Why FireBricks / Refractories Fail: Silica+Flux, Not Heat

[Forsythe]

Nov 8, 2022 6:17:59 GMT -8 Vortex said:

Nov 7, 2022 10:34:55 GMT -8 Forsythe said:

I'm not 100% sure what the max temp is in a DSR3, but trev has mentioned his Vortex afterburner has gone above 1400ºC, and he's experienced kiln shelf cracking.in his stove as well.

1286ºC was the hottest afterburner temp IIRC, and that was an exception as it was in a force 12 gale, my previous hottest was 986ºC. My Vortex stove is normally a slow and steady burn though, and only ever one load.

I did have a kiln shelf in the afterburner crack on the first firing. I have no idea what type of material they are though as they were just sold as 'kiln shelves'.

Oh, weird. I guess I misremembered the peak temp you had mentioned in regards to a stove glass convo. You’re totally right — 1286°C— I stand corrected.

[venison]

First off just another huge thank you to Forsythe for such a generous and much needed contribution!

Second, there are a couple different stove/heater applications mentioned in this post, so wanted to state the purpose of my own build, which is a pretty basic batch-box/riser RMH core.

So, after reading through most of these posts as well as the "Disintegration of Superwool" posts, I'm trying to think of what then would be an "ideal" RMH core, with the goals of simplicity, safety, cost effectiveness, functionality, and longevity.

I had been leaning toward building forms and using an insulating castable refractory (KAST-O-LITE) rather than IFB's, but after reading Forsythe's comments about the need to properly fire/cure these castables (and then looking at some of the instructions for the processes), I'm now leaning toward a core built of soft IFB (insulating fire brick) bonded with sairset (or similar), and coated on the fire face with ITC100, and perhaps ITC100 as a second coat over a maybe somewhat thicker coat of a high alumina refractory like mizzou. (Obviously everyone has different budgets, access to materials, etc, so this is partially based on what I know I can find in my area (Front Range of Colorado).

To summarize, if I'm understanding the gist of what's been presented here, the soft IFB provides the basic physical structure and a decent level of insulation, while the coating(s) provide some physical protection from wood/tool abrasion, as well as both IR reflectivity (ITC100) and, most importantly, resistance to the degrading thermo-chemical breakdown of the brick from the byproducts of high-temp wood combustion. Does that sound about right?

If so, is there an ideal temp rating/density/alumina content for the IFB's? I suppose there will be trade-offs going with one over another... higher temp rated bricks seem more dense, so would probably last longer but less insulative? I think for my application, that would not be a bad trade-off in that I will likely be burning multiple loads of wood over the coarse of a day, so some thermal mass for retained heat in the core I don't see as a bad thing. Also, along that line of thinking, what about a thin (3/4") "hard" firebrick core (still coated with ITC on the fire face, but wrapped in 2-4" of ceramic fiber blanket for insulation?
Also, any recommendations for a mortar? Seems something that would allow for the thermal expansions and contractions would be best? Also, maybe something that will seal and give structural strength, but still allow fairly easy de-construction for maintenance/repair/alterations?


Regarding a fire-face coating, if the ideal coating thickness of the ITC100 is only 1-3mm, a slightly heavier 1st coat of something like Mizzou or Kast-O-Lite over the IFB and under a thinner coat of ITC100 seems to me might make for a more physically tough surface (assuming they bond and don't delaminate).
Just thinking about all the knocks and abrasion of loading and tending wood over the years... I don't really want a wood burning heater that's so fragile it needs to be treated like eggs and in constant need of patching/rebuilding... (would rather just use a standard woodstove). Would a high alumina castable be a good candidate for this purpose? Are there other materials that might work better?

If a castable refractory was used, would the curing/firing of a thinner (maybe 1/8" - 3/16" or so?) layer be a little more tolerant of less than ideal control over the process (relative to casting the full structure)? Or maybe the ITC100 directly over IFB would be sufficiently tough? Anyone have thoughts or experience on this?
 
Obviously this is going to depend somewhat on the specific material, so here is the store I'm currently planning to source materials from: refwest.com/search.aspx
(my browser shows their website as being "not secure" but afaik it's not some giant risk, I hope anyway!)
Both Mizzou and Kast-o-lite are fairly high alumina, though if under a layer of ITC maybe that's no so crucial? I plan to give them a call to ask all these questions but if anyone here sees a product that might be better suited please say so:)

I built a propane forge quite a few years ago by coating ceramic fiber and board with a few pretty thin coats of one of these castables, though at this point I've forgotten which one it was... Basically I just let it air dry for a day or three, then fired up the forge, low flame at first then gradually increased, but was obviously not very scientific about the process. It has been holding up ok... my main goal was just to prevent the fiber becoming airborne, and it seems to have done that at least, but it's definitely pretty brittle and has cracked and is certainly not something that I would want to subject to the physical abuse of a wood burner.



In terms of some potential alternative materials... I've noticed some relatively affordable borosilicate glass tubes for sale on ebay... assuming sufficient physical protection from breakage, would these maybe be suitable for use as secondary air tubes (in order to avoid the problems associated with steel channel)?

Also, there is a ceramics supply store, and maybe a few more options in my area, where a variety of ceramics and other raw materials can be had: rockymountainclay.com/product-category/raw-materials/

Anyone see here any materials that might be advisable to add/use/experiment with as supplements/aggregate/etc to maybe improve an existing product given the particular purposes and needs of a wood burning RMH? Given the fairly wide variety of mixes that have been engineered by experts, I sort of doubt the possibility of this but figured it couldn't hurt to ask, esp since RMH's are a bit of a fringe use for these materials...

Many thanks!

[foxtatic]

My humble contribution to this discussion is a video I took of me probing around my DSR2 core made entirely out of sodium silicate coated ceramic fiber board. www.youtube.com/watch?v=ySsUlxaFjNI

I expected wear and tear in the firebox but wanted to test how durable the CFB was before adding a lining of firebrick. What I did not expect was how fragile the CFB is in the top box after the dozen or so burns. I believe I can now attribute this to the chemistry discussed in this thread (the reaction of wood ash, heat, and high silica content.) So based on this exploration, I will be applying Heat Guard 65% Zirconium coating and hoping for a nice extended service life of the CFB.

[masonryrocketstove]

Nov 24, 2022 11:41:48 GMT -8 venison said:

In terms of some potential alternative materials... I've noticed some relatively affordable borosilicate glass tubes for sale on ebay... assuming sufficient physical protection from breakage, would these maybe be suitable for use as secondary air tubes (in order to avoid the problems associated with steel channel)?

Definitely don't use borosilicate glass. Borosilicate's "permissible" operating temperature for any decent service life is only around 200-250ºC, and its top-use temperature is approx. 300ºC, but it doesn't last terribly long at that temp or above. (Borosilicate's flowability deformation temperature is approx. 525ºC... and you're likely to see at least 900ºC-1000ºC in a batchrocket. Even the lowest "clean-burning" wood fire temperature is around 600ºC, which is the minimum needed to crack all the off-gassing carbon compounds into CO2.)

Under labware conditions, borosilicate used up to 300ºC can be re-annealed to extend its service life. Re-annealing relieves residual stresses in the glass and helps to keep the boron dispersed between silica molecules. ...But at the upper end of its use-range, borosilicate steadily undergoes degradation until it eventually cracks and shatters, because the boron increasingly falls out of boron+silica glass solution. (True "glasses" are technically "liquids" from a molecular perspective...it's just that they don't flow much at low temps, but they also don't recrystallize when they solidify at low temps, either.)

Even if you were able to cool the borosilicate glass enough internally with incoming air to keep it from deforming at 525ºC, the hot-face side of the glass at the temperatures in a batch box will damage the glass solution of boron+silica beyond re-annealing repair ...and the ash from the wood at those temperatures will steadily fuse into the glass, reducing its thermal shock resistance as the glass turns ever more-and-more into soda-lime (common) glass.


Historically, the first primative glass produced by humans for thousands of years was made from a mix of wood ash [which was the source of the "glass-formers" soda (sodium) and lime (calcium)] + sand (silica) ...and that's where soda-lime glass gets its name.

"Glass formers" are fluxes that lower the melting point of silica and alumina, changing the tightly-packed crystalline structure to an amorphous, disordered, "fluid" with shapable plasticity at elevated temperatures... which allows the silica (or aluminosilicate) molecules to move and flow past each other for things like glassblowing and hot molding, etc.

Boron is also a glass-former flux, just like sodium and calcium, the difference is just that it stays in solution and dispersed among the silica molecules over a wider temperature stability range than soda or lime fluxes do... and thanks to its low coefficient of thermal expansion, borosilicate is more resistant to thermal shock over a wider range, too... but not up as high as ceramic "glass" like Robax, Pyroceram, or Neoceram.

...In the refractory industry, the ash slags which flux aluminosilicate are also referred to as "glass-formers" because they cause lowered-temperature melting and flowability ("fluxing") of the alumina and silica molecules in the refractory composition. And because the "glasses" they form have a higher coefficient of thermal expansion than the surrounding aluminosilicate refractory, they cause cracking.


Ceramic "glasses" like Neoceram, Robax, PyroCeram, etc. technically aren't true "glasses" from a molecular standpoint. They're actually mono-crystalline ceramics formed under heat and extremely high pressures. The pressure forces the alumina and silicate molecules to reorder themselves along a single, uniform plane, all with the same crystalline growth axis.

Under this pressure-induced monocrystalline structure, the crystal "defects" (multiple crystals growing in different directions along disorganized axiis,) which normally occur in ceramics, are not allowed to form.

Crystal defects refract and scatter light, making most ceramics and aluminosilicates like firebricks opaque (not see-through.) ...but when formed into a monocrystaline sheet along a single axis with no defects, it allows light transmission through the sheet of Robax / Neoceram / Pyroceram... making these refractory stove ceramics light-transparent (so you can see through them.)

We just call these transparent ceramics "glasses" because that's typically what everyone calls a transparent sheet of hard, brittle silicates, and because they can serve all of the same functions as glass panes do... but they can do it at much higher temperatures than true glasses (even borosilicate glasses) are capable of withstanding.

Over time, with exposure to heat and wood ash, you're likely to see some opaque spots form on the surface of Neoceram / Pyroceram / Robax ceramic-glasses used in close proximity to the burn chamber. These are spots where the ash fused into and fluxed the surface of the monocrystalline sheet. With some elbow grease and a mild abrasive, you can typically polish those opaque spots off of the surface. (It also happens to the ceramic glass tops of electric stoves / cooking ranges when spills trapped between the cooking pot and the glass top burn to ash and fuse into the glass surface, making it look dull, matte-textured, or scuffed.... so any kind of polish sold for glass-topped electric cooking ranges will also work to polish your rocket stove's ceramic-glass door or afterburner window.)

[Dan (Upstate NY, USA)]

Thanks for sharing.

Of these three what are their differences?

Robax / Neoceram / Pyroceram

[masonryrocketstove]

All three are good choices for stove doors and windows, because they're all lithium-aluminum-silicates with varying amounts of magnesium and titanium oxides (plus trace amounts of other minerals) to shift the crystal structure in subtle ways for different material strengths..

The main differences between them are their temperature rating and the color tint, with smaller differences in mechanical strength and thermal shock resistance (They all resist thermal shock, but Pyroceram is the standout. It can go from flame to ice bath quenching and back to flame all day long with no effect.)


Pyroceram and Neoceram (Nippon's first type) are very similar in their amber-tinted color, and have the same thermal ratings... so a lot of places sell them interchangeably, even though they're slightly different in some ways.


Pyroceram (made by Corning) is the most forgiving to bumps, drops, and extreme thermal shock. Also has the highest alkali resistance by a thin margin, so will resist ash fusing a tiny bit better than the others. Rated 800ºC Intermittent / 750ºC Continuous. Some people say it's a little more amber-yellow than Neoceram... but I can't tell the difference, personally.
Pyroceram's crystallinity leans closest to the lithium-aluminum-silicate mineral spodumene.


Neoceram (made by Nippon) is actually a family of several Neoceram(s) variations that Nippon makes. ...I can't keep up, myself. One type of Neoceram (the first type) is almost a copycat of Pyroceram...another type (type 2?) is the strongest for load-bearing like is needed for glass cooktops, most of which are made from Nippon's (second-type) Neoceram, which has that characteristic black stovetop glass color.
Neoceram Type 1 is the one intended for firebox doors and windows, and is also rated 800ºC Intermittent / 750ºC Continuous.
Its crystallinity leans closest to cordierite with its higher magnesium content, but it's still a lithium-aluminum-silicate, too.


Robax (made by Schott) is the most clear / least-tinted of the three, which is important to some people who dislike the amber color for fire gazing...but that shift toward colorlessness comes at a cost of a little bit lower temp rating: 760ºC Intermittent / 560ºC Continuous. The lower amount of lithium, magnesium, and titanium oxides make Robax's crystallinity lean closest to fused quartz, but it is still a lithium-aluminum-silicate as well. While it is mechanically very strong, it may also be the most susceptible to ash fusing (by a very tiny margin) due it its predominantly-silica makeup.

[iau461]

Tons of great info here. So much I doubt my comprehension. So, what seems like a simple question.

I'm contemplating a DSR3. I want it to last decades. What seems like a simple, fairly low-cost core build would be inexpensive (~40% alumina) hard firebrick, with all interior surfaces coated with ITC-100HT or equivalent. I'm willing to go to 70% alumina brick but it seems unnecessary with the effectiveness of the coatings at protection from slags. Am I wrong?

[masonryrocketstove]

I'd think the 70% alumina is probably better for insulating firebrick, because IFB is more porous. But dense brick doesn't have the same porosity vulnerability, so I'd think 40% alumina dense brick coated in ITC should be plenty durable.

Flipside is that coatings can make a stronger bond with rough or porous surfaces.. so maybe plan on inspecting and potentially re-applying the coating every 4-5 years?

Probably depends a lot on the amount of abrasion the coated surface sees. A sloped floor with an ash slot in the bottom might not need recoating as often or at all. Whereas as a flat bottom firebox that needs the ashes periodically scooped out might need a maintenence re-coat every few years, at least on the floor and sides where the ash scoop scrapes against them.

[iau461]

Perfect - thank you! Of course, other opinions still welcome!

[solobird]

Having no (easy) access to colloidal silica or commercial coating, would there be any other option for ceramic fiber (J-tube riser) ?

Could get: water glass, NaOH, KOH, zirconium silicate, bentonite
kaolin SiO2 47%, Al2O3 37%, Fe2O3 0,85%, TiO2 0,2%, K2O 1,1%, L.O.I. 13
pottery clay SiO2 77,0 %, Al2O3 17,8 %, TiO2 1,5 %, Fe2O3 1,1 %, clay, fire clay
Calcined Alumina AL2O3 99,30%
Blakite

This thread donkey32.proboards.com/thread/3287/basic-kaolinite-clay-geopolymer-chemistry mentions a few solutions, is any suitable ? 

[masonryrocketstove]

Apr 5, 2023 11:57:05 GMT -8 solobird said:

Having no (easy) access to colloidal silica or commercial coating, would there be any other option for ceramic fiber (J-tube riser) ?

Could get: water glass, NaOH, KOH, zirconium silicate, bentonite
kaolin SiO2 47%, Al2O3 37%, Fe2O3 0,85%, TiO2 0,2%, K2O 1,1%, L.O.I. 13
pottery clay SiO2 77,0 %, Al2O3 17,8 %, TiO2 1,5 %, Fe2O3 1,1 %, clay, fire clay
Calcined Alumina AL2O3 99,30%
Blakite

This thread donkey32.proboards.com/thread/3287/basic-kaolinite-clay-geopolymer-chemistry mentions a few solutions, is any suitable ? 

I think one of the general ideas is to avoid using something that adds sodium (Na) potassium (K) calcium (Ca) or iron (Fe)

From the ones you listed, I'm pretty sure the Blakite would be the best option since its a mortar that would stay stuck to surfaces as it dries, it would self harden, it's slag resistant, and it is very refractory rated to 1650ºC.

Plain clay slurry might work for soaking into and rigidizing a ceramic fiber riser in a large 7 or 8 inch J tube, since the larger CSA units could probably get hot enough fire the whole riser's length of clay-embedded fiber until it was vitrified.. at least enough to be hard and non absorbent. But regular clay isn't slag resistant, so it wouldn't protect from corrosion like refractory coatings do.. and plain clay would shrink and crack away from solid surfaces, so it wouldn't stick very well to the firebrick or castable part of the burn tunnel as it dried and hardened.

I don't know much about the geopolymers.. but just looking at the ingredients in them, I'm pretty sure they would be even less slag resistant than plain clay. Just a guess.

[solobird]

Have used Blakite for setting firebrick with good results, not sure about coating the inside of a ceramic fiber riser.
Whether it would crack, if prior treatment is necessary - i.e fumed silica or burning of the binders first with a gas burner.
It's air-setting, so no need for sintering temps.
If this would work I'd start the CF riser from the base level of the burn tunnel.

Would (meta)kaolin or zirconium silicate work as plain rigidizer if I don't have access to fumed silica ?

As for the alkali activated materials, the alkali is the activator. But have no experience, much less for applying them to CF.
High temp AAM
Geopolymer ceramic application

[masonryrocketstove]

Apr 6, 2023 10:57:41 GMT -8 solobird said:

Have used Blakite for setting firebrick with good results, not sure about coating the inside of a ceramic fiber riser.
Whether it would crack, if prior treatment is necessary - i.e fumed silica or burning of the binders first with a gas burner.

Are you using ceramic fiber batt (in a roll?) or ceramic fiber board? If it's CF board, you'll definitely need to burn the binder out before applying any rigidizer or coating. Nothing will stick to that phenolic formaldehyde resin they use. Its basically like a water-repellent plastic coating that you have to burn off before anything else will bind to CF board.

It's air-setting, so no need for sintering temps.

Exactly what I'm thinking too. What size J tube are you making? Depending on how big the system CSA is, some options might be able to fire in place and vitrify a rigidizer that isn't air-setting. Like the clay option. Clay is what's used for a rigidizer in those vacuum formed ceramic fiber tubes from Europe (the ones used in Loam Freemanship's DSR 3 that were discussed recently)

Would (meta)kaolin or zirconium silicate work as plain rigidizer if I don't have access to fumed silica ?

If you were to use kaolin, you'd for sure want to use the regular clay kaolin, not metakaolin that's already been fired to 750ºC to calcine it. Regular clay kaolin is hydrophilic and plastic (meaning it's wettable and shapable) so water will make the clay particles stick to themselves and to the fiber, and hold that shape as it dries. Metakaolin isn't a clay anymore, its more like a chemically-reactive dust that has no plasticity, and so water won't make the particles sticky.. they won't be moldable and won't stay in place in the fiber as the water evaporates.

Zirconium silicate also isn't wettable or plastic.. and it wouldn't vitrify either. There's a lot of problems with trying to use zirconium silicate that were talked about earlier in the thread here: donkey32.proboards.com/post/37770/thread

If you can get fumed silica, it's a pretty good option for rigidizing. Does have to be fired to cure it though since its not air setting like colloidal silica or Blakite, etc.

As for the alkali activated materials, the alkali is the activator.

Yes, I at least know you're right about that. For castable refractory, the Ca in calcium aluminate is the alkali activated cement binder that's needed to harden the green shape at room temperature. But that calcium is considered a liability after the castable is fired hot enough to sinter it, because room-temp alkali bonds break again at calcining temps around 700-950ºC, and then recombine again in new / different crystalline phases at the really high refractory temps >1000ºC. A lot of those highest temp alkali crystal phases expand when they form, which weakens the sintered cast's shape. Those phases also have a greater coefficient of thermal expansion through heat cycling than alumina does, and those things allow slag to penetrate more easily.

That's why most high alumina, high temp, slag resistant castable refractory is advertised as being "low-cement" castable. The less alkali activated cement it has, the stronger and more slag resistant it will be at service temps.

That's why I'm guessing most of the geopolymers would be less slag resistant, and probably lose strength the hotter they're fired. The alkali "polymer" chains break their room temperature bonds as you go past calcination temps, approaching vitrification and sintering temps.. and that frees up reactive alkali in the mix again to make expanding high-temp phases.

At least.. that's my understanding based on refractory chemistry and the limited amount I've gathered about geopolymer bonds. Maybe somebody with more experience in geopolymers can chime in.

[solobird]

Yes, using ceramic fiber mat, 1260C(2300F), 128kg/m3.

Planning 6" J-tube, feed tube and burn tunnel out of 1.2" vermiculite that I already have, maybe insulating firebrick on the outside. 2" ceramic fiber mat riser inside metal pipe for support. Have used CF in the past, just on the outside of some hard firebrick elements. This stove would work as an outside cookstove and oven, so no use in heating up mass, that's the main reason for trying to build directly out of some insulation material.

Your explanations and re-reading the info on this topic makes things clear. Thanks!
Also came across this forge insulation info while doing research.

Blakite it is then.
Should I apply it at a consistency of paint or keep it the way it is now, creamy ? how thick ?
Given that the CF will be coated, should I start the riser from the base of the burn tunnel or start from above ?

Insulating fire bricks from tore down stove based on Matt Walker's riserless, used for a few years in the summer time.
950C rated, SiO2 : 67-70 %, Al2O3 : 12-18%, Fe2O3 : < 2 %

firebox

port

riser and under cooktop

[masonryrocketstove]

Apr 6, 2023 10:57:41 GMT -8 solobird said:

High temp AAM
Geopolymer ceramic application

I had a chance to read over those papers, thanks for sharing those. It's interesting stuff with a lot of neat applications for sure. At least for these two geopolymer examples, they might make good thermal barriers, and even high temperature mullite ceramics, but would not make good refractories in a combustion zone, because of the iron content in the recipes used to make them.


Refractories and thermal barriers are both designed to withstand high temperatures, but they can have different applications with exposure to different operating conditions/ environments, so they need slightly different material properties to withstand those different settings.

A thermal barrier is meant to block heat transfer so it doesn't damage stuff outside an acceptable area. Thermal barriers primarily insulate or reflect conducted or radiated heat, usually as protection against waste heat or incidental / unwanted heat. Often that's stuff like emergency firestops, heat shielding, etc.

A refractory is also meant to withstand high heat, but in addition to reflecting / insulating, it may also absorb and retain heat directly within a heating zone itself, where that heat is intentionally produced. ..So they often have to withstand not only the heat, but also the byproducts of fuel combustion creating that heat. Very often that's carbon compounds in an oxygen-lean environment.

Aluminosilicates are pretty good at withstanding high heat in the presence of carbon, even when O2 is low.. but not when they're also contaminated with even very small amounts of iron, due to something called "the Boudouard reaction."


This article explains it better than I probably could: (not sure if you have to be signed in to an account to read it, so I'll copy/paste a snippet) www.industrialheating.com/articles/85752-engineering-concepts-co-disintegration-of-refractory-ceramic-p

"The mechanism for CO disintegration is well established. In the presence of iron compounds, carbon monoxide decomposes by the Boudouard reaction.
2CO -> C + CO2. [...] Carbon is deposited on the surface of the iron compound producing a volume expansion, which in turn produces cracking of the refractory. High permeability of the refractory will contribute to the rate of carbon deposition. Thus, as cracking occurs, the rate of carbon deposition increases and these cracks propagate under the increasing splitting force of the carbon deposit.

Carbon monoxide disintegration is controlled by the catalytic activity of the iron compounds and disintegration occurs between 572 and 1,292°F (300 to 700°C) and is most pronounced at temperatures of 842 to 1,022°F (450 to 550°C). CO disintegration will not occur when the free iron content is less than 0.1 weight percent (wt.%) or when the iron oxide content is less than 0.2 wt.% [...]

Many furnace refractory linings are produced from fireclays, which includes all refractory clays that are not "white burning." White burning clays are reserved for making dinnerware and are low in impurities, thus, are more uniform in color when fired. Fireclay refractory will contain by weight 25-45% Al2O3, 70-50% SiO2, 0-1% MgO, 0-1% Fe2O3, 0-1% CaO, and 1-2% Ti2O with an apparent porosity between 10 and 25%. These fireclays are often designated with respect to their alumina (Al2O3) content as either pyrophyllite (Al2O3,4SiO2,H2O) or kaolin (2SiO2,Al2O3,2H2O). For many heat treat furnace applications, an insulating brick with higher porosity is desired. The clay is then mixed with a fugitive organic such as sawdust, which is subsequently burned out during the firing to leave a porous refractory.

Strength and refractoriness is controlled by the alumina (Al2O3) content. A more refractory ceramic is obtained using higher alumina clay. During the firing stage the aluminosilicate clay converts to mullite, which grows as columnar grains. Impurities and excess silica form a glassy phase that is left between the mullite grains. High alumina clays will form interlocking mullite grains and this provides higher temperature strength. Firing temperature will control the degree of conversion to mullite [...]

CO disintegration is possible when Fe2O3 (hematite) is present after the firing. An excellent article by Chein and Ko [Ref. 2] shows that the size and distribution of hematite is critical in regard to CO disintegration. In general, the disintegration is more severe as the size of the Fe2O3 particle increases and as the concentration increases. In a reducing environment, the hematite grains are converted to magnetite (Fe3O4) and magnetite is a stronger catalyst for the Boudouard reaction. Also, the conversion of hematite to magnetite produces a 16 to 25% volume expansion and this contributes greatly to the cracking process. The amount of expansion depends upon oxygen partial pressure and temperature. It's important to note that the hematite conversion will also occur in other reducing environments such as hydrogen and water mixtures with out the deposition of carbon. Thus, control of the fireclay iron content and distribution is critical."

Unfortunately, the scientists who wrote the second geopolymer paper only noted the positive effect of iron on mullite formation, (which is true) ..but without considering the negative effect trace iron contamination has on mullite when heated in the presence of carbon monoxide.. That's a condition almost every refractory in a fuel burning environment has to tolerate well.. whether burning oil, gas, coal, or biomass.

Geopolymer materials composed of Si and Al have more potential as ceramics. As one of the most important clay minerals, kaolin has a high porosity, strong mechanical stability, and low thermal conductivity in geopolymers [22,50,71]. During the firing of kaolin, the type and quantity of secondary phases can have a significant impact on the thermal properties of the raw materials [48,68]. Iron oxide is very significant [49,72]; Fe2O3 can exist in raw materials as either mineral complexes or silicate structures. The addition of Fe2O3 in kaolin not only increases the quantity of mullite phase at lower temperatures (1050 ◦C), but also improves the crystallisation of mullite at higher temperatures [19,73,74].