294 



NATURE 



\Feb. lo, 1876 



this process the cost of melting I kilog. of platinum was reduced 

 to 20 or 30 centimes. According to a notice of Moigno,^ the 

 firm of Jose de.Susini and Co., in Paris, in the year 1867 

 prepared oxygen in this way at the low price of 0*85 fr. per 

 cb.m,, retransforming sulphurous into sulphuric acid. Instead 

 of the acid itself, Deville and Debray also proposed employing 

 sulphate of zinc : 100 kg. of anhydrous salt yielded them 6*8 

 cb.m. of oxygen (far more therefore than the best peroxide of 

 manganese), 22 kUogs. of sulphurous acid gas, and 5 1 kilogs. of 

 oxide of zinc. 



Wagner's statement ^ is worthy of remark, that in the year 

 1867 neither of these methods was in use in Deville's own 

 laboratory ; perhaps because the sulphurous acid evolved com- 

 plicated the working. 



We must not pass over Archereau's attempt ^ to employ 

 sulphuric acid in its cheapest combination as gypsum. He 

 asserted that heating pulverised gypsum with sand would pro- 

 duce silicate of calcium, setting sulphurous acid free, which he 

 partly condensed (as did also Susini) under a pressure of three 

 atmospheres, and partly removed by means of a thin paste of 

 lime. A manufactory established on these principles in Paris 

 did not work long.* Obviously the very high temperature 

 required is an obstacle. The production of this gas from one of 

 the oldest of all oxidising agents, saltpetre, was not employed on 

 acco'Ant of two drawbacks. In the first place, a quantity of 

 nitrogen is mixed with it ; and secondly, the temperature 

 necessary for its decomposition greatly increases the cost. This 

 last inconvenience was remedied by Webster's" adding oxide of 

 zinc to saltpetre : 20 lbs. of nitrate of soda and 4 lbs. of crude 

 oxide of zinc furnished 94*676 cubic feet of a mixture of 59 p. c, 

 of oxygen and 41 p. c. of nitrogen, while chiefly oxide of zinc 

 and caustic soda remained. The price of the oxygen contained 

 in this mixture, so useful for many purposes, without taking 

 into consideration the value of the solid residue, amounts ^ to 

 2'32 ; and allowing for the value of the remains, the price is 

 reduced to 078 fr. 



In no one of these methods appears one of the leading ideas 

 of modern industry, viz., the regeneration of the residues. 



The following plans were happier in this respect, and thereby, 

 in part, more successful. 



To combine chemically the oxygen of the air with a carrier of 

 oxygen that would easily give off the gas, and would be always 

 ready to take up and again to give off fresh quantities of oxygen, 

 just as mercury does when we transform it into its oxide and 

 retransform the oxide into the metallic state — that is the 

 problem of which the last few years have given an economical 

 solution. As early as 1829 Dingier jun.^ discovered that oxide 

 of copper as well as peroxide of cobalt and nickel, treated 

 with an excess of chloride of lime, generate oxygen gas, thereby 

 transforming it into chloride of calcium. In :he year 1845 

 Mitscherlich ^ observed that many other metallic oxides, viz. 

 peroxide of manganese, hydrate of ferric oxide, cupric oxide, 

 &c,, when added to a solution of chloride of lime, caused a 

 development of pure oxygen. In 1865, Th. Fleitman^ renewed 

 these observations with reference to freshly prepared sesquioxide 

 of cobalt, the smallest trace of which was sufficient completely 

 to reduce a concentrated solution of chloride of lime into 

 chloride of calcium and oxygen. For practical purposes he recom- 

 mended to heat to 70° or 80° a highly concentrated solution of 

 chloride of lime (which to avoid frothing over should be previously 

 cleared by filtration) with o"i to 0*5 per cent, of sesquioxide of 

 cobalt. By applying chloride of lime containing 35 per cent, of 

 pure hypochlorite, he obtained from twenty to thirty volumes 

 of oxygen in a regular stream ; and other observers, notably 

 F. Varrentrapp,^" confirmed these results, and commended their 

 industrial application. The sesquioxide of cobalt need not be 

 added ready formed ; any cobalt salt answers the same purpose, 

 and the sesquioxide employed with or formed by it soon settles 

 at the bottom and can be used again and again. 



For that very reason a cheaper oxide — cupric oxide, for 

 instance, as Bottger ^^ proposed — wovdd be of trifling advantage, 

 because its use demands -^^ a much higher temperature for decom- 



I Moigno, Mondes 1S67 ; p. 494. = Wagner, Jahresber., 1867, 216. 



3 Archereau, Dingl. pol. J. clxxvili. 57. * Wagner, Jahresber., 1867 ; 215, 



5 Pepper, " Chemical News," 1862 ; 218. 



6 Dupre, Compt. Rend. Iv. 736. 7 Dingl. pol. J. xxvi. 231. 



8 Mitscherlich, Pogg. Ann. Iviii. 471. 



9 Fleitman, Ann. Chem. Pharm. cxxxiv. 64. 

 '0 Varrentrapp, " Mittheilungen f. d, Gewerli 



Braunschweig, 1865, 1866; 72. 

 '^ Bottger, J. pr. Chem., xcv., 37S. 

 « Remsch, N. Jahrb, Pharm., xxiv., 94; Zeitschf. Chem., 1866, 31. 



. Gewerlie verein der Herzogthumas," 



position. The tedious work of preparing a clear solution of 

 chloride of lime may be avoided by adding^ small mor- 

 sels of paraffin, a thin layer of oil on the surface preventing 

 an overflow. There was still one evil to be grappled with. 

 Chloride of lime requires considerable quantities of water for its 

 solution, and consequently large vessels were necessary for the 

 production of even a moderate quantity of oxygen. For that 

 reason a Winkler * rejected chloride of lime, preferring to pass 

 chlorine through a thin paste of lime mixed with a little nitrate 

 of cobalt. By this modification a greater quantity of oxygen can 

 be produced in the same vessel, and there is no danger of the 

 liquid frothing over. The part which the metallic oxide plays in 

 these methods is easy to understand. It acts as a carrier of 

 oxygen, passing by turns into a higher and very unstable oxide, 

 and being reproduced in its original state. The hypochlorous 

 acid transforms sesquioxide of cobalt into cobaltic acid, which 

 instantly separates again into oxygen and sesquioxide of cobalt. 



CoaOa + 3 [CaCl(OCl)] = 3CaCl, + 2C0O3 

 2C0O3 = C02O3 + O3. 



Thus, part of our problem is solved. By the production of 

 oxygen the carrier of oxygen is reproduced. The oxygen, 

 however, thus obtained is not taken from the atmosphere, but 

 from lime. The solution of chloride of calcium resulting from 

 its preparation must be removed and replaced by fresh lime 

 water. The process is therefore an interrupted one, and in this 

 respect capable of economical improvements. These also have 

 been accomplished by methods which carry us back from the 

 wet to the dry way. 



Since 185 1 ^ Boussingault proposed baryta as a carrier of 

 oxygen, which, heated to redness in a porcelain tube and treated 

 with moist air previously freed from carbonic acid, became trans- 

 formed into peroxide of barium. A current of steam passing 

 over it reproduces hydrate of barium and liberates the oxygen ; 

 while an admixture of lime or magnesia prevents the fusing 

 together of the mass, and thus 75 gr. of baryta yield 4 to 5 lit. 

 of oxygen at every operation. 



In 1868 Gondolo* improved on this method by employing 

 iron tubes protected by an outside covering of asbestos and by 

 an inside layer of magnesia, and placed in suitable furnaces, the 

 temperature of which could be easily regulated. He further 

 added to the baryta a little manganate of potassium as well as 

 lime and magnesia. In this manner 122 alternate oxidations 

 and reductions were carried on in the same tube. Whether 

 it be the high temperature or other obstacles which 

 have prevented this method from being generally adopted, 

 it has certainly made no way as yet into practice, although 

 it has paved the way to final success. ^ Looking for carriers 

 of oxygen of a more useful sort than baryta, the chlorides of 

 copper were the first to strike the attention of chemists. The 

 facility with which they pass into oxychlorides of various com- 

 positions when exposed to the air, is the base of the manufacture 

 of a well-known painter's colour, Brunswick green. In 1855 

 Vogel proposed the action of muriatic acid on cupric oxychlo- 

 rides as a means of obtaining chlorine .° Mallet^ studied these 

 substances more closely, and founded on them a process of obtain- 

 ing both chlorine and oxygen. He discovered that cupric chloride, 

 treated with a current of steam, changes at 100° to 200° into several 

 oxychlorides, which, by means of muriatic acid, are not only at 

 once retransformed into the chloride and free chlorine gas, but 

 give off all oxygen at a temperature of only 400°, i kilog. of cupric 

 chloride yielding from 28 to 30 lit. of oxygen. Experiments on a 

 large scale produced from 3 to 3i cb.m. of oxygen, or from 6 to 

 7 cb.m. of chlorine, from loo kilogs. of chloride of copper in one 

 operation. As four or five operations can be performed in one 

 day, from 200 kilogs. of cupric oxychloride 15 to 18 cb.m. of oxy- 

 gen are producible daily. 



The apparatus employed consists of revolving retorts of cast- 

 iron lined with clay, and containing the cupric chloride mixed 

 with \ of sand or kaolin to render the mass less fusible. This 



J Stolba, J. pr. Chem., xcvii., 309. 



^ A. Winkler, J. pr. Chem , xcviii., 340. 



3 Boussingault, Compt. Rend, xxxii. 261 et 821 ; J. pr. Chem. lii. 480 U. 

 liii. 313 ; Dingl. pol. J. cxx. 120 u. 416 ; Ann. Chim. Phys. [3] xxxv. 5. 



"* Gondolo, Compt. Rend., Ixvi., 488. 



5 Robbin (Pogg. Ann., cxxii., 256) employed peroxide of barium in 



another form for laboratory use. He recommends a mixture of chromate of 



potassium (i mol.) and peroxide of barium (3 mol.) with dilute sulphuric 



acid, to obtain a very regular and abundant supply of oxygen : — 



aCrOa + 3H2O2 = Cr^Os + sHjO -h 60. 



fi Vogel. Wagn. Jahresber., 1861, 177. 



^ Mallet, Compt. Rend, Ixiv. 286, u. Ixvi. 349. 



