This method of analysis is generally applied to the determination of small percentages of an element or compound, and in this chapter the following will be considered:

(a) The Colorimetric Estimation of Copper.
(b) The Colorimetric Estimation of Carbon in Steel.
And though not coming under this head,
(c) The Volumetric Estimation of Copper by Potassium Cyanide.

This last method is of considerable importance to the metallurgical chemist, and, strictly speaking, cannot be introduced under any of the five heads laid down ; therefore, for convenience, it is introduced here.

(a) THE COLORIMETRIC ESTIMATION OF COPPER.

Apparatus, Reagents.—The usual apparatus, including two burettes or Nessler tubes. For preparing the comparison standard pure electrolytic copper is required. For analysis the student may obtain a sample of copper slag.

Method, Reactions.—A certain amount of pure copper is dissolved in a definite quantity of HNO3, and a known amount of NH4HO added. After dilution to a given volume, the solution is transferred to a burette. The slag is so treated that the copper is separated out, preferably in the metallic form, and this separated copper is treated exactly the same as the pure copper, and is transferred to another burette or tube without dilution. Distilled water is added to the assay solution till the colours in the two tubes are exactly the same. Note the volume of the assay solution; and as the value of 1 c.c. of the comparison standard is known, the value of the assay solution is easily calculated.

This method was originated by Heine. For other colorimetric methods for copper consult the works of Beringer, Furman, and Sutton.

The Comparison Standard. – In a flask with a small funnel in its neck dissolve 1 gm. of pure electrolytic copper in the smallest possible quantity of nitric acid. When dissolved, dilute and transfer to the 1000 c.c. flask, and make up to the mark at 16° C.

Ten c.cs. of this solution contain .01 gm. copper. To 10 c.cs. of this solution add 1 c.c. 16E. HNO3 and 4 c.cs. 20E. NH4HO. Transfer this solution to the Nessler tube, rinsing the beaker out with distilled water.

If a burette is used, the student must ascertain the volume of the burette from the 50 c.c. mark to the upper edge of the glass stopper. This may be done by filling another burette with water to the 50 c.c. mark. Fill also the burette to be used to the same mark. See that the jet below the tap is empty. Now run the water from this burette into the other burette, emptying the last drops by blowing through. Read the increase in volume, which is equivalent to the capacity of the burette tested from the 50 c.c. mark to the stopper. The second burette may also be checked in this way. The object of this is to enable the student to read the volume of liquid in the burette from the bottom upwards, and not from the 50 c.c. mark.

The Analysis.—Assuming that the slag contains about .2% of copper, 1 gram of the slag will contain .002 gm. copper. Weigh out about 5 gms. of the slag, and treat as instructed in the Chapter on Slags (Estimation of Copper in Slags, Part III.). The copper is thus separated as Cu2S. Wash thoroughly. Pierce the apex of the paper with a glass rod, and wash the precipitate through into a beaker. To the contents of the beaker add 5 c.cs. 16E. HNO3. Heat till the Cu2S is all dissolved. Then evaporate nearly to dryness to drive off excess of HNO3. Now add 1 c.c. 16E. HNO3 and 4 c.cs. 20E. NH4HO.

Transfer this solution to the Nessler tube or burette, rinsing out the beaker with a little distilled water. Now compare the tint of the solution in this tube with that of the solution previously prepared from 10 c.cs. of the standard. Dilute the assay solution with water, a little at a time, shaking with the thumb on the end of the tube after each addition. Continue the dilution till the tint of the solution appears the same in the two tubes. Read the volume of the solution in each tube. (If a burette is used, suppose it reads 40 c.c., then the contents of the burette = (50- 40) + x = 10 + x, where x = the contents of the burette between the 50 mark and the stopper.)

Calculation of the Results.—Assume that the ‘standard’ tube contains 22 c.cs. of standard solution, and that the ‘assay’ tube contains 25.5 c.cs. of solution when the tint is the same in each.
Now the 22 c.cs. of standard solution contain .01 gm. copper, or 1 c.c. = .01/22 gm. copper.
Then each c.c. of the assay solution will have the same value. Therefore the assay solution contains .01/22 x 25.5 = .0116 gm. copper, or .0116/5 x 100 = .232% copper.

(b) THE COLORIMETRIC ESTIMATION OF CARBON IN STEEL.

Note.—If the student cannot obtain a sample of standard steel (that is, of known carbon contents) of exactly the same manufacture and in the same physical condition as the sample to be analysed, this estimation may be omitted until the student comes to the section dealing with Technical Analysis (Part III.).

Apparatus, Reagents.—For comparing the tint of the assay and standard solutions two ‘ carbon tubes ’ are required. These are plain tubes closed at one end, and graduated from the closed end upwards in c.cs., generally from 0 c.cs. to 30 c.cs. For solution of the steel, two test tubes 150 mm. x 16 mm. are required, and for a water bath a beaker covered with a tin lid, through which two round holes 17 mm. diameter are made, may be prepared. Pure nitric acid free from Cl or HCl is used for dissolving the steel. The most suitable acid is one of S.G. 1.2.

Method.—The following method was first introduced by Eggertz in 1862 :—
When steel is dissolved in nitric acid, the carbon which sometimes at first separates out is at length dissolved and colours the solution brown, and by comparing the depth of colour with that obtained from a standard steel the carbon contents may be ascertained.

To secure accuracy, it is essential that the standard steel should be of the same kind and in the same physical condition as the sample to be examined. For instance, if examining a Bessemer steel, the standard steel should be Bessemer; and if the example examined has been hammered, tempered, or otherwise treated, the standard should have been subjected to the same treatment.

This method, then, is of much service when a large number of steels of one kind have to be tested. For the standard a sample of the kind under examination is taken, and the percentage of combined carbon is determined by the combustion method (see Technical Analysis, Part III.). Once the combined carbon has been accurately estimated in this way in a particular steel, hundreds of analyses of similar steels may be quickly performed without the tedious combustion, using the standard for comparison.

The student must remember, however, that a fresh standard must be used when dealing with a fresh kind of steel. It would lead to inaccurate results if, for instance, a standard Bessemer steel were used when examining an open-hearth steel. For further information the student is referred to The Chemical Analysis of Iron, by Blair.

The Standard Steel—As before mentioned, unless the student can obtain a sample of the same kind of steel as that which he is going to examine and knows the carbon contents of the standard, this estimation should be postponed until he is able to make the necessary combustion analysis. Then he will be able to ascertain the combined carbon contents of a steel; and taking this as a standard, he can proceed to examine other steels of the same kind by the colorimetric process.

The Analysis.—Weigh out carefully .2 gm. of the standard steel (drillings) and .2 gms. of the sample to be analysed. Transfer each sample to a test tube. Fill the beaker nearly half full of cold water. Cover with the tin lid, and insert the two test tubes through the holes. The steel is now to be dissolved in pure nitric acid of S.G. 1.2. If the steel is supposed to contain less than .3% carbon, use 3 c.cs. HNO3; between .3% and .5%, 4 c.cs.; between .5% and .8%, 5 c.cs.; between .8% and 1.0%, 6 c.cs. ; over 1%, 7 c.cs., and over 1.75%, 8 c.cs.

The right amount of HNO3 is now dropped slowly into each test tube, and then in the top of the tube is placed a small glass funnel. Place the beaker and tubes on an iron plate and heat to boiling. Boil till all the carbonaceous matter is dissolved. The time of boiling varies according to the carbon contents of the steels, and runs from twenty minutes with .2% to forty five minutes with over 1% carbon. The absence of small bubbles and the disappearance of any flocculent matter indicate complete solution. The tubes during the solution should be shaken now and then, to prevent the formation of a film of iron salts or oxide on the tube. Also the student must see that the solutions are not exposed to direct sunlight, which weakens the colour.

Now pour the standard solution into one of the carbon tubes, washing out the test tube with a little cold distilled water. Dilute to 10 c.cs. Pour the assay solution into the other carbon tube, wash out the test tube with a little cold distilled water. Compare the colours, holding the two tubes together in front of a sheet of white paper held against the light. Dilute the assay solution, shaking after each dilution, till the tint exactly matches that of the standard solution. Read the volume of the assay solution.

Calculation of Results.—Assume that the standard steel contains .32% combined carbon, and that the solution of the .2 gm. sample of the standard steel was made up to 10 c.cs., and that when the tints of the solutions exactly matched the volume of the assay solution was 16.6 c.cs., then the percentage of combined carbon in the steel is .32 x 16.6/10 = .5312%.

The student should continue his practice on the same sample of steel till the results do not differ by more than .02%.

(c) THE VOLUMETRIC ESTIMATION OF COPPER BY A STANDARD SOLUTION OF POTASSIUM CYANIDE.

Apparatus, Reagents.—The usual apparatus. For the preparation of the standard solution pure KCN (gold cyanide, 98%) is best, and for checking the standard, pure electrolytic copper foil is used. For precipitating the copper in the assay solution the sheet aluminium used in the iodide assay will again answer the purpose. For analysis the student may take the same sample as used previously for the iodide method.

Method, Reactions,—When ammonia is added to a solution of cupric nitrate a deep blue solution is formed, probably containing the compound Cu(NO3)2,4NH3. On adding to this solution a solution of KCN the dark blue colour disappears. The exact nature of this reaction seems to be a matter of some doubt; probably a double cyanide of ammonia and copper or potassium and copper is formed together with certain organic compounds formed by the complex reactions of the ammonia and cyanogen liberated (Liebig). Whatever the chemical reaction may be, their result is clearly evident; and by measuring the volume of KCN required, the percentage of copper present can, if the value of the KCN solution is known, be easily calculated.

Regarding the details of the analysis, the student will find on consulting Beringer and Sutton that some care is necessary when dealing with ores. The method here given is a modification due to A. H. Low (of Denver, Colorado), and as in it the copper is separated practically free from the ordinary impurities of copper ores, the effect of these impurities need not be considered ; but to ensure success the following points should be noted:—

1. For the same amount of copper, a concentrated solution requires more KCN for decoloration than a dilute one; therefore always work with approximately the same volume of solution.

2. A hot solution requires for the same amount of copper less KCN than a cold one, therefore always titrate at about the same temperature, say 16° C. to 20° C.

3. For the same amount of copper the volume of KCN will vary according to the time taken for the titration; therefore cultivate a uniform rate of titration for all cases.

4. As far as possible, weigh out such quantities of ore that approximately the same quantity of copper is always present in the assay solution. Also see that the final volume after titration is approximately the same in all cases.

5. See that the same quantities of acid and ammonia are present in the assay solution as were used in the solution of metallic copper for standardising the KCN solution. In general, see that all the conditions of standardisation are the same as those in the actual analysis.

The Standard Solution.—In practice the solution generally employed is of such a strength that 1 c.c. = .005 gm. copper. To prepare this solution weigh out 25 gms. ‘gold cyanide.’ Dissolve in distilled water and make up to one litre.

Checking the Standard. —Weigh out, in duplicate, about .2 gm. pure, clean, copper foil. Transfer each to a 300 c.c. flask. Add 5 c.cs. pure 16E. HNO3, When the copper is all dissolved add 80 c.cs. distilled water and then 10 c.cs. 20E. NH4HO. Cool under the tap and then titrate with the KCN solution. The student should roughly calculate how many c.cs. he will require. There is less danger then of overstepping the mark. At first run in 1 c.c. at a time and shake, and when about three-quarters the approximate quantity has thus been run in reduce the volume of each addition till finally, as the colour fades to a pale lavender, the solution is being added drop by drop; add water to bring the volume to 150 c.cs.; continue till the lavender colour is no longer seen on holding the flask against a sheet of white paper. The student should note the exact tint to which he is working, as he must titrate the assay solution to exactly the same degree.

Calculate the value of 1 c.c. of the KCN solution. The duplicates should agree within .00005 gm. copper.

The Analysis.—Weigh out such a quantity of the sampled and finely ground ore as will contain about .2 gm. copper. Pure copper pyrites contains about 30% copper, therefore .6 gm. will contain .2 gm. copper. With other ores the student must depend on his mineralogical knowledge, and until he gains experience he must frequently adjust the weight of ore largely on guesswork. The results of one analysis will, however, guide him when running the duplicate. Treat the ore exactly as was done in preparing the solution for the iodide assay, proceeding with the treatment until the copper is obtained precipitated in the metallic form by aluminium. After washing as before, dissolve in 5 c.cs. 16E. HNO3, add 80 c.cs. distilled water and 20 c.cs. 20E. NH4HO. Cool and titrate as before, making the bulk up to 150 c.cs. when the colour becomes faint.

Note the number of c.cs. used and repeat the estimation. Calculate the percentage of copper as usual. The results should agree within 0.1%.

Note.—The aluminium precipitation is in the older methods omitted, consequently on adding NH4HO when iron is present a heavy precipitate of Fe2(HO)6 is formed. If this be filtered off it is impossible to wash it free from copper, hence it is advisable to titrate with it present, but only on the condition that the cyanide solution has been standardised with a similar amount of iron present. For details consult Sutton. Used with this precaution, this modification gives with some practice results accurate within .2%, but where a higher degree of accuracy is required, the method just laid down (Low’s) should be followed. The results so obtained are sufficiently accurate for Ordinary work, but where absolute accuracy is required the iodide or Electrolytic methods are preferred.