Batch Melting Reactions

Three main stages: Chapter 5 of Paul's book

Stage 1 Characterized by absence of any melt

All free and most bonded water is removed

If any water or steam is present, there may be some hydrothermal reactions

Crystalline inversions occur

Organic materials burn, are otherwise oxidized, or decompose

Solid state reactions take place - new crystalline phases develop

Gases evolve, e.g., water, carbon dioxide, nitrogen, oxygen

Stage 2 - Melt is present

Melting reactions approach equilibrium

More crystalline compounds precipitate and are eventually disolved in the melt

Inorganic salts, if present, melt and decompose or partially dissolve in the melt

All gases, except fining gases, are liberated

At the end of the stage, the mixture consists of a melt with suspended refractory particles and gas bubbles

Frothing takes places because of the bubbles

Stage 3 - remaining refractory particles dissolve and bubbles are removed

Compatibility

To be able to mix glasses of two different compositions, the glasses must be compatible

That is to say, the two glasses must expand and contract at about the same rate.

The two glasses must also anneal at about the same temperature

If this is not so, the product of the mixture of glasses will crack - either right away or sometime later

Linear Expansion Coefficient

LEC is the property of changing size with temperature

LEC is a 'per' property

Other examples of 'per' properties

miles/hour, e.g., miles per hour

cost/quantity, e.g., $1.50 per dozen

density = mass/volume

LEC is the change in the length of a glass rod for a 1^oC change in temperature

With a = LEC, ´ L = change in length, Lo = original length, ´t = temperature change:

a =

If a rod changed from 100 to 101 mm as the temperature changes from 20^oC to 21^oC (or any other 1 degree C change), then the fractional change (LEC) is
1/(100 x 1) = 0.01

If a 1 meter rod changed 1 micrometer (1 x 10-6 meters) for a 1^oC change then 1 x 10-6 meters/(1 meter x 1^o) = 10-6, and LEC would be 0.000001

Common glasses have LECs in the range 0.000005 and 0.000013, i.e., 5 x 10-6 and 13 x 10-6 OR 50 x 10-7 and 130 x 10-7

LECs are usually reported as though the 10-7 were there, thus CaO with 1.63 is really 1.63 x10-7

These LECs seem like very small numbers, and they are

But these tiny numbers cause tiny stresses which cause glass to break!

Winkelmann and Schott and later English and Turner worked with LECs of the component oxides - these are the ones we've been using

Typical soda-lime glasses have been found to be compatible when the LECs of two different batches differ by 1.4 x 10-7 or less

Other glasses have larger tolerances, but smaller LEC difference is better

It is possible, if we had time, to calculate the adjustment in the composition of one glass batch so that it would be compatible with another glass batch, i.e., adjust the LECs of the two batches so they were within the tolerance

The proof is in the pudding

You can fuse two different glass samples together and look for stress lines using polarized light

You can fuse two glass beads and draw them out into a long thin 'wire' and see whether the thin 'wire' is straight or curves. If it curves, the two glasses are not compatible and the one on the inside of the curve has the larger LEC (pulling the other glass with it as it changes size)

Fining Process

Glass melts have bubbles - also called seeds

Fining agents (help) get rid of the bubbles/seeds

In the melting process

The lower melting oxides begin to melt first!

As liquid is formed and the temperature then can rise, the silica is melted

As the various solids begin to react with each other and decompose, gases are given off

CO₂ from the carbonates, CO₃^2-

H₂O from the hydrates, like borax, NaB₄O₇.10H₂O

N₂ from the nitrates, NO₃^-

Air trapped between the grains of the batch material may cause seeds - some references say yes, some say not a major source.

Sometimes seeds are formed as the molten glass reacts with the refractories (the high-melting vessels that the glass melt is contained in)

Two mechanisms

Buoyant Rise

Gases in seeds are less dense than the glass and will/should float to the surface

Buoyancy of the gas had to contend with the viscosity of the glass

Early in melt process, there is a lot of gas given off - and the bubbles coalesce and rise readily

Later, as reactions in the solid/melt near completion, the bubbles are fewer and more spread out - less chance to come together and rise

An old method was to dunk a block of wet wood or a potato - resulting steam created large bubbles which gathered the small bubbles

Newer - inject compressed air at the bottom of the melt

Gases diffuse into bubbles making them larger

As₂O₅ at the high melt temperatures Æ As₂O₃ + O₂

Sb₂O₅ does the same thing - at lower temperatures.

Na₂SO₄ also used - SO₂ gas formed

Solution of the seeds

At the high temperatures near the end of the reaction, the gases are not very soluble in the glass melt

Model is of cladding - a clad is a film of glass with gas in it

Like a halo around a silica grain

Forcing gas to be soluble in molten glass

As₂O₃, formed from the As₂O₅ reaction, scavenges O2 from smaller bubbles, draws the gas into the melt

Color in Glass

One obvious requirement - colorant must be stable at the high temperatures of melts!

Ions in the silicate matrices, i.e., network, produce colors by the same mechanisms as in pure pigments

Mostly polarizing of electron clouds in the metal ion-ligand bonding

Colors often the same in the glasses

Three basic ways to add colorant into the structure

Replace some of the Si atoms - inside the tetrahedral holes - the network formers - with transition metal ions, e.g., by Fe(III) or Co(II)

Replace some of the O atoms in the backbone structure, e.g., S^2- or one of the halides

Replace some of the Na or Ca ions, the network modifiers, e.g., Cu(II)

Sometimes hard to distinguish between replacing Si and replacing cations

Zumdahl pp 962-968 -crystal field theory

Shapes of d orbitals

Difference in 'splitting' with octahedral and tetrahedral geometries

Color seen is the complement of the energy absorbed!

Iron(II) and Iron(III)

Iron(II)

Blue

Present in reducing atmosphere or when reducing agents are present

Absorption primarily in IR and a 'tail' into the red

Iron(III)

Yellow to brown

Iron can exist as 4-coordinate or 6-coordinate

4-coordinate gives red-brown found in wine and beer bottles

Absorption in blue and green

Green color in normal window glass is a mixture of the Fe(II) and Fe(III)

Use of decolorizers to remove color

Iron a frequent impurity - to get colorless glass need to minimize amount of iron!

Add a component with complementary color - like the blue-haired women to keep hair from looking yellow.