NOW READY. Cloth, 316; Paper covers, 216.
(Postage, 4d. extra).

THE WHEAT PROBLEM:

Based on Remarks made in the Presidential Address to the British Association at Bristol in 1898.

REVISED WITH AN ANSWER TO VARIOUS CRITICS

By SIR WILLIAM CROOKES, F.R.S.
SECOND EDITION.

VITH PREFACE AND ADDITIONAL CHAPTER, BRINGING THE
STATISTICAL INFORMATION UP TO DATE.

With Two Chapters on the Future Wheat Supply of the
United States, by MR. C. WOOD DAVIS, of Peotone,
Kansas, and the HON. JOHN HYDE, Chief Statistician
to the Department of Agriculture, Washington.

OPINIONS of the PRESS.

"In his bulky volume Sir William reproduces the gist of the sensational Bristol Address, and supplements it with carefully prepared answers to his chief critics and confirmatory chapters on the future wheat supply of the United States."-Morning Post.

"The fuller examination of the problem as here conducted shows that Sir William Crookes did not speak unadvisedly with his lips."-Yorkshire Post.

"The problem is one of importance, and Sir William Crookes presents it to us fortified by the opinions of two American experts."-Manchester Guardian.

"Sir William Crookes's statistics seem to make good his alarmist statement."-British Weekly.

"In the present volume Sir William Crookes replies vigorously to his critics."-Liverpool Daily Post.

The book is a useful one to all interested in the production of wheat both from the commercial and scientific points of view."-Knowledge.

"It is a vital question, and considering the cheap issue of the volume all interested in the feeding of the millions ought to get it and read it carefully.”—Crieff Journal.

CHEMICAL NEWS OFFICE.

16 NEWCASTLE ST., FARRINGDON ST., E.C.

TANTALUM, NIOBIUM, AND TITANIUM :

OBSERVATIONS ON.*

THESE BODies.

1. Fusion with alkaline hydroxides: Ekeberg (Rose and others).

2. Fusion with potassium carbonate and borax (Wol. laston).

3. Treating with hydrofluoric acid or fusing with potas. sium bifluoride (W. Gibbs, Lawrence Smith, and Marignac).

4. Treating finely powdered minerals mixed with carbon with chlorine at a red heat (Ste.-Claire Deville). 5. Heating minerals mixed with carbon in an electric furnace (Moissan).

6. Fusion with potassium bisulphate (Berzelius, Rose). Lawrence Smith and Rose also use sodium bisulphate and ammonium bisulphate.

A short review of these processes and their respective advantages and defects is appended :

METHOD NO. I.-With regard to this process, Rose states that it would probably be preferable to any other if it did not require the use of silver crucibles, which are rather strongly attacked by caustic alkalis in a fused state. Another disadvantage is that only a very moderate heat can be used with silver vessels, and consequently there are produced acid salts insoluble in water which have to be fused again with fresh alkali to render them soluble. I have made many experiments on these lines with columbites, tantalites, and similar minerals, using, instead of the silver vessels, platinum crucibles lined on the interior with pure gold. With these crucibles higher temperatures can be obtained, but still the results are unsatisfactory. The fused alkali climbs the sides of the crucible very persistently and volatilises in white clouds, much insoluble acid salts are formed, and if manganese is present the gold crucible is attacked and gold is found in the solutions subsequently obtained. Wrought iron crucibles were also tried, but the "climbing" of the fused materials was even worse than with the gold crucible, and extended sometimes

The first part of this Paper appeared in CHEMICAL NEWS, vol. scv., p. 1 et seq.

almost to the bottom of the outside of the crucibles, though these were only half-full at the start. The following quantitative experiment was made on 5 grms. of very finely which was heated with successive charges of pure potassium hydrate containing 72·3 per cent of K2O. After each fusion the mass was thoroughly extracted with water, the dilute sulphuric acid mixed with sulphurous acid, the whole acids were precipitated from the alkaline solution with boiled, and the acids filtered, washed, and then strongly ignited, with addition of ammonium carbonate, to a constant weight. The matter insoluble in water-oxides of iron, acid salts, &c.—was also heated with dilute sulphuric and sulphurous acid, filtered, washed and ignited, weighed, and then fused again with fresh caustic potash. (See Table I.). This table shows that, starting with 5 grms. tantalite and 10 grms. of potassium hydrate, a fusion lasting three hours and ending with a red heat, only rendered 1895 grms. of the acids soluble in water out of a total of 4134 grms., or less than one-half, and that after a second fusion of the residue with 6 grms. more alkali there still remained insoluble o 84 grm. which contained tantalic and niobic equal to nearly 14 per cent of the total amount present, and that at least four successive fusions would be necessary to entirely decompose the mineral. The disadvantages of this mode of procedure are obvious, a very large amount of alkali is employed, and much time and manidefects arise from the fact that it is impossible to raise the pulation is necessary. I shall show further on that these temperature of the caustic potash sufficiently high to cause the easy formation of tantalates and niobates soluble in water. In fact, the alkali distils away and "climbs" the crucibles before the requisite heat can be obtained. After many trials I abandoned this method with some reluctance, since, as Rose points out, the metallic acids thrown down from soluble tantalates and niobates by precipitation with acids are always much more pure than those obtained by fusion with potassium bisulphate and exhaustion with hot dilute sulphuric acid. It is curious, however, to note that these successive fusions appear to cause a fractionation of the tantalic and niobic acids. In this tantalite the tantalic acid is in relatively very large amount to the niobic acid. Judging from the colour of the ignited acids, it would seem that the first fraction is almost pure tantalic acid, the second fraction being of a slate-blue colour is indecisive, but the third fraction would seem to contain most of the niobic acid, as the white acid becomes of a bright yellow colour when ignited; on cooling it becomes white again. This method does not free the metallic acids from tin, tungsten, &c.

METHOD NO. 2.-Fusing Minerals with a Mixture of Borax and Potassium Carbonate (Wollaston).—I have not

tested this process, since the introduction of sodium salts into materials containing tantalic acid gives rise to very slightly soluble salts, and is consequently objectionable. The effect of the boric acid is uncertain, but from what is known of its behaviour with tungstic acid and similar bodies it may be supposed that it would probably form complex compounds, boro-tantalates, boro-niobates, &c. Therefore its employment in the present state of our knowledge on these points does not seem advisable.

as

METHOD NO. 3.-Treating with Hydrofluoric Acid or Fusing with Potassium bifluoride (L. Smith, W. Gibbs, and Marignac). The action of hydrofluoric acid on the various tantalic, niobic, and titanic acid containing minerals appears to resemble its action on various silicates. It is well known that while some silicates are readily decomposed by this agent, others, such as zircon, andalusite, staurolite, &c., are but little affected by it. Lawrence Smith has shown ("Original Researches," p. 353) that samarskite, euxenite, and hatchettolite are readily and conveniently opened up by this reagent, and he points out that the rare earths are separated in a dense and granular form as insoluble fluorides, while the metallic acids go entirely into solution. But when he states that "there is no columbate which does not readily yield to this treatment," he is stating more than the facts warrant. I have tried this method on various columbites and tantalites, and have found that even when the pulverisation of the minerals was effected as he directs by levigation with alcohol, &c., the opening up of even 1 grm. of the mineral is a very tedious and lengthy operation. Some of the complex titano-niobates are but little attacked by hydrofluoric acid. Fusion with bifluoride of potassium has some advantages; its action is more powerful than that of aqueous hydrofluoric acid. When used in conjunction with potassium bisulphate it will decompose many minerals which are hardly attacked at all by hydrofluoric acid, such cassiterite, corundum, rutile, &c. (Clark, CHEMICAL NEWS, xvi., 232). The objections to these processes are:-That all the first operations have to be entirely carried on in platinum; the solutions, if cold and acid or neutral, can often be dealt with in ebonite vessels, and on a larger scale I have found the use of the paper-pulp utensils found in commerce coated with a hard resistant varnish very convenient. However, neither ebonite or the other utensils are practicable with alkaline solutions, especially when these are hot. Again, while the utility of hydrofluoric acid and fluorides as powerful agents (which are in many cases indispensable) cannot be denied, they are not substances which recommend themselves for use in the laboratory if they can be replaced by other bodies. Apart from the facts that their employment renders the ordinary laboratory utensils unavailable, and the deterioration of all glass vessels in their neighbourhood, their toxic effects are worthy of consideration. A long experience on a large scale with these bodies convinces me that the greatest care | is necessary in their use. The incautious inhalation of hydrofluoric acid, spilling it on the hands and nails, and the internal administration of only a few milligrammes of the soluble bifluorides is attended with serious consequences. When minerals have been attacked by fluorine compounds, bases such as iron, manganese, and tin pass into solution with the metallic acids, and are subsequently removed with difficulty. This is especially the case with

Acids have a slate-blue colour, hot or cold. Acids bright yellow when hot, white on cooling. Yellowish white.

82.68 per cent.

tin. Hydrofluoric acid and fluorides resemble oxalic acid in this respect, that they prevent the precipitation of tin by hydric sulphide. Hall and Smith's researches (Proc. American Phil. Soc., xliv., 181, 187, 190) show how tin kept turning up in the tantalic and niobic acids prepared from the double potassium fluoride salts, though their solutions had been treated repeatedly with hydric sulphide. METHOD NO. 4.-This process for opening up tantalites, niobites, &c., was recommended by Ste.-Claire Deville. I have had no experience of it as applied to minerals. The extreme volume occupied by the sublimed chlorides of tantalum and niobium and their contamination with ferric chloride are mentioned by Joly. H. Rose employed combustion tubes 5 feet long and nearly an inch in diameter to obtain 12 to 15 grms. of the sublimed chlorides. It appears to be more adapted for the working up of the acids obtained in the usual way than for minerals. It offers one great advantage that it separates the tin, titanium, germanium, and silicon in the form of very volatile compounds (zirconium will, however, be found with the tantalum and niobium chlorides), and up to the present time this chlorine treatment appears to be the only really efficient process for separating titanium from tantalum and niobium.

METHOD NO. 5.-Heating the Minerals mixed with Carbon in an Electric Furnace.-Moissan reduces the minerals to fine powder, mixes them with charcoal from sugar intimately with the aid of a little sugar syrup, calcines the mixture in a Perrot-furnace, and then agglomerates the mass by pressure. These masses are then heated in an electric furnace with a current of 1000 ampères at 50 volts pressure for seven to eight minutes. In this way Moissan states that the whole of the manganese and the greater part of the iron and silicon are volatilised. The product is a brittle metallic crystalline mass, which contains all the tantalum and niobium combined with carbon. The mass reduced to powder is attacked by a mixture of hydrofluoric and nitric acid, the iron is separated by ammonium sulphide, and the mixed tantalic, niobic, and possibly titanic acids are then separated by Marignac's method. Or alternatively he treats the fused mass broken up into coarse powder with dry chlorine gas at a red heat, as in Deville's process. Since few laboratories have at com

mand electrical installations capable of giving a current of 1000 amperes at 50 volts pressure, it is obvious that this mode of procedure can have but a limited application. I have found that tantalite and columbite in small pieces can be fused by means of an oxyhydrogen blowpipe, using a block of gas retort carbon as a support, into brittle fused buttons. These buttons, however, were not readily attacked by the mixture of hydrofluoric and nitric acid, probably because they contain a considerable admixture of fused or semi-fused tantalic and niobic acids.

METHOD NO. 6.-Fusion with Bisulphates of Potassium (Sodium or Ammonium).-Of all the methods proposed for opening up these minerals, the employment of bisulphate (or more properly pyrosulphate) of potassium, first suggested by Berzelius, seems to be the one almost universally adopted, especially for tantalites and columbites. Still, no one who has had much experience with it will deny that in many respects it leaves much to be desired. The fumes evolved upon heating any amount of the salt are extremely irritating, and the operations must therefore be

CHEMICAL NEWS,}

Jan. 1, 1909

conducted under a hood with a good draught. The minerals, especially in the case of tantalites, must be reduced to an extremely fine powder if anything like a fair opening up is to be obtained. Lawrence Smith, for instance, speaks of 1 grm. of columbite requiring ten to fifteen minutes constant trituration in a four-inch agate mortar. My own experience is that even when the division of the mineral is carried on to an extent inconvenient in ordinary laboratories a complete resolution of the mineral is rarely, if ever, attained at one operation. However long the fusion is continued, the acids obtained by extracting the melt with hot dilute sulphuric acid always contain basic matters, and especially iron and zirconia; even a second fusion in many cases by no means removes all these impurities. One must not be misled by the white appearance of the acids; they may be nearly white, and yet contain as much as 4 per cent of iron and other impurities. Adding a few drops of ammonium sulphide to the moist acids should not blacken them (lead or iron). The acids not ignited or strongly heated ought to be perfectly soluble in a hot solution containing their own weight of oxalic acid. Any insoluble matter indicates lime, lead, thoria, ceria, and yttria group of metals, &c. The quantity of saline matters introduced by the large amount of potassium bisulphate generally recommended is inconvenient when the basic matters have to be estimated, and as the bisulphate invariably loses a good deal of its sulphuric acid during the heating, there is a tendency to form in some cases extremely insoluble double sulphates, such as zirconium-potassium sulphate, which is, according to Fresenius, almost absolutely insoluble in water, and in hydrochloric acid if precipitated hot. Again, it has been shown, especially by Prior (Mineralogical Mag., xv., 83), that when much titanic acid is present along with tantalic and niobic acids, "on fusion with acid sulphate of potassium and treatment of the melt with a 5 per cent solution of sulphuric acid, practically the whole passed into solution." I can quite confirm this statement from my own experience; it has long been known that this method, recommended in most text-books as a means of separating titanic acid from tantalic and niobic acids, was imperfect, but Mr. Prior has demonstrated that in the presence of large amounts of titanic acid it is absolutely untrustworthy and misleading, and consequently the published analyses of minerals containing these three acids require in some cases considerable corrections. These minerals contain very frequently as constituents other metals, such as tin, tungsten, lead, antimony, and germanium, and when such minerals are treated with alkali bisulphates, the tin and tungsten certainly, and the antimony and germanium probably, become so firmly united to the tantalic-niobic acids that their separation becomes a matter of extreme difficulty. Berzelius recommended that the washed tantalic-niobic acids should be digested with ammonium sulphide to remove the tin, tungsten, &c. Rose ("Chimie Analytique Quantitative,” 1862, p. 456) found that this plan did not answer, and he recommended fusing the acids with six times their own weight (!) of a mixture of equal parts of dry sodium carbonate and sulphur. Why such an enormous excess of alkaline sulphide is used is not apparent. Rose states, however, that this fusion converts a good deal of the acids into insoluble sodium tantalates and niobates, and consequently the whole must be again fused with six times its weight of potassium bisulphate and treated with hot dilute sulphuric acid as before to free them from the combined soda. Rose says naively: "Ce traitement est un peu compliqué," but one fears that those chemists who have tried it would use stronger terms before they finished the operations. Dr. Edgar Smith says on this point:-"Our own experience leads us to say that the removal of tungsten and tin from columbium and tantalum oxides cannot be realised by digestion with ammonium sulphide. Indeed, not only did we find the fusion with the sodium carbonate and sulphur necessary, but that working on large quantities of material, as in our case, two or even three refusions with these reagents were

3

found necessary" (Proc. Am. Phil. Soc., xliv., 157). Blomstrand also (Journ. Prakt. Chem., xcix., 40) did not obtain favourable results with this method. He points out that if a low temperature is applied in the fusion the tin, tungsten, &c., are not removed, while if a higher heat is employed the tantalic and niobic acids pass into solution along with the tin and tungsten sulphides when the melt is treated with water. It is therefore apparent that the complete removal of the tin, tungsten (and probably also of antimony and any germanium) is an exceedingly tedious operation when carried out in this manner. It seems to be commonly assumed that the residue left on treating with water or dilute sulphuric, the result of a fusion of a tantalate or niobate with potassium bisulphate, is tantalic or niobic acid mixed with a certain amount of basic matters. But, as a matter of fact, this residue is really a tantalic or niobic sulphate (Moissan, "Chimie Minérale," ii., 139). Hermann (Journ. Prakt. Chem., 1872, 2), V., 66) found in the product obtained by fusing tantalic acid with potassium bisulphate, exhausting the melt with boiling water, and drying at 100 C., 85.25 per cent of tantalic acid, 523 per cent of SO3, and 9.52 per cent of water, corresponding fairly well with the formula 3(Ta2O5), SO3,9H2O. Rose points out that the acids can be washed continuously for several weeks without eliminating all the sulphuric acid. When the washed sulphides are collected on paper filters one observes that the paper becomes quite rotten, and even charred when dried at 100 C., from the liberation of sulphuric acid. It has been recommended that the acids should be washed with dilute ammonia in order to get rid of the sulphuric acid. I have tried this on the acids obtained from columbite by fusion with sodium pyrosulphate. After the acids had been well washed with dilute sulphuric acid and then with water only, they were stirred up with a slight excess of ammonia. This caused a great change in their appearance; at the beginning they were pulverulent, but the treatment with ammonia converted them into a flocculent state. Yet in this condition they filtered bright and fairly well. But the clear filtrate contained a surprising amount of the acids, which came down when the ammonia was just over-neutralised with hydrochloric acid. The precipitate was filtered off and washed. On examination it gave a very strong reaction for niobic acid with zinc dust, but only a very faint indication of titanic acid with peroxide of hydrogen; it was not examined for tantalic acid. It is clear that the acids (or rather the sulphates) are distinctly soluble in ammoniacal ammonium sulphate, and as this mode of procedure has been recommended in the quantitative analysis of the acids, a word of caution as to the error it involves is not out of place. It is not improbable that when the moist crude acids are treated with sulphide of ammonium in order to extract the tin, tungsten, iron, &c., that a certain amount of niobic and, perhaps, tantalic acid will be found to go into solution along with the tin and tungsten sulphides, as ammonium sulphate must be formed in this case also. I intend when time permits to follow up this question of the solubility of the acids in ammoniacal ammonium sulphate: possibly only niobic acid may thus be dissolved and not titanic or tantalic acid. The use of sodium pyrosulphate instead of the potassium salt seems advisable on the ground that it is far more soluble, and when rare earths are present it has not so much tendency to form insoluble double sulphates. It appears to me, however, to lose sulphuric acid at a lower temperature than the potassium salt. I have tried on some occasions ammonium bisulphate as recommended by Rose. I found that finely ground columbite can be opened up with this reagent, heating in Jena glass beakers, but tantalite seems but little attacked. I have never, however, been able to completely dissolve columbite in the fused salt so as to get a clear glassy mass which would dissolve in water temporarily to a clear fluid in the manner spoken of by Rose. There was always a residue which the most prolonged heating failed to dissolve. Vessels of fused silica are very

useful for these fusions with bisulphates, and are to be preferred to platinum, which metal is not unaffected by long heating with acid sulphates. Solutions of titanic hydrate in nitric and sulphuric acid seriously attack both platinum and platinum-iridium basins when they are evaporated in them. This is alluded to by Rose (Chimie Analytique Qualitative, p. 1028), but appears to have been overlooked since his time, as Fresenius recommends platinum dishes for the estimation of titanic acid in acid solutions ("Quantitative Analysis," p. 196). I have found that when pure titanic acid is fused with potassium or sodium carbonate, the melt dissolved in nitric acid, and the whole evaporated to dryness on the water-bath in platinum, on taking up with water and dilute nitric acid the insoluble titanic acid is quite grey in colour and the dish has lost in weight considerably. According to Rose, tantalic and niobic acids do not act in this way on platinum. I first noticed this action of titanic acid solutions on platinum in attempting to separate a mixture of a quantity of rare earth oxides from a considerable amount of titanic acid with which they were mixed, by dissolving in nitric acid and evaporating to dryness on the water-bath in a platinum basin, as one would proceed for the estimation of silica.

Requiring some quantities of pure tantalic, niobic, and titanic acids for various experiments, I was led to think whether it would not be possible to devise some better way of extracting these bodies from their native compounds, and especially from ordinary columbites and tantalites, so that as far as possible the use of fluorine compounds and the employment of platinum vessels should be avoided, and also that the very tedious means hitherto made use of for the separation of tin, antimony, iron, and other impurities could be improved. As my efforts in these directions have not proved unsuccessful, I offer the results of them to our readers. Desiring to obtain some materials uniform in composition so that experiments could be repeated under like conditions, I obtained some time ago parcels of minerals in some quantity. One of these was columbite; this I obtained from a dealer who had had it in stock for many years. It was marked "Columbite from Middleton, Connecticut, U.S.A." The weight of this was about 10 kilogrms. It was externally very free from foreign matters, and many of the lumps were highly crystalline and beautifully iridescent. Another material was massive tantalite, partly from Greenbushes, West Australia, and partly from the Northern Territories, Australia. These were also fairly free from foreign admixtures; the exterior of some of the pieces, however, was covered with a thin drab coloured adherent coating which could not be wholly removed by washing, and as it contained tantalic acid it appeared to consist of what is known as tantalic ochre. The weight of these tantalites was about 30 kilogrms.

The magnesium powder replaces the other metals owing to the heat of combustion of magnesium being greater than that of barium, calcium, strontium, and aluminium (see Trans. Far. Soc., iv., Part II., p. 130).

In order to commence this reaction, a high initial temperature is necessary, and this is supplied by the burning magnesium.

The free metal, as soon as it is produced, being in contact with air, immediately combines with oxygen and nitrogen.

With BaO, CaO, and SrO the product was quite yellow, and on moistening with water smelt very strongly of ammonia; wih Al2O, the product was black, but also smelt strongly of ammonia on adding water.

In the experiments with the oxides of the alkaline earths, carbonate, hydrate, and peroxide must be absent; if these are present the action is violent, and the formation of nitride is much decreased.

Calcium oxide, if the carbonate is present, gives no nitride, carbide, nor cyanide; barium and strontium oxides containing appreciable quantities of carbonates yield nitride, cyanide, carbide, and, possibly, cyanamide.

Excess of magnesium powder, as a rule, produces a larger yield of the nitride, possibly by union with the oxygen of the air leaving the nitrogen for the other metal. Estimations of the amount of nitrogen gave the following:Maximum yield.

I. BaO+Mg 2. BaO+2Mg 3. SIO+Mg 4. SrO+2Mg 5. CaO+Mg 6. CaO+2Mg 7. Al2O3+3Mg

Per cent of Ba3N2 = 24°4 78.4

= 43°33

64°5

Sr3N2

=

33.69

70.6

In the last column under the heading maximum yield is given the percentage of nitride that would be produced if the whole of the metal from the oxide became converted into the nitride, e.g., 3CaO+6Mg+30+2N=Ca3N2+6MgO.

The free metal after the reaction was always very small, never amounting to more than 3 per cent. That the nitride does not consist of magnesium nitride is evident from the following considerations:

When magnesium burns in air, either as the powder or the ribbon, less than 1 per cent of nitride is obtained, and when an oxide is reduced, the metal of which does not form a nitride, the yield of nitride is again less than 1 per cent. Magnesium nitride readily burns, and if a great heat evolution takes place the nitride would burn; now, when Mg reduces CaO the heat evolution is not excessiveMgO+Ca+3000 cals.

Mg+CaO

=

145,000 148,000

with such a reaction, as a large evolution of heat occurs :CaO+Mg MgO+Cu+111,000 cals.

37,000

148,000

In the first case we might expect that any magnesium converted into the nitride would remain as such, but from the study of two other reactions in which a similar small heat evolution occurs, no nitride is produced. The two reactions in question are:

(For one molecule of MgO 36,000 cals. being liberated). 2. SiO2+2Mg 2MgO + Si +40, 110. 215,690 296,000

=

(For every molecule of MgO 20,000 cals. being liberated).

Again, using excess of magnesium, although the yield of nitride is increased, the amount produced is never above that which could be produced from the metal obtained from the oxide. The above experiments are being repeated in vacuo, and in an atmosphere of nitrogen.

The Chemical Laboratory,
The Polytechnic, Regent Street, W.