US1091425A - Process of fixing nitrogen. - Google Patents

Description

J. E; BCHER. PROCESS of PIXING NITROGEN. APPLXOATION PIE UD 0OT.21, 191% 1 091,425 I Patented Ma.r.24.1914.

woe mbo@ b @WQ/wao Z. C

' UNITED ern-'rus PATENT OFFICE. N

JOHN E. BUGHER, or OOVENTRY, RHODE sneiimfsseuon 'ro NITROGEN PRODUCTS ooMrANYgA oonronerron or RHODE isLANn.

PROCESS or FIXIiie NITROGEN. i

A Specification of Letters Patent. Patented Mal'. 24,1914.

Applicationled October 21, 1912. Serial No. 726,924,v

This invention relates to the lxation of nitrogen and more particularly to animpioved process of preparing eyanogen coni- V pounds and deriiatives therefrom, together using the corresponding tree elements'.

with Valuable by -products incident to the production of such ,a compound or compounds.

The process-in question is essentially synthetic in character in its preferred forii, nd was suggested by my Work on the synthesis of cyanids and on iny processes 'for purifying' and preparing metals, in which nitrogen plays an important part. Applications ior Letters 'Patentfor these proeesses have been tiled in the United States latent (lliice, being there identified as zipplical'iong Serial Number 676,39?, dated 'February' S. 15112: Serial Number 690,886, dated April 15,1912; Serial Niiniher7lO,'58, dated July 22, 1912; and ,Serial- Nui'nbcr 711,211, dated July 24, 15,112'.

(lne ot' the equations i'epresenling the fundamental ideas ot catalysis and reversible reni-lions upon whieh these l'oiir applications are based is the li'ollowing':

Vl'ii the latter apliilication, particular attention is called to the fact that l was able lo convertczirhonales suchl as those o'l so diiun into eyanids, and the present applicaN tion is the one so` referred to in my seid application,Serial 'Number 711,211, in this connection. u,

One. of the objects of the present inveiin tion isvto synthesize cyanide from nitrogen and carbon by the use of compounds such as alkali carhonates, hydroxide, ctc., instead of following equation inziy he cited :is an eXf ample.; Y if 2) N a,CO-1-4C+N, :QNa UN+3CO- The empirical commercial preparation of The A( zyanogen derivatives from alkali com-- pounds dates back more than two centuries.v

From' about 1828 to 1845, much of the most -55 Y i' l A important scientific work on the preparationI oi" cyanids from atmospheric nitrogen and alkali carbonates Was done, and it was during this period that the formation of otassium cysnd in Ablast-'i'irnaees was a rma- 60 I tively noted. f

As early as 1839, Lewis Thompson claimed to have experimentally determined that a mixture of coke, cinders or coals, potassium mrbonate (potash or pearlash) and iron turiiings ground into a coarse powder and heated to full red heat in an Open crucible and in an Open fire, produces 'more cyanid than is obtained When air is notpres'- ent. His paper, in the English periodical- 7,0l The Mec/anios Magazine No. 822, page 92 primarily of interest as showing the addition of iron which e' has been used. in certain subsequent cases, He further posi-` ti rely states that the potash is decomposed 'lli` hy the iron, producing potassium, 'which being volatile, rises and combines' with the cai-hon 01": the coke and With the nitro en ot the. atmosphere, the oxygen of which as been removed by passing through the fire, or hy the eolie or cinders. He intimates that when iron is notI employed, a much higher teinjierature is required. He also states that the carbonaeeous metter employed may he worked over-again many times, sind even improved by each operation. His stateumut of the function 'of the iron, is, I believe, Widely at variance with the facts, as evidenced by the disclosures made in this and my above referred to patent applications; and. he certainly conveys no hint of the necessity for inaintain-- ing the proper surface, contact and other reaction conditions, to which I shall ,herein-l1" after rler. l 9 5 The.""Thoinpson, it he indeedy actually did- .forni any material quantity of eyaiiid by ignition of the potash with animal substance V 2.8 ounces of pure Prussian' blue Were obvcomments upon this that it is questionable peraturewas very much higher, gave with (German) relating to the progress of chemistry and mineralogy, vol. 21, pp. 80 and 8l, published in i842. After reviewing Thompsons .article in Dz'ngZers Polytechnic J om"- nel, vol. 73, p. Q83, he concludes in substance with the observation:

The question as to whether coke gives more cyanogen in this operation than cor* Responds to the nitrogen it contains, yet remains to be decided.

'Erdmann and Marchand-in the Journal fr Praktische Uitemv'c, vol. 26, p. 413- 1842, after reviewing the investigations of Fownes on the formation of cyanid from wood, charcoal or coke heated to a medium red heat with potassium carbonate, refer to Thompsons process substantially as follows; This process consists in igniting a mixture of two parts potash, two parts coke or bituminous coal (steinkohlen) and one part iron turnings in an'open vessel permitting lthe access of air. Hereby a greater yield of cyanid was said to be obtained than'hy the and iron. With one pound of pure potash,

tained.

Berzelius (Jahresbericht 21, page 80) whether the quantity of cyanogen is larger than could be produced from the quantity of nitrogen in the coke. Before this is shown by experiment, one cannot decide what importance Thompsons process has in its theoretical and practical relations. l

1 The statements of Fownes led us to have the experiments repeated. A. mixture of equal parts sugar charcoal and potassium carbonate was exposed toa protracted red heat in a nitrogen current which was preM pared from atmospheric air which assed Cyanid orma- A second experiment in which the teminitially abundant evolution of carbon monoxid which contained noticeable quantities of potassium vapor, only a doubtful trace of cyanid formation. f

Y In a third experiment, in which impure potash containing potassium sulfate was used, the presence of potassium sulfocyanid was shown quite plainly. Therefore, it seems as though cyanid formation as Fownes described it, may well take place but not Without having certain conditions which have not yet been exactly determined, fulfilled.

Bertelsmann in his ycmogen U01/)Wounds (Og/anverbz'ndungen), Berlin, 1906, page 62, states inpsubstance:

As early as 1839, Lewis Thompson showed that potassium cyanid forms when air is led over a red glowing mixture of colrc, potassium carbonate and iron turnings, however, it was not yet shown that the nitrogen came from the atmosphere and not in some way from the coke.

This .vriteraitcr referring to the investigations of Thompson, Fmvnes and Young, Erdmann and Marchand, Langlois, Rieken, Delbrck, and Bonson and Playfair, makes the following statements:

Industry very soon began to make use of these investigations and in the forties of the preceding century the first patents for the utilization of atmospheric nitrogen were taken out. Simultaneously, small works were erected to operate the new processes but these had to be closed after a few years. Similar industrial enterprises arose at a latery time but all hadthe saire negative result.

In 1845, Bunsen and Playfair in a very careful and comprehensive investigation of the action of al blast furnace in England, showed that potassium cyanid is formed in large quantities in the hottest part of the furnace. Furthermore, they claimed to have found that by heating charcoal prepared from sugar with pure potassium carbonate in nitrogen, in a gun barrel, potassium cyanid is obtained. In tliese experiments, iron was inadvertently prsent, in one case from the reduction of ores in the blast-furnace and in the other from the probable comn position vof the gun barrel used in the experiment. The theories which Bunsen and Playfair advanced to account for the phenomena were, however, inadequate, and the art does not appear to have been greatly ad-I 'vanced by their efforts. -For example, they clearly did not recognize the rle that'the iron played in the experimental part of their Work; but apparently considered potassium carbonate, nitrogen and carbon, theonly substances which were immediately n concerned 1n the chemical reactlon. 1ndeed,'1n

1845, Bunsen. proposed to effectuate their discoveries commercially by constructing a special blastdurnace for producing potassium cyanid .from potassium carbonate, charcoal, and air; but in -the accounts of this attempt, to which I have had access, I have found no reference to the use of iron in connection therewith, and such unsatisfactory results were obtained that the process had to be abandoned. The omission of iron (or its equivalent) in'itself, would` be sufficient to account for the failure of his process; since Without the use of such a material his ternperature of operation must have been prohibitively high, especially from a commercial standpoint.

The failure of Thompson and others to make the disclosures necessary for the suc cessful use of iron, was one of the factors rest synthesis of cyanids from atmospheric lnitrogen and carbon, not only in the case of carbonates such as potassium carbonate, but

even in the ease Where the free alkali metals fthemselves were used. This is true in vspite 'A olf-'the fact that `the following two thirds of a century represents a period of intense ac- ,tivity in the attempt to get a really satisfactory process for this purpose.

- Vast sums of money have been'spent in atl tempts to operate some of the processes de Scribedi'n the numerous patents which have been4 issued. In these patents almost innumerable disclosures and diferentiations have been made. In many cases these are contradictory and in some cases manifestly impossible. In some eases iron. was used, and in other cases not, with apparently nothing in favor of the cases Where iron was used." In fact, the eases which proved to be' more nearly successful were those in Which'the inventors did not add iron.

Theconfusions and contradictions both in theory and practice no doubt were due at least in part to the use of iron in the forni of retorts, tubes, furnaces, etc., in the processes, thus unconsciously having iron present with the reacting substances consciously einployed. In some other cases, inventors took no account of the presence of impurities 'such as ash in coke, coal, charcoal, etc

which under the 4conditions of the process might produce iron, silicon, calcium, aluini num, and other catalytic agents. In still other cases, not the slightest attention was paid to the formation of these same catalytic agents from the porcelain, fireclay or linings which were in Contact with the reacting substances. Certain of these cases which bear more directly on the disclosure set forth herein will be hereinafter considered.

Such lack of deiiniteness in the experimental conditions could lead only to conflicting theories of the process and ultimately to di'eat confusion in specifications. claims an disclosures, and all sorts of complicated steps were introduced to obviate real or Sonie of thcsc involved processes, as set forth in the printed descriptions of the saine, marked distinct retrouves sions as compared With the apparently more nearly correct and less involved work of Lewis Thompson, Bunsen and Playfan', and

'several others in the early history of the eyanid synthesis. Reference may be made in this connection to the- German patents 'to Alder, Nos. 12,851-1S,94524,3-14 and 32,334'. Alder, like Thompson. states that the addition of iron acts very favorably .n promoting eyanid synthesis. Alder, how ever, seems to have been led astray in his Work yand probably his lach of success was due to his failure to note what the previous experimenters had also over-looked, e'. e. the conditions of prime importance to success which `lshall hereinafter refer to in detail. As a result, Alder attributed the cyanid formation to the presence of carbon-containing gases such as carbon iiionoxid and hydrocarbons in the nitrogen.

l have found that quantitative formation of cyaiiid from, for example, alkaline can 'bcnates takes place readily Without the ad- -dition of hydrocarbons, carbon monoxid, 'or the like to the nitrogen. Furthermore, Alders disclosuresI concernino temperature conditions are very vague an in fact, add practically nothing to the state of the art as known at that. time.

According to the German chemist, Taiibei', in his article on Tests ofv Sulfurous Organic Substances for Nitrogen, published in the Iiejuorts of t/c German Ctcnrr'cl'o- {fiati/,1899, vol. Hl, Article 46T,Alders processes were abandoned after beingtried for six years, presumably on account of the great technical. difficulties vwhich were encountered 1 land his unsuccessful yeiit'orts have made so Lengleii`s--T7ic (fi/amd@ 111,7 cathy, Le Clercs translation, published in New York in 1906: while in Bertelsmanns Work-i77ie Tech- ;f/.wfug/.z/ of' Hic (ly/mogen C-ompozmcls, published in Munich and Berlin Ain the same year, sonic reference is made to Alders proeesses, but they are seriously questioned as to ltheir value or utility. Others made use of iron, nickel` cobalt and simi-lar metals iu their processes for attempting` to productx \'unid, but obtained more or less inconsistent and conflicting results` In fact so little success has herctoi'mo attended processes using iron, ctc.. as carriers ol carbon, that these attempts ccm not. only lo have been abamloncd but sounl have cvou gone so far as lo deny thc aclivitv 'of iron even in case oi' the complete synthesis from tree all-:ali metal, carbon and .nitrogenv Thus (Iy itner, iu his lluilcil titatcs lalcnl No. 577,937, dated lllarcl 2nd, '1897, und in his liritish Patent, New 12.218 oli ldt, states that ironA l those of Vllcrlclsmann and Robine and llenglcu, abovi` rolirrcd to, @-1 urll as the general 'scicntiliI4 literature and many of the patents themselves rclutimil iff. this arl. shows that numerousl attempts have bccn made to fiX atmospheric nitrogen and that although neither time nor cxiicnsc was spared. the resuits (with the possible exception of costly electric processes) have not been very satisfactory.

Bertelsmann states in substance in his book, above referred to, page 103. lines 1) ezt seq..:-

When one views the whole subject, it must he realized that while neither work nor expense has been spared upon the ilnoblem of utilizing atmospheric nitrogen, results have been attained which have been olf little practical value.

Robine and Lenglen in their work above noted, pp. :B2i-3:22, state:

The synthetic processes have the great advantage of producing,` potassium cyanid as a final product. But in most oi these processes this advanage is counterlnilanced by serious objections; the temperature required for the reaction is often verf.T high, from which it follows that losses volatinzatimi occur, bi-fsides a rapid wear and tear oi? the apparatus. lll'oreover. the conversion is very oit'ten incomplete.- and the reactions do not always take place :s simply as the 'theoryv would lead one to expect. Notwithstanding the munerous ei'forts of investigators and manutaetilrers, very t'ew ot' tl processes have had any reall;Y )jnactiealV applicatr; e. Not one el' the synthetic processes usingnitrogen and alkali metals has vet given satisfactory results.

Also:

Onlyv the methods which start with. an alkali (sodium), carbon, and amu'ionia, with intermediary formation of cyanid, seem have succeeded.77

lfroin this it will be seen that the use of atmospheric nitrogen and ont' the very cheap alkali carbonates for the manufacture oit eyanids had been practically abandoned by (and indeed long before) 1906. Since then, to the best olf' my knmvledge, no further efforts have been made lo use these inexpensive materials in a synthetic process of the kind in question. with the exception of the heroin described process.

The result of the abandonment of the use ot atmospheric nitrogen and carbonates, hydrates, or the like, in the practice of the art has led to the use of such expensive niaterials as vmetallic sodiumand ammonia. The most recent authoritative discussion oit the subject with which l am .acquainted is found in ft/)9771s Dictionary of .fippZ/ul (lizenz/stain Rev. Ed. vol. Il., London and New York, 1912. One. paragraph of the abstracts from the article on cvanids bv Dr.

Thomas Ewan, (Cassel Cyanide Coi., Glas` gow) found` on page 1913 of this volume under the caption: Production of Cyanide' from Atmospheric Nitrogen, reads as follows:- n

In every process now used for the pro! duction of cyanide, the nitrogen is derived from coal more or less directly. The production of cyai'nds from atmospheric nitrogen is a problem on which both mono)7 andyis evident that a satisfactory` commercialV method of producing cyanids fron atmospheric nitrogen andsa'lkah-earbonates for example, is not only desirable on account lofl the cyanide tliemselves,but even more so for`l the utilization oi atmospheric nitrogen. The.-` solution of these problems forms the subject i'natter of this application for Letter-s Patv ent, auf

attendant advantages ,ot such a'process are" numerous some ot the more importantare git/'en in the following list:V i

l. The process should use free atmospheric nitrogen; or, better still, air itself.

lt should make use of inexpensive raw materials g. sodium carbonate or bicarbonate l E. It should preferabl;7 employ the cheapest catalytic agents, c. iron.

rthe catalytic agent should he efficient, c. it should be capable of promoting the reaction with great rapidity even at relatively moderate temperatures.

The catalyst should preferably be capable of being used over and over again,- Without having to be renewed or purified.

G. The process should avoid rapid wear and tear of the apparatus.

7. lt should allow the use of a cheap heat conducting material, such as iron, for the re ceptacle used.

`8. It should preferably be capable oi' being conducted entirely b v heat generation in the apparatus itself, so that the outside of the apparatus could be embedded in non-conducting material to avoid loss by radiation and conduction and to avoid the. destructive oxidizing action of the air and furnace gases on the outside.

9. It should avoid loss of alkalis by "olatilization.

l()` It should avoid loss of cyanids by oxidation, or by the decarlourizing` action of the iron' or the like in the charge or in the container.

11. The yield of cyanid should preferably be subslantiallv quantitative,-oll of the alkali compound, nitrogen and carbon finally entering into combination.

1:2. These quantitative yields should be obtained easily. and consistently` when the specifications are followed.

the desirable characteristics and 13'. The process should be simple in character and rapid of execution so as not torequite unduly large or expensive plants nor involve excesssive expense for labor and maintenance.

14. It should be capable of yielding pure cyanids not contaminated with excess of alkaline substances, sullids, ferro-cyanids, cyanamids, etc.

15. rflieproduct from the furnace should be rich in cyanids.

16. The product should preferably be capable of being lixiviated very readily.

17. After lixiviation, when used, the moist residue of` thev carbon and iron (assuming iron to be the catalytic agent employed), 4should be capable of being used over dry or moist, with either dry or moistsodium bicarbonate as it coines from the a1nrnonia-soda process.

18..It should preferably be capable of converting certain of the impure alkali compounds, such 'as are obtained in various Ways in commercial processes, directly and Without purification` into pure cyanids.

19, It should preferably also enable usto convert sulfates, such as those of potassium and sodium, into cyani s in one furnace .operation by intermediate reduction and transformation into carbonates.

20. It should be possible' toy/recoverV the cyanid by distillation, sublimation 'or by draining it from the apparatus during the operation of the process.

21. The resulting product should preferably be in sucli condition that can be converted into ammonia directly with s team Without the necessity of lixiviation or separation of cyanid, thus making the iron and alkali carbonate mixture act both catalytically and cyclically; the carbon and the steam to be added under suitable regulation of temperature conditions-` 22. It is desirable, in certain cases, to have a simultaneous smelting process for producing the catalyst, e. g. iron, during the course of the operation of and by the process itself.

23. Similarly, in certain cases,vit Will be desirable to operate the4 nitrogen fixing process as a combined process iron and making cyanid.

24. For reasons hereinafter more fully set forth, it should be possible, when desired, to use the catalyst in solid forni, and to operate the process at onlyinoderately temperatures, suoli as lie below the melting@ Ipoints of copper and even of silver,

25. The process, in some ofv ts phases,

should be non-electrolytic, since lsuitable current is not alivaysavailable.

'26.' In thev last analysis, an 'edicient process of cyanid foranatio-n should enable us, very cheaply,- and vvitliout drastic temperature reuirenients, to transform air and water t rough the instrumentality of'carbon (coal,

"vention will be felt in a number for purifyllgigen or air as the case may be. s

coke', etc.) into pure ammonia, with such production of by-products as hydrogen, carbon monoxid, formic acid, oxalic acid, etc.

The hereinafter described process in its several aspects is adapted to fulfil or meet each and every one of the recited desiderata and hence broadly aims to economically iX atmospheric nitrogen in a form,whereby it may be utilized either with or Without subsequent operations, according to the nature of the ultimate product sought; and it is believed that the iniuences Iof the present in A of industries.

In the accompanying drawings, which form a part hereof and in which like reference characters designate like parts throu hout the several views, I have exemplifie. o ne form of apparatus in which my process may be electuated. It is to be understood, however, that this apparatus is but one of many adapted to the purpose in question, and that I am not to be limited thereto'in any Way;

nor indeed by the specific steps. or operation tions may be made therein withoutdeparting from the spirit of my invention. In general I may state, therefore, that my said invention is to be regarded as iimited only by the scope of the appended claims.

Description 0r App amtus.-Refei'riiig to the drawing: igure l is .a longitudinal vertical section of a mufiie furnace with retorts in position therein, Fig. 2 is an end elevation of said furnace; the blast lamp beingomitted for convenience ofV illustration. Fig. 3 is a fragmentary Section of one of the pipesorretorts, the section being taken on line III--III of Fig. 1'. Fig. 4 is a similar section of a retort'adapted for the use of air in the process.

Bric'kwork or other suitable material may bea-employed' in the construction of the mutlle 1, Aand onel or more iron pi es 2, serving as retorts, may extend direct y thgrethrough. Caps vor unions 3, serve to connectthe ends of the retorts to the pipes 45; t ose desig-v nated 4 being the pipes for supplr ing nitroy Y he gases, notably carbon monoxid, evolved during the A operation of the process are conveyed away viap'ipes 5.'

A blast lamp 6 or other suitable source.v of. heat may. be disposed in the preferably open yend of the muiie and thisliapparatus should.

vof course, be capable of .heating the retorts up to the reaction;temperatures` employed; which will be hereinaftervery Vfully discussed. J

Assuming thatthe process is to be conducted in such fashion as vto permit thel use of air, the latter'is led in through pipese and preferably first encounters one or' more masses 7 of charcoal or other suitable" mixture 0 which is next encountered by the gases flowing through the pipe or retort; since it is desirable to keep the charcoal away from the charge. During the course of the operation, the alkali `metal will be evolved in the manner hereinafter described, and if the charcoal is in contact with the ref active mass or mixture, it may, by capillary action, or from any other cause, remove a portion oi' such alkali metal from saidanass and thereby reduce the yield.

The mixture 9 preferably consists oi pulverulent coke, graphite, charcoal, coal, or the like, Very intimately mixed with finely divided iron. and sodium carbonate, or other inexpensive source of the metal forming the base of the cyanogen compound sought, or to be incidentally formed during the operation.

onsiderotions of importaacanletore considering the general reactions involved, I desire to state that I cannot too strongly emphasize the importance of considering what would appear, atiirst thought, to be merely mino-r considerations, but certain of which considerations are vital lto the success of the process, and the K@verlooking uit which has apparently led to the practical failure of the repeated attempts to effectuate the same. Non-observanccof these considerations has led experimenters in this art to apparently'partly succeed at one time, when by vaccident the conditions were propitious, and

' at other times to meet with total failure or with only such success as would not justify continued operation of the process on a commercial scale. y

` The present Work is based on a very large number of' experiments, but Iwill here refer particularly tothree comparative ones. In these experiments, sodium -carbonate and graphite were heated, with nitrogen, in one .case without the addition' of a, catalyze'r, in another with the addition Fof ,finely divided iron; and the two following eqluations are Avery important in considering these cases.

pared by passing air through a tube containing red hot copper, was passed through a tube containing the reaction mixture.

Eperz'ment 1.'-10 grams ot finely powdered graphite and 5 grams of sodium carbonate nerev heated in a current of nitrogen. to between 920 and 940O C. forabegtl-jgfty 'yellow color tothe llame.

- (the graphite being omitted minutes, in a co'pper tube of one half inch diameter. At this temperature, therewas a slow but steady evolution of carbonmonoxid and some Whitev fumes passed out of the tubes with this gas. These fumes imparted a steady though not very intense Experiment .9.-Sus ecting fumes were due to the'. ormation of the relatively volatile-,free sodium rather than to the volatility of the carbonate itself, Iheated 5 grams of sodium carbonate with 30 'grams of finely powdered iron in a similarv copper tube and in a current of nitrogen from 920 gradually up to 1,000" C. for ,fifty minutes. ln this test, the issuing ases did.

not impart the slightest trace of yel 0W color to the flame at any time,'nor could carbon monoxid be detected. This latter experiment withiron Vand sodium carbonate alone proves that sodium carbonate is not sensib y decomposed by finely divided iron at 1,000o C. nor is it volatile enough to give a flame test at the end of the tube.

ing, the copper tube containing sodium carbonate and graphite showed a minute quantity of white material which had condensed on the cooler part of the tube; which de.

positconsisted at least part-ly of freemetalthat these lic sodium. Thesefacts show that at about 920O to 940 C., small'quantities of sodium and carbon monoxid are formed according to equation 3. When these two components attam a certain small vapor pressure, the' reversible reaction comes to equilibrium; They. are then swept along with the nitrogen current and inturn another portion of` the carbonate decomposes, thus tending to restore the vapor pressures necessary 'for equilibrium. The contents of the last mentioned tube were treated with Water `and tested for sodium cyanid by means of the Prussian blue test.v Only traces could be detected even after the solution had stood for a day. This experiment shows that graphite, sodium and nitrogen yield practi cally no cyanid when heated to themodera-te temperature of920940 C. in the absence of a catalytic agent. And even here it is possible that the cop er in the tube or the impurities (Fe) in t e graphite may have exerted some catalytic influence.

E @apartment .Zt-A mixture of 10 grains of the graphite, l0 grams of finely powdered 4 an iron for a copper containing tube, all conditions were-kept as nearly'as possible like thoseof the first experiment. At a temper-` ature of 920"S)fi0fJ C. as measured by the make them reactive; and we also knowl that' did not escape with the, gases when iron `carbon,'an'd consequently is swept away by `quently become dissociated.A

pyroineter, there was, in this latter case, a steady absorption of nitrogen and the escaping gas burned with the characteristic blue flame of carbon nionoXid. Nitrogen was still being absorbed at the end of fifty-minutes i when the heating was discontinued. The escaping gases contained no fumes nor did they show the slightest trace of yellow color when led into the burner flame, at any time during the experiment. This very sharp contrast with the experiment in which no catalyzer was used shows that free sodium was used. The contents of the tube were then treated with water and the resulting solution was titrated with silver nitrate. It was found that'OU/b of the sodium carbonate had been converted into cyanid. Under the conditions of this experime1it,tlie velocity of the reaction was increased. at least several thousand-fold, even at this relatively low temperature, by the catalytic action of the iron with a consequent transformation of of the carbonate into cyanid in fifty minutes. A far more rapid formal tion of cyanid may be effected at, for eX- ample, higher temperatures, or by properly varying other conditions, as I shall hereinafter show. These and the further experi mental results below mentioned, enable us to formulate a theory which is in acco-rdance with the facts and is of great service in connection'with the process.

By means of Experiment l, in which iron was omitted, it was proved that reaction 3 takes place at 9200 to 940ML.- (and to a less extentv at even lower temperatures). Now when comparatively small quantities of sodium vapor and carbon mon'oxid are present at this temperature;l they tend to form sodium carbonate and carbon and thus reverse the reaction; establishing a chemical equilibrium.

-lVhe'n no catalyzer is present, the-sodium does not unite rapidly with the nitrogen and the current of nitrogen, so that We have very little formation of cyanid. When, however, finely divided iron is added to the mixture of carbon and sodium carbonate, the mixture being heated to the given temperature as per lExperiment 3, the iron will become carburized by dissolving some carbon;v thus forming a solid solution-ofcarbo-n in iron.

Theory of @mattoni-In chemical work, it is often necessaryV to dissolve solids to when substances are dissolved they very fre- These and manyl similar considerations point to they conclusion that if the very complex and ordinarily comparatively inert carbon can thus be dissolved'it will not only acquire mobility but perhaps may also become less complex; thus approaching the nascent condition. Iron is one of the best solvents known for carbon. Hencea solid solution', such as is obtained in the cementation process of manufacturing steel, will form very quickly at the surface of the metal; pro-l vided that itis not already there, as in carburized iron. 'The conditions atV this surface should be exceedingly favorable for chemif -cal reactions and the above mentioned results obtained by the use of finely divided iron are in accordance withthis view.

Presumably appart at least of the operation takes place as follows: The smalll quantity of sodium vapor, formedvaccording yto equation 3, and the nitrogen come inlcontact with the reactive surface of the carburized iron where they combine'""very quickly.- with lthe carbon dissolved/in the surface ofthe iron, forming sodium cyanid. vThefvafy por pressure of the gaseous sodium, is thus 35 reduced and consequently' another portion of sodium carbonate decomposesv to restablish theequilibrium. This new portion of sodium vapor in turn 4combines lwith ni,n trogen and carbon at the surface of the iron, l which has meanwhile become recarburizedn by taking up carbon from the 'carbonaceous masses in contact therewith, or from'the carbon in solutionin the mass of iron.' The iron hence acts as a catalyzer by u'ndergo- 95 ing a series of decarburizations and recarv burizations; resulting in the eliicient and continuous production of reactive carbon for the process. Thevery quick combina'- tion of sodium atthis catalytic surface re- 100,

duces the concentration of the sodiumyapor steadily, and thus materially'aids in, producing a continuous liberation of sodium with the consequent continuous, formation of sodium cyanid. This very simple theory, enables us to deal intelligently with the f many conditions which influence thfproc ess, Evidently the first requisite is to profduce as efficient a catalytic -or solution 'surface as racticabl'e and to maintain its elii- 11'0 ciency uring the process.: l

Even when the conditions of surface and contact appear to bef'correct or nearly so, at times a lapse of 10 or 1,5 minutes mayv 'occui before any vi orous reaction takes l115 placeD This may be ue, in 'some cases,y -especia'lly when working at a relati'velydow 4 temperature, to the time, consumed 1n car;` burizingy the' reactive surfaces ofthe catiiif-i lytic material; and probably alsoto changes in the relative disposition offluid and solid constituents of the reactive mass by, for ex- 1 ample, capillary action.' The theory of the operation must therefore be borne in mind while the a'ctualprocedure is really a com-125 promise betweenvarious factors. A t i one temperature one factor becomes dominant; at another temperaturesometimes` 'quite close to the first, another factor gains the f ascendency, and so on.

Should my theory be faulty, which however I do not think is the case, it at any rate supplies a working hypothesis Which enables us to achieve consistent and successful results.

Sources and condition of mately/2er, etc;- A ne state ofdivision of the riron exposes a large surface and hence acts favorably. Similarly a fine state of division of the carbon favors the quick carburization of the iron by providing thorough Contact at, what I may term, thev solution surfaces, or as I prefer'to term' it in the aggregatesurface The ironA and carbon should be mixed thoroughly for the same reason, except in those cases hereinafter referred to Where the carbon is supplied through the agency of a vapor.

It is of importance to consider both the relative and the absolute size of the particles of carbon and iron as well the extent to which they are in actual Contact.

The catalytic agent, e. g. iron, :if used in solid-form should'be quite finely powdered or in such other form as to expose, o-r to render possible the exposure of, a relatively enormous solution surface. Particularly is this the case when the operation is conducted at temperatures materially below the eutectic point'of the carbon-containing material.

If the iron be a coarsel powder, insufficient surface will usually be provided, under such 'conditionsunto permitof the process being conducted efficiently. If,on the other haue., it is too finely divided, there is some tendency to pack too tightly and thereby prevent dif' fusion of the gases and vapors involved. On the whole,.however, I prefer to divide the iron as finely as possible and to provide for diffusion in some suitable manner such as Will be hereinafter indicated.

When the particles of catalyticl material are relatively large, the surface -is correspondingly smffifand the area of Contact. is likely to beter; '..aiall for quick and eilicient renewal ofthe carbon content in the catalytic surface.. Under which I,have found convenient in eectuat-4 ingmy process, when nitrogen and sodium were heated with iron, better results were, in general. obtained when the iron was relatively ine and the graphitecomparatively coarse, than when the relative sizes were re- 55 versed, i. e. when' the iron was coarse and the aphiteiine. Thus iron Ph. Gv.' IV, which' 1s far inerthanr 100 mesh, gave very good' .results when used with graphite which had been passedr through a one hundred mesh sieve but which on the whole 'was not nearly 'Y so-nelydivided .as the iron.

An even better. result was obtained when said "fine iron was used with graphite of to 80 mesh'. 'lVhein however, the graphite was gf-nreslr used with this very fine iron the indicated conditions'.

the action was considerablyY slower. ThisI se ries of tests, show, therefore, a distinct maximum for the graphite of intermediate coarseness; as was to have been expected since when the particles are too fine the mass lacks porosity and the nitrogen and sodium v apor cannot so easily penetrate the interior of the 11i-ass whereby to efficiently contact with the reactive surfaces (or as I herein broadly term it, reactive surface). On the other hand, when the graphite particles are too coarse the `-fine heavy part1c1es of iron will tend to drop through the interstices and thus become segregated. The latter action would, in extreme cases, be equivalent to removin the catalyzer from the field of action. find that either of these two-extremes may substantially result in the failure to produce cyanid. We should, further, not only consider the sizes of the particles, but also their other general physical roperties. e. shape, smoothness or roug mess of surface, whether they are vcompact or spongy, regular" or irregular in shape, whether they arendhesive at the temperature of the operation, etc. For example, if the particles are coarse, spherical or globular, non-adhesive 4and smoothgthey favor segregation, while if, on the. other hand, they are fine, irregular, rough and preferably slightly sticky or adhesive the temperature of the operation, they obviously favor the contact and'surface conditions when once established.

It is to he understood that while I regard finely divided iron as generally preferable for use in the process, in view of the possibility of increasing the potency of certain other factors hereinafter referred to, I do noty wish to be limited to any particular size, condition or source of the catalyzer or in- 105 deed of the size,.condition or source of carbon for the same, except where,in certain of the appended claims such limitations are `specifically introduced. `For example, pulverulent iron may be produced directly in 11o the reacting mass itself from y iron compounds or ores, such as oxids, 'carbonntes and the like, e. g.,haematite, magnetite,'

siderite and iron scale.

A mixture of finely powdered'hematite1 I carbon and sodiumfcarbonate may be prepared in such proportions as to leave the iron, graphite and sodium carbonate in favorable proportions Ifor cyanid synthesis i after the excess of graphite had reduced the hematite to metallic iron with the evolution of carbon monoxid. Under these circumstances the iron becomes reduced very readily (even if the solid carbon particles should nut come in" thorough contact with the iron ore, reduction would still take place completely because the molten cyanid formed reduces iron oxids with the utmost ease and even earburjzes the iron at the same time). Further, if the temperature be sufficiently 1w high, sodium vapor will be formed, according to equation 3, which, of course, will act as a powerful reducing agent.

Iron scale (presumably l `e,(),) reduced very quickly and gave a very 'satisfactory yield of cyanid; leaving the resulting iron, after lixiviation of the cyanid, in excellent form for subsequent use with the process. Thus use may be made of a Waste product (iron scale), or ore Withthe simultaneous 'production of cyanid, metallic iron, substantially nitrogen-free carbon monoxid, and with the removal, in part at least, of impurities such as sulfur, alumina, silica and phosphorus from the reacting mass. i consider that. this may be of special interest in the iron and steel industry, especially as the process can be carried out in iron apparatus (which is not attacked by bas'f; materials) at relatively low temperatures with basic ie-agents which hence tends to remove' the tivo very troublesome elements, suitur and phosphorus, as Well as acid forming elements in general, including silicon. it presumably removes these elements as suliids, phosphids or phosphates, silicates, aluminates, etc., and probably some nitrid formin r elements in the form or nitride.

Ihe louT temperature, with consequent moderate reduction, to say Anothing the strong alkaline nature of the charge, tends to produce a very pure iron and at the saine time aiords a process for lining atmospheric nitrogen. Usually the iron thus obtained will be carburized but this may be decarburized very readily, asdescribed in my said pending application, Serial No. 676,399, by means of the cheap sodium produced from cyanid, as described in my said application Serial No. 710,758.

Carbon-The carbon may be supplied from a variety of suitable sources and in addition to those mentioned, which are solids, carbon supply-ingvapors may be used. Thus the hydro-carbone, c. g. petroleum, are ayailable for use in certain cases. It is highly desirable usually Vto substantially completely exclude the oxygen of the atmosphere from participation iff the reaction resulting in the product-ion oit' cyanid, since oxygen either converts the cyanid 'formed to cyanate; or oxidizes the iron and results in the production of ferrofcyanids or the like upon subsequentgliniviation; or even Worse, it may destroy the cyanid formed, with reproduction of carbonate.

The carbon has Oreater freedom of motion than' one migrlit at first .expect Its diusion is not limited to mechanical motion due to the action of gravity, agitation of the mass, melting of solid material with consequent sintering, Washing effect on the particles by currents of liquid due 'to surface tension, change of concentration, 'capillarity, distillation of liquids, etc. it is soluble` in the iron, or the like, used as a catalyzer, and can diii'use either in the torni ott solid .or liquid solution. It may, in effeet, be carried as liquid, vapor, or even as solid in the form of sublimate, by compounds such as cyanids, c. g. sodium cyanid. .it vthese latter come in contact with iron lacking in carbon, a reverse action may take place (equation libris helping to carbonize the mass of iron. mobility to the carbon such as would be attained if the carbon could diffuse as a vapor, thus giving to the carbon, such graphite, coke, charcoal, et, in the tube, essentially the same kin d of freedom that it would have it used in the form of hydrocarbons. l belindeed, this imparts a' lieve that this may be a matter of great imortance und-er some conditions ot' c anid` p synthesis as it does not require diiierences in temperature for 1ts manii'estation, e. g.

Nitrogen Mass A of Mass B of Na .y NHCN -v e (Na2COai-Fe+C) (Fe) becomes Nq carburi/ed For example, in the above indicated dispo-- sition of masses of naterial in the retort (the brackets indicating the respective spaced masses and the tube being indicated by dott-ed lines) We may form sodium cyanid in the usual manner at A. At B We may have decarburized iron. H now any cyanid be carried from A as a liquid, by distillation or by sublimation to B, the reverse l equation 4, Will take place at B; equation l being a reversible equation, the direction of the reaction depending on mass action, etc. The iron takes up the carbon vfrom the cyanid with the liberation of sodium and nitrogen. This is practically equivalent to transferring the highly non-volatile graphite, in which forni the carbon may be present in mass A, at temperatures substantially in the vicinity of the eutectic point. ot iron. and carbon, through the instrumentality of the intermediate formation of a volatile cyanid, to the mass B located in another portion of the tube or retort. course, the reaction can 'be completed at tiret; thus getting rid of A Zfali' metal-The source of alkali metal 13u or other base of the cyanogen compound formed, during the course of the process, is preferably some inexpensive compound such as sodium carbonate, sodium bicarbonate, sodium hydrate, or the like, but it is possible to employ initially free alkali metal or its equivalent. p

`With respect to sodium bicarbonate I particularly desire to call attention te' the fact that I have siiecessfully used the moist bicarbonate which comes from the ammonia soda process; and this Without previous drying. Indeed, very many impure alkali compounds Which may resultfrom various technical processes can be converted into pure cyanogen compounds. Forl example, mixtures of alkail'i" hydroxids, carbonates, acid carbonates, ac etates, oxalates, etc., can be used directly if iron, oxids, or hydroxids, or organic salts of iron, carbon or organic matter in general be present; and no, contamination of the alkali cyanids can occur since the substances of the classes'mentioned Aare 'so transformed in the process `that the only negative radical remaining Will be cyanogen. If more than one alkali metal be present, however, a mixture of alkali cyanide Will result. When the. cyanid obtained is to be transformed into ammonia as hereinafter described almost any other impurities may be present Without affecting the purity of the product, prov-idinff suitable precautions are taken. Alkali sul. ates, can also be used With much advantage in my process. Potassium sulfate is obtained extensively from the deposits at Stassfiirt, Germany, While sodium sulfate is" obtained in large quant-ity as a byi product from the action of sulfuric acid on sodium compounds, such chlorid.

(5) QNaCl-l-HZSCLzNa:SOH-ZHCL The suliid may noW- be converted into ca rbonate even Without renaming it. from the apparatus or separating' it from the carbon Witlivwhicli itis mixed. This may be done yby allowing` it to cool to a suitable temperatur and then passing a current of moist carbon dioxid through the ma. sym' the water in suitable proportions :nay even be added rdirectly to'tne mass containing;a the all-:ali sulfid and carbon before passing,r the carbon dioxid through the mass. The chemical bonate into cyanogen compounds per equation 4.

The cyanid thus produced can be utilized directly as cyanid vor it may be converted into ammonia by methods hereinafter de scribed Without previous removalfrom the carbonaceous mass.

The iron required for cyanid formation may also be added directly to the mixture of sodium sulfate and carbon, before heating` so that it possible to charge the tubes with. sulfates ani proceed to the formation of ammonia and carbonatos without removing the charge or adding solid material thereto.

I have shown that the transference of carbon from one separate mass to another'in a. tube or retort is possible, and will now consider the somewhat similar but not strictly analogous transference of sodium.

NSA-CO In this case a mass C of sodium carbonate and carbon (Without iron) is heated in u current of nitrogen to produce sodium and carbon monoXid according to equation 3. If this current of the three gases, nitrogen, carbon monoxid and sodium be then passed through. a mixture D of carbon and iron, cyanid forms, according to equation 4, by absorption and combination of nitrogen and sodium at the catalytic surface of the carbon and iron; While the carbor monoxid passes `out of the tube. I have frequently used this mixture of carbon and iron with rather ineffective reaction m'ir'tures to absorbvthe sodium vapor from the escaping gases, thus doing away completely with loss by volatilization, and the danger of explosions from free sodium When the massesvere lixiviated. It is obvious too that sodiumcarbonate, for

example, may be distilled over, from a mass of the same into contact with the material presenting the' reactive solution surface. I am of the opinion however that the bulk of the sodiumsupplied bymni'ss C to massl') passes over ifs free or mctallicsodium in va rateo'difiusion of the sodium vapor was4 aboutxafmilhon times as great as the velocity case the'reduction be, indeed probably is, enormous, as Ihave' indicated.

porous form. At 920 to 940 C. this sodium vapor would be formed quite slowly at C andbe carried to D by the nitrogen which at this temperature would usually make up the greater part of' the gas. MassvC being porous, as is also, of course, mass D, capillarity probably plays an important part in bot-h cases. An extended film of the liquid source of the alkali metal hence farms on thecar-I bon particles and this I shall, for convenience, term (Whether here or, asin other cases-0n the iron as well) a reactive film, since it decomposes under the action of a suitable agent e. g. heat to liberate the alkali metal. Presumably if the film at C is very extended the sodium'vapor quickly saturates the free space in the porous mass C, thus bringing about a condition of equilibrium. The rate of the sodium vapor would then be determined practically by the flow of the nitrogen current through the tube no matter Whether the masses in C and D were an inch apart or many feet apart. For example, if the nitrogen were passed through a one half inch iron pipeat a temperature of E320-94:0o C. or even somewhat lower, at the rate of about 100 cc; per second, it.would take a very long time to get a'substantially quantitative yield of cyanid at D; and this in spite of the fact that cyanid formation, equation 4, takes place practically. instantaneously at thi1s)very eflicient catalytic solution surface at Equation L tis apparently of the'fifth 0rder,l when we consider 4it as taking' place from right to left for the conditi "ns at about 900o C., and the rvelocity of this retarding reaction would hence vary as the fifth power gases, providing that it is of/this order. Hence if the catalytic or other action were efficient enough to reduce this vapor pressure down to one tenth of its formenvalue, the retarding effect would be decreased a `million-fold.-

I am aware that reactions of high order are rare (as stated by Nernstl on page 560 of his Theoretical C'zemz'st'ry, 4th'edition`,r New York, 1904); but even if the reaction in question were only of the decrease of the retarding effect 'would yet befah`undredfold; while in the lowest order', z'. e'. 4the first, said effect would still -be ten-fold. This demonstrates effect islarge and it nay .In the eXPerimentsabo-ve referred t0., the

of th nitrogen' current. Therefore when the components oi the reacting mixture are finely powdered, the sodium molecules have auch a .short distance to travel that the rate of the vapor pressure of the two second order the that in any ously noted. I am aware,

at which the sodium comes in contact with ,l the reactive solution surface is largely de` termined by the enormous molecular diffusion velocity rather than by that of' the rela-- tively slow sodium laden nitrogen current. Let us now consider the masses C and D to be superin'iposed. In this case the particles will be very close together, the masses will interpenetrate and cyanid forn'lation will take place far more rapidly than where the effective supply of sodium is wholly ldependent onthe rate `oi flow of the nitrogen current rather than upon the velocity of the molecules of the alkali metal. .F or cxample, in the already cited Experiment 3, such interpenetration of the masses C and D (i. c. such juxtaposition of the particlesvi'f iron and carbon, or such close juxtaiiosition of the reacting film and the reactive solution surface) .Wasprod'uced byl thoroughly mixing the charge of iron, carbon and sodium carbonate' previous to its introduction into thetube. The result was a yield of over of the theoretically possible quantity of sodium cyanid, in fifty minutes. This experiment shows that the sodium given oflI in close juxtaposition with the catalytic solution surface of carbon in iron. Temperatura-,Temperature also mustexert` an enormous influence when we consider its variousA effects. YIt increases the velocities of chemical reactions enormously, and it also increases the rate at; which the carbon goes into solution in the iron; as well as the rates of gaseous diffusion and development of vapors.v The temperature of the operation further intimately concerns thev surface and contact ferred to.

nearly to, the

bon' mixture is heated to, or

C. although it eutectic point (about 1150" probably accounts in large measure for the fail-.ure of many of the previous attempts.

e pulverulent iron may', 'if it has not repeatedly; Without operation since the treatment catalytic metal by removlng sulfur,-'oxygen,

and other impurities therefrom,

carbon solution need not conditions above. re-

the carbon been sintered into globules or worse, be used repulverization. Its condition may even be improved by the purifies the When the finely pulverized iron and carmay Vary somewhat) it becomes sintered 18%; 7%; 3%,oreven less than 1%. This' c as previhowever, that the be solid and lthat 'contemplate ell'ect-ing the process in Certain crises at. temperatures below even OO" C., since rubifliiun and cieeiiun carbonatos, for example, and espec z. ly mixtures of these: or of one or more or these with sofliuin or potassium carhoualeor auzilogously acting eomi'iouncls permit of* materially lowering thetemperature at vvliiel the nitrogen. nxiup; reaction takesl place ifprimi/:ipally,V lieve, by inf-reasing' tre liuiil eirlriililion). i .6; 1 l; 3 ally in the le action mixture) l same enil may also be yeral`-mobility oi" 'the riaet nti'aiuefl in ,fr nieaeure.

or momies, will m.ln vier.' el' the foregoing, there'lore, it is contient t iet the gein antl ite (minne e' pouenls must be emi-idc wel, i rclac-ive position el const' active uniss or masses may bei tlue to the iueltinfg olj alkali eciupouiu.: (or yeven the ifalalyzer,` to wit iron, itself) "with the consequent change in. the position olf the earth ries et' carbon zin/l iron, to the velocity o'l the nitrogen current, to the llrniiiz-'e'e oil" the molten conuiountlig unifier the iul'loence or" r:i\tity, to volatility, :flirtare tension, capillarity unil ille like. lll-rom hose elliaets` it is exf'hleut (hat '-iifjl Shape A or retorty i incl' in, the (leplll'e. oil' the 4L el the layer oit-reacting ni it-iue, siiirrnw ol the tube charge, rotation (either ci'inliniious y or at Liuterva sbr-ulti also be eo .ilerei'l. 'he .nirogtiu o mirri-5e,lepemls ou 'the velocit); ol the gas current uml the uioleffular ililliif-lion rate uil-- i (ler the e g er1iueiil-al eouilitious, Vsfor its chiel source o'l motion. ln a Heeoutlzuy wey il` might be carried, for example, in the 'torni ol liquid 1v-'uml or eyzuiul vapors auil these,

by reverse action, might again liberale the nitrogen to some exten s,

iu the torni 'mollen romporuule. or lorm of vapors oil suilr eompouiuls. lt, `will also, aller being liberated ents el the rif tion el the carbonate (or other con'iiimuul l nitrogen can also (li lluee to this surface very lllereof)7 equation L or #he hn-urlwirsil'ion ol Clo/suol, 0 llmli =u^l, have the' mim-1, treeiloii': olf motion' :ruil vpenetration the nitrf l' fi. c. from 'the velocity nll the our; current the molecular lill'usion velocity el the l'rce wiilium,

The effect of volatility has already been discussed miller the mobility of the sodium and carbon in the reaction mixture. It may also be of lgreat importance otherwise; as can be .seen lrcin the law of mass action and phase rule. lt also causes circnlation of liqi1i lsan l rapors, washing action upon particles, and changes concentrations and compositions of mixtures. Besides'this,-it may holy to keep temperature conditions uniform because of latent heat of vaporiza.- tion. The flowing of niolten matter, e. g. sodium carbonate, from whatever cause, must have much elleet as shown by the aboieconsitlerations. Flowing of liquid may even cause segregation, by uf'asliing away particles 0f arbon or irony and thus cause loss of ellivantageonely. For example, I have noticed that when eoiulitions 'were not quite right,

= beneficial results were frequently obtained by s'inply rotating the tube containing the rc. ion mixture, alwut 155i@ This 1s preunal/ly lue to the flowing of the liquid by and the consequent sur-tace changes. r reo ml the er'l t cete duc to surface tension or to relai i .ely minute lillerences in surface tensiona as vvell to that phaee of surface tension which generally termed capillaril and also the eonilil nis which, modify these, as o' mut-li import ace.

As the composition of the molten alkali mixture is constantly changing; during the y procesa the suriloce tension is presumably changed lilevviae, thus, tending to cause inoti. of the ynrticles in a well known manl'fie iinely powdered, porous reacting un' atl'vrijls exceptional oiiiportunities for lle. y to exert its influence. '.lhe powlei niass ol1 carbon an'rl iron will take up caliiiilaril'y the liquid independently or iaralely supplied thereto, or will similarly ii;-alte up the liquid which may have ilrainecl by gravity to the lower portions thereof all@ the. content of liquid remain ing in the ijlrainetl portions. hasbeen in whole or in part removed; the action being 'll'iesofliuui may more through the in inet 'dissimilar to the way in which a lamp 'ick takes up oil. ln this porous mass of l carbon anfl iron the liquid comes in cont-act 'with the very large surfez-,e of the iron, and incidentally, oli course; ol' the carbon; while rapidly, which obviously tends under the lwtevribetl conditions to fa vor the fmrmationy ol c ranid; the lihus oit li( uid more or less rapidly being` dissociated lo yieldfthc free raiiiliiini, which lutter may either react where ,3 eiei'ioy. (ln the oth ei-linnfl it may be caused to, or may eren inadvertently wash the particlos into more advantageous positions, thus.

liberated or elsewhere on the surface. Preferablv the thinner the film the more readily can the nitrogen molecules 'penetrate ity to reach the reactive surface of the carbon supplying iron. When th is film is too thick the nitrogen cannot Well penetrate and the reaction is impeded. Further, the more extensive the iilm surface, the greater is the liberation of free sodium per unit of time, together of course, with a correspondingly accentuated evaporation of the liquid compound itself, or of its decomposition products.

If 'too Amuch molten liquid is present, the pores or interstices will, of course, be more or less lilled or choked therewith, thus preventing the required contact of the gaseous materials. such as nitrogen and sodium vapor, with the catalytic solution surface. lVe mayrregard the nitrogen as burning the carbon and sodium to sodium c vanid, and if there be either insutlicient catalytic solution surface or an inadequately reactive or insufficiently extensive film. this peculiar combustion is retarded or in extreme cases maybe almost vwholly prevented.

, The above considerations concerning the mobilityy especially that due to capillarity,

Within the reacting mass led me to make a' considerable number of experiments of Which the following are typical: An effi- ,cient' mixture of 4.0% of graphite of 100 mesh, 40% finely powdered iron (P71. G. IIV), and 20% of freshly ignited sodium carbonate was prepared. A flat bottom iron retort. 70 mm. by 10 mm. was attached to a long piece of wire so that it could be inserted into a long pipe which was heated to the'desired temperature and through which a current of nitrogen was passing.

'Experiment LOne half a gram of the above reaction mixture containing carbon, iron, and sodium carbonate was tamped into the retort ina. uniform. layer. The retort was now pushed into the tube which was heated to 10700 C. Not more than a minute was allowed for it to come to the reaction.temperature` and it was then allowed to remain for five minutes more in the heated nitrogen. It AWas then Withdrawn to the cooler part'of the tube and allowed to cool in the current of nitrogen. Upon lixiviation, so little cyanid Was found that it could not he titrated accurately with silver nitrate solution. lThis experiment Was repeated with the exception that the tube was heated to 1000o the result being the same.

[imperi/nent lil-T he retort "was again charged with onehalf-gram of the aforesaid mixture of sodium carbonate, iron and graphite, and six-tenths of a gram (an-.equal volume, therefore; an equal depth of layer) of a mixture 'of equal parts of the 100 mesh graphite and4 the iron A(Ph. G. IV) was tamped on the surface of the reactive mixture containing the sodium carbonate. This was pushed into the tube containing nitrogen heated to.990o C. As before five minutes Was given for the reaction. The yield of sodium cyanid was 25% of that theoretically possible.

.l Experiment 6.-"1 he operation was noW repeated substantially as per Experiment 4,

at 900 C. but without tamping the ma- 75 terial. l-lere the quantity of cyanid formed was too small to be titrated with silver nit `ate Solution.

imperi/ment /`".--This experiment Was exactly like Experiment 5 with the exception 80 that twice as much (one and two-tenths grams) of the above carbon-ironvtamping mixture was used, vz'. c., the layer of tamping material was twice as deep as that used in Experiment 5. ln this case, I found that about 50% ofthe sodium carbonate had been converted into sodium cya-nid in tive min. utes at 990 C.

l'wperimeat"This experiment Was like Experiment 7 with the exception that the 90 one and two-tenths grams tamping mixture were placed in the bottom of the retortv and the five tenths gram reaction mixture formed the upper layer. The result Was substantially the same as that of Experiment 7, 2'. e., fully 50% of the sodium carbonate was converted into sodium cyanid in tive minutes at 990o C. (At another time the same yielded fully 40% in tWo minutes.)

Eperz'ment .Q -In this case the quanti- 10'0 ties of material used lWere exactly those in the preceding Experiment 8, but they 'Were thoroughly vmixed instead of being used in layers. ln this particular experiment the yield of cyanid was about 40% of the 105 theory, in live minutes at 990 C.

/Lznperimcnt 10.--This experiment was practically a repetition of that described under Experiment G, and is cited merely as an exception, since in this case, a small yield of cyanid, to Wit, 15% was obtained, which demonstrates that even under what appear to be idei'itical conditions, some rvariations in yield may occur. In the case in point, the cyanid formation may have been due t0 115 some condition such as lack of uniformity in the mixture, all parts not being uniformly saturated, or in other Words, somesutliciently effective film surface Was presn ent which during the course of the operation, once being presentwould tend to eX- tend itself, due. to capillary action, or -cir- .culation induced by changes in composition in the liquid phase.

i I have stated in the preceding experi- 125 ments, where yields were obtained, that these yields were fully 50%, etc. This phraseology is used advisedly since the yield was obtained by titration and in each case more or less of the. cyanid was volatilized during the course of the operation, especialiy owing to the extent of the surface present from which lvolatilization could take place, At different times Experiment 8 was repeatedV withy the exception that the period of operation was extended to one-half hour. In each case substantially no cyanid was found in the residue in the retort or tube, while further but little, if any, `carbonate remained, which in view of the preceding shorter term operations indicated that cyanid had been formed and had thereafter volatilized.

The following' experiments are of interestv as showing the effectiveness of the process at relatively low temperatures:

Eperimen 1].-This experiment was substantially like Experiment 8 with the exception that in the reaction mixture, one half of the sodium carbonate wasreplaced by potassium carbonate. The temperature of the operation was 840 C. and the cyanid yield in the residue, upon titration, and at the expiration of a thirty minute run was fully ln this connection' it may be of interest to note that the melting point of sodium carbonate is apparently about 860o C. while'that of potassium'Carbonate is about 880o C. vThe eutectic point of the mixture is about 690 C. f

Empem'ment 12.-This experiment was a repetition of Experiment 11,-except Ythat the temperature of the operation was 7800 C. and the timeforty minutes. The yield by titration was 22%.

Experiment 13..-This experiment was a repetition of Experiment 12, save that the temperature was 730 C. and the time fortyiive minutes. Yield by tit-ration somewhat over 2%. vObviously as the temperature was lowered the cyanid formation per given-unit of time decreased. Correspondingly, other things being equal, more elevated temperatures afford markedly greater yields per unit of time. Thus I have converted 443% of a given quantityof sodium carbonate int/p cyanid in two minutes, and over 1n tive minutes at 1060o C.

The efficiency of the catalyzcr in lowering the temperature of cyanid formation and 1nkcreasing the rate of its formation is' even more strikinel shown b f coninarinT the foregoing with the following lstatement of Dr. Ewan, .in T/wrpes Llczfoamy of AppZ-ed ULemz'stry, Revised Edition, Vol. II, 1912,J page 196:

A few temperature,measurements which the writer made with a platinum rhodium thcrrnocouple showed that potassium vapor is first evolved from a mixture of potassium carbonate and charcoal at about 1350O and that the formation of cyanid takes place very slowly at this temperature, the potassium cyanid volatilizing for the most? part. 1 Experiments 4, and 10 were made with l the very efiicient reaction mixture above referred to, bututheir yields were unsatisfac- Experiments 5, 7, 8 and 9 gave veryfrapid conversion ,of carbonate into cyanid. Presumably in Experiments 5 and 7 they car-A bonate (sodium) melted and then penetrated by capillarity into the upper layer of tain ing material, thereto be rapidly 'andeigciently converted into cyanid. In Experiment 8 where the reacting mixture constituted the upper layer, the same result was` attained by means of the capillary forces.

aided by gravity, while in Experiment 9 the same result as in 8 was attained by thoroughly mixing the materials before they were tamped into the retort. Also when the4 f very efficient reactive mixture was used inv the` thinner layer the yields varied from almost nothing (4 and 6) to a. comparatively poor yield in 10; presumably because the Aslight depth didA not permit 'of suicient drainage of the molten sodium carbonate fromv the' upper layers of the mixture s o as to provide suitable reaction conditions. On n l the other hand when a larger quantity, such as twenty-five grams was itam ed into a onehal:c inch iron pipe, presumab y a portion of Lherrnalteu carbonate in the upper portions of ,Lhefmixture drained to the lower portions :o :n to leave conditions in the upper partf ot ille mass very favorable for cyanid formation thereby automatically producing conditions substantially similar to those provided intentionally in Experiments 5 and 7.-

Use 0f apparatus shown by way of ewemp7z'jeoion.-Assuming now, that the con,

ditions prescribed havebeen observed in the,

disposition of material in the` furnace shown in the accompanying drawing, and thatan adequate current of nitrogen is flowing;l

through the tubes 2, each of the latter hav ing in this case a mass 9 of intimatelymixed finely powdered iron, sodium carbonate and graphite in placetherein. The actlons and reactions take place .as above described-,w1 A

sodium cyanidy is formed in the inyterstices' of the masses 9 and carbon monoxid gas flows ofl" together wth some excess nitrogen,

through .the tubes 5. It is to be understood,`

of course, that ,it is possible to operate q uanl titatively, so .that substantially all of the nitrogen, carbon and alkali metal compound@ are converted directly intocyanid; and al1 of this can be done below the melting'point of. @Oppem The cy'nid whichmay lbe obtainedv from 'the residue of the masses 9isfvsubstantially pure; v and as `cyanid is exceedingly fion, and hot flue gases are, in general, `direct functions of the time anoperation case. It is also evident that in many cases ".lvenient means for separating it from the .thesis, vwe find frequent references to the IJicinperatures, in from two to twentysfye such processes can ina year. it is evidenttha't such-au expeditious procy ess is an. enormously advantageous one even ywhat providing aZ considerable saving in advantageousjn'certain'cases to operate for a. short time only upon oa given mixture soluble in water, lixiviation affords a' conesidues of the reaction. y

Air may be used in lieu of nitrogen and ye arrangement of -materials andparts shown in Fig. 4 has been found a convenient one in such icases.

8, the operation and apparatus is substan-` tially the same in the case just discussed. It may be well to note, howeveigtha't when using air it is advisable to operate at a slightly higher temperature, .say 50 C., than when using nitrogen. v'lliecharcoal combines u'itli theoxygen content of the iniiowiug air so that a mixture'of carbon inonoxid and nitrogen, With possibly Sonie carbon dioxid passes on to the'. reactive mixture and the nitrogen reacts therewith, as before.

The heat necessary can be produced in the furnace tube or retort itself, thus almostentirely avoiding Wear and tear on the apparatus. y i

Tiff/ze fueron-In the art of cyanid syn-A yields obtained, but much less often to the time required for effecting the changes. I Wish toaemphasize thefvery great impor tance of this time element and have aimed constantly in developing. my process to make it'an expeditious one.

*When it is considered that the cost of labor, wear and tear of apparatus, interest charges, loss of heat by radiation, conduci'equires; We realize that the time factor may become, sometimes, of even greater importance than the ultimate or practicable yield. For example, the time given for certain processes is sometimes ten hours or more, with alleged yields of perhaps 50@ or much less, and` under very drastic temperature conditions. My process affords far greater '-yields, and at very moderate minutes according to the conditions selected: In other Words, my process will easily produce asmuch cya-nid in several days as From athis,

if it Were lequally affected 'bythe above items of expense, which, however, is not the .it is desirable to sacrifice alie yield somen time will result. Thus niay be found Withthe obyeca cf'producing, let us say, a

. Save for the provision of the oxygen con-y 1 suniing charcoal 7, and separating screens re-using the residue either alone or mixed with fresh material or materials, c. g. iron and sodium carbonate, as desired.

Rsum-To epitomize, therefore the features of the process Which, 'iyhen observed, lead to consistent results and commercial success in the production of cyanid:

(c)l A. material, e. g. iron, capable of readily dissolving as a catalytic agent. l I

(6)' This material should be present in such quantity and shouldbe in such condition 4as to afford an abundantsolution or reactive surface. y 'Y f.

(c) This surfacey should be maintained.` By this I do ,fot mean that necessarily the` same identical surface must be maintained, although even this is preferable, but that the extent of available surface consideredl as a' whole should be substantially maintained and that the carbon present therein ,should be maintained in sutiicient quantity to permit 0f an energetic and practically continuous reaction taking place betweenl said carbon and the nitrogen and free, oi' probably free, alkali metal. v

(zl)"l `he carbo-n should be present in such quantity and so disposed with respect to the active surface of the latter may be ade# quately and substantially continuously supplied.l To this end also proper contact must be effected and maintained between the carbonaceous material and said metal, unless the latter' be initially sufliciently carburized .to pei-mit of continuous operation fora sutli- .ciently extended period of time. e) The source of the alkali metal, or base of the compound sought, may either be intimately niixed with the catalytic material or' otherwise. The essential feature to be observed is that the metallic vapor yolecules must be developed or be supe if. in adequate quantity' and must-either be in contact with or be able to make contact with the car# `loon supplying solution surface at the same instant as the nitrogen molecules.

-oiiiy be extensive and properly carburized but it must be accessible to the molecules of the gases or vapors participating in the reaction. ln other )yoi'ds it must notbe clogged up oi covered with too thick ilins of liquid. Willich," therefore, the source of the base of the cyanogen compound sought is a material which is molten at the temperature of the operation. c. f/.vsodiuin carbonate. and' this'liquid islpi'sent in excess, a sufficient extent of effective solution surface must still be preserved by affording proper drainage facilities, or the equivalent thereof.

(g) -The mass should be suliciently porous to pei'init of the, at least, inodcratelyfree yield of 50%; thereafter removing the cyanid formed in any suitable manner and passage of nitrogen therethrough Whiereby carbon, must be employed carbon dissolving metal or agent that the re- (f) The reactive solution surface must not acid, oxalic acid, etc.

vthat the cyanid 'formation not only to supply nitrogen in adequatev euantity to a s'uilicient extent thereof to avor the rapid production of cyanid; but, also, to sul'li'cien'ly dilute and preferably to even dislodge or sweep away, the deleterious, inert carbon monoxid from the reaction surfaces.

(h) The mustv be such to afford the vrequisite molecular activity-I ind in practice that it is advantageous to operate at from, let us say 780C C. up to somewhat less than the eutectie peint of iron and carbon, when iron is' the solvent used, and it is in solid form. The expeditious production of cyanid in fairly appreciable quantity, by this method, can be eiiected however at. temperatures as lovv as, let us say, 730O to 780O C. The principal oint to be observed in lthis connection is that at the lower temperatures the mixture must be more intimate and complete and the effective reactive surfaces more extensive than at higher -temperatures. In other words, what We lose in temperature, must be made. up in some other Way. T he lower temperatures of course', tend to prolong the life of the apparatus. t l'lei'erring again to the itemized characteristics and advantages of the process as a whole it will be seen that items l to 25 inclusive have been discussed or are self-evidently possible from the preceding disclosure.

There remains to be presented a more particular discussion et the use of the process in forming ammonia, and in the produc tion of such ley-products as hydrogen, ormic Dr. Ewan states in his article in Thorpes Dictionary 0] Applied Chemistry, above referred to, that-'- i :i in every process now used for the production of cyanids, the nitrogen is derived from coal more or less directly.

This is equivalent to saying that the cyanids are, in general, preparedfrom ammonia. 4

In contrast to this, one of the main 9bjects of the present invention is the commercial production of ammonia through the instrumentality of cyanids produced from atmospheric nitrogen; thus making the prof duction of ammonia independent of the use of initial ty combined nitrogen. Indeed I contemplate effecting", in some cases, the production `ot ammonia in such a manner will be but momentary or evanescent, so that in effect the ammfinia may be produced 4not only syn' thetically but practically directly. This last Will be hereinafter more fully referred to while considering the phase of the process which involves the 'formation of cyanid as alkali metal cyanid in accordance with.

temperature of the operation lowing equation.

a definite and intermediatestep.

. Ammoaia.-r\ssuming that the apparatus shown by way of exemplication has beimy charged with iron, carbon and a refernbly inexpensiveVB alkali compound, al in Huib f able form or condition' and properly dirposed with respect to each other to yield u equation 4.

After the cyanid forming reaction hh! been etlected to the desired extent, the resi due consists chiefly of "a mixture .of iron,v {t} carbon and, for example, sodium cyani and this residue is especially Well adapte for'thc production of ammonia, without the necessity of previously removing the cyanid content thereof. For example, the mass n from the reaction may be cooled toa suit-v able temperature and a current of steam may then be passed over or through it. Ammonia will be produced according to the fol- The equations 4 and 9 show that We News, here, a cyclic process for producing'ammo'fr nia from air andWater through theinstrumentality of carbon ,by alternately passing nitrogen or steam through a mixturefconr sisting of iron, carbon and, for'examplii is sodium carbonate.. y

It desired, the carbon. monoxid and hydrogen may be burned with air, thus pw' .ducing heat, which may be utilized in thi process, and a mixture of carbon dioxid, 1N nitrogen and ivater. The Water condense! and the carbon dioxid and nitrogen togethqwith the ammonia already formed may be passed int-o a solution of sodium chlorid, as in the ammonia-:oda process. The catd 10i bon dioxid of the mixture is thus ultimately utilized to form sodium bicarbonate, or soy d iiun carbonate, leaving the nitrogen in l fairly pure form for use in the preparationr' r 5 of another portion ol cyanid according tj'lg equation 4. The process is admirably adapted for u,

with the ammonia-soda process as it'mlyf-j.

utilize the waste atmospheric nitrogen-from this process to furnish the pure ammonilgv-'i required in the latter process, very'cheaply.

The expense of ther ammonia as Well as the troubles.. arising from the mpuriti liable to be contained iu commercial ammoj nia are at present serious obstacles to the lil economical operation of the ammoniaprocess; but both of these diiculties'm-l.

obviated if my process is used in connection` therewith.` The sodium carbonate acts cyclicallyjig. tix nitrogen by beine;r transformed into cylnid (equation 4) and then the cyanid ii" changed buck into carbonate, (ecation 9),V

by the use ot steam. If desire we may start with 'the' cyanid instead of the carbonf .18|

lOk

atein 'this cycle. For example, a mixture of potassium cyanid, carbon and iron may be heated in a current of steam until a subc stantially quantitative yield oiE ammonia is obtained, The ste-ani is then displaced by a current oi? nitrogen and the temperature is raised to about i100O C. In a short time the action will be completed and the retort i ivill be found to contain much cyanid; the

potassium carbonate having again been transformed into cyanid Upon adding carbon as required, the process may be repeated tendency for cyanogen compounds to pass over with the ammonia. f

then the temperature is adequate but not excessive, very little ot' the cyanogen coin` pounds will pass over with the ammonia', and even these can be recovered very easily by using suitable scrubbers. If desired, the :uninonia may be obtained from the scrubber-s in anhydrous form.

My invention also ci'iiitemplates, as I have previous v intimated, the production of a'mnrnia iii suoli manner that the cyanid orinition will be transitory or evanescent. To so effectuate the process it is merely neces saiv to reduce the temperature of rthe operation to a point (approximately 7250 C.) where wliili` the reaction involved in equa tion 4 is proceeding` the reaction shown in equation t) may also occui. Under such conditions the catalytic material, iron, previously considered, is supplemented by the `alkali metal carbonate which now` may be The regarded as an auxiliary catalyzer. equation in'ay be represented thus:

in this case one of the catalyzers (c. g. iron) probably acts passively or merely as a solvent tor one or more ot' the actively participating elements; while the second or auxiliary catalvzer is itself actively involved in the reaction. The cyanid is probably `intermediately produced, and in effect simul- "with consequent low caloiiic intensity'. By

Nitrogen and its elimination fromV comy yZnisz'le gosen-,The disclosures in the process show in a number of'instances'tliat atmospheric nitrogen is obtained asia` Waste product, and that this nitrogen may, then serve as 'a source ofnitrogenfor the synthesis ot cyanids. Also, the disclosures show i that some commercial combustible gases conf' i. taining nitrogen maybe used to supply the nitrogen for cyanid synthesis With a oonsequent purification, freeing the gas from' nitrogen. For example, a gas may contain, let us say, 30 to 40% of combustible materigtl such as carbon mono-Kid, With lthe remaih er made up substantiallyl of nitrogen,

removing the" nitrogen from such agas in Vthe cyanid synthesis, the resulting gas would consist almost entirely of combustible` material and, it would be far more valuable on account of its high calorific intensity. l

' General remarks-Owing to the numer' ons possibilit-ies of the process, I have only described. some of the more important steps .90 involved, hence I wish` to be limited only by the general spiritoi' the above disclosures and bythe appended claims. c

For convenience, I have described most ot the experiments, and Written the corresponding equations, for sodium compounds but these descriptions are intended to apply to alkali compounds generally. and, Where applicable, to metals capable of performing like functions in the proc-iss.

In certain of the appended claims I have 1miv vused the term vapor and in such cases where not otherwise specicallyliniited or restricted, this terni is to be regarded as of suiticient breadth to cover, for example, both alkali metal vapor and the 'vapor of an alkali Vmetal compound, e. g. sodium carbonate,etc.

vreact-ion in which initially combined alkali metal, initially tree nitrogen and dissolved 't carbon participate to torni a cyanogen compound.A by bringing the molecules of 'the i1154 nitrogen into contact with the dissolved carbon present in an extended catalytic solu. tion surface, While at the same time elfectin a liberation of molecules of said halkali nietz from the material with which they are combined, in the vicinity of said extended solution surface and maintaininga-s'ipply .of nitrogen suiiicient to immediately combine said carbon in said surface therewith and with substantiallyv all of the alkali metal 1 125 effect directly as liberatedWhereby to sii stantially eliminate the tendency for thev latter, by means of its vapor pressure, tio establish a chemical equilibrium and retar 1 the further liberation of said metal. i

2. The prceess oit iixing atmospheric nitrogen which comprises eiiecting a v1 crous reaction in which initially combined alkali metal, initially free nitrogen and carbon dissolved in iron participate to form a cyanogen compound, by bringing the inclecules of the nitrogen into contactlwitli the dissolved carbo-n present in an extended surface of' said iron While at the same time efi A the alkali metal in effect directly liberated, whereby to substantially eliminate Athe tendency for the latter, by means ot' its' vapor pressure, to establishV a chemical equilibrium and retard the further liberation of said metal.

8. The process of xing atmospheric nitro gen which comprises eecting a vigorous re-4 action in which initially 'combined alkali metal, initially free nitrogen and carbon dissolved. in solid iron participate to form a cyanogen compound, by bringing the molecules of the nitrogen into contact with the dissolvedicarbon present in an extended `surface of said iron While at the same time effecting a liberation of molecules of said alkali metal from the material with which they are combined, in the vicinity of said extended solution surface and maintaining a supply oit nitrogen sufficientk to immedi atelycombine with said carbonio said surface and with substantially all ot' the alltali. metal in effect directly as liberated, `whereby to'substantially eliminate the tendency :tor the latter, bymeans of its vapor pressure, to establish a chemical, equilibrium and retard the further liberation oit said metal,

4. The process et' fixing atmospheric nitrogen Which comprises effecting a vigorous reaction in hvhich initially ycombined alkali metal, initially Ytree nitrogen and dissolved carbon participate to `term a cyancgen compound, by bringing 'the molecules ot t ie nitrogen into contact With the dissolved can bon present in an ext-ended catalytic solu tion surface, While at the same time eiiecting operation at a temperature below the eutectic point of the carbon solution.`

5. The process of fixing atmospheric nitrogen which comprises effect-ing a vigorous reaction in which initially combined alkali metal, initially freev nitrogen and carbon' dissolved in solid iron participate to form a cyanogen compound, by bringing the molecules oi the nitrogen into contact with the dissolved carbon present in an extended` suritace said iron While at the same time effecting a liberation of molecules of said alkali metal from the material. Wlth which they initially combined7 maintaining l` supply ot nitrogen suilicient to lmmedlately combine with said carbon in said Surface and with substantially all of the alkali metal in effect -directly as liberated, whereby to substantially eliminate the tendency for the latter, by means of its vapor pressure, to establish a chemical equilibrium 'and retard the further liberation of said m'etal and maintaining said extended solution surface by establishing the upper limit of the temperatine of the operation below the eutecti peint et the carbon-'containing iron. o. The process of fixing atmospheric nitro- Aeen which comprises effecting a vigorous reaction in which initially combined alkali metal, initially tree nitrogen and dissolved carbon participate to form a cyanogen compound, by bringing the molecules lof the nitrogen int-o contact with the dissolved carbon present in an extended catalytic solution surface, while at the same time eiectin a liberation or moleculesof said alkali met from the material with which they are com bined, in the vicinity' of said extended solution surface, elevating the temperature 0f the reacting mass above 700C' C. but below I' temperature,Whereat the process becomes in effective and maintaining a supply of nitrogen sufficient to combine said carbon in said surta ce therewith and witli'substantially all of the alkali metal ineii'ect directly as liberated, whereby to substantially eliminate the tendency forl the latter, by means of. its vapor pressure, to establish equilibrium and retard the further liberatien of said'metal.

7. The process ci vfixing atmospheric nitrogen which comprises effecting a vl op ons reaction in which initially combine n chemicalA vlis ltali metal, initally freenitrogen and cclv"v bon dissolved in iron participate to form s cyanogen.cempound, by bringing the molecules of 'the nitrogen into contact with the ltali metal from the material with which they are combined, in the vicinity ofsaid extended solution surface, elevating the temperature of the reacting mass above `700 C. but Ibelow ,ai temperature whereat the noemen l tially eliminate the tendent-y for the latter,

by means o't its 'vapor preses-ure, to establirih a chemical eipiilibriuin and retard, the turrher liberation olt said metal.

8. The process ot ixing atmrmpherie ni- VYtrogen which con'iprisee effect-ing a. vi You@` reaction in which initially oonibined alkali metal, initially 'lil-ee nitrogen and carbon dissolved in linely divided iron partiripate to form u oyanogen Compound, by lninging` the molecules of thenitrogen into contat-t lwith the dissolved carbon present in an ere tended surface of Suid iron While at the Same time eli'eeling a liberation of molecules of Said alkali metal from the material with ufnioh they are combined, maintaining' a Supply of nitrogen sufficient to combine 'with said can 'hon in said surface and with the liberated alkali mehr]r and substantially continually` reple ishig the carbon in said surface, so as to maintain the carbon eolution at sub- .Qtantially its lmaxii'num Strength. by providing pulverulent mareos of oarbomieeorm inaterinl in operative Contact with said finely divided iron7 throughout substantially the extent thereof.

S). The proceso of lining atnflospherie nitrogen which comprises etl'eetingf a vigorous reaction in which initially Combined alkali metal, initially free nitrogen and diasolred ooi-bon participate to form a eyanogen vom pound, by bringing the molecules ofthe nitrogen into Contact with the dissolved earl u;-npre$ent in an extended eotalytio solution siirfiieeql eli'ecting/a liberation of molecules of Said alkali metal'from the mate: rial with which they1 are Combined.v maintaining a bupply' ot nitrogen Combine with said carbon in soidurtaee and with substantially all of the molecules of eaid alkali metal encountering' said surface and repleniahing the carbon in said sui-taceao as to maintain the Carbon solution at. lubstnntially its maximum Strength.

i0. The 1 )rocess of ixing atinoeplnerio nitrogen which comprises eii'eoting a rig orons reaction at a temperature above4"0lo C; butv .below ateniperature whereat the proceso becomes ineti'eotii'e. in which" the alkali metal base or' compound oontziining oxygen, initially free nitrogen and carbon dissolved in iron participate to form :i cyanogen compound. by bringing the nioleenles ot' the'nitrog-en into rontar with 'the diesolved carbon prerent in nn extended fsuriaee of the iron` while atI the saine time elleeting a. liberation of the alkali metal from the material with which itoonibinetl. the oxygen ol said alkali ine-tal compound being combi-vieil with, carbon to sullieient to,A

'fresh carbon thereto ,troni a `form an oXid ot the latter element and maintaining an extended reactive surfaoe throughout the mass oi? Said iron by o0n- (hating the operation at a temperature belouY the euteetic point of theV carbon-containing iron.

ll. lhe proees of ixing atmospheric nii'r ren whieh ooinprisee effecting a liberation of alkali metal vapor from a coiny)ound ol' the saine substantially ureventin' 7 2D .the liberated metal. from exercising a reaction retarding vapor pressure upon the residue of said compound by interiningling a current of :tree nitrogen with the so liberated alkali metal and. reaetingga upon the mixture with carbon held in solution lin a ninos oit atalytie nmterial, to form a Cyanogen Compound of Said alkali metal.

lheprocess of fixing atmospheric nitrogen which comprises bringing initially tree nitrogen into Contact ywith dissolved Carbon present. in an extended catalytic solution surface and there eombiuingwsaid die- Solred carbon and said nitrogen with an elenient capable oi acting as the base of a C v-v anog'en compound. Said element being initially present in cheinieal combination with at least one other element. to form a. eyanogen compound of said first mentioned element, the while maintaining said Solution surface eiectire by efficiently :'supplying BUn-gl-XOll sour-re of said earbon.

1?). The process of fixing atmospheric nitrogen which lcomprises effecting a liberation of alkali metal `vapor from a compound or' the saine, lsubstantially preventing the liberated metal from exercising a 1'eacti0n` remi-ding vapor pressure upon the residue o'l Said compound by intern'iing'ling a cur-` rent of ,tree nitrogen with the so liberated alkali metal and reacting' upon the mixture with carbon held in Asolution in a mas` of catalytic material disposed in the immediate vicinity of said alkali metal Compound, to form :1 oyanogen compound of Said alkali metal. i

|14. The process 'of fixing' atmospheritl nitrogen whioh comprises eleeting a, liberation of alkali metal from a compound of the sinne, subntantially preventing the liberated metal from exeroieing' a reaction retarding etl'orl upon .the residue of said compound. by inter-mingling free nitrogen with the so liberated alkali metal and reacting upon the mixture with rnrbon held in solution in a unies ot catalytic material, to form a cyanonen eonrpound of Said alkali metal.

l5. The pim-ess of lixing atmospheric' nitrogen which comprises effecting a liberation of alkali metal from a eolripound ot' the same. suhetantially preventing the liberated metal from exercising a. reaction retarding' effort upon the residue ot' Said compound by iutermingling free nitrogen with the So

Images