Boron Suboxide Material and Method for Its Preparation

The invention relates to a novel boron suboxide material and to a method for its preparation. More particularly, the present invention relates to a new boron suboxide material which is much harder than boron suboxides previously reported, and to a method for preparing the new boron suboxide material.

A great deal of research has been devoted to developing synthetic superhard materials which have hardness values approaching that of diamond. The best known of these synthetic superhard materials is cubic boron nitride (cBN). Other very hard binary and ternary compounds made from light elements such as boron, nitrogen, oxygen, aluminum, silicon and phosphorous are also known. Among these, the compounds C.sub.3 N.sub.4, BNC, BPN, B.sub.4 C and BP may be mentioned. Another of these compounds which has been reported as having high hardness values is boron suboxide.

Boron normally has a valence of 3 and reacts with oxygen to form boron oxide having the stoichiometric formula B.sub.2 O.sub.3. Under suitable conditions, boron may react with oxygen and form compounds in which boron exhibits a valence less than 3. Powders of nominal composition B.sub.3 O, B.sub.4 O, B.sub.6 O, B.sub.7 O, B.sub.8 O, B.sub.12 O, B.sub.15 O, and B.sub.18 O have been reported as being formed by reacting elemental boron (B) with boron oxide (B.sub.2 O.sub.3) under suitably high pressure and temperature conditions. For purposes of this disclosure, the term "boron suboxide" refers to boron/oxygen binary compounds wherein boron has a valence less than 3. Since different varieties of boron suboxide have been reported by others, boron suboxide will be generally designated with the chemical formula B.sub.x O.

The formation of boron suboxide (B.sub.x O) and a description of its properties have been extensively reported in the literature. Most of the reports in the literature attribute the formula B.sub.6 O or B.sub.7 O to the boron suboxide being investigated. In some cases, the boron suboxide formed and the material being investigated may 0 consist of more than one phase. In U.S. Pat. No. 3,660,031 (Holcombe, Jr., et al.), a method of preparing a boron suboxide material is disclosed. According to this patent, the boron suboxide material is formed by reducing zinc oxide with elemental boron at a temperature in the range of 1,200.degree. C.-1,500.degree. C. The boron suboxide produced by this method is reported as having the formula B.sub.7 O. It is also characterized as having an average hardness value as measured with a Vickers indentor under a 100 g load (VHN.sub.100) of 3,820 kg/mm.sup.2 and a density of about 2.60 g/cc. The material is described as highly refractory, and suitable for use on surfaces subject to abrasion, e.g., grinding wheels, drill bits, machine tools, etc., and in structures employed in high temperature applications.

In U.S. Pat. No. 3,816,586 (Goosey), a method of fabricating boron suboxide products is disclosed. According to this patent, a mixture of elemental boron and boron oxide is cold pressed in a tantalum lined metal die. After the pressure on the compacted mixture is released, it is coated with a mixture of boron nitride and boron oxide, and is subjected to a second pressing step while heating at a temperature sufficient to melt the boron oxide in the compacted mixture. This is followed by a cooling step and another hot pressing step. The boron suboxide product made by this method is reported as being a smooth, sound boron suboxide article, free of flaws and contaminants, and suitable for a variety of applications, e.g., in cutting tools, ceramic bearings, and grinding media. Upon analysis, the boron suboxide product gave 80.1 wt. % boron and 19.9 wt. % oxygen which corresponds to the stoichiometry of B.sub.6 O. It was also reported as having a density of 2.60 g/cc and a Knoop hardness under a 100 g load (KHN.sub.100) of 3000 kg/mm.sup.2.

Other reports on methods of preparing boron suboxide and the properties of this material are the following:

1. R. R. Pasternak, "Crystallographic Evidence for the Existence of B.sub.7 O", Acta Cryst. 12 (1959), 612;

2. S. LaPlaca, B. Post, "The Boron Carbide Structure Type", Planseeberichte Fur Pulvermetallurgie, Bd. 9, 1961;

3. H. F. Rizzo, W. C. Simmons, and H. O. Bielstein, "The Existence and Formation of the Solid B.sub.6 O", Journal of the Electrochemical Society, Jan. 1963;

4. F. A. Halden, R. Sedlacek, "Growth and Evaluation of Boron Suboxide and Zirconium Dioxide Single Crystals", Stanford Research Institute, Menlo Park, Calif., Jan. 1963;

5. H. Tracy Hall and Lane A. Compton, "Group IV Analogs and High Pressure, High Temperature Synthesis of B.sub.2 O", Inorg. Chem. 4 (1965) 1213;

6. W. C. Simmons, "Progress and Planning Report on Boron Suboxide B.sub.6 O", Air Force Materials Laboratory, Mar., 1968;

7. H. Wrerheit, P. Runow, and H. G. Leis, "On Boron-Suboxide Surface Layers and Surface States of B-Rhombohedral Boron", Phys. Stat. Sol. (a) 2, K125 (1970);

8. E. V. Zubova, K. P. Burdina, "Synthesis of B.sub.6 O Under Pressure", Dokl. Akad. Nauk. SSR, 197 (5) (1971) 1055-1056;

9. C. E. Holcombe, Jr., and O. J. Horne, Jr., "Preparation of Boron Suboxide, B.sub.6 O", Journal of the American Ceramic Society--Discussions and Notes, Vol. 55, No. 2 (1971) 106;

10. D. R. Petrak, R. Ruh, B. F. Goosey, "Preparation and Characterization of Boron Suboxide", National Bureau of Standards Special Publication 364, Solid State Chemistry, Proceedings of 5th Materials Research Symposiums, issued Jul. 1972;

11. W. H. Rhodes, A. J. DeLai, "Research on Development and Fabrication of Boron Suboxide Specimens," AVCO Corp., prepared for: Air Force Materials Laboratory, Aug., 1972;

12. R. R. Petrak, Robert Ruh, and G. R. Atkins, "Mechanical Properties of Hot-Pressed Boron Suboxide and Boron", Ceramic Bulletin, Vol. 53, No. 8 (1974), 569-573;

13. P. M. Bills and D. Lewis, "Non-stoichiometry of Boron Suboxide (B.sub.6 O)", Journal of the Less Common Metals, 45 (1976) 343-345;

14. V. S. Makarov and Ya. A Ugai, "Thermochemical Study of Boron Suboxide B.sub.6 O", Journal of the Less Common Metals, 117 (1986), 277-281;

15. C. Brodhag and F. Thevenot, "Hot Pressing of Boron Suboxide B.sub.12 O.sub.2 ", Journal of the Less-Common Metals, 117 (1986), 1-6;

16. Tadashi Endo, Tsugio Sato, Masahiko Shimada, "High-pressure synthesis of B.sub.2 O with diamond-like structure", Journal of Materials Science Letters 6 (1987), 683-685;

17. David Emin, "Icosahedra Boron-Rich Solids", Physics Today, Jan., 1987.

18. Andrzej Badzian, "Superhard Material Comparable in Hardness to Diamond", Appl. Phys. Lett. 53(25), 19 Dec. 1988;

19. W. E. Moddeman, A. R. Burke, W. C. Bowling and D. S. Foose "Surface Oxides of Boron and B.sub.12 O.sub.2 as Determined by XPS" Surface and Interface Analysis, v. 14 n5, May, 1989, p. 224-232.

All of the aforementioned patents and publications are incorporated herein by reference.

These publications and patents are in agreement that boron suboxide compounds can be produced according to the following chemical equation:

at moderately high pressures (e.g., 1,000 to 6,000 psi) and moderately high temperatures (e.g., 1,400.degree. C. to 2,200.degree. C.). The specific form of B.sub.x O actually produced depends on the process conditions and the ratio of elemental boron to boron oxide loaded into the reaction cell. Alternatively, boron suboxide compounds may be produced according to the following chemical equation:

wherein M=Mg or Zn, at temperatures ranging from 1200.degree. C. to 1500.degree. C. without applied pressure. Depending on the starting proportions of boron and metal oxide, the formula B.sub.6 O has been attributed most frequently to the boron suboxide compound formed, although some researchers report the formation of B.sub.7 O and other boron/oxygen compounds.

Generally, the hardness and density values reported for boron suboxide in all of the aforementioned publications have been in good agreement. Average KHN.sub.100 values of 3400 to 3600 kg/mm.sup.2 were reported by Simmons, and KHN.sub.100 values of 3400 to 3500 kg/mm.sup.2 were reported by Petrak et al. The KHN.sub.100 value reported in U.S. Pat. No. 3,816,586 (Goosey) is somewhat anomalous and this may be due to the presence of another phase in addition to B.sub.6 O. In any event, none of the researchers have reported a KHN.sub.100 value greater than 3600 kg/mm.sup.2 for boron suboxide. The densities reported by these researchers have been from 99.5 to 100% of the theoretical value. Furthermore, based on crystallographic data, most of the researchers have ascribed a rhombohedral unit cell to the boron suboxide that was produced.

While the reported hardness values are quite good, it would be desirable to produce a superhard form of boron suboxide having a hardness value which more closely approaches that of diamond which ranges from KHN of 6000 to 9000. More particularly, because the raw materials to form boron suboxide are inexpensive and the process is relatively simple, it would be desirable to produce a superhard form of boron suboxide which has an average KHN.sub.100 hardness value of at least about 3800 kg/mm.sup.2, and preferably in the range of about 4000-4500 kg/mm.sup.2.

Recent crystallographic studies of B.sub.6 O and B.sub.7 O have revealed that both possess the same crystal structure. This indicates these are the same defect phase and include all intermediate compositions. If used in a generic sense the term B.sub.6 O in this application henceforth refers to any composition between B.sub.6 O and B.sub.7 O, inclusive.

In accordance with the present invention, a new boron suboxide (B.sub.2 O) material is provided which comprises at least 95% by weight of B.sub.6 O and has an average KHN.sub.100 hardness value of at least about 3800 kg/mm.sup.2. Preferably, the new boron suboxide material has an average KHN.sub.100 hardness value above about 4000 kg/mm.sup.2, typically about 4250 kg/mm.sup.2, which is well beyond the average hardness values previously reported for any form of boron suboxide.

According to the invention, the novel form of boron suboxide is produced by reacting stoichiometric amounts of elemental boron and boron oxide in accordance with the chemical equation:

under carefully controlled conditions. The reaction is carried out in a specially designed cell at a temperature of about 1900.degree. C. to about 2100.degree. C., and at a pressure in the range of about 3,000 to about 4,000 psi, for a sufficient time to produce a form of boron suboxide which is at least 95% by weight B.sub.6 O and has an average KHN.sub.100 hardness value of at least about 3800 kg/mm.sup.2.

Furthermore, while the B.sub.2 O.sub.3 starting material in the above chemical equation is assumed to be essentially pure (e.g., about 99.9% pure), surprisingly it has been found that if the elemental boron used as a starting material is less than about 95% pure a harder B.sub.6 O material can be formed. It was found that the impure B contained about 6% Mg. It is believed that the Mg aids the sintering of the boron suboxide, so that an improved form of B.sub.6 O is produced which has greater hardness. The quantity of boron used is that suggested by equation 3 and was not adjusted to compensate for its lack of purity. Transmission electron microscopic examination of this harder material reveals a more defect-free structure than previously reported forms of B.sub.6 O. An alternate method of producing BsO of high hardness is to use pure raw materials and separately add a suitable quantity of the desired sintering aid. Other materials such as Ca and Y are also expected to act as sintering aids.

FIG. 1 illustrates the specially designed cell which is used to produce the novel form of boron suboxide.

FIG. 2 is a photomicrograph of a sample of the new superhard form of boron suboxide under a magnification of 1330.times..

FIG. 3 is a plot of the X-ray diffraction pattern obtained with copper K-alpha radiation of the sample shown in FIG. 2.

FIG. 4 is a photomicrograph of another sample of the new superhard form of boron suboxide under a magnification of 1330.times..

FIG. 5 is a plot of the X-ray diffraction pattern obtained with copper K-alpha radiation of the sample shown in FIG. 4.

FIG. 6 is a plot of average hardness (KHN) values for a variety of materials including the new superhard form of boron suboxide.

FIG. 7 is a photomicrograph at a magnification of 1330.times. of a comparison sample of boron suboxide which is not superhard and was made using high purity boron (99.9% B).

FIG. 8 is a plot of the X-ray diffraction pattern obtained with copper K-alpha radiation of the comparison sample shown in FIG. 7.

The novel boron suboxide material of the present invention is made by hot pressing a stoichiometric mixture of elemental boron and boron oxide powders in a specially designed cell. The boron oxide (B.sub.2 O.sub.3) powder is of high purity (e.g., 99.9% pure), while the elemental boron powder is desirably of lower purity (e.g., about 90-95% pure, preferably only about 90-92% pure). The main impurities in the 90-92% boron powder are Mg (ca. 5-6%), oxygen in combined form (ca. 2-3%), and small amounts of other ingredients (e.g., 0.15% Si, 0.1% Fe, 0.08% Mn, 0.05% Al, 0.01% Ca, 0.01% Cu). Desirably, the boron powder should contain at least 1% by weight of Mg or another sintering aid.

The boron and boron oxide powders are mixed together in stoichiometric proportions according to equation (3). It is important that the two powders be mixed very thoroughly and for this purpose it is desirable that the mixture be placed in a container with tungsten carbide balls as a mixing aid and hexane or other non-oxygenated dispersion liquid (1/3 by volume) and tumbled overnight to ensure intimate mixing. This slurry should be dried followed by oven heating at about 100.degree. C. to evaporate any remaining hexane. This is followed by screening to remove large agglomerates.

A mixture of boron and boron oxide powders prepared in this manner can be converted into the novel boron suboxide material by hot pressing the mixture at temperatures in the range of about 1800.degree.-2200.degree. C. and at pressure of about 2,000-6,000 psi while encapsulated by hexagonal boron nitride or other suitable barrier material. For this purpose, the specially designed cell shown in FIG 1 may be used. The cell comprises a graphite mold having outer wall 12 and inner wall 14. A cylinder 16 made from hexagonal boron nitride forms a part of the inner wall 14. The boron/boron oxide mixture 18 is positioned in the graphite mold within the hexagonal boron nitride cylinder. The boron/boron oxide mixture is then sandwiched between hexagonal boron nitride plates 21, 22. The hexagonal boron nitride plates 21, 22 are set between layers 25, 26 of hexagonal boron nitride powder. Another mass of hexagonal boron nitride powder 24 is confined between a second set of boron nitride plates 20, 28. Graphite pistons 32, 34 press against the hexagonal boron nitride powder 25 and the hexagonal boron nitride plate 28 thereby compressing the boron/boron oxide mixture while encapsulated with hexagonal boron nitride or other suitable barrier.

When the components are assembled and loaded with premixed powders as described above and subjected to pressures of about 2,000 to about 6,000 psi and temperatures of about 1800.degree. C. to about 2200.degree. C. for periods of about 1 to 3 hours, the novel boron suboxide material is formed. Typically, a 30 g sample of stoichiometric boron/boron oxide is subjected to a pressure of 3,240 psi and a temperature of 1960.degree. C. for about 3 hours to form boron suboxide which is at least about 95% by weight of B.sub.6 O and has an average KHN.sub.100 hardness value of at least about 3800 kg/mm.sup.2 and more typically about 4250 kg/mm.sup.2.

Since the scatter of average KHN.sub.100 values for a given material is no more than about 25-50 kg/mm.sup.2, and the average KHN.sub.100 values of the thus produced boron suboxide material is at least about 200 kg/mm.sup.2 and more typically about 650 kg/mm.sup.2, greater than any previously reported value for a boron suboxide material, it is concluded that the material produced in accordance with the present invention is a new superhard form of boron suboxide. X-ray diffraction, electron microprobe analysis and transmission electron microscopy confirm that this new superhard form of boron suboxide comprises at least about 95% by weight of polycrystalline B.sub.6 O and has a more perfect crystal structure with fewer defects than softer forms of boron suboxide also examined.

Boron suboxide made in this manner and having a KHN.sub.100 hardness of at least about 3800 kg/mm.sup.2 is in the form of a solid mass when it is removed from the cell. It can then be crushed into granules and sized by techniques well known to those skilled in the art. The powder can be used as an abrasive grit, or the grit can be bonded to tools which require highly abrasive surfaces, for example, grinding wheels, drill bits, machining tools, etc.

The invention will now be described by reference to the following examples.

A 20 g sample of boron/boron oxide powder was placed in a container containing tungsten carbide balls and hexane liquid (1/3 by volume). The 20 g sample consisted of 14.26 g of elemental boron and 5.74 g of boron oxide (B.sub.2 O.sub.3). The boron oxide powder was 99.99% pure, while the boron powder comprised 91.03% by weight boron, 5.76% by weight magnesium, and 2.81% by weight oxygen, the balance being trace other elements.

The powder mixture was tumbled overnight, placed in drying trays, and then heated at about 100.degree. C. to produce a dry, thoroughly mixed sample. Twenty grams of this sample was placed in a cell similar to that illustrated in FIG. 1 and was heated at 2000.degree. C. under a pressure of 3650 psi for about three hours to produce the new form of boron suboxide.

The boron suboxide material was in solid form as it was removed from the cell. The boron suboxide produced in this manner was given the designation NM086.

The NM086 sample was tested for hardness with a Knoop indentor and was found to have an average KHN.sub.100 hardness value of 3790 kg/mm.sup.2. The NM086 sample was also visually inspected under an optical microscope at a magnification of 1330.times.. FIG. 2 is a reproduction of the image seen through the microscope. It shows light gray areas containing primarily B.sub.6 O and less than 5% Mg and W, dark gray areas containing approximately 10% B, 55% O and 35% Mg, and white areas containing approximately 95-96% B, 1-2% W, and 2-3% Mg.

A sample of NM086 was also subjected to X-ray diffraction analysis. FIG. 3 shows the results that were obtained. The sharpness and position of the peaks demonstrates that the NM086 sample consisted of a crystalline product with lattice constants of a=5.38435A.+-.0.00051A and c=12.3175A.+-.0.0062A when indexed as a hexagonal unit cell. Transmission election microscopy (TEM)

photographs and electron diffraction patterns of the NM086 confirmed the presence of fewer faults in this sample than found in softer samples of B.sub.6 O.

It was concluded that NM086 was a new superhard form of boron suboxide which was much harder than any previously known form of boron suboxide, and which consisted primarily of polycrystalline B.sub.6 O with relatively few defects or dislocations in the material.

The conditions of Example 1 were repeated except that this time the boron/boron oxide sample weighed 18.5 g and was subjected to a temperature of 1960.degree. C. at a pressure of 3241 psi for about three hours. The sample was given the designation NM138 and was tested as in Example 1.

The NM138 sample was tested for hardness with a Knoop indentor and was found to have an average KHN.sub.100 value of about 4250 kg/mm.sup.2. The sample was also inspected visually under a magnification of 1330.times. with an optical microscope and a photograph thereof is shown in FIG. 4. The same microstructure and constituents are seen in FIG. 4 as were seen in FIG. 2. The NM138 sample was also subjected to X-ray diffraction analysis and the results thereof are shown in FIG. 5. These tests confirm the conclusions reached in Example 1.

The average Knoop hardness of the NM138 sample was measured under various loads. FIG. 6 illustrates the results obtained. Also illustrated are average Knoop hardness values that were measured for various other well-known abrasive materials under the same indentation conditions. FIG. 6 shows that the average KHN value for the NM138 sample of boron suboxide is higher than that of polycrystalline cubic boron nitride at a load of 100 g, and is much higher than the hardness values measured for all the other abrasive materials at all loads. The greater decrease in hardness of some materials at 400 g load is believed to be due to porosity in the test material.

In order to compare the results obtained for the new superhard form of boron suboxide with previously known forms of boron suboxide, Example 1 was repeated except this time essentially pure (99.9%) elemental boron powder was mixed with pure boron oxide. In this case, the mixture was sub]ected to a temperature of 1960.degree. C. and a pressure of 3200 psi for 3 hours. All other conditions were the same as in Example 1. The sample produced in this manner was given the designation NM176 and was tested as before.

The average KHN.sub.100 value obtained for the NM176 sample was only 1568 kg/mm.sup.2. A photograph of the sample under a magnification of 1330.times. is shown in FIG. 7. This photograph shows the presence of considerable porosity. This indicates that the material is not well sintered. The X-ray diffraction pattern shown in FIG. 8 confirms that NM176 is not a well crystallized material as many of the peaks are broad compared with the X-ray diffraction patterns of the NM086 and NM138 samples. This may indicate an incomplete chemical reaction and the presence of intermediate phases.

In another set of experiments, NM086 samples were subjected to a friability test. The NM086 samples were removed from the die, crushed and then sized. The grit was then sub]ected to a friability test. The same friability test was conducted on a sample of single crystal cBN and on an Al.sub.2 O.sub.3 grit made from seeded gel.

The NM086 material was found to have desirable friability characteristics as shown in the following table.

The friability index is a measure of toughness and is useful for determining the grit's resistance to fracture during grinding. The values in this table are the percent of grit retained on a screen after friatesting. This procedure consists of a high frequency, low load impact test and is used by manufacturers of diamond grit to measure the toughness of the grit. Larger values indicate greater toughness. The data show that the NM086 material has friability characteristics similar to that of the cBN tested (BZN-I from General Electric Co.).

In another set of experiments, grit made from NM086 material was subjected to a grinding test. In a first test, the NM086 material was used in making a resin bond grinding wheel. Wet grinding of a D-2 steel workpiece was conducted at a traverse speed of 600 in./min and a crossfeed of 0.05 inches. The results for the NM086 material and for the Al.sub.2 O.sub.3 (Norton SG) are shown in Table 2.

A similar vitreous bond grinding wheel test was conducted using another sample of NM086 material. In this case, the workpiece was 52100 steel and the infeed was 0.001 inches/second. The results are shown in Table 3.

The results from the grinding experiments demonstrate that B.sub.6 O having a KHN.sub.100 value of at least about 3800 kg/mm.sup.2 can act as an abrasive grit. However, the results of the tests shown in Table 3 are somewhat disappointing in comparison to cBN and Al.sub.2 O.sub.3 (Norton SG). More recent experiments have shown that the reason B.sub.6 O was sometimes inferior to cBN and seeded gel is because of a probable reaction between the iron in the workpiece and the boron in the B.sub.6 O.

In an Al390 turning test a B.sub.6 O cutting tool insert (SNE433) performed similarly to a Si.sub.3 N.sub.4 insert with the same geometry.

An experimental design was also conducted wherein the factors pressure, temperature, boron to oxygen ratio and purity of boron powder were examined to determine their effects on hardness. Pressure was found to affect hardness the least, i.e., hard materials could be produced over a range of pressures. As a result of this finding, it is postulated that it may also be possible to produce B.sub.6 O materials with hardness greater than KHN.sub.100 =3800 kg/mm.sup.2 with other apparatuses which are capable of operating at pressures greater than 6000 psi (for example pressures up to 50,000 psi which are typical of hot isostatic pressing.)

While the invention has been described by reference to specific embodiments, this was for purposes of illustration only. Numerous alternative embodiments which are within the scope of the invention will be apparent to those skilled in the art.