Note: Descriptions are shown in the official language in which they were submitted.
~03889~L
This invention relates to ceramic materials.
In one aspect, the invention resides in a ceramic
material including at least 90% of a single phase compound in
the form of a ~-phase silicon nitride lattice in which the
silicon in the lattice has been partially replaced by aluminium
and the nitrogen has been partially replaced by oxygen.
In a further aspect the invention resides in a method
. . .
of producing a ceramic material comprising mixing not more than
75% by weight of high active surface area alumina in powder
form of particle size less than lO microns, or a compound of
aluminium which decomposes to give the requires alumina at the
elevated temperature of the process, with powdered silicon
nitride of particle size less than 20 microns and sinteriny the
mixture at a temperature in the ran~e o 1550C to 2000C for
at least 30 minutes to produce the ceramic material.
Preferably the sintering operation is accompanied by
pressure.
Preferably the starting materials are surrounded by a
protecting medium while at elevated ternperature, such as powdered
boron nitride.
Preferably the silicon nitride powder is less than 5
microns particle size.
Preferably the alumina powder is less than l m.i.cron
particle size.
More preferably the alumina is 0.5 microns particle
size.
In a modification of said further aspect of the
invention, the silicon nitride powder is produced by nitriding
silicon powder in the presence of the alumina powder or said
com~ound of aluminium, the atomic ratio of silicon to aluminium
being greater than or equal to 1 : 3 and the nitriding tempera-
ture being between 1250C and 1600C.
~L~38891
Preferably, where the atomic ratio of silicon to
aluminium is less than 3 : 1, the method includes the further
step of raising the temperature to a value in excess of 1600C.
Conveniently, the further heating step is also con-
ducted in a nitriding atmosphere.
Alternatively, the further heating step is conducted
in a separate furnace within a protective medium, such as boron
nitride powder.
Preferably, the further heating step is conducted at
at least 1700C, preferably 1900C and more preferably 2000C.
Preferably, the starting materials are less than 5
microns particle size, and more preferably less than 1 micron
particle size.
Conveniently the components are cold pressed to shape
prior to nitridiny.
The accompanying drawing is a graph showing the
change in the a and c unit cell dimensions of a ceramic material
produced by a met'nod accordiny to one example of the invention
as the alumina content in the starting materials is increased.
In a first example of the invention, silicon nitride
powder consisting of at least 8~% of the a-phase material was
rnixed wi-th high purity ~-alumina powder having a mean particle
size of less than 1 micron and a high surface area and ~:eac-
tivity 90 that the mixture contained 27.5% by weight of the
alumina. The mixing operation was effected by wet ball milling
the powders in iso-propyl alcohol for 72 hours until the average
particle size of the mixture was 5 microns. The mixture was
then dried and placed in a steel die between steel punches,
and pressed at room temperature at 2000 lbs/square inch to form
a self-supporting pre-Eorm o~ approximately 3/4 in. diameter by
3/4 in. in length. The preform was then transferred to a gra-
phite hot pressing tool comprising a die with a 1 inch diameter
-- 3
1~)3889~L
bore, a plug and a punch, all surfaces presented to the die
cavity having been sprayed with boron nitride powder to a depth
of 0.002 to 0.005 ins. Before inserting the preform into the
cavity a 3/4 in. layer of fine boron nitride powder was poured
into the cavity onto the plug, and the preform was placed in
the centre of tlle cavity and pushed into the bed of powder so
that the powder rose into the annular space between the preform
and the die walls. More fine boron nitride powder was then
poured onto the preform until the layer was approximately 3/4
in. above the top of the preform. The graphite punch was then
assembled into the die cavity and onto the bed of boron nitride
powder, and pushed to compact the powder. By this means the
preform was embedded in a compacted protective environment
during the subsequent hot-pressiny operation. The boron
nitride used in the example was hexagonal powder sold by New
Metals and Chemicals Limited, having a particle si~e of the
order of 5 microns and termed Grade FL/T. The mixture was then
pressed at 2.5 tons/square inch at a temperature of 1700C for
1 hour, the temperature being raised from room temperature to
the hot pressing temperature over a period of 20 minutes while
the pressure was raised from an initial 500 lbs/square inch at
room tempera~ure to full pressure at approxima~ely 1500C. The
product was allowed to cool in the dies while under pressure.
The resultant product was subjected to X-ray analysis using
monochromatic CuKa (Hagg-Guinier focusing camera, KCl standard)
and was found to consist predominantly of a si~gle phase silicon
aluminium oxynitride ceramic material having a crystal structure
based upon that of ~-phase silicon nitride, but having increased
cell dimensions, there also being present a small percentage,
less than 5~, of an unidentified phase.
The above example was then repeated at other pressing
temperatures within the range 1600C to 2000C, the pressing
-- 4 --
1~)38891
at 2000C being maintained at fu~l pressure and ~emperature for
1/2 hour. X-ray analysis of the resultant products showed that,
while pressings performed at the lower temperatures gave pro-
ducts which contained a small percentage of an unidentified
pllase in addition to the single phase ceramic material obtained
previously, the unidentified phase was absent from the X-ray
traces of products obtained from pressings performed at the
upper temperatures, nearer to 2000C. Two samples were pressed
at 1800C, one for 1 hour at temperature as in the first example,
the other for 3 hours at temperature, without any difference
being observed in the resultant products.
Further work performed on time at temperature showed
that at tlle lo~er ternperatures 1600C to 1700C, incomplete
reaction occurred if the time under pressure was below approxi-
mately 30 minutes, consequently at the lower temperatures it
proved preferably to maintain at temperature for at least 1 hour,
while at the upper temperatures 1800C to 2000C, complete reac-
tion had taken place by maintaining under pressure for approxi-
mately 30 rninutes.
By way of comparison, the procedure of the first
example was repeated, but with the hot pressing temperature being
below 1600C. In this case, the reaction between the silicon
nitride and alumina to produce the single phase ceramic material
obtained previously was found to be incomplete such that it was
not found to be possible to produce more than 90% of the single
phase ceramic material even after prolonged heating under pres-
sure and, for example, when the hot pressing temperature was
1500C, the product contained only about 40% of said material.
In a second example of the invention, a similar
mixture to that of the first example was prepared, but in this
case the mixture was arranged to contain 50~ by weight of alumina.
Samples of this mixture were then hot pressed, by the method of
-- 5 --
1~38~39~
the first example, at varying temperatures within the range
1600 to 2000C and the resultant hot pressed products were sub-
jected to X-ray analysis. Ayain it was found that the products
resulting from pressin~ at the lower temperatures not only con-
sisted of at least 90~ by volume of the single phase compound
obtained above but also contained a small percentage of an un-
identified phase. However, as ln the preceding example, it was
found that the unidentified phase was absent from the X-ray
traces of the products obtained at the upper pressing tempera-
tures, in the order of 2000C.
In a third example, the procedure of the second exam-
ple was repeated, but with the silicon nitride/alumina mixture
being arranged to contain 70% by weight of alumina. In this
case also, X-ray analysis of the resultant products showed that,
with the hot pressing temperatures being the lower part of the
range, the resultant products consisted predominantly of a single
phase silicon aluminium oxynitride ceramic material, but now
the products contained some free alumina as well as a small
percentage of an unidentified phase. However, as the hot
pressing temperatures approached 2000C, it was found that the
resultant products contained no free alumina and the unidenti-
Eied phase was absent rom the X-ray traces.
In a fourth example, the method of the third example
was repeated with a silicon nitride/alumina mix containing 75
by weight of alumina and hot pressed at 2000C for 1/2 hour.
. The resultant product contained more than 90% by volume of a
single phase silicon aluminium oxynitride ceramic material,
about 5~ alumina and a trace of an unidentified phase.
By way of comparison, the method of the fourth example
was repeated wi-th silicon nitride/alumina mixes containing more
than 75~ by weight of alumina. }lowever, it was fo~mcl that even
when hot pressing was performed at 2000C the resultant product
-- 6 --
1038~9~
contained some free alumina in addition to the substantially
single phase ceramic material obtained previously, said single
phase material occupying less than 90~ of the ceramic phase.
In a fifth example of the invention, the procedure
of the first example was followed with a silicon nitride/alumina
mix containing 20% by weight of alumina and hot pressed as be-
fore at varying temperatures within the range 1600C to 2000C,
with hold times of 1 hour for all except the 2000C pressing
which was helf for 1/2 hour. The resultant products contained
the single phase material obtained above together with 10~
unidentified phase for the pressing at 1600C, the quantity of
t~le unidentified pllase reducincJ with increasing pressing tem
perature until samules pressed in the region of 2000C contained
no unidentified phase.
In a sixth example of the invention, the procedure of
the fifth example was repeated with the starting mixture con-
taining 10~ by weight of alumina. The same pattern of resultant
products as in the fifth example was obtained.
Similarly in a seventh example using a 2% by weight
of alumina mix, products included at least 90% of the single
pllase material referred to above.
In all the examples quoted above the bulk density of
the products obtained was of the order of 3.04 g/cc.
While in the examples given, high ~-phase silicon
nitride was used, some examples were repeated with a low composi-
tion and little difference was observed in the comparable pro-
ducts.
The high active surface area alumina used in most of
the examples was high purity ~-alumina supplied by La Pierre
Synthetique Baikowski, France, known as type GE 30 of mean
particle size 0.5 microns with a surface area greater than 1 sq.
m/gm. ~iowever, alumina supplied by the Aluminium Company of
-- 7 --
103889~
America, types XA16 and XA17 was also used successfully. Also
alumina supplied by Degussa and termed gamma-alumina was used
successfully.
While in all the examples given alumina has been one
of the starting materials it will be appreciated that compounds
of aluminium which decompose to give alumina at the hot pressing
temperature could be used for e*ample aluminium hydroxide and
aluminium nitrate. For example, 7 gms of the silicon nitride
of the previous examples was added to a solution of 42 gms. of
aluminium sulphate in 75 mls. of water. To this mixture, 22.5
mls. of ammonium hydroxide (0.880) was added and mixed for 18
hours. After decanting and washing, the precipitate was dried
and hot pressed in the normal way at 1700C for 1 hour to pro-
duce a 95~ silicon aluminium oxynitride together with 5~ of an
unidentified phase.
It will be observed from the above seven examples that,
as the alumina content is increased the silicon in the tetrahed-
ral framework of silicon nitride has been partly replaced by
aluminium while simultaneous replacement of nitrogen by oxygen
has occurred. With large degrees of replacement it was observed
that, while the ~-structure was retained, the unit cell dimen-
sions were substantially constant and some free alumina remained.
The graph shown in the drawing demonstrates the cell dimensional
change with increasing alumina content.
In all the examples the surfaces of the hot pressed
product were found to have an adherent boron nitride layer which
was removed in subsequent shaping operations.
In producing a substantially single phase ceramic
product by the method described above, it is desirable to ensure
that the hot pressing temperature is greater than 1600~C, or
more preferably greater than 1700C. Alsim where a 100% single
phase product is required, then it is desirable that the hot
-- 8 --
1~)3889:~L
pressing temperature should be greater than 1900C and more
preferably be of the order of 2000C. It ~ill, however, be
appreciated that the upper temperature lim.it is governed by
parameters such as the dissociation temperature of the ceramic
product and the strength of the hot pressing tools.
Also, the time for which the pressure at maximum
temperature is maintained should preferably be greater than
30 minutes, and its upper level will obviously be determined by
economics and/or degradation of the of the sample.
It will be noted in the examples given, the silicon
nitride/alumina mixtures were protected within the graphite dies
during the hot pressing, by embedding the mixtures in boron
nitride powder, in addition to the conventional spray-coating
of the tools with boron nitride. Otherwise, it was found, at
the upper hot pressing temperatures, above 1800C, some diffi
culty was experienced in removing the hot pressed examples from
the tools and in some instances some degradation of the surface
of the sample occurred. However, the additional protection
afforded by embedding the samples in boron nitride powder alle-
viated this difficulty. However, it will be appreciated that at
the lower hot pressing temperatures, below 1800C, the conven-
tional spray coating of boron nitride on the tools provides
satisfactory protection.
It will be appreciated in the examples given that,
whilst the silicon nitride/alumina mixes were introduced into
the die cavities as preforms, by arranging the boron nitride
powder as preforms, the mixes could be introduced in powder
form.
In an eighth example of the invention, 14 grams of
the silicon nitride powder of the earlier examples was mixed
witll 13.6 grams of high purity a-alumina powder having a mean
particle size of less than 1 micron and a high surface area and
-- 9
103889~l
and reac-tivity, as sold by La Pierre Synthetique Baikowski,
France, and 0.14 grams of amminium alginate powder. To the dry
intimate mix was added 36 ml. of water and the mixture was then
mixed on a roller mill for 1 hour and allowed to stand. The mix-
ture thus obtained was of slip-casting consistency and was slip
cast in a plaster-of-Paris mould into the shape of a crucible.
The slip was dried, removed from the mould and placed in a
graphite reaction tube which was lined with an alumina tube,
one end of the alumina tube being closed by a pressed plug of
alumina powder. The alumina tube was then half filled with
the fine hexagonal boron nitride powder used in the ear:Lier
examples, whereafter the crucible was placed on the boron nitride
powder away from the alumina tube walls, and more boron nitride
powder was poured over the crucible until it was completely
bured. The other end of the alumina was then closed with a fur-
ther pressed plug of alumina powder and the assembly was heated
at a rate of 90C per minute to 1700C and held for 1 hour.
After 1 hour at the sintering temperature the tube was allowed
to cool, and the resultant product was found by X-ray analysis
to consist predominantly (90%) of a single phase silicon
aluminium oxynitride ceramic material.
~ he crucible as produced by the above method was
found to have some adherent boron nitride when removed from the
surrounding boron nitri~e protective medium which was removed
by sand-blasting.
In a ninth example of the invention, silicon nitride
powder consisting of at least 85~ of the ~-phase material was
mixed with high purity ~-alumina powder having a mean particle
size of less than 1 micron and a high surface area ancl reactivi-
ty. The mixing operation was performed by wet ball mil:Ling thepowders in iso-propyl alcohol for 72 hours until the average
particle size of the mixture was 5 microns, the final mixture
-- 10 --
~L~3885~
being arranged so that the atomic ratio of silicon to aluminium
in the mixture was 9 : 1. When the wet ball milling operation
was completej the mixture was dried and a 100 gram sample was
placed in a steel die and compacted to form a self-supporting
preform at a pressure of 2000 lbs/square inch. The preform was
then removed Erom the steel die and placed within boron nitride
powder in a graphite die lined within an alum1na tube as in the
first example. The assembly was then heated over a period of 20
minutes to the required sintering temperature, which in this
1~ case was 1700C. After 1 hour at the sintering temperature, the
~-~ tube was allowed to cool and the resulting product was found, by
X-ray analysis, to consist predominantly of a single phase sili-
con aluminium oxynitride ceramic material having a crystal
structure ~ased upon that of ~-phase silicon nitride~ but of in-
creased cell dimensions, the product also containing a small
amount, less than 5%, of an unidentified phase.
, The above example was then repeated at other tempera-
tures within the range 1600 to 2000C. It was observed that,
- while products produced at the lower temperatures within the
20 range contained a small percentage of an unidentified phase, the
unidentified phase was absent from the X-ray traces of products
obtained at the upper temperatures, that i5 nearer 2000C,
these products consisting entirely of the single phase material
defined above.
By way of comparison, the procedure of the ninth exam-
ple was repeated, but the sintering temperature was now held
below 1600C. In this case, the reaction to produce the single
phase ceramic ma-terial defined above was found to be incomplete
and, for example, when the sintering temperature was 1500C,
3~ the product only contained 40% of said material.
In the tenth to fifteenth examples of the invention,
the silicon nitride and alumina uowders of the first example
~L~3889~L
were used to produce mixtures which contained 20, 25, 40, 50, 60
and 70~ by weight of alumina respectively. In each case, subse-
quent treatment of the mixture followed the procedure of the
first example and the sintering operation was carried out at
various temperatures between 1700C and 2000C. It was found
that where the sintering reaction was performed at the lower
temperatures in the range, each:of the resulting products was
found to consist predominantly of a single phase silicon alum-
inium oxynitride as obtained in the previous examples, together
in each case with a small percentage, less than 5~, of an uni-
dentified phase. In addition, some free alumina was found to
be present in the products formed from the 60% and 70~ by weight
of alumina mixtures. ~lowever, where the sintering reaction was
carried out at the higher temperatures in the range, that is
nearer 2000C, it was found that the unidentified phase was ab-
sent from each of the products and moreover that the alumina was
absent from the products obtained by sintering the mixtures con-
taining 60% and 70% by weight of alumina. Thus in each case the
product consisted entirely of the single phase ceramic material.
Inventigations were also carried out on mixtures which
contained above 7S~ by weight of alumina, but these were found
to contain more than 10% free alumina even when sintered at
2000C.
It will be appreciated, examples eight to fifteen dif-
fer from the earlier examples in that the sintering operations
are not accompanied by pressure. However, while the resultant
products have lower bulk density figures of the order 2.7 gms/cc,
i-t is found that similar deductions to those of the examples one
to nine may be made in respect of control parameters to ensure a
product containing at least 90% by volume of a silicon aluminium
oxynitride is obtained. Thus, it is desirable to ensure that
the sintering temperature is greater than 1600C, or more
- 12 -
~038891
preferably greater than 1700C. Where a 100% single phase pro-
duct is re~uired, it is desirable ko ensure that the sintering
temperature is greater than 1900C or is more preferably of the
order of 2000C. It will of course be appreciated that the
upper temperature limit is governed by parameters such as the
dissociation temperature of the ceramic products and the strength
of the tools.
It will also be noted that in examples eight to fif-
teen the samples were embedded in boron nitride powder during
sintering, as was of course the case with the earlier examples.
It was found that without this protecting medium, degradation
of the surface of the samples occurred, which was especially
pronounced at the upper temperatures, above 1800C, such that
a~ove 1900C the sample could not be tested. ~lowever, embedding
the samples in boron nitride powder alleviated this difficulty.
However, it will be appreciated that at sintering temperatures
of the order of 1700C such coatings are not necessary but would
obviousl~ be preferred if large complex shapes are being sin-
tered.
While boron nitride is preferred as the protecting
medium it will be appreciated that mixes includiny boron nitride
could be employed, as could other pro-tecting means such as
yaseous media, for example nitrogen adjusted to appropriate
partial pressure. Obviously, it is more convenient when choosing
a powdered protecting medium to choose one which does not sinter
at the hot pressing temperature of the process. Any adherent
protective medium to the hot pressed component is easily re-
moved, such as by grinding, sand blasting etc.
In a sixteenth example of the invention, fine silicon
powder, as sold by Murex Limited under the trade name Superfine,
and having a mean particle siæe of 3 microns was wet mixed in
iso-propyl alcohol with high purity alumina powder having a
- 13 -
1~3~1519~
mean particle size of less than 1 micron and a high surface area
and reactivi-ty. The mixture was produced so that the atomic
of silicon to aluminium was 3 : l and, when mixing was complete,
the iso-propyl alcohol was removed and the resulting powder mix-
ture passed througll a 60 mesh sieve. llwenty grams of the sieved
mixture was then introduced into a rectangular, silicon nitride
boat of length 3 inches and width 2 inches and the powder was
tapped to compress it slightly. The boat was then placed in an
alumina reaction tube, which was subsequently evacuated and re-
filled with forming gas until the gas pressure within the tubewas approximately that of ambient atmospheric pressure. The
temperature witllin the reaction tube was then raised over a
period of ~ hours to the required nitriding temperature, which
in tllis example was 1400C, and was retained at this temperature
for six hours, forming gas being allowed to pass through the
reaction tube at a rate of half litre/minute throughout the
heating operation. The reaction tube was then allowed to cool
over a period of 8 hours and, on removal from the tube the resul-
tant product was found, by X-ray analysis, to predominantly of
a single phase silicon aluminium oxynitride.
The above example was then repeated with the sintering
temperature being varied between 1300 and 1600C, all the other
conditions remaining the same. In each case, the product, as
determined by X-ray analysis, was found to consist predominantly
of the single phase ceramic material of the first example. Also
the product was again found to contain a small amount of an
unidentified phase.
In a first modification of the sixteenth example, the
same procedure was repeated, but the nitriding temperature was
now held just below 1300C. In this case, although the product
of tlle sintering reaction was found to contain the silicon
aluminium oxynitride defined above together with a small
-- 1~ --
~3889~
: percentage of an unidentified phase, the product also included
~ free silicon nitride and free alumina.
: In a second modification, the products of the six-
teenth example and the first modification thereof were heated
to 2000C in graphite dies and then allowed to cool. The resul-
tant materials were subjected to X-ray analysis and were found
to consist entirely of the siliaon aluminium oxynitride defined
above, the unidentified phase b=eing absent from each X-ray trace
and no free alumina or silicon nitride being present in the
material produced according to the first modification of the
sixteenth example.
In a third modification, the procedure of the six-
teenth example was again repeated, but this time the nitriding
temperature was arranged to be in excess of 1600C. The same
single pllase ceramic material defined above was found to be
contained in the product of the sintering reaction, but the
; product was also found to include some aluminium silicate and
aluminium nitride.
In a further modification of the sixteenth example,
the silicon/alumina mixture was mixed with an acrylic dispersion
in water to make a mixture of extrudable consistency. A green
preform produced from the mixture was heated to drive of~ the
water and burn off the acrylic dispersion. The porous product
on heating in a nitrogen atmosphere at 1400C was converted to a
single phase Al'- Si -n -O compound and having the expanded
~-phase silicon nitride cell form.
The initial mixture of the above example together with
a small quantity of an alginate deflocculating agent when tem-
pered with additional water was found to have excellent slip
casting properties, and a slip was cast by conventional slip
casting techniques in a plaster-of-Paris mould to form a small
crucible. The slip cast crucible was removed from the mould
- 15 -
~03~91
and placed in a nitrogen atmosphere. The temperature was in-
creased from room temperature to 1400C in six hours and held
for six hours. A crucible containing preclominantly the single
phase compound containing Al - Si - N -O of the expanded ~-phase
cell form was produced. It will be appreciated that other means
; for mixing the silicon and alumina may be employed, for example
the two materials could be flame-sprayed onto a release agent on
a former, and the article so formed could then be removed and
nitrided.
In a seventeenth example, the silicon and alumina pow-
ders of the slxteenth example were wet mixed in iso-propyl alco-
hol so that the relative atomic proportions of silicon to
aluminiunl in the resultant mixture was 1 : 1. Subsequent pro-
cessing of the rnixture then proceeded as in the si~teenth exam-
ple and, after nitriding at 1~00C, the resultant product was
found to consist predominantly of the single phase ceramic
material defined above. The product was, however, also found
to contain a small percentage of an unidentified phase together
with a small amount of free silicon nitride and a large amount
of free alumina, such that the silicon aluminium oxynitride con-
tent was less than 90% of the ceramic phase. The product of
the sintering reaction was buried in boron nitride powder in a
graphite die and heated to 1700C, whereafter an X-ray analysis
was performed on the resultant material. The analysis showed
that the material still contained a small percentage of an
unidentified phase, but consisted of at leas-t 90% by volume of
the required single phase product, the unreacted alumina having
' been removed. This second sintering step was then repeated with
a further product produced according to the seventeenth example,
but in this case the heating within the boron nitirde in the
graphite die was performed at 2000C. Again an X-ray examina-
tion of the resultant material was performed and in th:is case
- 16 -
.
` `
~L~3~389~.
the material,was found to consist entirely of the required sin- '
gle phase proauct, the unidentified phase being absent from the
X-ray trace.
In a modification of the seventeenth example, the
sintering temperature was arranged to be 1300C and in this case
the ceramic product was found to contain free alumina and free
silicon nitride in addition to the silicon aluminium oxynitride
produced in the sixteenth and sëventeenth examples. Two samples
of this product were then heated in boron nitride powder in a
graphite die to 1700C and 2000C respectively and the effect
of the heating operation investigated by X-ray analysis. In the
case of the sample heated to 1700C, the resultant product was
found to consist substantially entirely of the single phase
Material of the first example. A small amount of an unidenti-
fied phase was also present in this sample. However, with the
same which was heated to 2000C, it was found that the product
consisted entirely of the single phase material of the first
example.
In an eighteenth example the procedure of the six-
teenth example was followed in which the silicon to aluminiumatomic ratio was 1 : 3. As in the sixteenth example the two
stage heating process was followed, but in this case the second
heating step at 1700C was performed with the product of the
first heating step being embedded in boron nitride powder. The
product was found to contain approximately 90~ silicon aluminium
oxynitride together with approximately 10~ alumina.
In a nineteenth example the procedure of the eight-
eenth example was followed with the silicon to aluminium atomic
ratio being 7 : 1, the product obtained being more than 90
silicon aluminium oxynitride.
It was observed in the sixteenth to nineteenth exam-
ples that with product subjected to temperatures of the order
- 17 -
1~31989~
1900C to 2000C the specimens evidenced some degree of surface
degradation if unprotected at these temperatures. This difficul-
ty was alleviated by embedding the samples in boron nitride
powder although it will be appreciated that other protecting
media such as silicon carbide powder or even gaseous media such
as nitrogen of controlled partial pressure could have been used.
In the examples sixteen to nineteen,~~alumina has been
used as the starting material. It will, however, be appreciated
that as in the earlier examples, compounds of aluminium which
decompose on heating so as to provide alumina at the nitriding
temperature could have been employed.
In performing the method described in examples sixteen
to nineteen it is desirable to ensure that the nitriding step
is performed at temperatures less than 1600C, but preferably
above 1250C, and more preferably between 1300C and 1500C and
most preferably at about 1400C. Also, for mixes where the sili-
con to aluminium atomic ratio is less than about 3 : 1, it is
desirable to raise the temperature of the nitrided product above
1600C and preferably above 1700C.
While in examples sixteen to nineteen the product
after the nitriding operation has been removed from the ni~riding
furnace, and heated to the higher temperature in a separate
furnace, the complete heating cycle could be conducted within
the nitriding furnace with the temperature being raised direct-
ly to the final sintering temperature. However, the rate of
increase of temperature must be controlled such that the silicon
nitride of the reaction does not decompose faster than the rate
at which the silicon nitride reacts with the alumina. This is
dependent upon many parameters including final temperature,
particle size, bed geometry, bed density, heater configuration
and nitrogen potential.
Also where an entirely single phase product is
- 18 -
~ ~3~9~
required, the nitriding step should be followed in all mixes by
raising the temperature above 1700C and more preferably above
1900C, in particular to about 2000C.
. . .
- 19 - .
