kopp in 1857, and by Hermann Krätzer in 1887.

In 1892, Rudolf Von Wagner distinguished soda, potash, double (soda and potash), and fixing (i.e., stabilizing) as types of water glass. The fixing type was “a mixture of silica well saturated with potash water glass and a sodium silicate” used to stabilize inorganic water color pigments on cement work for outdoor signs and murals.

Solutions of sodium silicates can be produced by treating a mixture of silica (usually as quartz sand), caustic soda, and water, with hot steam in a reactor. The overall reaction is

2x NaOH + SiO2 → (Na2O)x·SiO2 + x H2O

Sodium silicates can also be obtained by dissolving silica SiO
2 (whose melting point is 1713 °C) in molten sodium carbonate (that melts with decomposition at 851 °C):[19]

x Na2CO3 + SiO2 → (Na2O)x·SiO2 + CO2

The material can be obtained also from sodium sulfate (melting point 884 °C) with carbon as a reducing agent:

2x Na2SO4 + C + 2 SiO2 → 2 (Na2O)x·SiO2 + 2 SO2 + CO2

In 1990, 4 million tons of alkali metal silicates were produced.

So lots of paths to make it, and some fairly easy for a low tech society with lots of sand and natron to follow.

General Intro to Silicate Chemistry

My first search results led me to this wonderful site:

Their “about” box says they are dedicated to nanotech and interesting chemistry. Here’s a bit from the end of it:

The new 3,5 years long page in ASDN history started in the summer 2007 when the leading Russian nanotechnology tool maker, NT-MDT signed an agreement with Nano & Giga Solutions for the development of ASDN.NET into an educational web portal for atomic structure and fundamentals of nanotechnology. Generous support from NT-MDT helped to develop series of tutorial pages, including very popular ones, such as Chemistry of Love and Silicates.

From the beginning of 2016 ASDN becomes an internet gateway of the Center for Computational Design of Materials and Devices (CDMD) at the National Research Tomsk State University in collaboration with international education and research community. Activities of the Center cover three main areas: fundamental and applications oriented research, development and implementation of educational programs, and development of new computational tools for academic and industrial applications. We continue with development of tutorial web pages as well.

So lots of good stuff there to explore. Seems lots of ceramics are used in nanotech and semiconductors and the idea of custom made geopolymer substrates is interesting to them. On the silicates page, I found a very nice intro along with some very useful links, several of which I link to further down:

It has a very readable intro to the nature of silicate chemistry, the species and crystal structures involved, and what to expect in each category. Then at the bottom a bit of geopolymer discussion:

Geopolymers is another example of artificial semi-crystalline materials. The name “Geopolymers” has been introduced by J. Davidovich and is applied to a wide range of alkaline – or alkali-silicate-activated aluminosilicate compounds. They can be obtained from amorphous clays and alkali-silicate solutions without any melting under high temperatures and avoiding significant CO2 emission. The hardening process of the alkali activated aluminosilicates is based on formation Si-O-Al and Si-O-Si bonds under ambient conditions. It is likely that these products have a highly significant commercial potential, but this technology is just being developed today.

Geopolymer Institute

Which brings up the point that the Geopolymer Institute doesn’t like the word geopolymer being extended to a lot of alkali catalyzed things that are still rocky, but where they assert it isn’t really a polymer. So just FYI be advised the word now has a couple of conflicting definitions. With very different chemistries. But they all make rocks ;-)

Why Alkali-Activated Materials are NOT Geopolymers ?
27 Jul 2017

Many scientists and civil engineers are mistaking alkali activation for geopolymers, fueling confusion, using them as synonyms without understanding what they really are.

To sum-up: Alkali-Activated Materials (AAM) are NOT Polymers, so they cannot be called Geo-Polymers. Geopolymers are NOT a subset of AAM because they are not a calcium hydrate alternative (no NASH, no KASH). They are two very different and separate chemistry (a hydrate/precipitate that is a monomer or a dimer versus a true polymer). Those who claim that both terms are synonyms are promoting a misleading scientific belief. Learn why by watching these three videos.

In his four recent keynotes at the Geopolymer Camp 2014, Geopolymer Camp 2015, Geopolymer Camp 2016 and Geopolymer Camp 2017, Prof. J. Davidovits explained why Alkali-Activated-Materials are not Geopolymers, or why alkali-activation is not geopolymerization. We have selected all the sequences that had been dedicated to this issue in the GPCamp-2014, 2015, 2016 and 2017 keynotes. These new videos are titled: Why Alkali-Activated Materials are NOT Geopolymers. You will finally understand why they are two different systems.

They have 4 videos that claim to describe the chemistry and why it’s different. I’ve not yet watched them but they imply the chemistry details are in there…

Part 1 (2014): AAM are not geopolymers, two different chemistries

Prof. J. Davidovits explains the main differences between AAC (Alkali-Activated Cement or Concrete), AAS (Alkali-Activated Slag), AAF (Alkali-Activated Fly Ash) and Slag-based Geopolymer cement, in terms of chemistry, molecular structure, long-term durability. In a second part, on hand of the industrialization of Slag/fly ash-based geopolymer cement/concrete implemented by the company Wagners, Australia, he focuses on the results provided by the carbonation testing data obtained for ordinary Portland cement, AAS and EFC (Slag/fly ash-based geopolymer). The tests were carried out at the Royal Melbourne Institute of Technology RMIT in Australia. Geopolymer behaves like regular Portland cement, whereas AAS gets very bad carbonation results.

So, OK, lots of ways to make “liquid stone” and he wants to claim he knows which ones deserve the geopolymer name. I’m more interested in the “How To”s of it all, but whatever… So watching the videos is on my todo list.

Henley’s DIY Formulas

In some ways, this is the biggest “win” for the DIY guy at home just wanting to play with something that works:


This section is from the “Henley’s Twentieth Century Formulas Recipes Processes” encyclopedia, by Norman W. Henley and others.

Manufacturing Artificial Stones

The following is a process of manufacture in which the alkaline silicates prepared industrially are employed.

The function of the alkaline silicates, or soluble glass, as constituents of artificial stone, is to act as a cement, forming with the alkaline earths, alumina, and oxide of lead, insoluble silicates, which weld together the materials (quartz sand, pebbles, granite, fluorspar, and the waste of clay bricks). The mass may be colored black by the addition of a quantity of charcoal or graphite to the extent of 10 per cent at the maximum, binoxide of manganese, or ocher; red, by 6 per cent of colcothar; brick red, by 4 to 7 per cent of cinnabar; orange, by 6 to 8 per cent of red lead; yellow, by 6 per cent of yellow ocher, or 5 per cent of chrome yellow; green, by 8 per cent of chrome green; blue, by 6 to 10 per cent of Neuwied blue, Bremen blue, Cassel blue, or Napoleon blue; and white, by 20 per cent, at the maximum, of zinc white.

Chrome green and zinc oxide produce an imitation of malachite. An imitation of lapis lazuli is obtained by the simultaneous employment of Cassel blue and pyrites in grains. The metallic oxides yield the corresponding silicates, and zinc oxide, mixed with cleansed chalk, yields a brilliant marble. The ingredients are mixed in a kind of mechanical kneading trough, furnished with stirrers, in variable proportions, according to the percentage of the solution of alkaline silicate. The whole is afterwards molded or compressed by the ordinary processes.

The imitation of granite is obtained by mixing lime, 100 parts; sodium silicate (42° Be.), 35 parts; fine quartz sand, 120 to 180 parts; and coarse sand, 180 to 250 parts.

Artificial basalt may be prepared by adding potassium sulphite and lead acetate, or equal parts of antimony ore and iron filings.

To obtain artificial marble, 100 pounds of marble dust or levigated chalk are mixed with 20 parts of ground glass and 8 parts of fine lime and sodium silicate. The coloring matter is mixed in proportion depending on the effect to be produced.

A fine product for molding is obtained by mixing alkaline silicate, 100 parts; washed chalk, 100 parts; slaked lime, 40 parts; quick lime, 40 parts, fine quartz sand, 200 parts; pounded glass, 80 parts; infusorial earths, 80 parts; fluorspar, 150 parts. On hardening, there is much contraction.

Other kinds of artificial stone are prepared by mixing hydraulic lime or cement, 50 parts; sand, 200 parts; sodium silicate, in dry powder, 50 parts; the whole is moistened with 10 per cent of water and molded.

A hydraulic cement may be employed, to which an alkaline silicate is added. The stone or object molded ought to be covered with a layer of fluosilicate.

A weather-proof water-resisting stone is manufactured from sea mud, to which 5 per cent of calcic hydrate is added. The mass is then dried, lixiviated, and dried once more at 212° F., whereupon the stones are burned. By an admixture of crystallized iron sulphate the firmness of these stones is still increased.

An Open Source Site

This link has a couple of formulae that look more like Alkali Activated {stuff} so maybe not strictly a geopolymer, but still of interest. OTOH, I’m not quite clear on exactly what the distinction is that is being made and how to map it to ingredients used. I also don’t really care a whole lot as I just like the idea of making liquid stone of any sort.:

Sample Recipe

The making of alkaline solution, 12 hr before mixing, slowly! dissolve 320gm sodium hydroxide (pure lye, as in a “drain cleaner”) into a liter of water. This should be stirred in slowly, with care, wearing gloves and goggles as it is very caustic. This mix will generate some heat while dissolving.

After the lye solution is fully dissolved (12hr) mix one part lye solution with 2 1/2 parts sodium silicate. (available at pottery supplies)
Basic recipe #8

4 1/2 parts metakaolin 1/2 part lime (type s) 8 parts aggregate (sand mix) alkaline solution as needed (about 1/3 the amount of metakaolin and ash, by weight)

Mix all the dry ingredients together then stir in just enough alkaline solution to make a stiff mix. Keep the liquid content as low as possible. Cure like concrete, warm and moist.

Class C Fly ash can replace the metakaolin and lime, if its type F fly ash replace only the metakaolin.

taken from: [1]

Here is a Dr. Michel Barsoum’s formula for a man-made limestone. To a high pH water, add limestone powder, diatomaceous earth and a very small amount of lime. Form and cure at 90 degrees Centigrade. This yielded a stone with a compressive strength > 20MPa and able to withstand 2 months of submersion in water. He stressed that the one of the keys to geopolymerization is the rapid dissolution of silica and that this is best achieved using diatomaceous earth as a silica source, not clay. Because natural earth materials vary, unlike standardized commercial products, the formulae and properties of the final polymerized stone building products may vary greatly from region to region. (Link below.)

Some have described geopolymer as a man-made zeolite; some as man-made limestone. It could also be likened to a man-made soil duripan, especially if volcanic ash is used as a amorphous silica source. Clearly, a range of local natural earth materials can be successfully used, but a period of formula testing is necessary.

Potassium vs Sodium

One of the bits I’ve still not quite sorted out is if sodium or potassium makes the best or easiest geopolymers. Sodium Silicate is the one in the stores, but lots of sands are made of potassium silicates. It’s a place where some of the articles address it, but in slightly obtuse form and ATM I’m 1/2 a bottle of Sauvignon Blanc into it and fine detailed chemistry is not in my “mood” right now… so I need to read these more carefully later.

Alkali silicate binders: effect of SiO2/Na2O ratio and alkali metal ion type on the structure and mechanical properties


Taisiya Skorina, Irina Tikhomirova

Taisiya Skorina is also the author of the page above that has a rather fetching photo of her at the bottom. Looks like Russia has more ladies in chemistry these days than when I was taking it in the USA. Now I wish my Russian was better ;-0

First Online: 14 March 2012

The influence of SiO2:Na2O molar ratio and the nature of an alkali metal (Na vs. K) in commercial aqueous alkali silicate on the microstructure, textural properties, phase composition, and hydrolytic stability of an alkali silicate binder have been investigated using scanning electron microscopy, nitrogen adsorption/desorption technique, X-ray diffractometry, thermal analysis, and dissolution tests. It has been found that microstructure and textural properties of the alkali silicate binder depend both on silica to alkali molar ratio and type of alkali metal (Na vs. K). Sodium silicate binder obtained from commercial silicate solution with lower SiO2:Na2O molar ratio (2.2) exhibits a globular microstructure of silica xerogel with high content of micropores, whereas the binder formulated with SiO2:Na2O molar ratio 3.2 is characterized by more open cluster structure with lower content of micropores. It is observed that surface specific area estimated by Brunauer, Emmett, and Teller method and mesopore volume obtained by the Barrett–Joyner–Halenda method for sodium silicate binder are substantially higher than those for potassium silicate binder. The ultimate hydrolytic stability of the sodium silicate binder increases slightly with increase in the silica to alkali molar ratio within the studied range. Decreasing in SiO2:Na2O molar ratio and replacement of sodium silicate solution by potassium silicate solution in the corresponding filled composition lead to the improvement of mechanical properties and decrease in open porosity.

So if I’ve read that right, the K based (and more heavily KO2 ratio mixes) have a more solid and rock like finish while the Na based and more SiO2 heavy ratios have a more open, weaker, and porous nature. I think…

The Redit Folks Thread

Then there’s someone on Redit who’s made some rocks:

Posted byu/Anenome5
3 years ago
Here is the Recipe for making Geopolymer Concrete, go wild

I have not yet perfected the geopolymer formula, though I have learned a good bit about what to do and what not to do. I plan to put these into a short monogram and release it for everyone to try.

It was very difficult for us to discover the formula but I’m quite willing to share :)

Let me dig out my notes here…

These are the proportions by weight for our geopolymer concrete that tested out at ~5,000+ PSI. These proportions are for a 6,000 grams batch.

101.8 grams of 14-molarity solution Lye (sodium-hydroxide). (This means 41g of lye and 60.7 grams of water). Be careful when mixing this together. Start with a plastic cup of water, 60.7g of it, and then add about half the lye. It will heat the water almost to the boiling point. If you see bubbles forming that’s okay, just stir and let it cool. Once it has cooled a good bit, say 5 minutes or so, add the rest of the lye and stir until it dissolves as well. If you dump in all the lye at once it can boil and sputter and send caustic lye back at you, and it will burn you. If it burns you, wash the spot with water for 10 min. And be careful, because lye can burn your skin in such a way that it will do damage long before you feel any pain, so be careful.

This is the only dangerous step in making geopolymer concrete, and it’s about as dangerous as making soap, which also uses lye.

255.7 grams of Waterglass (sodium-silicate).

15.15 grams of superplasticizer. (Geopolymer concrete turned out to be plastic enough on its own that we omitted this from future batches as unnecessary. It’s generally fairly loose. This is one of its problem! Makes it hard to prepare for spraying and plastering, but perhaps with the addition of nylon fibers it can be made thicker.)

1848 gram of mixed aggregate (sand and 7mm gravel). One point on this, we began ommitting the rock and using pure sand and still obtained a high strength value, but I suggest you play around with the ration of rock to sand and try to find a good medium point. We cut back on aggregate compared to the first pour because the first pour was extremely rocky and wouldn’t even fill the mold we had. The first pour had 1715g of rock and 734.3g of sand. This mix with all sand and no rock came out very beautiful and strong, but it could be made stronger with some rock most likely. This would be a good thing to try out. Also, this rock and sand should be measured out at its wet-weight, not dry weight. So make sure it always has some water in the bag to keep it hydrated. Otherwise dry aggregate will suck water out of the alkali-activator and possibly cause a failed pour when you begin to mix them together. One more note, do not use beach sand, you want some kind of granite-sand or mason-sand. Don’t use beach sand, it results in significant strength loss.

1013g of type-F, low-calcium flyash.

41g of water. One thing we learned was to not play around with the water ratio. You can’t make geopolymer thicker or thinner by adding or taking away water like you can with normal concrete. Instead this will cause the chemistry to fail. The chemical ratios have to be kept fairly consistent. That’s why I say try nylon fibers as a thickener rather than trying to play with water ratios. We did a lot of playing with water ratios and had a lot of failed pours that failed to set-up.

Mixing Process:

Measure out and combine the damp aggregate (sand, rock) into a plastic bucket (do not use metal bucket). Measure 41g of water add it in. Mix the sand and rock for several minutes until everything is well uniformly wet and mixed using a mechanical stirrer of some sort.

Measure 60.7g of water, put into a plastic container.

Measure 41g of solid lye pellets. Don’t leave these standing in the air too long because they will absorb moisture from the air and become gummy.

Pour about half of the lye into the water and mix with a wooden stirrer. Allow the lye to cool down as you mix, then add more lye until it absorbs. Be careful not to add so quickly that it begins to first bubble and then boil. You should be able to feel the heat on the outside of the container and can use that to judge. If mixing large batches of lye solution you will need to mix these the day before and allow them to come down to room temperature before continuing. Cover the lye solution and continue.

Measure out 255.7g of liquid waterglass (36.5% sodium-silicate, 62.5% water). Immediately add it to the cooled lye-solution and stir together.

Pour the solution into the aggregate and mix for several minutes with a mechanical mixing paddle. We used an aluminum-tipped mortar mixing paddle on the end of a drill. The lye will off-gas hydrogen if it comes into contact with just about any metal, but we felt that once it was mixed in with the flyash and aggregate that it wouldn’t be as active against the metal. The alternative was to try to coat the paddle somehow, and that wasn’t a good option as we thought it would surely wear off into the mix. A tough and strong plastic-coated paddle would be idea.

Spray the molds with Pam cooking spray as the mold release (or use any similar mold release, but don’t use petroleum jelly, it’s been known to interfere chemically with geopolymer).

Let it sit for a few minutes, then pour the mix into a mold. I suggest wooden or silicone molds that can survive the heat of curing. We used 2.5″ cube molds made of wood and previously coated in silicone caulk. Note: ideally you would de-gas the mix in a vacuum chamber to get rid of any entrained air before pouring.

Cure the geopolymer in a pre-heated oven at no more than 200° Fahrenheit. Any hotter and it will negatively affect the strength. At 200°F it cures in 4 hours. At 85°F it will cure in 24 hours. Any analogous range and length between works too (ie: you could try 120° for 12 hours). It does not need to be covered or kept wet while curing.

Remove from heat when the time is up and remove from the mold (further heat will not hurt or help it). It is now cured and has about 90% of its final strength. Within 3 days it will have 95% of its full strength, and 99% within a month.

A note about flyash:

On the sidebar you can find details for ordering a flyash type-F sample from Boral free of charge. However if you’re ever in doubt there’s a simply test you can perform. If the flyash is high calcium, it will heat up when mixed with a little bit of water. Calcium compounds in both concrete and type-C high-calcium flyash are what cause both concrete and type-C flyash to cure themselves by generating their own heat, what’s known as the heat of hydration.

If you add a bit of water to a good amount of flyash (say the size of a cup) and it stays completely cool, then you have a low-calcium type-F flyash that is possibly a good fit for this recipe.

If you have a choice, the lower the calcium content the better. 2% calcium flyash is about as good as can be hoped for. I performed this recipe with 5% flyash that was available to me.

Good luck!

And just so there’s no confusion, I am releasing this info under the MIT license:

The MIT License (MIT)

Copyright (c)

Permission is hereby granted, free of charge, to any person obtaining a copy of this document, to deal in the document without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the document, and to permit persons to whom the document is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.


[… down in comments: -E.M.Smith]

Please, this is not geopolymer but alkali activated materials. This has nothing to do with geopolymers. Some civil engineers are wrongly using the word geopolymer as a synonym of alkali activated fly-ash thus creating confusion in the scientific community. The chemistry and chemical reactions are different. Please do a favor to our open community and use the appropriate wording. In your text, replace the word geopolymer with AAFA or other acronyms. See the Geopolymer Institute website for a plain explanation: Why Alkali-Activated Materials are NOT Geopolymers ?

Otherwise, great formula. Thank you for sharing ! I know it will help a lot of students.
level 2

I think you may have jumped the gun here.

I watched the video, and agree with its assessment, however it’s not talking about the geopolymer concrete I’ve made here. He focuses on slag-based materials.

There is no slag in my formula up above.

The other major aspect is that Davidovits and many others in this field are focusing on type-C flyash and geopolymer which does not require alkali-activation because high calcium content allows it to generate its own heat via reaction with water in order to cure.

However, type-F flyash and does require alkali activation due to extremely low calcium content (2% – 5%). If you simply add water to it, nothing happens. It does not heat up, it does not cure, because it has no calcium, it is calcium that produces the heat of hydration to cure type-C flyash.

Any time you see him talking about chemical formulas that include calcium (Ca) you know he’s talking about type-C flyash and not type-f based geopolymer. He does not like the word phrase ‘alkali activation’ for reasons that I make clear below, but he also does not care about type-F flyash, which is the only kind that I care about, because it’s produces the only kind of geopolymer that can survive at sea long-term.

The term “alkali activation” has a very bad sound to the ear of a concrete engineer, because it is a phrase used to refer to a major problem with regular concrete, which can become damaged by alkali environments, and this is called “alkali activation.” This may be part of the reason why geopolymers have been slow to be taken up by industry, a sort of bias against them due to association with that phrase to the uninitiated. Davidovits want concrete engineers to start using geopolymer, thus he has banished that phrase to try to overcome the confusion:

geopolymer cement is sometimes mixed up with alkali-activated cement and concrete, developed more than 50 years ago by G.V. Glukhovsky in Ukraine, the former Soviet Union.[5] They were originally known under the names “soil silicate concretes” and “soil cements”. Because Portland cement concretes can be affected by the deleterious Alkali-aggregate reaction, coined AAR or Alkali-silica reaction coined ASR (see for example the RILEM Committee 219-ACS Aggregate Reaction in Concrete Structures [6]), the wording alkali-activation has a negative impact on civil engineers. Nevertheless, several cement scientists continue to promote the idea of alkali-activated materials or alkali-activated geopolymers. These cements coined AAM encompass the specific fields of alkali-activated slags, alkali-activated coal fly ashes, blended cements (see RILEM Technical committee DTA).[7] However, it is interesting to mention the fact that geopolymer cements do not generate any of these deleterious reactions (see below in Properties).

Ultimately what makes a material a true geopolymer is whether zeolite structures are formed in the resulting material, and by all account the above listed recipe does into produce alumino-silicate zeolitic structures.

If you want more information on the means of making type-F geopolymer, with advice of Davidovits himself, check out this PDF I found where they made type-F geopolymer and inspected it on the microscale and chemically to verify its having turned into a true geopolymer, and also presented is a formula similar to the one I present above:

In this community we’ve got to make a very clear delineation between type-C flyash or metakaolin-based geopolymer concrete that most of the world is pursuing and researching, which is good for making buildings and airport-runways and doesn’t require “alkali activation”, and type-F flyash-based geopolymer that is good for seasteading and does require alkali activation.

Even now while research into and use of high-calcium geopolymers are rare, research into and use of low-calcium geopolymers are even more rare because they don’t need its sea-proof properties the way we do.
level 3

You are right to quote the wikipedia page on geopolymer cement which is a very good introduction to this topic, IMO.

The other major aspect is that Davidovits and many others in this field are focusing on type-C flyash

In fact it is the opposite, Davidovits only works on type-F because with his method, he only gets flash set with type-C and, according to him, the way to solve this problem is too complicated that it ruins the advantage of using fly-ash (he said that at a Q&A session during one of his webinars). But scientists may find a way with time, maybe.

Also, I would like to add that Davidovits has also released his formula to the public. Check this paper at the Geopolymer Institute Library GEOASH: ambient temp. hardening of fly ash-based geopolymer cements I invite readers to compare, perform testing and make their recipe.

The key element to make a great room temperature hardening, user friendly mix of cement, is in mastering the process. The order of mixing, the preparation, creating the geopolymer binder first, then add the rest, are the essential keys to manufacture a very good product. Otherwise, people do rubbish, full of blooming, high carbonation, leachate and other bad problems. I strongly invite people to watch Davidovits webinar, especially the 7th topic on fly-ash. Geopolymer Web Workshop Webinar Watching these videos was a great eye opening to me and I immediately made fantastic improvements.

One more thing. I would like to share my experience. Now I use potassium instead of sodium. First, the mix is more fluid (I add less water), I get much better physical properties and less chemical problems in the end. Like most people, I thought using potassium was more expensive than sodium. It is true if we substitute 1kg of sodium silicate with 1 kg of potassium silicate. But this is not the rule with geopolymer cement. The real truth is I use up to 2.5 x/weight less potassium silicate than sodium silicate in my mix to get the same fluidity and reactivity (the latter depends on your fly-ash and how you prepare your mix). So, the final price is the same and I solve many problems caused by the use of sodium!!! Please, try it, it is worth it.

Sorry for not being more precise as you do, but I am working for a company now and the results are very promising in terms of properties, standards compliances (very important) and workers’ safety (a less corrosive mix, you can dip your finger in it and it does not burn your skin).
level 4

Interesting, thank you. I’ll have to dive back into Davidovits’s material. We were sure he’d given up on type-F entirely, but that seems to have been old info or a mistaken conclusion. Strange that even in the doc you reference he talks about Ca bonding, even though the Ca percent is extremly low in type-F.

However, the remaining question for my purposes is salt-water durability. This Davidovits definitely is not focused on. He talks a lot in that doc about zeolitic AAM, and in studying Roman concrete it was found to be zeolite minerals that were able to resist seawater attack so well.

So regardless all of this, if the method I’ve used above is a superior saltwater material compared to what Davidovits has moved onto, and thus far I have no reason to think others since he’s deprecating zeolitics in this piece, then I may well end up sticking with this formula. But I’ll have to do more research to make the determination or not.

Thanks for bringing these ideas to our attention.

Edit: This quote from the piece seems to indicate what I’d said above, that his preferred geopolymer method does not produce zeolitics structures, which is the primary thing I’m trying to generate for saltwater resistance:

In the zeolitic procedure, Na-aluminium-silicates (mainly zeolitic products) are formed as a result of the alkaline and thermal activation. The method implies the dissolution of the fly ash particles in such a way that the original mineralogy is significantly modified. Fly ash aluminosilicate glassy spheres are dissolved. Figure 4 shows the clear decrease in the background hump of the diffraction patterns between fly ash and NaOH 12M (alkaliactivation). This results in new species, mainly chabazite-Na (NaAlSi2O6·3H2O) and sodalite (Na4Al3Si3O12(OH). Quartz, mullite and magnetite are low reactive phases only partially involved in the zeolitization and remain as relict mineralogy of fly ash. KOH is not an optimal activator in this conventional method since the degree of reactivity is lower than with NaOH. The (Ca,K)-based geopolymeric method is performed at room temperature and entails a low degree of dissolution since only the surface of the starting materials is taking part in the reaction. For the Geopolymer pattern in Figure 4, the original mineralogy of fly ash is not significantly modified. The (Ca,K)-poly(sialate-siloxo) amorphous matrix results from the inter-geopolymerization of the fly ash aluminosilicate glassy spheres (slight decrease in the background hump), with the alkaline solution and the slag. (Nr 22, p8)

Ultimately we’ll have to do ocean durability tests to see how each material compares.

Edit2: Oh wait, he adds in a lot of extra calcium with the slag–he essentially converts his type-F flyash into a type-C slurry, adding in some 10% Ca that way:
level 1


Have you tried using bauxite instead of fly ash? Is that cost prohibitive?
level 4
1 point ·
3 years ago


With two minutes of research to go on, it seems bauxite has an unfavorable ratio of silica to alumina, which would throw the chemistry way off. You want at least double the silica to aluminum.
level 5

Sure, but finding a source of silica to get the ratios right should be a non-issue. I thought about saying “and silica”, but thought that might be implied. The reason I brought it up (which I suppose is also non-obvious) is that fly ash is in finite supply. Hopefully we’ll stop burning so much coal, and the supplies will be even “more finite”.

I’ve always assumed that based on the name, cinder blocks once used fly ash as a primary component…
level 6

In Conclusion

So it looks like there are several direct ways to make “liquid stone” and “geopolymer” of various kinds without too much fooling around reinventing it.

Now I just need to find my saved materials and buy another bag of sand ;-)

Subscribe to feed


About E.M.Smith

A technical managerial sort interested in things from Stonehenge to computer science. My present "hot buttons' are the mythology of Climate Change and ancient metrology; but things change...
This entry was posted in Earth Sciences, History, Science Bits and tagged , . Bookmark the permalink.

33 Responses to Geopolymer – some formulas, methods, and links

  1. jim2 says:

    Sodium silicate is used to seal concrete after it has set up, in this case not part of the concrete formulation. Lithium silicate is more expensive, but penetrates deeper into the concrete. The lithium cation is smaller than either sodium or potassium, which is probably why that is.

    Other uses:

  2. E.M.Smith says:


    From that second link this line caught my attention:

    “In the “Joosten Process”, sodium silicate and calcium chloride are separately injected in the subsoil. They combine to form an insoluble gel that sets and seals the pores and soils with low resistance.”

    So just mix sodium silicate and calcium chloride (used as a drying agent and sold in marine shops as it sucks water out of the air to keep your boat interior at lower humidity) and you get an “insoluble gel” (gel – in silicate speak – is the glossy amorphous stuff that binds aggregate together in compounds like concrete… think rocky glass like solid).

    Hmmm….. Wonder if you get calcium silicate and salt as the reaction product… Would be interesting to mix some calcium chloride with sand, then add in the sodium silicate and wait.

    It also ought to mean that magnesium chloride would also work, and potentially some others too.

  3. E.M.Smith says:

    Oh, and forgot to include this one in the article:

    But it’s “science direct” so all you get is the abstract unless you pay up. The thrust of it seems to be that some geopolymers need a good washing to finish up “pure”.


    Ion exchange properties of pure potassium-based geopolymer (pure K-GP) formulated with Si/Al = 2.1 and K/Al = 1.0 ± 0.02 atomic ratios have been studied. For pure K-GP, the maximum ion-exchange level of K+ with respect to Na+ and Cs+ is 77 and 61%, respectively, as determined by exhausted exchange with 1 M solutions of appropriate nitrates at 22 ± 2 °C. The geopolymer shows a slight preference for K+ over Cs+ as indicated by the normalized Cs/K-GP isotherm obtained under noncompetitive exchange. Parent geopolymer yields a pure geopolymer when it is extensively washed with deionized water until neutral pH is obtained and equilibrated with aqueous solution of KNO3 resulting in changing in composition and textural properties. As calculated from N2 sorption–desorption isotherms, surface specific area and cumulative pore volume for parent K-GP are 5.5 m2 g− 1 and 0.01 cm3 g− 1, respectively, whereas for pure K-GP surface specific area and cumulative pore volume are 270 m2 g− 1 and 0.4 cm3 g− 1, respectively.

  4. jim2 says:

    Another use is to make refractory objects, like a furnace. I want to make one to melt metals, make cement, and other stuff. This article features perlite, but vermiculite can also be used.

  5. cdquarles says:

    Now add critical point water to your list, EM. :)

  6. cdquarles says:

    Oh, a reminder. When you burn the remains of pretty much every biological organism, you get some sodium carbonate and potassium carbonate, as well as some nitrates, in the ash, if it doesn’t get hot enough to dissociate the carbonate. In that case, you get a bit of sodium peroxide and potassium oxide. Add some water, and you get lye (sodium hydroxide) or potash (potassium hydroxide). Hmm, potash … I wonder why that became the common name ;) .

  7. cdquarles says:

    Oh, about salt, of course you get salt. You are doing acid + base with exchange reactions. Whether you get sodium chloride or not, depends. ;)

  8. p.g.sharrow says:

    take vast piles of seaweed dry and burn along with sea shells or limestone, take the ash 15% and mix with 85% dry and powered river clay, mix 15% with 85% river sand and add water to dampness and ram into shape. Will cure into sandstone…pg

  9. Jason Calley says:

    Hey, E.M., you say: “It also ought to mean that magnesium chloride would also work, and potentially some others too.”

    You may also be interested in Magnesium Phosphate cements.
    also, for historical background:

    Click to access MgO-GENERAL.pdf

  10. E.M.Smith says:


    Fascinating… Nice to know my instincts were correct; just wish I’d bothered to look into it enough to know others had been doing it for thousands of years! ;-)

    It’s looking more and more like there’s a dozen or two ways to make “liquid stone” and many of them are well known to a small trade group!

    This line is particularly interesting:
    ” Zinc, calcium, copper and other elements also form similar compounds.”

  11. jim2 says:

    ” Zinc, calcium, copper and other elements also form similar compounds.”
    All form charge 2+ cations.

  12. jim2 says:

    RE the TecEco piece. Phosphate is much more valuable as a fertilizer. Can you imagine the tons of phosphate that would be locked up in various structures were the phosphate type used? Portland cement requires only cheap limestone and clay which is why it is favored. I’m not convinced it can’t be mixed with cellulose, either. Portland cement, the completed mixture, is hydrophilic as is cellulose. In fact, if water gets under a slab of concrete, it can migrate through the slab carrying with it soluble ionic compounds, a process known as efflorescence.

  13. jim2 says:

    Here is an “alternative” building method just found. Paper is cellulose, so it appears to be OK to mix it with cement.

    “Papercrete and paper adobe are also becoming popular alternative building materials. Papercrete is simply shredded paper, sand, and cement mixed in an industrial type blender.”

  14. Larry Ledwick says:

    Here has also been some tinkering with fiber reinforced concrete where they include chopped glass fiber in the concrete mix to improve its tensile strength without re-bar. In principle this has been used for hundreds of years with fiber mixed with adobe brick and horse hair in mortar.

  15. E.M.Smith says:


    Some folks I know used fiberglass reinforced concrete for a building in Ecuador. For some esoteric reason they thought metal would upset the energy flows… but using fiberglass is a “well known method” so that was OK.

    Adobe brick typically uses straw as one of the components. FWIW, under heat and pressure, straw bonds to itself and makes a strand board like material. At least at one company was making it commercially in the last decade.

  16. Larry Ledwick says:

    Yep heat and moisture makes lignin very plastic (why they use steam to bend wood) a similar “cement” the supposedly magic ingredient for the bricks in the great wall of china was rice starch as an extra binding agent.

    The chinese also used straw fiber in rammed earth construction for walls, some of them are still standing. They would lay a layer of slightly damp clay and straw about 2″ thick in a form and then ram it, then add a new layer and repeat, building the wall as tall as necessary. Saw a video of (Mongol /Chinese) showing the method, still used today in the remote areas.

  17. p.g.sharrow says:

    At Meadowood at Albany, Ore. we made a grass straw press board with heat and pressure but the resin that was used was a cyanoacrylate that was set by the heat and moisture in the press. I know of no strawboard made without resin of any kind except those made in a paper mill. I first got involved with strawboard in shaping it with steam and pressure. Later I got involved in the manufacturing of the board as well as a rice strawboard plant in Colusa, Ca. At a wheat Strawboard plant in the Red River area, they used a soy based resin.
    As far as I know all plants have been closed and scrapped during the financial collapse in the early 2000s …pg

  18. E.M.Smith says:

    has an inset table showing several with “lignin bound” board. Which see::

    so at least some are heat / moisture / pressure lignin bound.

  19. E.M.Smith says:

    Interesting company site:

    Don’t know where they are but they make metric panels so likely not USA…

    Compressed Straw Board
    13 November 2017

    Compressed Straw Board (CSB) is a wall construction and lining material made in a continuous process by heating straw under pressure in a Stramit International machine.

    A solid board is produced which is 100% pure straw, without extra binder, the lignin in the straw being sufficient to act as a binder during the process.
    Normally it is covered with a paper cover during the process.

    CSB is produced in solid self supporting boards in thicknesses of 40 or 60mm and 800 or 1200mm widths, at selectable length between 1.5 and 3.9m.

    Clearly several of the makers in the above chart using various resins and binders (likely to give a superior product) but just using natural lignin can give a pressboard of OK strength for some uses.

  20. p.g.sharrow says:

    @EMSmith; I’m afraid your information is a bit out dated , 1 May 2005, I personally know of Meadowood and Prairie Forest as well as the Colusa riceboard operation. I believe Dow closed down their Canadian plant, It’s board was crappy. and the “Strawmet” system is a franchise based on selling equipment to people to make their straw Bail board, Kind of like a narrow bale of straw with a paper hide like a really thick gypboard. If you were building you wouldn’t want to use it.
    Most composite wood products including paper were once made from straw but had one major draw back, the cost of gathering the needed materials. Piggy backing on the lumber industry material gathering meant wood waste materials were much cheaper as input materials then straw.
    We never could overcome that advantage. Gathering and transportation of straw was a killer. Just too low a density of material to justify the cost. The Meadowood panels and the Prairie Forest boards were beautiful products that could have made it if we could have properly capitalize them…pg

  21. jim2 says:

    Many of the “green” products intended to replace the tried-and-true solutions don’t perform as well and are usually more expensive, including labor involved. Rammed earth construction is very labor intensive as are some other “green” construction methods. That being said, in remote locations some of these methods can be more cost effective due to the high cost of shipping to a remote location.

  22. E.M.Smith says:


    Outdated? Um, it’s what’s currently “up” at the sites and one is dated 2017… Besides, my purpose was not to be giving a current state of the art summary, but just to show “one or more” examples of no binder used. I’ve done that.

    I make no claims about what technology is better, or will succeed in the market place, or has desirable characteristics. It’s basically just a case of you said “does not” and I’ve said “these guys do”. That is all. No “dating” need apply.


    Also note I’m not advocating “green” solutions. ALL I did was a single sentence saying heat and pressure can cause straw to self bind. That’s just a fact, not an advocacy. Not an encouragement. Not an endorsement. Not a statement of economic viability. Not “going green”… etc. etc.


    It may well be a crappy product (or it might be great for some things – I’ve never used it). It may well be too costly (or cheaper than anything else if made locally) – I’ve never bought it. (I like real wood). The context of this was about cements and reinforcements and binding to cellulosic materials. Like mud plasters binding to wood in “wattle and daub”, or straw fillers in adobe. I just noted you can get this straw stuff to stick to itself.

    (My private muse “vision” behind the mention, what triggered it, was more wondering about making an open mat of straws loosely bound so you could just tack it up and do a kind of “wattle and daub” construction using it in the mud or cement plaster – not as a strand board replacement in flooring or such.)

    So out leaked the known fact that straw can be pressed and heated into a stuck together stuff; and that some folks had actually done it. That does NOT mean I’m suggesting anything at all about the best construction practices of the world, or economics of production, or advocating ‘going green” or dozens of other flights of fancy. It isn’t saying “this is currently being done to great effect and is the norm of the industry – state of the art type stuff.”. ALL it says is that straw sticks to itself under heat and pressure and some folks have done it at scale (so it isn’t a hypothetical or lab only thing).

    So please stick to criticizing me for what I DO say, not for what is imagined I’ve said but never did:

    Straw can be made to stick to itself.
    Some companies have done that in the last 10 years.
    Then, when challenged, citing 2 places with 3 companies that actually did / do it in conformance with my “last 10 years” window.

    Nothing more was asserted.

  23. jim2 says:

    EM said @

    Not a statement of economic viability. Not “going green”… etc. etc.

    My comment was not directed at you. Just an observation from another old guy.

  24. E.M.Smith says:

    Looks like it’s alive and well in India, China, Bangladesh with 12 providers:

    6 more without the space:

    3 listed in the USA:

    A Canadian cabinet maker using strawboard with an isocyante binder to make kitchen cabinets (seems it is more water resistant than regular particle board made with formaldehyde resins)

    What looks to be the UK standard supplier:

    Stramit International (Strawboard) Ltd

    The Stramit strawboard technology was invented in Sweden in the 1930s. It was brought to the UK in 1945 by the Mosesson family, who established a factory and research centre in Suffolk, registered the trademark internationally, and have subsequently set up factories around the world under licence.

    Stramit is produced by a patented process of heat and pressure, that utilises the resinous substances within the straw.

    The commercial success of Stramit as an alternative partition material to gypsum board (eg 300,000 houses in the UK) resulted in its being awarded a British Standard (BS 4046 Compressed Strawboard).

    Stramit has a long history of pioneering organic based building materials, has undertaken extensive research in the gathering, storing, and utilisation of straw as a raw material for construction, and is enthusiastic about being a member of the EU consortium ISOBIO.

    And I’m sure there’s more to be found if more than 5 minutes were put into it…

    Again: I am NOT advocating it. I AM saying it exists.

    (Though I’m more interested in the idea of making an open straw lattice as wattle replacement)

  25. E.M.Smith says:


    Fair enough.

  26. Jason Calley says:

    Hey Jim2! “Phosphate is much more valuable as a fertilizer. Can you imagine the tons of phosphate that would be locked up in various structures were the phosphate type used?”

    Hmmmm…. maybe bad, maybe not. Portland cement with metal reinforcement has a working lifespan somewhere between 50 to 100 years, but some of the old magnesium phosphate structures may last 100- years or more. That longer lifespan may be more than enough to compensate for having tied up the phosphates. (There is an old reinforced concrete bridge in Alabama which I have visited. It was built about 1920 or perhaps slightly later. The bridge is closed to traffic, but the cement is crumbling from long term spalling of cement from the rebar. Looks like architectural leprosy!)

    “I’m not convinced it can’t be mixed with cellulose, either. Portland cement, the completed mixture, is hydrophilic as is cellulose.”

    Yes, cement can very well be mixed with cellulose. Bucky Fuller built a cement wall with straw reinforcement back in the 1920s. I have read of structures in India and Asia using bamboo fibers for the same purpose. My personal favorite (based strictly on what I have read, no actual personal experience) these days would be basalt fibers. Expense, strength, and lifespan all superior and plus we have cubic miles of basalt available. Worth doing a search on.

    And E.M., one of the things I read about long ago was a group using straw to make insulation panels. Instead of pressing the straw into rigid panels, they used a technique which they claimed was common in old England during the days of wattle and daub. The straw is completely wetted with a thin slurry of water and clay. The straw is then packed into a form where it dries. The clay covered straw has a lifespan of centuries and as the clay dries it binds the straw together into a block. (Think of a kind of straw styrofoam, moderately rigid, but not enough to be structural.) If you are making a post and beam type house, you could perhaps use such panels for infill and then stucco over. Naturally, using a thin cement slurry instead of clay might be better.

  27. E.M.Smith says:


    Basalt rebar:

    Doesn’t have the problems of iron rebar rusting out and causing your concrete to fail…

    Also like that straw insulation clay sintering-like idea. Now I’m thinking about how to blend that with posts made of cement and basalt or posts made of straw reinforced concrete or strawmat & daub or…

  28. E.M.Smith says:

    Generally speaking, using something as a soluble consumable (lots of tons over millions of acres) permanently consumes much more of it than using a smaller amount in a nearly permanent construction (and where the material can be recovered later).

    I’d have no worries at all about using phosphate to make cement. Besides, seabirds crap out tons of it all the time. Many of the best mines are based on old seabird pooping grounds.

  29. jim2 says:

    Phosphate cement provide corrosion protection to steel. Work is in progress on 1000 year portland.

  30. cdquarles says:

    Hey, Jason. I know that bridge, if you’re talking about one that was replaced a few years ago. All of the similarly constructed bridges that I know of, where many were in Birmingham, have been replaced because of iron corrosion damaging the concrete.

  31. E.M.Smith says:

    Anywhere with rebar in concrete and road salt has a major bridge corrosion issue. They really need to look at things like the basalt rebar or phosphate cement in making the replacements, but they won’t. “Not standard practice” nor lowest bid, so no government will go there…

  32. Larry Ledwick says:

    They have moved part way there, some projects are using epoxy coated rebar or even stainless steel rebar. Both still probably coming from the traditional suppliers or rebar at a higher cost, rather than seeding a new industry by using an alternative material not native to the steel fabrication industries.

Comments are closed.