CA2191771A1 - Ceramic production process - Google Patents

Abstract

A process for the production of O'-SiAlON ceramics from silicon metal, nitrogen and clay, the process comprising heating the silicon metal and the clay in a flowing nitrogen atmosphere to a temperature sufficient to react the components to form O'-SiAlON and wherein the clay participates in the reaction as a source of aluminium and silicon.

Description

W0~ 33700 2~ 77 ~ CERAMIC PRODUCTION PROCESS

Technical Field The invention comprises a process for the production of O'-SiAlON's and composite ceramic materials rrmt:lining O'-SiAlON's.

Ba.k~. ' Art The term SiAlON, or silicon ~ mininm oxynitride, ~ nll~.laacs a family of <~mrolmrl~ or phases comprised of the elements: silicon, :llnminillm~ oxygen and nitrogen. Each phase is described by a rnrnr~cirinn range for which that particular strucrure is stable. O'-SiAlON is stable over the ~ r~ range: Si2.XA!YOt+XN2-X
where 0 c x ' 0.4. O'-SiAlON has a similar structure to silicon oxynitride with equal amounts of :llnminillm and oxygen ~..i,,l;l..l;"g for silicon and nitrogen Je~ into the silicon oxynitride structure. The amounts of ~l........ ,.:.. .;... and oxygen which can be ~ d into the structure inctease with ~,................ -n..~ At 1600~C, x can be as high as 0.2. At 1900~C, x can be as high as 0.4.

SiAlON's are advanced ceramic materials which exhibit useful properties such as high strength and hardness, low density, wear resistance and corrosion resistance, and are able to retain these properties at high I ~ . .l, -- . . . ~s. SiAlON's are used in ~r.~O~
and for a variety of ,~ g ~rrli~ticm~ such as cutting tools, spray nozzles and ~ pump seals. The exact properties of a given SiAlON depend on the chemical ;011 and fabrication variables, such as purity, grain size and shape, and the 2 1 ~
~0 9~/33700 r~ .7r~
method of fPhrir~finn O'-SiAlON has similar properties to silicon oxynieride which include excellent resistance to oxidation and thermal shock. Silicon oxynitride is commonly used as a refractory material.

SiAlON's and O'-SiAlON's are known ~u~ ou~ and common methods for producing them include:

(i) Reaction Sintering. Mixtures of two or more of the following: Si3N4, SiO2, Si~N~O, Al~03 and AIN, are sintered at 2 1600~C under a nitrogen atmosphere, usually in the presence of a rare earth sintering aid such as Y2O3 or CeO7. This process involves expensive raw materials and high Ir~ m~ra~ but allows good control over the ~ and purity of the product.

(ii) G~ Ul.T~ ..,C/ Reduction. Alllminllcilir atp materials are blended with carbon and fired at - 1350~C under a flowing nitrogen ~LI~ L..r. This process is described as c~.b.,l}~ l reduction because the carbon acts by reducing the ~h~minrcilir~tP, allowing nitridation to occur. This process involves cheap raw materials and lo~ver furing u ...I.. .,o,..~ than for reacrion sinrering but impurities in the ~1. " "; ,.~ ir~tP can degrade the properties of the product. The process is difficult to control because it involves stopping a reaction at a specific point prior to ~u...pl~.ion.

(iii) ~. ' ~ Syn~hesis. A mixture containing silicon metal powder is ignited under a nitrogen d~--loa~,L..c. The energy evolved by this strongly exothermic 2~77~
WO ~.i/33700 r~
~ nitridation of silicon ~I~Jpdg.llC5 a reaction front through the resction mixture. This method is very rapid and energy efficient but is difficult to control.

Methods (ii) and (iii) both yield SiAlON powders which must then be formed and sintered to obtain a ceramic body. Method (i) is the most commonly used method for preparing O'-SiAION. As is apparent from the above known methods, in order to get good control over the . . .r. .~ ;OI~ and purity of the product expensive raw materials andlor extreme reaction conditions are required.

European patent application EP 0153000 to Kemlecott Corporation discloses a bonded material consisting of granular silicon carbide and a bonding phase ~ g Si3N4 which has been modified by the presence of oxygen and ~Inminillm Also disclosed is a method for the ~ l --- e of the bonded material which includes the use of bentonite clay as a ~ IdlJ' binder. The amount of this bentonite clay in the raw starting material is disclosed as being about 0.5% by weight. The starting materials are disclosed as being silicon carbide, ~ mini~m powder, silicon powder plus optional processing aids and temporary binders. The use of temporary binders, and thus the bentonite clay, is optional. There is no disclosure of the use of a clay material as a basic :~lnmini--m and silicon source, nor is there disclosure of the use of clay to facilitate clipr~ring or extrusion.

Canadian patent 1,156,685 to Dresser Col~ldLioll is directed to a method for producing nitride bonded refractory shapes in situ from a mixture .. ,~ ug silicon, crude clay, and a graded refractory aggregate. The intention is to form nitride phases in situ to avoid the cost of ~Jur~ hll~hlg silicon nitride for example. The amount of 21~7~
~'~'0 9513370(~ P( ~
crude clay used in the process is stated as being between 1 and 5Yo and the }efractory aggregate may be fire cday amongst other options. The refractory aggregate used in tile process is a non-reaaive bulking agenr and the crude clay facilitates ~ . n during pressing and facilitates the formation of ,B'~SiAlON and silicoll oxynitride. The examples ~see Table 1) do not form a SiAlON when reac~ing the basic ~ on~..
together. There is no disclosure of the production of O'-SiAlON. There is no disclosure of the intention to form O'-SiAlON as the major or sole product. Clay is dl;l,.~ lr added at very low levels and there is no intention to use clay as a basic source of silicon and ~ minillrn in the production of O'-SiAlON or to facilitate slipcasting or extruding.

United States patent No 43~050~ to Societe Europeenne des Produits Refractaries.
This United States patent disdoses a method for producing ,~-SiAiON's by heating in a nitrogen ~.. ~, ,~l.~. ~ ~ l. ,.. 1l ed elements obtained by drying a paste t ,, ' ' ~, a silico-aluminous material, carbon, and fine particles oÇ a ligneous material. The silico-aluminous material can be a clay such as a kaolin clay. It is essential to the procéss that carbon, preferably carbon black, and a ligneous material, such as sawdust be used. There is no disclosure of the use of silicon metal and the prooess relies on the reduction of the silica-aluminous materiai by carbon. There is no disclosure of the intention to form O'-SiAiON as a major or sole product.

EuropeanpatentapplicationO317980toTheNortonCompanydisclosesaprocessfor the mlmlf~ re of a refractory material by mixing a~ ,Li~e amounts of silicon metal powder, fine reactive alumina, a fine silica source (eg bentonite clay), a binder and a liquid tlicFercqnr forming a shape and heating the shaped body in a nitrogen wo ~s/337no 2 1~1 7 7 ;~ ~s ~
~ dLIUO~l,h~ . The bentonite clay is disclosed as being used in only a 0.5-~ wlo amount.
Fine reactive alumina is an essential component of the process. There is no disclosure of the clay c o~ Jbn..lt being an essential part of the reaction sequence or as the source of the alumina ~ulll~un.,ll.

The process of the present in~ention relies on the use of a clay material as a basic source of process ~u~lpoll~ms and the ability of the clay material to react with the other process components to form O'-SiAlON under suitable reaction conditions.
Further, the use of clay in the process facilitates the formation of shaped ceramic bodies ~unl~l;sillg O'-SiAlON.

It is an object of the invention to proYide an improved process for the n~ ~nllfl~nre of O'-SiAlON ceramic materials and composite ceramic materials ...nl ~ g O'-SiAlON.
Summary of the Invention In broad terms the inYention in a first aspect comprises a process for the IJrbdu~lioJl of O'-SiAlON from a mixture of silicon metal, nitrogen and clay, wherein the process comprises heating the silicon metal and clay in a flowing nitrogen ..LII.o~h~.c to a ~ u ~ I, . sufficient to react the ~ - ~. . .a ~ to form O--SiAlON and wherein the clay participates in the reaction as a souroe of ~ mininm and silicon.

In broad terms the invention in a second aspect comprises a process for the ~ production of O'-SiAlON from a mixture of silicon metal, nitrogen and clay, wherein the process comprises dd~yLb~yLIdllg the day, mixing the ddlyllb~ylated clay with ~'0 9~133700 2 1 ~ 1 7 7 1 ~ ~ n the silicon metal, and hea~ing the combination under a flowing nitrogen atmosphere --tu a t~ ,u.~al~lle sufficiellt tO react the mixture to form O'-SiAlOh- and wherein the dehydroxylated clay pallic;~ L~a in the reaction as a source of sllnminillTn and silicon.

Preferably the clay is the sole source of alnminillm Preferably silica is included in the mixture.

Preferably the clay is a hydrated layer alnmin(~cili~

Preferably the clay is selected from kaolin, montmorillonite, halloysite, bentonite and pyrophyllite clays.

Preferably the clay is dLhy~l~u~yl~ d by pref ~1- ining Preferably the clay is dehydroxylared by heating to a 1~,..l.. l nl"~ between about ~00~C and about 800~C.

Preferably the ~u~ Jol.~.lLb of the mixture are present as fine powders.

Preferably the mixture contains between about 5C~o and about 60% by weight of clay and more preÇerably between 10% to 40tYc by weight of clay.

~1771 ~ Preferably the mixture contains, by weight, about 20% to 50% clay, Oq/o to about 25%
silica and about 409'o tO about 609~o silicon metal and, more preferably, 30%-40% clay, 10%-20% silica, and 40%-S5% silicon metal.

Preferably the fiowing N2 ;ILlllo,~ c comprises ~ 0.5% oxygen and s 0.59/o water vapour, and more preferably ~lOppm oxygen and c25ppm water vapour.

Preferably the N2 flow rate is between about 3 and about 20 ml. min 1. g 1 by weight of silicon in the sample, and more preferably between 3 and 15 ml.min-t.g-t.

Preferably the mixture is heated to between about 1100~C and about 1900~C, more preferably between 1300~C and 1900~C, more preferably between 1300~C and 1600~C and most preferably to 1450~C to form the O'-SiAlON.

Preferably the ~.J~ lo~ are heated at a rate of up to about 10~C per minute, more preferably between about 0.2~C and about 10~C per minute, more preferably between 1~C and 7~C per minute, more preferably between 1~C and 6~C per minute, and most preferably between 2 and 5~C per minute.

Preferably the . u~ n~ are held at the required n ~ . . u ., . until the reaction is complete, preferably between about 4 and about 60 hours and more preferably betvveen about 4 and 12 hours.

~ Preferably additive ~ul~ olllld~ are included to enhance the reaction and/or to promote sintering.

WO gs/337~ 7r Preferably the additive compounds are selected from Y~03, Fe~,04, ZrO2, ZrSiO,~, MgO
and ,B-SiAlON.

Preferably the product formed contains greater than 75~b O'-SiAlON by weight of product and more preferably greater than 80% O'-SiAlON.

In more limited terms a third aspect of the invention comprises a process for the production of O'-SiAlON from a ~u~ lcll~ mixture ~ullllJI;a;llg~ by weight, 40~
60% silicon metal, 20uio-50~~6 clay, and 0%-25% silica, the process comprising the steps of:

(a) dehydroxylating the clay prior to inclusion in the ~ .o~ mixture;

(b) heating the ~,UI~IfJUn..ll mixture at a rate of 1~G to 6~C per minute, to a Itu.,~e,~lu.c of between cl~hct~nri~lly 1300~C to 1900~C under a flowing N2 aurlospllere llaving preferably 5 0.5% oxy~en and preferably c 0.5~fO water vapour and more preferably 510 ppm oxygen and '25 ppm water vapour;
(c) holding the u . ,.~ " . c between 1300~C and 19U0CC for up to 12 hours; and (d) .c~u~..;..g the product.

In broad terms the invention in a fourth aspect comprises a process for the prûduction of a composite ceramic material including O'-SiAlON from a mixture of, un.l,ol.e.ll, comprising a ceramic material and an O'-SiAlON forming l U~ 1).Ø1t collll,illdlioll of silicon metal, clay and nitrogen whereirl the process comprises heating the ceramic material, silicon metal and clay in a flowing nitrogen d~llloa~Jt~.c to a ~ u.,c WO !15~33700 2 1 g 1 ~ 7 ~
sufficient to form the O'-SiAlON composite ceramic and wherein the clay participates in the reaction as a source of silicon and ~ tliniilm.

Preferably the component combination further comprises silica.

Preferably the clay is the sole source of ~ minillm in the m inllfr~mre of the O'-SiAlON.

Preferably the ceramic material included in the mixture is selected from silicon carbide (SiC), alumina (Al203), silicon nitride (Si-,N4~, SiAlON, ~:irconia (ZrO2), ~ircon (ZrSiO.~) and silica (SiO2).

Preferably the ceramic material included in the m;xture is coarser than the other mixture components which react to form the O'-SiAlON.

Preferably the ceramic material included in the mixture will constitute up to 75~h by weight of the mixture and more preferably between about 40% and about 70% by weight of the mixture.

In more limited terms a fifth aspect of the invention comprises a process for the production of a composite ceramic including O'-SiA.ON from a mixture of ~,uUIpom,l~L:~ ~UIII~il;a;llg~ by weight, up to 75~~ of a ceramic material and 25% or more of an O'-SiAlON forming mixture, wherein the O'-SiAlON forming mixture comprises, by weight, 40%-60~~ silicon metal, 20%-60% clay, and 10%-25% silic,-"
the process ~ ull~ ;ahlg the steps of:

2~17~ ~
9sl337~ ..c-~ , ~a) heating the component mixture at a rate of between ~ I ..a .~ 1.0~C to 6~C
per minute, tO a ~ LUI~ of between cllh~r:~mi~lly 1300~C to 1900~C under a flowing N~ atmosphere having 5 0.59'u oxygen and 5 0.5~'o water vapour;
(b) holding the IL~ IUIC: between 1300~C and 1900~C for up to 12 hours; and (c) recovering the product.

Dcscription of the Drawings The attached Figures show X-ray diffrac~ion patterns of O'-SiAiON products formed by the process of the invention. The standard O'-SiAlON and SiC X-ray diffract;on XRD patterns were supplied by the International Centre for Diffraction Data, USA
and the product X-ray diffraction patterns were obtained from a Phillips 1700 Series Diffractometer controlled by Phillips APD1700 software.

In the Figures:

Fi~ I shows a ~un.udlison between the X-ray diffraction pattern for the produa formed by the process of Example 2 and a standard O'-SiAlON pattern; and Fig 2 shows a ~ulll~dl;~n between the X-ray diffractive pattern for the composite product formed by the process of Example 5 and standard O'-SiAlON and SiC
patterns.

Fig 3 shows the mass curves of the O'-SiAlON reaction using raw and dehydroxylated clays.

wo s~/33700 ~ 7 7 1 1 ~ ".~
~ Figs 4 and 5 show XRD peak intensity v t~ ule for the O'-SiAlO N reaction.

Fig 6 shows the effect of the heating rate on the O'-SiAlO N reaction profile.

Fig 7 shows XRD peak intensity v heating rate of the products of the O'-SiAlO N
reaction in Example 8.

Fig 8 shows the effect of increashlg the SiC content in the production of SiC/O'-SiAlO N l,Ullll~U~iLe:~ produced in Example 9.

Fig 9(a) and (b~ show XRD analyses of the products of the reactions in Fig 8.

Fig 10 (a) (b) and (c) show behaviour diagrams for the SiAlO N family of compounds.

Fig 11 shows XRD results of O'-SiAlO N formed in Example 10 and plotted in Fig 10.

Figs 12 and 13 show the effect of additi~es on O'-SiAlO N ~lodu~ioll.

Fig 14 shows the XRD analysis for (a) O'-SiAlO N fired with no Y203 and (b) O'-SiAlO N fired with 2wt% Y203.

Figs 15 and 16 show the effect of a second ceramic ~u~llpon~l., in the ~ludu~ ~iun of O'-SiAlO N composite products.

Fig 17 shows the XRF analyses of the samples processed in Example 16.

2i~771 WO 9513370U . ~
Fig 18 shows the composition of O'-SiAlON reaction mixtures prepared for the various clay minerals in Example 16, rhe weighl changcs during reaction, and XRD
analyses of thc products.

Fig 19 shows the effect of firing le~ ,..dml~ on the production of O'-SiAlON from Example 17.

Detailed Description of the Invention The process of the present invention is a novel process directed to the preparation of O'-SiAlON from silicon metal, clay and silica (SiO~). The raw material should be in fine powder form. The clay may be dehylllu..~ldled prior to use, however retention of the clay in its natural plastic fonrl will allow the mixture to be more readily tormed into a desired shape pr;or to firing.

The raw materials may be blended by standard techniques such as ball milling or the like as will be known in the art. Following blending, if the clay is retained in its natural form, the raw materials can be fabricated into the desired shape by traditional methods of pressing, slip casting, or extruding and the more advanced methods including isostatic pressing and injection moulding as is known in the art. These methods may involve the use of pressing aids, binders, .l;~ and lubricants for example. The mil~ture is then placed in a furnace or kiln or the like and heated under aflowingnitrogenatmospheretoL~ ,..dLul~:sgreaterthanl3004Catand~ u~ Le rate and held at this ~lu~ dlu~e until the reaction is complete. Four to eight hours is generally sufficient although longer times may be required at lower ~

WO~ 33700 ~ 773 ~ For u~uuu.~r-idl processes, to ensure complete reaction the sample may be held at ~.U~ .UU~: for times as long as 60 hours or more, although it is likely 12 hours ~vill probably be sufficient. This will depend on a number of factors, such as the size of the sample and lrll",r,.~ and dLulos~l~elic variations within the furnace, as will be apparant to a person skilled in the art. It may also be beneficial to hold the material at a lov.~er ~.u~ ule than desired to form the O'-SiAlON for a period of uy to 8 hours. This will result in a two tiered firing regime which may be desirable uullllu..~ially to allow completion of the surface nitridation reaction. This lower holding Lt~lllp.ld~Ulc will preferably be between 1100"C and 1300~C. The nitrogen flow rate should be as low as possible, but should be~ sufficient to replace nitrogen consumed by flle reaction, and maintain an atmosphere with preferably 5 0.5%
oxygen and preferably c 5% water vapour inside the furnace or kiln. To form very pure O'-SiAlON the flowing nitrogen dllllo~lJL~Ic should have as low a level of oxygen and water vapour as possible. Levels of l 0 ppm oxygen and ~5 ppm water vapour or below, are preferred. For the sample sizes as used in the examples, nitrogen flow rates of between about 3 and ~0 ml. min h g-l by weight of silicon in the sample were found to suffice. The oxygen and water vapour content of the atmosphere should be kept to a minimum as these factors can affect the process by attacking the unreacted silicon powder. During the reaction the nitrogen in the furnace atmosphere becomes incol ~,or~i~d into the product via a nitridation reaction giving an increase in density. The product formed is primarily O'-SiAlON although small amounts of silicon nitride, mullite, and other SiAlON phases may also be formed.

2 t WO 95/33700 1 ~ L J.~. l I .
Ir has been found that d~hydiv.~yLLiull of the clay prior to mixing with the silicon meral is beneficial to the production of O'-SiAlON in pure form. For example at about 500-800~C, clay undergoes this dllyllu~d.llion reaction and forms a reactive ;uu . ,,.~.I;AIr as is well known in the art. This d~hydlu~ylal;ull will also occur when the process is carried out in a single firing (ie without del,~llu~-ylation of the clay prior to mixing with the other process ~u~ ol1e.l~a). For example, if kaolin clay is used in the process then a reactive dlllol~,hous i..t~ e is formed whicll is commonly referred to as meta-kaolin. This amorphous il.~ l.lclid~e reacts with silicon metal and nitrogen from the flowing nitrogen .l~llloa~ll.le~ to form silicon Alnmininm oxynitrides ~SiAlON's) under suitable conditions. Formation of the reactive clay h.~ lcli~e facilitates the use of relatively low ~ urc~ in the process.

The d.hyllw~yld~ion of the clay prior to mixing with the other raw materials may be achieved by methods such as precalcining as will be known in the art. If the d~ Lyllu~ylatioll step is not done prior to mixing the clay v,rith the silicon metal and silica it appears that tlle water released as the clay d~h~l~u~cyL..~s on heating can partially prevent rhe silicon from reacting (see Example 6).

While the removal of water from the clay prior to mixing w ith the other ~ Jù~ a in the mixture will enhance the production of the O'-SiAlON product this step is not essential to the process. Dehydl Ul~y l~..i.,g the clay prior to mixing increases the purity of the O'-SiAlON formed in the process however the benefits of clay with regard to slip-casting and extruding for example, will no longer be available.

~i9~71 wo~sl33700 r~ Lbs.~ ~s, ~ In the process of the invention the clay material provides a source of silicon and ~I..minil1m for the reaction. While it is possible to include other sources of all.mini.-m, the process of the present invention does not require this and the clay used is preferably the sole source of l~lllminillm for the reaaion. As a result of the extensive use of clay the provision of raw material for the process is cheaper than for prior art processes and the complexity of the procedures involved in producing O'-SiAlON products is lowered.

Some reaaion between the reaaion ~u,.",on~ occurs during the heating stage, and therefore the holding step for up to eight hours is optional. E~owever, the reaaion ~rl'pl~'r:l~pc berween 1400"C and 1450~C as tl1e silicon begins to melt (at approximately 1414~C). Holding the furnace at a l..;llJ~l~I~UlC greater than 1450~C
is also optional but may be used to force the reaction to completion, or to sinter the body to obtain better densities. A heating rate of between 1~C per minute and 10~C
per minute has been found to be suitable however the heating rate should preferably be between 1~C and 7~C per minute. More preferably the heating rate should be between 1~C per minute and 6~C per minute and most preferably between 2~C and 5~C per minute. The process will however work at heating rates as low as 0.2~C per minute and this may in faa occur as, on a co,.",~ ;al scale, large furnaces can be slow to heat resulting in such lower heating rates.

The preferred tC,ll~ UlC range to which the mixture of components is heated is berween 1~00~C and 1600~C as, at higher rr. ~ r~ C~ cpeci~liced furnaces or kilns may be required. lf impurities are present in the silicon then it's melting point may be lowered and in such cases reaaion may proceed at lower n ~ cs~

O g~33700 Tc~ aLul~S as low as about 1100~G are envisaged as being possible. For example, in thc case of Fe particularly, there is a eutectic between E~e and Si at about 1190~C
which would allow the Si to melt at the lower le~ ,.dLUI~ and thus allow the reaction to proceed. While it has beell found that natural clay products containing iron impurities may be usefuliy used in the process of the invention, the presence of Ca, Na, and Ba impurities have been found to hinder the reaction. Addition of small amounts of ZrO2 also permits the reaction to proceed at lower ~ ,.. cs.

As will be readily apparent to a person skilled in the art, the type of furnace or kiln used must be able to maintain a controlled internal a~ulO~ .c at the L~ a~u~
required. Any type of furnace or kiln which is capable of this may be used.

Ceramic or ceramic composite bodies formed and reacted at 1300-1600~C can be sintered at higher u ~ (1600-1900~C~ to increase the density and reduce the porosity of the product. This sintering can be performed in a second firing or as a second, higher ~ alule~ stage of a single firing step. Sintering will be i no...l.~ni. d by an amount of shrinkage in the product.

Examples of reactions to produce O'-SiAiON (x= 0.2) from New Zealand China Clays (NZCC) premium grade halloysite and BDH light kaolin are shown in equations in (i)and (ii) le~e._Li~,ly. Theamountofeachrawmaterialmustbebaiancedtopro~ide the correct Si:Al:O:N ratio for the desired O'-SiAION as will be Imown in the art SiO2 is required to achieve the d~lO~ l;àLe stoichiometry. In some cases where clays 2~9~771 WO 9.';/33700 . ._1 " . L7~
are parricularly siliceous or contain large amounts of native free silica, it may not be necessary tO add silica tO achieve the aplJIO~ L~ stoi~hi~ltn. try, as in the case of the pyrophyllite, bentonite and rheogel clays in Example 16.

-2H2O,+N2 (1) 2(AI2O3.2.41 SiO2.2H2O) + 27 Si + 4.18 SiO2 = 2Si18AI2OI2N18 NZCC halloysite clay heat (inchlding native free SiO2) -2H2O, +N2 (2) 2(AI2O3.2.09 S;O~.2H2O) + 27 Si + 4.82 SiO2 = 2S;~BAI2OI2N18 BDH light kaolin clay heat (including native free SiO2) If the correct balance of raw materials is used then the produaion of O~-S;~ON in the resultant ceramic is ".~;",::rA This balance of material urill be able to be calculated readily by a person skilled in the art and will depend largely on the type of clay used in the reaction. The process of the present invention can produce a product cnnraining over 75~f0 O~-SiAION and is capable of producing a product ....,~ ug over 80% of O'-SiAlON. The p~ ..ta,~,e of day used can be as low as 5% by weight and, in such a case, the plupolLiollD by weight of silica and silicon metal will be d~U~ d~IY 40~~ and 5S~o ILD~.ai~ Iy. The preferred p~.~m~c makeup of the starting CO~ OII~..L mixture is however between 20~~0-45~~0 clay, 10~/0-25~/O silica and 40~/0-60~~ silicon metal.

As will be apparent to a person skilled hl the art a variety of clays may be used in the process. Clays that can be used are the hydrated layer alllm jnncili~at~c generally. The preferred clays are the kaolin clays however the pyrophyllite and the monrmnrillnnitl~

WO ~5133700 PCTINZ951000~0 clays, such as bentonite, may also be used. Other types of clay, such as illites for example, may also be used however most will conraiu a variety of impurities such as ~, Na, Ca, Mg and Fe together witll the Alnmino~ r~ content. These impurities may detrimentally affect the purity of the SiAlON product formed by the prooess of the invention. In some cases however, for example with Mg and Fe, the innpurities may have positive effect.

The process can also be used to fabricate composite ceramics containing O'-SiAlOIN.
Jn these processes the O'-SiAlON is used to bond together grains of other ceramic materials such as silicon carbide (SiC) (see Example 5), alumina ~A1203), silicon nitride (Si3N~), SiAlON, zirconia (ZrO~), zirconia (ZrSiO~) or silica (SiO~). These bonded materials take little or no part in the reaction chemistry. They will preferably be coarser than the raw materials which react to form the bonding phase, and will preferably Collstitutc up to 75g~o of the starting mixture and thus of the resultant fired ceramic. Preferably the addirional ceramic will constitute bet~veen 40% and 70% of the mixture. The additional ceramic is preferably coarser in order to get better packing of the grains prior to firing w hich allows higher fired densities ro be obtained.
Further, as the additional ceramic is coarser it is less likely to influence the O'-SiAlON
reaction. This is particularly relevant in the case of silicon carbide and silica. The additional ceramic material is bonded by a matrix of O'-SiAlON formed by the other cu~ Jon~lL~ in the starting mixture (ie the clay, silica (if present), and silicon). These other romronrnrc will therefore constitute between 25r~ and 995~o of the total starting mixture and will be present in the preferred p~ $g~ amounts that have been discussed previously.

WO 9513370U ~ 7 l 1 ~ When proceeding with the prucess of mRmlf ~rnring a composite ceramic ~ont!linine O'-SiAlON there appears to be no significant benefit in dehydlu,.ylali"g the clay material used prior to mixing the raw materials. D~ Ly d~u;~y laiion of the clay material prior to ~ u ~ ~ c of the non-composite o'-siAloN ceramic~ as has beell discussed previously, enhances the production of the O'-SiAlON ceramic by removing water which can attack the silicon and thus detrimentally affect the process. It is thought that the water content of ehe clay has less effect when l" ,.,..r~ the composite ceramic materials due to a dilution effect based on the presence of additional materials (ie the additional grains of the other ceramic materials). When using lower amounts of additional ceramic however, dllylllu,~.yldlion of the clay prior to mixing may be of benefit. This can be seen in Example 9 and in Figures 8 and 9a. The natural properties of the clay material can thus be more readily utilised to preform the desired shape of the composite ceramic prior to firing.

The reaction to form the non-composite O--SiAlON is generally ~ by an amount of shrinkage when the natural clay material is used in the process. When the clay material has been d~h~Lu~ylalcl prior to use in the process such shrinkage is reduced. When the process is used to form the composite ceramic material and natural (non-d~h~.l,u,.yl~lcd) clay is used this shrinkage can become negligible allowing near nett size shapes to be formed. As a result of the use of natural clay in the process there is a great flexibility in the shape and size of ceramic ~u~
which can be produced by the process of the present invention.

2~17~
w~s/33700 r~ ,7'. ~~, The method of the present invention is capable of producing either ceramic bodies or ceramic powder containillg the O'-SiAlON in a single firing step. As will be apparent when clay material is dehydroxylated prior to use the process will then have two firing steps The first to dcllr 11 u~r L~le the clay and the second to form the ceramic material r~lnr:lining O'-SiAlON. To make a ceramic powder the reaction will proceed without an emphasis on the ~ierlcifir~rif)n of the resulted ceramic. For example, the starting materials can be formed into pellets, reacted to form a soft ceramic pellet of O'-SiAlON which is then ground into a powder. This may be used as a supply of O'-SiAlON powder f.or use in other processes. For example the powder could be formed and sintered with or without sintering aids such as Y2O~ or the like, to form fully densified ceramic bodies.

Various additiYes can bc used in order to enhance the process of the invention or to promote sintering. It has been found that iron oxides (sucb as Fe3O~) zirconia ~zro2)~
zircon (ZrSiOf) and ,~-SiAlON promote the reaction. Yttria ~Y20~) is known to be a useful sintering aid for O' SiAlON.

o 95133700 1'~ .'0 Examples EXAA~PLE 1 A 10g mixture of: 36.9% New Zealand China Clays Premium Grade Halloysite Clay;
49.1~~o P~ as~ d 4D Silicon; and 14.0~o Superfine Quartz Powder ~supplied by Commercial Minerals Ltd);

was blended by ball-milling with ca. 400 g of 10 mm diameter SijN4 balls and 70 g of hexane in a one litre high density polyethylene (HDPE) bottle l:or 20 hours at ca.
150 rpm. The hexane solvent was removed by rotary cvapor~lLiom Samples of powder were blended with 5% oleic acid to act as a pressing aid and binder. Ethanol was used to disperse the acid, and was later allowed to evaporate off under ambient conditions. The powder was brushed through a 295 l~m sieve to granulate it, and then pressed at ca. 7~ MPa in a 20 mm diameter die. The pressed discs were heated at 0.5~C.min~l to 400~C for 30 minutes under nierogen .~ lO~p',~l~ to remove the oleic acid. Two 1.3 g discs were fred in a tube furnace on a bed of sialon granules under an ;~i...o",l.cle of "oxygen-free" nitrogen (c10 ppm ~2~ C 25 ppm H20, supplied by BOC) - flowing at cc 50 ml min-h The samples were heated at 2~C.min ' to 1450~C and held at that u .~ ollc for 2 hours.

:
2~ ~17~1 95/33700 , ~ 1~ ~7~
The discs increased in mass by ca. 18~~ during the firing, and an analysis of the products by X-ray powder diffraction (XRD) revealed primarily O'-SiAlON witll traces of B-SiAlON and X-Phase SiAlON.

EXAMPLE ~ ~

A mixnlre of: 30.7~~o deLydlo~ylaLed bDH Light Kaolin Clay (d~L~ llu~LyldLI d at 800~C for one hour);
50.1Q~o Permascand 4D Silicon; and 19.2~/o Superfine Quartz Powder (supplied by Commercial h,liilerals Ltd);

was prepared and fired as in Example 1.

The discs increased in mass by ca. 22~ during the firing, and a linear shrinkage of 4.5% was measured. An analysis of the products by X-ray powder diffraction (XRD) (see Fig 1) revealed primarily 07-SiAlON with a trace of X-Phase SiAlON. The bull~
density and open porûsity of the fired pellets were measured by evacuation and water saturation:

bulk density = 2.i)6 g.cm3 (74~,6 of theoretical) open porosity = 26.8%

7 7 ~ .
wos.~./33700 F~l.~L ..~ ~.
~ EXAMPLE 3 A 10g mixture of: 30.4% dehydroxylated BDH Light Kaolin Clay (dehydroxylated at 800~C for one hour);
49.6% P..~ dnd 4D Silicon;
19.0% Superfine Quartz Powder ~supplied by Commercial Minerals Ltd); and ~.0% Sigma Yttrium Oxide (Y20 was yrepared and fired as in Example 1.

The discs increased in mass by ca. 22% during the firing, and a linear shrinkage of 10.8% was measured. An analysis of the products by X-ray powder lliffraction (XRD~
revealed primarily O'-SiAlON with B-Si3N4 and a trace of X-Phase SiAlON. The bulk density and open poros;ty of the fired pellets were measured by ~v~.U~.~iOII and water saturation in water:

bulk density = 251 g.cm3 open porosity = 9.0%

2~L~i7~1 WO 95/337011 T.~

A mixture of: 30.4% dcLydLo~yL~Itd BDH Light Kaolin Clay ~dcl~yll~u~ylllL~d at 800~C for one hour);
4g.6% P~l,.,as~dl,d 4D Silicon;
lY.0~~ Superfine Quart~ Powder (supplied by Commercial Minerals Ltd); and 1.0n~o Pronalys AR Magn~cil~m Oxide ~MgO~;

was prepared and fired as in Example 1.

The discs decreased in mass by ca. 1.5~,'o during the firing, and a linear shrinkage of 10.8% was measured. An analysis of the products by X-ray powder rliffrarti~Ll (XRD~
revealed primarily O'-SiAlON with traces of B-Si3N~ and X-Phase SiAlON. The bulk density and open porosity of the fired pellets were measured by e ~ a~ud~iou and water c ~ l rs~ n -bulk density = 2.39 g.cm3 open porosity = 4.7%

EXAMPLE S

This example L,~cu~ dLt5 the fabrication of a composite ceramic product (SiC~SiAlON) formed by slip casting. The ~upO~Liulls of clay, silicon and silica used ~4 ~917~1 W0 95~33~00 r~ c !
~ in the starting mixtu}e being sufficient to form O'- SiAlON which bonds to the additional ceramic (SiC) to form a near netr size composite ceramic shape. A mixture of:

6095o Navarro 36-grit silicon carbide ~SiC) 14.4% New Zealand China Clays Premium Grade Halloysite Clay 19.2% Simcoa Silicon Dust (mean size = 3.5 ~m) 6.4% Pemco-32S Fused Silica (SiO~) was blended to a slurry with ca 25~~o water and 0.2% Dispex N40 dispersant in a Hobart mixer. This slip was used to cast a 425 mm long closed end tube (ca 1.6 kg).
This slip was used to cast S bars (175 x 25 x 25 mmm, ca 300 g dry weight) and S
cubes (~0 x 50 x 50 mm, ca 240 g dry weighr). The pieces were heated under flowing "oxygen free" nitrogen (45 I.min ~) to 1450~C and held at that L~ d~UIc: for 6 hours.

A shrinkage of ca 0.5% was measured during drying, but there was no significant size change during firing. Analysis of the products by X-ray powder rliffr~ n ~XEtD) (see Fig 2) revealed primarily SiC and O'-SiAlON with a trace of ,B-Si3N4. The bulk density and open porosity of the fired pieces were measured by e ~ a~uaLiOIl and water saturation. The S bars were used to determine the modulus of rupture by the 3-point method (ASTM C 133), and the S blocks were used to determine the cold crushing strengttl. A piece (50 x 8 x 8 mm) cut from one of the broken bars was used for measuring the thermal expansion, using a Harrop Labcllat~ s ,1;1 1..........

2~

21 ~:Lri71 WO '~5/337(1~ . ~, I /, .~, 7.. , ~.. ~. . ~, average bulk density = 2.60 g.cm-3 (85% of theoretical) average open yorosiry = 13.4%
modulus of rupture = 27 MPa cold crushing strength = 138 MPa thermal expansion = 2.6 x lo-6 ~C-1 from 25'~C to 100~C

3.8 x 10~.~C-1 from 25~C to 1000~C

At Lt~ Lul~s below 1600"C, O'-SLAION is stable over the following ~um~Joailio range: Si2.,,Al~OI +~N2 ~ where 0 < x ~ 0.2. Mixtures (24g) containing r~ aS~ alld 4D
silicon, Superfne O~uartz Powder (from Commercial Minerals), and either BDH
kaolin, NZCC halloysite, dchrllu~rldled BDH kaolin or dellrdiuArld~tl NZCC
halloysitc, were prepared with the correct croi~hiom~try to give O'-SiAlON with x= 0.2. For example:

2(AI2O3.2.095iO2.2H~O) + 27Si + 4.82SiO2 s 25il8AI2Ot2NI, BDH kaolin (including free quartz) These mixtures were blended by b~ll milling in I litre HDPE bottles for 20 h under hexane (170g) with Si3N~ balls (950g, 10 mm diam). The solvent was removed by rotary evaporation, and samples of the dried powders were lightly pressed (8 MPa) to form pellets (1 g, 10 mm diam.), and heated under "oxygen-free" nitrogen at 2~Cmin l to 1450~C for 2 h in a Mettler Therm~hAlstn~P The resultirlg mass curves are plotted in Figure 3.

~177~.
WO95/33700 r~ L~SI'C~
~ These mass cun~es show 3 main features:

(i) weight loss between room teul~ d~l.L~ and 500~C.
(ii) an initial nitridation reaction beginning at ca 1200~C, which by ~ u~ alison with the well known silicon nitride reaction bonding process, is probably due tû surface nitridation ûf the silicon powder.
(iii) a rapid nitridation reaction at cn 1~00~C, comparable with that observed in the silicon nitride reaction bonding process.

The samples containing raw (non-dehydroxylated) clays showed a weight loss at c~
SOO~C consistent with the loss of structural water from the clays.

The initial nitridation reaction at 1200-1300~C is stronger for the dehydroxylated clays~andappearstobesuppressedforthellon-dellylllu~yla~edclay~particularlywhen ball-milled. This is probably due to a water-silicon reaction during milling or the early stages of firing (<500~C), giving the silicon an ;~ . ."~r~ oxide coating which prevents surface ni~ri~:~ril~n. SUI~ Jn of the surface n;~l;d~iOII reaction was also aL.ull.~,~l,ied by the formation of unreacted silicon nodules in the product, and therefore should be avoided where possible. These nodules can be greater than I mm in diameter and, once they form, have such a low surface area that if any nirrirl .rion does occur it is extremely slow.

~ () g~/3370(l 2 ~ ~ ~ 7 ~
EXAh,fPLE 7 Samples of the reaction mixtures containing raw kaolin and dcLydloAyldLed kaolin prepared as described in Example 6 were heated at 2~C min ~ under nitrogen to various Lcally~ad~ultS betweell room Ir~ 0ll c and 1450~C. XRD analyses of the products are shown in Figures 4 and 5. The intensities of the strongest XRD peaks for all identified phases are plotted on a log scale. The results are non-~ud--LiLaL;Y~

butgi~eanindicationofthed~,~,cc.ldll~eand~:l;,dyycdldllceofdif~erentphasesduringthe nitridation reaction. For simplicity, a-silicon nitride and a-SiAlON, and ,B-silicon nitride and ,~-SiAlON have been grouped together as a-phase and ~-phase, Icayc. iivrl~. Notable features of these graphs shown in Figures 4 and 5 are:

(i) the appearance of O'-SiAlON by 1250~C, (ii) the ~li 7dyy~ dnce of quartz by 1 350~C, w;thout the dyy~ul .In~e of ~ric~ .h~lit.-, (iii) the appearance and disdyy,c..d~ of mullite and a-phase between 1050 and 1450~C, except in Figure ~ where some mullite remained in the final producr, (iv) the dlJpedl.1llcc of ,B-phase at 1~50-1350~C, ~v) tbe appearance of X-phase at 1350-1450~C, and (vi) the residual silicon in the final product vvhen using raw kaolin (see Example 6).

The a-phase probably represents the silicon nitride surface coating which forms on the silicon grains early in the nitridation reaction and is associated with the weight gain at 1200-1250~C (see Example 6). Alpha-silicon nitride is the expected product from this surface reaction. Thii Q-phase disappears at higher t~ ~lly~dLulc~ and is probably reabsorbed into the melting silicon and reformed as,B-phase, X-phase and O-SiAlON.

2~1 771 wo gs/33700 r~ L~
In the sample containing raw kaolin, where the 1200-1250~C weight gain is ~u~Lesscd, less ~-phase was expected, and was observed, but the difference is not great.

EXAI~PLE 8 Samples were fired at different rates to examine the effect of the heating rate on the reaction chemistry.

The O'-SiAlON reaction mixture containing dchyd~ yl.lL~ :I kaolin as used in Examples 6 and 7 was lightly pressed (8 MIPa~ to form discs (1 g, 10 mm diam~, and heated under "oxygen-free" nitrogen at 0.S, 2, 4, 6, ~ and 10~Cmin l, to 1450~C for 2 h in a Mettler Therm-)b-l~nre The resulting mass curves are plotted in Figure 6, and XRD analyses of the products are shown in Figure 7.

As the heating rate increases towards 5~C.min~l, the relative amounts of mullite"l~-phase and X-phase impurities in the products decline, and then begin to increase again at higher heating rates. At heating rates greater than 6~C.min-', nodules of residual silicon begin to appear in the product, and this can again be associated with the su~ n of the surface nitridation reaction. The overall reaction proceeds so rapidly that this surface reaction does not have a chance IO occur. These results suggests that the ideal heating rate is ca 5~C.min~k 2~

wo ~sl3370~1 2 1 ~ ~ 7 ~ .~3 .

Effect of SiC on the Nitridation Reaction The O'-SiAlON reaction mixrure containing raw kaolin (as used in Example 6 and 7), was blended with 0%, 50% and 80% silicon carbide (H C Starck, A20, 0.5 jum) in a mortar and pestle, and the resulting mixtures were heated at 2~Cmin-1 to 1450~C for 2 h under "oxygen-free" nitrogen in a Mettler Thermrib~l~nre The resulting mass curves are given in Figure 8, and the results of XRD analyses of the produas are given in Figure 9a.

As the proportion of SiC in the mixture increases, the amount of residual silicon in the product decreases (Figure 9a), and the weight gain at ca 1200~C associated with surface nirrid~rirn increases (Figure 8). If a reaction between the water of dehydroxylation and the silicon is p~ ,.lL;llg the surface nitridation of the silicon, then thc presen- e of SiC may be acting to dilute the mixture and lower the water vapour pressure, or it may itself oe reacting with the water vapour. The addition of alumina in place of SiC had a similar effect, suggesdng thar ir is simply a dilution effect.

As the FJ1 OPOL ~ion of SiC in the mixture increases, the pLOpOI ~ ns of a-phase and ,B-phase impuritics in the products also increase cignifir~lnily, This implies that the SiC
is actively participating in the reaction~ and therefore tlle effect should decline with 7 ~
WO ~15/33700 1~, 1 /lv; rS)t - - ~
coarser grades of SiC. To confirm this, the O'-SiAlON reaction mixture contain;ng raw kaolin was blended with different sizes of silicon carbide: 0.5 ~m (H C Starck A20), 32 ,um (Navarro C5 80 grit) and 180 ~m (Navarro CS 400 grit); and fired asabove. The results of the XRD analyses of the products are given in Figure 9b.

As expected, as the size of the SiC increases it becomes less reactive and the proportion of ~-phase formed in the product decreases.

Note that small samples fired hl a Mettler Thermobalance experience very fresh gas and this tends to enhance the amounts of impurities such as ~3-phase and mullite in the products. This is a size effect, and in production scale futnaces it is negated by the sheer bulk of sample.

This example illustrates the range over which the proportions of clay, silica and silicon can be varied, while still producing ~ dc,.~ ly O'-SiAlON as the product.

At t~rnr~-r~tllres below 1600~C, O'-SiAlON is stable over the following ~
range: Si2.,A]XO, .,.~N2.~ where 0 ~ x c 0.2. This range of ~ ~ ..l,o- ;o~.~ is .~.c ~e..~cd in the SiAlON behaviour diagram (Figure 10).

The behaviour diagram in Figure 10 has been constructed by reacting 3 or 4 c ~ JollcuL mixtures of SiO2, Si3N4, Al203 and AIN at 1700~C. In the case of the ~ 177 1 U'0 9~/3371\0 r~
present invention a different process is used and the firing l~ Lu-c is lower, so rhe behaviour d;agram rmay uot apply.

Mixtures containing NZCC halloysite clay, Permascand 4D s;licon and Commercial Minerals superfine quartz powder in varying ~lopulliolls~ were hand blended in a mortar and pestle under ethanol, and the solvent was removed by ~V.IpOI~iOll. The theoretical fired ~ ~u~ o~;l;ons of these mixtures are plotted on the SiAlON behaviour diagram tFigure 10a). Note that the process of the present invention is restricted to the triangular section of the diagram defined by the raw materials; SiO2, clay (in the case of this example, NZCC Halloysite) and Si (equating to Si3N~). Ille dried powders were lightly pressed (8 MPa) to form pellets (10 mm diam., 2g), and heated under "oxygen free" nitrogen at 2~Cmin-1 to 14.~0~C for 2 h in a tube furnace. The products were analy~ed by X-ray Powder Diffraction (XRD), and the amount of O'-SiAlON in each of the products was estirnated from XRD peak heights. The results are presented in Figure 11 (sample codes refer to ~nmrn~;~inns indicated in Figure I Oa).

All of the samples were completely reacted in that no unreacted silicon remained in any of the products. The products formed from each mix, and their relative proportions, were largely consistent with what was expected from the behaviour diagram. The estimated percentage yields of O'-SiAlON are plotted in Figure 10b.
From this plot, the target composition needs to be within the range: Si~.,.l.8,AI

0~333~O(0~8~5olN~Is~208~ to obtain 2 80% O'-SiAlON. The exact amounts of silicon, silica and clay needed to achieve this will depend upon the exact rnmrocirion of the particular clay be;ng used.

wo 95/33700 ~ 7 7 ~ L~.,. ~ .
~ EXAMPLE I ]

The effects of various additives on the formation of O'-SiAlON were i"~ ~.Li~ LC~d to assess their potential as catalysts or sintering aids, and to examine the effect of common clay impurities such as Ca and Na.

A standard O'-SiAlON reaction mixture was prepared by ball-milling Permascand 4D
Si, dehydrated NZCC halloysite clay and superfne quartz powder for 20 h under hexane with Si3N4 balls as described in Example 6. Samples of this mix (5.8g) were remilled for 4 hours with BaO, CaO, ZrO2, Na2CO3, Fe304 and Y203 at a level of 1.5 mol9~ of cation by theoretical weight of O'-SiAlON product. The dried powders were lightly pressed (8 MPa) to form pellets (1 g, 10 mm diam.), and heated under "oxygen free" nitrogen at 2~Cmin l to 1450~C for 2 h in ,3 Mettler therm~h~310nre. The products were analyzed by XRD. The results are presented in Figure 12, and mass curves are plotted in Figure 13. Of the additives that were tried, only Fe304 and Zr~2 promoted reaction. Na2CO3, BaO and CaO appeared to hinder the O'-SiAlON
reaction, and large amounts of unreacted silicon remained in the products. One possibility is that the Ba, Ca and Na are fluxing the clay at such a low u ..,I...-.....
that the pellet sinters before it can nitride. Nitrogen is then unable to get to the silicom In support of this is the observation that these three pellets showed the greatest shrinkage during firing (see Figure 12).

Yl03 appeared to retard the reaction slightly, but the XRD analysis indicated that it gave a more pure O'-SiAlON product (Figure 14). Significant shrinkage during firing 219 1~ rl ~
WO ~)513370U . ~
(14.4%) was also observed. It may therefore be valuable as a sintering aid for the O'-SiAlON reaction.

O'-SiAiON ~u~ JO~ 5 were hlv~;a;~dL~d by preparing and firing mixtures of the raw O'-SiAiON mix with various ceramic materials. These were combined in a ratio to give a 50 50 mix of O'-SiAiON and the second ceramic ~ Fine powders were used to highlight possible iilLt:l[.~ C of the second ceramic ~u,.-,uO~ in the O'-SiAlON reaction.

A standard O'-SiAlON reaction mixture was prepared by ball-milling P~L.l.a~..d 4D
Si, halloysite clay and superfine quartz powder for 20 h under hexane with Si3N~ balls as described il Example 6, Samples of this mix were hand-blended with alumina (Buehler 0.3~m y-Al~O3), silicon nitride (HC Starck LC10), sialon (Bensalon S-011), zircon (opaciline) and zircollia (Tosoh rz-o) under ethanol using a mortar and pestle.
The dried powders ~were lightly pressed (8 MPa) to form discs (1 g, 10 mm diam.), andheatedundernitrogenat2~Cmin-ttol450~Cfor2hinaMettlerThl,,...~l..,l-..l~

The products were analy~ed by XRD. The results are presented in Figure 15 and mass alrves are plotted in Figure 16.

In each case, the reaction yielded p~edu~t~h~ ly O'-SiAlON, ie none of the ceramic materials that were tried cignifir!~nrly hindered the formation of O'-Si41ON. The possible excep~tion was A12O3 which increased the amount of ,B-phase in the product.

~1~ 1 7~.1 WO 9S133700 P~
~ This may not occur with coarser Al2O3. In contrast, the ,B-SiAlON, ZrO2 and ZrSiO4 greatly Icc~ r~t~d the reaction to form O'-SiAlON (Figure 16).

This example demonstrates the ~b~ iu-l of a composite ceramic product (SiC/O'-SiAlON), formed by extrusion.

To a mixture of: 50% Navarro 220-grit silicon carbide (SiC);
16% New Zealand China Clays Premium Grade Halloysite Clay;
2% Bentonite SM23 supplied by Commercial Minerals;
24~'o Simcoa Silicon Dust (mean size = 3.5 ~m); and 8% Pemco - 325 mesh Fused Silica (SiO~);

was added sufficient pressing soap (a mixture of water, kerosine, oleic acid and TEA
as is known in the art) to give a u~ ..l. y suitable for extrusion.

This mix was used to extrude 5 cylindrical rods (25 mm diam x ca 200 mm, ca 275 g dry weight). These rods were heated under flowing "oxygen free" nitrogen ~45 I.min l) to 1450~C and held at that u ~ e for 6 hours.

Analysis of the products by X-ray powder diffraction (XRD) revealed primarily SiC
and O'-SiAlON with a trace of X-phase SiAlON. The bulk density and open porosity of the fired rods were measured by c. ~u~dOII and water cqt~lrqtion. The modulus of rupture of the S rods was measured by the 3-point method.

7 ~
WO 95/3370tl P~, I /I~L.7.!.~, ~
average bulk density = 2.28 g.cm3 (76% of theoretical) avcrage opcn porosity = 23.3%
modulus of rupture = 71 MPa This is an example of O'-SiAlON reaction bonded body. which is sintered in a second firing to form a fully dense body.

A standard O'-SiAlON reaction mixture was prepared by ball-milling Pc ....as.dud 4D
Si, dehydroxylated BDH kaolin clay and superfine quartz powder, as described in Example 6. A sample of the dried powder was formed into a bar (62.6 x 8.8 x 7.0 mm, 4.98 g) by uniaxial pressing at 25 MPa. The bar was fired in a graphite resistance furnace on a bed of SiAlON granules under a flowing "oxygen free" nitrogen d~llloa,,l-~.~ at 2~C.min~l to 1450~C and held at that ~lll,U..dLu~ for 2 hours.

During reaction bonding, the bar increased in mass by ca 22q~o, and a linear shrinkage of 7.7% was measured. The bulk density and open porosity of the sintered bar was measured by e.a.uc~ and water saturation:

bulk density = 2.11 g.cm-3 (75~~ of theoretical) open porosity = 24.8%

The bar was then sintered by firing in a Si3N., powder bed undcr "oxygen free"
nitrogen in a graphite resistance furnace to 1700~C for I hour.

WO !~S133700 2 ~ ~ ~ r17 ~
~ During sintering, the bar decreased in mass by ca 13~o, and a linear shrinkage of 8.7~,'o was measured. An analysis of the products by X-ray powder diffraction (XRD) revealed O'-SiAlON with faint traces of mullite and X-phase SiAlON. The bulk density and open porosity of the sintered bar was measured by evacuation and water saturation. Thermal expansion was measured using a Harrop Labo.~llories dil~loult:L~I. The modulus of rupture was measured by 4-point method.

bulk density = 2.80 g.cm3 (100~o of theoretical) open porosity = 0.0%
thermal expansion = 2.2 x 106.~C-I from 25 to 100~C
3.3 x l06.oC-~ from 2j to 1000~C
modulus of rupture = 375 MPa This is an example of fully dense O'-SiAlON body prepared by sintering preformed O'-SiAlON powder.

A standard O'-SiAlON reaction mixture was prepared by ball-milling r~. I..a ,.d.ld 4D
Si~ d~ hydlo~yla~td BDH kaolin clay and superfine quart~ powder, as described in Example 6. The dried powder (23.42g) was blended with water (14.35g, 38%), extruded through an 8 mm diameter hole, dried, and broken into pellets ca 10-25 mm in length. A 10 g sample of these pellets was fired in a 50 mm diameter tube furnace on a bed of SiAlON granules under flowing "oxygen free" nitrogen (ca 150 ml.min ') at 2~Cmin-1 to 1450~C and held at that ~cm~ ulc for 2 hours. A weight gain of 21~7~ i 9513370~) r~ lfi~-7~
23.9% was observed, and the product was found to be primarily O'-SiAlON with rraces of mullite and ,B-phase. The fired pellets were crushed to 1.5 mm in a WCpercussion rnortar, and ball-milled in à I litre HDPE bottle for 20 h under ethanol (170 g) with Si3N4 balls (g50 g, 10 mm diam). Oleic acid (5% by weight of O'-SiAlON) was added and the mixmre milled for a further 15 min. The milled slurry was filtered througll a 10 ~m filter cloth, and the solvent was removed from thefiltrate by rotary evaporatiom The dried powder was brushed through a 295 ~m sieve to granulate it. A sample of the powder was formed into a bar (S8.5 x 8.2 x 6.0 mm, 4.68 g) by uniaxial pressing at 25 MPa, followed by cold isostatic pressing at 350 MPa. The bar was heated at 0.5"C min l to 400~C for 30 min under nitrogen ILlllua~ c to remove the oleic acid, and sintered in a Si3N4 powder bed under "oxygen free" nitrogen hl a graphite resistance furnace to 1700~C for 1 hour.

During sintering, the bar decreased in mass by c~ 11%, and a linear shrinkzLge of 18%
was measured. O'-SiAlON was the only phase detectable by X-ray powder diffraction (XRD). The bulk density and open porosity of the sintered bar was measured by evacuation and water saturation. The modulus of rupture was measured by the 4 point method.
bulk density = 2.79 gcm~3 (100~~ of theoretical) open porosity = 0.0'Yo modulus of rupture = 424 MPa ~'77 WO~133700 2 1 ~ ~ P~

This example illustrates the variety of clay minerals that can be used to prepare O'SiAlON.

Nine clay mineral samples were selected and analysed for major oxides by X-Ray fluorescence ~XRF) ~Figure 17). These minerals were calcined at 1000~C for 1 hour, and hand-ble[lded in a mortar and pestle under ethanol, with d~ upl idLc~ amounts of Perm ~ nd 4D Si and Commercial Minerals superfine quartz powder to give O'-SiAlON (Si2 XAIXO~+xN2 x) with x = 0.2 where possible. In some cases, the Al203:SiO2 }atio in the mineral was too low to form x = 0.2 O'-SiAlON. The dried powders were lightly pressed ~8 MPa) to form discs ~0.5g, 10 mm diam), and fired in a tube furnace at 2~C min ~ to 1450~C for 2 h on a bed of SiAlON granules under a flowing a~ ua~ .c of "oxygen free" nitrogen (30 ml.min~lg~~). The weight change was measured, and the products were analysed by X-ray powder diffraction ~XRD) ~see Fig 18).

This example illustrates the range of firing t. lllp~ldLu.~ s which can be used to form O'-SiAlON.

The O'-SiAlON reaction mixture containing deL~J~u~yLI~d kaolin as used in Examples 6 and 7 was lightly pressured ~8 MPa) to form discs (0.5 g I0 mm diam.).
Discs were heated on a bed of SiAlON granules, under a flowing "oxygen free"

~ ~ 7~ ~
~'0 9~/33701~
nitrogen atmosphere ~50 ml.min-hg-1) at a heating rate of 2~C.min l to the following t~ nUlCD. 1175~C, 1200~C, 1250~C, 1300~C, 1350~C and 1450~C, and held at those lru~p~.dLulcS for 8 hours. The results from XRD analyses of the products are shown in Figure 19.

In general, discs fired at higher ~ LdLuLc., gave more pure O'SiAlON. Discs fired for 8 hours at ICIII~ d~ ,D less than 1300~C did not react to completion; unreacted silicon remained in the products. However the reaction may go to completion at Lclll~ d~ulcS lower tharl 1300~C if the furnace is held at LCul~,.dLulc for periods longer than 8 hours.

The foregoing describes preferred forms of the invention and it is to be nn~l~r~ood that the scope of thc inventioll is not to be limited to the specific fomls described.
Modifications and variations as will be obvious to a person skilled in the art may be made tO the forms of the invention as described without departing from the spirit or scope of the invention as defined in the attached claims.

Claims (33)

1. A process for the production of O'-SiAlON from silicon metal, nitrogen, and clay, the process comprising heating the silicon metal and the clay in a flowing nitrogen atmosphere to a temperature sufficient to react the components to form O'-SiAlON and wherein the clay participates in the reaction as a source of aluminium and silicon.

2. The process according to claim 1 wherein the clay is the sole source of aluminium.

3. The process according to claim 1 or claim 2 comprising dehydroxylating the clay and mixing the dehydroxylated clay with the silicon metal and heating the combination under a flowing nitrogen atmosphere to a temperature sufficient to react the mixture to form O'-SiAlON,

4. The process according to any one of the preceding claims further comprising the addition of silica.

5. The process according to any one of the preceding claims wherein the clay is a hydrated layer aluminosilicate mineral.

6. The process according to claim 5 wherein the clay is a kaolin, montmorillonite, pyrophyllite, bentonite or halloysite clay.

7. The process according to claim 3 wherein the clay is dehydroxylated by precalcining.

8. The process according to claim 3 wherein the clay is dehydroxylated by heating the clay to a temperature between about 500°C and about 800°C.

9. The process according to any one of the preceding claims wherein the components of the mixture are present as fine powders.

10. The process according to any one of the preceding claims wherein the clay is present in an amount of between about 5% and about 60% by weight.

11. The process according to claim 10 wherein the clay is present in an amount of between about 20% and about 60% by weight.

12. The process according to any one of the preceding claims including by weight, about 20% to about 50% clay, 0% to about 25% silica and about 40% to about 60% silicon metal.

13. The process according to any one of claims 4 to 12 including about 30% to about 40% clay, about 10% to about 20% silica, and about 40% to about 55%
silicon metal.

14. The process according to any one of the preceding claims wherein the flowing N2 atmosphere comprises ~ 0.5% oxygen and ~ 0.5% water vapour.

15. The process according to any one of claims 1 to 13 wherein the flowing N2 atmosphere comprises ~ 10 ppm O2 and ~ 25 ppm water vapour.

16. The process according to any one of the preceding claims wherein the mixture is heated to between about 1100°C and about 1900°C.

17. The process according to any one of the preceding claims wherein the heating rate is between about 0.2°C and about 10°C per minute.

18. The process according to claim 17 wherein the heating rate is between about 2°C and about 5°C per minute.

19. The process according to any one of the preceding claims wherein the components are held at the required temperature for between about 4 and about 60 hours.

20. The process according to any one of the preceding claims wherein additive compound are included to enhance the reaction and/or to promote sintering.

21. The process according to claim 20 wherein the additive compounds are selected from Y2O3, Fe3O4, ZrO2, ZrSiO4, MgO and .beta.-SiAlON.

22. The process according to any one of the preceding claims further comprising the addition of a ceramic material to produce a composite ceramic including O'-SiAlON.

23. The process according to claim 22 wherein the added ceramic material is selected from silicon carbide (SiC), alumina (Al2O3), silicon nitride (Si3N4), SiAlON, zirconia (ZrO2), zircon (ZrSiO4) and silica (SiO2).

24. The process according to claim 22 or claim 23 wherein the ceramic material is coarser than the other mixture components which react to form the O'-SiAlON.

25. The process according to any one of claims 22 to 24 wherein the ceramic material is present in an amount of up to about 75% by weight of the mixture.

26. The process according to claim 25 wherein the ceramic material is present in an amount of between about 40% and about 70% by weight of the mixture.

27. A process for the production of O'-SiAlON from a component mixture comprising, by weight, 40%-60% silicon metal, 20%-45% clay, and 10%-25%
silica, the process comprising the steps of:

(a) dehydroxylating the clay prior to inclusion in the component mixture;

(b) heating the component mixture at a rate of 1°C to 6°C per minute, to a temperature of between substantially 1300°C to 1900°C under a flowing N2 atmosphere having ~ 0.5 oxygen and ~ 0.5% water vapour;

(c) holding the temperature between 1300°C and 1900°C for up to 12 hours; and (d) recovering the product.

28. A process for the production of a composite ceramic material including O'-SiAlON from a mixture of components comprising a ceramic material, and an O'-SiAlON forming combination of silicon metal, clay and nitrogen wherein the process comprises heating the ceramic material, silicon metal and clay in a flowing nitrogen atmosphere to a temperature sufficient to form the O'-SiAlON
composite ceramic and wherein the clay participates in the reaction as a source of aluminium and silicon.

29. The process according to claim 28 wherein the O'-SiAlON forming combination further comprises silica.

30. The process according to claim 18 or claim 29 wherein the clay is the sole source of aluminium for the production of O'-SiAlON.

31. The process according to claim 18 or 19 wherein the ceramic material is selected from silicon carbide (SiC), alumina (Al2O3), silicon nitride (Si3N4), SiAlON, zirconia (ZrO2), zircon (ZrSiO4) and silica (SiO2).

32. A process for the production of a composite ceramic including O'-SiAlON from a mixture of components comprising, by weight, up to 75% of a ceramic material and 25% or more of an O'-SiAlON forming mixture, wherein the O'-SiAlON forming mixture comprises, by weight, 40%-60% silicon metal, 20%-45% clay, and 10%-25% silica, the process comprising the steps of:

(a) heating the component mixture at a rate of between substantially 1°C
to 6°C per minute, to a temperature of between substantially 1300°C to 1900°C under a flowing N2 atmosphere having ~ 0.5% oxygen and ~ 0.5%
water vapour;

(b) holding the temperature between 1300°C and 1900°C for up to 12 hours; and (c) recovering the product.

33. An O'-SiAlON containing product when produced by the process of any one of the preceding claims.