Jump to content

Recommended Posts

Crystalline Glaze Chemistry


Several have inquired about this specialty glaze in recent months, so I thought I would post an introduction into crystalline chemistry. The very first thing to do with this glaze is: throw out your glaze calculator and UMF spreadsheet: they do not apply to this glaze nor can you bring it into unity.


Basic  cone 6 through 10 crystalline recipe is:

50% ferro frit 3110, 25% zinc oxide, 25% silica 325m, and a seeding agent (see later).

^ Lithium carbonate flux additions discussed later.


This is the basic crystalline glaze recipe, regardless of what cone you are firing to. However, crystalline glaze requires high fluidity of the glaze, so a high cone five is the minimum temperature level. It can be fired under five, but it takes a fair deal of chemistry to achieve it; so perhaps later it will be discussed. The reason you see so much variation in this basic recipe is because so many variables effect its outcome. Most of the variations are artistic preferences being inserted to control crystal size, population, and growth patterns. Likewise, the wide variance in firing ramp holds are due to kiln size, crystal development, peak cone temperature, and in many cases because the kilns were never calibrated by using cone packs or pyrometers to adjust thermocouple readings. An example: a ramp hold is stated as 2002F, but that same temperature could be 1992 or 2018F in your kiln.


Frit 3110 is the standard and most reliable frit used in the crystalline recipe. There are frits used such as Fusion Frit 644, and others. If Ferro Frit 3110 is not available in your region, then use the chemical analysis of 3110 as a guide to compare available frits. Match the chemistry of your available frit as closely as possible to the chemistry of frit 3110.


50% of the basic recipe is frit, it alone makes up the glassy matrix of the glaze. Although silica is being added to this recipe, that silica addition serves a specific function other than being a glass former. Yes, the silica will add to the glass content, yet it is being added for feedstock for crystals to form. Frit is ground up glass that has already been melted at high temperatures; so the chemistry required to form glass in a glaze is not a consideration. Crystalline glaze requires a thick application of glaze, and crystals require a thick layer of molten glass to grow in. So lowering the frit level in this basic recipe also means you are restricting the glass level in which they grow. You rarely see the level of frit deviate from this basic recipe, and when it is adjusted: those adjustments are minimal. 


Crystalline glaze or crystals are chemically known as zinc-silicate crystals. Crystals form an ionic bond between zinc and silica: which is why equal parts of each comprise the basic recipe. Yes zinc adds opacity to the glaze and silica is a glass former: but neither addition is present in the recipe for those reasons. Zinc and silica are added and adjusted to form crystals, the additional glass formed or opacity added are secondary reactions. So the adjustments you see in various recipes are to manipulate crystal growth and population, and secondary “glaze chemistry†is not a consideration.


Nerd (TJA 2017) 

Edit 8-12-2017 for spacing and grammar.

Edit 8-13-2017  Explanation of temperature in first paragraph amended.

Edited by glazenerd
Link to post
Share on other sites
  • Replies 74
  • Created
  • Last Reply

Top Posters In This Topic

Top Posters In This Topic

Popular Posts

When I first became interested in crystalline glazes (in a school setting where we had digitally controlled kilns) I had only a manual kiln in my home studio. Fortunately my kiln had "infinite switche

Tom,   The comment quoted below suggests this is about more than facts for you. I'm not willing to get into a pissing competition about this stuff. I offered up my suggestions and you come back w

Um, band gaps don't expain the spikes. I've never heard of them used outside of discussing conductivity. Ions are ions. I think that's where the confusion is coming from. Reactions happen when ion

Zinc oxide comprises 25% of the basic recipe, and is used to control crystal size and population. The basic recipe insures some crystals will grow, and is adjusted up or down to control their development. However, if you drop below 23-24%; your crystals will be few and far between, and their size will be minimal. To increase size and population, the recipe is adjusted to a maximum of 30%. At 30%, there will be an explosion of crystals to the point where they grow into and on each other. The usual adjustments are 26-27 and 28% to increase population and size. While artistic expression determines the final level of zinc, it is a generally accepted rule that some open field of color separates the crystal population.


The other important point of crystalline chemistry is the zinc variety used in the recipe. Most potters do not realize there are three primary varieties of zinc. Yellow zinc, as it is commonly called in pottery is sphalerite zinc ore, which is the most impure variety of zinc. Even after being mined and processed, the maximum zinc content typically averages 80%. However it is the most commonly used zinc in crystalline glazes, primarily because it is the most widely available. It should also be noted, that it is the only zinc variety that can be used above cone 7, now that Cerox has been taken off the market.


White zinc is processed from pure zinc ingots through a process of vaporization, and the resulting vapor fumes are collected. This French method of processing either pure zinc ingots, or metallic zinc produces a white zinc powder with a purity level above 99%. While the purity does play a minor role in the overall outcome of this glaze: primarily because the impurities found in yellow zinc often produce random spike and needle crystals in the glaze. However, white zinc has strict limitations on how high it can be fired before it begins to vaporize out of the glaze: typically around 2260F. Cerox was the only French process zinc that could handle be fired to cone 10, and it was discontinued two years ago. Blue/grey metallic zinc is the third variety, but it is very rare to see this type of zinc used in crystalline recipes.


Important chemistry notes about both types are zinc are as follows: yellow zinc requires more flux at lower cone temperatures than white zinc. If you over flux white zinc, it will vaporize at lower temperatures than stated above. Use only yellow zinc if you plan to fire above cone 7.


All zinc additions are used to control crystal size and population. Lowering below 25% will decrease their size and crystal population. Raising zinc levels above 25% will increase population and to some degree their size. Often you see recipes with 26.5 or 25.9% zinc, these types of specific recipes are potters dialing in the zinc level to produce the population and size that satisfies their particular expressions. In starting out your crystalline journey, start by using 25, 25.5, 26, or 26.5 (.50) additions until you reach a crystal population that you like: then make 0.25 adjustment to dial in your final recipe.


Nerd (TJA )


Edit 8-12-2017 for spacing.

Edited by glazenerd
Link to post
Share on other sites

Silica additions are used to control crystal growth, and like zinc when you hit a certain minimal level: development drops off. If you want to keep crystal population low, and their size smaller; then silica additions should run around 14-16% of the recipe. If you want to grow larger crystals, then additions between 18-22% are commonly used. Increasing silica additions are also used to control crazing in this glaze, which is notorious for crazing due to its chemistry. Do not concern yourself with controlling crazing issues until you have adjusted the recipe to suit your personal taste.


Titanium dioxide is the seeding agent used to nucleate crystals. The black specks you see in the B&W photo below are tiny specks of TiO2. Most often I only use 0.10% of titanium to seed my glaze, and that is all that is required. However, you often see recipes calling for nearly 10% of titanium, or other additions such as rutile or ileminite that are both high in titanium. These additions are used for other reasons other than nucleation of crystals. Most commonly they are used to alter color, but more so because reduction firings are planned after the primary glaze firing. These levels of titanium are used to produce a purple color in reduction firings later on.


Lithium carbonate is used as the primary flux in crystalline glazes. Lithium, silica, and zinc all have hexagonal crystal lattices on an atomic level. Secondly, lithium has a much lower molecular charge than zinc or silica, and therefore does not interfere with the ionic bonding required to produce zinc-silicate crystals. This molecular balance is most often described in articles on crystalline glazes as; “lithium has an affinity for zinc and silica.†The other fluxes that are compatible in crystalline glazes are calcium and boron. Both of these fluxes are added as secondary fluxes, and usually in very low levels. Many feel that small amounts of boron aids crystal growth, and I have found that to be true. Calcium is used more to produce a mottled background field in the glaze, more so than the typical use as a flux.


Lithium additions are dependent on the peak firing temperature and the zinc variety used in the recipe. At cone six, 3% is used with white zinc, and 4% is used with yellow zinc. As the peak temperature rises, the lithium additions decrease. At cone 10, less than 1% is added and often none is used if a hold at peak temperature is incorporated in the firing schedule. Lithium additions are a balancing act: this glaze has to be fluid enough for crystals to form, but not so fluid that all of the glaze runs off into the catchers. So it is common to see lithium additions stated as 3.25 or 3.75; which only means that recipe has been dialed in to the peak firing temperature used by that potter.


Firing schedules are broken down into two primary functions; the first being peak temperature used to melt any given recipe: the second being ramp holds at lower temperatures that actually grow the crystals. Almost all firing schedules seem to be vastly different from each other, and that is because of kiln size, kiln calibration (or lack thereof), and type of crystals being developed. Typically when a cone 10 peak temperature is being used, the corresponding ramp holds run between 1998F and 2032F. When cone six peak temperatures are being used, the corresponding ramp holds run between 1930F to 1976F.  These temperatures are highly generalized and are meant only to give you some sense of reference.


Kiln size effects crystals, proximity to the elements effects crystals, and hot and cold spots within the kiln chamber effects crystal development. Sneezing while loading the kiln effects crystal development, they are very finicky little creatures. (pardon my humor) The fact of the matter is, it takes a considerable amount of time to dial in the best ramp hold temperature for your kiln: and every kiln will be different.  So start in the middle of the temperature ranges given above and move up or down from there. After you first test firing you produce spikes and clusters, make a ten degree downward adjustment in your ramp hold temperature. This should produce what are called “maltese†crystals, from there make 5 degree downward adjustments until round floret crystals form. Unless you like spikes or maltese crystals: then make no adjustments at all.


All colorants used in crystalline glazes are metallic oxides: stains are not used and will not work to color crystals. Copper, cobalt, manganese, iron, and nickel are the most commonly used. Like most glazes, as you move up the ladder from low to high percentages, the depth of color increases. I would recommend you start with just copper and iron until you dial in your basic recipe. Both are reasonably predictable and much cheaper to use while testing. Once you dial in your basic recipe, then the sky is the limit in how you blend various colorants. It is very common to use multiple colorants to produce effects in this glaze, you can find all kinds of variations and their results online. Colorant additions gets into a rather complex explanation about how they interact with each other, so I will save that discussion for later.

Clay: use only a good quality porcelain to start your crystalline journey on, it will produce the best and most predictable results. Crystals will grow on certain blends of stoneware, but you can explore that once you learn the basics.


8-12-2017 Edit addition: Glaze application rate for vertical surfaces is typically 0.75 grams per square inch. Application for flat surfaces typically runs 0.50 grams per square inch. There are slight differences in the recipes if you are firing vertical or horizontal pieces. Lower zinc levels slightly on horizontal surfaces because none of the glaze runs off. Increase zinc additions on vertical surfaces because 20-30% of the glaze runs off into catchers.


Enjoy the journey. 


Nerd (TJA )

Tom.... this will be the last technical post I make for awhile. my head hurts thinking about this stuff:)


Edit 8-12-2017 for spacing and addition of glaze application rate information.

Edit: 8-13-2017 opening comment on silica corrected, original was a bit confusing.

Note: additional links to resources may be added to these posts in the future. They will appear as "edits".

Edited by glazenerd
Link to post
Share on other sites

ZNO levels

Tile 1 : 25% zinc    Tile 2: 26% zinc   Tile 3: 27% zinc
Zinc is used to control population and crystal size.

Baby Blue

The black dots in the center of these crystals *nucleation site: is titanium dioxide, used as a seeding agent.

crystal 01


Adjust ramp hold temperatures until "floret: crystals form.

The inner crystal was made by holding at the first higher ramp hold for two hours, the smaller outer ring was made by lower the ramp temperature by 100 degrees and held for 20 minutes. Crystal growth rings are controlled by the length of the ramp hold, and by raising and lowering ramp hold temperatures. If you are firing in larger kilns (above 4 CF), then do a fast drop from peak to 50 degrees below targeted first ramp hold: then program a 500F an hour climb back to the actual first ramp hold temperature. By doing this, you will even out the heat in the kiln, and produce a more uniform product.




Edit 8-12-2017: the test tiles in the top picture were fired to cone 8. The blue floret picture in the bottom picture was fired to cone 6.

Edit 8-12-2017: additional firing information added.

Edited by glazenerd
Link to post
Share on other sites



Fire a test bar of the porcelain you plan to use and compare it to this porcelain chart. It should be bright white, like the round test tile on top. If it has a grey cast as shown, then you have higher levels of magnesium in the clay; and I would advise finding a different body. If it has a brown to dark brown cast, then it has high iron; and I would advise a different body. If it has a light tan to almond appearance; then it has higher levels of titanium which is acceptable for use.



Link to post
Share on other sites

Understanding Ramp Holds (growing temperatures)


Assuming you have read through the introduction, it is time to insert more specifics about firing schedules. You will learn over time to read crystal formations, and how they form will tell you how close or far you are away from achieving that perfectly round floret crystal. As DW alluded to in another post: there is an upper and lower ramp hold temperature that will produce these crystals. The higher end ramp hold will produce larger, coarser formations, and the lower will produce smaller, finer threaded crystals. The thread was started for those who have just entered into, or who want to try firing this specialty glaze.This information is not found in books, and us crystalliers usually guard it rather fiercely.


Most all crystalline books give a splattering of recipes, a firing cycle, and pictures of the results. It is standard crystalline practice to post dazzling pictures, then really not fully explain how to get there. The firing cycle is not that complex once you understand how to use it. The problem is finding the exact growing temperature for YOUR kiln. The book might say 2 hour hold at 1975F, but that could translate to 1950 to 2000F in your kiln. In addition, the type of zinc, type of silica used, and the type of clay it was fired on all effect final hold temperatures. There is a way to cut a year of testing out of your life trying to find that exact temperature for YOUR kiln.


Cry. 1&2


After your first firing, you pull a test tile out that looks something like this. The beginnings of crystals, some smaller or bigger than others: which is common for crystalline glaze. These crystal formation are talking to you rather you realize that or not. Formation 1 is about 25-30 degrees off from being a perfectly round floret crystal, and formation 2 is about 20-25 degrees off from that same goal. Crystal threading (as shown above) are an indication that your formula is okay, it is just a matter of finding the right temperature. These formations are telling you to lower the temperature 20-25 degrees and fire again.


As mentioned above, crystals will form in a lower and upper ramp hold cycle. When first starting out, aim for the lower ramp hold temperature first. One, it is easier to find, and two it cost less money to experiment in this range. So after adjusting your ramp hold downward in an effort to find those floret crystals:


12 X 16 arch2

Now you have a mix of maltese formations (also called axe heads), nearly round florets, and a few random crystals here and there ( normal). You are now within 3-5 degrees of forming the perfect crystal Once you dial in the final adjustment, you will have a crystal like this:

crystal 01

This crystal was formed at 1938F for two hours in MY kiln, and the temperature may be slightly different in YOUR kiln.
Nerds' Rule of Sixty   ( I hear it is pottery custom to name new things after yourself.)
Once you have your lower ramp hold temperature dialed in: simply add 60F to that temperature and fire again to grow those large crystals that everyone knows and loves. This is the reason I teach potters to establish their low end growing temperatures first, because after that it becomes a simple matter of adding 60 degrees (+/- 5-8 degrees) Adding sixty degrees produces this:

910 Test

Roughly a four inch crystal with several ramp holds programmed in. The other reason you need to know your low and high end growth temperatures. The large rings were produced by the high temperature hold, and the smaller rings were produced by the lower end ramp hold. If you want to put a feathered cap as the final outside ring, again just drop 60 degrees below your lower ramp hold temperature and hold for 20 minutes.


From Bisq:


Your typical firing schedule to 2050F

130F to 2225-2230F ( for white zinc) or 130F to 2240-2245F for yellow zinc.

<adjustments to peak can be made pending Lithium additions>

From Peak: full drop to either upper or lower hold ramp for 2 hours. 


* if you are firing in a kiln over 4CF, it is advisable to drop from peak to 50 degrees under target ramp hold, then add an additional segment of 500+ to target ramp hold. This evens out the temperature in the kiln and you will produce more uniform results.


Drop 60F from low side ramp hold for 20 minutes an add an outer finish ring.


Want growth rings:


From peak- full drop to first high ramp hold for 2-3 hours.

Full drop to lower ramp hold for 30 minutes

500F to high ramp hold temp for 30 minutes

Full drop to lower ramp hold for 30 minutes.

Full drop to 60 under low ramp hold for 20 minutes.

Kiln off.


You mix your ramp cycles anyway that pleases your artistic expressions.




Edit: this post originally marked: saved for future use.



Edited by glazenerd
Link to post
Share on other sites

Mixing and Application of glaze.

The basic recipe of 50% frit, 25% zinc, and 25% silica produces a glaze that: a. Settles out of suspension quickly. B. Rapidly dehydrates upon application. C. Tends to crack and powder when dried. Given these inherent properties: adding a glaze suspender is necessary. The most common addition is bentonite at 2-4% of total recipe batch weight. Some potters prefer using CMC gum, while others make ball clay additions up to 10% of total batch weight. There is no hard set rule for which product makes for better suspension: availability and preference usually predict selection. Regardless of product: sieving wet glaze is crucial to remove larger particles of frit primarily.

Adding a suspender to control settling of ingredients is one point: but these additions are used to control rapid drying after application as well. You will learn after your first test application just how quickly bisq pieces pull water from this glaze. Many potters wet sponge or dunk their bisq pieces to help slow drying and make application smoother. CMC gum is often used because it extends drying time,  in addition to acting as a brushing medium. Brushing or spraying are the only methods of application due to the required consistency of this glaze.

The actual amount of water (ratio to glaze powder) has never been discussed in the handful of books available on this specific glaze. Most reference water additions to create a " heavy cream" consistency. If brushed this glaze needs to be much thicker and applied heavier than typical glazes. If sprayed; then thinned down to pass through the gun: but additional coats applied to mimic that if brushing. You can brush or spray successive coats without waiting for the prior coat to throughly dry. Typically the coat applied at the foot of the piece will be nearly dried by the time you reach the lip of the piece. Otherwise just wait a minute or so; then apply next coat. You will find drying cracks and a powdery texture when dried: both are normal. Do not attempted to fix drying cracks unless actual flakes of glaze fall off: if so simply touch up that area with fresh glaze.

After you have finalized your recipe and selected a suspender: mix all ingredients dry including colorants. ( mix CMC in water separately to hydrate: do not add dry CMC powder to dry glaze). Start water additions at 60% of dry glaze weight. EX: 100 grams of dry glaze times 60% water = 60 grams of water. This should produce a heavy paste. Is so, then add 10% more water and check consistency: repeat if necessary or reduce to 5% additions until a heavy cream consistency is achieved. The type and amount of suspender creates variances from potter to potter. Keep a record of the amount of water added on your first test mix: then use that measured % thereafter. Glaze can be thicker if applied to green ware because of low absorption. Water content needs to be higher on bisq because of high water absorption. 

Application: for vertical pieces- 0.65 to 0.75 grams per square inch of surface. For flat pieces such as tile or low profile plates with large foot ring: 0.40'to 0.50 grams per square inch. As a practical,point of reference: you have to apply according to the high degree of glaze run at peak temperature. That said: the lower 1/3 of a vertical piece receives roughly 25% of the glaze ( 2-3 coats). The middle 1/2 to 2/3 of piece receives 40% of the glaze (3-4 coats) and the top 1/3 receives 35% of the glaze (4-5 coats). The glaze will flow from the top to the bottom: so application rates follow this known parameter. It will take some time to learn application and the actual number of coats will vary depending on glaze consistency and your own personal application technique. Some potters slather on one coat, some apply thinly: just adjust coats to your style.

note: attach piece to catcher before glazing. Handle actual piece as little as possible once it is glazed. Remember: glaze applied to green ware can be mixed with less water. Glaze applied to bisq needs more water because of absorption. Wet sponging or dunking bisq before applying this glaze is helpful.



Edited by glazenerd
Information added to reserved post.
Link to post
Share on other sites


The other technical posts that I have done had  this same format, it does make it easier to compact all of the information. I do not plan on publishing any books on crystalline glazes, so I do not have to worry about pulling it down because a publisher gets pithy about it.


I will have to edit it some more, my grammar stinketh!! The thoughts are crystal clear in my head, but somehow gets muddled when I try to put them on paper.

Edited by glazenerd
Link to post
Share on other sites

Okay, since I’ve got a few minutes, Tom, here goes. Here’s my addendum.


I feel the need to say that all glazes can, in theory, be crystalline—it’s just a function of cooling slowly enough. Calcium matte glazes are crystalline, as are some incarnations of tenmoku, kiseto, and woodfired ware can be as well, if properly downfired (I’m thinking of Tony Clennell’s MFA woodfire work).


As Tom has stated, zinc silicate is the crystal at work, and it is indeed an ionic bond. Like salt. And I’m going to use salt and water throughout this addendum as a flawed analogy to explain what’s going on. Remember when you were like 8 and grew a sugar or a salt crystal on a string in a cup? That’s exactly what’s happening with a macro glaze. You’ve got a solvent (the frit), a solute (the zinc silicate), and then you have the seed crystal to get things going.


The frit functions like the water in my analogy. When the crystals are growing, the glaze is extremely fluid and it runs like a fugitive. And in the same way you can’t really grow salt crystals in ice, the glaze has to be completely fluid.


One thing Tom failed to mention was the nature of the frits themselves. Head over to Tony Hansen’s digitalfire and have a look at Ferro frit 3110, Fusion frit 644, and Fusion frit 413. These are all used for macrocrystals. Note that there is a high alkali content in the form of soda, there’s a high silica content, and extremely low alumina, and hardly anything else. The fusion frits have no boron or lime, just soda, silica, and just enough alumina. They’re high expansion, low viscosity glazes.


Doesn’t really matter how you get there, but you’ve got to have a base something like that.


Regarding zinc oxide, more important that the source of the zinc (my pottery supplier just sells one kind—precalcined), is whether your zinc oxide is calcined or not. Tony Hansen has a good discussion on this and the pros and cons of both forms. The TL;DR Some don’t calcine, some do, it has big connotations in how the glaze goes on and fires. Store in an airtight container, though, or time and moisture will negate your work if you do calcine.


Regarding Tom’s comments on lithium carbonate. Firstly, lithium, silica, and zinc oxide don’t all have hexagonal crystal structures. The vase majority of silicaTE crystals have tetrahedral structures. Zinc oxide DOES have a hexagonal structure, sortof, but it would be misleading to state that. Lithium carb is monoclinc.


Lithium is also not the primary flux. Sodium is, from the frit. Lithium is an auxiliary flux used to improve the surface quality, increase viscosity, while not contributing to the already high glaze COE.

I must also correct Tom’s statement that lithium has a much lower charge than zinc or silicate. It has a much lower e l e c t r o n e g a t i v i t y, but it has an extremely strong positive charge, which is why it is an extremely powerful flux. This is why, like hydrogen above and sodium below, its ion is written Li +1. Also, it absolutely interferes with the silica lattice, that’s why it is a flux. If it didn’t, it would be inert and would contribute nothing to the melt.


I should also add that titania isn’t the only possible seeding agent. Back in the day, people used to plant (if you’ll pardon the metaphor) hemimorphite or willemite seeds (both hemimorphite and willemite are forms of zinc silicate) in the glaze to achieve the same effect, albeit less reliably.


So how does it actually work? Like the sugar or salt crystal from when you were 8. Sugar and salt have a greater solubility in water when the water is hot, than when cool. So you parent or teacher got some water hot, and you poured in your crystal starter and stirred and stirred until no more would dissolve. You’ve made a saturated solution. As the water cools, hat solution becomes super saturated and the excess sugar, salt, or whatever precipitates out. To get your super cool rock crystal, you needed a seed crystal to work as the foundation. And then you needed time. The more saturated you could get the solution, the more crystals


It’s just the same with a macrocrystalline glaze. You dissolve your silica and zinc into the glassy matrix of the frit, cool it such that it becomes super saturated (while still fluid), and the crystals have time to grow.


I hope this helps the discussion going forward.


Edited to corrwct typo.

Edited by Tyler Miller
Link to post
Share on other sites
Guest JBaymore

Back in the day, people used to plant (if you’ll pardon the metaphor) hemimorphite or willemite seeds (both hemimorphite and willemite are forms of zinc silicate) in the glaze to achieve the same effect, albeit less reliably.


Years and years ago.... in a galaxy far, far away, I worked at Mass College of Art with colleague Charles Abbott.  Charles was quite famous in his day (studied at Cranbrook under Maija Grotell and then at Alfred in the 40's), and he was a "glaze guru" guy.  He was known for his macro-crystalline glazes back when you controlled the cooling cycle by babysitting the kiln and turning the banks of elements on and off over and over to hold the temps. (Yeah..... crazy stuff.)   He utilized just this technique that Tyler mentions to seed many of his crystalline works.





Link to post
Share on other sites

Back in the day (whenever that was...) you could get granular zincite with which to seed macrocrystalline glazes. It was scraped off the insides of the chimneys of zinc smelting factories in communist Poland. Since the collapse of communism, the Polish factories either closed or were modernized so there is no more chimney scraping.

Link to post
Share on other sites

For the record, Tom intentionally left out the complicated kinetic chemistry involved in crystalline glaze, and as stated in my opening comment: "introduction".


Tyler> I am not sure what to make of your post, other than it signals to me that you have never fired crystalline glaze before. All the chemistry you alluded to is misapplied, because the terminology is all related to A-phase temperatures. The silica atom is indeed a tetrahedral chain, but the molecule is trigonal in its A-phase structure, and hexagonal in its B-phase. Ionic bonding occurs when silica is in its B- phase (above 573C).


I have yet to see a crystalline glaze recipe with sodium or potassium additions. I assume you comment about sodium being the primary flux is related to the basic chemistry of frit, and in particular the sodium content. However, frit is fired much higher than we melt glaze, and the glass is rapid cooled and ground. Your assertion that sodium is the primary flux because frit has high sodium dismissed that the sodium atom was thermally disassociated at around 2190F (as confirmed by the U of I study) and only sodium cations remain. The fluxing action of sodium has already occurred, and the sodium molecule has been completely broken down, leaving only cations.


To further explain, Paulings scale of electronegativity would apply of the ionization energies of sodium were still present. However the valence bond of the sodium atom has been thermally disassociated. Once you break the valence bond by exceeding the melting temperature of any given element, (thermal disassociation) you then deal with band gaps 


NERD;s Band Gap Chart

SiO2       0.033            Silica

Li2O       0.024            Lithium

Na2O      1.790           Sodium

K2O        0.830           Potassium

CaO        0.469            Calcium

TiO2       0.184            Titanium

Al2O3     0.054           Aluminum

ZnO        0.283            Zinc

Mo          2.160            Molybdenum

You are correct however that sodium and potassium both form ionic bonds, in lieu of the covelant bonds most often associated with common glazes.  As you can see by this band gap chart prepared by UC Berkely (2200F) lithium is below the band gap of silica: and both sodium and potassium are above it. Which is why sodium and potassium should be avoided as glaze additions because their cubic crystal structure will pull silica away from zinc, and form sodium/potassium crystals that appear as spikes and clusters.


There has been a decades old debate among the crystalline community in regards to calcining or not calcining zinc. Very few who fire crystalline on a regular basis calcine their zinc: and those who do swear it makes a difference. I have experimented with both, and will say at cone 6 it does seem to make a difference, and at cone 8 and above: none. Actually Tony Hansen does not fire crystalline glaze but relies on the information provided by Fara Shimbo on his site. Fara was a firm believer in calcining zinc, but she also specialized in cone 6 firings.


I know of a half a dozen seeding agents that can be effectually used, but again this was an introduction: not the final examine.


And no it does not act like sugar in water forming crystals. Ionic bonds are complex electro=chemical molecular bonds. If it was that easy, the glaze makers would be selling it by the gallon, and every potter would fire them. Zinc, like the crystals they form are HCP (hexagonal closed pack) crystalline structures. Not sure why you affirmed that zinc was not hexagonal.


350px-Willemite.pngHexagonal closed packed crystalline structures. Zinc is a HCP crystal, silica is hexagonal, and together they for the ionic bond of HCP crystals.



Applying the chemistry and corresponding analysis of elements at room temperature, or below 573C, before the phase change of A quartz to B quartz is not applicable. Applying the fluxing abilities of sodium or potassium that has been thermally disassociated is also not applicable.




Sorry to those who are just trying to learn about crystalline glazes, I tried to keep it simple for you.

Link to post
Share on other sites



The comment quoted below suggests this is about more than facts for you. I'm not willing to get into a pissing competition about this stuff. I offered up my suggestions and you come back with an ad hominem that I've not fired crystalline glazes before. That's not a game I'm willing to play. I'll discuss facts and science with you, but I'm not going to play ego games. Lots of people fire crystalline glazes without knowing the chem and glaze companies indeed sell ready mixed crystalline glaze by the gallon.


Re: sodium. It is irrelevant what form the sodium is in. Sodium in feldspar is tied up like a frit. Indeed, all usable sodium is tied up that way in ceramics, whether thermally dissociated in the ground under geologic pressure or fritted in a factory in Mexico. Soluble forms aren't usable due to their tendency to effloresce. If you removed the soda from the frit, you'd have a non-fluid glaze at temp. It is the causa sine qua non est. That is, without it, no melt, no glaze, no crystals.


If you could put together a glaze from raw materials that could supply the numbers the frit and additives do, you'd have a macrocryatalline glaze just the same. They fire frits higher in the factory because it takea forever for a vat of the stuff to off gas and fully incorporate. It's the same reason Roman and medieval glass batch was fired for days before use. The fusion takes forever in comparison to a thin cross section of glaze on a pot.


Band gaps aren't relevant unless you're planning to use your pot as a semiconductor. If you are, that's your prerogative, but you don't need a data on the energy required to make an electron conductive to make a glaze--not even a crystalline one.



Re quartz and zinc structures. [edit: explaining silica in terms of tetrahedra allows explanations of more of its forms, crystalline and not. It's more didactically useful, bridging the behaviour of quartz and phyllosilicates in clay. /edit] In the case of beta quartz, the struxture class is 6, 2, 2 trapezohedral, which is a hexagonal system but describing it as an hexagon crystal lattice is opaque and unhelpful. As is describing zinc oxide as hexagonal without qualification--it's a wurtzite crystal structure which explains more about its structure and properties than hexagonal or hexagonal close packed. But again, this info is also unnecessary unless you want to exploit some wildly exotic properties. Piezoelectric ceramics anyone?




It's probably important to point out that your zinc silicate there is willemite. One crystalline structure, but there are three, including a phyllosilicate--i.e. a clay like crystal. Willemite's structure doesn't explain how things melt. It explains crystal formation on the way down, but in the melt, it's moot--everything's amorphous.


Regarding sugar and salt in water. I'm not sure if you're aware, but salt dissociates into its ions due to the polarity of water. It is pretty close to the same mechanism at work in a crystalline glaze, but at different temperatures. Those same complex electrochemical bonds, ionic bonds, are what hold salt together and break apart in water. Acid-Base chemistry at its simplest relies on the same principles.


When a glaze is at temp and fully liquid, those ions are active and reactive, just like ions dissolved in water. It all comes down to time, temperature, and the associated transformations that occur. Just like growing a salt crystal on a string.


This isn't high level stuff, and making it out to be will only drive people off trying it. 50% frit of some kind mentioned above (though fusion f644 is garbage), 25% silica, 25% zinc oxide, colourants and a well calibrated kiln with a good downfiring program are all you need. Anything else is bluster. The real reason people don't fire crystalline more? They're expensive and time consuming to fire, messy to clean up, and not hugely functional.


Now, let's talk about making macrocrystalline glazes easy for people while providing accurate, helpful information. Okay?


Edited to complete a sentence. Edited a second time, as marked, to better explain a thought.


Tyler> I am not sure what to make of your post, other than it signals to me that you have never fired crystalline glaze before.

Edited by Tyler Miller
Link to post
Share on other sites

Flux Test

This is a potassium/sodium test I did last April. The spike and cluster crystal are formed because both elements have higher band gaps than lithium; causing the silica to form an ionic bond with them, instead of the zinc. Every potter who fires crystalline regularly has experimented with different additions in search of the magic bullet that grows mega-crystals. Just as every potter who fires crystalline glaze has learned; absolutely do not add sodium or potassium to the glaze recipe. So Tyler; when you started talking about sodium, I sorta knew.

910 Test


The sodium in clay even effects crystal development: even though a nice floret crystal formed: the threads are open and coarse.


Refractive Test

Been when you custom mix your own clay, keeping the sodium under 1% molar: the crystals develop their true (HCP) hexagonal closed pack shape. The threading is very dense and compact.
*John: by the way, baby blue was the one Ron Roy loved so much at NCECA 2016 in KC.
It was interesting that semiconductors was used in response to band gaps. Actually, that is where I learned the true nature of ionic bonding; in reading technical uses of zinc silicate. Every mother board and micro processing chip are coated with zinc-silicate because they neither give off or conduct electrical charges. So in apply that to crystals, it became clear that sodium and potassium had to go because both have high electrostatic charges. I could not figure out why semi conductors were even mentioned until I googled "band gap", then it became clear why.


Link to post
Share on other sites

Um, band gaps don't expain the spikes. I've never heard of them used outside of discussing conductivity. Ions are ions. I think that's where the confusion is coming from. Reactions happen when ions of greater or lower electro negativity are introduced. These concepts took forever to sink in.


In science, what you've done is failed to attempt to disprove your hypothesis. Sodium has a larger band gap because it is less electrically condictive than lithium, but lithium is less electronegative, having a stronger positive charge. All the time. Reading up on the band gap of gold will be instructive. Non-reactive metal, highly conductive. Band gaps explain eletroconductivity, not electronegativity. Even when thermally dissociated, ions are ions. Edit: Put another way, solid state physics doesn't have much to do with liquid state chemistry in this case.


My theory about your spikes: Zinc oxide is a flux. The frit only has so much sodium. The more silica you add without adding anythking else, the higher the melt temp. Zinc Oxide then acts as a flux, like it did/does in the old Bristol glazes. But too much causes a saturation and if the glaze cools slowly, a crystalline form precipitates out. If you add more sodium the added silica and zinc becomes locked up with that instead, reducing the amount of the zinc silicate to form its crystals and inhibiting their growth. You're altering how much salt you can dissolve into the water in your cup before it becomes super saturated.


To get back to my original point regarding your initial comments re: lithium. This doesn't make the lithium any less powerful a flux, and its small band gap just means its a better electrical conductor than sodium--as it ahould be, it's the quintessential metal. Despite being an R2Oit demonstrates some properties of an RO flux. That's where I think the answer lies. Not with band gaps. Electronegativity is the principle at work here.


But no one in the crystalline community discusses this stuff this way. I lurk the crystalline forums (like I do most ceramica forums--cone 6 electric, wood fire, etc) the information is useful. Reduction effects and alternative crystalline glazes would be cool to learn more about. What do you know about molybdenum crystals, Tom?


Editted to delete a redundant sentence and expand another. Second edit to correct typo and gloss a term.

Edited by Tyler Miller
Link to post
Share on other sites


I am not about to disprove my theory, took me ten years to come up with the one I have.


I do like the whole concept of "conductivity". I remember many years back reading up on that topic, but every abstract I read always went back to electro negativity: but I think your terminology is more applicable. You use salt and water, and it always used magnets; atomic polarity is no different than currents in the world of big objects. (Ours) so in application, zinc is the big magnet in this glaze that needs to bond with the smallest magnet- silica. Sodium and potassium have the energy to attract the silica if those levels are too high. Lithium has less magnetic pull than silica, making it the safe flux. So yes, I do like your terminology of conductivity; it is a concept that most potters can grasp.


Saturation is not a concept that I have given much thought to, but I will. In relation to zinc and silica, it does have some basis. It is after all, the adjusting of zinc and silica that potters use to control(laughs) crystal development.


Thank you for the link, read it many years ago but lost it when my laptop took a dirt nap.



Edit: correcting the auto corrections of this ipad..

Edited by glazenerd
Link to post
Share on other sites



Your feelings about your theory remind me of something Socrates said in Plato's dialogue "Theaetetus." Socrates explains to his interlocutors that he is a midwife of ideas, facilitating their birth, and evaluating whether they would thrive. Advising them whether or not to keep the idea or discard it--an emotionally difficult thing for the idea's "mother." Now, I'm no midwife of ideas, but I do advise you to reconsider theoretical attachment, since it is one of the mechanisms from which bad science arises. For every major breakthrough study there has been, there are (or should be) hundreds of copies attempting to disprove the breakthrough. Theae don't get press, and the perception of science is skewed as a result. There are few, if any, lone geniuses toiling away at the next major breakthrough. Einstein is fetishized as such a genius, but he almost always worked with colleagues, and he spent a lot of his career actively trying to disprove his findings. "God doesn't play dice with the universe" was a reflection of his hatred of quantum mechanics--a field he inarguably helped found. My point is that you shouldn't get too attached and avoid confirmation bias--a true nerd shows commitment to truth and learning (though I'm not fond of the term, I was beaten up as a teen while that word was applied to me).


Re: conductivity. I was arguing for electronegativity, conductivity isn't relevant. Band gaps, are, again, the amount of energy it takes to move an electron to the conductive band. Metals either have no band gap (like gold), or a very small one. Insulators have very large band gap values, which is why you can't run a current through an insulator. Semiconductors have medium sized band gaps and this is how they're useful in electronic components like transistors and diodes. They work like solid state switches. This has little to do with ionization and chemical reactions. Your copper wire doesn't ionize when running a toaster, or all manner of anarchy would ensue. Might I suggest you're confusing ionization energy with band gaps? The former being how much energy it takes to remove an electron?


The magnets analogy is good to explain electronegativity, keep using it.


To everyone else, the non-sciencey version of the article I posted explains how the round crystal shape of a macro glaze is formed. They explain it shows up everywhere, plastic bags, obsidian, etc. The theory they state (which isn't 100% accepted) is that it comes about through a tension between different phases in the melt. This causes very specific branching behaviours irrespective of crystalline structure. Why the same patterns show up sometimes in wal-mart bags. In the crystalline glaze this is a tension between soda fluxed glass and zinc oxide fluxed glass.


I used the term saturation, because that's how it plays out in our experience but it's not quite saturation. High boron high fired glazes will separate out as well, only, nothing pretty results.


This article demonstrates how a glass system of fluxed with soda, borate, and zinc oxide (which will readily both phase separate and devitrify) can be exploited to create opal glass. It also explores additives (among which is lithia--Oops! wrong article! no lithia in this one) and how they modify this effect.






Would you be willing to post a typical downfire schedule and explain they rationale behind each step? I'm also curious about what research you'd done with molybdenum crystals and how they differ from conventional macrocrystalline glazes. Moly's capacity for expressive crystalline glazes is remarkable.


Edited to correct typos and an error.

Edited by Tyler Miller
Link to post
Share on other sites

Let me pop in for a moment with an observation from my admittedly more limited experience than Tom's, while staying out of the chemistry food fight (which I know only enough of to be wrong...). You asked about a typical downfiring schedule. There is no such thing that I know of. The crystalline growing window is between approximately 1800F and 2000F. One must hold in that window for a number of hours while the crystals grow (assuming any seeded in the first place). It is my observation that the crystals grow faster and with a more spikey texture in the higher end of the growing range and slower with a softer more flowery texture toward the lower end of the range. One can raise and lower the temperature for intervals within the overall holding period and cause the crystals to take on variations in the texture, including bulls-eye targets caused by repeated up and down tweaks - always staying within the growing window.

Link to post
Share on other sites



At the end of my original post I stated: this would be my last technical post for awhile. ( I was not kidding.) been way too absorbed in the science and it is time to get out of this rabbit hole. Eyed a very large oak tree growing next to the lake up the road, I heard it calling my name.

So no Tyler, I am done posting about chemistry at this point. We will have to agree to disagree. Although I did find some of your points intriguing.



Link to post
Share on other sites

Dick White,


What I was hoping for was a non-technical educational example that show the decision making process. I realize that there's no cookie cutter firing program, there's nothing hard and fast about ceramics. But it can help to see the decision making process involved in firing and how that informs artistic/studio practice.


In glass textbooks, there are usually firing schedules included, despite the fact that firing schedules are manufacturer and batch specific. It just helps open up an otherwise opaque world of technique.


I apologize, Tom. I feel I've been too heavy handed. I genuinely was trying to help, add to, and sometimes correct your understanding of this stuff. Your posts are always welcome. If you like, I'll back off and give you a wider berth and more interpretive charity.



Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Create New...

Important Information

By using this site, you agree to our Terms of Use.