Before attention was directed to the adsorption of gases on the surfaces of solids much work was done on the “occlusion” at higher temperatures. Above 400° C. solution usually occurs rapidly, and because of the decrease in surface by sintering, the adsorption is negligible compared with the absorption. In this paper, investigations on the sorption of hydrogen on copper are described at temperatures intermediate between 25° C. when adsorption is the principal phenomenon and 200° C. when solution has become important. Over this range of temperature both adsorption and absorption have been measured. On bringing the hydrogen into contact with the copper there was always an immediate fall in pressure attributable to adsorption, followed by a slower fall as absorption proceeded. This latter process, of course, became quicker at higher temperatures.
Experimental
.
Apparatus
.—The apparatus used was almost exactly the same as that described in the previous paper. The only difference was that instead of the thermostat at 25° C. a furnace was used. A copper tube about 50 cm. long and 5 cm. in diameter, wound with nichrome wire, had placed inside it, for half its length, a tightly fitting iron tube with thick walls (1 cm.). The adsorption bulb went into this half, and the low conductivity and large thermal capacity of the iron hindered fluctuations in temperature from reaching the bulb. In the lower half of the furnace, where the absence of the iron tube allowed the temperature to vary promptly with a changed heating current, was a thermoregulator bulb containing air, connected by capillary tubing to a U-tube in which mercury made contact with a tungsten point. On the other side of the U-tube a system of the same volume, with a bulb immersed in a thermostat, counteracted the effect of alterations of room temperature. An extra U-tube of mercury enclosed nitrogen around the spark gap to prevent dirtying the mercury surface by oxidation. With this arrangement the temperature could be kept constant to within half a degree for any length of time. Temperatures were measured by a chromel-alumel thermocouple calibrated at the boiling points of suitable liquids.
I. Thermo-electric behaviour of aqueous solutions with mercurial electrodes
