US3692478A - Method of manufacturing silicon carbide whiskers - Google Patents

Abstract

WHISKERS OF APPROXIMATE DESIRED DIAMETER ARE GROWN ON A UNIFORM ROUGH SURFACE OF A SUBSTRATE BY MEANS OF A NUCLEATING SUBSTANCE DUE TO THE PRESENCE OF CRYSTALLITES PROJECTING FROM THE SURFACE THE LATRAL DIMENSIONS OF

WHICH AT THE SURFACE APPROXIMATELY CORRESPOND IN SIZE TO THE THICKNESS OF THE CRYSTALS TO BE GROWN.

Description

p 1972 w. F. KNIPPENBERG ETAL 3,692,478

METHOD OF MANUFACTURING SILICON CARBIDE WHISKERS Filed Feb. 2, 1970 INVENTOR. W. F. KNIPPENBERG G.VERSPUI 0 v /Mk United "States I)atent @1 hoe 3,692,478 Patented Sept. 19, 1972 9 1 Int. Cl. C01b 31/36, 21/06; Clllf 7/02 US. Cl. 423-345 2, Claims ABSTRACT OF THE DISCLOSURE Whiskers of approximate desired diameter are grown on a uniform rough surface of a substrate by means of a nucleating substance due to the presence of crystallites projecting from the surface the lateral dimensions of which at the surface approximately correspond in size to the thickness of the crystals to be grown.

The invention relates to the manufacture of whiskers.

Whiskers are to be understood to mean herein crystals having a maximum average thickness of 100 ,um. the length of which is at least 10 times larger than the thickness.

As is known, whiskers can be formed by growing in a gaseous phase on a substrate, a strong anisotropic crystal growth being promoted by nucleation of the substrate, that is to say such a treatment that crystallisation nuclei are formed on places on the substrate surface separated from each other.

As is known, elongated crystals can be obtained in all kinds of dimensions by using a substance, a so-called nucleant, which stimulates the formation of a local crystallisation nucleus. This is in particular the case in the manufacture of crystal whiskers by VLS-growth.

VLS (Vapour-Liquid-Solid) crystal growth has already been described in Transactions of the Metallurgical Society of A.I.M.E. 233. (1965), 1053. In VLS crystal growth a substance to be crystallized (or its components) is taken up from a gaseous phase in drops of a substance nucleant locally provided on a substrate and in which the substance to be crystallized is soluble, and said substance is deposited via the drops on the substrate in the form of elongated crystals. This method is, of course, useful for growing whiskers and, as is known, has been applied for manufacturing whiskers of different substances.

For that purpose, a substrate is nucleated by sputtering or vapour-deposition of a substance, from which upon heating prior to or during the crystal growth very fine drops are formed which can each be active separately, as a liquid phase necessary for VLS-growth.

Cases of whisker growth are also known in which nucleation of the substrate surface takes place via the gas atmosphere. It is sometimes found, as is also found in the case of VLS-growth, that the grown crystals, may comprise at their free end a spherical part which consists at least partly of the nucleant used, so that it may be assumed that in this case also VLS-growth has taken place, although, no VLS-growth has been carried out deliberately.

The formation of a spherical part at the ends of the crystals may be found sometimes, even in whisker growth, in which a nucleant was provided on the substrate neither by sputtering or vapour-deposition, nor by the supply via the crystallisation atmosphere. It may be assumed, that in these cases also VLS-growth might have taken place, in which crystal nuclei have formed from impurities present in the starting material for the crystals, in material from which the substrate or the apparatus are constructed and/ or in the gas atmosphere.

Moreover, there Were cases, in which either intentionally or unintentionally nucleation of the substrate has been produced with as a source the impurities present in the apparatus whereas no spherical part containing a nucleant was observed at the free end of the grown crystals. However, also in these cases the possibility of a nucleation and crystal growth mechanism as in VLS- growth need not be out of the question, because in cases in which VLS-growth has been identified in some of the crystals formed the characteristic spherical ends are lacking. This may be a result of evaporation and/or reaction in the crystallisation atmosphere.

Finally the possibility exists that the nucleant only fulfils a function at the beginning of the crystal growth and is certainly not active as a liquid phase in a VLS-growth during the whole crystal growth.

In all the above-described cases of crystal growth with intentional or unintentional nucleation of the substrate, it is required, in order to obtain a good yield of thin whiskers of slightly varying thicknesses, that the substrate should be densely populated with crystal nuclei of equal size.

Of course, this requirement cannot be fulfilled as such if the nucleation of the substrate is produced by impurities accidentally present.

However, in many cases, also the intentional addition of a nucleant via the gaseous phase does not open the possibility of controlling the thickness of the crystal, and often results in varying thicknesses and non-uniform population of the substrate with crystallisation nuclei.

In those cases of forming whiskers in which VLS- growth has been identified and in which the substrate is nucleated with drops of a nucleant, in the conventional manner of providing the drops, a dense and uniform population of drops of equal size cannot be realized as such.

For example, by powdering or smearing a substrate with a finely divided substance of equal grain size, in a subsequent heating step a dense uniform population of the substrate with drops of equal size is not obtained as such in all circumstances. Actually, upon heating, powdered particles which are in contact with each other may flow together, so that in addition to drops which correspond in volume to that of the provided grains, larger drops will also be formed. If nucleation is effected by vapourdeposition of a substance, suitable for VLS-growth, followed by heating, the vapour-deposited layer will divide in individual drops but usually the sizes of these drops will vary and, moreover, may form an irregular, poorly dense pattern of crystallisation nuclei on the substrate.

It is an object of the present invention to avoid the above-mentioned drawbacks associated with the nucleation of substrates for growing whiskers.

The invention relates to a method of growing crystals in the form of whiskers on a substrate in a gaseous atmosphere containing the substance to be crystallized or its components, in which strong anisotropic crystal growth is stimulated by nucleation of the substrate, characterized in that the crystals are grown on a substrate which is rough at least superficially due to the presence of crystallites projecting from the surface the lateral dimensions of said crystallites at said surface approximately corresponding in size to the thickness of the crytsals to be grown.

As a result of this uniform rough structure of the substrate surface it is achieved that, independent of the fact whether the nucleating substance is supplied in a solid, liquid or vapour form, a dense and uniform nucleation of the substrate will always take place.

The nucleant may be applied in an elementary form but in that case variations in the choice of the crystallisation circumstances in connection with the volatility of the nucleant, that is to say in the choice of the temperature, the partial pressure of the substance to be crystallized or its components and the overall pressure of the gas atmosphere, are limited.

In many cases, however, it is of advantage to supply the nucleant in the form of a compound which can be converted into a nucleant by disproportioning or other chemical reaction because in that case various substances of a varying volatility are available as a result of which there is a greater possibility of adapting the crystallisation conditions in a correct manner.

Also for replenishing the nucleation via the gas atmosphere during the crystal growth, in so far as the nucleant, necessary for the crystal growth, is fully or partly lost by side reactions or evaporation and dissipation in the gas atmosphere, it is of advantage, for the same reasons as mentioned above, to use the nucleant in the form of a compound which can be converted into a nucleant by disproportioning or by the other chemical reaction. D

In the cases in which the nucleation of the substrate is carried out via the gaseous phase, nothing can be established visually as regards the density and uniformity of the nucleation of the polycrystalline substrate. Only the uniformly dense growth of whiskers of slightly varying thickness can be established. I

Clearly visible, however, is the density and uniformity of the more macroscopic nucleation as is used in intentional VLS-growth.

When the nucleant or a compound which can be converted into a nucleant is provided on the substrate by vapour-deposition or by sputtering, the formed continuous layer will be divided into separate drops upon heating and form a dense uniform pattern of drops Wind: is determined by the uniformly rough structure of the substrate surface.

This is the case also when the nucleant or a compound thereof is provided on the substrate in a finely divided form by spraying or smearing. Upon heating, mainly a pattern of drops is formed which is determined by the lateral dimensions of the crystallites in the plane of the substrate surface. This result is even promoted by providing the nucleant in the form of powder of slightly varying grain size and particularly having a gram size which corresponds at most to the said transverse dimensions of the crystallites.

Of course the substrates should be constructed from a material which is resistant to the gas atmosphere in which and the temperature at which the crystal growth is carried out.

Substrates which consist at least superficially of polycrystalline material of the above crystallite size can be manufactured by sublimation or by gas reactions in which, of course, parameters, such as vapour pressure and temperature, should be suitably chosen so as also to obtain the required roughness.

Substrates are preferred, however, which consist of powder agglomerates which can be obtained by sintering, pressing or adhering with a binder in any suitable form, since in the manufacture thereof the crystallite dimensions and the surface roughness can easily be controlled by the choice of the grain size and the grain shape of the starting material.

If, however, there has been proceeded so that a porous material is obtained, this provides the additional possibility that the supply of the nucleant and of the substance to be crystallized or its components may be carried out via a gaseous phase through the substrate.

Finally the manufacture of cohering powder agglomerates provides a favourable possibility of forming nucleated substrates in one operation by starting from a powdered substrate material, in which the nucleant or a compound which can be converted to a nucleant is distributed evenly in the form of a powder having a grain size corresponding to the thickness of the crystals to be formed.

A particular advantage of powder agglomerates as substrates is that it can simply be obtained by means of a suspension of the raw materials on a support.

The invention will now be described in greater detail with reference to the accompanying drawing and a few specific examples.

EXAMPLE 1 A plate of 25 x x 2 mm. is formed by sintering silicon carbide powder having a grain size of 1 am. at 2200 C. for 30 minutes. The plate is then atomized on one side with iron powder having a grain size smaller than 1 ,am. and the excessive iron powder is wiped oif.

As is shown in the diagrammatic cross-sectional view of FIG. 1 of the accompanying drawing, a substrate 1 thus formed which is nucleated with iron powder 2 in a graphite crucible 3 is placed on a bed of quartz powder 4.

The assembly is arranged in a quartz tube 5, through which hydrogen is led at a rate of 1 litre per minute and heated at 1280 C.

Whiskers of silicon carbide, diameter 1 m., length 2 cm., grow on the substrate in 10 hours. The yield is mgms. of whiskers per plate.

EXAMPLE 2 Aluminum oxide powder (pro analysis) having a grain size of 0.01 to 0.1 ,um. is mixed with ethyl acetate and compressed to tablets of 10 x 100 x 1 mm.

As shown in the diagrammatic plan view of FIG. 2, a few of these tablets 11 serving as substrates and aluminum oxide crucibles 12 filled with aluminum 13 are placed on a bed of quartz grains 14 in an aluminum oxide boat 15. The assembly is surrounded by a tube of aluminum oxide 16 through which a hydrogen current of 1 litre per minute is led, and heated by means of an oven at 1300 C. The substrate is nucleated via the gaseous phase with aluminum.

Ribbon-shaped aluminum oxide whiskers formed from Al oxidised by SiO grow in 6 hours and have a width smaller than 0.1 ,um., a thickness of 0.01 m, and a length of 2 cm., which at their end, have an Al head.

The substrate is dense and uniformly populated with whiskers. The yield is 200 mgms.

EXAMPLE 3 The substrate used in this example is a commercial product consisting of a graphite plate provided with a grown pyrographite layer of known orange peel structure having projecting graphite cones of approximately 10 am. diameter (Chernie, Ingenieur Technik, nr. 39, vol. 14 (1967), 833).

A plate of 100 x 20 mm. of this material is smeared with powder of carbonyl nickel having a grain size smaller than 1 m.

As shown in the cross-sectional view of FIG. 3, this plate 21 nucleated with nickel is laid on a graphite crucible 22 as a cover in which crucible quartz powder 23 is present. The assembly is arranged in a tube 24 of quartz and a flow of 75% nitrogen with 25% hydrogen is led through at a rate of /2 litre per minute. By high frequency heating of the crucible for 60 hours at a temperature of 1250" C., whiskers of silicon nitride, diameter 10 m, length 10 mm., grow on the lower side of the substrate 21 in a uniformly dense population.

EXAMPLE 4 Onto a quartz plate the surface of which is provided with an adhesive layer consisting of a rubber adhesive, a powdered mixture of sand and gold in a weight ratio of 1:10 having a grain size of 30 m is spread and nonadhered powder particles are flicked off. In this manner a monograin layer of sand and gold is formed on the quartz plate.

The resulting substrate nucleated with gold is arranged in a quartz tube.

A gas-current consisting of hydrogen with 2 mol percent SiCL of 20 litres per hour is led through.

Upon heating at 1050 C. for 60 hours a uniformly dense growth of silicon whiskers having a length of 1 cm. and a diameter of 30 m. is formed on the substrate.

EXAMPLE 5 Zirconium oxide powder, having a grain size 1 m is mixed with 1% by weight of Fe O grain size 0.1 m. A suspension of the mixture obtained is spread on a quartz plate in the form of a thin layer.

The resulting substrate is arranged in a quartz tube, and a gas current consisting of hydrogen with 3% propane is led through said tube at a rate of 1 litre per min.

The iron oxide is converted in a uniformly dense nucleation of iron by the hydrogen upon heating at 1280 C. Furthermore, silicon monoxide is developed from the quartz so that a crystallisation atmosphere containing silicon and carbon is formed, as a result of which whiskers of silicon carbide, having diameters of 0.1 ,uIIL, lengths of 1 cm., grow at the locations of the iron nuclei.

What is claimed is:

1. A method of growing crystals in the form of whiskers which comprises forming a substrate by sintering silicon carbide powder having a uniform grain size at a temperature of about 2200 C. to form a plate thereof; atomizing the plate on one side with iron powder to nucleate it, said iron powder having a grain size which is smaller than the silicon carbide powder, and wiping ofl any excessive iron powder from the plate; placing the thus formed substrate in a graphite crucible resting on a bed of quartz powder, arranging the crucible with the substrate resting on the quartz powder in a quartz tube, and flowing hydrogen gas heated to a temperature of 1280" C. through the quartz tube; whereby whiskers of silicon carbide having a diameter of 1 micrometer and a length of 2 centimeters are grown on the substrate in response to strong anisotropic crystal growth stimulated by the nucleated substrate.

2. A method as claimed in claim 1, wherein the hydrogen flow is one liter per minute.

References Cited UNITED STATES PATENTS 8/ 1965 Merkel et a1. 23223.5 5/ 1971 Blakeslee et a1 148--1.6

U.S. Cl. X.R. 23-625; 423-400 mg UNITED STATES PATENT oTTTcE ERTiFlATE OF QGRREQTWN Patent No. 3692478 Dated September 19 1972 Inventor(s) WILHELMUS FRANCISCUS KNIPPENBERG ET AL It is certified that error appears in the above-identified patent and that said Letters Patent are hereby corrected as shown below:

-1 In the heading: before "Filed Feb. 2, 1970" insert assignors to U. S. Philips Corporation,

New York, N. Y.

Col. 3, line 1, "temperature" should read temperatures Signed and sealed this 10th d f April 1973,

(SEAL) Attest:

EDWARD 1 I.PLETCHER,JR. ROBERT GOTTSCHALK Attesting Officer Commissioner of Patents

Images