should sum up to give the original weight of the sample. The solution remaining
after filtering off the silver chloride was treated with potassium hexacyanoferrate
(phlogisticated alkali) to precipitate any traces of other metals. By finally adding
alkali carbonate, any earths could be precipitated.
Silver ores may also be composed of silver, sulphur and arsenic. In this case,
boiling with dilute nitric acid gave a white powder and a solution. As before, silver
was precipitated from the solution with sodium chloride. The white powder was
treated with Aqua regia to dissolve the arsenic (III) oxide, which could be
re-precipitated with water (as As 2 O 3 ). It is likely that the use of Aqua regia gave
losses of arsenic, as a portion of arsenic was most probably oxidised to soluble
arsenic acid. The sulphur remaining could contain traces of silver chloride, which
was extracted from the residue with aqueous ammonia.
The next type of ores contained silver, sulphur, copper and arsenic. Boiling with
dilute nitric acid gave a white powder and a solution containing copper and silver.
Bergman noted that silver and copper could not be efficiently separated by sodium
chloride, since the precipitate contained copper. Instead, silver was precipitated by
addition of a known amount of copper. Copper could finally be precipitated with
iron or alkali carbonate, taking care to subtract the amount of copper used to
precipitate silver. The white powder obtained in the first step was treated with
hydrochloric acid to dissolve the arsenic (this was probably safer than using Aqua
regia, as stated above). Arsenic(III) oxide was re-precipitated with water. The
sulphur was treated with aqueous ammonia to detect traces of silver chloride or
copper salts (which give a deep blue colour with ammonia). If the silver ore
contained antimony, the latter could be separated in the same way as arsenic.
The final type of ores was silver combined with hydrochloric and sulphuric acid.
These ores were powdered and heated for a day with hydrochloric acid. Sulphate
was precipitated from the solution with barium nitrate. The amount of sulphate
could be calculated from the weight of the barium sulphate precipitate, but in this
case Bergman’s determination of the composition of barium sulphate was very far
from the true value: Bergman reports that it contains 15% acid, while the true value
is 41%. If Bergman failed to determine the composition of barium sulphate, he was
more successful in other case, as can be seen from Table 23.1.
A mistake that Bergman made, and which was not so easily realised at the time,
was that he precipitated several metals as carbonates (copper, tin, nickel and
cobalt), but these metals may form hydroxycarbonates of varying compositions, and
were therefore not suitable for accurate quantitative analysis. In these cases, it is not
easy to judge Bergman’s accuracy. Another problem faced when attempting to
reproduce Bergman’s experiments is the fact that any access to carbonate raises the
pH, causing an increase in the solubility of copper due to complexation. The
precipitation of iron as Prussian blue is also difficult to perform; due to the small
particle size, the precipitate is often very difficult to collect quantitatively by filtration. Although latter authors managed to get more accurate results, Bergman’s
great authority caused his values to be trusted for a long time [40].
23.4 Mineral Analysis in Solution
321
after filtering off the silver chloride was treated with potassium hexacyanoferrate
(phlogisticated alkali) to precipitate any traces of other metals. By finally adding
alkali carbonate, any earths could be precipitated.
Silver ores may also be composed of silver, sulphur and arsenic. In this case,
boiling with dilute nitric acid gave a white powder and a solution. As before, silver
was precipitated from the solution with sodium chloride. The white powder was
treated with Aqua regia to dissolve the arsenic (III) oxide, which could be
re-precipitated with water (as As 2 O 3 ). It is likely that the use of Aqua regia gave
losses of arsenic, as a portion of arsenic was most probably oxidised to soluble
arsenic acid. The sulphur remaining could contain traces of silver chloride, which
was extracted from the residue with aqueous ammonia.
The next type of ores contained silver, sulphur, copper and arsenic. Boiling with
dilute nitric acid gave a white powder and a solution containing copper and silver.
Bergman noted that silver and copper could not be efficiently separated by sodium
chloride, since the precipitate contained copper. Instead, silver was precipitated by
addition of a known amount of copper. Copper could finally be precipitated with
iron or alkali carbonate, taking care to subtract the amount of copper used to
precipitate silver. The white powder obtained in the first step was treated with
hydrochloric acid to dissolve the arsenic (this was probably safer than using Aqua
regia, as stated above). Arsenic(III) oxide was re-precipitated with water. The
sulphur was treated with aqueous ammonia to detect traces of silver chloride or
copper salts (which give a deep blue colour with ammonia). If the silver ore
contained antimony, the latter could be separated in the same way as arsenic.
The final type of ores was silver combined with hydrochloric and sulphuric acid.
These ores were powdered and heated for a day with hydrochloric acid. Sulphate
was precipitated from the solution with barium nitrate. The amount of sulphate
could be calculated from the weight of the barium sulphate precipitate, but in this
case Bergman’s determination of the composition of barium sulphate was very far
from the true value: Bergman reports that it contains 15% acid, while the true value
is 41%. If Bergman failed to determine the composition of barium sulphate, he was
more successful in other case, as can be seen from Table 23.1.
A mistake that Bergman made, and which was not so easily realised at the time,
was that he precipitated several metals as carbonates (copper, tin, nickel and
cobalt), but these metals may form hydroxycarbonates of varying compositions, and
were therefore not suitable for accurate quantitative analysis. In these cases, it is not
easy to judge Bergman’s accuracy. Another problem faced when attempting to
reproduce Bergman’s experiments is the fact that any access to carbonate raises the
pH, causing an increase in the solubility of copper due to complexation. The
precipitation of iron as Prussian blue is also difficult to perform; due to the small
particle size, the precipitate is often very difficult to collect quantitatively by filtration. Although latter authors managed to get more accurate results, Bergman’s
great authority caused his values to be trusted for a long time [40].
23.4 Mineral Analysis in Solution
321