from a burette into a known quantity of the cyanide solution. For this purpose a standard solution of silver nitrate is usually made up, of such a strength that the result may be read off without calculation.

Such a standard solution of silver nitrate may be made by dissolving 3.262 grams of silver nitrate (AgNO3) in distilled water and making up to 1 litre and this is the standard adopted and referred to in the following pages. (Should the least milkiness develop while dissolving it is an indication that the water is not sufficiently pure). The reaction taking place between the silver nitrate and potassium cyanide may be represented by the following equation: AgNO3 + 2KCN AgK (CN)2 + KNO3

=

Thus 169.9 gm. AgNO3 saturates 130.2 gm. KCN

2.5 gm. KCN

or 3.262 gm. AgNO3 saturates Therefore 1 cc of above solution (=0.003262 gm. AgNO3) saturates 0.0025 gm. KCN. If 25 cc of mill solution is taken for titration, then 1 cc of nitrate solution = 0.0025 gm. KCN = 0.01% or 0.1 kilo KCN per metric ton or 0.2 lb. KCN per 2000 lb. ton of solution. Similarly, a solution containing 6.525 grams of silver nitrate per litre may be made up, 1 cc of which contains 0.006525 grams of AgNO3 equal to 0.005 gm. of KCN. Then if 10 cc of mill solution be taken for titration, 1 cc of nitrate solution = 0.005 gm. of KCN 0.05% 0.5 kilo per metric ton or 1 pound per 2000 lb. ton of solution; while if 50 cc of mill solution be taken, 1 cc of nitrate solution 0.1 kilo per metric ton or 0.2 lb. per 2000 lb. ton of solution. Probably the most common strength for the standard silver nitrate solution is 13.05 grams per litre, and is such that if 10 cc of mill solution be taken for titration 1 cc of nitrate solution equals 0.1% KCN or 2 lb. per ton. This strength, however, is not recommended for reasons which will be stated later.

=

=

0.01%

=

Free Cyanide Determination.-Two methods are in use for this purpose, and there is a third that may be used on occasion, but it should be understood that there is no direct method yet devised which will give better than comparatively correct results.

If the determination, however, can be relied on to be relatively uniform and consistent, that is all that is required for practical working.

Method No. 1.-Take 25 cc of the solution to be tested and place in an Erlenmeyer flask, add a few drops of 5% solution of potassium iodide free from alkali, and slowly run in the silver nitrate from the burette, shaking the flask meanwhile, until a distinct yellow opalescence appears. Then read off the number of cc of silver nitrate used and multiply by 0.01, and the result will be the amount of free cyanide in the solution in terms of percentage. To get the cyanide in kilos per metric ton of solution multiply the number of cc of silver nitrate by 0.1 or if the cyanide in terms of lb. per 2000-lb. ton of solution is desired multiply the number of cc of silver nitrate by 0.2. If the cyanide strength is over 0.2% KCN it is better to take only 10 cc of the cyanide solution for the test. In this case, when using as the standard solution 3.262 grams AgNO3 per litre, after the burette reading. has been noted move the decimal point one place to the left and divide by 4. For instance if B = burette reading in cc then

BX0.1

B

=

4

% KCN or

=

4

BX 2
4

B

=

2

kilos per metric ton or = lb. KCN per 2000-lb. ton of solution. Method No. 1 is not to be recommended when the solution contains zinc (as it usually does), because the results are not uniform. When there is no free alkali in the cyanide solution the value obtained for free cyanide may approximate the correct one, but if any free alkali is present (and it almost always is, in a working solution), the cyanide reading will be in excess of the correct figure, and this excess error will be greater in proportion as the amount of free alkali is greater, until, if the latter happens to be about 0.08% (in terms of NaOH) or over, the reading supposed to represent free cyanide will coincide exactly with the amount of "total" cyanide, as shown in the method to be described later.

✓ Method No. 2.—In place of the foregoing the following is recommended, as giving the most correct and consistent reading

for free cyanide under ordinary conditions. It needs a certain amount of practice to stop at the exact end-point of the reaction, but precision is easily and rapidly acquired.

Take 25 cc of the solution to be tested. (If not absolutely clear and colorless to transmitted light, filter and re-filter until it is.) Pour into a bulb parting-flask with the neck cut short, Fig. 1, or into a small conical parting-flask, Fig. 2, first having seen that the

FIG. 1.

flask is as clean and transparent as the solution, for which purpose wash with some weak hydrochloric acid solution and rinse well. Do not add water to the solution to be tested, because it will tend to dissociate some of the zincpotassium cyanide and make the free cyanide reading too high. Place the burette opposite a window with a good FIG. 2. light (but not in the direct rays of the Parting Flasks. sun), and fix a black card vertically about 4 inches behind the spot where the flask is to be held, so that the light will strike the surface of the solution over the top of the card leaving the solution itself in shadow. Then, keeping the eye on a level with the flask, watch the behavior of every addition of silver nitrate as it touches the solution, not keeping it in continuous agitation but allowing the bluishwhite cloud formed by each addition of silver to hang for a second or two before shaking and dissolving. The finish is indicated by the first appearance of a bluish haze dulling the original brilliancy of the solution. With a very little practice the operator will find he can check himself and others within a tenth of a cc on the burette. To ensure the success and delicacy of this reaction it is important that the silver nitrate solution should be dilute, and for this reason the standard already recommended (3.262 grm. AgNO3 per litre) is to be preferred to the usual standard of 13.05 grm. per litre. With a strong silver nitrate solution the tendency is to precipitate flocks instead of the dispersed opalescence necessary for a clearly defined end point.

Number of cc of standard AgNO3 X 0.01 = % free KCN or -X 0.1 kilos KCN per metric ton of solution or X 0.2 = pounds KCN per ton (2000 lb.) of solution.

=

(NOTE). It sometimes happens (that is, in the case of certain solutions) that the addition of a few drops of potassium iodide solution before titration will render this first haziness more easily distinguished and may give a slightly lower reading than without The reason of this phenomenon is not clear and the writer is not prepared to offer any explanation of it. In such cases, however, there is no harm in making use of it so long as it is clearly understood that it is the first faint blue-white opalescence that is to be looked for and not the yellow tint.

Walter Virgoe1 has pointed out that cyanide solutions containing copper, when titrated with silver nitrate without potassium iodide indicator, yield a result higher than the true value for free cyanide, owing to part of the copper-potassium cyanide being recorded as free cyanide. Thus, while in presence of zinc and free alkali the addition of neutral potassium iodide as indicator gives a higher and less correct reading than is recorded without the indicator, when copper is present without zinc the opposite is the case and the addition of neutral iodide indicator gives a lower reading than without, this lower reading being the correct one. So, when zinc is absent and copper present, method No. 1 should be used.

Method No. 3.-When copper and zinc are present together both tendencies are displayed in proportion to the relative amounts of each metal, and the result is thereby complicated. In this case, it is advisable to make titrations both with and without KI indicator and to base solution control on the lower of the two readings; or a method devised by Clennell2 for determining free cyanide may be tried, though the writer has not always found it to give concordant results.

First determine protective alkali, by Green's method, carefully running in the acid drop by drop and recording the number

1 Proceedings I. M. M. (London), Vol. X, page 102.

2 Chemistry of Cyanide Solutions, page 25 (Second Edition).

of drops needed. (For this purpose and for the subsequent neutralization a mineral acid should be used, say decinormal nitric acid, because oxalic would precipitate the lime in the cyanide solution as oxalate, which would mask the end-point in the test for cyanide.) Then measure out another 25 cc of solution to be tested and run into it the same number of drops of decinormal acid as was previously recorded. The whole of the protective alkali in the cyanide solution is thus exactly neutralized so that it does not militate against the use of the iodide indicator. To the test solution so neutralized add at least 2 cc of neutral 5% solution of potassium iodide, and titrate with standard silver nitrate to first white flocculent precipitate. Calculate free cyanide as already explained.

The writer has found this white end point rather indeterminate, but the yellow end point is well marked and the readings uniform though the latter gives a higher and probably less correct result than the white end point.

In the use of this method the quantity of KI added has an important bearing on the result of the titration. The writer has found that smaller quantities than that stated give proportionally higher cyanide readings, though the amount of iodide may be materially increased without much apparent effect on the result. Total Cyanide Determination.-The term "total" cyanide is generally limited to mean zinc-potassium cyanide + free potassium cyanide + hydrocyanic acid. To determine this, make up an indicator consisting of 4 grams caustic soda and 1 gram potassium iodide in 100 cc of distilled water. Take 25 cc of the solution to be tested, add 3 to 5 cc of the indicator, and titrate with standard silver nitrate until a distinct lemon yellow opalescence appears.

=

Number of cc AgNO3 used X 0.01 % "total" cyanide. Keep stock bottle of indicator well stoppered to avoid absorption of carbonic acid, which will tend to make the test cloudy with carbonate of lime. Should a whitish precipitate of zinc dense enough to mask the color come down before the yellow tint appears Clennell recommends the addition of a little ammonia before beginning the titration, but this is usually unnecessary.