Sciencemadness Discussion Board

DIY Alumina refractory

Rainwater - 12-3-2022 at 19:34

The recipe im using has served me well over 5 years. The prices of the ingredients made me jumpnwhen i saw them. Complete bs.

So, after a busted crucible full of precious metals, it's now time to construct a new kiln floor before i rip out the old one and recover my cash. I did some searching here and could only find low temp setups or ridiculously high temp setups. So, I thought I'd share what I do in the hopes it could be useful to someone. And maybe I can improve upon my setup.

This recipe is good for my needs. And cheap. Dirt cheap.
1500c 100+ firings with minimal degrading. Ramp rate 15c/min, electric heat.

Now, the true purpose of this post is ...
These ingredients are difficult to source in my area.
Alumina is $90 usd for 10 kilos, quick lime is sold in 50lb bags $30usd.
Aluminum hydroxide is unavailable locally and $35+shipping a kilo online.
Which would total my repair to 180 bucks + a lot of work

Alumina firebrick are $12 each, need 16 to fix my floor properly
So thats $192 + tax + 2 hour drive, one way. But not much work

Scrap aluminum and limestone are all thats needed, but a few extra steps are involved. My local scrap yard lets me buy aluminum for 20% over their cost.
So i got $16 dollars worth of extruded aluminum scrap. 10kg
I picked up some lime chips from the hardware store, $4 for .4m³
And refilled the propane tank, $35
So im $55 bucks in.

aluminum can be converted into aluminum hydroxide.
Aluminum hydroxide can be converted into alumina
Limestone, when heated, becomes calcium oxide.

Ingredients
All powders are 100 mesh or finer.
Quantities are by mass

A propate burner with high pressure regulation, firebox, or jabod to act as an oven/kiln
Temp controller. This is optional of you like drinking beer all day and tending the fire for 8 hours.

75% - alumina powder.
10% - Aluminum Hydroxide powder.
10% - Silica sand
5% - Calcium Oxide
5~8% water
Graphite powder (optional)
3% perlite (optional. Increases r-value, decreases max operating temp)

First, I construct a form from 2x4 wood. Internal dimensions are 5x9x3.5. Bigger can be done but is prone to cracking

The form is wetted and coated with graphite to act as a release agent.

All dry ingredients are mixed before water is added
The mixture should be damp and crumblely
It takes roughly 500g of the final mixture poured into the form and compacted with a hammer to fill it.
If i had a hydrolic press, it would be a lot better, and I'd just use alumina.
2-hour set time followed by a 4-step pre firing
100c 2 hours.
300c 1 hour.
500c 2 hours
750c 2 hours

If you can control the ramp rate properly, 3c/minute till 750, then soak for 1 hour.

After the prefire, the bricks are placed in the propane furnace, soaked @ 1900c for 30 minutes.
After they cool, their sanded to their final size.

The purpose of the calcium oxide is to strengthen the brick until the aluminum hydroxide decomposes and fuses. Afterward, it serves no purpose and actually causes problems if it gets wet. It can be replaced with portland cement but I don't like the results. It seems to crack easier after the final firing.

During the prefire, the aluminum hydroxide decomposes into a wet sticky mess that helps hold the brick together until it can be properly fused. As well as creating bubbles inside the brick that aid in insulation.

Aluminum to hydroxide.
A deep stainless steel pot will serve as the cathode+ and the scrap aluminum will be the anode-
Suspend the anode into the pot without touching it. Fill the pot with water, and add Epson salt. I use a 2.5qt pot and add 150g epson salt. More salt does not increase the current. KHSO4 works best, and H2SO4 works but damages the cathode. The more soluble the sulfate, the better the results you get.

Your electrolyte will turn gray,
1kg of aluminum will make about 2.8kg of aluminum hydroxide.
2.8kg of aluminum hydroxide will make about 1.8kg of alumina.
Assuming perfect efficiency
540000 columbs are needed, which adds up to 10 amps of current for 300 hours
Or around 62 hours of electrolysis per brick.

You will not be able to filter the solution easily.
Pouring it in a tall cylinder and letting it settle, then siphoning off the upper water layer is the best method i have found. The water layer contains most of the magnesium sulfate and is returned to the cooking pot for further electrolysis.

Drying in a baking dish in an oven set to 150c/300f for an hour produces a nice gray powder. I believe the color comes from contamination in the source metal.

Making calcium oxide.
Heat calcium carbonate (lime stone) to about 900c

Fantasma4500 - 14-3-2022 at 10:10

i would urge you to try with maybe calcium silicate instead of sand, or even sodium silicate- you probably know already a mix of sand and sodium silicate can work for some medium-efficient refractory setups
reacting sodium silicate with just CaCl2 will give you immediate precipitate of calcium silicate, dries up to a nice uniform powder. i tried making mixes of sand, NaSiO4 and CaSiO4 but found the CaSiO4 ones to be crumbly- but with your mixes im sure this would change as it contains many other things
iirc calcium silicate is incorporated into kiln cement mixtures along with .. was it sand powder, and definitely sodium silicate

as for the graphite, i would mess around with adding some paper pulp instead, as you mentioned having gas cavities, albeit in a small amount aids the refractory quality, that would indeed be helped with adding paper pulp to it
its made easily by dumping newspaper and similar paper waste into a bucket of water for a few days then stirring it, scooping it, squeezing and drying, it can then be ground up fine with a blender or similar- coming to think of it, sawdust would probably do just fine.

i remember AvE one time produced some excruciatingly refractory carbon material by taking a piece of sandwich bread and putting it into an airtight steel container, loaded with argon, then torching it for a while. just one slice of this was enough to cover up for an oxy-acetylene flame directed at his hand

the aluminium hydroxide is very tricky to deal with, i remember a fellow chemist dealt with it in order to make Al2O3 by simply drying it out and then heating it in oven, but its probably best if its in a kind of gelled state before the refractory material is finished

another source of aluminium oxide is dirt, up to 40% by my vague memory, ontop of that simple dirt could probably also do quite well. the aluminium in dirt may be extracted using sodium hydroxide forming sodium aluminate.
clay is also composed of materials that could prove useful however unless you can mine your own clay this might not be very profitable. maybe calcium hydroxide could be reacted with aluminium in the mix? the problem would probably be particle sizes, coffee grinder aluminium would maybe work but tedious work, and bulk aluminium powder could be difficult to outsource, even with particle sizes outside of the pyrotechnics use frame

since it appears youre evaporating off large amounts of water ill toss in one of my insomnia ideas for rapid liquid evaporation: steel pipe, steel sponge, hotair gun and liquid inlet + crude mesh at the bottom of steel pipe
a liquid is pumped into the tube, coming in from the side, the hot air blows it down into the tube where it collides with the steel sponge which acts to increase surface area
the hot air rapidly causes the whole lot to evaporate, the device requires some manipulation of air flow/temperature aswell as liquid flow, the system can be flushed by turning off the hotair gun and letting the liquid to be evaporated flush out the other parts, or simply the sponges can be swapped

"It's clearly a budget. It's got a lot of numbers in it." -George W. Bush

Rainwater - 14-3-2022 at 17:10

I got my +/- signs crissed crossed in the electrolysis setup.
One improvement I have made to the electrolysis setup is to use a steel pipe. 4in diameter, which is clamped in place and suspended inside of a 5 gallon bucked. Then, place my scrap tubing inside of it. This is a much better cathode due to its closer proximity.
Quote:
As for the graphite, I would mess around with adding some paper pulp instead

Sorry for the confusion. The graphite is not added to the mix. It's only used as a mold release agent. At the temperature I normally run, it will turn to co2 and float away
Quote:
i would urge you to try with maybe calcium silicate instead of sand

The sintering temp of alumina is 2600c far beyond my ability.

Silica sand aka Silicon dioxide, has a melting point around 1700c.
I get it as " pool filter sand".
That is the primary fusing agent that holds my brick together after the final firing.
I use a propane forge burner to heat my furnace, 1900c is the max i can reach. Even that is difficult. Usually only reaches 1850. But anyway, when the molten silica mixes with the alumina it freezes and holds everything together.

Not sure about calcium silicate.... ive found confilicting information on google
no sintering information given except 1 page saying 600c-800c
https://cfpub.epa.gov/si/si_public_record_report.cfm?Lab=NRM...

I like the pool filter sand better than play sand. But i have used river sand and other than color, couldn't notice a difference in the final product.
Quote:
the aluminium hydroxide is very tricky to deal with, i remember a fellow chemist dealt with it in order to make Al2O3 by simply drying it out and then heating it in oven, but its probably best if its in a kind of gelled state before the refractory material is finished

I dry my into a powder. Usually, boil all the water off on the stove until it's a thick paste, then finish it in the oven overnight. My research says that it must be heated to over 600c, don't remember the exact number. Before it turns into an oxide,

Mechanism and Kinetic Parameters
of the Thermal Decomposition of Gibbsite Al(OH)3
by Thermogravimetric Analysis
Attachment: a131z3p62.pdf (821kB)
This file has been downloaded 308 times
Quote:
another source of aluminium oxide is dirt, up to 40% by my vague memory, ontop of that simple dirt could probably also do quite well. the aluminium in dirt may be extracted using sodium hydroxide forming sodium aluminate.
Georgia Red Clay here. Iron oxide rich.
How is the sodium removed from the sodium aluminate?

[Edited on 15-3-2022 by Rainwater]

Fantasma4500 - 15-3-2022 at 07:23

Mechanism of the Precipitation of Alumina from Sodium Aluminate Solutions
https://pubs.acs.org/doi/pdf/10.1021/ja01225a010

so this forms hydrated Al2O3- reacting sodium aluminate with acid

reacting sodium silicate with acid forms SiO2, and it can be very fine particles ontop of that
when i made sodium silicate from SiO2 kitty litter i would get a block of gel at the bottom of the bucket i made it in, appeared to be some sort of aerogel- which is around the highest level of insulator you can get.
i just simply poured in 20% NaOH solution, about twice the volume of the kitty litter. temperature was around 20*C coming to think of it, the kittylitter SiO2 you can buy is probably some type of gel already, maybe you can somehow stick that to the surface of your bricks?
about aerogel- theres some demonstrations of it on youtube, i remember one where they took a redhot glowing block of it and could hold it in their hands because the outside of it wasnt really warm while the core was glowing

electrolysis is very straightforward and may give a decently pure product but it can be very tedious
as for keeping the electrode parts apart, just put in some pieces of wood, having all of it balanced and set up can be a pain sometimes

as for CaSiO4, you say sintering, it says here that it decomposes at 1100-1300*C, which for your setup isnt really wild.
THE DECOMPOSITION OF TRICALCIUM SILICATE IN THETEMPERATURE RANGE, 1,000° -1,300° C
https://nvlpubs.nist.gov/nistpubs/jres/7/jresv7n5p893_A2b.pd...

maybe barium silicate would be the next thing? barium silicate would be a lot heavier and would then also hold more heat, which would translate into worse efficiency. its listed as 1.67g/cm3, and melting/decomposition temperature 1600*C

WGTR - 15-3-2022 at 08:17

I'm just going to throw a few bits of information in here in case they're useful. I normally sinter small alumina/bentonite (90%/10%) ceramics by soaking at 1200-1220°C for 1-2 hours. The porosity is high but there is also low firing shrinkage and the end result is very tough. If I drop small fired articles with this composition on the table they just bounce without breaking. I also use an alumina/Veegum T (97%/3%) composition with the same firing schedule and results. The alumina powder is sourced from Aardvark through a local pottery supply store. The high melting temperature of alumina is actually a good thing when it comes to refractory bricks. Zirconia is another thing people like to use.

Alumina powder is expensive partly because it is a pure material. You can also source alumina from clay (a silicoaluminate), which is the normal way of obtaining it. I picked up a 50lb. bag of Edgar Plastic Kaolin for $20 recently. That has about 45% alumina (after firing), the rest being silica. You can add extra alumina to this if you want to change the ratios.

Silica comes in different mesh sizes. The reason is that when firing a clay composition the silica will dissolve to an extent depending on its mesh size and the firing schedule, leaving various amounts of free silica in the fired article. This affects various properties of the ceramic such as fired strength and temperature coefficients, etc.

Using a super-plasticizer like Veegum T is great for plasticising/binding a non-plastic material like alumina, but it greatly increases the drying time. Wet alumina by itself may dry out in room air in a few hours, but with a few % of Veegum T this increases to days or weeks. In other words, one can't mix a slurry of this stuff with alumina and pour it into a mold as large as a brick. It might never actually dry. It would also shrink to about half its original size if it did finally dry. It would have to be dried in thin slabs, powdered, and semi-dry pressed to avoid this problem. Bentonite is much better regarding the drying problem, but it takes more of it and the final article is less pure (This is for high alumina technical ceramics).

I'm not sure how insulative your brick composition is, but it looks like a good mix to me. Thanks for sharing it. The CaO acts like a cement for green strength, especially in combination with the silica and water. This probably isn't practical (or necessary) for brick making, but I melt hard microcrystalline wax into green alumina articles for green strength. After doing that you can carve it with a knife, drill holes, etc, before firing.

I may try out your idea for making alumina powder if I need large amounts of it.

Rainwater - 15-3-2022 at 17:28

Quote:
I'm not sure how insulative your brick composition is

an r value between 2-3 per inch. That led to the optional perlite. Which doesnt give consistent results. I think its the distribution but never looked into it.

Funny story
The first build, I used a single layer of bricks, long side up, to line the table the kiln sat on. It was a disaster.

I was in my driveway doing my first burn. The first one always smokes.
My table heated up...
Cars began to stop in the street and watch
There was a white cloud with a lot of cussing from a man in a big red truck with perty lights who would not lissen when i said just wait for it to cool, NO WATER.

But I did get to 1000c. Calculated time was 45 minutes, real time was 4 hours

After that, I rebuilt with a 3 layered base.
The first layer is air channels, and the second is my mix with perlite, the third is without perlite. which solved the problem. I left a slot in the floor so I could probe the layers, the air gap gets 70c, and the perlite reaches 800c with a fire of 1500c.
Warn up time is about 45~60 minutes.

2 240v 25ohm MoSi2 type heaters if anyone is asking. Not cheap, but I've gotten almost 500 hours without an issue. 2200/5400 watts. (Custom controller required. Have to be started at a half voltage until 400c.)

Fantasma4500 - 16-3-2022 at 08:28

talcum is Mg₃Si₄O₁₀(OH)₂ - hydrated magnesium silicate
if aluminium oxide is great, i bet magnesium oxide would be great too

There was no change in the crystal structure of the talc up to 800° C. At
800 to 8400 C the t alc decomposed to enstatite, amorphous silica, and wat er'
vapor. The enstatite gradually changed to clinoenstatite around 1,200° C, and
the amorphous silica changed to cristobalite around 1,300° C, giving clinoen- statite and cristobalite as final products.


talc:Mg₃Si₄O₁₀(OH)₂
enstatite:MgSiO3
amorphous silica:SiO2
clinoenstatite: MgSiO3
cristobalite:SiO2

Cristobalite is a mineral polymorph of silica that is formed at very high temperatures. It is used in dentistry as a component of alginate impression materials as well as for making models of teeth. It has the same chemical formula as quartz, SiO₂, but a distinct crystal structure.

sucks i dont have a backyard, i would weld up a small brick mold and mix up some NaSiO4 with talcum and wood pulp

maybe look into how gas concrete is made, maybe you can make some adaptation of that? ive seen homemade devices that relies on pressurized air but also a gelling agent
now combine that with your alumina or alumina yielding mixture and you could produce very low weight bricks at home- high pressure air is what i'd be trying to dodge adding in pieces of combustible organic matter

out of curiousity what did you really pay for those MoSi2 heating elements? ive looked around the internets a few times and never found any actual prices on em

Rainwater - 16-3-2022 at 08:55

I got them about 2 years ago, total was around 450-500. W shaped 28in each. The controllers are ridiculous so i just built one with a few relays and an esp32.

Cold they have a restance of 14-18 ohms so i put them in series to reduce the average voltage to 120. Once the element reaches 300-350c the resistance is stable and full voltage can be applied by switching the relays to operate the elements in parallel.

Edit
For anyone wanting to reproduce my results.
Here is my CYA statement, considering safety practices, hazards, and liability.
Don't do this without proper equipment, training, and safety protocols.
I take no liability for the actions of persons who attempt any procedures outlined here.
Bllaaaa blaaaaa do not try at home...

Using sulfuric acid or epson salt to create the aluminum hydroxide by electrolysis releases hydrogen sulfide gas. Stinks/toxic bad for you. I don't know if this is a normal byproduct or unique to the impurity in my reagents, or the current density of my setup
I have found that using an electrolyte composed of potassium bisulfate does not produce any noticable smell

Aluminum hydroxide, aluminum oxide, calcium oxide, and silicon dioxide are respiratory hazards, and long-term exposure can cause serious health problems and death.

Use properly rated breathing protection when handling these materials and cleaning the work area.

Acids BURN. Bases BURN. 1900c shrapnel cuts, penitrates, and burns.

Rise rates and temperature control are very important to allow the outgassing of decompossion products and water.

Be prepared for exploding bricks.
Be prepared for complete dusting( when a brick turns to powder in the kiln )
Stack bricks in a proper pattern, so the complete failure of 1 does not collapse the heap being fired.
Be mindful of the weather. Ive had perfect firings get ruined because of rain shattering my firepit.

[Edited on 16-3-2022 by Rainwater]

timtinker - 10-6-2022 at 04:12
Quote: Originally posted by Rainwater  
This recipe is good for my needs. And cheap. Dirt cheap.
1500c 100+ firings with minimal degrading. Ramp rate 15c/min, electric heat.


Hi there Rainwater etal,
I noticed your reference to a refractory as “dirt cheap’ so I thought I would contribute another idea that is literal dirt cheap refractory. I hope I am not too late for this discussion and that it is on the most appropriate topic.

I tinker with DIY wood-fired/forced air firing of stuff that needs a suitable combustion chamber that is heat and chemically resistant. With some success, I have experimented with sodium silicate-based refractories (with aluminium oxide, talcum powder, garnet abrasive grits, sand (no good), clay, and crushed basalt dust.

The last one on the list was found by accident. It is essentially fine crushed and sifted loamy soil with minimal clay content (crab-hole-mounds or farm post-hole-soil) mixed with sodium silicate. They all seem to stand up to 1,000C and most go well above this temperature except for those with added sand. Here is a link to my post on the subject:
DIY refractory

I have only recently purchased a thermocouple with an upper limit of 1,300C and even so it is difficult to know what temperature my refractory reaches down near the charcoal (where my precious thermocouple does not dare go).

Right or wrong I have concluded that my mix of DIY sodium silicate (DIY low SiO2/Na2O ratio est 1.84) with crab-hole-soil (or post hole) has been the most promising refractory for my furnace making. It can be fired to a rock-like refractory at about 500C without a great dwell time to apparently cure throughout. It is cheap and abundant and makes a solid refractory, but it, unfortunately, is not a good insulator for a furnace build. It is more appropriate as a hot face or liner.

This soil silicate refractory can puff or swell (intumescence) during the first hardening firing to ~500C. I think it is steam bubbles that form in the amorphous silica gel or polymer as the chemically bound water in the silicate is the last to be removed as the refractory is fire curing. The image below is of a test puck (T6) that puffed this way, but it is still hard enough to sharpen a knife on it.



Unfortunately, I found this to be a hit-and-miss way of forming insulation bubbles and it can result in uncontrolled steam explosions (does that qualify as madscience?)!

I have added perlite or vermiculite to my refractories to provide insulation. They speed the drying/curing and seem to create thermal shock resistance. Unfortunately, these additives start to melt at ~1,000C. I think the sodium from the sodium silicate act as a flux in this regard. However, it can be used behind a hot face where temperatures are much moderated.

I next tried clay (commercial white school clay) with fine paper fibres dispersed in it (paper clay). It at first seemed to fire OK as small test samples. However, large items eventually failed when exposed to repeated hot and cold cycling. I think this was because the clay body did not get hot enough for long enough to vitrify throughout the clay body. This will be an ongoing problem when building a furnace with it.

By contrast, the soil silicate only needs ~500C or even less to become stable. I have lined my current furnace with this refractory as a render. The furnace can melt white student clay and cause it to slump. I think this is about 1500C or maybe more. At the same time, the furnace wall has gone glass-like and smooth, but it has survived. Even if it has softened on the surface, it has not slumped, as I think the solid layers below the surface provide support against slumping.


Another ceramic on my list is crab soil silicate and paper fibres (egg cartons). It has all the wonderful tolerant molding & shaping, quick-drying without cracking, easy firing, insulation and thermal shock resistance properties of paper clay. I think it has the distinct advantage of being able to be converted into a stable refractory (throughout its core) with modest firing temperatures of about 500C. This means that, unlike the clay bodies, it can simply wait in a stable state within a kiln wall, ready to be fired to its ultimate temperature on its hot face. I have a post on the subject:
Soil paper silicate refractory

Could this soil/silicate/paper refractory be a good DIY refractory for forges, furnaces and kilns?

Hydrated lime (Calcium hydroxide) with sodium silicate is another refractory that is suggested in this Youtube video
Lime refractory video

At$12/ 20kg bag of fine lime powder, I thought this would be much easier than processing crab hole soil. I have tried to make the refractory glue and castable refractory as suggested in the video and neither worked. They just fell apart. Here is my post on my testing of it:
Lime silicate testing


Lastly, I must tell you about my little accidental refractory discovery that I have not seen reported by anyone else. It is great for filling holes and packing around fittings in ceramic things that get hot.

I paint (or scrub) sodium silicate onto aluminium cooking foil with a toothbrush. I add a little aluminium oxide as an abrasive to scratch the oxide film on the foil to speed the reaction, baby talcum powder for a little extra viscosity and the smell of roses during the cure and some iron oxide for good colour to observe the cover of the silicate. None of the extras is essential.) Then I fold it over and paint again and fold…… Then I stuff the laminated aluminium it into cracks and holes to form a self-curing and expanding refractory filler. As it is scrunched into place it starts to heat up, expand, bubble, blow off hydrogen gas and slowly go solid. The cure can be further accelerated with a flame. The burning of hydrogen gas is an added bonus! The surface can be finished with a render of aluminium oxide and silicate as a tough render.

I hope this little contribution qualifies as madscience and may encourage more dirt-cheap DIY refractories.
Tim


Fantasma4500 - 3-10-2022 at 05:57

i just wanted to chime in and say that flyash contains about 25% Al2O3 (and its all super fine powder)

https://www.researchgate.net/profile/Purwanto-Khusnan/public...

some CaO present as well so it should be very easy to turn into concrete

Fantasma4500 - 30-10-2022 at 01:06

https://en.wikipedia.org/wiki/Fly_ash_brick
it has already been done