By THEODORE W. RICHARDS and CHARLES R. HOOVER. (Concluded from p. 7).

The Solution and Quantitative Neutralisation of the Sodium Carbonate.

HAVING been carefully weighed in the weighing bottle containing dry air, into which it was transferred in the bottling apparatus, the sodium carbonate was now dissolved and neutralised in pure dilute hydrobromic acid. This solution was provided in carefully weighed quantity, almost exactly necessary to neutralise about 5 grms. of sodium carbonate, the amount usually employed in a determination. Because the method of dealing with this standard acid solution was somewhat unusual, a brief description is fitting.

Specially designed weighing burettes containing the desired quantity of acid solution (about 200 cc.) were em ployed to weigh it. A large amount of the solution had been prepared shortly beforehand and preserved in a bottle of resistant glass, from which it could be removed by a paraffin coated and lined glass syphon. Air was admitted to the bottle after passing through a U tube containing a solution of the same nature and concentration as that in the bottle. When the burette was to be filled, after thorough agitation of the solution (which was at room temperature), a portion was drawn off suficient to wash the interior of the syphon, the last of this being added to the U-tube, replacing the solution which had previously been in the tube. Small portions of the acid solution were now run into the dry weighing burette, which was closed and gently shaken until it was thoroughly rinsed, and filled with the vapour of the solution. The burette was then filled, with the syphon-tip inserted far into its interior, to avoid evaporation. Subsequently the burette was carefully wiped with a slightly damp, perfectly clean linen cloth, and thereafter touched only with platinum forceps or linen cloth. After standing with the upper cock nearly closed in the balance-case for one hour, the weight was taken by substitution, using a precisely similar, partly filled companion burette as substituting tare. Before filling a burette the bottle containing the acid solution had been placed for some hours near the balance, so that long waiting was unnecessary.

The required amount of solution was next removed from the burette and placed in the bowl of a very large platinum retort, the graduations upon the burette allowing one to determine to within o'r grm. when the required weight had been removed. The burette was then returned to the balance case, and in not less than an hour its weight was determined as before. It was found that three such complete manipulations could be carried out in succession, each consisting of weighing after one hour in the balance case, removing burette to the titration room, replacing and removing the tip from the burette outlet, and re-weighing, without any change in weight exceeding a mgrm. rule of course in actual analysis only one such operation was carried out.

As a

As will be detailed, the concentration of this solution was exactly determined by reference to silver, the solution being weighed out for this purpose in successive portions, just as for the neutralisation, but in order to not interrupt the narrative the discussion of the neutralisation will be continued.

As already stated, the exactly weighed acid solution was placed for the neutralisation of the sodium carbonate into the body of a large platinum retort, very kindly given by Prof. J. M. Crafts, and the weighed boat with the fused carbonate was carefully lowered into the solution. The

Contribution from the Wolcott Gibbs Memorial Laboratory of Harvard College. From the Journal of the American Chemical Society, xxxvi.i, No. 1.

condenser tube attached, and the whole was set aside for some hours, usually over night. The condenser tube was packed with scraps of platinum foil to serve as baffles, and the lower end placed into a gold flask, which served to collect any solution projected in minute drops during the evolution of gas. When the fused salt was completely dissolved, the retort, with condenser attached, was very gradually warmed during several hours, until at last boiling commenced. The supply of heat was then decreased, and the condenser tube and gold flask were removed and rinsed into the retort, which was now covered, placed in an electric oven or bath, and warmed, while very pure air was passed in through a hard glass tube bent to deliver the air just above the surface of the liquid. During the evaporation thus conducted the temperature of the oven was maintained at a point slightly above 100°, actual ebullition being prevented by the cooling effect of rapid vaporisation. When the volume of the solution had been reduced to about one-half the heating was discontinued, and the retort was cooled in a current of pure air. Methyl red was added to the cool liquid as indicator, showing in general an alkaline reaction. This outcome had been arranged beforehand in weighing the acid, in order to prevent possible loss of bromine during the evaporation.

The next step was the completion of the titration, using a very dilute solution (approximately o'01 N) of hydrobromic acid, which had been prepared by diluting the more concentrated acid and standardised by titration against dilute carbonate-free sodium hydroxide and by precipitating with silver nitrate. This dilute standard solution was weighed in a smaller set of weighing burettes, and added drop by drop to the solution in the retort. For this purpose the platinum cover of the retort had been removed and a glass plate having two small openings put in its place; through one of these openings the tube admitting a rapid stream of pure air was passed, while the other served to admit the tip of the weighing burette. When a distinct acid reaction was attained the burette was removed, and if the amount added was comparatively large the burette was weighed and emptied. If, however, the amount added was small, the burette was placed in the balance case and removed for a second addition after the carbonic acid formed by the previous addition had been expelled. As with the stronger solution, we found that it was possible, using care not to touch the burette with the hands and allowing but little change in the temperature of the room, to work with the burette for several hours without producing a significant change in the weight.

The expelling of the carbon dioxide produced by the addition of dilute acid was usually complete within onehalf hour at a temperature just below boiling, with a rapid current of pure air passing over the solution. In no case was there a change in the depth of indicator colour after the solution had been heated for one hour. Consequently this was the usual time of heating the retort, after the addition of a portion of dilute acid. Experience taught that a certain depth of pink was required in order that the solution would be left neutral when the excess of carbonic acid had been expelled, and this helped materially to shorten the tedious process of reaching the end-point by this method. In short, this method of adding successive small portions of very dilute hydrobromic acid was continued until an end-point slightly on the acid side of exact neutrality was reached, which remained unchanged when cold after at least an hour's heating, frequently with slight boiling during the last stage. It was thought for a time in the preliminary work that the end-point might be determined in a hot solution, but the action of the indicator is less reliable at the higher temperature under the conditions concerned, and the difficulty of manipulating the hot solution, especially because of the obstruction of clear vision by the drops condensed on the glass plate, caused us to cool the liquid in order to attain the accuracy demanded.

In two titrations an equivalent amount of p-nitrophenol was added to the methyl red. This mixture is very well

adapted for the titration of all strong acids and bases, working from either acid or alkaline solution. With a concentration of ionised hydrogen of 10-8 the mixture shows a distinct yellow colour, at 10-7 the colour is a very faint greenish yellow, almost colourless, while at 10-6 the pink of methyl red appears. In hot solution it is less reliable than methyl red alone, tending to make the solution appear too alkaline, but it resumes its normal reaction when cooled to room temperature. It cannot be depended upon in solutions containing more than 15 or 20 per cent of dissolved substance, especially in sulphate solutions. The weight of the indicator substance required is very small-from o'05 to 0.68 mgrm. of methyl red sufficing for 100 cc. solution, and the amount transformed, when a distinct pink is noticed, is not more than one-half this, as determined by Noyes's method of superimposing solutions. Hence, considering the high molecular weight of the indicator, it is certain that no weighable amount of acid or base was required to combine with the indicator. Confirming this conclusion it was found that one drop of o'or N solution of acid containing 0.03 mgrm. of bromine would change the reaction from distinct alkaline to distinct acid. For yellow or red indicators the eye is much more sensitive to a small change in colour when the solution is placed in the bottom of a platinum vessel than when it is contained in glass or porcelain. The methyl red, of course, gives the final evidence as to the true neutral point, but for rapid work the mixture of indicators can be used to advantage, since it gives more warning of approach to the end-point from either acid or basic side. After the sodium carbonate had been neutralised with hydrobromic acid, the sodium bromide solution resulting was quantitatively transferred to a glass-stoppered precipitating flask and diluted, and the calculated quantity of silver, having been dissolved with great precaution in nitric acid, was added. The exact end-point of the precipitation was determined in the usual nephelometric manner.

The above described method of reaching the end-point by gradual addition of very dilute hydrobromic acid was employed in five of the six determinations. In the third experiment (No. II in Table II.) a very slight excess of hydrobromic acid was added during the first operation. This excess was titrated very carefully with dilute carbonate-free sodium hydroxide solution, which had been standardised against both dilute and concentrated hydrobromic acid solutions. The fact that in this experiment essentially the same quantity of hydrobromic acid was used as in the others suggests that the danger of loss of bromine during evaporation from an acidifying liquid is very slight.

Besides thus finding the weight of silver needed to precipitate the ionised bromine balancing the ionised sodium from the sodium carbonate, we made separate determinations of the amount of silver needed to precipitate given weights of the same solution of hydrobromic acid, and also in four cases weighed the silver bromide which resulted. By giving the exact weights of silver and of bromine corresponding to a given weight of the hydrobromic acid solution, these determinations made it possible to compute, in a way entirely independent of the other determinations, the amount of silver and silver bromide corresponding to given weights of the solution used for neutralising various portions of sodium carbonate, it being assumed, of course, that the neutralisation was correctly performed. The agreement of these results with the others showed thatat least as far as the determination of the bromine was concerned-the earlier results were trustworthy. They showed, moreover, that no significant amount of hydrobromic acid was lost on the average during neutralisation and evaporation, for the silver analyses were made in one case after evaporation and in the other case without any such treatment. Several of these analyses were made in the midst of the earlier series; indeed the intention was to make them alternately.

The question having arisen as to the possible action of the dilute hydrobromic acid on the resistant bottle in

17

which it was kept, experiments were instituted on purpose to test this point, for such action would have had a highly deleterious effect upon the result, introducing sodium bromide from a foreign source. The final distillation of the acid had occurred on March 20, 1913, and not long afterwards the solution was made up in its final form. All the analyses were concluded before May 10, an interval of less than seven weeks. Six months after that, the acid having remained in the same bottle, two portions of 60 grms. each were separately evaporated in a quartz dish, and the barely visible trace of residue was dissolved in water, transferred to a small platinum crucible, dried and weighed. In each case the residue weighed only O'I mgrm., and the solution in water was found in the nephelometer to contain an amount of soluble bromide equivalent to o'04 mgrm. sodium bromide. The amount of sodium bromide formed from the flask during the first six weeks in question must, therefore, have amounted to less than o'or mgrm., a negligible quantity. It should be stated that this flask had previously been used for dilute hydrobromic acid for several years; probably a new bottle would not have yielded so satisfactory a result. Incidentally, in order to reduce the results to the vacuum standard, the specific gravity of sodium carbonate was determined by weighing it in a pycnometer under toluene, which had been found in two closely agreeing experiments to have a density at 20° of 0.8661 referred to water at 4°. Thus, 9:7256 grms. and 5.62691 grms. of salt were found to displace 3'32581 grms. and 1.92439 grms. of toluene respectively at the same temperature. The two values for the density of the salt were, therefore, 2:5327 and 2'5325 respectively. This value is distinctly higher than the results of other experimenters. For example, Schroeder found values ranging between 2:43 and | 2:51 ("Landolt-Börnstein," 1912, 180). The difference is doubtless due to the more compact condition of our fused material and the absence of air held within the pores of the substance.

Table I. reports the weights of silver ar.d silver bromide obtained from given weights of the standard hydrobromic acid solution. All the analyses are recorded, excepting two which met with accidents and could not be brought to completion.

Thus, 100'000 grms. of silver bromide are found (from the two averages) to correspond to 57'4446 grms. of silver. Baxter (Proc. Am. Acad., 1906, xlii., 210) found 57'4453 grms. very nearly the same figure. If we assume Baxter's value (because determined in a simpler way) to be correct, our average-4'77442 grms. of silver-corresponds to 8.31125 grms. of silver bromide, or very nearly the result (8 31135) found directly, as given above. The average between these averages, or 8:31130 grms. of silver bromide, must represent very nearly the true amount of precipitated halide to be obtained from 100 000 grms. of the hydrobromic acid solution. This weight of silver bromide corresponds to 3.5815 grms. of hydrobromic acid gas-a figure which gives the percentage composition of the acid, enabling us to know the density of the solution and apply accurately the large correction to the vacuum standard in weighing.

sulphuric acid and influence the composition and physical condition of the finished product.

Much phosphate rock contains organic matter which consumes a certain amount of sulphuric acid, but owing to the various forms in which this material occurs it is almost impossible to determine except by actual trial the quantity of acid required to decompose it.

Silica is acted upon only indirectly by sulphuric acid. Calcium fluoride, which is present in nearly all phosphate rock, is acted upon by sulphuric acid, resulting in the formation of gaseous hydrofluoric acid. This gas in turn acts upon silica and silicates, producing silicon tetrafluoride. The silicon tetrafluoride is decomposed by water, forming hydrofluosilicic acid and silica. The presence of fluorides is objectionable because of the obnoxious fumes evolved in treating with acid; otherwise this impurity is not objectionable.

Compounds of iron and aluminium are the most dreaded of the impurities occurring in phosphate rock. These elements when present in small quantities are very apt to cause a certain amount of reversion to take place, and when present in large quantities may render the product sticky and unfit for use. By careful handling, however, phosphates high in iron and aluminium compounds may be made to produce high-grade acid phosphate.

Carbonate of lime, which is present in nearly all phosphate rock, is a rather desirable impurity when the quantity is not excessive. The decomposition of this compound by sulphuric acid is attended with considerable heat which promotes chemical reaction between the more slowly acting substances in the mass; moreover, the calcium sulphate produced therefrom acts as a drier for the acid phosphate. In the manufacture of acid phosphate the rock is first ground to pass a 60-mesh sieve, and then mixed with an equal weight (approximately) of "chamber acid." The quantity, strength, and temperature of acid used have an important influence on the quality of the product.

After thorough mixing in a cast-iron pan the material is discharged into a "den" just below the mixer or into a car which takes it to a shed and dumps it on a pile. When the "den" system is used the reactions take place rapidly, and the product can be dug out in twenty-four to thirtysix hours, practically ready for shipment. The method of emptying the "dens" by hand, however, is attended with some risk owing to the poisonous nature of the fumes evolved from the freshly-made acid phosphate, and to the danger of large masses of the material falling on the labourers.

In the "open-dump" system the acid phosphate requires a long time to reach its maximum availability, and unless it is properly made may never be fit for use.

The storing of acid phosphate in large piles for protracted periods sometimes causes reversion owing to the pressure on the material in the lower part of the pile; this pressure also tends to compact the material. The storing of wellmade acid phosphate in medium-sized piles, however, should cause no ill effects.

Properly made acid phosphate should require no artificial drying, since the calcium sulphate formed in the process takes up the water to form gypsum. It is nearly always necessary, however, to disintegrate and screen the material before shipping. This is often done by simply throwing the product upon inclined screens, but sometimes disintegrating machines must be employed.

Acid phosphate is sold on the basis of its so-called available phosphoric acid, and has a value of 40 to 56 cents per unit. The marketed product contains from 14 to 21 per cent phosphoric acid, depending on the raw material used in its manufacture.

Double acid phosphate is made by treating phosphate rock with sufficient diluted sulphuric acid to produce free phosphoric acid, and then using the phosphoric acid thus obtained to decompose a fresh batch of rock. The final product contains from two to three times as much phosphoric acid as ordinary acid phosphate, and is very useful in the making of concentrated fertilisers.

19

THE CYANIDE PROCESS.* By FRANK S. WASHBURN.

THE fixation of atmospheric nitrogen has had commercial application for the past ten years. During his period different types of processes have been conceived; four, the Serpek, Haber, Arc, and Cyanamid, have seriously attempted commercial exploitation; two, the "Arc" and "Cyanamid" processes, have proved to be commercially practicable in the broad sense. Of these two the latter may justly claim to be of great economic importance to the world because the unit of fixed nitrogen in its first product is capable of being more cheaply produced than by any other present-known process. and because this first product has a direct and valuable use of itself and constitutes the raw material for a varied and important line of derivatives into which, for the most part, it can be cheaply transformed.

The annual productive capacity of the existing plants employing the "Arc" and "Cyanamid processes is between 90,000 and 100,000 net tons of fixed nitrogen, which, at the normal average value of the product at the factory door, represents 25,000,000 dols. This is divided in the ratio of two-thirds to "Cyanamid" and one-third to the "Arc" process. The American Cyanamid Company's factory at Niagara Falls, Canada, has alone a capacity in fixed nitrogen approximately half as great as the total world's installed capacity by the "Arc" processes.

More than three thousand technical articles dealing with almost every conceivable phase of nitrogen fixation have appeared in different languages. The patents applying to the subject in all its branches number into the thousands. So far as the author of this paper knows, nothing has appeared in the nature of a differentiation of processes based on the economics of the world demand for nitrogen. It is only for a limited developmental period that the theoretical, technical advantages and limitations of related processes constitute a practical basis for hazarding opinions as to their ultimate commercial value. It is also true that until an art has had some years of commercial application it is impossible to analyse and draw sound conclusions from the economics of the situation, the final arbiter in determining the social value of any process.

The art of atmospheric nitrogen fixation is now old enough to enable one to see the economic aspects of the case, and therefore, in casting about to determine what part of this vast subject to deal with in this paper, the author thought it would be of interest and possibly of value to treat herein the broad commercial outlook for the various fixation processes, and in order to limit the matter to conditione more or less well known to all of us, conditions on the North American continent, for the most part, will be the only ones considered.

Two nitrogenous compounds, nitrate of soda and sulphate of ammonia, constitute in value 80 to 85 per cent of the total raw nitrogen compounds produced throughout the world. The value of their combined production in 1913 was between 190,000,000 dols. and 200,000,000 dols. It is estimated that 80 per cent went to agriculture, corresponding to an annual per capita tax on the 600,000,000 inhabitants of the earth of 33 cents.

Nitrogen has four main uses: as an agricultural fertiliser, as nitric acid and derivatives therefrom, as ammonia, and in dyes and various other chemical compounds. Measured in the quantity of contained nitrogen, and in view of the restricted opportunities in the United Stotes for immediately developing the chemical and dye industries on a large scale, we need consider as bearing decisively on atmospheric nitrogen fixation processes only the three firstmentioned uses, namely, for crop fertiliser, for nitric acid and its derivatives, and for ammonia as such.

The farmers' purchases in nitrogen used east of the Mississippi River last year approximated 75,000,000 dols.

From the Chemical Engineer, xxi., No. 5.

More than 90 per cent of fertiliser nitrogen is in the form of ammonia. The normal annual consumption of nitric acid in the same region is about 7,000,000 dols. The price paid per unit of nitrogen is about the same for each, and the percentage annual increment of consumption is about 10 per cent in each case. However, the importance of the fertiliser field in comparison with the nitric acid field. in their bearing on the fixation of atmospheric nitrogen, is not fully nor even approximately measured by this ratio of ten to one.

The use of nitric acid is along established lines looking comfortably to the future for a steady growth, with no serious limitations hampering modest development and no great unsatisfied human need demanding great development. It is essentially a small business as great industries go. The fertiliser industry has a very different significance. The population of the United States from 1900 to 1910 grew from 76,000,000 to 92,000,000, an increase of 21 per cent. Crop production increased 10 per cent in the same period. Exports of wheat and flour decreased from 31 per cent of their production to 13 per cent. The importation of foodstuffs and live animals practically doubled, while imported manufactured foodstuffs more than doubled. Beef cattle production fell off 32 per cent as the population increased 21 per cent. These figures are strikingly reflected in the increased cost of foodstuffs, which from 1896 to 1912 amounted to 80 per cent in the United States, while the general advancement in the cost of living during this period was 59 per cent in the United States, 40 per cent in England, and 40 to 45 per cent in continental Europe.

These facts point to the necessity of increasing crop yields without additional labour-and that defines fertiliser. European yields per acre cultivated are 50 to 100 per cent greater than American yields. The use of fertiliser in Europe, per acre cultivated, is enormously greater than in America. The average increase in yield in Germany for the past twenty years has been approximately 60 per cent, in the United States 20 per cent. Abroad this increase is attributed to intensive farming, better selection of seed, the rotation of crops, and, to the extent of 50 to 70 per cent of the increase, to the use of commercial fertilisers.

to include the low as well as the high-priced materials, there results of necessity a product of which 85 to 90 per cent is useless dead weight, only the small remainder having fertiliser value. Forty per cent of the annual fertiliser product in value is in the nitrogen constituent drawn from materials that are either by-products or wastes, such as sulphate of ammonia, slaughter-house refuse, fish scrap, vegetable refuse, waste of sugar works and distilleries, hair, leather, wool, hoofs, and horns. The phosphoric acid constituent is secured from acid phosphate analysing from 14 to 16 per cent P2O5, and as a general rule constituting one-half of the complete fertiliser mixture. Few of these raw materials in the broad sense are satisfactory. They are borne with because up to the present time there are no others commercially available. They pay the uttermost limit in the way of transportation charges to the point of their preparation, and after being mixed they bear a further great transportation burden to the consumer's railway station. There is a relatively large expense for mixing, storing, drying, bagging, and selling. After the farmer has received his goods at the railway station, he must haul over spring roads, often almost impassable, 1000 pounds of material in order to get 150 pounds of plant food. It is one of the recognised, recurring, contributing causes for small crops, particularly in the South, that the roads are so bad in the spring, due to rains, that farmers cannot haul the necessary amount of fertiliser.

Notwithstanding the consumer pays for nondescript materials, requiring only to be mixed in the crudest manner, a price 60 per cent greater than their spot value, the fertiliser industry as a whole, which prepares and sells the product, is conducted without measurable profit.

Unless revolutionary developments in some vital respect come to the aid of this great industry in the United States, any marked improvement is impracticable. It would appear, for reasons already set forth, that possibly the matter of greatest economic importance in the United States to-day is that conditions shall be brought about by which the farmer will use a vastly greater quantity of fertiliser than is now used. We believe the most important single contribution to this end would be a high-grade chemical salt containing nitrogen and phospossessing in a high degree the many requisites of a good fertiliser, such as being non-hygroscopic, non-toxic, finely granular in form, non-leachable in the soil, and readily convertible by nature's forces into the organism of the plant. The process for the fixation of atmospheric nitrogen that will contribute the nitrogen content to such a fertiliser compound to the greatest net advantage may be justly viewed as the most important, not only to society, but also therefore to those who exploit it.

The fertiliser industry to-day suffers many handicaps. These originate in part from the farmer's limited capital for properly conducting his business, his lack of knowledge which closes to him what would otherwise be avenues of advancement and prosperity, and in part also from the fact that there is no vital, broadcast propaganda and educational bureau conducted by the fertiliser industry itself.

The greatest handicap is possibly in the material used. The industry is carried on by collecting at a great number of places throughout the agricultural regions east of the Mississippi a great variety of materials, each containing some one of the so-called plant foods, namely, nitrogen, phosphorus, and potash, the last representing only about 10 per cent of the total. For the most part these materials are low in fertilising value, and when mixed in a manner

Discussions of the ultimate commercial superiority of a favourite process and the consequent complete annihilation of every other have been almost interminable. These have proceeded apart from any broad conception of time, place, the demands of mankind, or the quantitative application of the product. They generally take the form of rival claims for cheapness in the cost of production, and this without much reference as to whether the unit cost applied to a form of nitrogen for which there was any particular or large demand, and without reference to whether the product was to be produced at a place and in a form enabling it to be transported to the consumer.

The Serpek process was discussed by volunteer adherents without knowledge that per unit of nitrogen fixed the necessary expenditure of heat energy from electricity alone is as great as the total electric energy consumption in cyanamid. Therefore, no suspicion was aroused that possibly nature had not placed the necessary raw material, bauxite, within practically shipping distance of equally necessary cheap water powers. The limitations of the by-product, upon which the commercial possibilities of the process depend, were not questioned either as to the