close

Вход

Забыли?

вход по аккаунту

?

GLASS FORMATION IN THE SODA - SILICA - TITANIA, AND SODA - SILICA - PHOSPHORUS-PENTOXIDE FIELDS

код для вставкиСкачать
The Pennsylvania State College
The Graduate School
Department of Ceramics
Glass Formation in the Soda, Silica, Titania and
Soda, Silica, Phosphorus Pentoxide Fields
A Thesis
by
Alexis George Pincus
Submitted in partial fulfillment of the requirements
for the degree of
Doctor of Philosophy
August 194-0
Approved:
\/J
/J
Professor of Glass Technology
\0J
Department^of Ceramics
ACKNOWLEDGMENTS
I am g r e a t l y indebted t o Dr. E. D. T i l l y e r , Director of
Research, and t o other executives of the American Optical Company for
making i t possible f o r me t o pursue graduate v/ork and for extending to
me the use of t h e i r l a b o r a t o r i e s and equipment.
Professor W. A. Tf/eyl has been most generous and helpful i n
guiding t h i s research and b u i l d i n g up ray knowledge of g l a s s .
I have also
profited from conversations and studies vd.th Professors N. ¥ . Taylor
and B. E. "Warren.
4
TABLE OF CONTENTS
I. Introduction
1. Purpose
2.
Preliminary Considerations
II. Review of the Literature
1.
Phosphorus PentoxLde in Glass
2.
Titania in Glass
3. Soda, Silica Glasses
III. Procedures
IV.
1.
Raw Materials
2.
Crucibles
3.
Melting and F a b r i c a t i o n
4.
Measurement of P r o p e r t i e s
5.
Chemical Analyses
Results
1.
Soda, S i l i c a , Phosphorus Pentoxide "Field
2.
Soda, S i l i c a , T i t a n i a F i e l d
3.
V.
a.
Experiments with Kelting Procedure
b.
Limits of Glass Formation
c.
P r o p e r t i e s of Titania-Containing Glasses
Color Changes i n Ite^O, S i 0 2 , ?2°5 Glasses
Discussion
1.
Earls'- Theories
2.
"Warren's Contributions
3.
Primary Glass Forming Oxides
44
4. Binary Glasses
45
5.
The Problem of Bimiscibility
47
6.
Ternary Glasses
4.9
7.
Soda, Silica, Titania Melts
50
8.
Soda, Silica, Phosphorus Pentoxide Melts
53
^
VI. Summary
VII. Bibliography
A.
Glasses Containing Phosphate
B.
Glasses Containing Titania
C. Discussion
ILLUSTRATIONS
Figure
1
Between Pages
Densities and Refractive Indices of Glasses i n
t h e System Na 2 0-Si0 2 .
22
~ 23
2
Batches i n the Field Na 2 0, S i 0 2 , P 2 0 5 -
27-28
3
Typical Softening Temperature Curves
30 - 31
4
Typical Dilatometer Curves
30 - 31
5
Properties of Na^O, S i 0 2 , P 2 0^ Glasses
31 - 32
6
Batches Prepared i n the Soda, S i l i c a , Titania
Field
33 - 34
Ultraviolet Absorption Spectra of a 20;'i Ti0 2
Glass Batch Melted at Indicated Temperatures
36 - 37
7
INTRODUCTION
The l i t e r a t u r e of g l a s s technology contains abundant m a t e r i a l
on the synthesis of g l a s s e s by the i n t e r a c t i o n of s i l i c a with a l k a l i s and
with oxides of b i v a l e n t elements.
Much information also e x i s t s about
the glasses formed w i t h i n the t h r e e component f i e l d s :
the t r i v a l e n t oxides of boron or aluminum.
soda, s i l i c a , and
Only meager evidence i s
available about the i n t e r a c t i o n of sodium s i l i c a t e glasses with the
oxides of four or f i v e valent elements.
The purpose of t h i s t h e s i s i s t o contribute t o the knowledge
of the glass-forming f i e l d s by examining t h e e f f e c t s of additions of
t i t a n i a and of phosphorus pentoxide on some fundamental cliaracterisbics
of s o d a - s i l i c a g l a s s e s .
2
Preliminary Considerations —
Valence is an important factor,
often a dominant one, in the synthesis of glasses. Silica glass owes its
remarkable practical properties to the fact that the four valent silicon
cation (Si^) surrounds itself with four strongly held bivalent oxygen
anions (02~*). To make a more easily melted glass, monovalent alkali
ions are introduced.
Then to produce more durable and workable glasses
bivalent alkaline earth ions are also included.
Ions which can have higher valences are known to contribute
various desirable properties to glasses even when they are present in
relatively small amounts. Among those which have been extensively used
are A l > , B 3 + , Zr^ + , A s 3 + and As^ + , sb^ + and Sb^ + , and Fe 3 + .
In the present state of our knowledge, it is not clear how
these higher valence elements enter into the structure of glasses. X-ray
diffraction lias not yet been developed to handle glasses with more than
two components, and is afflicted with other limitations.
Physicists may later develop techniques for definitely determining the atomic and molecular arrangements in complex glasses. But
at present, it seems to the writer that what is wanted is accumulation of
facts about the whole range of glass-forming fields, the factors limiting
their extent, and the reactions of oxides to form glasses from the viewpoint of the chemical forces involved.
Titania and phosphorus pentoxide were chosen for this particular study because they offer certain unique features of both theoretical
and technical interest. Titania can exist as either a colored or a
colorless ion in glass, and thus provides a self-indicator of its position
3
in the vitreous structure. Recent advances in the technology of preparing titania for pigment purposes have made cheaply available a highly
pure raw material. This has led to renewed interest in the value of
titania additions for improving the tarnish resistance of glasses,
vitreous enamels, and glazes. Titania-containing glasses tend to develop
yellow coloration, control of which has remained a problem.
Phosphorus pentoxide, like boric oxide, is a glass former itself, and it will be useful to c ompare the glass-forming characteristics
and the reactions of these three oxides.
Silica-free phosphate glasses
are finding ever increasing uses because of their many unique physical
and chemical properties. Yet no reports have been published which disclose the extent to which these characteristics can be imparted to or
merged with those of silicate glasses.
In choosing the physical properties to be measured on the
glasses surveyed in this work, consideration was given to the factors:
characteristics of practical interest in determining the utility of
specific glass batch additions, and characteristics of theoretical interest
in disclosing facts about the chemistry and structure of glass.
Softening temperature, thermal expansion coefficient, and
penetration modulus were chosen as measures to some extent of the strength
of the glass. Comparison of the temperatures of melting, softening, and
annealing of a glass furnishes information about its working range.
Thermal expansion plays a dominant role among the factors v.hich control
the heat shock resistance. Penetration modulus measurements by the
A
recently developed Knoop-Peters' method has been shown to give excellent
guidance on the practically important property of scratch hardness.
Refractive index was measured because i t furnishes information
about the packing of atoms and because i t indicates the possible usefulness of the ingredient under investigation for glasses with special
optical properties.
Chemical durability i s one of the fundamental factors i n the
application of g l a s s .
Even though the simple glasses concerned i n t h i s
study are too soluble t o be of practical use, evaluating t h e i r relative
durabilities indicates the effects which the t h i r d constituent might have
when used i n commercial glasses.
I t i s particularly interesting to
consider the relative acidities and the influence upon chemical balance
of the higher valent cations.
The changes i n the colors and absorption spectra of representative coloring ions caused by variations i n the base glass composition
i s a fascinating field of aesthetic and practical i n t e r e s t , and has been
developed by Weyl into a valuable tool for investigating the structure
and chemistry of glass.
5
REVIEW OF THE LITERATURE
Phosphorus Pentoxide in Glass -
Only fragmentary information
exists about the extent to wliich clear glasses can be formed from silica,
basic oxides, and phsophoric oxide. Use of bone ash or other source of
phosphate to make opal glasses is an old art, dating back to the 14th
century at least (6). Such glasses invariably contained calcium and had
different characteristics than those to be considered here, for calcium
phosphate is insoluble in the silicate melt and separates as immiscible
liquid droplets wliich may later crystallize (19).
P20c frequently appears in analyses of gold and copper rubies.
Weyl1 suggests the following explanation of this practice: Phosphate
glasses have low dielectric constants and their weak solvent action is
further indicated by the sharp absorption and emission spectra -which
coloring ions produce when dissolved in them. The above mentioned liquid
droplets of phosphates within a silicate melt, therefore, provide interfaces at which precipitation might occur. This precipitation causes
formation of nuclei and influences those colors in glasses vihich are due
to colloids or to undissociated compounds.
Commercial glasses are generally too complex to give more than
hints about the limiting compositions for glass formation. Opalescent
or opal compositions which have been quoted include those listed in
Table I, indicating that about 5% P20r is the limiting amount used, and
that as little as 3% leads to opalescence.
Personal communications.
TABLE I
Weight Percentage Compositions of Phosphate Opal Glasses
1+
2+
3++
4++
5++
Si0 2
57.4
70.3
71
67
66
p2o5
4.6
2.9
3
6
3
Na
3.4
10.9
1
4
7
K20
12.0
0.9
16
14
—
CaO
10.2
4.7
9
9
4
PbO
12.4
6.9
-
-
-
—
3.4
-
-
15
Al203
—
—
—
mm
3
ZnO
—
——
—
•"
2
B
2°
2°3
Inwald (14) found that calcium phosphate dissolved i n ordinary
silicate glasses to give a clear glass at low temperatures provided that
alkali or boric oxide was present i n significant amounts. Large quantities
of calcium phosphate could be dissolved i n ordinary glass at very high
temperatures, but on cooling the phosphate precipitated.
He concluded
that replacement of Si0 2 by P 2 0 5 i s limited to 0.25 molecule, which i n
the ordinary range of soda-lime-silica glass compositions amounts to
about %*
+ Knapp (17)
++ Kreidl and Weyl (19)
7
As for the corresponding solubility of s i l i c a i n sodium phosphate glasses, Huttner (12) said that up to 5% Si0 2 can be dissolved i n
sodium metaphosphate and s t i l l form a clear glass.
Stanworth and
Turner (26) described four experimental melts i n t h i s region (Table I I )
and stressed the great influence of small amounts of basic oxides i n
aiding formation of glasses i n the silicon phosphate region.
TABLE I I
Stanworth and Turner's Melts
Glass No.
•
1
£&
76.43
Si02
8.26
Na20
13.90
2
82
1£)
8
3
65
28
10
4
64
23
13
Remarks
Melted a t 1400 t o e a s i l y
poured glasses -which
suffered only s l i g h t
d e t e r i o r a t i o n by exposure
t o the atmosphere for a
week.
No glass even a t 1600.
No appreciable difference
from Glass No. 3
The relative effects of boric oxide, t i t a n i a , and phosphoric
oxide on the water solubility of a soda-lime-silica glass lias been i n vestigated by Nagai and Takahashi (22).
In a base glass (72.6$ Si0 2 ,
0,60$ RgOo, 12.73$ R0, and 14.00$ Na-^O) which corresponds to a commercial
formula, up to 8% of the s i l i c a was replaced by each of these oxides i n
turn.
The resulting glasses were tested by extraction at water bath
temperatures and i n an autoclave.
In the water bath t e s t , a l l three
substitutes improved the durability of the parent glass, boric oxide
being best and phosphorus pentoxide the l e a s t favorable.
In the auto-
8
c
clave t e s t , the phosphate-containing glasses developed opacity and were
most strongly attacked.
A point about glasses containing both s:LLica and phosphorus
•which must be mentioned i s the patent claim of Grimm and Huppert (7)
that transparent glasses of highly desirable properties can be obtained
provided sufficient alumina i s present i n proportion to the phosphate.
Some of the examples T/hich they quote are given i n Table I I I .
Berger (2)
claims that easily reducible oxides, particularly antimony, are helpful
in aiding compatibility of PgOr and Si0 2 .
Partridge (24) has patented high alumina compositions of lower
p
2°5
corr en-
k k £ ° r sodium or mercury vapor resistant glasses (Table IV)
and Stanworth (27) i n the same laboratory has recently published continued studies i n t h i s field.
In the composition range 45$ silica, 35$
alumina, 15$ lime, and 5$ boric acid, up to 3$ P ^ substituted for
B203 prevents devitrification on cooling and aids drawing into rod form.
Wien Al203 i s l e s s than about 28$, P ^ produces very viscous glasses
covered with scum.
Silica-free phosphate glasses provide material for a long
chapter in themselves, but they are outside the scope of this t h e s i s .
Reviews of t h e i r history and applications have been written in recent
years by Schmidt (25), Stanworth and Turner (26) and Khapp (17,1B).
They are being used today for t h e i r special refraction - dispersion
properties (11) for low-softening temperature glasses (13), for alkali
vapor resistance (8), for fluorescent tubes (16), for glasses with high
TABLE I I I
Compositions Given by Grimm and Huppert
Si02
41.6
37.5
35.3
29.6
14.7
10.0
p2o5
24.5
22.2
21.1
26.2
34.9
23.3
%°3
17.6
15.9
15.1
3J8.3
25.0
33.7
Na20
16.3
9.7
23.0
25.4
25.4
1.0
K20
—
MgO
—
—
—
—
10.0
CaO
—
—
2.0
¥3
—
—
14.7
20.0
—
TABLE IV
Compositions Given by Partridge
2
1
29.3
49.7
2°5
7.9
3.3
A1 2 0 3
20.1
26.3
B203
1.8
5.0
MgO
5.2
2.7
CaO
7.0
BaO
25.4
2.9
ZnO
5.2
Si0 2
p
—
6.8
10
ultraviolet transmitting and special absorption properties (9, 10, 1,
2, 20) and have been suggested for containers for hydrofluoric acid (3).
Sodium phosphate glasses are finding extraordinary applications in water
treatment and as a hydrogen ion conditioner (23) Calcium phosphate glasses
are being developed for fertilizers, especially under the sponsorship of
the T. V. A. (21).
Titania in Glass -
Many of the fundamental facts about the
utilization and advantages of titania for glasses were disclosed in a
British patent issued to George and Charles Leuchs in 1893 (41). Although they stressed the applications to glazes and enamels, their statements have been found to be equally applicable to a wide range of glass
compositions.
Messrs Leuchs found that titanic acid can replace a portion of
the silicic acid in glazes without resultant coloration provided that the
batch is free of coloring oxides and is melted in an oxidizing atmosphere.
They asserted that the titanic glazes fuse more easily than glazes containing only the equivalent of silicic acid and are much more resistant
to the decomposing action of humidity and acids than the corresponding
silicic glazes, even when the silica content lias been reduced to a few
molecules.
Glazes containing only one molecule of titanic acid and two
molecules of silica referred to one molecule of total alkali plus bivalent oxides are said to be as acid resistant as those high in silica,
and yet to melt as easily and run as well as glazes containing boric
11
acid or fluorides or lead.
In addition they exhibit high specific gravity,
refraction and luster, and thermal expansion.
When used in large proportions, titanic acid was found tc produce "dulled glazes," and it was mentioned as a possible substitute for
or in combination with tin dioxide, phosphates, and fluorides -for opacification. Higher temperatures lead to a more or less transparent glass,
colored yellow to brown by solution of the iron.
Specimen batches mentioned are:
Transpare n t Glazes
Quartz
60 - 360
Boric acid
62 - 83
Titanic acid
20 - 80
20 - 30
Nitrate of soda
15-30
15-30
Soda
116
Carbonate of Lime
33 - 50
]Snamels
240 - 360
60 - ISO
—
—
80
30 - 1B0
60-83
80 - 160
15 - 30
15-30
116
116
116
100
100
33 - 5 0
"In these compositions there can be substituted wholly or partly:
soda by potash, calcium by aluminum, barium, strontium, magnesium, zinc,
bismuth, antimony, thallium, lead, and for coloration by cobalt, nickel,
chromium, iron, manganese, copper, uranium."
General papers on the utilization of titania in enamels, especially of the acid resistant type, have been presented by Vondracek (51)j
Landrum and Frost (40), Button and Wagner (35), KLnzie and Plunkett (39),
and Aldinger (29). Generally 3 to 8$ Ti0 2 i3 used in acid-resistant
enamels, though as much as 14$ has been recommended in an alumina-free
batch (39).
in many ways titania would be an ideal opacifier, but as the
12
leuchs .'patent pointed out, i t tends to dissolve and opacification i s l o s t ,
pother difficulty has been the yellow cast of titania-containing enamels.
A German patent j u s t issued t o the Kaiser Wilhelja. I n s t i t u t e (38) declares
that more than 12$ A l ^ must be present i f Ti0 2 i s to act as an opacifier.
In 1912, Thomas (49) made claims that fused s i l i c a glass could
be melted at lower temperatures and have superior properties i f 0.5 to
2% titania or zirconia was included.
Apparently the claims have not been
substantiated by l a t e r developments.
Bunting (32) has made a phase rule study of the system Ti0 2 SL02 in which he found t h a t no compounds are formed between the two oxides,
but a eutectic exists at 1540°C. i 10° having the proportions 89.5$ Si02,
10.5$ Ti0 2 .
Badger and Doney melted titania-alumina-silica glasses at l680°C.
and 2000°C, obtaining"fairly homogeneous glass" in the high, s i l i c a
corner above 70$ at 1680°.
At oxyhydrogen flame temperatures clear glasses
resulted along the alumina-silica side of the triangle between 60 and
90# Si0 2 wherever Ti0 2 "was more than about 5$.
Attempts to take advantage of the refraction characteristics
of titanium have occupied many investigators, even as far back as the
vrork of Harcourt, reported by Stokes i n 1871 (47) and 1875 (43).
In
phosphate glasses t i t a n i a was found to cause noteworthy dispersion at the
blue end of the spectrum.
In s i l i c a t e glasses t i t a n i a showed considerable
dispersive power, but l e s s notable differences b etween the blue and red
ends of the spectrum.
13
The l a r g e amounts of t i t a n i a -which can be dissolved i n t r a n s parent glasses was revealed i n a patent issued t o Titanium Pigment Co.
(50) i n 1920 which claimed i n t r o d u c t i o n , of Ti0 2 i n t o glass batches i n t h e
ratio of not l e s s t h a n 25$ of t h e whole.
An example i s given a s :
Calculated analysis
69
Si0 2
43
Borax
10
B203
3
1^003
29
NagO
13
CaO
7
CaO
5
Ti0 2
45
Ti0 2
31
Si0 2
.
I . G. Farbenindustrie (36) claimed the use of t i t a n i a t o y i e l d
glass of 1.70 r e f r a c t i v e index derived from a batch such a s :
Si02
100
Ti02
71.5
K2C03
47.1
KNO3
17.2
CaC03
42.8
As 2 03
0.3
•which gives a c a l c u l a t e d analysis of roughly 43$ s i l i c a , 30$ t i t a n i a ,
17$ potash, and 10$ l i m e .
A l a t e r p a t e n t by t h e same concern (37) claimed t i t a n i a glasses
vdth 20 - 48$ T i 0 2 , 20 - 35 a l k a l i , added mainly as potassium n i t r a t e i n
order t o keep the g l a s s c o l o r l e s s , and t h e balance acid oxides.
14
Berger (31) used t i t a n i a to improve the resistance to tarnish
and devitrification of high index barium glasses.
I t i s pointed out
that 25$ of titanium oxide makes the glass assume a brownish color,but
nmch less than 20$ suffices to strongly reduce the solubility in acids
and correspondingly the l i a b i l i t y of the glass t o become stained.
The
examples which t h i s patent specification gives are tabulated i n Table V.
TABLE V
Compositions of Tranish Resistant Barium Glasses
Si02
32.0 .
32.0
32.0
B203
8.0 ••
8.0
7.5
AI2O3
•
—
0.5
Na20
0.2
—
—
CaO
10.0
10.0
10.0
BaO
43.5
37.2
36.2
ZnO
3.0
3.0
2.0
PbO
0.3
6.5
0.3
Ti02
2.5
3.0
11.0
As203
0.5
%>
V
—
0.3
0.5
100.0
100.0
100.0
1.653
1.671
I.696
51.4
47.3
42.0
15
Results of t h e i r researches into the effects of t i t a n i a on the
properties of glass were published by Sheen and Turner i n 1923-24 (46).
Some preliminary experiments were carried out to establish the fusibility
of mixtures containing t i t a n i a at the ordinary glass-melting temperatures
of 1400-1500°. 30, 50, 70, and 80$ Ti0 2 mixed with s i l i c a did not fuse,
but increasing sintering was observed up t o a dark brown mass at 80$
Ti02. A mixture consisting of 6 Ti0 2 • 2 CaO sintered to a hard brown
cake. Two soda, t i t a n i a , s i l i c a batches were t r i e d .
15 Wa20 formed a dark brown glass.
70 Si0 2 , 15 Ti0 2 ,
64 Si0 2 , 27 Ti0 2 , 9 Na20 melted to
a very dark blue, almost opaque glass.
Finally t i t a n i a was substituted
molecularly for s i l i c a i n a Na 2 0: CaO: 6 Si0 2 glass, adding 1, 1.5, and
2 molecules which are equivalent t o 16 t o 31$ Ti0 2 .
All three formed
dark brown glasses.
For the batches used for preparation of t e s t specimens, Sheen
and Turner melted at 1400 - 1500° for five hours six glasses starting from
6 Si0 2 • 2 Na^O i n which Ti0 2 was substituted for Ka20 i n steps of 0.1
molecule from 0.1 t o 0.6 Ti0 2 .
The compositions of t h e i r glasses are
quoted i n Table VI and some of t h e i r data on physical properties in
Table VII.
From t h e i r experimental data the authors concluded that r e placement of soda by t i t a n i a decreases solubility in boiling water, i n creases the annealing temperature, considerably reduces thermal expansion,
and increases refractive index.
No decrease in the solubility i n alkali
16
•vras observed u n t i l 10$ T i 0 2 a f t e r which these glasses are superior t o high
silica g l a s s e s .
The t i t a n i a - c o n t a i n i n g g l a s s e s did not e x h i b i t maxima o r
.minima on the composition — physical property curves.
Extended d u r a b i l i t y measurements on t h e s e same glasses were d e s cribed i n 1926 by MLmbleby and Turner (33), who pointed out t h a t the
t i t a n i a improved t h e r e s i s t a n c e t o v/ater and hydro c h l o r i c acid, but not t o
sodium hydroxide o r sodium carbonate s o l u t i o n s .
Nagai and Takahashi's (22) comparison of b o r i c oxide, t i t a n i a ,
phosphorus pentoxide, and s i l i c a i n a soda-lime glass has already been
mentioned (Page 7 ) .
Their t i t a n i a - c o n t a i n i n g - g l a s s e s had the compositions
given at the bottom of Table VI.
Im a continuation of t h i s study Nagai and Masuda (44) prepared
eleven glasses i n two s e r i e s s t a r t i n g from a s o d a - l i m e - s i l i c a glass having
the composition:
72$ S i 0 2 , 15$ Na 2 0, 13$ CaO.
In t h e f i r s t s e r i e s t i t a n i a
replaced lime i n t h e amounts 2 , 4, 6, 8, 10, and 13$, i n the second s i l i c a
was replaced i n t h e amounts 5 , 10, 1 5 , and 25$.
Members of t h e f i r s t
series are c a l l e d t i t a n i t e g l a s s , because i t i s said t h a t t h e t i t a n i a combined with t h e s i l i c a .
Members of the second s e r i e s are c a l l e d t i t a n a t e
glasses because "In t h i s case t i t a n i a combined with soda o r lime as
titanate."
This c l a s s i f i c a t i o n i n t o t i t a n a t e and t i t a n i t e glasses f i t s i n w i t h
a coBment by D i t l e r (34) t h a t t i t a n i c acid being amphoteric can replace
either s i l i c i c acid or alumina i n s i l i c a t e s .
The glasses were analyzed chemically, then t e s t e d and compared
for degree of c o l o r a t i o n , s p e c i f i c g r a v i t y , water s o l u b i l i t y by the grain
17
TABLE VI
Chemical Analyses of Experimental Glasses Containing Titania
Code
No.
Turner & Sheen
753 a
Si02
J^_
Na
Al 2 03
Ti02
_*?2°3
2°
74.53
0.95
2.04
0.11
0.04
22.36
b
75.22
0.89
3.01
0,10
tr.
20.71
c
76.30
0.87
4.52
0.19
0.16
18.06
d
74.86
0.52
6.38
0.12
0.12
17.32
e
73.56
1.15
8.06
0.15
0.58
16.40
f
73.00
1.06
9.39
0.19
0.81
15.14
754 a
73.42
0.53
1.94
0.13
0.12
23.68
b
72.83
0.44
3.01
0.15
0.3J3
23.16
c
72.82
0.29
4.95
0.10
0.16
21.82
d
73-30
0.61
6.88
0.11
0.26
18.73
e
72.61
0.65
10.03
0.16
0.07
16.31
f
72.41
0.75
14.91
0.15
0.06
11.90
g
70.52
1.84
19.45
0.17
0.06
8.24
Nagai and Takahashi
SCN- T I
68.30
1.11
2.45
0.22
12.65
14.54
SCN - T i l
65.88
1.14
5.06
0.08
13.41
14.13
SCN- T i l l
63.02
1.05
7.59
0.14.
13.01
14.90
method under atmospheric, or 5 and 10 kg/cm2 pressures, thermal expansion
and softening point with a Chevenard dilatometer, and degree of devitrification exhibited after heating at 700, 800, 900, and 1000CC. for 30 minutes.
IB
The results are summarized i n Tables VIII and IX.
The authors' con-
clusions vrere:
1.
The t i t a n a t e glasses are stronger and more stable i n various
respects than the t i t a n i t e glasses.
2.
In the t i t a n i t e series, the glass containing lime and t i t a n i a
in nearly equimolecular ratio has a clear tendency to show
maxima or minima in several chemical and physical properties.
TABLE VIII
Chemical Analyses of Nagai and Hasuda's Glasses
CaO
MgO
Na20
13.32
0.30
14.39
1.98
11.39
0.21
14.79
0.11
4.02
9.09
0.27
14.94
1.29
0.08
6.12
7.11
0.21
13.07
71.83
1.33
0.10
8.03
5.02
•0.24
13.39
SC(T)N - 10
72.31
1.14
0.13
9.39
3.24
0.30
13.05
SC(T)N - 13
71.79
1.32
0.10
12.87
0.30
0.27
13.35
S(T)CN - 5
66.87
1.27
4.96
13.41
0.30
13.19
S(T)CN - 10
62.04.
1.04
10.32
13.31
0.28
13.00
S(T)CN - 15
56.91
1.21
14.91
13.28
0.31
13.38
Code No.
Si02
A1 2 0 3
pe203
SON - 01
70.24
1.12
0.13
SC(T)N - 2
70.30
1.23
0.10
SC(T)N - 4
70.27
1.30
SC(T)N - 6
72.12
SC(T)N - 8
Ti0 2
-
TABLE IX
Properties of Nagai and Masudafs Glasses
Code No.
Color
Density
Thermal Expansion
Coefficient 20-54O°C.
x 10-6
Softening
Point
°C.
Attack by
N/50 H2S04
SCN-01
Colorless
2.548
7.57
579
9.07
SC(T)N-2
Light grass green
2.529
6.67
600
7.95
SC(T)N-4
Yellowish brown
2.516
6.56
618
8.12
SC(T)N-6
Light yellowish brown
2.515
10.11
591
10.33
SC(T)N-S
A l i t t l e stronger
2.509
9.06
593
10.31
SC(T)N-10
Brownish yellow
2.645
5.54
620
7.55
SC(T)N-13
A l i t t l e stronger
2.654
5.20
640
6.31
S(T)CN-5
Yellowish green
2.551
9.75
612
7.32
S(T)CN-10
Light brownish yellow
2.618
10.00
622
5.02
S(T)CN-15
Brownish yellow
2.662
10.15
670
4.17
S(T)CN-25
Purplish brown
—
—
—
1
s
A
20
Morey and Merwin (42) i n 1932 mentioned t h a t they have studied
numerous glasses i n the soda, s i l i c a , t i t a n i a field, but t h e i r results
have not been published up to t h e present time.
In his recently published
monograph (43, p . 393) Morey quotes the compositions and optical properties in the unannealed state of five of these glasses:
Si02
NagO
Ti0 2
%
V*
71.8
23.5
4.7
1.520
53.1
68.7
22.1
9.2
1.543
47.6
38.5
31.5
30.0
1.666
31.3
52,6
17.4
30.0
I.676
30.2
20.1
29.5
50.4
1.798
23.4
Sawai and Suda (45) studied the influence of t i t a n i a on a sodalime-silica glass containing cobalt:
Si02
71.82
*h°3
0.92
CaO
4.61
C03O4
5.93
Na^O
(16.67)
They concluded that the thermal expansion coefficient rises with increase
in t i t a n i a and the effect i s greater above the transformation temperature,
Replacement of s i l i c a by t i t a n i a seems to increase the hardness.
Acid
resistance i s also improved, but a l k a l i resistance i s proportionately
lowered.
21
Soda. Silica Glasses — As the base line from which almost a l l
commercial glasses are derived, and also as the basis for the important
soluble s i l i c a t e industry, glasses formed by the interaction of soda and
silica have been extensively investigated.
The classic works in t h i s
field are the phase equilibrium studies from the Geophysical Laboratories,
supplemented by Morey and Bowen's careful measurements of density and
optical properties.
The published researches of Gelstharp, of Peddle,
of Turner and his coworkers at Sheffield, and of Finn and his collaborators
at the Bureau of Standards have also been of fundamental importance.
The
available information about glasses i n t h i s f i e l d has recently been
critically surveyed by Morey (43) and only brief references to pertinent
material will be given here.
In following the discussions about sodium s i l i c a t e glasses, i t
is helpful t o bear i n mind the relationships between t h e i r molecular
formulae and percentage compositions because many authors write mainly
in terms of empirical r a t i o s .
values are set down below.
For convenient reference some of these
Since the molecular weights of s i l i c a (60.1)
and soda (62.0) are nearly the same, the molecular and weight percentages
are approximately equal.
Molecular Formulas
Na20
Si02
6
1
1
5
1
4
1
3
1
2
1
1
2
1
Weight Percentages
Na20
Si0 2
85.7
14.3
17.0
83.0
20.5
79.5
74.5
25.5
34.0
66.0 ..
50.9
49.1
67.7
32.3
22
I t i s generally known t h a t around t h e m e t a s i l i c a t e r a t i o e a s i l y
ciystallized, very soluble glasses are formed, although by strong quenching
i t i s possible t o produce g l a s s e s with as l i t t l e as 32$ S i 0 2 .
In the d i -
to t r i s i l i c a t e ranges, t h e b e s t melting, most s t a b l e glasses are formed,
but by 75$ s i l i c a the" sand d i s s o l v e s slowly and above about 80% S i 0 2
production of c l e a r glass i s hindered by d e v i t r i f i c a t i o n or incomplete
solution.
Liquidus temperatures f o r g l a s s e s i n t h e binary system sodasilica may be derived from the f a m i l i a r phase equilibrium diagram a f t e r
Morey, Bowen and Kracek (43, p . 3 8 ) , but such temperatures i n d i c a t e mainly
the region through wliich t h e melt must be cooled r a p i d l y .
To obtain
rapid melting and good q u a l i t y , i t i s necessary t o f i l l and hold t h e
batch at about 1350°C. with t h e whole range of these g l a s s e s .
Density and r e f r a c t i v e index measurements i n t h i s f i e l d have
been summarized by Morey and Mervdn i n the graphic form which i s reproduced as Figure 1 .
Finn and h i s co-workers have obtained thermal expansion data
for nineteen of t h e s e glasses ranging from 52.15 t o 82.76$ S i 0 2 .
Similar
data from Turner's l a b o r a t o r y are summed on pages 275 - 7 of Morey's
book.
2 20
/Vd^O
60
WEIGHT
P£R
?0
SO
O f A / .5 02
SiO,
-3c. 1.—The D m - i t i c s ami H«*ir:n-ti\t* Iiulu-i- of CI1U>M-.> in the S \ > u m
Na a 0-SiO a , as Determined by Several U1JM rvers. Aitt-r Morey ami Mcrwm.
FIGURE
23
PROCEDURES
Limiting composition ranges for glass formation were f i r s t
roughly established, mainly by adding increasing amounts of the t h i r d
constituent to sodium s i l i c a t e f r i t s and melting at the lowest temperatures
found to be p r a c t i c a l .
Then representative points throughout the glass-
forming fields were selected, and batches were calculated to yield glasses
of the corresponding compositions.
Raw Materials of silica.
A high grade glass sand was used as the source
Soda was derived mainly from commercial soda ash of good
quality, and i n some cases p a r t i a l l y from sodium metaphosphate.
The
source of t i t a n i a was an excellent quality of low iron titanium dioxide.
All of these raw materials were pure enough that after preliminary
checks on the composition of the stock, the theoretical formula could be
used i n batch calculations.
A small correction was applied for 2.3$ of
moisture in the soda ash.
Crucibles — Iyanite crucibles were slip-cast from a body made
up of:
Celo lanes Iyanite (200 mesh)
70
Kentucky d a y Mines No. 1 Ball Clay
20
Georgia Kaolin Co. G-l Clay
10
Water
20
Sodium metasilicate
0.1
Soda ash
0.1
This body has excellent characteristics throughout manufacture and use.
24
pelting and Fabrication -
Batches t o yield 100 grams of glass
were weighed into glass b o t t l e s and tumbled u n t i l apparently homogeneous.
It has been found that such mixing i s good enough for small batches and
makes grinding in a mortar or mill unnecessary.
Melting was carried out i n a Globar electric furnace with a
consequent strong oxidizing atmosphere. Furnace temperatures were
measured v&th a noble metal couple and controlled with an induction type
voltage regulator or with a potentiometric controller - recorder.
The batches were f i l l e d into the crucibles at about 1300°C. to
get rapid reaction and t o prevent foaming over in the highly viscous
region. Then the temperatures were adjusted to the range which gave a
good fluid s t a t e , and the melts were held there for several hours u n t i l
homogenization seemed adequate.
Since no chemical agents to assist fining could be used i n these
experimental melts, i t was necessary to use s t i r r i n g , heat shock, r e melting, and other techniques to obtain satisfactory quality with some
of the extreme glasses.
Further handling of the melts depended upon the type of t e s t
specimen which was needed.
For optical properties 50 mm. discs were
cast in graphite molds and annealed in an electric muffle furnace at
about 550°, then l e f t to cool overnight.
If these discs s t i l l showed
strain, they were reheated t o slightly higher temperatures and for longer
times. For softening point and thermal expansion specimens, rods were
drawn out of the glass i n the crucibles with a controllable speed apparatus
25
after the melts had been cooled to the working range. The thermal expansion
specimens, which were about 4 mm. thick, were then annealed.
For chemical analyses, chemical durability, and i n some cases
refractive index determinations, an annealed disc was powdered to 100 200 mesh screen sizes.
In some cases three kilogram batches were melted i n large crucibles,
quenched by pouring into water or into graphite molds, then powdered,
mixed, and used as bases for colored glasses.
Measurement, of properties — Softening temperatures were measured
by the fiber elongation method which i s i n general use in t h i s country as
it has been propounded by Littleton.
Duplicate determinations had to
check within 3°C.
Thermal expansion coefficients and c r i t i c a l temperatures were determined with a Chevenard dilatometer.
Refractive indices and optical dispersions were measured where
possible on a polished surface with an Abbe refractometer.
YJIth a few
badly striated specimens i t was necessary to measure their refractive
indices by the immersion o i l method with a microscope and select the
most representative value to the second or third place after the decimal.
"Hardness" or penetration modulus was measured on polished surfaces with an apparatus based on that developed by Peters and Knoop at
the Bureau of Standards.
Preliminary testing indicated that a l l of the soda,silica, P205
glasses were too soluble to justify quantitative appraisals of t h e i r
chemical durabilities.
Therefore, their durabilities were appraised
26
roughly by examination of polished surfaces after several months storage.
The glasses were then classified into one of four groups:
1 —
commercial grade
2 —
slight attack
3 —
film formation
4 —
hygroscopic
Chemical Analyses —
was as follows.
The scheme of a n a l y s i s which was used
Loss on i g n i t i o n of t h e powdered sample was determined
in a l l cases as some of the g l a s s e s were very hygroscopic.
Heating a t
105°C. was u s e l e s s because t h e powders gained i n weight during t h i s
treatment.
Analytical values were based on the i g n i t e d sample.
S i l i c a was determined by the usual double dehydration method.
Any residue a f t e r t h e hydrofluoric a c i d t e s t was added t o the main
f i l t r a t e , and alumina was then separated i n acetate buffered solution as
the phosphate and i g n i t e d t o A1P0,.
Usually the E 2 0 <j content of t h i s
substance was checked by fluxing i t with sodium carbonate, dissolving
in n i t r i c acid, and p r e c i p i t a t i n g as phosphomolydate.
Total P20<j was
determined on a separate sample which had been put i n solution by
attack with hydrofluoric and p e r c h l o r i c a c i d s .
After adjustment of
ammonium and n i t r a t e concentrations, t h e phosphomolydate p r e c i p i t a t e was
brought down.
F i n a l determination was by s o l u t i o n of the phosphomolybdate,
reprecipitation vd.th magnesia mixture, then i g n i t i o n of the p r e c i p i t a t e
to Mg2P207»
27
Great difficulty was experienced in obtaining satisfactory
determinations of Na^O, and these were finally omitted because of lack
of time. This throws the entire error of the analysis upon the soda, and
this is particularly objectionable because the samples may contain
appreciable weights of water. The most promising approach to valid soda
determinations seems to be a modified J, Lawrence Smith method.
Fig. Z
Batches in The F/e/ej
<&
Code
o C/ear
Glass
*
Non-v/freous
®
Separations
28
RESULTS
Soda, S i l i c a , Phosphorus Pentoxide
Field
Batches which were prepared during t h e studies on t h i s t r i a x i a l
field are p l o t t e d i n Figure 2 .
The symbols i n d i c a t e t h e p o i n t s a t which
clear glasses were obtained by ordinary melting and cooling techniques,
or where formation of c l e a r g l a s s was prevented by r e f r a c t o r i n e s s , rapid
d e v i t r i f i c a t i o n , o r opalescence.
I t can be seen t h a t c l e a r g l a s s e s were obtained from combinations
of soda and s i l i c a w i t h i n t h e l i m i t s which have been previously described.
The preliminary meltings i n d i c a t e d t h a t additions of phosphate t o a high
s i l i c a f r i t , approximating t o t h e t r i s i l i c a t e r a t i o , l e d t o c l e a r glasses
at f i r s t but with an i n c r e a s i n g tendency t o d e v i t r i f i c a t i o n u n t i l a t
25$ P20*; a powdery white mass was obtained which would not melt even
around 1500°C.
An i n t e r e s t i n g observation was t h a t during melting of
borderline glasses t h e proportion of crystal l i n e m a t e r i a l increased with
increasing temperature.
Adding phosphorus pentoxide t o an intermediate f r i t ,
approximating
t o the d i s i l i c a t e r a t i o , produced c l e a r glasses at somewhat higher P 2 0^
contents, but once again c r y s t a l l i z a t i o n l i m i t e d the amount which could
be introduced.
Additions of phosphorus pentoxide t o a m e t a s i l i c a t e f r i t
led t o t u r b i d opalescent glasses which were very hygroscopic.
To round out the survey, a few batches were melted with compositions
lying i n t h e high P 2 0rcorner of t h e t r i a n g l e .
Not even 5$ of s i l i c a
29
could be dissolved in a metaphosphate glass (70$ P205,30$ Na^O), but
7&th higher proportions of P20/j
out trouble.
as much as 15$ Si0 2 was dissolved with-
70$ P205 with 20 or 25$ of Si0 2 led t o clear melts, but
with tendency to formation of a surface scum.
these melts opalized.
"With ordinary cooling,
Introduction of Si0 2 notably increased the already
pronounced hygroscopicity of the sodium phosphate glasses.
I t i s well known to chemists that s i l i c a i s insoluble i n sodium
metaphosphate glass, for that i s the basis of the microcosmic s a l t bead
test for the presence of s i l i c a .
Representative points were now chosen within the glass-forming
field, and a group of specimens were prepared for measurements of
physical properties.
Table X .compares the calculated compositions of
these glasses with the values found by chemical analysis.
Table XI
brings out the thermal expansion coefficients and various temperatures
which are significant for the working and annealing of these glasses.
Table XII sums up the measurements of physical properties:
refractive
index, optical dispersion, penetration moduli, and relative chemical
durabilities.
The glasses which l i e on the borderline between homogeneity
and separations exhibit the roughening and brittleness which has been
described as characteristic for glasses opacified by calcium phosphate
(28). There i s a notable raising of the softening temperatures on substituting P20tj for soda or for s i l i c a , and t h i s also has been previously
ascribed to the calcium phosphate additions rather than recognized as
an attribute of phosphate i n i t s e l f (19).
29 a
Table X
Compositions of Na-jO, S i 0 2 , P 2 0c Glasses
GLass No.
Calculated Percentages.
Si02
NajjO
P2°5
Analyses
Si02
P
2°5
A1 2 0 3
P 31
80
20
—
80.9
—
0.3
P6
75
25
—
74.4
—
0.2
P 70
70 !
30
—
71.9
—
0.4
P13
65
35
—
66.2
—
1.8
P 74
60
40
—
63.5
1.4
P 52
55
45
—
58.4
1.3
P 48
15
20
5
74.2
5.1
3.5
P7
71
24
5
72.9
4*./
0.3
P 51
65
30
5
66.4
4.1
1.8
P15
62
33
5
65.2
5.5
0.6
P 32
55
40
5
57.0
4-9
1.3
P66
50
45
5
51.7
5.3
1.6
P 50
69
24
7
69.3
5.4
2.0
P 16
& \
X*
10
50.8
10.6
1.3
P53
55
35
10
56.3
9.5
2.2
P54
50
40
10
57.0
5.2
1.4
29 b
Table XI
Significant Temperatures f o r Na 2 0, S i 0 2 , P205 Glasses
( a l l i n °C.)
Thermal
Dilatometer
Expansion Coefficients
Curve Transitions
Room t o
Within
Upper
Lower
Standard
Lower AnAnnealing
Annealing
Annealing
nealinK Temp. Range
mass No. . Softening
520
485
8.9
622
514
468
12.0
24
590
499
430
15.1
33
P 74
480
413
16.3
43
P 52
470
398
17.6
39
—
P31
P6
—
P70
P13
P 43
730
—
478
9.2
P7
650
494
420
12.8
34
P 51
619
498
400
14.1
38
P15
614.
496
412
14.3
35
P 32
572
P 66
576
460
—
18.4
31
P 50
658
P16
622
—
420
19.9
—
P 53
625
495
412
17.0
26
P54
29 c
Table XII
Properties of N a ^ , Si0 2 , P20r Glasses
Relative Chemical Durability
1 - Commercial
2 - Slight attack
Glass No.
Refractive
Index ND
Relative Reciprocal Penetration 3 - Film Formatj
Dispersion
Modulus
4 - Hygroscopic
P 31
1.437
62
—
2
P6
1.497
59
336
3
P 70
1.499
57
322
3
P13
1.504
54
309
3
p" 74
1.507
—
246
+3
P 5?
1.513
—
3163
+3
P 48
1.482
58
400
2
P7
1.490
59
338
-2
P 51
1.495
58
325
3
P15
1.480
56
312
-2
P 32
1.507
—
225
+3
P 66
1.507
52
226
+3
P 50
1.484
57
356
2
P 16
1.498
57
269
-3
P 53
1.496
57
208
2
P54
1.494
50
167
3
30
Further evidence of the i n c o m p a t i b i l i t y of P205 and S i 0 2 i n t h e s e
compositions was found during t h e measurements of physical p r o p e r t i e s .
Even where P20t; a d d i t i o n s were not high enough t o cause d e v i t r i f i c a t i o n ,
the r e s u l t i n g g l a s s bore within i t s e l f a strong tendency t o
crystallize.
This tendency manifested i t s e l f during subsequent reheating at comparatively low temperatures.
"With the softening point measurements
instead of the u s u a l smooth hyperbola with a sharp t r a n s i t i o n p o i n t , t h e
phosphate-containing s i l i c a t e glasses gave i r r e g u l a r curves with a
choice of tangents p o s s i b l e .
Figure 3A and B shows two specimen d e t e r -
minations which i l l u s t r a t e the smooth and uneven types of softening
curves.
Likewise, with thermal expansion measurements, deviations from
the t y p i c a l curves for glasses were obtained because of the increased
tendency t o d e v i t r i f i c a t i o n i n the presence of phosphate.
Figure 4
contrasts t h e s t r a i g h t n e s s and sharp t r a n s i t i o n s of t h e curve f o r a
sodium s i l i c a t e g l a s s with t h e rounding of t h e curve for a borderline
phosphate-containing g l a s s .
All of t h e above evidences of a dual phase nature were most
pronounced i n t h e b o r d e r l i n e g l a s s e s , p a r t i c u l a r l y i n t h e high soda end
of the f i e l d .
At lower P 2 0* concentrations (7*here the phosphorus could
f i t i t s e l f i n t o t h e network) n i c e l y homogenous c h a r a c t e r i s t i c s were
obtained.
A thought arose "whether the l i m i t e d amount of compatibility
vshich was found between phosphate and s i l i c a t e i s due t o the small
amounts of alumina present i n a l l of these g l a s s e s .
Stanworth and
x1/)
""
so
1.C
•
i
»
.
iilijiiii iiiijIHi iiiiiiiii ! : ; ; i : : : : •nnnn H-HIH: ::: : i: : i: HHl'H:
::::
Iiii ;;;: liidlljl
llll'lll.'
iiii
iiii
Iii:
/IHIH: S d a r&jjk r P£3 isfli Serf*
Hi? iii':..!;:;, i'liil'
iii:
: :| ;
Hi:
iiii
•Hi
Hi: # ;
IIII
::::i •. I'lli'li:
i
''',' :::: iiij iiii
i: 11
ilii ;;;;
uji
;;;!
iii!
• t V, ; j •;
III:
»t-«
ill:
iii;
IIII
jlii
Ill:
Hii
111P
IIII ||f|
t
iiii III: H;i P ii'|
:il:
;;. ';::;
£
pj: ill!
.v3.t
Ilil;
II-1
HHiiH;
P
Jit!
Hjt
iiii
IMpl^iil^iliL
P
iii; 1
P.
;;!: HI:
:;;i i l l !
in;
iiiiliiii
P•iii
w
IIHliiii •>••
iiii in; ilii
iii;
H|| I:!;
:ili!;!i:
ill
:
P ' i)
...:!;;:.
iiii
: :: :
•
IppP .i'.iin;;
i i i . i:: H::!:
:;::£;;:: : : : 1 : : : :
;;::
::;i :::/.:::: .:: ! ;:: ..,,...,.
:::i
.:;!;::: iliijlili T-rri-Httr ii.i J ; ; ;
ill! yfij:;:; iil:j'.:i ;:•!::.:
iiiillil;
;; j ' t * - • •
:::: ::•:: :::: Hi:
•::: :::::::;: •.;r.i:::y.'•j/v.':\': H!liHr: iiHiiiii
^
liiirH- llllr:-H'
HHjHir rr:
~H"H"
%l.O
~:iilHT ~Wv
'::;; li;inil
^1
Illljlpl l l l l i l l l l
'ffi.Y-\ ;::;;;:;; i:i|:li: ; ; ; ; i : : : ;
-£
: :
• ;
: : : : ) • • : : :
pp-
PjiP
TUTTTTTT
1.04-rff- P
0 iiir
<9
IIII
|1
iiPili
/
Li~:
iiii
frH
•Hi
....
:::: I: :fi jp»i^~v^
'.'-.'•jilt*.
Iiii inij/i:
nil
..•:l:;.
111
iillp!
IpfHr
IB III! IB
;;:liiii P'P
z
3
77 Ar1G
it >7
H : |I ; ,:
;:J.:..i
:
; : • : ! : • :
:
:
•
•
:
:
:
:
•
:
•
-
.
-
.
-
•
: - . : ! : • : •
: ; ; : ! : : : •
Ppi
- • - . ; —
.'•:'•!:.•'•
• - : : ; ! ; • : •
111!
1|| iiliii;!!
;..H.:.; Hi-i-.H
;;i;i::
".:;:::.
. .;..
•*••::•"
•niiiili
.::.!...:
;;;; iin iiiiiii:: ::::;:::: :::.;:::: ::::!:...
—..;-...
•
!
~
H
!
H~"
H
|
"
~
Tn~:
. .::.:::
"'.''::.:
Iiii';:;;
.::;!:•:.
iiii iii;
-p-
:
rllllHr: 'iilii --
iiiiliiii.
^
dT
h iinutes
Ill'illl;
HHiHH iprlr
;::
ii';':
::l'!l::.
: 1:': I
::.H.H. —-;-:•• ....;._.
:*:."::
--HI-- :
ili'ii-ll
6
~"T~; T-":j"!i
lllli; : :; :'ni; : ':
7
Time in Minutes
Fig. 3.-
Typical
Scftenwg
Temperature^ Curves
A. Glass P7V~ Sodium Silicate
i
20 100
300
500
Temperature
700
300
m °C.
3. Glass P66~ Phosphate
v\
Containing
v.
X
/
[
.0
V) 0
\
•
<5 '
20 I0Q
300
500
700
300
Temperature ;* *C.
Fia Hr Typical Dilatometer
Curves
31
Turner (26) had emphasized the importance of even smaller amounts of
basic oxides on the formation of silicon phosphate glasses. Therefore,
one borderline composition: Si0 2 55,
Na 2 0 35, P20/j 10 was melted from
pure raw materials in a platinum crucible. No noticeable difference
in melting behavior from the corresponding batch melted in a kyanite
crucible was observed, and a good-looking transparent glass was obtained.
Its refractive index was 1.4-938 which compares with the value 1.4.986
found for the same batch melted in kyanite.
Bearing in mind the complicating factors of the alumina contents and the uncertainties caused by the strong unmixing tendencies,
certain deductions can still be made about the effects of P20/j additions
upon the properties of sodium silicate glasses.
Melting temperatures were raised somewhat by P^O^ additions,
but for the proportions used no notable changes were observed in the
fining or working characteristics of the glasses. As can be seen
from Table X, attack upon the aluminum silicate refractory seems to
be much more affected by higher melting temperatures than by P 2 0 5
substitutions. Possibly the lesser attack of the acidic P20£ than the
basic soda upon the silica of the refractory was more than compensated
by the greater affinity of P205 for alumina.
Sodium phosphate glasses have extremely low softening
temperatures. The metaphosphate was found to
soften at 344°C. and
an 80% P205, 20% Na20 glass at 400°C. Yet introduction of P205 into
sodium silicate compositions raises the softening temperatures rapidly
vdiether silica or soda is being replaced. The relative effects may be
Properfies
of A/a^qSiOzj /ZQf-
muFFfu a i«»rn c o „ N. Y. NO. 3 4 4 A
M
TRIANGULAR CO-OROINATC
M»DtL m u . S. *.
Gfosses
f^/q.
S~
Cor?Tfr?uect
too
KcurrcL a
EIIEK
CO..
N. Y.
NO.
TRIANOULAR CO-ORDINATE
HADE IN U. 8. A.
S44A
32
observed from F i g . 5A i n which softening temperatures are i n d i c a t e d f o r
the compositions found by chemical a n a l y s i s .
For convenience i n p l o t t i n g ,
the small contents of Al 2 0o have been added t o the Na 2 0.
From t h e
evidence a v a i l a b l e i t ' also appears t h a t the P20e shortens the temperature
interval from melting t o softening, but lovrers t h e transformation temperature so much t h a t t h e i n t e r v a l from the softening t o the transfoimation
temperature i s increased.
P20«j additions r a i s e the thermal, expansion coefficient
rapidly
vhen s u b s t i t u t e d f o r s i l i c a , but not q u i t e as much as a comparable i n crease i n Na20 does.
The r e f r a c t i v e index f a c t o r for Pn0,- has a lower value than
that for Na 2 0, and i s approximately equivalent t o t h a t for S i 0 2 , but
s l i g h t l y lower.
The values f o r p e n e t r a t i o n moduli b r i n g out i n a very i n t e r e s t i n g
way the s e n s i t i v i t y of t h i s measurement to changes i n composition.
Listed
below, as a b a s i s for comparisons, are t h e values found a t the U. S.
Bureau of Standards f o r liohs' scale minerals and for commercial o p t i c a l
glasses.
In the s o d a - s i l i c a s e r i e s of glasses i t can be seen t h a t the
moduli decrease continuously with increasing soda from a hardness of
about 5 for the t r i s i l i c a t e glass t o a hardness ea_ual t o t h a t of f l u o r i t e
at 42JS lla 2 0.
P 2 0r additions r a i s e t h e modulus inhether s u b s t i t u t i o n i s
for s i l i c a or f o r soda, e s p e c i a l l y a t t h e 5% P20^ l e v e l .
33
Penetration Moduli Values Found a t Bureau of Standards
Molis' Scale
H i n Kg. p e r mm.2
Sample
2
3
4
5
Gypsum
Calcite
Fluorite
Apatite
Parallel to axis
Perpendicular to axis
Albite
Orthoclase
Crystalline Quartz
Parallel to axis
Perpendicular to axis
Topaz
Artificial Corundum
Diamond
6
7
8
9
10
" Glasses
Fused Silica
Borosilicate crown
Ordinary crown
Dense flint
Very dense flint
32
135
163
360
430
490
560
710
790
1250
1635
8200
H
4-75
4-72
447
340
289
Soda, S i l i c a , T i t a n i a F i e l d
In Figure 6 are p l o t t e d t h e c a l c u l a t e d compositions for the
batches i n t h i s f i e l d which were melted a t various stages during the
study of t i t a n i a - c o n t a i n i n g g l a s s e s .
Points corresponding t o the
glasses prepared by Turner and Sheen and by Morey and Merwin are also
included.
Experiments with Melting Procedure — Production of glasses
from these t h r e e components i s complicated by t h e f a c t o r t h a t t h e r e
i s an upper as well as a lower temperature l i m i t v&thin which homogeneous
melts can be obtained.
This fact has not t o ray knowledge been pointed
out before for t h i s or other f i e l d s .
Some tendency of t h i s s o r t was
3
I
>c
1
I I
5
5 - ! fc
ai
0
o
*3 <0
<ff
.2 -
i : .
0 "
f
i p s
*
i.
*
x ®
*l B
34
noticed upon addition of phosphate to sodium silicate glasses, but vdth
titania the phenomenon is more marked.
The melting range of phosphate-
containing glasses is limited by their high viscosities and melting
ranges and by the rapid increase in the volatility of P20*j above 1300°C.
Titania additions to sodium silicate glasses, on the other hand, lower
the melting range and lead to extremely fluid melts. • There is a range
of several hundred degrees within which a soda, titania, silica melt
is fluid enough for complete reaction and fining. It was necessary to
decide what portion of this range should be used for the actual melting.
If titanium dioxide is mixed vdth a sodium silicate glass,
solution of the titania in the liquid is slow. If sand, soda ash, and
titanium dioxide are mixed, then heated rapidly, the first liquid will
form in the proportions given by the extremely low eutectic in the three
component field.
The mineral Ramsayite, Ka^O, 2 Ti0 2 , 2 Si0 2 (IB, 47,
35% by weight respectively) has been said to have a melting point at
about 624°C.:L and this compound probably forms a eutectic vdth the other
components which melts somewhere near this temperature. Because of the
fluidity of this liquid and the high density of titania, too high, a
temperature favors segregation.
This leads to crystallization and
coloration which are usually observed radiating from the bottom of the
crucible. Melting above a certain temperature level also favors the
yellow titania discoloration.
To bring out these effects of temperature of melting upon the
character of the glass, the following series of melts were prepared.
1
Niggli, P., "Das Magma and Seine Produkte," Part I., Akod. Verlag.,
Leipzig. 1937. page 8.
35
Six batches were carefully weighed to have the same calculated composition Si0 2 60, Na 2 0 20, Ti0 2 20. The batches were then melted separately
at temperatures ranging from 1150 to 1425 °C. To insure homogenization,
they were all stirred by bubbling air through the melt. After one hour
at the stated temperature, the melt was poured into a heated graphite
mold, then transferred to a muffle furnace, annealed at about 550°, and
allowed to cool in the furnace overnight.
After one hour at 1150° there was still a large proportion of
the titania which had not reacted.
The specimen had a yellow tint and
many bubbles. The glass was crushed and reheated at the same temperature
for two and one half hours in the same crucible. Only a trace of undissolved material now remained and the quality was much improved though
striae and fine bubbles still were present. However, the yellow tint
persisted.
The 1200° melt likewise was of poor quality after one hour's
melting, mainly because of large bubbles. It was crushed and remelted
for one and one-half hours. Now a water white specimen was obtained
vdth a negligible trace of yellow stain at one edge. A mass of very
fine seed remained, but the specimen is unusually free of striae.
At 1260° a similarly striae-free, but seedy glass was obtained
vdth one melting for one hour.
A weak yellow tint appeared. By 1315°
this yellow tint had deepened somevAiat, but the quality and refractive
index remained about the same.
The 1375° specimen was distinctly yellow,
and the bubbles were larger but less numerous. Striae began to appear
throughout the mass.
4
36
Melting at 1425° carried on this striae development to such an
extent that the first specimens were a maze of whorls. The glass contained relatively few bubbles, and peculiarly enough was less yellow
than even the 1315° melt. Upon remelting for two hours at the same
temperature, a very seedy specimen with a yellow cast was obtained.
Striae were still quite bad.
As a check on the color changes of these specimens ultraviolet
absorption spectra photographs were taken vdth a quartz spectrograph.
In Figure 7, the absorption spectra of the six glasses are. compared vdth
photographs of the arc taken through a 5% sector for the same exposure
tame. The wavelengths at wliich the transmissions of the successive
glasses were equivalent to 5% were measured as given in Table XIII.
These values indicate how slight a shift in the absorption will produce
visible coloration. Refractive index measurements are also included
in the table as a check on constancy of composition.
In the light of these experiences, the melting procedure selected
was to place the entire batch in an oversize crucible, then to raise
the temperature slowly and observe where fusing took place at an
appreciable rate. The furnace temperature was then held in this region
until melting was complete, stirring the liquid continually with air
bubbled in through a refractory tube.
Ordinarily the melt was held at
this temperature for two hours, but in many cases attack on the crucible
was so severe that pouring had to be hastened.
Limits of Glass Formation —
The field within vhichjclear
gLasses can be obtained was roughly delimited by melting batches spaced
36 a
Table XIII
Observations on A Titania Glass of Constant Composition Melted
at Varying Temperatures
Glass No.
Melting
Temperature
i n ° C.
Refractive
Index
N
D
Wave l e n g t h a t
Vihich U. V.Transmission
i s 5%
Millimicrons
Color
P 79
1150
1.6112
343
Yellow t i n t
P80
1200
1.6043
342
Colorless
P81
1250
1.6052
342
"Weak yellow t i n t
P 82
1315
1.6054
343
Yellow t i n t
P83
1375
1.6038
344|
Yellow
P 78
1425
1.602
345^
Yellow t i n t
37
to vary the proportions of the t h r e e oxides i n 10$ s t e p s .
On t h e high
silica s i d e , t h e l i m i t i n g f a c t o r was incomplete s o l u t i o n a t t h e temperature
used.
On t h e h i g h soda s i d e , i t was corrosion of t h e c r u c i b l e , and
possibly the f i e l d could be extended by meltings i n platinum.
On t h e
high t i t a n i a side t h e l i m i t i n g f a c t o r was immiscibility o r r a p i d d e vitrification.
The i m m i s c i b i l i t y , which was p a r t i c u l a r l y noted i n two 60$
Ti0 2 glasses with 20$ and 30)1 S i 0 2 , took the form of a watery t o p l a y e r
•which c r y s t a l l i z e d immediately during cooling and a pasty bottom l a y e r .
Melts low i n t i t a n i a were very seedy, and i t was hard t o find
the temperatures a t which r e f i n i n g would go on b e s t .
As t h e t i t a n i a
content i n c r e a s e d above 10$ and the batches became more f l u i d , bubbles
became l e s s and l e s s of a problem.
From Figure 6, i t can be seen t h a t the s e r i e s melted by Turner
and Sheen l i e s a t the boundary on t h e high s i l i c a side of t h e g l a s s fornring f i e l d .
Morey and Merwin's glasses are d i s t r i b u t e d through t h e
glass-forming region, i n d i c a t i n g t h a t t h e y have found about t h e same
l i m i t s as t h e w r i t e r .
P r o p e r t i e s of Titania-Containing Glasses — Data on physical
properties were obtained only on a l i m i t e d number of g l a s s e s .
In t h e
absence of chemical analyses only a few generalizations can be made
about the e f f e c t s of s u b s t i t u t i o n s of t i t a n i a f o r s i l i c a or f o r soda.
Inspection of Table XIV b r i n g s out s t r i k i n g l y t h e pronounced
effects of t i t a n i a on r e f r a c t i v e index and dispersion.
The p e n e t r a t i o n
moduli are increased t o values which are comparable with commercial
Table XIV
Soda, Silica, Titania Glasses
Best
Melting
Temperature
Refractive Relative
Index
Reciprocal Penetration
Dispersion
Modulus
%
Di
1300°C.
1.528
51
397
3
Glass No.
Calculated Percentages
Na20
Ti02
Si02
P 61
65
30
5
P 62
60
35
5
• 1300
1.526
50
402
3
P86
70
20
10
1300
1.536
51
424
+2
p 60
65
25
10
1275
1.55
48
380
+2
PS5
60
30
10
1250
1.543
48
3r#
3
P84
50
40
10
1200
1.553
46
336
3
P 64
60
25
15
1300
1.577
42
473
P 80
60
20
20
1200
1.605
39
470
P87
50
30
20
1150
1.59
—
454
P 91
40
30
30
900
1.677
37
395
3
P 96
40
20
40
1150
1.716
—
466
1
P 95
30
30
40
900
1.71
—
364
-2
P99
20
30
50
1025
—
—
375
2
.
Chemical
2
FIGURE 7
ULTRAVIOLET ABSORPTION SPECTRA
of
A 20$ T i 0 2 GLASS BATCH
MELTED AT INDICATED TEMPERATURES
soda-lime glasses (450), but the chemical durability is'not. To make
this type of glass practically usable additional constituents would
have to
be added for improving weathering ability.
Dilatometer curves were obtained on the glasses listed below.
Glass No.
oC x 10°
Transformation Temperature
P85
13.0
478
F80
10.3
520
P91
15.2
458
P95
15.4
436
These results indicate that alkali content is still the
dominant factor in determining thermal expansion.
As in Nagai and
Masuda's series where titania was substituted for silica in a
soda-lime glass, titania has a higher expansion factor than silica.
Thus, in the series P31, P35, P91, where Ti0 2 is substituted for Si0 2
in 20$ steps, ot x 10^ changes from 8.9 to 13.0 to 15.2.
Above the transformation temperature the expansion of titaniacontaining glasses rises very steeply as would be expected from the
fluidity of titania melts. With P92, for example, the <x x 10 6 for
the transformation interval is 82.
39
Color Changes in Na 2 0, Si0 2 , P20^ Glasses
Tfeyl, Pincus and Badger1 have pointed out the intensification
of coloration by vanadium in sodium phosphate glasses as compared with
sodium silicate glasses. It was thought that it would be of interest
to make a similar comparison for other coloring ions.
To this end suitable amounts of the following colorants were
melted vdth sodium phosphate or sodium silicate frits:
Colorant
Source
Parts per 100 frit
Chromium
Cr203
1
Manganese
MnO
1/2
Iron
Fe
2
Nickel
2°3
Ni 3 0 4
0.2
Cobalt
G03°4
0.1
Uranium
U0 3
2
It must be remembered that melting conditions may have as
much effect as chemical composition in determining the color obtained.
As far as possible, these conditions were kept uniform by melting in
the same electric furnace, for the same times, at temperatures which
gave comparable conditions. Introduction of oxidizing or reducing
agents through the raw materials was avoided by adding the colorants
as the oxides.
The results are summarized in Table XV. The familiar dark
green of chromium does not change. Manganese goes from the deep
1
J. Amer. Ceramic Soc. 22. 374-7 (1939). "Vanadium as a Glass
Colorant."
Table XV
Comparison of Colors in Silicate and Phosphate Glasses
Base Glass No.
and Type
Chromium
P6 - trisilicate
P13 - disilicate
Iron
Cobalt
Nickel
Uranium
Dark green Purple
Yellowish
green
Dark blue
Brown
Yellow color
and fluorescence
Dark green Purple
Greenish
yellow
Dark blue
Brown
Deeper yellow
Pinkish
blue
Grayish
yellow
Pale green
yellow fluorescence
Similar
Greenish
yellow
Green no
fluorescence
Manganese
P 35 - diphosphate Dark green Light purple
Pink fluorescence
P 36-metaphosphate Dark green
Blue t i n t
Pink fluorescence
Pale brown
Flesh
vO
P3
40
purple shade in the sodium silicate glasses to a barely visible purple
in the metaphosphate glass and a very faint blue in the 80$ P20r glass.
Meanwhile fluorescence appears. Iron changes from the familiar yellow
green to brown tints, and cobalt from blue to a pinkish hue. Nickel
is converted from dark brown to yellow tints, and uranium exhibits a
striking transition from fluorescent yellow to a dark green resembling
the chrome green and devoid of fluorescence.
The next step was to observe the effects upon the colors of
minor additions of P20r to the sodium silicate glasses or of Si0 2 to
the sodium phosphate glasses. In glass P 7, sodium trisilicate plus
5$ P20c, no changes took place except for a brightening of the uranium
color. Similarly vdth glasses P 15 and P 16, which approximate sodium
disilicate plus 5 and 10$ P205 respectively, only the uranium color
varied.
More significant effects were obtained vdth sodium phosphate
glasses into which silica had been introduced.
Three of these glasses
were used:
Glass No.
P205
Calculated Composition
Na20
Si0 2
P 25
80
10
10
P 24
70
15
15
P 2
70
10
20
In a l l t h r e e of t h e s e the manganese i s n e a r l y c o l o r l e s s .
In P 24 t h e blue t i n t i s completely absent, and the b e s t pink f l u o r escence was obtained.
The brovai t i n t of t h e i r o n was a t a minimum
41
in P 2.
Cobalt was bluer in P 25, about the same pink i n P 24 as i n
p 35, and pink or black i n P 2 as immiscibility developed.
Nickel's
color became a brighter yellow i n P 25 and s t i l l more i n P 24, but i n
p 2 i t varied from yellow to black with the immiscibility development.
Uranium i s the same green i n P 25 as in P 35, but lightens to a yellower
green in P 24 and P 2.
DISCUSSION
An attempt w i l l be made t o explain the experimental f a c t s
in terms of the atomic p i c t u r e of g l a s s , t r e a t i n g the constituents as
ions.
The p i c t u r e must be obviously simplified, f o r i t neglects such
factors as non-ionic bonds, p o l a r i z a t i o n , secondary f o r c e s , and t h e
variation of the p r o p e r t i e s of an ion vdth i t s surroundings.
Vfliile no
detailed references w i l l be given during the p r e s e n t a t i o n , i t i s t o be
understood t h a t the general background i s based on t h e work of Warren
and h i s school wliich goes back t o t h e writings of Goldschmidt (57)
and of Zachariasen (66).
The l a s t two from frankly t h e o r e t i c a l reasoning, though based
on a background of sound f a c t s from c r y s t a l s t r u c t u r e s , derived r u l e s
for the formation of g l a s s e s .
Using these r u l e s and c r y s t a l s t r u c t u r e
data about the oxygen polyhedra formed around t h e various cations, i t
was found t h a t t h e following cations might be expected t o form oxide
glasses:
t r i v a l e n t B, P, As, and Sb i n decreasing order of a b i l i t y
to form g l a s s j t e t r a v a l e n t Si and Gej pentavalent P, As, V, and
possibly Sb, Cb, and Ta,
An oxide glass of the general type A ^ O
must contain as
framework appreciable amounts of the above cations o r of other cations
vihich can replace them isomorphously.
Zachariasen thought t h a t the
l a s t q u a l i f i c a t i o n added only Al + + + for S i + + + + and s t a t e d t h a t highly
charged cations l i k e T i + + + + would tend t o produce d e v i t r i f i c a t i o n .
glass v d l l have the most advantageous p r o p e r t i e s i f the value of n
A
43
(the number of glass-forming cations per oxygen) l i e s between 0.33 and
0.50 for t e t r a l i e d r a l and around O.67 for a t r i a n g u l a r network.
The
additional c a t i o n s , A, which were assumed t o occupy the holes i n the
network, must have small charges and l a r g e r a d i i so t h a t t h e repulsive
forces between them and t h e B cations are low.
Warren's Contributions — By b r i l l i a n t experimental and mathematical methods, Warren has since produced confirmation of t h i s p i c t u r e
and gone on t o explain additional phenomena of glasses (52, 53, 54,
62, 63, 64).
Warren's work has accomplished two ends.
First, inter-
atomic d i s t a n c e s and coordination numbers have been d i r e c t l y determined
for c e r t a i n simple glass s t r u c t u r e s .
Second, these distances and co-
ordinations have been shown t o be so close t o those of the corresponding
c r y s t a l l i z e d compositions t h a t a l l of the data of c r y s t a l chemistry has
become a v a i l a b l e for extending the scope of the s t r u c t u r a l chemistry
of g l a s s .
This l a s t i s p a r t i c u l a r l y fortunate because Warren's method
i s very l i m i t e d i n i t s a p p l i c a t i o n .
So f a r a maximum of t h r e e kinds
of atoms i s the most t h a t can be handled, the atoms must be of low
atomic number so t h a t fogging of t h e film due t o fluorescence produced
by the x-rays w i l l not be a f a c t o r , and t h e r e must not be too great
a difference i n i o n i c r a d i i between t h e two cations or the s c a t t e r i n g
power of one w i l l mask the effects of the other.
Nevertheless, the potency of t h i s new approach t o the s t r u c t u r e
of glass i s apparent from the number of reasonable explanations which
i t s a p p l i c a t i o n has achieved of f a c t s which previously could not be
(,
44
sufficiently explained.
Before talcing up the applications of these
concepts t o t h e present f i n d i n g s , i t w i l l be i n t e r e s t i n g t o consider
some cases where t h e y have been applied t o simpler systems.
Primary Glass Forming Oxides — Because i t s valence and coordination number are both four, the s i l i c o n i n s i l i c a glass surrounds
i t s e l f firmly vdth four oxygens a t t e t r a h e d r a l p o s i t i o n s .
In Pauling's
terminology t h e e l e c t r o s t a t i c bond strength Si-0 valence/coordination
= 4/4 ° 1 (60).
Each oxygen vdth i t s valence of two can then a t t a i n
the two p o s i t i v e ion bonds which i t r e q u i r e s .
A continuous three
dimensional s t r u c t u r e r e s u l t s wliich i s hard, strong, r e s i s t a n t t o
attack by water, and softens at high temperatures t o a very viscous
liquid.
Boron likewise forms a s t a b l e oxide glass because i t s coordination
number and valence are i d e n t i c a l .
again equals u n i t y .
The e l e c t r o s t a t i c bond s t r e n g t h 3/3
But since the coordination value i s only t h r e e ,
the oxygons a l l l i e i n one plane around the boron, and the glass i s
more s o l u b l e , s o f t e r , and melts at a lower temperature than s i l i c a
glass.
The l i q u i d i s q u i t e viscous, however, aid has a f l a t t e r
viscosity-temperature r e l a t i o n s h i p than s i l i c a .
Phosphorus v d t h a valence of five lias a coordination of four
oxygens.
The excess oxygen makes unnecessary a continuous oxygen-
phosphorus network, and -the r e s u l t i n g glass has t h e high s o l u b i l i t y ,
low melting temperature, and other p r o p e r t i e s which such
favors.
a structure
The observed f a c t t h a t PJJOJ during heating undergoes several
a l l o t r o p i c changes wliich lead t o decreased vapor pressure and higher
45
viscosity and t h a t i t forms a glass of lovrer s o l u b i l i t y a f t e r p r o t r a c t e d
heating a t red heat have not been explained by the simple atomic p i c t u r e
and some form of polymerization i s u s u a l l y p o s t u l a t e d .
I t i s i n t e r e s t i n g to note t h a t B203 and Si0 2 are poor s o l v e n t s .
Their continuously l i n k e d s t r u c t u r e makes i t d i f f i c u l t to introduce
metallic oxides, such as c o b a l t , n i c k e l , and chromium.
P 2 0r, on the
other hand, i s an excellent solvent because of t h e r e a c t i v i t y of t h e
unsatisfied bonds.
Boric oxide and s i l i c a are miscible i n a l l proportions to form
glasses because t h e oxygens can s a t i s f y t h e i r valence requirements with
either boron or s i l i c o n without i n t e r r u p t i n g the continuous network.
The consequence of adding B20o t o a s i l i c a glass i s t o weaken the
structure p r o g r e s s i v e l y by t h e proportion of t r i a n g u l a r planar groups
which replace t e t r a h e d r a l groups.
Not enough evidence e x i s t s t o allow p i c t u r i z a t i o n of what occurs
when P205 r e a c t s vdth S i 0 2 or B 2 0o.
Reactions betvreen t h e acids l e a d s •
t o s i l i c o n phosphate and boron phosphate.
That i s the phosphoric acid
i s so strong t h a t i t causes the others t o act as bases.
Corresponding
behavior vrould probably occur i n the higher temperature r e a c t i o n s betvreen
the oxides.
Binary Glasses — Adding non-glass forming oxides such as Na20
and CaO t o one of these t h r e o primary g l a s s formers also r e s u l t s i n a
competition f o r oxygen bonds i n t h e melt with the competitive a b i l i t y
l a r g e l y determined by t h e i r "ionic p o t e n t i a l , " t h a t i s t h e cliarge on t h e
ion over i t s r a d i u s .
Values of i o n i c r a d i i and p o t e n t i a l s for ions
46
commonly used in glasses are listed in Table XVI. While the application
of this concept of ionic potential to glasses is quite new, having been
first used this year by Warren and Pincus (64) in discussing miscibility
and immiscibility in certain fundamental ceramic systems, it had previously been used by crystal chemists to explain such fundamental problems
as acidity and basicity.
The reaction betvreen alkali oxides and silica occurs by breaking
of a silicon-oxygen bond and the attaching of this silicon to an oxygen
from an alkali instead of a re-forming of the original bond.
The alkali
ion does not enter into the tetrahedral framework but is loosely held
in the open spaces which the network structure brings about. Thus, Na +
ion is usually found with about six oxygen neighbors. Meanwhile the
addition of extra oxygen to the silicon-oxygen network has produced a
significant change in its characteristics. Instead of a strong, continuous
structure vdth every oxygen linked to two silicons, there now exists
for each alkali ion introduced one oxygen which is bonded to only one
silicon.
This kind of oxygen is of great importance in this theory of
glass. A sharp distinction is made betvreen the oxygens which are bonded
to two silicons SiO Si, and those which are bonded to only one SiO-.
Warren has always referred to these as double bonded and single bonded
oxygens, but as the term single bonded has a different significance to
chemists it is avoided here.
Because of the presence of the SiO- type
of oxygen the melt is a better solvent and the glass has a "weakened"
46' a
Table XVI
Sizes and Potentials of Cations Used in Glasses
Cation
Li +
Na +
Ionic Radius
Valence/Radius
or
"Ionic P o t e n t i a l "
Eb+
Cs +
0.60
0.95
1.33
1.48
I.69
1.7
1.1
0.75
0.68
0.59
Be++
Mg + +
Ca*+
Sr*+
Ba.++
0.31
O.65
0.99
1.13
1.35
6.5
3.1
2.0
1.8
1.5
Zn++
Cd ++
Pb + +
0.74
0.97
1.21
2.7
2.1
1.7
B+++
A1+++
0.20 •
0.50
0.6Q
0.90?
Sb+++
Si++++
15.0
6.0
4.3
3.3
0.41
0.68
0.80
1.01
9.8
5.9
5.0
4.0
Ge++++
Sn++++
0.53
0.71
0.84
7.5
5.6
4.8
p+++++
As + + + + +
0.34
0.Z.7
0.62
14.7
10.6
8.1
Ce++++
47
structure. By the time enough P^O has been added to reach the metasilicate proportions, R 2 0 • Si0 2 , the silicon-oxygen ratio has become
1 : 3, and out of every four oxygens two can be bonded to only one
silicon. Sodium metasilicate crystallizes easily, has a low softening
point, and in other ways gives evidence of a "weak structure" vdth some
of the characteristics of P20r.
The, Problem of Immiscibility —
Although the SiO- oxygens have
been used to explain certain properties of an alkali silicate glass, the
alkali ions must be considered also. They are trying to satisfy their
valency requirements by surrounding themselves vdth a sufficient number
of single bonded oxygens. At low soda contents there are not enough
single bonded oxygens to provide satisfactory surroundings for all of
the sodium ions.
A certain tendency to separation into two liquids,
one higher in alkali content, therefore exists in alkali silicate
melts. From the liquidus curves as summarized by Kracek (59), Warren
and Pincus concluded that the tendency to separation into two phases
increases vdth increasing ionic potential from caesium (0.59) to lithium
(1.7) (See Table XVI), but that the separating forces are not strong
enough to overcome the simultaneous tendency of the silicon to bond to
all the oxygens which are available and keep them in a single phase.
Their picture neglected the electrostatic repulsion forces between
an S i + + + + and the neighboring S i + + + + ions which would have to be considered in a more detailed analysis.
With barium ions the ionic potential is likewise not strong
enough to bring about immiscibility, although the S shape of the BaOSi0 2 liquidus curve indicates that a very strong tendency to separation
48
does exist. The other alkaline earth ions, however, possess ionic
potential values at which they can make very definite demands for oxygen
bonds, that is for a definite place in the structure of the liquid.
As
a result the melt must separate into two liquids: one nearly pure Si0 2
and the other vdth a definite R0-Si02 composition which satisfies the
geometry of the configuration of the oxygens around R ions.
By
approximate calculations, Warren and Pincus found that this structure
is probably (1) two Ca + + ions to each SiO-oxygen and (2) several SiOoxygens around each Ca ++ , besides enough SiOSi oxygens to make up the
requisite coordination number of six to eight.
With more RO present than in this end member, enough SiOoxygens are present to satisfy both the silicon and the R ions within
the structure of a single liquid phase. But the large number of such
oxygens (two per R ion) leads to very fluid melts -which crystallize
relatively easily.
When reactions between boric oxide and alkali or alkaline earth
occur, the interpretation must be altered because of the fact that the
boron uses the oxygen supplied by the reactant to change over from three
to four-fold coordination as much as possible. This leads to a strengthened
structure up to a certain point and explains the maxima observed in
composition vs. physical property plots of boron containing glasses. In
the cases of alkaline earth-boric oxide systems an immiscibility gap
exists similar in extent to those vdth silica. Beyond the two liquid
region, melts with strong tendencies to crystallize are obtained, although
not so strong as with silicates.
49
Binary reactions with P20r present a much different picture
than those with Si0 2 or B 2 0^, due once again to the already existing
incompletely satisfied oxygen bonds. Bivalent oxides react with P 2 0 5
to form single liquid phases over the whole range. They are much more
easily cooled to glasses than the corresponding silicate or borate
melts (54).
Ternary Glasses —
M
Since the interaction of alkali vdth silica
or vdth boric oxide leads to SiO- oxygens, the resultant melts have
much better solvent power than the primary oxide. In many ways the
reactions of a high alkali silicate resemble those of P20c. But a
remarkably small addition of alkali oxide suffices to change the
characteristics of silica. Thus, less than one per cent of soda will
make lime and silica miscible in all proportions.
Warren ands Pincus were unable to present any quantitative
explanation for this effect, but pointed out that even the small amount
of SiO- oxygens introduced by the one per cent of Na 2 0 may change the
energy picture enough to permit satisfaction of calcium's oxygen needs
within a single liquid.
The role of alumina in ternary silicate glass structures has
not yet been established.
It is knovai from the crystalline silicates
that aluminum can occupy either a sixfold or fourfold oxygen coordination
position.
In the fourfold coordination it substitutes for silicon.
The valence relations are balanced by the simultaneous introduction of
an alkali ion, which occupies holes in the framework.
behavior can be postulated for the aluminum in glasses.
Similar dual
50
Soda. Silicaf Titania Melts —
The results for this thesis show
that in the presence of the SiO- oxygens introduced by the alkali, large
proportions of titanium can find a position in the structure of a silicate
melt.
This position is stable enough to lead to homogenous liquids
and to sufficient viscosity around the liquidus so that it is easy to
obtain glasses on normal cooling.
Since it is not possible to determine
the atomic arrangements in these glasses directly by x-ray methods, it
is necessary to hypothesize about what these arrangements may be.
From analyses of crystals containing titanium, it is known
that T i + + + + can have a coordination number of either four or six oxygens.
The radius of the T i + + + + ion (0.68A) is much larger than that of S i + + + +
(0.41) and isomorphous substitution of titanium for silicon does not
appear likely.
In fact, 7J". L. Bragg stresses this in his review of
the investigation of the "Atomic Structure of Minerals" by means of
x-rays (55).
On page 96 may be found the statement:
"Six-coordination has been found for titanium in every case
where a structure has been fully worked out, and in many silicates
the evidence is strong that titanium replaces the cations aluminum
and magnesium in six-coordination, and never silicon in fourcoordination."
However, N. W. Taylor has assigned a coordination number of
four to titanium in certain spinel structures (61),
The picture derived from the present research is that titanium
enters into the structure of the soda, titania, silica glasses by going
into fourfold coordination but it can be forced out of the continuously
Liiked network and become a "hole-filler" vdth a coordination of about
six oxygens.
51
In spite of the large amounts of Ti0 2 which could be introduced
into these glasses, it is not essential to require that the titanium
be a part of the glass forming network. For even the glass with the
highest titania content (50$)has an oxide formula Nao.,46 Ti 0.44
si
0.33
0 which indicates enough Si atoms to provide a continuous framework.
0.33 is the lower limit in Zachariasen's AaBnO formula (42). But the
characteristics of the glass suggest that most of the titaniums are in
the framework rather than in the six-coordination positions. Such
evidence is as follows.
In the four-coordination, the titanium could surrender its
four valence electrons completely, one to each of the four oxygens, and
thus reach the rare gas configuration.
It has been pointed out that
the transition elements are colorless when stripped to the rare gas
state (56).
Upon impoverishment of the melt in oxygen, the titanium begins
to drop in valence, and probably will change its position in the glass
structure. There is an intermediate state before the titanium is
reduced to the trivalent form, for the latter's purple color does not
appear except under very strong reducing conditions.
Instead, the
electrons which are not completely removed by single bonded oxygens
vibrate between the titaiium and its six oxygen neighbors. The
frequency of these vibrations are such that the absorption starts at the
ultraviolet edge of the base glass and extdnds continuously towards the
visible.
As the impoverishment in oxygen proceeds, the number of
electrons thus vibrating increases and the distances over which they can
52
move increase in variety.
Consequently, the number of frequencies at
vMch absorption can take place increases until they include some in
the visible violet and the yellow color appears and deepens. Analogous
behavior is shown by the colored KMhO/, RjCrO^ and KVO3.
Weyl (65) has pointed out similar changes in absorption spectra
with changes in coordination number for cobalt and nickel. The color
consequences were observed in Table XV.
In the higher coordination
possible in phosphate glasses as compared with silicates, cobalt changes
from an intense blue color to a pale pink, and nickel from a dark
purple to a weak yellow, i. e., in the direction of less color in each
case.
The ease of breaking of the fourfold oxygen to titanium bond
is shovoi by the slight reductionnecessary to cause the yellow discoloration.
It further accords vdth the development of the yellow at higher temperatures,
and explains the striae formation vdth increasing temperature of melting.
In this respect, six-fold coordinated titanium (as a diole-filler) behaves
like Ca + in tending to promote a separate liquid phase.
The instability of the four-fold Ti-0 bond also accounts for the
low melting temperatures and high fluidities of glasses in this field.
Apparently the bonds are breaking rapidly and to a large proportion in
the melts, but re-form rapidly enough during cooling so that the viscosity
reaches a suitable level before crystallization can set it.
The strong electric fields around tlds titanium coafiguration
probably explain the abnormally high refractive index factor of titanium
in relation to its atomic weight.
53
The quoted value of t h e i o n i c r a d i u s of T i + + + + i s based on
minerals i n which i t coordinates s i x oxygens.
In t h e four-coordination
where t h e valence e l e c t r o n s h e l l s are more completely s t r i p p e d away,
t h e r a d i u s vrould be somewhat l e s s , t h a t i s more n e a r l y approacliing t h a t of
Si
and t h e s u b s t i t u t i o n becomes more p l a u s i b l e .
The i o n i c p o t e n t i a l
value for T i + + + + would then be g r e a t e r t h a n t h a t i n Table XVI, but
still
enough smaller than t h a t for S i + + + + so t h a t t h e t i t a n i u m could contest
s e r i o u s l y vdth t h e s i l i c o n f o r t h e oxygen bonds.
0.0.0.0.0.
Soda. Silica, Phosphorus Pentoxide Melts — With P
a
different set of reactions develops because its ionic potential (14.7)
is appreciably greater than that of S i + + + + (9.8).
(PO,)
is generally
considered to be an acid radical in which the bonding is covalent, while
silicon is considered to be in an intermediate position and can be regarded
either as the center of an (SiO/)
acid radical or as a very.small
cation coordinating four oxygens (55, page 34).
The consequences of the
high potential of P + + + + + are that it can take away from the silicon
the oxygen donated by the sodium.
The sodium ion will migrate to the
P0» region because of the abundance of oxygen bonds available there.
At lower temperatures these relationships lead to a strengthened
structure and a higher viscosity.
But at higher temperatures and also
at higher P 2 0 5 concentrations the p + + + + + begins to go after the silicon's
own oxygens as well, vdth the result that unmixing is promoted.
54
Compatibility w i t h i n an oxide melt i s then not j u s t a matter of
valence, or of whether i t i s "framework formers" or " h o l e - f i l l e r s "
which are being blended, or even of balance betvreen acid and b a s i c oxides,
but i s determined by the r e l a t i v e i o n i c p o t e n t i a l s of t h e c a t i o n s v d t h
respect t o t h e oxygens and t o the s t r u c t u r e s and linkages t o which they
lead.
55
SUMMARY
The limits of proportions of glass formation under ordinary
melting conditions have been determined for two triaxial fields based
upon the soda-silica join.
Certain physical properties have been
measured on representative glasses.
One field (soda, silica, titania) has been previously investigated only at high silica contents. The outlines of the rest of the field
were roughly implied, but this investigation more definitely brings out
the limiting factors and compositions.
The other field (soda, silica,
phosphorus pentoxide) has not previously been described, except for four
melts in the high P20r region.
A remarkable phenomenon was discovered in the melting behavior
of batches in both of these fields. An upper as well as a lower
temperature limit definitely exists vdtliin which melting is best carried
on. Vdth the P 2 0 5 melts too high a temperature favored crystallization
of a refractory phase. With the Ti0 2 melts it led to discoloration,
striation, and eventually devitrification.
Ti0 2 forms glasses within the triaxial field over a surprisingly
large range and up to 50$ Ti02.
P205 additions to sodium silicate
glasses appear to be definitely limited, reach 10$ at the maximum, and
the glasses which are obtained were almost all found to possess strong
tendencies to devitrify.
The P205 series of glasses were all very soluble. It was found
that Po0r additions raised the softening temperatures and increased the
56
brittleness, facts which have been known about phosphate opal glasses,
but vihich. had been ascribed to the calcium phosphate rather than to the
P20c as such.
The TiO^ series glasses were remarkable for their high fluidities,
freedom from yellow discoloration when melted at low enough temperatures,
and high refractive indices and dispersions. Their chemical durabilities
improved proportionately vdth the titania content.
The experimental findings have been discussed from the viewpoint
of structural chemistry and explained on the basis of relative "ionic
potentials" of the cations. It is postulated that titanium occupies a
fourfold coordination position in the glassy network.
Cn impoverishment
of the melt in oxygen by too high temperatures, reduction, or too low an
allcali content, the titanium is forced to change to a six-fold coordination
position in the holes of the framework. Discoloration and increased
tendency to devitrify are the results. The incompatibility of P20^
and Si0 2 in these glasses is ascribed to the high ionic potential
(acidity) of the p + + + + + enabling it to take the alkali-donated oxygen
away from the silicon and also to disturb the other silicon-oxygen bonds.
BIBLIOGRAPHY
A.
Glasses Containing Phosphate
Berger, E . , (Schott), "Heat Absorbing Glass," U . S . I , 961, 603
(June 5, 1934).
Berger, E . , (Schott), "Heat Absorbing Glass," U. S. 2, 194, 784
(Mar. 26, 1940).
Buck, E. C , "Vitreous Composition and A r t i c l e s Made Therefrom,"
U. S. 1 , 570, 202 (Jan. 19, 1926).
Cedivoda, F . , "The Phosphate Glasses," Chen. - Ztg. 2£, 347
(1901), J . Soc. Glass Tech. 2 , 122-3A (191B).
F r i t z , Helene, "Phosphoric Acid Glasses and Glazes," Ker. Rund. 26,
195, 247 (1918) J . Soc. Glass Tech. 2, 124A (1918)G8nsenmuller,T;yHistory of Phosphate Glasses," Glaslmtte 68, 873-5
(1938).
Grimm, H. G., and Paul Huppert ( I . G. Farberdndustrie), "Vitreous
Compositions o r Glasses," "u. S. 1, 964, 629 (6/26/34), U. S. 2,
100, 391 (11/30/37).
Grimm, II. G., and P. Huppert ( I . G. Farbeidndustrie), "Phosphate
Glasses f o r P a r t s Exposed t o Vaporous Allcali or Other Light-Letal
L i q u i d s , " Gor. 654, 925 (Aug. 17, 1935).
Hood, II. P . , "Substance Transparent t o U l t r a v i o l e t Light and Method
of Producing the Same," U. S. 1, 830, 903 (Nov. 10, 1931).
Hood, H. P . , "A New Ultra-Violet Transmitting Glass," Science 64,
281-2 (1926), Glass Ind. 7, 287-8 (1926).
Hovestadt, II.,"Jena Glass and I t s S c i e n t i f i c and I n d u s t r i a l
Applications," LacUillan, London, 1902 (Translated by J . D. aid
Alice E v e r e t t ) .
Iluttner, K., "On the Action of Phosphoric Acid on S i l i c a and
S i l i c a t e Glasses," Z. anorg. Cher/.. g9_, 216-24 (1908), Chen. Abs.
2 , 2917 (1908).
Farberdndustrie, I . G., "Laproveucntc i n Phosphate Glass," B r i t .
498, 049 (Jan. 3, 1939).
14.
Inwald, 0 . , ^Studies on the Conposition and Properties of Phospliate
Glasses," Inaug. D i s s e r t . Berlin 1900, Chen. Ztg. 2£, IO58 (1901)
J . Soc. Glass Tech. 2, 125 (1912).
15.
Kaufraann, Yf. and E. Bungartz, "Glass Permeable to Ultraviolet."Rays,"
U. S. 2, 031, 958 (Feb. 25, 1936).
16.
Kaufmann, TT., L. Eckstein, and K. Rosenberger ( I . G. Farberdndustrie)
"Photoluminescent Glass," U. S. 2, O42, 425 (May 26, 1935).
17.
Khapp, 0 . , "Phosphate Glasses," Glaslmtte 68, 794-7 (1938), Chem.
Abs. 33, 5613 (1939).
18.
Knapp, 0 . , "Phosphate Glasses," Glass 1£, 487-90 (1938) Die Glashutte
68, 778 (1938), chem. Abs. 21i 1896 (1939).
19.
Kreidl, II. J . , and Yf. A. Tfeyl, "Opal Glass," Glass Ind. 21, 57 (1940).
20.
Long, B . , "Glass and Method of Producing I t . " U. S. 1, 749, 823
(Mar. 1 1 , 1930).
21.
Llaclntire, W. H., L. J . Hardin, and F. E. Oldham, "Calcium Lletaphospliatc F e r t i l i z e r s . Chenical Composition ond P r o p e r t i e s , "
Ind. Eng. Chem. 29, 224-34 (1937).
22.
Nagai, S. and K. Talcahasld, "Chemical Conposition and T.rater S o l u b i l i t y
of Glasses. I l l " , J . Japan Ceran. Assoc. 44, 143-53 (1936).
23.
P a r t r i d g e , E. P . , "The Chemistry of the Metaphosphates," Dual Service
Hows, June 1937.
24.
P a r t r i d g e , J . H., "Resistant Glasses for Modem E l e c t r i c Discliarge
Lamps, I . Glasses for High Pressure Mercury Discharge Lamps,
I I . Glasses, Resistant t o Sodiuu Vapor," J . Soc. Glass Tech. 19,
266-74, 274-8 (1935).
25.
Cclmidt, R., "ilevrer Developments i n Spccirl Glasses for the Field
of Lighting Technology," Glastech. Ber. 15 , 89-99 (1937).
26.
Stanworth, J . E. and Tf. E. S. Turner, "Kotos en the P o s s i b i l i t y of
' Preparing Glasses Containing P o 0 r , Al o 0^, BoO,, end SiOo." J . Soc.
Glass Tech. 2 1 , 3&S2 (1937) *~ J
" '
" '
27.
Stanworth, J . E . , "Glass Manufacture for E l e c t r i c Laups. The Development of Special Glasses for Mercury Discharge Lar..ps." J . Soc. Glass
Tech. 23, 268~£0 (1939).
28.
Zschiixier, E . , K. Hesse and L. Stoess, "Roujheniri- of Calcium Phosphate Glass for Lighting Purposes.," Sprechsaal 5C_, ; i 3 - 7 , 529-34
"(1925).
B.
Glasses Containing Titania
29.
Aldinger, R., "Titanium Dioxide i n the Enamel Industry," Glashutte
68, 290-1 (1938).
30.
Badger, A. S . , and L. M. Doney, "Glass Melts i n the Systems ZirconiaAlumina-Silica and Titania-Alumina-Silica," Glass Ind. 2 1 , 309
(1940).
31.
Berger, E . , (Schott), "Improvenents i n Optical Glasses," U. S. 1,
943, 051 (jan. 9, 1934).
32.
Bunting, E. N., "Phase E q u i l i b r i a i n the Systems Ti0 2 , T i 0 2 - S i 0 2 ,
and T i 0 2 - A l 2 0 3 , " Bur. Stand. J . Res. 1 1 , 719-25 (1933).
33.
Dimbleby, v . , and W. E. S. Turner, "Relationship Betvreen the
Chenical Composition and the Resistance of Glasses to the Action
of Chemical Reagents. I . " J. Soc. Glass Tech. 10, 325 (1926).
34.
D i t l e r , S . , "Degree of Oxidation of Titanium, i n S i l i c a t e s , "
Holder-Pichlcr-Tempslcy, A. G., Vienna and Liepzig 1929. 42 pp.
Coram. Abs. % 474 (1930).
35.
Button, 0. A.., and D. ? . Wagner, "Titanium and Zirconium Compounds:
Their Application i n Enamels," Ceran. Ind. 2^., 24-5 (1935).
36.
Farberdndustrie, I . G., "Glass," Ger. 556, 895 (Oct. 24, 1929),
Fr. 703, 671 (Oct. 14, 1930), Chen. Abs. 25, 4677 (1931).
37.
Farbenindustrie, I . G., "Production of Colorless Glass Containing
T i 0 2 , " Ger. 622, 346 (Nov. 7, 1935), Ceram. Abs. 15, 240 (1936).
38.
Kaiser Wilholm I n s t i t u t e for S i l i c a t e Research, "Glass, Enamels
Opacified by T i 0 2 , " Ger. 638, 737 (1940).
39.
Kinzie, C. J . , and J . A. Plunkett, "Titanium Compounds and Application Thereof i n Vitreous Enamels," J . Amer. Ceran. Soc. 1 3 ,
117-22 (1935).
40.
Landram, R. B. and L. J. Frost, "Titanium Enamels, 11 J. Aner.
Ceram. Soc. 3 , 316-21 (1920).
41.
Leuchs, G. and C. Leuchs, "Improvements i n the Manufacture of
Glazes, Enamels, and the L i k e , " B r i t . 9 6 H (I893).
1,2. Morey, J. Y.r., and II. E. Lervdn, "Relation .octwecn the Jompositiou
and t h e Density and Optical Properties of Glass I . Soda-LimeS i l i c a G l a s s e s . , " J . Opt. foe. Auer. 22, 632-62 (1932).
43.
Morey, G. W., "The Properties of Glass," Reinliold, N. Y., 1938.
44.
Nagai, S. and M. Masuda, "Chemical Compositions and Water Solub i l i t i e s of Glasses, V I I I . " J . Jap. Ceran. Assoc. 46, 525-31 (1938).
45.
Sav/aij I . , and K. Suda, "Properties of Glasses Containing Titanium
and'Cobalt Oxides," J. Soc. Chen. Japan ^ 1 , 719-20, Suppl. binding
356B (1933) Ceran. Abs. 19_, 10 (1940).
4.6.
Sheen, A. R., and W. E. 3 . Turner, "Sono Observations on the I n fluence of T i t a n i a on the Properties of Glass," J. Soc. Glass Tech.
8, 137-205 (1924).
47.
Stokes, G. G., B r i t . Assoc. Reports 33T (1871).
43.
Stokes, G. G., and J. Hopkinson, B r i t . Assoc. Reports
49.
Thomas, F . , "Silo;dde, A Valuable Substitute for Quarts Glass,"
Chen.-Ztg. 36, 25-7 (1912).
50.
Titanium Pigment Co., "Glass," B r i t . 176, 430 (Oct. 29, 1920).
51.
Vondracek, "Titanium as an Opacity Producing M a t e r i a l , " Sprechsaal
42, 593 (1909).
26T (1875).
(
C.
DISCUSSION
52.
Biscoe, J . , and B. E. Warren, !'X-Ray Diffraction Studies of SodaBoric Oxide Glass," J . Amer. Ceram. Soc. 2 1 , 287-93 (1938).
53.
Biscoe, J . , C. S. Robinson J r . , and B. E. Warren, "X-Ray Study
of Boric Oxide-Silica Glass," J . Amer. Ceram. Soc. 22, 180-4 (1939).
54.
Biscoe, J . , A. G. Pincus, C. S. Smith J r . and B. E. Warren, "X-Ray
Study of Lime-Phosphate and Lime-Borate Glass," J . Amer. Ceram.
S o c , i n press.
55.
Bragg, W. L . , "Atomic Structure of Minerals, I t h a c a , Cornell
University P r e s s . 1937.
56.
Ephraim, F . , "Inorganic Chemistry," 3rd English Edition. P.C.L.
Thome and A. M. Ward. London, Gurney and Jackson. 1939.
57.
Goldschmidt, V. M., "Crystal Structure and Chemical C o n s t i t u t i o n , "
" T r a n s . Faraday Soc. 25_, 253 (1929).
5&* Greig, J . W., "Immiscibility i n S i l i c a t e M e l t s , " Amer. Jour.
Science 1£, 1 - 154 (1927).
59.
Krocek, F . C , "The C r i s t o b a l i t e Liquidus i n t h e Alkali Oxide
S i l i c a Systems and the Heat of Fusion of C r i s t o b a l i t e , " J . Amer.
Chem. Soc. 52, 1436 - 42 (1930).
60.
Pauling, L . , "The P r i n c i p l e s Determining the Structure of Complex
Ionic C r y s t a l s , " J . Amer. Chem. Soc. 51, 1010 (1929).
61.
Taylor, N. W., "Crystal Structure of the Compounds Zn2 TiO/,
Zn2SnO^, Ni 2 SiO^ and NiTi03. Z.f. Phys. Chem. B9_, 241-64 {1930).
62.
Warren, B. E . , H. Krutter and 0. Momingstar, "Fourier Analysis
of X-Ray Patterns of Vitreous S i 0 ? and B^Oo," J . Amer. Ceram.
Soc. l g j 202-6 (1936).
63.
Warren, B. E. and J . Biscoe, "Fourier Analysis of X-Ray P a t t e r n s
of Soda-Silica Glass," J . Amer. Ceram. Soc. 2 1 , 259, 65 (1938).
64.
Warren, B. E. and A. G. Pincus, "Atomic Consideration of Jjamiscibility
i n Glass Systems," J . Amer. Ceram. Soc. i n p r e s s .
65.
Weyl, W. A., "Chemistry of Colored Glass," Glass Ind. 1 8 .
73-78, 117-20, 167-71 (1937).
66.
Zachariasen, W. H., "The Atomic Arrangements i n Glass," J . Amer.
Chem. Soc. 5_A> 3841-51 (1932).
Документ
Категория
Без категории
Просмотров
0
Размер файла
3 355 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа