Note: Descriptions are shown in the official language in which they were submitted.
8Z4
The invention relates to producing glass articles having
a composition range which reduces the breaking tendency substantially.
A further embodiment of this invention discloses a heat-treatment proced-
ure which also reduces the breaking tendency of partially leached articles.
The present invention describes a method of producing glass
articles in which the composition of the surface layers of the walls is
richer in silica than the central layers. According to this invention a
phase separable composition in alkali borosilicate or germanate systems
having 0 to 5 mole percent alumina is melted and formed in the conventional
manner. Suitable compositions have been described in previous patents:
U.S. Patents Nos. 2,106,744; 2,215,039; 2,355,746 (Hood and Nordberg);
Nos. 3,843,341 (Hammel) and 3,938,974 (Macedo and Litovitz).
Depending on the composition, the glass is heat-treated at a
temperature from about 500 C. to 700 C. for a few seconds to several weeks
such that it separates into two phases having a totally interconnected
microstructure with an average thickness of 100 - 2000 A preferably be-
tween 150 and 500 A. These dimensions are much smaller than the wall thick-
ness of the preformed article.
One of the phases is mostly covalent and will be referred
to as the "hard" phase. The other is predomonantly ionic in nature and
will be referred to as the "soft" phase. It has been found advantageous
to heat treat the glass at as low a temperature as possible but one which
is consistent with reasonably short heat treatment times. This reduces
the possibility of any deformation of the preform due to viscous flow.
The soluble phase is then leached out by a suitable leaching
solution only up to a definite depth leaving an unleached central layer
in the glass walls. Since the glass article is not leached throughout
its mass but only in the surface layers, we call it "the partial leaching
technique" in contrast to other processes where the leaching is carried out
throughout the mass of the article.
-2-
- ~ :
.
~9~8Z4
3.
We h~ve discovered that in order to increase the survival
proba~ility in the subsequent stages of the process, it is ad- '
vantageous to wash the partially leached article with a suitable
organic solution which should contain less than 5~ weight per-
cent water, preferably less than 25~ water. The organic contain- j
ing solvent media should ~e one that preferentially keeps in solu-
tion the oxides of boron and other leaching reaction products of
the silica-poor phase which'are formed during leaching, without
substantially attacking the silica~poor phase. While not intend-
ing to be-bound by the veracity of the mechanism, we theorize that
removal of the oxides of boron and other leaching reaction pro- I'~
ducts from the porous layer greatly increases the probability for
the survival of the glass article through the process. If only
water is used as a washing media, the silica-poor phase is attacked l~
during washing, generating constantly new reaction products and j,
preventing the porous glass fxom bei'ng cle~ned. The organic
components of the media are w~ter ~iscible and include the
lower molecular ~eight' aliphatic alcohols, containing from 1 to
5 carbon atoms, or acetone or mixtures thereof, ~ere found to
be especially`good, methanol and ethanol being preferable. The
organics in an aqueous media decrease the rate of attack of
5;/ic~
water on the slilca-poor phases' while still permitting the l
porous glass to be washed. Merely washing with water as de- ~¦
scribed by Hood and Nordberg results in excessively large
breakage rates in the later stages of the process, making it
economically impractical. By using the approach of the present
invention, i.e., washing with methanol at room temperature
for times varying from 5 minutes to 24 hours, depending on the
thickness of the leached layer, it is possible to reduce the
breakage rate to a level where the process becomes economically
feasible. The washing temperature is not critical. For
economic reasons alone, it is desirable to carry out the washing
at or near the room temperature.
After washing, the article is dried as usual and consolidated
at tempera'tures in the range of 800-1000C, depending on the com-
8'~
position of the porous glass in the surface layers on solid glass objectssuch as tubes and bottles. The thickness of the silica-rich surface layer,
which has greater strength than the remainder of the article, varies with
the intended use. For example, the thickness may range from 0.1 mm for uses
where minimum abrasion is present up to several mm for uses where severe con-
ditions will be encountered. In each case, silica-rich surface layers of
the desired thickness can be produced on all exposed surfaces of the art-
icle regardless of shape and the resultant article will have improved phy-
sico-chemical properties associated with the presence of silica-rich sur-
faces. The glass articles produced according to this invention can be usedto make cooking utensils, radar domes, air and space craft windshields,
windows for structural applications, containers such as jars, bottles and
pipes, and chemical ware such as reaction vats.
We have further discovered that the breakage rate is substan-
tantially reduced provided the composition is selected from the following
ranges (in mole per cent):
Broad Preferred Most Preferred
SiO2 50-66 58-65 60-62.5
B2O3 28-42 30-35 32.35
R20 3.5-9 4-7 4.5-5.5
2 3 0-1 0-0.5
f~ 0.4-0.85 0.5-0.75 0.6-0.7
where R2O is the total concentration of all alkali metal oxides and f is
the ratio of the sum of K20, Rb2O and Cs20 contents to R2O content.
We have also discovered that:
i) The most preferred composition range yields the highest
survival rates for a given article. As one goes to the preferred and then
to the broad range, good survival rates are still obtained. However, the
survival probability decreases from the preferred range to the broad range.
ii) The survival probability increases as the ratio, x, of
thickness of porous layer to wall thickness of the article increases for
any given composition and processing conditions.
-4-
.
: ,
10~824
5.
iiil The ~urvival probability is high, provided the heat
treatment ~o cause phase separation is chosen according to the
following table:
- ~ - road Prefexred
Temp. (C) 480-550 500~530
Time (hrs.) 1/2-200 5-1~0
Finally, it is a well-known fact that the~survival probability
decreases as the size of the article increases. Therefore, whether
in practice one uses the most preferred composition and/or heat-
treatment ranges or the broadest ranges depends on the ratio,x,
ar.d the size of the article. For small siæe and larger x, the broad
range may be most economical. On the other hand, large article size
and small values of x clearly require the most preferred range.
EXAMPLES
1. A mixture having the composition (in mole %) 4 Na2O, 4 K2O,
36 B2O3 and 56 SiO2 was melted-and stirred to produce a homogeneous
melt from which rods were drawn having a diameter in the range of 7
to 8 mm. The rods were heat-treated at 550C for 1.5 hrs. to
cause phase separation and then the furnace was cooled.
The rods were partially leached at 95C with 3N HCl. The
leaching time was chosen to be two hrs. which corresponded to a
leached layer of about 1 mm in thickness. The partially leached
rods were washed in methanol at room temperature for a period of
24 hrs. and dessicated at room temperature for 24 hrs. The rods
were heated at a rate of 1C/min. up to 150C and then at a rate
of 2C/min. up to about 850C when they consolidated. The glass
article thus produced was clear in appearance and had a surface
layer 1 mm thick containing more than 90 mole % of SiO2.
Photoelastic measurements showed that the surface layex had
a uniform compression of 24,000 psi. The process may also be carried
out using ethanol, propanol, acetone and aqueous mixtures containing
50 wt. % methanol and 75 wt. ~ methanol instead of methanol in the
above example.
109'~8'~:4
6.
2. A-m~xture having the composition ~ir. mole per cent? ~'
61-.3 SiO2, 33.4-B203, 1.8 Na20 and 3.5 K20 was melted and stirred t
to~p'roducê' a~homogeneous melt from which rods were drawn having
diameters in the range of 7 to 8 mm. The rods were heat-treated
at~ 515C fo'r 100 hrs. to cause phase separation and the furnace
wa~s-cooled.
The rods were par~ially leached, washed, dried and consoli-.
da-ted as described in Example 1. Breakage rate was found to be
l-ess than 5%. The finished rods had a surface layer about 0.50 mm'
t~ick in compression of about 30,000 psi. These rods were abraded
an~ tested in three point bending. The modulus o~ rupture was
fo'un~d to be 40,000 psi.
`. _
ll
