The present invention relates to a method of produc¬ ing metal composite powder such as for example cemented carbide.

Swedish patent 502 754 discloses a method of coating hard constituent powders with Co and/or Ni with the polyol process disclosed in US 4,539,041 and used today for the manufacture of cobalt and nickel metal powders with a small particle size. According to said Swedish patent hard constituent powder in suspension in a polyol solution containing a suitable salt of cobalt and/or nickel, during reduction of said metals by the polyol obtains a cobalt and/or nickel metal precipitation on the surface. The metals are precipitated with a quite even distribution over the surface of the carbides without forming separate islands. However, the require¬ ment of a large excess of ethylene glycol and the tech¬ nical difficulties in separating and removing the ethy- lene glycol oxidation by-products interfere with the economy of the process. A mixture of byproducts are formed during the reduction which complicates separation of un-reacted ethylene glycol. A suitable strategy to improve the process economy is to recycle or decrease an excess amount of ethylene glycol needed for the cobalt or nickel reduction.

It is an object of the present invention to improve the cost efficiency of the process for coating hard constituent powders with cobalt or nickel metal using the polyol process disclosed in SE 502 754.

In a first embodiment of the present invention the consumption of ethylene glycol is economized by inter¬ rupting the reaction before any by-products are formed and completing the reduction by heat treatment in the dry state. A complete reduction of the Co^ ⁺ in the

intermediate phase on WC can be accomplished by reduc¬ tion in H2 at 550 °C for 24 hours. In this way only a stoichiometric amount of the ethylene glycol is consumed and the excess of ethylene glycol is prevented from being polluted with the oxidation byproducts that are formed when Co^ ⁺ is reduced in solution. The ethylene glycol can thus be re-used several times without purifi¬ cation.

In a second embodiment the ethylene glycol that has been used for reducing one batch of Co (OH)2 to cobalt metal on WC is re-used in an identical cobalt reduction. If an excess of about ten times of ethylene glycol is used there is sufficient un-reacted ethylene glycol left in the reaction mixture to reduce an additional amount of Co(OH)2- It has been found that the byproducts pre¬ sent do not interfere with the reaction. Re-using ethy¬ lene glycol in a reaction period of one hour is more ef¬ ficient than re-using in a five hour reduction period. The cobalt in the powders is reduced by H2 in the dry state.

In a third embodiment the excess of ethylene glycol is reduced by a factor of two and more. Normally an excess of ten times the stoichiometric amount of cobalt is used to achieve adequate stirring of the highly viscous suspension. Preferably, the excess is reduced to five times and even to three times the stoichiometric amount. In this case the stirring of the suspension becomes less effective since the mixture becomes highly viscous. However, the stirring has been found to be sufficient to give a rather even distribution of the precipitated cobalt metal on the WC surface. By using an excess of three to four times of the ethylene glycol complete reduction of the cobalt is still achieved but the distribution of the cobalt metal on the WC surface is less even than when a larger excess is used. Decrease

of the ethylene glycol excess is easier accomplished when WC is kept in suspension compared to when (Ti,W)C is kept in suspension where an excess of no less than five times should be used. In the case of nickel the reduction is generally faster, the yield is higher and the reduction seems to require a smaller excess of ethylene glycol.

Preferably, the process economy is optimized by a combination of separation of the intermediate solid compound and WC from the ethylene glycol mixture after 15-60 minutes of reaction and reduction of the interme¬ diate solid compound in the dry state by H2 at 550 °C for around 24 hours combined with re-use of the ethylene glycol . This would best be accomplished in a crosscur- rent mode.

In an alternative method colloidal Co(OH) ₂ is pre¬ cipitated from an aqueous solution of Co(CH3COO)2 4H20 by addition of NH ₃ or a hydroxide. A precipitate is formed on the WC surface which is separated from the so- lution and reduced by heat treatment in the dry state. The invention has been described with reference to WC and Co but can also be applied to Ni, (Ti,W)C and other hard constituents.

Example 1

94 g WC was suspended in 120 ml of ethylene glycol in a 500 ml stirred glass reactor equipped with a ther¬ mometer and an air-cooled condenser for the removal of volatile by-products while most of the ethyleneglycol was recycled. 10.07 g Co(OH)2 was added while stirring. The excess of ethylene glycol was ten times the stoi¬ chiometric amount. The amount of dry substance was 44 weight %. The suspension was heated above 180 °C and was kept at this temperature for the given reaction time. The solid phase was then separated from the ethylene

glycol by centrifugation, washed with ethanol and dried overnight at 40 °C.

The product mixtures obtained after the investigated reaction times of 30, 45, 60, 75, 90 and 120 minutes consisted in all cases of two partially mixed solid phases . One grey phase of WC and one phase that varied in colour with the reaction time from pink after 30 minutes of reaction to purple after one hour and then back to pink after 75 minutes of reaction and finally the phase turned brown via reddish-brown after 120 minutes of reaction. The residual ethylene glycol phase was in all cases turbid. After several days of sedimen¬ tation the ethylene glycol became clear with a gelati¬ nous brown phase at the bottom. The colour of the ethylene glycol had turned yellow after 30 minutes of reaction and yellow-brown after 45 minutes of reaction. After 75 minutes the residual ethylene glycol had obtained a dark brown colour.

Two different phases could be distinguished in SEM- microscopy. The phases were mixed to some extent but there were also particles of around 10 μm present con¬ sisting of a phase different from WC and cobalt metal.

The X-ray powder diffraction showed that the strong¬ est peak of the intermediate complex of Co^ ⁺ and ethy- lene glycol could be detected in all samples. After 90 minutes of reaction the strongest peak from cobalt metal started to become distinguishable.

The yield of cobalt in the samples varied between 80 and 94 %.

Example 2

The sample reduced during one hour was used for further experiments where the solid phase was reduced by heat treatment in the dry state. The samples used for the reduction by heat treatment in the dry state were

reduced in ethylene glycol for one hour before separa¬ tion and consisted of two partially mixed solid phases: one grey WC phase and one pink Co ²⁺ -ethylene glycol complex phase. After reduction under H2 atmosphere at 550 °C for 24 hours the sample appeared homogeneously grey in colour. SEM-examination showed that there were spherical, pre¬ sumably cobalt metal particles present as well as parti¬ cles around 10 μm consisting of a phase other than WC and similar to the particles present before the reduc¬ tion.

In the X-ray powder diffraction spectrum only WC and cobalt metal (cubic) were detected.

Example 3

28.2 g of WC was suspended in 35 ml of ethylene gly¬ col that had already been used in one identical reduc¬ tion using the same apparatus as in Example 1. 2.97 g Co(OH) 2 was added while stirring. The excess of ethylene glycol was ten times the stoichiometric amount and the amount of dry substance was 44 weight %. The suspension was heated with adequate stirring and allowed to boil for five and one hour respectively. The solid phase was then separated from the ethylene glycol by centrifuga- tion, washed with ethanol and dried at 40 °C overnight. The sample obtained after re-using an ethylene gly¬ col mixture from a five hour reduction in an additional reduction of the same amount of Co(OH)2 for 5 h con¬ sisted of two partially mixed phases: one grey WC phase and one white-pink phase. In the SEM two different phases could be distinguished that were partially mixed but a phase other than WC also formed separate islands of >10 μm. The X-ray powder diffraction spectrum showed no cobalt metal present. Only peaks from WC and from un- identified phases were found.

The sample obtained after re-using an ethylene gly¬ col mixture from a one hour reduction to reduce the same amount of Co(OH)2 again for one hour also consisted of two partially mixed phases, one grey WC phase and one pink phase. In the SEM the sample looked the same as the sample obtained after one hour of reduction in fresh ethylene glycol. In the X-ray powder diffraction spect¬ rum no other peaks than those for WC could be detected. The total yield of cobalt in these two samples was around 87 %.

Example 4

18.8 g WC was suspended in 11.5 ml ethylene glycol in a 250 ml stirred glass reactor using the same appara- tus as in Example 1. 2.02 g Co(OH) 2 was added while stirring and the suspension was heated until boiling. The excess of ethylene glycol was five times and the stoichiometric amount and the amount of dry substance was 62 weight % . The reaction mixture was allowed to boil for 5 hours and the solid phase was then separated from the ethylene glycol by centrifugation, washed in ethanol and dried at 40 °C overnight.

The same reaction procedures were then repeated but with the excess of ethylene glycol reduced further to 9 ml corresponding to an excess of between 3 and 4 times the stoichiometric amount (some ethylene glycol was lost during the distillation) and the amount of dry substance was 68 weight %.

In both experiments using five and three to four times excess of ethylene glycol for five hours of reac¬ tion the samples appeared to be homogeneously grey. SEM- studies showed that there were spherical cobalt metal particles present on the WC surface and no other sepa¬ rate phase present. The spherical particles appeared to be somewhat more evenly distributed on the WC surface

when an excess of five times of ethylene glycol was used as compared to when an excess of three to four times was used.

In the X-ray powder diffraction spectra only peaks from cobalt metal and WC could be detected in both samples. The yield of cobalt metal seemed to decrease when the excess of ethylene glycol was decreased and was 85 % in these experiments as compared to 94 % when ten times excess of ethylene glycol was used.

Example 5

26.7 g of (Ti,W)C was suspended in 35 ml of ethylene glycol that had already been used in one identical reduction, (the ethylene glycol contained a small amount of H2SO4 to increase the solubility of i(OH)2) . The ex¬ cess of ethylene glycol was five times the stoichiomet¬ ric amount and the amount of dry substance was 44 weight %. 5.69 g Ni(0H)2 was added while stirring. The suspen¬ sion was heated with adequate stirring and allowed to boil for 5 hours. The solid phase was then separated from the ethylene glycol by centrifugation, washed with ethanol and dried at 40 °C overnight.

The precipitate was homogeneously grey in colour. The SEM micro graphs showed an even distribution of spherical particles with a particle size of around 1 μ on the (Ti,W)C surface. In the X-ray powder diffraction spectrum no other peaks than those of nickel metal and (Ti,W)C could be detected. The yield of nickel was 99 %.

Example 6

13.489 g Co(CH ₃ C00)2 4H ₂ 0 was dissolved in 50 ml water by heating. 47 g WC was suspended and a solution of 4.07 g NaOH (stoichiometric amount) dissolved in 10 ml water was added drop-wise during five minutes while stirring. The reaction mixture was stirred for another

20 minutes and the solid phase was then separated from the water by filtration, washed with water and dried at 40 °C overnight.

The sample appeared to be homogeneously dark in col¬ our with a quite even distribution of an amorphous phase on the surface of WC. In the X-ray powder diffraction spectrum only the peaks from WC and Co(OH)2 could be ob¬ served. The yield of cobalt was 87 %.