BACKGROUND OF THE INVENTION

This invention relates to a process for the production of powdered metallic cobalt and, more particularly, relates to a process for the production of powdered metallic cobalt including ultra fine powdered metallic cobalt by reduction of cobaltous ammonium sulphate solutions.

Much of commercially available cobalt powders are prepared by a method wherein cobalt oxalate, precipitated from a suitable cobalt salt solution, is decomposed and reduced in a partially reducing atmosphere at elevated temperatures to give metallic cobalt powder. The resulting cobalt powder is of high purity but has a fibrous morphology and is not free flowing. End users recently have expressed interest in high purity free flowing cobalt powder as a replacement for the high purity fibrous powder in powder metallurgy applications.

A method for the production of cobalt from aqueous cobaltous ammonium sulphate solutions by reduction wit _h gaseous hydrogen at elevated temperatures and pressures was disclosed in a paper entitled The Hydrometallurgical Production of Cobalt published in the Transactions, CIM, 65(1962), 21 - 25 by . Kunda, J.P. Warner and V.N. Mac iw. In the commercial production of metals by this method, there are two basic stages in the reduction process: an initial "nucleation" stage followed by a later "densification" stage. In the nucleation stage, reduction is initiated and fine metal particles or nuclei are forme _d in the solution. In the densification stage, metal is precipitated from solution onto the preformed "seed" particles to produce larger particles. This latter step is repeated until the powder reaches the desired size.

In order to initiate the formation of the metal particles during the nucleation stage, a nucleation catalyst must be added to the aqueous metal salt-containing solution. The method developed by Kunda et al, and in commercial use by Sherritt Gordon Limited, uses a mixture of sodium sulphide and sodium cyanide to promote nucleation

of cobalt, powder. This method can be used to produce powders of as small as 25 microns in size; however, the powder is relatively high in sulphur and carbon content (0.3 to 0.8% C and 0.2 to 0.5% S). When powders of finer size are required, the carbon and sulphur levels normally are higher since fewer densifications result in less dilution of the initial carbon and sulphur in the nucleation powder.

In addition to the potentially high carbon and sulphur levels reporting to the product powder, the use of sodium cyanide is undesirable because of its toxic nature.

It is a principal object of the present invention to provide a process for the production of spherical or nodular cobalt powder having an average particle size less than 25 microns, as measured by FSSS, with low carbon and sulphur contents.

It is another object of the present invention to provide a process for the production of ultra fine, i.e. submicron, free flowing, spherical or modular cobalt powder which powder has particular utility as a binding material for cemented carbide for use as a cutting tool.

A further object of the present invention is the provision of a process which does not require sodium cyanide for the nucleation of fine cobalt powder. SUMMARY OF THE INVENTION

The process of the present invention obviates the need for sodium sulphide and sodium cyanide for the nucleation of fine cobalt powder, it having been found that the production of fine metallic cobalt powder suitable for use as seed in the preparation of coarser powder can be precipitated from ammoniacal cobaltous sulphate solutions by the addition of a soluble silver salt, preferably silver sulphate or silver nitrate, as a nucleating catalyst, in the presence of suitable organic compounds such as bone glue, polyacrylic acid and bone glue/polyacrylic acid mixture to control growth and agglomeration of the cobalt particles. This process for the production of cobalt

powder comprises adding to a solution containing cobaltous ammonium sulphate having an ammonia to cobalt mole ratio of about 1.5 to 3.0:1, a soluble silver salt such as silver sulphate or silver nitrate in an amount to provide a soluble silver to cobalt weight ratio in the range of 1.0 to 10 g of silver per 1 kg of cobalt to be reduced, adding bone glue and/or polyacrylic acid in an amount effective to prevent growth and agglomeration of the cobalt metal powder to be produced, and heating said solution to a temperature in the range of 150 to 250°C with agitation under a hydrogen pressure of 2500 to 5000 kPa for a time sufficient to reduce the cobaltous sulphate to cobalt metal powder.

The process of the invention for producing cobalt powder having an average size less than 25 microns comprises three stages consisting of an initial nucleation stage, a reduction stage and a final completion stage. The nucleation stage, which serves as an induction period, typically requires up to 25 minutes, the reduction stage (reducing period) for reducing most of the cobaltous cobalt in solution requires up to 30 minutes, usually about 15 minutes, and the completion stage (completion period) for removal of last traces of cobalt in solution typically requires 15 minutes.

We have further found that an ammoniacal cobaltous sulphate solution having a molar ratio of ammonia to cobalt of about 2.0:1, a soluble silver concentration of at least 1 g of silver per kilogram of cobalt and a mixture of animal glue and polyacrylic acid in an amount of about 0.01 to 2.5% of the weight of cobalt, can be reduced under hydrogen pressure with an induction time of less than 10 minutes and a reduction time of less than 10 minutes, to produce ultrafine cobalt powder having an average size less than one micron.

In its broadest aspect, the method of the invention for the production of cobalt powder from a solution containing cobaltous ammonium sulphate thus comprises

adding a soluble .silver salt in an amount to provide a soluble silver to cobalt weight ratio in the range of 0.3 to 10 g of silver per 1 kg of cobalt to be reduced, adding bone glue and/or polyacrylic acid in an amount effective to prevent agglomeration of the cobalt metal powder to be produced, and heating said solution to a temperature in the range of 150 to 250°C with agitation under a hydrogen pressure of 2500 to 5000 kPa for a time sufficient to reduce the cobaltous sulphate to cobalt metal powder.

More particularly, the process of the invention comprises adding ammonia to a solution of cobaltous sulphate containing a cobalt concentration of 40 to 80 g/L to yield an ammonia to cobalt mole ratio of about 1.5 to 3.0:1, adding a soluble silver salt such as silver sulphate or silver nitrate to yield a silver to cobalt weight ratio of about 0.3 g to 10 g silver:1 kg cobalt, adding a mixture of bone glue and polyacrylic acid in an amount of 0.01 to 2.5% of the weight of the cobalt, heating said mixture to a temperature in the range of 150°C to 250°C and agitating said mixture in a hydrogen atmosphere at a total pressure in the range of 2500 to 5000 kPa until cobaltous cobalt is reduced to cobalt metal powder.

In a preferred embodiment of the process of the invention for the production of submicron cobalt metal powder, the process comprises adding ammonia to a solution of cobaltous sulphate containing a cobalt concentration of about 40 to 80 g/L to yield an ammonia to cobalt mole ratio of about 2.0:1, adding silver sulphate or silver nitrate to yield a silver to cobalt weight ratio of about 0.3 g to 4 g silver:1 kg cobalt, adding a mixture of bone glue and polyacrylic acid in an amount of 0.01 to 2.5% of the weight of the cobalt, heating said mixture to a temperature in the range of 150° to 250°C, preferably about 180°C, and agitating said mixture in a hydrogen atmosphere at a total pressure in the range of 3000 to 4000 kPa preferably at about 3500 kPa during a nucleation period and reduction period of less than 20 minutes for reduction of cobaltous

cobalt to ultrafine. cobalt metal powder. BRIEF DESCRIPTION OF THE DRAWINGS

The method of the invention will now be described with reference to the accompanying drawings, in which:

Figure 1 is a process flowsheet of the process of the invention; Figure 2 is a photomicrograph of fibrous ultra fine cobalt powder well known in the prior art produced by decomposition and reduction of cobalt oxalate; Figure 3 is a photomicrograph of ultra fine, substantially nodular cobalt metal powder produced according to the process of the present invention; Figure 4 is a graph showing relative expansion of cobalt metal powder of the present invention; and Figure 5 is a graph showing relative expansion of cobalt powder produced from cobalt oxalate as illustrated in Figure 2. DESCRIPTION OF THE PREFERRED EMBODIMENT

With reference to the flowsheet of Figure 1, a solution of cobaltous sulphate may be prepared in step 10 by adding cobalt powder to an aqueous sulphuric acid solution, as is well known. Iron present in the solution is removed by addition of air for oxidation of iron at a pH greater than 6.0 and a temperature in the range of 50-70°C in step 12 and precipitated iron oxides removed by liquid/solid separation 14 and discarded.

The cobaltous sulphate solution essentially free of iron is fed to an autoclave reactor in step 16 in which concentrated aqua solution is added to provide a pH of about 8.0 to 10.0. Typically, ammonia is added to a cobaltous sulphate solution having a cobalt concentration of about 40 to 80 g/L to provide an ammonia to cobalt mole ratio of about 2.0:1 to 2.5:1.

A soluble silver salt, preferably silver sulphate or silver nitrate is added in a ratio of about -ϊ to 10 g of silver per 1 kg of cobalt to be reduced, preferably about 2 to 4 g of silver per kg of cobalt to be reduced.

A mixture of organic materials such as bone glue, gelatin or polyacrylic acid is added for agglomeration control, and the mixture heated with agitation to a temperature in the range of 150 to 250°C, preferably about 180°C, with agitation under an applied hydrogen atmosphere of about 3000 to 4000 kPa, preferably about 3500 kPa, for a time sufficient to reduce the cobaltous sulphate to cobalt metal powder.

The agglomeration and growth control additives, preferably a bone glue/polyacrylic acid blend, are added in an amount of from 0.01 to 2.5% by weight of the cobalt.

The resulting slurry is transferred to liquid/solid separation step 18 for removal of ammonium sulphate and the cobalt metal powder is washed by addition of water. The washed cobalt metal powder is passed to a wash/drying step 20 in which a further water wash is conducted followed by the addition of alcohol for a final wash and drying prior to packaging 22.

The process of the invention will now be described with reference to the following non-limitative examples.

EXAMPLE 1

Cobalt nucleation powder was made in a one gallon laboratory reduction autoclave using procedures which parallel commercial nucleation procedures. All runs used 115 g/L C0SO4 nucleation solution. Solution volumes to provide 80 g/L Co were charged to the autoclave along with the polyacrylic acid and the silver salt. The autoclave was then sealed and purged with hydrogen. NH4OH was introduced into the autoclave after the hydrogen purge was complete. Standard reduction conditions of 190°C and 3500kPa total pressure resulted in complete reductions in about 15 minutes.

- ~ ~

A standard test using Na ₂ S/NaCN as catalyst produced powder in 15 minutes after a 30 minute induction period. The powder, which analyzed for 0.18% C and 0.18% S, was 100% minus 20 micrometres (microns) and had a Fisher number of 1.65. Test results are shown in Tables 1 and 2.

Table 1

Table2

Tests using 10 g of AG ₂ S0 ₄ per kg of contained Co produced powders in 15 to 20 minutes after induction periods of 20 to 45 minutes. The powders analyzed 0.1 to 0.2% carbon, 0.002 to 0.007% sulphur, were 100% minus 20 micros and had Fisher numbers of 1.25 to 2.40. These results indicate that silver salt is an acceptable alternative to the conventionally used Na ₂ S/NaCN catalyst.

EXAMPLE 2

Cobalt nucleation tests were conducted in a one gallon laboratory autoclave using procedures which parallel commercial procedures described above with reference to Figure 1. A calculated volume of cobalt plant nucleation solution to provide 80 g/L Co was added to the autoclave along with silver sulphate and a mixture of bone glue and polyacrylic acid. The autoclave was heated to 160°C, and a hydrogen overpressure of 3500 kPa was applied and maintained until the completion of the reduction. A temperature increase of 10 to 20 Celcius degrees was recorded during the reduction. Reduction times of

SUBSTITUTE SHEET

30 to 60 minutes were observed.

Seven tests were carried out in which an initial nucleation was followed by multiple densifications using cobalt plant reduction feed to determine the growth rate of the powder and the effect of densification on the carbon, sulphur and silver contents of the powder. Densifications were conducted as follows: hot (170°C) cobalt plant reduction feed solution was charged into the autoclave containing the nucleation powder; and hydrogen applied until the metal values were reduced.

Upon completion of the reduction, the end solution was flash discharged and the autoclave recharged with fresh feed solution. The additives tested to control particle growth in the densifications were polyacrylic acids such as sold under the trade-marks "ACRYSOL A-l" and "COLLOID 121" and a mixture of bone glue/polyacrylic acid.

The organic additives were made up as stock solutions containing 10% by weight active ingredient and added by pipette as required.

The levels of AG,S0 ₄ and additives used in the nucleation stages densification stages and the results of the reduction tests are reported in Table 3. Table3

SUBSTITUTE SHEET

Three further nucleation tests were conducted to determine the effect of increasing the level of bone glue/ polyacrylic acid additive on the degree of powder agglomeration. The results are recorded in Table 4. Table4

The degree of agglomeration decreased significantly as the additive addition rate was increased from 5 to 20 mL/L with optimum results obtained at an addition rate of 5 to 10 mL/L.

EXAMPLE 3 Two plant trials were conducted in a cobalt plant reduction autoclave using silver sulphate and bone glue/ polyacrylic acid to produce nucleation powders. Trial 14, conducted with bone glue/polyacrylic acid added at the rate of 3.0 mL/L, produced powder with a Fisher number of 2.75 and an average agglomerate size of 22 microns. This powder received

SUBSTITUTE SHEET

about 30 densifications of cobalt plant reduction feed and produced commercial S grade cobalt powder. The second trial (Trial 15) conducted with the bone glue/polyacrylic acid, added at the rate of 1.6 mL/L, produced agglomerates in excess of 150 microns in size which were leached to remove them from the autoclave.

Changes and results of the plant trials are reported in Table 5. Table 5

A standard plant nucleation using NaCN/Na ₂ S catalyst with bone glue/polyacrylic acid added at 1.5 mL/L, yielded nucleation powder approximately 15 microns in particle size. Laboratory nucleations conducted in a one gallon autoclave using NaCN/NaS catalyst required l5mL/L bone glue/polyacrylic acid to yield similar sized nucleation powder.

EXAMPLE 4

360 L of an aqueous solution containing 112 g/L cobalt as cobaltous sulphate was transferred through a filter to a clean 1000 L autoclave to yield 40,000 g cobalt as cobaltous sulphate and sufficient water was added to bring the volume to 850 L. Concentrated aqua ammonia solution was added to the cobaltous sulphate solution to give an ammonia to cobalt mole ratio of 2.5:1. This addition, 135 L of 215 g/L aqua ammonia solution, provided a pH in the range of 8.0 to 10.0 in the autoclave.

A mixture made by combining 170 σ of silver sulphate, 1 L liquid bone glue, 0.33 L polyacrylic acid, and 1 L aqua ammonia to 6L of water was added to the autoclave and the mixture heated with agitation to approximately 175°C. The autoclave was pressurized with hydrogen to 3500 kPa.

SUBSTITUTE SHEET

A standard plant nucleation using NaCN/Na2≤ catalyst with bone glue/polyacrylic acid added at 1.5 mL/L, yielded nucleation powder approximately 15 microns in particle size. Laborarory nucleations conducted in a one gallon autoclave using NaCN/NaS cataylst required 15mL/L bone glue/ polyacrylic acid to yield similar sized nucleation powder.

EXAMPLE 4

360 L of an aqueous solution containing 112 g/L cobalt as cobaltous sulphate was transferred through a filter to a clean 1000 L autoclave to yield 40,000 g cobalt as cobaltous sulphate and sufficient water was added to bring the volume to 850 L. Concentrated aqua ammonia solution was added to the cobaltous sulphate solution to give an ammonia to cobalt mole ratio of 2.5:1. This addition, 135 L of 215 g/L aqua ammonia solution, provided a pH in the range of 8.0 to 10.0 in the autoclave.

A mixture made by combining 170 g of silver sulphate 1 L liquid bone glue, 0.33 L polyacrylic acid, and 1 L aqua ammonia to 6 L of water was added to the autoclave and the mixture heated with agitation to approximately 175°C. The autoclave was pressurized with hydrogen to 3500 kPa.

The nucleation stage (induction period), reduction stage (reduction period) and completion stage (completion period) for reduction of cobalt ions to cobalt powder required less than 30 minutes at which point the solution concentration was less than 1 g/L cobalt. Table 6 below shows the induction time and reduction time to be less than 10 minutes.

TABLE 6

Test 4 Induction period ■= 1 minute

Reduction period = 7 minutes

Completion period = 15 minutes

The end solution contained less than 0.4 g/L total metals at a pH of 8.4. The powder was washed, dried and analyzed with a yield of 38 kg cobalt. The size distribution and chemical composition are shown in Table 7

TABLE 7 Test Ni 0 S C Microtra ^""

% % % % D-90 4 0.176 0.76 0.0055 0.159 3.96

EXAMPLE 5 The test conditions of Example 4 were repeated with the exception that only 60 g of silver sulphate were added, compared to 170 g of silver sulphate in Example 4 (i.e.33%), to a charge of 40,000 g of cobalt ascobaltous sulphate. The induction time was 4 minutes and the reduction time was 10 minutes for a yield of 34 kg cobalt.

The size distribution and chemical composition are shown in Table 8.

TABLE 8 Test Ni 0 S C Microtrac (microns) FN

% % % % D-90 D-50 D-10

5 0.169 0.74 0.006 0.142 5.21 3.06 1.12 1.09

The yield dropped to 34 kg cobalt powder and the average particle or agglomerate size increased.

EXAMPLE 6

The test conditions of Example 4 were repeated with the exception that only 0.25 L liquid bone glue was added, compared to 1 L liquid bone glue in Example 4 (i.e. 25%), to a charge of 40,000 g of cobalt as cobaltous sulphate.

The induction time increased to 23 minutes and the reduction time to 57 minutes. The size distribution is shown in Table 9.

TABLE 9

Test Microtrac (microns FN

(microns)

D-90 D-50 D-10

6 45.7 21.07 7.92 4.35

The induction and reduction times increased substantially to a total of 80 minutes with an increase in the average particle and agglomerate sizes.

EXAMPLE 7

The test conditions of Example 4 were repeated with the exception that 0.5 L liquid bone glue was added, compared to 1 L liquid bone glue in Example 1 (i.e. 50%), to a charge of 40,000 g of cobalt as colbaltous sulphate.

The induction time was 5 minutes and the reduction time was 32 minutes for a yield of 39 g of cobalt. The size distribution is shown in Table 10.

TABLE 10

Test Microtrac (microns) FN

(microns)

D-90 D-50 D-10

7 14.48 6.43 2.81 1.60

The average particle size distribution increased to well over 1 micron compared to Example 4.

EXAMPLE 8

The test conditions of Example 4 were repeated with the exception that the charge of cobaltous sulphate was increased to 50,000 and the silver catalyst increased to 210 g to maintain the same ratio of silver to cobalt.

The induction time was 7 minutes and the reduction time was 6 minutes for a yield of 49 kg cobalt. The size distribution is shown in Table 11

TABLE 11

Test Microtrac (microns) FN (microns)

D-90 D-50 D-10

8 4.17 2.40 0.96 0.89

EXAMPLE 9

The test conditions of Example 4 were repeated with the exception that the charge of cobaltous sulphate was increased to 50,000 and the silver catalyst decreased to 140 g to maintain the same ratio of silver to cobalt.

The induction time was 3 minutes and the reduction time was 6 minutes for a yield of 51 kg cobalt. The size distribution is shown in Table 12.

TABLE 12

Test Microtrac (microns) FN

(microns

)

D-90 D-50 D-10 7.68 4.19 1.92 1.25

Table 13 provides a summary of test results described in Examples 4-9. Reduction times in excess of 10 minutes, due for example to a reduction of silver sulphate catalyst or a reduction of the organic additive below optimum amounts, resulted in an increase in the Fisher Number above 1.

TABLE 12

S UBSTITUTE S H~~l

Figures 2 and 3 give a good visual comparison between submicron substantially spherical or nodular cobalt powder produced according to the present invention and the fibrous or rod-like cobalt powder produced by the well-known oxalate process.

With reference to Figure 3, the cobalt powder illustrated as produced according to the process of the invention has a substantially spherical or nodular shape and an average size of 0.6 to 0.8 micron. The shape provides superior flow characteristics to aid in mixing for preparation of consistent blends used in the manufacture of cemented carbide and diamond cutting tools. The uniform spherical shape and submicron size provides a high surface area, in excess of 2.0M ² /g, which results in improved sintering properties with high sintered densities.

EXAMPLE 10

Table 14 provides a summary of physical testing of ultra fine cobalt produced according to the present invention and extra fine cobalt produced from oxalate. The two cobalt powders were compacted at 5T/cm ² into rectangular green compacts, placed in a Netzch™ Dilatometer under an argon -5% hydrogen atmosphere and the green compacts subjected to a sintering profile from 100°C to 1050° at 10C°/minute and held at 1050° for 20 minutes.

TABLE 14

SUBSTITUTE SHEET

The gree density of ultra fine cobalt of the invention was about 4% greater than extra fine cobalt from oxalate and the sintered density of the ultra fine cobalt of the invention was 100% compared to 97% for the extra fine cobalt from oxalate.

Dimensional changes as represented by relative expansion were recorded during sintering and are represented by Figures 4 and 5, Figure 4 showing relative expansion of the cobalt powder of the invention and Figure 5 showing relative expansion of the cobalt powder produced from oxalate. The cobalt powder of the invention densified at a lower temperature to a greater final density than the cobalt powder from oxalate, the powder of the invention approaching 100% of theoretical density at 850°C while the cobalt powder from oxalate approached 97% of theoretical density at about 1000°C.

EXAMPLE 11

Tests were conducted to produce ultra fine cobalt powder using silver nitrate as a nucleating agent. The autoclaves were equipped with dual axial impellers and set to run at 860 rev/min. The reductions were carried out at 180°C under applied hydrogen pressure to a total pressure of 3500 kPa. The test solution was prepared by dissolving atomized cobalt in sulphuric acid and then sparging the solution with air once the pH had risen to over 6.0 in order to remove any dissolved iron. The solution contained 116.4 g/L cobalt, 0.286 g/L of nickel and less than 0.0002 g/L iron.

Swift's ™ animal bone glue, a colloidal protein containing approximately 50% by weight solids, and Acrysol A-2™, a solution of polyacrylic acid in water, were used. Eight L of a glue/acrysol mixture was made up by mixing one litre of glue, one litre of aqua, 0.33 mL of Acrysol A-2 and 5.66 L of water. The resulting light yellow suspension was sealed in an airtight container and used for all of the tests except Test Nos. 14 to 22.

Experimental conditions are provided in Table 15:

Table 15 Experimental Conditions for Cobalt Reduction Tests

- 18 - The induction and reduction times together with particle size determinations are listed in Table 16.

Table 16Reduction Times and Size Analysis of Cobalt Powders

The chemical analyses of the cobalt powder are given in Table 17 .

Table 17 Chemical Analysis of Cobalt Powder Samples

Preliminary tests to establish standard parameters were conducted as follows. 0.74 g of silver nitrate, predissolved in 50 mL of concentrated aqua, was added to 1.0 L of cobalt solution and 1490 mL of distilled water. 313 mL of concentrated aqua was then added to the cobaltous sulphate solution followed by 39 mL of the bone glue/acrysol mixture. The slurry was charged into an autoclave and reduced at 180°C under 3500 kPa hydrogen pressure. When the reduction was complete, the autoclave was cooled and the solids discharged. Typical total induction and reduction times were 30 to 35 minutes, inlcuding a 15 to 20 minute induction time.. The product powders typically contained 0.25 to 0.28% Ni, 0.36 to 0.38% Ag and had Fisher Sub-Sieve Size numbers in the range of 1.0 to 1.2.

With reference now to Table 15, which tabulates variables in operating conditions, Tests Nos. 1 to 6 show the effect of ammonia additions at various reaction temperatures. For each test, 856 mL of cobaltous sulphate solution and 1340 mL of distilled water containing 0.636 g of dissolved silver nitrate were charged into the reduction autoclave together with 39 mL of bone glue/acrysol mixture. The autoclave was then sealed and purged twice with 1000 kPa hydrogen. The contents were then heated to the preselected temperature in the range of 25°C to 180°C as indicated and 258 mL of concentrated aqua was then pumped into the autoclave. The temperature was then raised to 180°C if necessary and the reduction carried out as previously described. The aqua thus was added under an inert atmosphere to eliminate oxidation of the cobalt by air and subsequent formation of cobaltic ammine complexes. With the exception of Tests Nos 1 and 6, in which the ammonia was injected at 180°C, the reduction times (see Table 16) were significantly shorter than those observed in the standard test. The particle size analysis of these samples also showed a decrease, particularly in the Fisher

number which dropped from over 1.0 to an average of 0.73 for Test9s Nos. 2 to 5. Both Tests Nos 1 and 6, which were prepared by injecting the aqua at 180°C and immediately applying a hydrogen overpressure, had longer reduction times and substantially larger particle sizes.

The remaining Tests Nos. 7 to 10 to be described, were conducted with the ammonia added at 25°C in the manner indicated with reference to Test No. 5.

Tests Nos. 7 to 10 show the significance of ammonium sulphate presence in the head solution. The conditions of Test No. 5 were carried out with the addition of reagent grade ammonium sulphate in concentrations of 50, 150, 250 and 350 g/L (NH ₄ ) ₂ S04 prior to the injection of ammonia. The induction and reduction times showed a direct correlation with the amount of ammonium sulphate added. Both the induction and reduction times increased, with no reduction after 60 minutes, with an increase in particle size as measured by both Fisher number and Microtrac.

The effect of bone glue/acrysol additive dosage was assessed in tests Nos. 11, 12 and 13. The amount of additive solution added to the reduction charge was reduced from 39 mL to 29 mL for Test 11, to 19.5 mL for Test 12 and to 10 mL for Test 13. It was observed that both the induction and reduction times increased as the additive volume was decreased (see Table 16). Particle size analysis of the product powders also showed a similar inverse correlation between average particle size and the amount of additive, according to both Microtrac measurements and Fisher number analysis (see Table 15). Test No. 11, prepared using 29 mL of additive instead of 39 mL, closely resembled the samples prepared in the previous set of tests Nos. 1 to 6 indicating that there is a plateau level beyond which increasing the additive dosage has no beneficial effect. These results show that the glue/polyacrylic acid mixture has an influence on both the reduction times and on the size of the product cobalt powder.

In the series of tests Nos. 14 to 22, the ratio and amounts of the bone glue and the polyacrylic acid were varied to determine what influence each had on the reduction times and product particle size. A typical additive mixture for a test was made up as follows. The selected quantities of bone glue and polyacrylic acid were added to a solution of 7.5 mL aqua in 42.5 mL of distilled water. The mixture was agitated until it was homogeneous, at which point 29 mL was added to the autoclave charge. The additives used in each test are listed in Table 14. In tests Nos 14 to 18, in which the level of polyacrylic acid was held constant and the amount of bone glue was varied, the total reduction time varied inversely with the amount of bone glue added. Test No. 18, in which no glue was added produced no cobalt powder even after one hour. The particle sizes of the powders produced in these first five tests show a similar inverse relationship, the particle size increasing as the quantity of bone glue was decreased. This trend is evident in both the Fisher Numbers and the Microtrac values (Table 16).

In Test Nos. 29 to 32, in which the quantity of bone glue was held constant and the amount of polyacrylic acid was varied, a direct relationship existed between the total reduction time and the amount of polyacrylic acid added. In these latter tests, the variation in the amount of the additive did not affect the Fisher number but did impact on the Microtrac values. A general inverse relation between reduced additive level and increased D50 with constant Fisher number is apparent, which is indicative of increased agglomeration. It should be noted that the sample prepared with no polyacrylic acid was severely agglomerated and resembled steel wool when removed from the autoclave.

The effect of ferrous and ferric iron on the reductions was assessed in Tests 23 to 28. Neither ferrous nor ferric iron increases had an apparent effect on the Fisher numbers but the Microtrac values increased in both

cases, indicating increased agglomeration. The ferrous iron reported to the cobalt powder whereas not all ferric iron reported to the cobalt powder. (Table 17).

In tests Nos 29 to 32, the effect of varying the amount of silver added to the charge was examined. The first test No. 29 was carried out using the standard test previously described as a standard reference. Subsequent tests Nos. 30, 31 and 32 were conducted with 0.477 g, 0.381 g and o,159 g of silver nitrate, representing 75%, 50% and 25% respectively of the original weight. Results of the individual tests are given in Table 16. It was observed that the reductions proceeded as normal with no increase in reduction times as the silver content was decrease .

In the series of tests Nos. 33 to 36, the effect of varying the ammonia to cobalt mole ratio from 2.0 to 2.6 to 1 was examined. In the first three tests (Tests Nos. 33, 34 and 35, conducted at mole ratios greater than 2.0 to 1, the reduction times were approximately constant but noticeably longer in comparison to the fourth test carried out with a mole ratio of 2.0 to 1. The particle size analysis and the chemical analysis of the product cobalt powders showed no correlation with the ammonia to cobalt ratio. A mole ratio of about 2 to 1 of ammonia to cobalt thus provides effective reduction.

Tests Nos. 37 - 40 were conducted to determine the effec of cobalt concentration on the size of the product powder. Cobalt concentrations of 45 to 50 g/L were used and for each concentration two tests were conducted. For the first test, only the ammonia concentration was increased, in order to maintain an ammonia to cobalt mole ratio of 2.2 to 1, while for the second test, the amounts of silver nitrate and glue/polyacrylic acid added to the charge were raised in proportion to the increase in the amount of cobalt. Details of the tests are given in Table 15.

In spite of the larger quantity of cobalt to be reduced, the total reduction times of all four tests were not significantly different than those observed in previous tests for charges containing only 40 g/L cobalt. The particle size data also show that no significant increase in average particle size of the powder occurred as a result of using the higher concentrations, even when lower quantities of silver and organic additives were used. In fact, the two samples from the tests run at 50 g/L cobalt are actually finer than those prepared at 45 g/L and finer and less agglomerated than most of the samples prepared in previous tests at 40 g/L cobalt. These results indicate that acceptable ultrafine powder at high production rates can be prepared by using higher concentrations of cobalt in the autoclave charge.

The ultra fine cobalt powder of the present invention has particular utility as a major constituent of matrix material in the manufacture of diamond cutting tools such as rotary saw blades, wire rope saw ferrules and grinder cups which may contain up to about 95% by weight cobalt, the balance diamond grit typically larger than 12 microns and various combinations of bronzes, brasses, nickel, tungsten and tungsten carbide to provide desired ductility, impact resistance, heat dissipation and abrasion resistance characteristics. The ultra fine cobalt reacts with the diamond particles during sintering to form a strong bond with diamond particles in the form of cobalt nodules bonded to the diamond surfaces without altering diamond to carbon. In that almost 100% of theoretical density of the ultra fine cobalt powder is achieved at 850°C, effective matrix sintering and bonding can be accomplished at below 1000°C, in the preferred range of 750° to 1000°C, to bond dense cobalt to the diamond particles below the temperature of about 1000°C above which diamond becomes brittle.

It will be understood that other embodiments and examples of the invention will be readily apparent to a person skilled in the art, the scope of the invention being defined in the appended claims.