Equilibria Investigation of the Bucher Process for Nitrogen Fixation Citation Hainsworth, William Richard (1918) Equilibria Investigation of the Bucher Process for Nitrogen Fixation. Master's thesis, California Institute of Technology. doi:10.7907/a17y-hn74. https://resolver.caltech.edu/CaltechTHESIS:07242024-153823595 Abstract No abstract. Item Type: Thesis (Master's thesis) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1 January 1918 Record Number: CaltechTHESIS:07242024-153823595 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:07242024-153823595 DOI: 10.7907/a17y-hn74 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16587 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 25 Jul 2024 16:57 Last Modified: 25 Jul 2024 16:57 Thesis Files PDF - Final Version See Usage Policy. 10MB Repository Staff Only: item control page

.THESIS . EQUILIBRIA nrmsTIGATION OF THE BUCEl~R PIDCESS FORNITROGEH FIXATION. By WILLIAM B. HAilTSUOmrH Prepared as a Partial Fu.lfilment of the Requirements for the Degree of ]Jaster of Science in Chemistry. THROOP COLL.EGE OF TECID:rGLOGY. 1 9 l 8 EQUILIBRIA INVESTIGATION OF THE BUCHER PROCESS FOR NITROGEH FIXATION. The purpose of this investigation was to obtain data which would aid in increasing the efficiency of the Bucher process for the production of sodium cyanide. The indications are that this process when perfected v.vill compete favorably with other proce·s-ses for the fb:ation of atmospheric nitrogen, by fUrnishing ammonia products with a minimum outlay for apparatus, labor and raw materials. The historical development of the process, wh.ich dates from approximately 1840, has been completely outlined by Bu.cher "(l}. Du.ring the period of development, however, very little work has been done upon the equilibrium conditions of the reactions involved; hence the process has not been developed to its fullest extent. Essemials of the reaction. The fundamental reaction of the process is m.2 co3 + N2 + 40 ~ 2 NACE' + 300 - 138, 500 cal.(l). The presence of a catalyzer such as iron or nickel is required if an appreciable yield of sodium cyanide is to be obtained in a reasonable time. That the reaction is reversible has already been proven; the equilibrium point is far to the carbonate side of the reaction at temperatures near 860° c. The commercial operation of the process has been considered in detail by Bucher (l}; only cut and try methods of control, however, have been used in h~s development. The large heat content change (-138,500 cal.} of the reaction indicates that the equilibrium constant changes rapidly with the temperature. It was therefore thought probable that an accurate determination of the equilibrium constant at various temper~tures would lead to a knowledge of the most favorable temperatu.re for the COlllllercial o-peration of the process, an:l the pres ant thesis describes the attempt to obtain this information. In this work the equilibrium constan~ K of the reaction was determined over the tanperature range 898° to 1016° c. This constant is the ratio of the cube of the partial presm.re of carbon monoxide over the reacting mixture, to the partial pressure of nitrogen over the mixture for- a given composition of gas admitted to the reaction chamber. The greater the value of K the great- er is the tendency of the reaction to go tONard t.h:e production of s()ditim cyanide. The experiments described below show that the value of K increases very rapidly with the tenrperature, and indicate that as high a temperature should be used in comnercial operation Of the Bucher process as is consistent with comnercial limitations. Experimental. The bomb The reaction chamber, or bomb, as it will hereafter be called, is shown at A, Fig I. It consisted of a short piece of seamless steel tubing about 2-1/4'' in diameter, and varying in length from 4 to 9 ", the ends being closed by plates into which tubes for the entering and exit gases were autogenously welded. After welding on one of the and plates the joints were tested for leaks, a thin layer of steel wool was placed on the bottom of the tube am the charge inserted. .Another layer of the wool was then placed over the charge and the other end plate welded in place. leaks. The entire bomb was then tested for In the later experiments the thermocouple tube was placed within the gas entrance tube, which made it necessary to charge the bomb with this tube in place, to prevent difficulty in getting the tube into the bomb. The steel tubing served well as a reaction chamber, although the length of the run.was limited to the time necessary for the tube to oxidize through and develop leaks. In only one case, however, did this factor make any material difference in the results obtained. Furnace A Heraeus platinum-.vound electric furnace was used to heat the bomb, variations in temperature were obtained by means of a transformer which was capable of giving voltages in steps of 0.5 from Oto 110. The first; t\vo or three runs were made using only asbestos boards to prevent radiation from the In these runs the temperature was measured open ends of the :f'urnace chanber. by means of a thermocouple placed between the interior furnace wall and the bomb, moving the tip of the couple from one end of the tube to the other showed a large variation in temperature. Since it was kn.own that a small vari- ation in tEmperature caused a large change in the equilibrium constant at temperatures near 1000°0, it became evident that a much more uniform heating of the bomb would be necessary. To avoid unnecessary radiation at BB, Fig I, end plug containing heating coils were used, thus supplying heat to balance that lost by radiation. To make the construction of these plugs simple and at the same time to determine with more accuracy the true temperature of the c1mrge, the thermocouple was inserted in the bomb through the gas entrance tube, the joint between the couple tube and gas entrance tube being made tight by means of the rubber tube c. \'Ii th this arrangement it was necessary to have only one hole in each end plug. The details of construction of the plug are sho,vn in the assembled drawing Fig. II. They were made according to the gen-4- eral principles outlined in the Bulletin of the Bureau of Standards (3), having alternate layers of heat conducting and heat insulating nnterials so arranged that the conduction waa perpendicular to the :path of radiation, thus tending to transfer the heat to the fUrnace wa,11s rat.her than to the air. At the outside end of the plug a variable source of heat was applied, thus theoretically 81C.8.bling the radiation loss to be counteracted at any temperature. Nichrome wire, protected by a layer of almidum cement, was used as the heating element and a · flange of asbestos was placed outside the nichrome wire to insure the complete covering of the open ends of the furnace chamber. With the plugs in place, duri~ a run, the temperature gradient inside the bomb was obtained by moving the tip of the thermocouple from one end of the bomb to the other by one centimeter steps, and reading the corresponding tanperature by means of a potentiometer indicator. These data enabled curves to be plotted showing the temperature of various parts of the charge. The temperature variation in the central part of the bomb was small, not exceeding 6 or 7°0, when a potential difference of 35 volts was used across the end plugs. The difference in temperature between a few centimeters at the ends of the bomb run the center of the bomb was much larger, often amomiting to about 15°0. Since the steel wool occupied sane space at the ends, it is pro- bable that the variation in the charge itself was not over 6 or 7°C. Temperature measurement The thennocouple first used was of iron and constantan. This combina- tion did not prove satisfacto~J for temperatures over 900°0, conseq:rently a cauple of platinum and platinum- 1070 rhodium was used for tanperatures above this point. The calibration curve supplied with the couple was used rather than a calibration made in the laboratory. -5- .,., .,_ c:: i ."'1 ~ ~ ;:g · ~ ~ C/\ ~ '.\ ~ [l) , ~ j:;; en • ),. (/\ ~ ~ r-" r,, C) ~· ::'!'~ :,,- : ~ J ~ '·~- i~~1 _ .,.. 1 . ~ 3l ~ ..... ; :r ... "11. ;:,-'!:-'-" [ i\:;t Jai o~m f ; .. .... V ~ ~ ~ .,..rC --..,. ~::.. q·-1. Jo ~ ~ ~ I t • ~~J(JI . f ii 'S.. -· " ' ~ (I) ~f~ -6- The wires were placed in quartz tubes enclosed in a piece of small seamless steel tubing; this tube was inserted into the bomb-thermoco:uple tube. A Leeds and Northrup potentiometer indicator# 36310 was used to measure the potential difference set up in the thermocouple. Figure II is an assen.bled drawing of the bomb, f'Urnace and end plugs as used in the later experiments; the various pa.rts were constru.cted as indicated in the above discussion. Gas .Analysis Since a kncmledge of the composition of the gases in equilibrium with the solid materials is necessary to compute the equilibrium constant K, an accurate means of analyzing the equilibrium gas mixtures was necessary. tire investigation the gas analysis apparatus gave trouble. Throughout the en- A description of the several methods used anl a typical calculation of data obtained by each are given below: 1. A combustion pipette containing an electrically heated platinum coil was first used for the determination of hydrogen and carbon monoxide, knovm quantities of oxygen being added to the gas mixture to enable combustion to take place. The gas was freed from moisture before each l'eading by causing it to pass through a calcium chloride tube. Carbon dioxide was determined by absorption in KOH solu- tion, and excess oxygen by means of a phosphorus pipette. considered to be all nitrogen. The residual gas was The apparatus is shown diagrammatically in Fig. III. In making an analysis the following procedure was used: Before drawing a sample into the measuring burette A, the capillary tubes and header were evacuated. This was done by the aid of a two-way stop cock on the measuring burette, which allO\'ted the burette to be used as a pump. -7- Since ercury was used in the burette, a good vacuum could be obtained in this way. ir was then drawn into the burette and the volume read. When the stop cock was urned so that the evacuated capillary tubes were connected with the burette a ortion of the air rushed in to fill the evacuated space. :1.ining in the burette was then reruh :1.e volume of the capillary. The volume of air re- The difference between the readings gave All the air was then again pamped out of the capil- 9.ry, and the gas sample drawn in and measu:red. It was then treated successive- '! with potassium hydroxide in Pip~tte :a, phosph9rus in pipette C, oxygen in !le combustion tube D, potassium hydroxide in B, and again with phosphorus in c. The analysis data and calau.lations for a typical sample are sho\m elow: DATA SHEET .ead.1ng Temp. Press • Vol. 219 urette ur & Capillary 21.9 ol. of Ca;eillary 22.4 nitial Vol. 22;7 fter KOH fter Phos. 22.8 ample 21.9 02 20.2 fter comb. fter KOH 21.e ;fter Phos, 22.1 22;5 urette ur.& Capillary 22.5 748~0 748.0 53.l 50.2 748.0 748.0 748.0 777.0 747.8 747.7 748.0 748.0 748.0 69.7 69.8 68.3 100.0 77.5 74.l 49.8 34.9 32.5 Total Vol. Stan. Vol. Contraction. 72.1 72.2 70.7 65.7 65.7 64.2 102.4 ~n.o o.o o.s 79.9 76.5 52.2 31.8 29.6 73.2 69.7 47.6 23.8 3.5 21.9 2.6 CALCULATIONS The O"Jcygan used was Of' the following ccmpos1tion: N2 CO2 02 1.39% 2.31 96.32 100.02% a) Oxygen added• 97.0- 64.2 )) CO = 32.8 cc. 32.8 x .0139 o.5 cc of N2 32.8 x .0231. o.s cc of CO2 = fonns CO2 by combustion, therefore -9- 2.2 = 3.5 - o.s 2.7 co• no. of cc of CO originally preseat since 2 c~+ 02 __...,. 2 co2 (c) 52.2 - o.5: 51.7 cc originally present. H2 was detennined by excess contraction over that due to the "6iirning of the CO preseilt~ (d) Total contraction after combustion= 23.8 oc Contraction due to CO: 2.7/2 ~ = excess contraction 22.6 cc The hydroge11 originally present was then 22.6 x 2/3 : 15.1 cc since 2 Hz+ 02 ......,. 2 H2 0 Final results: = = CO2 : O.O cc o.o % Oz: 0.5 11 o.a co : 2.7 n : 4.1 Mz =47.1 n :71.8 H2 :!5.1 " 65.4 " :; 23.O 99.7% Use of the Ca CL2 tube ma.de the calculations of vapor tension unnecessary• .Analyses Of the same sample ma.de by the above method cauld not be made to consistently check with. one another and the total percent varied from 94 to 106. This was probably due to oxidation of some of the nitrogen in. the combination pipette. 2. In a second modification of the gas analysis apparatus, the combustion pi- _pette was replaced by a tube containing powdered cupric oxide heated to 285° C after the manner described by :Burrell and Oberfell (4). Potassium hydroxide and phosphoru.s pipettes were used as in the preceding method. The arrange.111ent of the apparatu.s is sh~vn in Fig. IV. The oxidation of CO and H2 was carried out in the CuO tube A. This tube was heated by means of a small electric :furnace wound with advance wire and capable of attaining a temperature of 325° Con 25 volts. The cupric: oxide was rei1ewed, after being reduced, by passL"lg ox:ygen slowly through the heated -11- tube into the au.."'{iliary pipette B and bact: again. In this type o:f apparatus the contraction in volume of the gas sample after passine over the heated cupric oxide was taken as a direct indication of the hydrogen present. The CO in the sample was burned to CO2 without change of volume and was detennined by absorbing, in KOH, the CO2 formed. A typical calculation from data obtained by this method is given in the following table, DATA SHEET Reading Temp. Press. Vol. Eurette 21.6 21.6 748.6 748.6 42.2 39.2 Bur. & Cap. Vol. of Cap. Total Vol. Stan. Vol. Contraction. 3.0 ·2 1.6 21;6 21.6 21.6 21.6 21.6 21.6 21.6 21.6 Initial After KOH " Phos. II CuO 11 KOH " euo KOH Burette Bur. & Cap. Yol. of Cap. II 748.9 748.9 749.0 749.1 749.1 749.2 749.2 749.2 749.2 50.1 so.1 50.0 48.7 5.3 5.3 4.4 53.l 53.l 53.0 51.7 8.3 0.3 7.4 48.5 48.5 48 .4 47.2 7.6 7.6 6.8 o.o 0.1 1.2 39.6 o.o 0.0 3.0 Final Results: . o1 p - o.o. cc - o.o 0.2 - 0.1 = 2.s - 1.2 -39.6+ o.8: 83.3 14.0 - 6.8 100.0 % 48.5 cc - - Analyses of s~les of the same gas by this method gave varying results for the hydrogen and carbon monoxide content, the variation beine as high as ten per cent in some cases. came close to 100%. The total percent for each analysis, however, a~vq.ys The analyses yielded suspiciously large amounts o:f hydrogen, and since the arrangement of the apparatus necessitated cooling the cupric oxide after passing the gas through. and before reading the volume, it was thought possi- ble that some carbon dioxide was absorbed by the cooled Cllpric oxide to form cupric carbonate. This absorbed CO2 would appear in the calculated results as hydrogen, in the above method of analysis, and thus introduce considerable e rror. Table I gives the re~ilts of analyses oft.he same gas sample by the oxygen canbustion method and the cupric oxide method. The results are too irreg- ular to be of use in an accurate determination of the equilibrium constant. Es- pecially the . percentage of carbon monoxide mu.s t be detennined, as this va.Jlue is cubed in the expression :!hr the constant. A gas sample of known composition was ma.de and analyzed by the Cu.O method, but the results gave no better checks than the analyses shown in the table. Other modifications of the apparatus consisted of the substitution of a pipette containing alkaline pyrogallol solution for the pipettes containing phosphorus and potassium hydroxide. This solution removed the oxygen and ca rbon dio~ide at the same time, thus avoidi ng unnecessary operation. The final apparatus was essentially that shmrn. in Fig. IV with exand caption that the potassium h.ydroxideAphosphorus pipettes were replaced by one pipette containing allcaline pyrogallol. In future work the apparatus will be so ar- ranged that the cupric oxide tube is kei,t hot at all times, the gases o:n:rtained within it being drawn out by means of a mercury pump, and added to those in the measuring burette. The hot tube can then be cut off completely from the rest of the apparatus while the gas volume is being measured. It is hoped that this type of apparatus will eliminate aey possibility of cool euo absorbing co2 • As it is, calculations based on the results obtained thus far can be only ap- proximated and of use only in drawing general conclusions. ':J9.7 Tofa/. '::J8.b so.'CJ /2.6 I 3.<iJ 13,0 8,Z. D qa~. 6-7 £ J.. 100. / . 5'"2.€ S7o /00,b 13./ 13.4 13',5 7.6 t1ver. 'a.5 /2.3 14.7 -tube TABLE /01. 3 Sf.I 5/.3 52.0 Nz. " 100,3 13~ 13.7 VJ.::J /5.4 l~.6 13.S ?.~ C Hz I/. I co ,, 13.0 ■ 7.7 B " /3.3 Oz ,. . 7.3 CO2. A "lo van7p/e Ox<jqen V5inq C ornbv~-t,on /00,/ l~O 51.2. /2.3 (00.0 /5".5 62.7 zz. ~.I S.5 6.5 13./ B A f-v/::Jc::. 8.3';) 3.98 C. app. b,ned Com-· 100.0 11.7 58 .5" SB.'u9 16,:)7 66.67 'J. 7 / 3,'..J8 I I. I 6,0 aver C-vpr1c ox,de ' The charge Preliminary work was begun upon the question of the reversibility of the reaction by first using a charge made up by igniting sodium ferrocyanide , the reaction being- The cyanide, carbon and iron are fo:nned by this process in intimate contact with one another, and should constitute a satisfactory charge for the bomb. Experi- ence showed, however, that some other compound than cyanide was also formed from the ferrocyanide by i gnition (2). The charge used in the next fmv experiments consisted of iron, charcoal and sodium cyanide intimately ground together am placed in lqVers with plugs of steel wool separating them. - Thus it was kept porous, intimtely mixed and in su ch a coniition that a large amount of surface was exposed. After heat:- ing and passing in carbon monoxide the charge was rapidly cooled and a portion of it tested for sodi um carbonate. present. A large amount of this compound was found to be Since Bucher has shCYNn that t he reaction proceeds also in the opposite direction under proper conditions , the reversibility of' the reaction i s established. .Analyses of' the gas mixtures in equilibrium with the chart;e did not, how- ever, give sufficient- reliable data for a caluulation of' the equilibrium constant .K. The charges used in experiments following the above have been of t wo types: (1) Sodium cyanide, sodium carbonate, iron and carbon in definite proper- tions were intimately ground together, and placed with separating plugs of steel wool in the reaction chamber; (2) A. mixture similar to the above was treated with a small amount of water and passed through a meat grinc1.er , producing long pencils of the desired composition. These pencils were dried in an oven at 90 - 130 ~ C for eight hours and then transferred to a dessiccator until ready for use. Hydrolysis probably took place to some extent in accordance with the follovril1g reaction: NACli+ 2 H20 ~ NAOH + HON rather than according to the reaction. NACN +- 2 1!20 -l' NHz + NACOOH since the odor of a.nmonia could not be detected. ticeable on the surface of the pencils. anide or sodium hydroxide. A thin white coating wa.s no- This coating was probably either cy- In the event that it was the hydroxide, it is known that the relation of carbon monoxide to nitrogen in the equilibrium mixture would not be altered, since the only effect of the presence of sodium hydroxide muld be the formation of hydrogen through reaction with the carbon monoxide, .Analys is Of the Charge In future work it will be desirable to analyze the charge both before and after the run. A methOd based on suggestions by Dr. Gra.b.a.-rn Ed.gar was test- ed and found satisfactory for the accuracy desired. In this method the sample to be analyzed, oo:nsisti:ng essentially of' a mixture of sodium carbonate, sodium cyanide and sodium hydroxide, with a possible small amount of sodium, is first treated with a small amount of water cooler than 30°C (to avoid the formation of R4~e(CN)5) in an evacuated flask which is connected to the gas analysis apparatus. .4ny sodium present will react with the water to form hydrogen, the volume of which is measured in the gas analysis apparatus. From this volume the anount of sodium originally present can be calculated. The aqueous solution obtained after filtering off the iron and carbon is then treated with an excess of' o.5 N Ba (N03) 2 solution and filtered. It is best to perform the filtration under a bell jar containing air free from co2 and to wash the precipitate with freshly boiled water. -15" This step precipitates all the carbonate as Ba C03 acoording to the reaction: Na.2C03 + Ba{lI03) 2 ~ J3aC03 The solubility of Ba.~ 3 is 1.86 x 10- 3 2 Nrum3 • grs. in 100 grs. of solution at 16°C {5). Si,nce l arge excess of water was avoided in the first step it is apparent t.ba,t practically all the carbonate present will be precipitated. The filtrate obtained from the Ba(H03 )2 treatment is diluted to 500 cc and a 50 co sample titrated with a standard solution of 0.1 N.Agwo3 until a permanent turbidity is formed. This point is best observed by placing the contain- er over a piece of blaclr glazed paper. The burotte is then read. The reaction which takes place result s in the production of the soluble Na.Ag(CN) 2 according to the equation: 2 I-ra.CN+ .AgU03 ~ Na.Ag(CN} 2 -t- Na.N03 • As soon as all the {CN)- ions are used up in prodttc:t.llz Ma..Ag(Cn) 2 the following reaction takes place upon the addition of more AgN03 • At this point the permanent turbidity appears . AgCN is soluble to the extent of 2.2 x 10 solution (5). -5 grams per 100 gr ams of Na.Ag{C.N) 2 in solution however produces fewer Ag+ ions t han AgCN in solution, consequ.ently no .AgClT will be precipitat ed until al l the(ClT}ions present have r eacted to form Na.Ag(at0 2 • As soon as thi s c onvers ion i s cllnplete an excess of .Ag+ ions from the AgN0 3 will cause the precipitation of the .AgCN. The amount of AgN03 required t o produce the :first s ign of a precipitate of AgCJ:{ affords a basis for calculation of the amount of (CH)originally present, one .Ag+ being equivalent to 2 (CN}-. All of the Na...4g{CH) 2 is t hen converted into insoluble AgCN by adding an amount of .AgN03 equal to the amount requi red to produce the first tu:r- bidity. Further addition of Ag:l!03 causes the follovring reaction to take place: 2 NaOH + 2 .AgN03 ~ Ag20 + Na.N03--t-H2 o. Enough .AgN03 is added t o insure the complete precipitation of the hydroxide as This compound. is soluble to the extent Of 2.15 x 10•3 grs. in the brown Ag20• 100 grs. of solution (5) and is consequently practically all precipitated. The solution containing the precipitated Ae 2o and AgCN is then treated with standard 0.2 N HN03 solution until the brown color disappears, ani the amount used is· recorded. The Ag20 goes into solution according to the following reaction: A.g20 -- + 2 mro ~A.gN03 + H20• 3 The cyanide, however, does not dissolve as it is only slightly soluble in d'ilute HN0 3 solution. The solution is then filtered to renove the cyanide, and titrated back with standard 0.1 N :naOH solution to the point of the reapP3arance of a br0vm color. In this way the exact amount of acid required to react with the Ag2o is found, anl in consequence the amount of Na.OH originally present is calculable. Since this last end point is not distinct the reading is taken at the point of the first appearance of the brown color. The precipitated Ba.003 from the Ba.(N03 ) treatment is dissolved in 2 standard IDT03 solution and the amount used recorded. The solution is then boiled for five minutes and titrated back with standard HaOH solution, using phenolphthalein as an indicator. The first reaction between the carb~aa.te and acid i s as follows: BaC0 3 + 2 I:Il.lJ0 3 -4--Ba(N03} -t- H2C03 • 2 This reaction is caused to go completely to the right by boiling the solution and decomposing the H2co3 formed. Knowing the acid equivalent of the NaOH solution the amount of' m.ro3 required. to react with the carbonate can be found. This then enables the calculation of the amount of carbonate origi nally present. A sample taken from the charge after a run was analyzed according to t he a·bove system of analysis and gave the following results: DATA ~TD Cl!.l,CUL.ATIONS 15.5 grs. Sample taken - l.Tre?,ted the sample with cold water- no hydrogen was evolved. 2.Added 75 c.c of 1 N Ba.{N03 )2 solution, filtered anl washed. . 3.Made up the filtrate to 500 c.c and titrated. 50 c.c of this solution with .05 N AgNO"v solution. . • Then 10 X (2 X ,0463) X 49 X .05: 14.6% NaCl[ 15,5 Added an excess of AgN03 solution. Added Oe2113 N HN0 3 solution • .Amount used: 16.73 c.c. Filtered a:nd titrated back with NaOH solution, 1 c.c = 0.4982 c.c of mro solution. • 3 .Amount used: 19.60 CoCe 16.73 - (19.60 x o.4982): 6•96 c.c of acid used. Then 10 x 0.00696 x 40 x 0.2115 : 3.62 %of ]faOH. 15.5 • 4. Added m:ro 3 , 1 c.c = .0112 grams of Na.2co 3 , to the l3a{N03 } pre cipitate • .Amou.Jit used:: 153.59 c.c. Boiled and titrated excess with NaOH, l c.c :: 0,4982 c.c of mro 3 solution. 153.59 - (93.80 x o4982}:: 106.86 c.c of acid used in reacti11g with the :saco3 • This amount is equivalent to 106.86 c.c of 0e2115 N Na2co • 3 Then 53 x 0.2115 x 0.1069: 7.73% NA.2COr. 15 5 3 • Analysis of the sample:- NaCN = 14,6%, NaOH: 3.62%, NaC0 3 : 7.73%. A duplicate a.'Ylalys is of the above sample gave ch.eeks within o.5 %for the MaOR , and within o.3% for the carbonate and cyanide. Entering gases The gases passed into the reaction chamber consisted essentially of a mixture of carbon monoxide and nitrogen. The first p~riments were car- ried out using practically pure (over 95%) carbon monoxide. This, of course, necessitated the consumption of a large amount of the gas before the equilibrium mixture of carbon monoxide and nitrogen was attained. In later ex- periments, basing the proportions on the analyses of the gases obtained in the equilibrium samples, it was found advisable to use a mixture of nearly equal am.ou:hts of carbon monoxide and liitrog-en. :Ebcperimeni; s in which near- ly pure carbon monoxide was admitted to, the bomb, and others using a mixture of equal amounts of carbon monoxide and nitrogen would than enable the eq,uili brium of the reaction to be approached both from the nitrogen and the car- bon monoxide side; for it was found that the nitrogen ;n the equilibrium gas mixture a.t about 1000°0 wasr.considerably less than 50,%, The general arrangE!llent of the apparatus for a typical rur. is mown in Fig• .v. In this figure, A is a container for the entering gas- eous mixture, and is equipped with a manometer. enabled the The two-way stop cock B use of any other gas without disconnecting the container. Two of the towers shown at O contain alkaline pyrogallol, and the other two concentrated sulphuric acid; these free the entering gas from carbon dioxide, oxygen and moisture. The stop cook D enables the rinsing 0£ the washing tcmers with the gas be£ore starting a ru.n. The pressure in the reaction chamber within the ru.rnace E is measured by the mercury manometer G, and desired vacuum being attained by means of a pump connected at F. Hand I are the sample tubes and gas analysis apparatus respecuively. A typical run consists of the following operations: 1 • .Mixing the cllarge to be used in the bomb and filling the -19- F19ure. Ya . r N () -zo- gas reservoir with the proper gas mixture. 2. Filling and welding the reaction chamber w ith subsequent tests for leaks throughout. 3. Pumping out of the bomb the gases which a.re formed during t he fiv e or six hour heating ~riod, in which time the temperature rises from room temperature to about 1000 9 C. The proportions of nitrogen and carbon monoxide in this gas depend on the proportions Of' cyanide and carbonate present in the original charge. 4. Admitting the gas mixture to the bomb until equilibrium at the desired tempera. tu.re is .a ttained. 5. Drawing the samples from the equilibrium mixture. methods, were employed. Tw o general First, the floo method, which required from twenty to forty- five minutes, in which the gases are admitted very slowly at slightly over atmospheric pressure and drawn from the bomb at practically the same pressure, thus insuring the format ion of an equilibrium mixture. This method was finally abandoned as the concentrations of the CO and N2 gases were thought to change as they gradually passed from a white heat to the cooler end of the bomb. To obviate this difficulty a second method was devised. The charge was allowed to come t o equilibrium with the gases pres ent in the bomb. The bomb was then cut off from the entering gas by closing the stop cock E and the gases within 1 . the 1,!omb were allcmed to escape into the previously evacuated sample tubes H by quickly opuning the stop cock ~. This stop cock was then closed, and E1 opened slowly, prepa.rato~J to taking another sample. The equilibrium gas mixture samples ~ere obtained in about 2 to 3 seconds in this manner and probably more nearly represented the true equilibrium mixture. 6. The analysis o:f the equ.ilibrium gas samples,as soon as possible after taking, to enable a better control of the ran and te lessen the possibility of contamination by leakage. T.be average time consumed in making an analysis with the cupric oxide tube apparatus was about thrae quarters of an hour. It was i.rnposs ible to make runs longer than 40 to 48 hours; at the end of this time the iron bomb usually oxidized through and developed lealrn. The results relating to the determination of the equilibrium constant are tabulated in Table II. Ela.ch r".tn is separately shown and the type of' analysis apparatus indicated. Examination o:f the table reveals the fact that the discrepancies in the gas analyses are large, and that in many cases accuracy su.fficient for a useful determination of the value of K was not obtained. Experiment IV gave probably the best results of the series, and upon the da.ta obtained in this experiment the calcv.lation of the equilibrium constant and other data is based, not with the idea of establishing accurat~ values, but more to obtain a general knowledge of the prevailine conditions. Calculations The derivation of the heat of reaction is as foll~vs: From the data recorded in Eicp. IV, the table given belov1 is obtained: - Tue 973 989 1008 1010 1010 1012 1016 1016 .1£ 9~47 X 10 It 19.7 32.5 " 41.3 " ti 31.9 45.0 "It 64.8 l?glO K 3 29e5 -22- 3e9763 4.2945 4e5119 4.6160 4.5038 4.6532 4.8116 4.4698 '. f= x;en'rnent . .. e,harqe • rn/x-tvre •. 4 qr:, . -1- .. (6 ,, 16 ·~ r-l~CN C Nat.C03 Ft! llllm- _· 111■1111■~-~.~ ..... 1■111111 - , ·.Ill ~-- " ,. . ~: ■ -· ~ -•·· == , ■ ►-- ■It : .. f ' .' .. ·, . II■ : - =~=== :• . ; . :■ ~ : . . . . . •. CC N, S3.7 ~o L:./-4.~ '7,:;, C..01. ~ 02. O.~ ">o Hz. f. f ">o The values of T and log K are plotted in Fig. VI. They may be assumed 10 with an accuracy equal to that of the data, to lie on the straight line sh~vn. Assuming thls linear relation, the slope of the curve will be found from the plot to be 0.0176. substituting this value fo:r in the equation d loglOK : ~ dT RT 2 .6H = o.0176 x 2.303 ~ 1.98 x (1273) : 130,000 calories d logJ{ : dT gives d log K 10 dT 2.303 for the heat change Of the reaction From Landolt-Bornstein-Roth Tabellen, pgs. 859 and 855, the heats of foi-mation of NaCN, CO, and Na2 c~3 from their elements are given as 22,600, 29,000 and 270,800 respectively. From these data the heat change of the Bucher reaction at room temperature can be calculated as H: 270,800 - 2(22600} - 3(29,000): 138,600 cal., a value in fair agreement with that given above. The agreement becomes still better if account is taken, in the following manner,, of the fact that one value refers to room temperature and the other to lOOO"C. L . The molal heat capa.ci ty of N~C0 3 is given by Landolt Bomst ein, pg. 758, as 0.27 x 106-: 29 cal. per degree. The molal heat capacity of :NamI is not given but may be calculated by the fonµala given by Noyes ( 7) to be MC PNaCN The molal heat ca1%1,cities of carbon, nitrogen and carbon monoxide at " high temperatures are given in Landolt-Bornstein, pgs. 751 and 773 as 4.3, 3.6 and 6.8 respectively. Na C0 + 2 3 The change in molal heat capacity when the reaction n2 + 4 C "'-➔ 2 NaCN ~ 3 CO occurs will therefore be 29. + 4(4.3} - 2{14.2) - 3(6.8} = 4.2 cal. 2(3.6~ The heat content decrease attending the· Deqvees Cenh9rqd e . I I I i ---!----' I ._I ____ . -- 1 .- _J I- -- __ , ------ I 1- i I . - - __ i - the react ion at 1000°C \Vill therefore be approximately 4.2 x 1000 :: 4200 calories less at 1000°C than at room temperature, if no account is ta.ken of the latent heat of fusion of sodium carbonate ro1d sodium cyanide. The value, ,, -138,600 cal. from Landolt-Bornstein should on these assumptions be decreased by 4,200 cal., giving - 134,400 cal. as the heat content decrease at l000QC. This figure compares favorably with that {-l-30,000 cal.) derived from the experimental data of this thesis. The mean value, -132 ,000 cal., m.ay be assumed provisionall y for the heat content decrease near l000°C. Using th~s value the r~tio of the equilibrium constant¾ at any absolute te~perature T, to the constant K at 1273°A, may be calculated, by use of the equation 1273 log JSt1 Ll H I{ 2.303R T,, /;. 4 Talcing K 1000 as 2.9 x 10 (see plot Fig.VII) gives 1z2,ooo 100~ K 10 T:: (2.3} (l.98) 1273 T-1273 ,t- log K - 22.8(1- ~J+4.45 1273.T 1273 - The use of this fo~la, assuming the heat of reaction to renain constant over the tenperature range, gives the following values for Kat various temperatures: --1 oc -1L 1.62 24.6 321. 3290. 800 850 900 950 1000 1050 1100 29,000. 206,000 1,294,000 -25- SU.MM.ARY In this them:.s has been presented a description of the apparatus and methods employed in a determination of the equilibrium conditions of the Bucher perature. The results obta:lned, while not entirely satisf'actory, show that the equilibrium constant K of the reaction is about 1.6 at 800°0, 320 at 900°C; 4 6 3 x 10 at 1000°0, and 1.2 x 10 at 1100°0; and that the heat content decrease when the reaction goes toward the production of sodium cyanide is 132,000 .:=l::=.2000 calories. REFERENCES Jolm E. Bucher, Jor.Ind.Eng. Chem. Vol. 9, No. 3 (1) Fixation of Nitrogen. (2) Wie Equilibrium in a Process for the Production of Sodiu.rn Cya.~ide, Paul De Vries lYanning, Throop College of Technology. Thesis 1917. (3) Prc:duction of Temperat ure Uniformity in an Electric Fu.mace. Bulletin of Bureau of Standards, Vol. 10, no. 4. (4) Use of Copper Oxide in Fractionation combustion of Hydrogen and carbon monoxide in gas mictures. Burell and. Oberfell. Jor.Ina.• Eng.Chem. Vol. 8, P• 228. (5) " Landolt-Bornstein-Roth Tabellen. (6) Dictionary of Chemical Solubilities. (7) Theoretical Chemistry. A. M. Corney. Noyes and Sherr.ill. A. W. Gray.