Raku Sealing Tests

A set of 8 tumblers were thrown using Potclays Smooth Raku Clay.
The glazes used were: (to add)
After scrubbing clean, the pots were weighed regularly until no more weight loss indicated that they had dried out fully from the washing - this took just under 4 days. During washing, they absorbed about 50% of the weight of the pot in water, e.g. a 400g piece would weigh 600g after washing.

The tests applied are as follows:

  1. fill with water, leave 24 hours, pat dry and measure any weight gain - this is a short term test suitable for tableware that is used occasionally
  2. repeat the above daily for a total of 7 days - this is a longer test suitable for vases, on the basis that flowers will be fresh for about a week. And by testing daily it shows if a point is reached where no more water is absorbed. The pots are placed on dry kitchen towel, and a note is made if the towel feels wet or dry
  3. let the pieces dry until there is no more weight loss, then put in a pan with 5% soda ash and simmer for 6 hours, pat dry and measure any weight gain - this is my standard dishwasher test (corresponding to 250 dishwasher cycles), and will test both the waterproofness of the pot in regular use, and its resistance to a caustic dishwasher environment

After the waterproofness tests, additional tests will be done to determine how stain resistant the pots are, and whether or not they affect the taste of food.

The following sections discuss the applications and results. Click on an image to see it full size.

1 - Control Pot

This pot was not treated with anything, so it gives us a baseline measure of how water resistant an untreated pot is.
As expected, it was fully saturated by the end of Day 1, and remained in that state thereafter.

2 - Walnut Oil

Here I'm applying walnut oil, which is a food safe oil used on wood as well as on pots.

The left hand pic shows the control pot (left) and the walnut oil pot pre-treatment (right), sitting on a dining chair I made from elm and ash that was also treated with walnut oil - green woodwork meets Scandi design!

The right hand pic shows the application. I've followed the green woodwork approach of heating the oil to make it more fluid, and so to penetrate the pot more. Being a bit smaller than a chair, I've put the beaker into the pan, and rolled it around to ensure it all gets immersed. I took it out after about 10 minutes - be aware that the oil has a smoking temperature of 160C, so it is easy to burn yourself when taking it out! After taking it out, I've wiped off any oil left on the surface with a cloth. I've repeated this 3 times, once a day, to ensure the pot is fully impregnated.

Walnut oil takes a day or two to dry, but then a month or more to cure fully, so it'll be a while before I test this one for waterproofness! Curing is accelerated by warmth and bright light (though the light presumably only affects curing on the surface).

There are some walnut oils sold specifically for wood finishes, but their benefits over a good quality pure supermarket oil are debatable. Lower quality supermarket oils, with more additives, take longer to cure.

After 5 weeks the aroma of walnut oil was still present. It did diminish somewhat during the 7 day test.

This worked well in both the 1 and 7 day tests. It appears that by the end of day 2 it has absorbed just over 5% of its weight in water, but then doesn't absorb any more thereafter. It is one of only 3 products not to be leaking water after a week.

3 - Lithofin MN Stain Stop

This is described as a food safe impregnator for water and oils - I'm not sure if it'll result in your pots having babies! But it coats the pores and capillaries in the pot, reducing water absorbency by 90-95%, and claims not to affect the colour or appearance.

I chose this as one of the lower priced food safe worktop sealants. It is silicone/silane/siloxane based solution (same technology, different words to describe it) that forms a barrier against penetration by liquid water, but allows water vapour to pass through.

It is solvent based (87% of the content is VOCs), so should be applied and left to dry outdoors or in an area with good ventilation and away from naked flames (it has a flash point of just 36°C). Also, skin contact should be avoided and nitrile safety gloves and safety glasses should be worn.

The pic shows the control pot (left) and the Stain Stop pot pre-treatment (right).

I applied it with a cloth, pouring some into the beaker and swiping it around to be absorbed, and wiping onto the outside with the cloth. It is very readily absorbed, and also goes slightly frothy. After 15 minutes I repeated the process. Then, within 30 minutes of the initial application, I removed all the material still on the surface with dry kitchen towel - this is important as it is designed to soak into the surface, and not remain on it. After 24 hours I gave it a final rub down with a dry microfibre cloth. It is touch dry within 2 hours and fully dry after 24 hours. After 5 days it has achieved full effectiveness.

The bare clay on the underside of the pot stayed matt and the colour appeared unchanged. In the glazed areas the level of gloss may have increased slightly, and I think the coppery areas on the outside glaze became more coppery.

After 5 weeks there was no aroma.

This performed well in both 1 and 7 day tests. Though the amount of water absorbed slowly increased through the tests, the rate of absorption was possibly slowing down towards the end, and it did not seep any water by the end of 7 days.

4 - Drytreat Waterborne Dense Stone Sealer

This was recommended to me by a potter who has used it for sealing pots. Until recently it went under the name of Meta Crème.

This is a silane based sealer, impervious to water but letting water vapour through, and claiming not to affect appearance. The manufacturer says it is safe for food contact. It takes 5 weeks from the time of application to achieve full resistance. It comes with a 15 year warranty if applied by one of their approved firms, so should be durable.

It is water based, so less fumes on application, but it should still be applied outdoors or in a ventilated area, and gloves and safety glasses worn. It has a flashpoint of 94°C.

The left hand pic shows the control pot (left) and the Waterborne pot pre-treatment (right). The right hand pic shows the pot with the cream applied.

It comes as a cream that you brush on - I dabbed it on with a small foam brush. They don't indicate a thickness (and I found it hard to apply an even thickness) but I reckon it was about 2mm thick. Being a cream, it adhered to the vertical surfaces quite well, but it will be difficult to apply to the inside of pots with a small entrance that then flare out. The instruction are to leave it for 1 - 4 hours, then remove the excess using dry absorbent cloths. The longer it is left, the better the penetration, but also the harder it is to remove the excess from the surface. I left it on for 4 hours, but every hour I revisited it to dab some of the cream that had slid off the bottom back on to the pot, and to top up as needed. It didn't seem to be soaked up into the pot in great quantities, unlike the solvent based sealers, though in time some of the cream went from white to clear. At the end of the 4 hours I wiped off the excess cream with a dry cloth, and then rubbed the surface with dry cloths until all was removed as far as I could tell. They do warn that if left on too long it may dry on and leave a soapy finish, but you can avoid this by moistening the cloth with a bit of the sealer and rubbing to soften the dried material. I did find it got caught in pits in the surface, I don't know if this will be visible when fully cured. Also, they day after applying the sealant the pot had developed a slightly soapy feel, so I gave it a rub down with a dry cloth.

After 5 weeks there was a slight hint of an aroma. This had gone after the 7 day test.

Whilst this product seemed to limit absorption to around 7.5 - 8%, this was not low enough for the pot to be waterproof, and it seeped water.

5 - Tung Oil

Tung oil comes from the nuts of the Chinese Tung tree, and it has been used for sealing and waterproofing everything from wooden ships to paper umbrellas from at least 500BC. As with walnut oil, the oil cures over time, forming a solid polymer. Curing takes about 30 days.

Whilst there is total agreement on the cured oil being food safe, reports vary on the uncured oil. But, as cases of skin reactions, allergies and poisoning of dogs and horses are documented, I'd assume that there is some toxicity in the uncured oil.

I bought the pure oil, and diluted it 50:50 with White Spirit (AKA mineral spirits, mineral turpentine, turpentine substitute, and petroleum spirits for those not speaking English English), and then saturated the pot by first immersing the outside for an hour, and then the inside - very little was absorbed when immersing the inside and, as the glaze was extensively crazed, I assume this means that the clay body had become saturated. I then removed the pot from the liquid, wiped off the excess, and carried on wiping until I couldn't see or feel any surface film. At this point the glaze seemed glossier, and the colours on the outside glaze brighter, than before, but the unglazed base stayed matte. Note that if you are wiping the oil on with cloths, they may spontaneously combust if left in a heap. They should be spread out flat individually until dry, and then disposed of in the normal waste.

The left hand pic shows the control pot (left) and the Tung oil pot pre-treatment (right). The right hand pic shows the pot immersed in a slightly bigger pot of diluted tung oil, to seal the outside.

Over the longer term, tung oil is the main constituent of traditional yacht varnishes, and these break down in hot, sunny climates due to the varnish being attacked both directly from the sunlight, and also reflected off the water. This is obviously much more intense than the exposure in most households, but it does suggest that if there is a surface layer of tung oil on the pot, this may eventually degrade and lose its shine. Oil that has impregnated the pot won't be affected, as it is not exposed to daylight.

After 5 weeks there was a strong aroma. This did diminish significantly during the 7 day test.

This product seemed to limit water absorption to about 9.5%, but this is well in excess of what is required to make the pot waterproof. It would be interesting to do further tests, to see if changing the application method to that used with the walnut oil would improve the results.

6 - Eposeal 300

This is a 2 part epoxy sealer from SP Systems, part of Gurit. It's been my standard sealer for things like the insides of vases in the past, just because I had some left over from a boat building project (though I finished the last bit off with this project). Note that the epoxy was well beyond its stated shelf leaf so, although it seemed to perform perfectly well, it may not be as good as a new batch.

Whilst epoxy resins are generally deemed to be inert and non-toxic once cured, this product is not sold to any sector where food contact is likely, so it has not undergone any formal safety tests. Also, like all epoxies, if exposed to UV light it will eventually go from clear to chalky - probably not a problem on the inside of a vase that is protected from direct sunlight, but if some of the epoxy is on the outside surface then this could be a problem in the longer term.

You mix up the two parts 1:1 by volume to form a very fluid liquid. With these small tumblers, I managed to immerse the whole tumbler, and left it there for 2 hours - you can see the air bubbles that have risen to the surface in the photo. Then, as with the others, I took it out and rubbed all I could off the surface. It became surface dry within an hour or two, and gave a more matte finish to the glazed surfaces (though this effect decreased as it dried), with the bare clay staying matte. It then takes 7 - 10 days for all the solvents to evaporate and for the resin to cure. Note that, unlike most other epoxies, the mix will stay liquid and usable for a couple of days if kept in an airtight container in a cool location.

The left hand pic shows the control pot (left) and the Eposeal 300 pot pre-treatment (right). The right hand pic shows the pot immersed in a container of the mixed Eposeal, with bubbles rising to the surface.

This product has a very high proportion of VOCs, and the fumes are not good for you, so either wear a respirator with a filter against organic gases (brown in the UK and EU), or do the work outdoors. Also, wear nitrile gloves, protective clothing and safety glasses. If it gets on your skin, use an epoxy remover, not an epoxy solvent - the solvent will help the resin permeate your skin. Also, avoid naked flames, as its flash point is -4°C. It is being replaced with Ampro Seal, which is solvent free so much less harmful during application, but also a lot more viscous so I'm not sure if it will penetrate as readily.

After 5 weeks there was no aroma.

This was by far the best product, the only one to keep water absorption below 1%. However the high level of fumes are a problem during application, and although probably inert it has not been formally certified as food safe.

Note that Eposeal 300 has been discontinued. The only similar epoxy is CPES (Clear Penetrating Epoxy Sealer), to the best of my knowledge. The replacement product from Gurit is of much higher viscosity, so is unlikely to seal as well.

7 Nanoman Stone + Brick / Liquid Quartz

This product is manufactured by Nanoman and sold as their Stone + Brick sealant, but it is also bought in, rebadged, marked up and resold as Liquid Quartz by Made of Australia. The information here is an amalgam of the content of both web sites. I bought the Nanoman product through Amazon UK

The sealant is quartz based - whether it is a silane or siloxane, or something else, isn't clear from the information on either website or on the SDS, and we are also told that it involves nanoparticles. Whilst both Nanoman and Liquid Quartz web sites primarily refer to it as a surface sealer, in places they also both refer to it as a penetrating sealer. This is a bit confusing as the mode of action of the two approaches is quite different: the surface (or topical) sealers form an impermeable skin in the surface which, when breached, allows water to enter the body of the pot; penetrating sealers aim to either fill all the pores with the sealer, or coat the pore surfaces with a water repellent layer. But even if it is a surface sealer, there will be some penetration on a porous surface. The combination of the short dip time and the warnings about surface abrasion affecting performance make me believe that it is primarily a surface sealant. Life expectancy is given as 3 to 7 years, depending on conditions.

Note that Made of Australia say it should not be used on glazed pieces, including raku, so this may not be a totally fair test - though I'd expect the nanoparticles to be able to enter and seal the crazed and porous raku glazes. Anyway, we'll see how it fares when it comes to waterproofness tests later...

The left hand image shows the control pot (left) and the Nanoman pot pre-treatment (right). The others show product shots of Nanoman and Liquid Quartz.

Application is easy, dipping the pot for 20 seconds then wiping off excess liquid to prevent pooling as it dries, and then polishing glazes surfaces with a cloth after 15 minutes. It rapidly became touch dry, and is fully cured after 5 days. It looks as though the metallic glaze may be slightly brighter after treatment, but not by much if anything.

To my surprise, it had no noticeable waterproofing effect compared to the control. Perhaps the crazing in the glaze had cracks that were too small for it to penetrate.

After 5 weeks there was no aroma.

8 - Milk

Sealing with milk has been recorded by potters in a variety of locations, using a variety of techniques. Milk is the last sealant to be tested. Archaeologists have found evidence of pots being sealed with milk in early Neolithic times (4000BC) in Europe, and more recently (relatively speaking) it has been used in Tibet, Ethiopia and SW North America, so I think it fair to assume that the practice was pretty widespread, at least until the development of glazes.

There is some debate as to whether it is the milk fat or the protein that seals the pots. The archaeologists say it is the fat, but the proteins in milk (especially casein, but also whey) have long been used as adhesives and sealants, particularly in woodwork (e.g. Stradivarius used it to glue his violins together around 1700, and the glue joints are still performing well), so they probably play a part as well. It seems that potters either dunked the hot pot in milk, or soaked it in milk and then heated it. I've looked at how milk degrades when heated - it is complex, and I haven't worked out the benefits. Many casein adhesives don't use heat, but add lime to polymerise the casein, so that would be another approach to follow.

Here I'm soaking the test piece in milk over 24 hours, longer than the others as the milk is more viscous and there is no heat to help it penetrate. As the milk doesn't totally cover the pot, I'm rotating it every couple of hours to get an even coating. Then I'm putting the pot into an oven heated to 250C for an hour to seal it - giving off that aroma of burning milk!

To my surprise, it had no noticeable waterproofing effect compared to the control. Perhaps the crazing in the glaze had cracks that were too small for it to penetrate. Also, milk is used as a glue or sealant with a variety of application methods, so perhaps a different method would work better.

After 5 weeks there was a very slight aroma. This had gone by the end of the 7 day test.

Silica, Silane and Siloxane

There are a variety of silica based sealing technologies. They all benefit from using very small molecules, that can penetrate more deeply than epoxies, urethanes and acrylics, and so are not reliant on a waterproof surface skin on top of the pot surface. Both silanes and siloxanes react with the silica in the pot to form a siloxane resin film that is hydrophobic, repelling liquid water but letting water vapour through. This is possible because of the inter-molecular bonds in liquid water, meaning that it cannot get through, whereas the unattached water vapour molecules can.

Silanes are the base molecule for siloxanes and, being smaller, can penetrate more deeply. In curing, they react to form siloxane polymer chains, but need a high pH (alkaline) to do this. Siloxanes have larger molecules, consisting of 3 or 4 silanes bonded together, so they cannot penetrate as deeply, but in curing they form the same siloxane polymers without needing a high pH environment. Another difference is that the silanes are volatile during application and curing, whereas the siloxanes are not. Silanes generally do not affect the surface appearance, whereas siloxanes may darken it slightly.

Soak Tests

Each pot was weighed, filled to the brim with water, and left to soak for 24 hours. The water was then poured out and the surface patted dry with a cloth, the pot was weighed again, and the water absorption calculated as a % of dry weight. This was repeated 7 times, so ultimately representing a week's soak test.

The first set of results represent the standard porosity test, where a weight increase of less than 0.5% indicates a totally waterproof pot. The 7 day test is one that is used for checking the waterproofness of vases, on the basis that this is how long a bunch of flowers is likely to last.

Pot No.12345678
SealantControlWalnut OilStain StopDrytreatTung OilEposeal 300NanomanMilk
Dry wt (g)256.8275.9261.1272.6267.8255.2260.9255.1
1 day wt286.8290.1263.9292.4292.6256.2292.2284.6
1 day %
2 day wt287.3291.3268.4292.9292.9257.1292.5284.6
2 day %
3 day wt287.2291.4272.2293.3293.1257.2293.0285.0
3 day %
3 day paperwetdrydrywetwetdrywetwet
4 day wt287.3291.4274.8293.1292.8256.7292.9285.0
4 day %
4 day paperwetdrydrywetwetdrywetwet
6 day wt287.8291.8277.3293.9293.3257.2293.2285.3
6 day %
6 day paperwetdrydrywetwetdrywetwet
7 day wt287.8291.8277.7293.7293.5257.6293.7285.6
7 day %
7 day paperwetdrydrywetwetdrywetwet

Other Sealants

There are many other sealants for brick or stone, though not all have been certified as food safe. Here are results on some of them carried out by a builder: www.pavingexpert.com/sealants-for-paving
Other vegetable oils that polymerise on drying, as opposed to going rancid, should be tried. This includes almond, coconut and linseed (flax) oils.