work showed just how exhaustive their approach had been. They firstly varied
the source of ammonia. Experiments 1–36 used commercially available
solutions of ammonia together with ammonium nitrate, whist for 37–49,
gaseous ammonia was taken from a cylinder. Different sources of ammonia
generated different flow rates but this was regulated by overcoming the
resistance of a manometer of controllable height.
The decomposition and development reaction tubes were always side by
side in the oven. These tubes, (60 cm in length) were therefore always at the
same constant temperature (within 5° at 1000 °C) and it was here that the
catalyst resided, so that the gases never passed over catalyst in cooler parts of
the oven. The gases were also allowed to remain in these tubes for hours to
ensure that equilibrium was reached. Various materials were also used for
these tubes such as quartz and china, the general idea here being to make the
tube less permeable to the diffusion of gases that they felt was a fault
22 with
the Nernst set-up. Temperature measurement was via a thermocouple and
galvanometer, one end of the thermocouple lay in the oven—pushed into
both tubes—the other lay in ice. This device was calibrated using the gold
melting point at 1062 °C giving an accurate galvanometric determination of
oven temperature.
Four different catalysts were used viz., iron, nickel, chromium and manganese and these were prepared in the laboratory to control their purity.
23 All
the catalysts, except nickel, were ‘raised’ on iron free asbestos. Prior to use, the
asbestos was, for ‘day after day’, boiled with concentrated hydrochloric acid
and heated in a hydrogen current for at least 10 h, followed by renewed
boiling with the acid. The asbestos was impregnated with the catalyst, and the
amount used varied depending on the experiment. In the case of iron,
between 10 mg and 1 g were examined. Between experiments, the apparatus
was cleansed of any undecomposed ammonia by flushing it overnight
through auxiliary washing flasks containing 20% sulphuric acid with glass
wool filter plugs to prevent any acid vapour remaining in the apparatus
afterwards. The remaining part of their description of the experimental setup
concerned quantitative aspects of the ammonia determination, and this aspect
of the paper was critical because Nernst was unhappy with ‘Haber’s’ earlier
experimental accuracy when determining the low values of ammonia present
at 1 atmosphere equilibrium pressure.
Immediately after the decomposition reaction tube were two sequentially
placed Volhard flasks modified so that they required only 15 cc of titration
acid to fill them. In this respect the flasks were originally charged with 10 cc
of 0.02N HCl to which 5 cc of distilled water
24 were added. Two such flasks
were also placed just after the development tube. In each case, the second
116
D. Sheppard
the source of ammonia. Experiments 1–36 used commercially available
solutions of ammonia together with ammonium nitrate, whist for 37–49,
gaseous ammonia was taken from a cylinder. Different sources of ammonia
generated different flow rates but this was regulated by overcoming the
resistance of a manometer of controllable height.
The decomposition and development reaction tubes were always side by
side in the oven. These tubes, (60 cm in length) were therefore always at the
same constant temperature (within 5° at 1000 °C) and it was here that the
catalyst resided, so that the gases never passed over catalyst in cooler parts of
the oven. The gases were also allowed to remain in these tubes for hours to
ensure that equilibrium was reached. Various materials were also used for
these tubes such as quartz and china, the general idea here being to make the
tube less permeable to the diffusion of gases that they felt was a fault
22 with
the Nernst set-up. Temperature measurement was via a thermocouple and
galvanometer, one end of the thermocouple lay in the oven—pushed into
both tubes—the other lay in ice. This device was calibrated using the gold
melting point at 1062 °C giving an accurate galvanometric determination of
oven temperature.
Four different catalysts were used viz., iron, nickel, chromium and manganese and these were prepared in the laboratory to control their purity.
23 All
the catalysts, except nickel, were ‘raised’ on iron free asbestos. Prior to use, the
asbestos was, for ‘day after day’, boiled with concentrated hydrochloric acid
and heated in a hydrogen current for at least 10 h, followed by renewed
boiling with the acid. The asbestos was impregnated with the catalyst, and the
amount used varied depending on the experiment. In the case of iron,
between 10 mg and 1 g were examined. Between experiments, the apparatus
was cleansed of any undecomposed ammonia by flushing it overnight
through auxiliary washing flasks containing 20% sulphuric acid with glass
wool filter plugs to prevent any acid vapour remaining in the apparatus
afterwards. The remaining part of their description of the experimental setup
concerned quantitative aspects of the ammonia determination, and this aspect
of the paper was critical because Nernst was unhappy with ‘Haber’s’ earlier
experimental accuracy when determining the low values of ammonia present
at 1 atmosphere equilibrium pressure.
Immediately after the decomposition reaction tube were two sequentially
placed Volhard flasks modified so that they required only 15 cc of titration
acid to fill them. In this respect the flasks were originally charged with 10 cc
of 0.02N HCl to which 5 cc of distilled water
24 were added. Two such flasks
were also placed just after the development tube. In each case, the second
116
D. Sheppard