CATALYTICALLY ACTIVE ANODE ^" This invention relates to a catalytically active anode,, Many suggestions have been made for improving anodes, especially in brine electrolysis. Thus, as one example among many, US Patent Speci ication θ4θ93^ discloses an anode having a conductive substrate on the surface of which there is coated a

"metal suicide", the metal being one of Pt, Pd, Ir, Eh, Cr, Co, Ni, Ru, Ti, V, Zr, Nb, Hf, Ta and V.

The activity of these anodes has now been discovered to stem from the formation of certain compounds thereon in use, and the invention consists in the making and exploitation of this discovery to make better anodes. Thus, certain "metal suicide" anodes have been discovered to develop, in use, superficial metal silicates which are electrocatalytically active.

Accordingly the invention consists in a catalytically active anode comprising at its surface an at least semiconductive water- insoluble metal silicate. Preferably, the anode is made by dispersing the metal silicate in a non-conductive binder, and applying the dispersion to a conductive substrate. Preferably, the silicate is reduced sufficiently to improve its electrical conductivity. Preferably the binder is hydrophobic, for example polytetrafluoroethylene, but the dispersion as a whole should not be excessively hydrophobic.

The metal silicate may have a spinel or olivine structure, both of which are reported to have been synthesised. The olivine is convenient. It is preferably cobalt orthosilicate (Co SiO. )

_eέ _t or may be nickel silicate (Ni SiO, ). Other silicates which could be used are those of other transition metals such as iron, or of copper or precious metals. Mixed silicates may also be used. The substrate is preferably in the form of an electrically conductive sheet (foraminate or otherwise), slit metal or mesh, preferably of nickel. If of iron, the substrate preferably contains under 1.5% (more preferably under -0.3%) of carbon.

__ OMPI

Preferably the weight ratio of binder to metal silicate is from l-3 _> more preferably from 1:5 _> to 1:20 and the dispersion is preferably applied to give a coating of from 5 ^m 9 to 50mg of the metal silicate per square centimetre of the substrate. The invention extends to a catalytically active anode made as set forth above, and to an electrolytic cell including a cathode and such an anode. The invention further extends to a method of electrolysis using this cell, especially of aqueous solutions, e.g. aqueous alkali. The invention will now be described by way of example. MAKING AN ANODE

Although known.' syntheses of transition metal silicates include precipitation from sodium silicate solution, and gelling using tetraethyl orthosilicate, it was found that solid state sintering gave the best results.

Cobalt nitrate was ground with a stoichiometric amount of silica with a pestle and mortar. The mixture of cobalt nitrate and silica was transferred to an evaporating basin, gently heated to allow the nitrate to dissolve in its own water of crystallis- ation and then vigorously heated by bunsen burner to dry and decompose the nitrate to its oxide. The resultant powder was o , ground again and then heated at 1000 C for 24 hours, to promote the reaction 2Co0 + SiO —>Co SiO.. The product was ground again, washed with 5M KOH and then with concentrated HN0 to remove any unreacted silica or cobalt oxide, then washed with water, dried, ground again and put in a dry sample bottle. That the product was cobalt silicate (olivine structure) was confirmed by X-ray powder diffraction.

It was found that the cobalt silicate prepared had a specific - —*1 —."1 conductivity of 2 x 10 — _ cm or lower, which was far too low to contemplate its use as an electrocatalyst. Therefore it was decided to try to improve the conductivity by standard methods used in semiconductor electrochemistry. It was found that heating

OMPI /., WIPO

under argon did not improve the •conductivity, but the conductivity of cobalt silicate heated at 800 C for 2 hours under hydrogen rose

-2 -1 -1 to 6 x 10 - > cm . X-ray photos (and subsequent electrochemical performance) showed that the silicate had not been reduced by any degree to the metal, or to metal oxide and silica. (Parallel experiments on iron silicate when treated in the same way showed some iron Present, and nickel silicate was completely reduced to metal and silica. However the literature on the reduction of silicates by CO indicates that if the conditions are altered then conducting non-reduced iron and nickel silicates may be obtained. ) A mixture of the treated cobalt silicate, polytetrafluoro¬ ethylene and water was dispersed in an ultrasonic bath, and the dispersion painted onto a nickel wire mesh 1 cm square. The electrode so formed was dried and cured for 1 hour at 300 C. Initially a cobalt silicate: polytetrafluoroethylene ratio of 10:3 was used, but as this proved to be too hydrophobic, the present electrode used a ratio of 10:1, giving a cobalt silicate loading of 21 mg/ /cm2.

ELECTROLYSIS An aqueous solution of 5 KOH was electolysed at 40 C in a cell having a nickel mesh cathode and the above anode. (The anode had previously been anodised for 2 hours at 2V vs the dynamic hydrogen electrode, to activate it, possibly by forming higher oxides on its surface and hydrσphiusing it. ) The current was held at 1 amp. Oxygen was evolved at the anode, at which the voltage vs. the dynamic hydrogen electrode rose from an initial' 1700mV, neglecting early fluctuations, to 174θmV after 40 days, possibly due to formation of a poorly conductive nickel oxide at the interface between screen and cobalt silicate. An identical cell (but with cobalt silicate loading of l4 mg/cm ) was run at 25 C, 50 C and 70 C at up to 2-g-A/cm . At

2 2A/cm , the iR-corrected voltages at the anode vs. the dynamic hydrogen electrode were 1710mV, l650mV and l620mV respectively.

(Parallel experiments on iron silicate suggested that the

2 iron silicate might be corroding at up to 6 mg/cm /week at

2 lA/c , before and after which the voltage was l68θmV and 1720mV respectively. )