What is Glass?
Making Batch

Rev. 2002-12-15, 2004-01-06, 2005-02-10, -12-19, 2007-02-10, -04-30,
2008-01-15, 2009-05-04, 2010-12-08, -12-20, 2011-01-23

Back to Sitemap

Local Links



Melting Point




Colored Batch


Remote Links
Why is glass transparent?
Flowers Out of Glass
A Long, long article on the Glass Flowers at Harvard and their problems
with considerable discussion of glass chemistry at an easy level
Statistical Calculation and Development of Glass Properties
A large website of glass property calculations


Glass is sand or other sources of Silicon Dioxide, like flint, with other ingredients added to make it melt easier. The result is an amorphous solid - non-crystalline - that melts over a range of temperatures.  The fewer ingredients added, the higher the temperature it melts at and the less it expands and contracts. Almost pure sand would result in quartz, which melts at over 3000F (1650C), next down is borosilicate glass (Pyrex), lower down from there are soda/lime glasses, which melt for working at 2000-2100F (1090-1150C).  Glazes for clay and enamels for glass and metal are forms of glass and can melt as low as 1150F  Sodium Silicate (Water Glass) dissolves in water.  Handling glass batch chemicals requires safety practices including cleanup, keeping chemicals safely, and using a respirator.
In addition to lowering the melting point, chemicals being added can change the color and widen or narrow the melting range.  Glass batch mixed for making bottles or window glass mechanically has a narrow (or short) temperature range so it goes from being molten to being able to stand without sagging over a couple of hundred Fahrenheit degrees while glass formulated for manual blowing has a wide (or long) range of up to 400F degrees between thick fluid and too firm to work any more. 2010-12-20

An alternative to melting batch is to remelt broken glass, referred to as cullet.  Often, some cullet is added to a batch melt because it assists in melting the batch and saves money.  An all-cullet melt requires buying cullet in bulk which is done from firms mostly in the mid-Atlantic states where pressed and blown glass objects produce a reliable glass scrap. For a long time the primary supplier of cullet from this source was Gabbert Cullet.   Recently certain sources, especially on the west coast are melting specifically for sale as cullet (no objects made before), some of which comes from China.  Cullet melt can be done with broken window or bottle glass collected locally, but this glass sets quickly and is more awkward to work on the pipe.  Cullet as a product costs more than batch, especially considering shipping, but savings occur in energy use since cullet does not have to be taken to as high a temperature to melt and fine and the corrosive effects of some batch ingredients are absent.  Thus, a furnace does not have to be built to as high a standard and fuel piping and maximum usage are reduced.  Cullet also permits a moderate design for electric melt since corrosive gases can effect elements when melting batch and the higher temps are nearer the failure limit of elements. 2005-12-19

Melting Point - Glass is an amorphous solid which means that it is without a crystalline structure.
Crystalline structures have a specific melting point, amorphous do not.
Different formulas of glass have different ranges of softening/melting and, in fact, by manipulating the chemical contents one can make a glass that goes from molten to stiff over a narrow temperature range, which is desirable for mechanically making bottles, or over a wide range, which is good for manually blowing glass while making lots of little changes. In glass working the former is called short and the later long glass.
Soda-Lime glass formulated for glassblowing from a furnace will start to sag and sink into other pieces of itself at about 1300F and the pieces will melt flat with each other in a kiln at about 1500F, but it will be hard to drag (a design method) in the middle of the range. The glass will be workable on the pipe but resistive at about 1700F and be like thick honey in gathering from the pot at about 2000F. A glass blower can adjust the viscosity from thick to thin for gathering purposes by setting the furnace temperature from 2000 to 2100 or 2150 when it is getting close to trying to gather room temperature olive oil.
Lead crystal glass is very long, staying soft enough to work while cooling.  When I took a class at Corning, the tech assistant was given the chance to work with Steuben crystal in the shop, he reported that is was soft for so long that he had trouble getting used to it.
Pyrex, a borosilicate glass, needs to be much hotter to flow easily.  Torch workers (lampworking) work it with oxy-propane and can take it and out of the flame to control its stiffness.  Almost no one hand works boro from a furnace where it requires protective gear and is so near the limits of the materials that the furnace does not last very long. 2010-12-20
FORMULAS - Further down this page is a formula for making a glass that is widely sold for art glass blowing and which melts to an easy to work, very clear glass. Ten ingredients resulting in 12 testable chemicals are listed, but the first four ingredients make up 94.4% of the product (by weight.)  Some of the other ingredients are added to make other ingredients behave, ease melting, or to balance color. Below the formula are instructions for melting the batch. This link [BROKEN] NOTES ON GLASS (John M. Rusin) has an excellent discussion of various aspects of glass.
At right is the basic soda lime recipe used in American Glass Practice for discussion and further development.  Except for the sand, each of the other ingredients can be changed and others added with varying effects on the glass. The melted glass totals 1954 lbs., a shrinkage of 311 lbs.,  most of which loss as gas in the reaction to make glass, such as carbonate giving off CO2. Sand
Soda Ash
Lime (burnt)
1500 lbs.
550 lbs.
150 lbs.
65 lbs.

Batch is expected to be cooked at a specific temperature or temperature range. Different ingredients melt at different temps and cooking at too low a temp can cause some ingredients to melt out before they can do their job of fluxing to make the other ingredients melt more easily. Too high a temp can burn off ingredients. A parallel would cooking bread in too hot an oven so the outside is brown or black while the inside is soggy or cooking cake without preheating, so as the oven comes up to temp the ingredients that cause rising make oversized bubbles.

In choosing the sand, the maker of the glass looks for sand pits that can provide a blindingly white sand. That is because of the very last chemical in the right column below: Fe2O3. The ordinary sand you find on the beach or sand box is a shade of brown and it is so because of the rusted iron in it. Iron oxide in glass gives it a green color. In fact, the easiest color to get in glass (and the hardest to get rid of) is green. Glass melted in the early colonial days is predominantly green (shaded to blue as a second choice) because there was iron in the sand. Window glass, viewed edge on, is green for the same reason (and because the chemicals to reduce the color are costly.)  In fact, one of the most popular glass objects, the green glass Coca Cola bottle, is that color because it is easy to make. (Not because a green slag volcanic glass is added to it as gullible travelers to the southwest are sometimes told.)

Soda ash or potash are added to lower the melting point of the glass mix, while limestone brings durability and the feldspar is a flux to make the melting process go forward more easily. In the history of glass, various ingredients have substituted for the soda ash. The first source for alkali ash was burned plants rich in the right chemical. Another source is chemicals remaining from boiling (or evaporating) sea water.

One way to make a (small) batch of glass would be to use white sand from a hardware store of the kind that is used in ashtrays in hotels, substitute whiting (calcium carbonate) used in plaster formulas for the limestone, and ... for soda ash buy it from a ceramics supply firm (which can also do calcium carbonate.)

Here are some old recipes, from Glass Gaffers of New Jersey, pp 101-102

Window Glass
100 lb Sand
33 lb Lime
30 lb Soda
10 lb Salt

Carboys Green
(Large bottles)
100 lb Sand
35 lb Lime
37 lb Soda
20 lb Marl
15 Salt

Vials & bottles
100 lb Sand
35 lb Lime
35 lb Soda
15 lb Salt

"Flint Glass" (Clear)
160 lb. Sand
28 lb. Lime
200 lb. Perl Ashes
16 lb. Salt
5 lb. Arsenic
8 oz. Manganese
50 lb. Cullet

And one table from NOTES ON GLASS (John M. Rusin) (see link for analysis)

Common Name

Chemical Name

Chemical Formula

Glass Component

Wt % of Ingredient


Silica or Silicon Dioxide




Soda Ash

Sodium Carbonate





Calcium Carbonate




Note that discussing glass from batch is complicated by the fact that parts of ingredients will cook off. Note above that the difference between the ingredient formula and glass component in the carbonates is CO2 which passes off as a gas.

Common Names and Chemicals
Sand Silicon Dioxide SiO2  
Silica Silicon Dioxide SiO2  
Black Tin Stannous Oxide SnO  
White Tin Stannic Oxide SnO2 Opacifier
Limestone Calcium Carbonate CaCO3  
Whiting Calcium Carbonate CaCO3  
Soda Ash Sodium Carbonate Na2CO3  
Pearl Ash Potassium Carbonate K2CO3  
Pot Ash Potassium Carbonate K2CO3  
Flint Silica (Silicon Dioxide) SiO2  
Feldspar Alumina w/Soda or Potash K2 or Na2Al2O3:6SiO2 Flux
Fluorspar Calcium Fluoride CaF2 Flux

What makes glass? One of the problems/factors in making glass is the difference between what goes in and what is in the result.  Looking at the list of ingredients, it can be noted that many have CO3 as part of the formula.  However glass is made of oxides, so during the process of cooking of glass CO2 is given off.  Other compounds evolve other gases, in particular, fluorine in making opals. What this means is that the formula for making a glass batch is not the same as the chemical analysis of the glass that is used for determining coefficient of expansion.  Traditionally, batch formulas are given starting with 100 pounds of sand - not 100 pounds of mix or percentages.  To determine the final analysis, the pounds of  ingredients are reduced to pounds of effective material. There is a fixed percentage for each ingredient which can be determined from the chemical formula and molecular weights, but is also available from sources such as Henry Halem's Glass Notes, along with a further discussion of the process. 2006-06-29

The purpose in presenting the following material is to suggest what is involved in making glass from scratch and melting such a formula. Some details of the formula have been omitted, but other formulas will be added below from other sources. Most of the added ingredients are fluxes to make the glass melt at lower temps and clarifiers to remove color provided accidentally by small imperfections in the purity of ingredients (like iron in sand.) Please note that melting on a wrong schedule can cause fluxes to melt out early or ingredients to boil off.

SPRUCE PINE, N. C. 28777 PHONE 704-765-9876
(This material is from information provided by Spruce Pine Batch about their glass batch. Copyright Spruce Pine Batch (TM), www.sprucepinebatch.com )

We mix glass batch for the studio glass artist. All batch is agglomerated (pelletized) for easier, safer storage and use. All materials are selected and tested for use by studio artists- not large factories. Our agglomerated batch is prepared by weighing and mixing the materials which compose the formula of the glass, then processing in a rotating disk pelletizer which forms larger granules, balls and pellets. The primary purpose of agglomeration is the prevention of separation (unmixing) of the materials during shipment and storage. However, it also has the advantages of increasing safety and convenience by reducing the dusting of the materials, most of which are finer than 200 mesh. In addition, some researchers have found agglomeration results in energy savings and reduced furnace wear when compared with loose batch. We will custom mix your formula if it does not contain lead, price depending on material and quantity.
We stock a standard glass batch which was designed by Dominick Labino for use by studio glass workers. 

Labino (TM) Batch Formula  
Soda ash  
Lithium carbonate  
Sodium nitrate  
Zinc oxide  
Barium carbonate  
Antimony oxide  
(Exact quantities omitted  
at request of SPB  
Theoretical composition (%)  
SIO2 72.0
Na2O 14.9
K2 0.4
Li2O 1.0
CaO 7.9
MgO 0.15
Al2O3 1.8
BaO 0.5
Sb2O3 0.22
ZnO 1.0
F 0.1
Fe2O3 0.026

English and Turner coefficients
of linear thermal expansion 87.3 x 10-7 What does this mean?
Actual measurements 90+-1 x 10-7
Annealing point 477C (890F)
Softening point 662C (1224 F)
Melt glass around 2250 to 2400

Melting Spruce Pine Batch
(This information was provided by Spruce Pine Batch.)
SPRUCE PINE. N. C. 28777, PHONE 704-765-9876
Suggested Melting Schedule for Batch Pellets (Melting directions recently revised, see web site for current suggestions. www.sprucepinebatch.com )
This schedule is based on experience gained from melting in a 120 lb pot furnace which melts at a rate
of about 20 lb/hr. A larger furnace would melt at a faster rate.
Preheat your furnace to 2250F (1232C). Hopefully you have a thermocouple or other means of determining the temperature. Maintain 2250-2300F (1232-1260C) thru the melting stage. Melting at a higher temperature sometimes leaves scum.

Method 1: Charge 15 lb into the empty furnace (any kind). [heat the furnace] When the batch surface is flat, but not yet glassy (looks like cottage cheese), charge 25 lb. [and melt] That will take about 1 hour. Continue melting 25 lb charges the same way until the furnace is full. Each cycle will take a little over 1 hour.

Method 2: Fill an empty tank furnace or invested pot furnace with batch. This method should not be used for a free standing pot furnace. When the batch surface is flat, but not yet glassy, fill the furnace again. Since you have more batch to melt, this will take longer than in Method 1. Continue filling the furnace and melting the charges the same way until the furnace is full. There will be less time between charges as the charges get smaller.
Stirring the batch is not necessary for either method, but may accelerate the melting some. The stirring would be done about midway in each melting cycle.

Approximately 2-3 hours after the last charge (longer for charges larger than 25 lb), the temperature will rise to about 2400F (1315C). We do not think that you have to get over 2400F.
At this time, start taking small gathers and look at the bubbles. When the bubbles are all larger than pin heads, turn the temperature down to 50-100 (28-56C) below your working temperature to squeeze the bubbles out. This will take about 6-8 hours. However, the furnace can be left at this temperature for longer periods, overnight if you wish. If you have a larger furnace, you can turn it down sooner because the furnace will retain the higher temperature for a longer time. Some furnaces may squeeze OK if you turn it down to your working temperature.

When the glass is bubble free and at least 1 to 2 hours before working the glass, turn up the heat to your working temperature. A 2100F (1184C) working temperature is a good place to start. You may prefer to work the glass either hotter or colder.
Method 1: Use all of the glass in the furnace and repeat the process.
Method 2: Charge the furnace after each working session. Fill the furnace with batch and turn the temperature up to about 2400 F to melt and then just follow the fining instructions. Cords may tend to build up in the bottom of the furnace, it is recommended that you clean the furnace out at about weekly intervals, or at the longest, every 2 weeks.

[end of SPB copyright material]

Heating a batch on the wrong schedule - using the wrong temperatures and times at various temperatures can cause the glass to behave differently because the flux may melt first and actually run off the pile of batch before contributing to the melting of other materials.

There was a nice little example here of a batch being developed, but the guy who posted it to a discussion group decided after giving permission to use it to take away his permission, even though his name was given in the material.

This is a formula e-mailed for my use, this is a simple batch formula, with no added complications.

27# 2oz SILICA 325 [mesh]
8# 11 oz. DENSE SODA ASH
15 oz. BORAX (5 mol.)
2# 4 oz. POTASH


 "Janet" wrote on rec.crafts.glass
> I was wondering why all glassware is not made from 'crystal'.
> Surely it doesn't cost much to add a bit of Lead oxide to the glass mixture?
> Or even Barium (I think it's Barium they use instead of Lead because it's less toxic?)
> Even a lot of quality glasses that one sees in shops are not crystal but plain glass. Why?
> Thanks in advance.

 Lead oxide makes a glass "long" - workable over a wide temperature range.  When I took a class at Corning, the tech assistant was given the chance to work with Steuben crystal in the shop, he reported that is was soft for so long that he had trouble getting used to it.  Most glassware is made from "short" glass that goes from molten to stiff over a narrow temperature range because it is blown into a metal mold, the neck being burned off with a small torch flame (shown by a small bead on the lip at the last place to melt free.)   The desire is to get the piece blown, out of the mold and standing on its own for annealing.

From: "Jeff" <>
Subject: experiment
Date: Monday, January 14, 2002 2:02 PM

after my recent success is casting silver i'm in the mood to experiment with glass and i have a question for this group. i understand that glass is basically silica with impurities added to it to do various things like make it melt at a lower melting point, and one recipe i ran across was this one --
lime 11 parts
sand 63 parts
sodium carbonate 26 parts
my question is - can glass be made in small quantities by simply mixing these things together and firing them with an acetylene torch ? if so, should they be powdered in a mortar first or will they melt by simply being mixed together as course material ? i understand that the glass is a solution of those materials and it is not a reaction of any kind.
btw that recipe above, the page i found it on said that it was for green bottle glass, there was another one that had some sodium nitrate for removing the iron (or so i read) to make it clear. i'm sure someone has done this, make glass from sand, how did it work out ?

From: "Jeff" <>
Subject: Re: experiment
Date: Tuesday, January 15, 2002 3:01 PM

and the answer is ... yes! you can indeed make green glass by mixing one part hydrated lime from the hardware store with three parts sodium carbonate from the spa & pool store and 6 parts white sand. I did it with an acetylene torch, just mixed up the raw components on a piece of paper and poured them into a small pile on top of a fire brick and torched them until they melted together. the result was a bubbly green glass that looks a bit like a soft chewy wad of candy with tiny bubbles all in it. i do not know where the bubbles came from but I observed that the liquid almost seemed to be boiling, was the torch too hot for it, or was there water in the fire brick ? were the components inside the glass reacting chemically in some way ? I do not know the answer, but i know that the bubbles got bigger and then would pop and more bubbles would form.

the material or glass does not act in the same way as silver when fired with a torch. silver suddenly gets to a certain temperature and then it all kind of melts into this puddle that pulls itself together into a bead. the glass did not do that, it melted but did not have any will to come together as a puddle, it just melted. to get it into something like a puddle i had to rely on the "pushing power" of the torch's flame to nudge the glass into a little area from the outside.

I don't think it's a good idea to powder the components any smaller than they come, that is, I wouldn't crush the sand up any more than it was to get it all to melt better. the reason is that even though the materials were coarse the flame on the torch tended to blow some of the material around and make it difficult to melt. the torch had to be held well away from the pile of material and gradually moved towards it until some of the material started to melt and then the full force and heat of the torch could be used to melt the material.

you can tell that the result is some kind of glass, but it does not look like "glass" like being clear and see through and shiny and solid, etc. it looks more like bubbly melted plastic that is frozen except that it does have a glass like shine to it on the surface. i do not know what temperature the material melted at but it seemed to melt quicker than silver to me.

From: "TWB" <>
Subject: Re: experiment
Date: Tuesday, January 15, 2002 3:52 PM


What you made is headed toward being glass, but it's not finished. When glass ingredients melt, there are gases given off. That's probably what looks like boiling. When you're melting a tank of glass ingredients, you get to see the whole process -- from dry ingredients plus heat to what you saw, and then on to a frothy pot of partly melted glass, to a hot-as-hell pot of sure-'nuff glass. Then in that last stage of very hot, fairly clear glass you gather samples and pour them on a flat surface to observe the bubbles in the glass. First you see zillions of tiny bubbles, and then later you see fewer, larger bubbles. Then finally you "squeeze" the bubbles out of the glass by dropping the temperature of the furnace appropriately. The bubbles disappear into the glass. Magic!

Like a friend of mine used to say, you need to get it hotter, longer.
But that's hard to do on top of a firebrick. ;<)

Tom Bellhouse - NL
From: "Jeff" <>
Subject: Re: experiment
Date: Tuesday, January 15, 2002 4:46 PM

interesting, thanks for the response tom. so you melt all of this in some sort of oven then, are there plans out there for building an oven that will be hot enough to melt the materials into liquid glass like that ? what kind of container is the glass held in inside of the oven ? I would guess that the oven is made with a few layers of fire brick and that it has some sort of heat source inside like electric coils or something. do the materials have to be mixed together after they have melted ? or does the mixing while they are dry do the trick ? sorry for so many questions I'm just really curious about it. thanks for the response even if you do not have time to respond again!

TWB wrote:
> Jeff,
> What you made is headed toward being glass, but it's not finished.
Hi Mike,

Sure, you can use it.  Check the spelling, though -- gases? gasses?
Speaking of that, are you going to GAS in Amsterdam?


Mike Firth wrote:
>  I would like to install jeff's comments on my batch.htm page, if he gives
> permission, and would like to include your reply, if you do so.  Can i
> include your comments?

Glazes - Glazes are glass-like mixes that are applied to pottery/ceramics to provide decoration and seal the surface so that vessels will hold liquids without leaking or holding contaminants.  Below are two clear glazes from the site  http://digitalfire.com/4sight/material/index.html along with their summary version of the ingredients in float window glass.  The formulas and numbers to the right of the percent (%) sign are the content of each compound in the ingredient which are listed highest percent first.
While glass and glazes have a considerable percentage of SiO2, glazes tend to have more fluxes to ease melting 200-02-02

Clear Base-cone 10 (ox. & red.)
Potash feldspar-27 Potash Feldspar Formula: Na2O.Al2O3.6SiO2 - SiO2 64.76, K2O 16.92, Al2O3 18.32
Whiting-20 Calcium Carbonate CaCO3
EPK-20 EP Kaolin Plastic White Firing Kaolin - % SiO2 45.73, Al2O3 37.36, Fe2O3 0.79, TiO2 0.37, K2O 0.33, P2O5 0.24, CaO 0.18, MgO 0.1, Na2O 0.06
Flint-33 Silica SiO2 - % SiO2 94.00, CaO 3.00 
Wide Base-cone 4 to 10  
Potash feldspar-40 Potash Feldspar Formula: Na2O.Al2O3.6SiO2 -%  SiO2 64.76, K2O 16.92, Al2O3 18.32
Gertsley borate-40 Plastic Calcium Borate % B2O3 25.00 CaO 24.00 SiO2 14.00 MgO 4.00 Na2O 4.00 Al2O3 2.00 K2O 0.50 Fe2O3 0.50
Ball clay-15 Ball Clay Highly Plastic Fine Particle Clay SiO2 59, Al2O3 25, TiO2 1, Fe2O3 1, K2O 0.9, Na2O 0.4, CaO 0.3, MgO 0.3
Flint-10 Silica SiO2 % SiO2 94.00, CaO 3.00
Float Glass Float Glass % SiO2 74.48, Na2O 13.09, CaO 10.56, MgO 0.19, Fe2O3 0.06


Contact Mike Firth