<<

THE DEHYDROGENATIQN OP TO PRODUCE AND , USING AS A DEHYDROGENATIQ N AGENT

DISSERTATION

Presented in partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

by

Ralph Eugsene Morningstar, B.S* !\ The Ohio State University

1952

Approved by ACKNOWLEDGEMENT

Tlae author expresses h.is sincere appreciation for the advice and counsel of Dr. L. K. Herndon, not only in con­ nection with, this thesis, hut also for his guidance and personal interest in other phases of the author's profess­ ional training.

t

S&OS 9 8 TABLE OF CONTENTS

Indications page i Introduction 1 Related Literature 3 Discussion of the Literature 53 Thermodynamic Calculations 55 Construction and Operation of Equipment 68 Operation of Vaporiser 71 Calibrations 79 Analytical Methods 89 Experimental 110 Effect of Temperature and Sulfur Concentration 111 of Ethylene by Sulfur 128 Effect of Space Velocity on Reaction of Ethane and Sulfur 133 XTse of Catalysts for the Reaction of Ethane and Sulfur 137 Temperature Gradient through Reactor 146 Discussion of the Data 149 Conclusions 163 Autobiography 168

• • • \

■■// ■■ ^ THE. DEHYDRO DERATION OF ETHANE TO PRODUCE ETHYLENE AMD ACETYLENE, USING SULFUR AS A DEHYDROGENATION AGENT

Indications: . - 1 . Apparatus was cona truciied for the purpose of studying the reaction of ethane and sulfur . This apparatus consisted of two parts: a, A vaporizer for the generation of sulfur vapor. Ethane was bubbled through a hath of molten sulfur at a high temperature in order to obtain the desired mixture, \ b, A reactor furnace consisting of a porcelain tube in. a high temperature fuf- nace. This furnace was capable of tempera­ tures up to 20G 0°F, 2 . No appreciable dehydrogenation of ethane occurr­ ed in passage through the sulfur vaporizer. Such reaction as did occur led to the formation of acetylene, with no evidence of ethylene formation, ^ 3* The conversion of ethane to ethylene was found to be a function of temperature, with a maximum conver­ sion of 50^ being obtained, at l600°F. in the absence of sulfur vapor,

VI '■ 14^ 14-. The conversion of ethane to ethylene decreased with the addition of sulfur vapor to the reaction mix­ ture* Thus at 1565°F. and a ratio of sulfur:ethane of l.ij., a conversion of 357° was obtaihe d, compared to ij.6.57° in the absence of sulfur. This was due to further de­ of the ethylen e. 5 * The conversion of ethane to acetylene increased with an increase in the sulfur vapor content over the range studied. A maximum conversion of ^0% was obtained at a sulfur-ethane ratio of 1.14. and a temperature of 1565°F. 6. The conversion of ethane to acetylene was found to be independent of temperature except at high tempera­ tures and high sulfhr:ethane ratios. Under these con­ ditions the conversion decreased with increasing tem­ po rature. This effect was attributed to dehydrogenation to c arbon• 7. The conversion of ethylene to acetylene was found to be less efficient than conversion of ethane to acetylene. At 1560°F. and a sulfur :e thane ratio of 1 .1+, a conversion of 30% was observed. 8. In the range of contact times in the reactor of 1 to 6 seconds, no significant trent in the dehydrogena- tion of ethane by sulfur was observed. s . ; . i

9. Lump pumice, Porocel, and chromium on activated alumina were investigated as catalysts for tlie reaction of ethane and sulfur. No significant increase in conversion occurred except in the case of lump pumice at lIj_50oP. The conversion obtained in this case was no "better than that obtained in the absence of catalyst at l600°P. 10 . In general, the trends observed in this study corresponded to those predicted by thermodynamic cal­ culations. No significance was attributed to the rela­ tion between actual and calculated conversions because of uncertainties in the thermodynamic data used. ,

I XNTRQDUGTIOH: .. The importance of both ethylene and acetylene as raw materials for chemical manufacture cannot he overemphasized. Both these are obtainable from natural * both and higher gases, by known means. In the case of acetylene, however, the familiar carbide route is still gen­ erally employed. For a number of reasons the carbide process is preferred, most particularly because of the greater safety and relative ease of transportation of solid . With the increasing trend to locate chemical manufactur­ ing near sources or on supply lines, the transportation factor is less important, and the use of nat­ ural gas as a raw material is rapidly expanding. In general, two methods of manufacture are employed. The first and most widely described, usually for the manu­ facture of acetylene, involves the and reforming of ranging from methane to light oils. Electric discharge is often used to supply the required energy. In the second method, controlled oxidation is used to form unsaturates, usually from an ethane rich feed gas. Various means of con­ trolling the reaction are used. o \ : ^=9 It is the object of the present investigation to study a different type of oxidation process, the oxidation, or de- hydrogenation, of ethane by means of sulfur* With this milder oxidizing agent, the f o m tion of unsaturates in good yield and In a readily recoverable form should he obtained. By­ product will be oxidized to sulfur and re­ turned to the system* RELATED LITERATURE ?. Both ethylene and acetylene are finding increasingly wide -use as chemical raw materials. Smith and Holliman

Harold M. Smith and W, C. Holliman, ’'Utilization of Natural Gas for Chemical Products", 1*0 , 7547* S. Bureau of Mines, (19470, chart 7.

summarize some of the uses for ethylene: ’’Ethylene is used as a fuel for cutting and , as an anesthetic, as a refrigerant and as an accelerator for growth sod food ripening. In addition, it is an Intermediate in the manufacture of the following items with uses as indi­ cated: acenaphthene (dyestuff intermediate), and anhydride, acetylene, alkylates (motor and aviation gasoline), , , , butylene, diisopropyl (high aa ti-knock motor fuel), ethyl alcohol, , ethyl ether (anesthetic, solvent), ethyl halides and polyhalides, ethyl mercaptan, ethylsulfhric acid and diethylsulfate (ethyl alcohol Intermediate) ethyl and ethylene nitro compounds, ethylene chlorohydrin (disinfectant, insecticide, solvent), ethylene halides, , , - , , Intermediates for plastics and resins, , , neohexane (high anti-knock motor fuel), oxalic acid, polymers of ethylene (synthetic rubbers and plastics, lubricants and additives), styrene, , vinyl and vinylidene chloride (textiles and plastics), and a host of others”. Smith and Holliman ~

Ibid, chart 1+

also present a comprehensive summary of the general methods of the preparation of hydro­ materials. nl. ( thermal, catalytic, electric).; The splitting of the hydrocarbon into smaller or into carbon and. hydrogen by heat alone (pyrolysis), with the aid of catalysts, or by electric discharge, generally accompanied, especially in pyrolysis, by recombination, of some of the. products into new compounds. 2 . Oxidation (thermal, catalytic); reaction of the hydrocarbon molecule with , air , or oxygen-containing compounds, activated by heat or catalyst, whereby oxygen is introduced into the hydrocarbon molecule, or the molecule is changed to and hydrogen, and hydrogen, ca? finally to carbon dioxide and . £!?. {3 3. (thermal, catalytic, pfeotolytic): Reaction of the hydrocarbon molecule with a halogen (, , bromine, ) activated by heat, light or cata­ lyst, whereby one or more halogen atoms replace an; equivalent number of hydrogen atoms in the hydrocarbon molecule. Ij.. Nitration (thermal, vapor phase): Reaction of the hydrocarbon molecule with nitric acid, accelerated by heat and pressure, whereby a nitro group, NOg, is introduced into the hydrocarbon by replacing a hydrogen atom. Rs Sulfurisafelon : Reaction o f th© hydrocarbon molecule with sulfur or to form sulfur-containing compounds such as organic , mercaptans, , thiophenee. * 6. Desulfurization (catalytic): Removal of the sulfur atom from a sulfur-carbon-hydrogen molecule to-f&rs-e-sulfur- free molecule. Activated by catalyst. 7. Hydrogenation (Catalytic, thermal catalytic): Ad­ dition of hydrogen atoms to an mole­ cule to produce one or more saturated molecules. Activated by catalyst or heat and catalyst. Known as destructive hydro­ genation when original unsaturated molecule is "cracked" to form more than one subsequently hydrogenated smaller molecule, or non-destructive when no cracking occurs. 8* Dehydrogenation (catalytic, thermal): A farm of con­ trolled decomposition whereby hydrogen atoms are removed from hydrocarbon molecules to form less highly saturated molecules. Known as destructive dehydrogenation when original' molecule is "cracked" to form more than one smaller molecule, or non-dest­ ructive when no breaking

Benjamin T. Brooks, "The Chemistry of the Nonbenzenoid Hydro­ ", Second Edition, Reinhold Publishing Corp., New York

(1950) pp k 70 ff.

"The chemical utilization of acetyl one began in Germany in 1910 where chlorinated solvents, tetrachlorethane and tri- chlorethylene began to be manufactured, followed shortly thereafter by the production of by reaction with water, by means of a catalyst in solu­ tion. This was quicklybfollowed by the of acetaldehyde to acetic acid and the manufacture of and acetic anhydride. Industrial acetylene chemistry started in Canada in 19ll|. and in the United States a few years later. Thus, for fourteen years after its commercial production on a large scale, acetylene, like pe tr oleum oils and ole fin-rich oil gas which had beeiy&bundantly available for a much longer time, was regarded only as a special fuel and not as a raw material for valuable chemical syntheses. "Research, in the field of acetylene chemistry was greatly expanded'during both World Wars I and II, as will be indicated in the following pages - - - "Methods of Preparation. Acetylene is made industrially only from calcium carbide., although its formation from methane and other hydrocarbons by thermal decomposition at high tem­ peratures (q.v.) continues to receive serious consideration. When ethane, or other paraffins, or petroleum oils are employ­ ed, the ethylene which is also produced aids the economics of the process." It is beyond the scope of this paper to describe in even very brief detail the present state of acetylene utility. A few brief statements by Brooks

Ibid, pp. I4.7i4.-I1.88

will give an indication of its varied applications. "The polymerIsationlof acetylene contrasts markedly with the polymerization of ethylene and vinyl compounds, which form linear polymers of high molecular weight. Ethylene has no tendency to form cyclic hydrocarbons (cycloparaffins) at 377- 393°• Acetylene, however, under thermal conditions, gives benzene (and other hydrocarbons) as noted above, and under other conditions gives 1he solid substance, or substances, v cupreae, and under special catalytic conditions, gives good yields of vinyl acetylene, divinyl acetylene, and the very interesting product cyclo-octatetrene •--- "Vinyl acetylene H-C£C-CH»CH2 is a low-boiling colorless liquid, 5*5°^ • - - - It reacts with, , in the presence of many metallic chlorides, by 1-Ij. additions to form first 1-chloro-2,3-butadiene, CH201-CHC5GH / HC1 — CH2C1-CH=C-CH2 The latter product rearranges in tie presence of cuprous salts to 2-chloro-l,3“butadiene or ""•-- "Hydropolymerization* Under the conditions of hydro- polymerization, or polymerization at elevated temperatures with hydrogen and a catalyst, a series of branched chain hexenes, hexadienea, and octenes is formed*-- "The hydrogen atoms in acetylene are very reactive in much the same way as the hydrogen atoms in the CH2 group in cyclopentadiene, reacting readily to form sodium compounds, displacing hydrogen, and reacting with G-rignard alkyl mag­ nesium halides* -- "The of under some conditions Is not quite as reactive to addition reactions as the ethylene bond. However, acetylene may be catalytically hydrogenated to ethylene with good yields* -- "The addition of hydrogen chloride to acetylene to form is favored by cuprous chloride, mercuric chlo­ ride, and other catalysts. --- ; A..'J •’ adds to acetylenes at JL|_0° to 70°C to give good yields of additive products. --- ’’The chlorination of acetylene is carried out industrial­ ly for the production of tetrachlorethane. -- •’ results from the addition of one mole of acetic acid and the addition of a second mole gives ethylidene diacetate. Vinyl acetate, with vinyl chloride, goes into the manufacture of the well known vinyl plastics and ethyldene di­ acetate is an intermediate in one of the industrial processes for making acetic anhydride. * ‘'" is more readily made hy passing acetylene into a mixture of arsenious chloride and a solution of mercuric chloride in either 18% hydrochloric acid or saturated solution. T As indicated above, ethylene and acetylene may be pro­ duced by dehydrogenation of paraffins, such as ethane, by either catalytic or thermal processes. Such processes are widely used in commercial applications at present. A discussion of the cat­ alytic processes is given by C-rosse and Ipatieff.

Aristid V. G-rosse and V. N • Ipatieff, Ind. Eng. Chem. J2. t 268-72 (19l|.0)

"The advantages of the catalytic process over purely thermal cracking.are the much larger' overall yields of the corresponding olefins and a much, higher reaction velocity. This is demonstrated for normal, and isobutane in Table I. TABLE I COMPARISON OF CATALYTIC AND THERMAL TREATMENT Temp. Press. Contact Conver- Concn. of Production iig./Kg./ Time sion of ox Cj.Hq in Rate of O}C. r i sq.cm. n w i <3 Sec. o n r>v, Charge ^ TO ot/. Exit Q Gas %of nCj.HR i T.TO Vol, TT_ n S- Catalytic for n- 600 1 2 39 39 152 600 1 1 30 2k 230 600 1 O.5 20 17 310 Thermal for n-Butane 600 1 77 50 3.6 0.7I+ 600 7 36 50 3-5 7-7 650 1 11 50 3.0 3.5 650 7 9 50 It-.o 40.0 Catalytic for Isobutane 600 1 2 I4.I 31 156 600 1 1 31 25 239 600 1 0,5 2 0 17 310 Thermal for Isobutane 650 1 12 50 18 2 4 . 1 650 7 12 50 3 4 14-5 a Volumes of olefin (at 25° C . and 750 mm.) per volume of Re­ action space per hour. The thermal results were obtained from an article by Egloff, Thomas and Linn. Even though the thermal results may be somewhat improved by the use of higher pressures and tern- r * 9 peratures, tlie nonselectivity of the thermal reaction always causes them to lag behind those obtained with catalysts* "A suitable catalyst for an economic process must fulfill the following requirements: 1 * It must have the ability to split off hydrogen selectively without cleavage of the carbon-to-carbon bond. This is a se­ vere requirement since the energies of the bonds (Pauling's values ) Linkage Gal*/Mole Electron Volts C-H 87,300 3*70 c-c 58,600 2.54 involved, favor the carbon-to-carbon cleavage. It is in­ tensified by the fact that high temperatures of the order of 500-750°C. must be used, since at lower temperatures the equilibrium lies almost completely on the paraffin side-— and the reaction velocities are low. 2 . It must be easily regenerated when fouled by a carbon de­ posit. 3 * It must have a useful of many hundreds of hours. 1|. It must be low priced. "These requirements are fulfilled by selected solid catalysts containing minor molar proportions of the of the transition metals of the VI (e.g., Chromium and ), V (e.g. vanadium) and IV (e.g. titanium and cerium) groups of th.© periodic system, supported on carriers of relatively low catalytic activity (e.g. aluminum and oxides). "These catalysts are highly selective and direct the conversion in accordance with the general hehydrogenation equation,

CnIf e / 2 — > QnH2n / H2 under proper operating conditions. The scission of the carbon- to-carbon bond, leading to the formation of carbon, methane, and other degradation products, has been almost completely suppressed." A description of the thermal process for ethylene production is given by Schutt*

H. C. Schutt, Chem. Eng. Prog. lt2.No. 5, 103-16, (19I4.7)

"The production of ethylene from light, saturated hydro­ carbons involves essentially two phases of operation or unit processes* the conversion of the charge by pyrolysis and the separation of the conversion product. A simplified flow diagram of such a unit is shown in Figure 1 which illustrates specifically the arrangement necessary for practicing re­ cycling operation. "Fresh feed is combined with a selected fraction of the pyrolysis coil effluent and charged ;bo the furnace where it is I f -■ :• j .-'-.x heated to the proper temperature at which the desired or maxi­ mum obtainable conversion takes place simultaneously with the attainment of the coil exit temp eratu re. A quench system is devised to reduce-rapidly the temperature of the reactant at the coil outlet and subsequently cool it completely to as low a temperature as is conveniently and economically attainable with prevailing cooling water conditions. The quenching med­ ium may be either oil or water circulated in a closed system from which the heat is removed by direct heat interchange or cooling. If water quench is employed, no effective heat interchange or recovery is possible. "The cooled conversion products are then compressed, dehydrated and separated by straight fractionation into a residue gas stream of .hydrogen and methane, ethylene product stream and a selected* fraction suitable for recycling. A liqu­ id fraction Is withdrawn containing the higher molecular weight hydrocarbons formed during the reaction as well as some light­ er unsaturated hydrocarbons which cannot be recycled without endangering the continuity of the pyrolysis coil operation. "The low temperature fractionation method for separa­ ting the conversion products has been chosen for this illus­ tration of the recycling pyrolysis unit as it offers great flexibility for high ethylene and recycle recovery, both con­ ducive to maximum yield and production of ethylene from a given quantity of available feed s to*ck. Of course, other m m j & z __ M M ntOWQT

QUENCH SYSTEM.

RECYCLE ACCUMULATOR

^ “"QUENCH MEDIUM ACCUMULATOR RECYCLE

Fig* 1. Flow Diagram of Che*. Dept*, Ohio State University Pyrolysis Unit Chew* 950 Schutt} H»C*| Chew* Eng( July 12, 1952, Eugene Momingstar

P u t * No. I f -f o ... :■ ,j„ O methods of light hydrocarbon separation, such as selective and nonselactive absorption or charcoal adsorption as primary steps, each combined with several fractionation steps, may effect the desired segregation of the conversion product and enable the successful practice of recycling pyrolysis operation# Prac­ tical process and engineering design aspects of various light hydrocarbon separation methods applying to ethylene recovery have been discussed# This phase of ethylene plant design and operation is mentioned here only to define the recycle stream and emphasize the importai ce of its recovery in maximum quan­ tity and the correct constituency with respect to ethylene yield and pyrolysis coil operability. '•typical commercial plant performance data are given in Tables 2 , 3* and 1^, showing variation in fresh feed composi­ tion and the range and complexity of the pyrolysis coil feed composition which may be encountered in recycling operation# Table 2 applies to a unit operating essentially on an ethane-propane fresh charge# The conversion per pass is moderate, and the stream separation is such that neither maxi­ mum ethylene nor recycle recovery is achieved. In fact,the data apply to the initial run of a new installation where at­ tainment of ultimate production capacity usually is not attempt­ ed immediately -- "A small amount of steam is admixed with the charge in all cases, its effect upon the conversion appearing to be • a r°i negligible. The principal reason for the use of steam is the protection it affords if the charge contains small amounts of hydrogen sulfide or other impurities which may be present in the original feed or have passed through a treating system. Without the presence of steam, sulfur corrosion may be acute under operating temperatures pre­ vailing in the pyrolysis coil, especially in sections where no carbon scale forms on the tube wall. The use of steam has been made a general practice in pyrolysis coil design and operation, even when the feed stock is absolutely sulfur-free. "The pyrolysis coil3 are designed for a continuously rising temperature gradient up to tine outlet and no con­ stant temperatures soaking tubes are employed. The reaction must be completed when the outlet temperature is reached and changes in degree of conversion are effected simply by raising or lowering the coil outlet temperature. In cases where the heat of reaction is relatively low, as in pro­ cessing a highly unsaturated or high molecular weight feed stock, the coil is so arranged that the tubes near the outlet are located in a zone of relatively low heat inten­ sity or partly shielded from the source of intense heat. The overall conversion reaction in the pyrolysis coil, as well as the progressing reaction at any stage of conversion of the reactant, is endothermic. ’’T h e development of die process cycle for recycling operation and the design of the pyrolysis coil proper in­ volve the application of numerous chemical and mechanical engineering principles, each of which could form the subject of extensive experimental exploration and scientific analysis for this particular application* Inasmuch as the theme of this paper is the production of ethylene and is of "broad scope, it is intended to cover the "basic process information, important design criteria and operating exper­ ience which haye led to a practical and expedient design method for commercial pyrolysis units ---- "Consideration of heat-transfer rates, pressure drop and surface to volume ratio of tubes of various sizes to the conclusion thgit an elevated pressure of from 50 to IpO 3bs./sq. In* gauge, and a reaction time in the order of 0*7 to 1*3 sec* will prevail In commercial pyro­ lysis coils. Experimental data available in the litera­ ture covering these conditions are relatively scarce and not very specific In the definition of reaction time. The applicable literature information has been amplified by specific experimental tests to establish the product distribution for the pyrolysis of ethane, propane, pro­ pylene and butylene, components of the reaction coil charge ■when producing ethylene from an ethane-propane mixture in recycling operation- The analysis of the experimental data has furnished the apparent reaction rates of these hydrocarbons and plant operation established conversions of ethane and propylene relative to propane, the key constituent of die reactant mixture- A general proced­ ure for establishing the recycle, the pyrolysis coil charge and reaction products has been described- The method used in stepwise analysis of the reactant flow through the cracking coil has been given* The com­ mercial units designed on thi s basis have shown per­ formance characteristics in close accord with the design conditions.” The literature on the reactions of light hydro­ carbons (ethane) with sulfur is not extensive- Ellis

Ellis, Carleton, ”The Chemistry of Petroleum Derivatives" The Chemical Catalogue Co., New York (193^1-) PP • ^-28 ff- gives a summary of the types of reaction that may occur between sulfur and hydrocarbons, and the reaction pro­ ducts to be expected* "Elemental sulfur may exist as such in crude petro­ leum or may be formed by the decomposition of complex organic sulfur compounds during the distillation processes. At such, temperatures, or at higher temperatures as in cracking operations, elemental sulfur itself reacts with petroleum hydrocarbons to yield various sulfur compounds, A review of the possible reactions which may take place should therefore be of interest. "The paraffin hydrocarbons of lower molecular weights generally react very slowly, if at all, with sulfur. Spanier has shown that n-hexane and sulfur do not react appreciably when heated for 2lp hours at 210°.0, A t 300° C s*id over longperiods of heating, n-butane and n-heptane yielded, with sulfur, some and thio- phene derivative's, respectively. At high©*' temperatures, however, reaction does take place as Fletcher, Wheeler, and McAulay have found that me thane at 1100°C. reacts with sulfur to form carbon . They also stated that at such high temperatures, hydrogen sulfide behaves in the same way as sulfur vapors.*1 Between the two extremes of no reaction and com­ plete degradation of the hydrocarbon to carbon disulfide as given by Ellis, there should be an intermediate react­ ion level to yield partially oxidized (dehydrogenated) products. Rasmussen et al describe the type of products s»« that might be obtained:

Rasmuss^a-, H« E., Hansford, R. 0 ., . and Sachanen, A. N., "Reactions of Aliphatic Hydrocarbons with Sulfur", Ind. Eng. Chem. ^8, 376-82 (19ii-6 )

"The reactions of sulfur with various types of hydrocarbons and hydrocarbon mixtures (petroleum frac­ tions) have been extensively studied by many investigat­ ors, who have generally reported that the principal pro­ ducts of the reaction are hydrogen sulfide, carbon di­ sulfide , and complex sulfurized hydrocarbon derivatives, including polymethylene sulfides and other products, rncs tly of unknown structure.--- "The reaction between sulfur and hydrocarbons may be compared to that of oxygen and hydrocarbon in that uncontrolled reaction conditions In each case to analogous end products: hydrogen sulfide-carbon di­ sulfide and water-carbon, dioxide. The essential dif­ ference between the two reactions is one of degree, sulfur being milder in Its reactions than oxygen and hence more controllable. Under certain conditions, less critical than for oxygen, It should be possible to obtain attractive yields of primary or secondary products in \' which the original carbon skeleton Is preserved In the reaction of sulfur -with, a hydro carbon. The present paper describes a feasible method for controlling the reaction between sulfur and such hydrocarbons as butane, pentane, and hexane to give the corresponding olefins, diolefins, and thiophene or alkylthiophenes• T,If a hydrocarbon such as n-butane is bubbled through molten sulfur at a temperature in the range J00- !j.00o C ., the exit gas stream will contain considerable amounts of hydrogen sulfide but no olefins. Dehydro­ genation has evidently taken place, but the unsaturated products have undergone a secondary reaction with sulfur (or perhaps partially with hydrogen sulfide) and largely remain in the molten sulfur. If the butane is passed in­ to a zone of sulfur vapor at l+ljl^C. and the mixed gases are heated together to about 600-6^0^0., the product gases may contain appreciable amounts of butylene, buta­ , and thiophene, provided the reaction time is not too long. However, the reaction tube will soon become .plugged with a heavy carbonaceous deposit, resulting from the decomposition of a heavy sulfurized tar formed in the reaction. nBy preheating the reactant vapors separately to about 600°C., introducing them rapidly through a mixing ■ r '■ o o nozzle Into a reactor tube at 600°C., and quenching the product stream rapidly, the process may continue for long periods without interruption* With a proper choice of temperature, reaction time, and sulfur concentration, yields of butylene, butadiene and thiophene may be sub- f stantial* / / The7react!ons involved appear to lead stepwise to the formation of * thiophene:

9l|.H10 / — °2jH8 / H 2 S (X) °1|.H8 / is2 — % H 6 / H2S <2)

°4 H 6 / S2 — < W * H2S C 3) "Actually, these equations represent the ideal <■ case, but o then reactions also occur* Some of them may be controlled and minimized, resulting practically in about Q0% conversion of butane according to equations 1, 2 and 3 - that is, withdrawing butylene, butadiene, and thiophene as useful products, or about 50%con­ version of butane to thiophene alone for the overall reaction* "The formation of thiophanes according to the reaction . ’ ' / S2 - C^HgS / H2S- reported by Friedmann who worked at lower temperatures ■ jL and higher pressures was not observed in the present investigation. "Other reactions which occur simultaneously with the above reactions are: (a) degradation by thermal cracking of the initial hydrocarbon to lighter hydro­ carbons , (b) complete sulfurization of a portion of the primary reaction products to carbon disulfide, and (c) secondary reaction of a portion of the reaction pro­ ducts with additional sulfur or with hydrogen sulfide to form a .complex tar of high sulfur content. "By proper choice of temperature, time, and sulfur concentration, reaction a and b can be largely minimized. Reaction c_ however, appears to proceed about as fast as the ‘overall reaction leading to thiophene formation since, under the best conditions attained, the amount of tar and thiophene are about equal. Practically all of the sulfur in the major by-products, hydrogen sul­ fide and tar, can be recovered for recycling to the pro­ cess by burning the tar and enough of the hydrogen sulfide to to give the reaction:

2 H£S / S02 — 1& S2 / 2 H20 (!(.) "A prerequisite to the formation of the thiophene ring is a linear chain containing at least four carbon atoms* n-Butane will give thiophene, but isobutane will not; n-or isopentane will give ms th.ylthioph.enes, and all of tlie aliphatic hexanes will give dimethyl thiophene s cr ethyl thiophene s • Hydrocarbon s lower than do not yield thiophene but can be dehydrogenated to olefins. No diolefin was found in the reaction of propane or pro­ pylene with sulfur, and no appreciable amounts of acetyl- enic hydrocarbons were found in the range of conditions investigated with any of the hydrocarbons. The lower hydrocarbons, especially the olefins such as ethylene and propylene, are readily converted to carbon disulfide. Methane requires more severe conditions of time and tem­ perature and forms principally carbon disulfide. "In the operation of the process under the preferr­ ed conditions using n-butane charges there appears in the product stream a mixture consisting of butylene, buta­ diene and thiophene along with ' unreacted butane, hydro­ gen sulfide, tar and minor amounts of carbon disulfide, lighter cracked hydrocarbons, and bottoms heavier than thiophene. The thiophene, butadiene, and butylene may be withdrawn and the balance of the C]^_ in the product stream recycled. --- "Hydrocarbons lighter than yielded only the corresponding olefins and carbon disulfide. In general, the lighter the hydrocarbon homolog, the higher the react­ ion temperature necessary to achieve dehydrogenation, other things being constant. The dehydrogenation of ethane to ethylene required a temperature of about 760°C. for a reaction time of 0.1 second and a sulfur ratio of 2 to 1 by weight to give a yield. The reaction temperature under similar conditions and for similar yields of pro­ pylene from propane was 704°c* & charge consisting of ethylene or propylaae and sulfur gave only carbon di­ sulfide as the main product. Table I gives examples of. results obtained in charging lighter than hydro­ carbons to the process. "Higher hydrocarbon homologs than 0^ give the homologous olefin-diolefin-alkylthiophene reaction. The reaction of degradation competes more strongly, however, the higher the molecular weight of hydrocarbon charge. On the other hand the higher the molecular weight, the more readily is dehydrogenation with sulfur achieved but also the mere complicated is the reaction in that com­ plex sulfur compounds are formed in larger amounts. Pen- tane and hexane have both been successfully used as charge materials in the process to yield me thy Jfc thiophene and di­ me thy I'-o'r ethyl thiophene, respectively, as well as the TABUS I RESULTS OF CHARGING HYDROCARBONS AND MIXTURES TO THE PROCESS Conversion of Products and. Reaction Approx. Re­ Wt. Ratio, Hydrocarbon Yields per Hydrocarbon Temp* action time, S to Hydro­ Charged per Pass Charged •C. Seconds carbon in Charge Pass n-Butane 650 0.07 1.0 25.4 15.5 C4H4S 2-Butene 650 0.07 ' 1.0 53.7 28.7 C4H4S Mixed n-butylenes 652 0.07 1.0 53.3 27.2 C4H4S 30j6 butadiene, ( TO# n-butane 651 0.07 1.1 42.4 22.1 Cj & S Isobutane 649 ’ 0;07 0.5 22.3 14.3 iso-C*H* Propane 704 0.1 2.0 17.6 15.0 C3H6 Ethane 760 0.1 2.0 15.9 14.4 C2H* Propylene 704 0.1 2.0 13.9 18*2 Ethylene 760 0.1 -2*0 m 3 7.9 OS*

aExpressed as moles of hydrocarbon used to form the product Indicated and degradation products per mole charged, x 100

Expressed as moles of hydrocarbon charge converted to the product indicated per mole charged, x 100. The difference between columns 5 and 6 represents degradation to other products than those indicated, chiefly tarj column 6 divided by column 5 in each case would be an approximate calculated ultimate yield* ' . ;o corresponding olefins and diolefins. The process where these materials are charged in minor proportions, along with n-butane, to yield a mixture of products which can then be separated by fractionation has also been success fully carried out, Further work on hydrocarbons higher than is being conducted in this laboratory, "The superheating of sulfur vapor, to temperatures in the vicinity of 700°C. as well as the handling of high ccncentrations of hydrogen sulfide at elevated tem­ peratures have presented severe corrosion problems. Ex­ tensive tests were conducted to find corrosion resistant alloys for the conditions used. Of the conventional stainless steels, the high chromium type was found to be superior to the chromium- types. For this reason 27% chromium steel (low carbon, no nickel) was used ex­ tensively since it is available in tubing of all sizes and can readily be machined. High- steels, such as duriron, were found to be superior to high chromium steels, but they can only be cast and are non-machinable A possible solution to the problem is aluminum-coated or calorized steels; or aluminum-containing steel alloys, perhaps also coated with aluminum* An alloy containing 20% chromium and 2.% aluminum was found to be superior to 27$ chromium steel. Aluminum-coated or calorized steels (both, carbon steels and alloys) were found to be more corrosion resistant than any of the alloys just men­ tioned* The problem in connection with the calorized or aluminum-coated steels is to obtain a coating free of sur­ face imperfections such as slag or oxide inclusions or pinholes. If such defects are present, they in variably cause localized failures. Many failures of aluminum coated or calorized installla ti ons have been reported in connection with other projects, probably largely traceable to surface defects. Recent improvements in the method of applying the aluminum coat may decrease the frequency of such failures. The corrosion resistance of aluminum coated articles- Is probably due to a protective\ coating v of aluminum sulfide initially laid down on the aluminum rich alloy, which is present on the calorized article or which soon forms from the aluminum-coated article as a result of diffusion of the aluminum into the parent metal when the specimen is heated." The literature on the reaction of light hydrocar­ bons with sulfur or sulflir compounds is not extensive, and some reliance must be placed on information obtained on thermal reactions as guides far study, with the ■ thought that the use of sulfur as a hydrogen acceptor should im­ prove the yield of unsaturates-and permit the use of low­ er tempe r atur es • the .following references are cited with this thought in mind. Thermodynamic studies of the equilibrium between ethane and ethylene has been reported by Pease and Durgan •

Pease, Robert N. and Durgan, Elford S., J.A.C.S. 50* 2715-8 (1928)

"Bone arid Coward have shown that ethane is rather rapidly decomposed at 675°> ^ <3- that the primary pro­ ducts of decomposition are ethylene and hydrogen; and it is well known that ethylene and hydrogen combine quanti­ tatively to form ethane in the presence of catalysts from room temperature up to at least 550°. The indica­ tions are, therefore, that the reaction C2H5 B C2H1|. / ^2 is reversible somewhere within the temperature region 550 and 675°* and- that equilibrium should be determinable Indeed, Berthelot has reported that dissociation and for­ mation of e thane both occur at a dull red heat. These results have encouraged us to attempt the measurement of equilibrium in this reaction. — - "Our equilibrium measurements were made at 600, 650 and 7OO0. Observations were also made at 500 and 550° r but this extreme slowness of ithe reactions and other factors rendered the measurements unsatisfactory. At 600 ■ I *+4 . JL and above equilibrium- could-b© approached more rapidly. Methane formation could not be avoided entirely, and at 7 00° -would seem to have reached serious proportions. However, the consistency of our results inclines us to the belief that our final equilibrium values are not far from the truth. "Our final experimental results are presented in Table I. For each experiment we give the composition and pressure of the gas mixture as introduced into the reaction bulb, and the composition and pressure after heating. "Three other experiments carried out at 600° with mixtures initially close to equilibrium gave for the equilibrium constant the values O.O309, O.O307, O.O315. "Our data indicate that the equilibrium constants at 600, 650, and 7OO0 are 0 .0310, 0 .082, and 0.20 re­ spectively. The plot of the corresponding values- of log K against l/T gives a straight line. The data are sat­ isfactorily reporduced by the equation F * -RTlnk - 31,224-28.88T. "According to the above equation, the heat of dissociation of one mole of ethane into ethyloae and hydrogen at 6OQ-7OO0 is -31,244 cal. From the combus- TABLE 1

EQUILIBRIUM DATA

Heating Pressure Composition of Gasjo time,min. Mixture Atm. C2 V H2 C2H6 OH4 I2 Katm = KpX at 6oo°

m .30 Initial O.673O 5*80 92.00 2.20 • ' Einal O.7915. 13.55 17.65 65.50 1*55 1.75 (0.0289) 30 Initial .9770 4840 50.20 «■ 1 4 0 Pinal .6620 17.05 25.95 1 5 .5 0 10.35 1.30 (0.061+7) 30 Initial .9115 4 .3 5 16.05 67.95 I.6 5 i Pinal .9260 14.55 14*75 62.80 6.15 1.75 O.O316 at 6500

6 Initial O.5555 5.80 92.00 wm 2.20 Pinal .7290 20.35 24.40 47*15 6.70 1.35 (O.O768) 8 Initial 1*0040 49.00 49*60 - 140 Pinal 0.71^5 21.40 25.20 35.85 15.25 2.30 (O.IO75) 6 Initial 1.0205 19.80 21.70 56.75 1.80 Pinal 1.0500 19.20 20.35 5O.9O 7.90 1.70 0.0806 6 Initial O.969O 20.85 25.05 5245 1-55 . ^ n Pinal O.9785 17.30 23.60 47.6O 10.65 0.85 0.0839 at 700u 5 Initial O.9975 24.75 30.95 42.40 - 1.90 Pinal I.O925 15.60 26.10 22.55 34.10 1/65 0.197 5 Initial 0.96^0 34.20 35.95 28.10 mm 1.70 ■ Pinal *9735 17.45 26.15 22.65 31.85 1.95 0.196 r; **,3 tion data of Thomsen and Berthelot the heat of dissocia­ tion of ethane may he calculated to he -51*200 cal. (Thomsen) or -57*100 cal. (Berthelot), the average being -5^,200 cal. Although adequate specific heat data are lacking, the indication is that our value at 600—700° might he reduced hy 1000-5000 cal. at room temperature, making it approximately 28,000-50,000 cal. as compared to the above average of -5^.,000 cal. which also is for room tem­ perature. "It is difficult to estimate the accuracy of our results. The most serious source of error is indicated hy the presence of considerable amounts of me thane among the products of some of the experiments. The corres­ ponding side reaction undoubtedly has displaced equilib­ rium to some extent. However, in view of the facts that approximately the same endpoint is reached from either side and that our calculated heat o'f reaction is not un­ reasonable, we are inclined to believe that our data are not greatly in error. "

C2H6 m C2E[^ / H 2 has been measured at 600, 650, and JOO0 • Equilibrium was approached from both sides. Some un­ certainty exists owing to the simultaneous formation of methane, but the indications are that the equilibrium constants at the three temperatures are 0.0310, 0.082, and 0.20 respectively. The themal cracking of paraffins to olefins, es­ pecially ethylene, and to acetylene, is described exten­ sively in patent literature. Two typical examples are presented here in some detail. The first is an example of non-catalytic thermal dehydrogenation, and serves to emphasize again the control necessary in thermal crack­ ing processes. The example cited is that of Robinson.

Robinson, Sam P., "Hydrocarbon Conversion Process", U.S. Pat. 2 ,1^39,023, April 6, I9I4.8.

"The conversion of hydrocarbons to other less saturated hydrocarbons of lower molecular weight can be readily affected at temperatures above about 1000°F. Such reactions are ai do thermic, so that it is necessary to sup­ ply large quantities of heat, at these high temperatures, in order to obtain an extensive conversion, per pass, of the charge stock. However, the resulting unsaturated hydrocarbons, such as olefins, diolefins, acetylenes, etc., are quite reactive at these same high temperatures, and readily take part in polymerization or condensation reactions which are highly exothermic. As a result, in order to convert a given hydrocarbon material to less saturated hydrocarbons of lower molecular weight not only must the hydrocarbon material be heated to a high temperature level, but the reaction conditions must be closely controlled. This control must be effective not only to prevent loss of desired products by secondary reactions, but also to prevent a conversion process from getting out of control as a result of extensive and self accelerating exothermic reactions of such unsatur­ ated hydrocarbons. I have now found that a continuous conversion process of this nature may be readily control­ led by a rapid cooling, or quenching, of the reaction. Reaction times for the desired endothermic reaction are dependent upon the temperature of the conversion and vary, mere or less inversely with the temperature, from a few seconds (less than a minute) to as low as about 0.01 second. At temperatures in the region of, and below, 1O0O°P the undesired exothemic reactions do not take place rapid­ ly and a suitable upper limit for the cooling can generally be found in this region. In most instances a relatively compile x mixture of hydrocarbons result s from the desired conversion, and a relatively complex combination of com­ pressors, separators, coolers, condensers, fractional •v. O distillla tion columns and all usually associated equipment is necessary tollaolate an individual desired unsaturated hydrocarbon product in relatively pure form. --- "It Is an object of this invention to effect a con­ version of hydrocarbons at a high temperature. "Another object of this invention is to convert other hydrocarbons to ethylene. "Still another object of this invention is to con­ vert hydrocarbons at a high temperature to lower boiling unsaturated hydrocarbons with a minimum formation of high-boiling hydrocarbons. — - "A preferred embodiment of my invention will now be discussed in connection with t he accompanying drawing, which is a diagrammatic flow sheet illustrating schemat­ ically various pieces of equipment which can be used in the practice of my invention. Referring now to the draw­ ing, a suitable hydrocarbon charge stock, such as a material comprising a major portion of propane together with some ethane, is introduced through line 10. This Is composed of a net charge, introduced through line 9, and of a recycle stream passed through line 55* 3-a often preferred that this materially be substantially free from sulfur compounds, such as hydrogen sulfide and \ mercaptarns, and when so desired a sulfur-containing charge METHANE

53

f 37 STEAM

WATER V

JSIHAKE^PROPAHE RECYCLE Diagram of Hydrocarbon Conversion Process Chen. Rngr. Dept., Ohio State University Rbbinson, S.P., U.S.Pat* 2,439,023 Chen. Eng.950 April 6, IW July 12, 1952 Morningstar «r^O

' t O may be treated to reduce its sulfur content. Ttiis charge stock, at a pressure not greater than about IpO pouhds per square inch, absolute, is passed through an expansion valve or orifice 11 to preheating coil 12 and cracking or dehydrogenating coils 13 aa d lip, which are situated in furnace 15 • In coils 13 and lip the charge is heated to a temperature of about 1350 to about l650°P. for a time sufficient to give an optimum yield of ethylene. This will generally be about 1*5 about 0.15 second. In actual plant operation, a furnace 15 will have a plural­ ity of preheating and cracking coils operated In parallel and by means of an expansion valve or expansion orifice 11 at the inlet to each set of coils, the flow through the sets of coils will be more or less uniform. To the hydrocarbon stream, prior to its introduction to the re­ action 3d ne, steam is added through line 16 . Enough steam is used to passivate catalytic metal and depress formation. In some instances about 0.1 to about 0 »5f° by weight of sulfur will have a similar effect. However, when about 0.2 to 0.5 mol of steam per mol of hydro­ carbon is used, the effect of the decreased partial pres­ sure of hydrocarbon reactants Is beneficial, and leads to optimum ethylene production, and serves to minimize polymerizing of olefins produced. It is preferred that, in the cracking coils 13 and llj., tubes of a small dia­ meter be used, receiving direct radiant heat and with the diameter of the tubes and the flow of reactants through the tubes so correlated that a lower than usual rate of heat transfer through the tube walls takes place. This r esults in having more uniform gas temperatures throughout the cross-sectional area of the cracking tube and increases the ratio of heat transfer area per cubic foot of gas handled. Effluents of cracking coil 1I4. are passed through line 17 to a tar separator 2 0 . Immediately upon the exit of the gases from c racking coil II4., water is injected through line 18 to cool the gases as rapidly as possible, to a temperature' below, about 1000°F. Unless this is done, beneficial results from careful design of the cracking coils and control of flow rates with heat transfer rates to produce optimum ethylene production will be lost, be­ cause ethylene is highly reactive and readily polymerizes at the reaction .temperatures. — - "My Invention will be further illustrated by the following example. In connection with this example, re­ ference Is made back to various pieces of equipment il­ lustrated In the drawing. A gaseous C^-C^ hydrocarbon mixture is Introduced through valve 9* having a composi- tion shown in the accompanying table. To this is added a recycle stream from line 55, also shown in the ac­ companying drawing, and steam to give a to tal furnace feed passing through expansion orifice 11 as shown in the table. Compositions in the table are in mol per­ cent. This mixture is subjected to cracking at a max­ imum temperature of about lij.70oP. under an exit pres­ sure of about 20 pounds per square inch absolute, to give an effluent through line 17 having the composition shown in the "table, prior to the introduction of quench water. Sufficient water is directly injected through temperature controlle d valve 19 to the stream immediate­ ly upon its exit from the cracking coil, to bring' the » t - tempe rature down below 1000°P. Additional water is then added through line 22 to bring the stream temperature to about 300°P. A small amount of tar is removed through line 23, and the gaseous effluent is passed through line §5 to water quench tower 26. A stream of water, intro­ duced through- 50 &t about 100°P. cools the gases passing through line 35 to about llj.0oF. When an abnormal inter­ ruption in service causes the temperature of the gas pas­ sing point 21 to rise from its normal temperature of about 300°P to abou t 1000°P., valve I4.I immediately opens to discharge the gas passing from quench tower 26, and valve If. 3 immediately causes the in furnace 15 to be kilUe d. "Cooler 36 cools the gas to about 100°P, and con­ denses most of the remaining water, which is discharged through line If. 5 • The remaining gaseous material is passed through line If.7 to separating means 50 wherein it is compressed to a pressure of about 600 pounds per square inch absolute, cooled, and subjected to a series of fractional distillation steps. The hydrogen as dis- - charged through line 62 has the composition shown in the table, and the methane fraction discharged through line 53 has the composition shown. The resulting ethylene stream, removed as a product of the process through line 51, is substantially pure, having the composition shown in the table •" An example of catalytic thermal dehydrogenation is given by Thacker.

Thacker, Carlisle M. "Dehydrogenation of Hydrocarbons", U.S. Pat. 2 ,35i+,892, Aug. 1 , 1 9 Ijlp.

"Various catalysts have been tried and the prior art and literature disclose a xnumber of catalysts useful in dehydrogenation of hydrocarbons. Known catalysts are un­ satisfactory, however, for the reason that the amount of conversion produced thereby is not sufficiently high to warrant commercial use; or the efficiency of the catalysts is so low that undesirable products, such as coke, are produced in large quantities in addition to the desirable unsaturated hydrocarbons; or the life of the catalyst is too short. "I have, discovered that dehydrogenation of low boiling saturated hydrocarbon gases can be effected with a high percentage of conversion and high efficiency of conversion by using a catalyst containing a difficultly reducible metallic oxide gel, whie,h possesses extended surfaces and a highly adsorptive capacity iS©r gas, and a metal, metal oxide, or compound, which in combination with the gel imparts thereto the selective property of dehydro genating hydrocarbons. When a catalyst component of these two classes is used, the results obtained are entirely unexpected since neither component alone will produce results even closely approaching that of the combined components. "In accordance with my invention, the activated alumina may be prepared by precipitating the tri-hydrate from an aluminate solution and calcining the precipitate R r§ at temperatures of from 300 to 800°C. The methods of pre­ paration are fully set forth in the patent to Barnitt; No. 1 ,868,869 and Derr, No. 2 ,015,593. A well known acti­ vated alumina is that sold by the Aluminum Company under the trade name TAlorco* (Grade A). The alumina gel sold under this trade-mark has a large surface which makes it particularly active when used in conjunction with mild dehydrogenating catalysts. "As metals and oxides that may be used in combina­ tion with the gel are: a. Readily reducible metals of Group 1 of the per­ iodic table, such as , and ; b. Oxides of metals of Group IT of the periodic table, such as beryllium, magnesium and zinc; c. Oxides of difficulty reducible metals of Groups V, VI, and VII of the periodic table, such as vanadium, chromium, tungsten, molybdenum, and manganese. "Combinations of the aforementioned metals and oxides In conjunction with the gel may be used. For example, good results are obtained by depositing copper tungstate or magnesium and chromium compounds on activated alumina. De- hydrogenating catalysts such as , cobalt, and nickel have been tried in conjunction with activated alumina, but these catalysts are unsatisfactory for the reason that they tend to decompose the hydrocarbons into coke and hydrogen instead of simply splitting off hydrogen to form olefins,--- "In order to more clearly understand the manner in which the catalyst is prepared, the following examples are given:-- "Example 2 . Chromic acid. (19.6 grams) was dissolved in water (100 cc) and. added to activated alumina (Z|.00 grams) that had previously been heated at 110-120°C. for 2 hours. The mixture was well stirred and then dried overnight in an electric oven at 11Q-120°C. The dry material was screened through an 8 to 14 mesh sieve and. .reduced in a stream of and hydrogen by gradually increasing the temperature to 250°C. over a period of 1 to 2 hours, and then continuing the reduction at 2$0°C for two hours. The catalyst was then heated, for 15 hours in dry hydrogen at 450°C., after which it was ready for use in the dehyd.ro- genation of hydrocarbons.-- "Example L. A gas consisting mainly of ethane with a small amount of impurities was contacted with a catalyst consisting of chromium oxide d.eposited on activated alumina prepared in accordance with Example 2 , at a temperature of 600° C . and at a space velocity of 521. The gas was dried prior to contact with the catalyst by passing it through calcium . The reaction gases were analyzed and showed an ethylene content of 1 3 . and a hydrogen content of 18.5$. The yield of ethylene per cubic foot of charging gas was 14% and methane formation was practical!;/- neg­ ligible . "Example 5 . Another run was made using the same gas'and. same catalyst with the exception that the gas was saturated with water vapor at room temperature. At a tem­ perature of 550°C. and a snace velocity of 515, the re­ action gases contained, only 3.4% of ethylene and 1 . 8 % of hydrogen. The yields of ethylene based on the charging gas was 1%. A run made under substantially the same condi­ tions with dry gas produced 9.2% of ethylene and 8 . 7 % of hydrogen with a yield of 7.5%, thus showing the detrimental effect of water vapor. : "Example 9 . In this run dry ethane was contacted with a catalyst consisting of activated alumina only at a temperature of 600O C . and a space velocit;?- of 165 . The re­ action gases obtained only 1% of ethylene and 1 . 1 % of hy­ drogen. "In the examples given above, space velocity is de­ fined. as volume of gas entering converter per hour at 0° and at 760 ram. pressure per unit volume of reacting space. The runs were all conducted at substantially atmospheric pressure.-- "A number of runs were made to determine the effect of various catalysts giVen in the above examples when not \ supported, on activated alumina. The results of these runs are given in the table below: .... Temp­ Ethylenfc Hydrogen erature Space in Reac- in React- Catalyst °C. Vel. tion gases ion gases Yield Chromium 600 337 2.6 2.3 Oxide 675 337 3.8 2.8 1.4 673 334 4.8 .3.5 2.7 Magnesium Oxide- Chromium Oxide (1 :1 ) 700 337 8.4 6.5 7.3 Magnesium Oxide- Chromium Oxide (1 :3 ) 700 337 9.2 6.6 7.5 Copper 675 341 3.1 2.8 0.7

In each case the catalyst was supported on pumice except in the case of copper, and in that case the copper was unsup­ ported. "These results clearly show that the activity of the oxides and the gel when used separately is far below the activity of the combined catalysts and that the activity of the combined! catalysts is far greater than the additive re­ sult of the two. In the case of magnesium oxide-chromi$*m oxide, although the yield appears to be as good as the yield obtained when the catalyst was supported on the alum na gel, the activity of the catalyst actuallj1- was far less when not supported on the gel because in the latter case, the temperature of contact was 700© as against 350° in the run where catalyst was supported on the alumina gel and the A ' 1- . . t space -velocity was much less in the case of the catalyst which was not supported on the gel. Increase in reaction temperature greatly increases the reaction and likewise decrease of space velocity increases time of contact, and consequently, the amount of conversion. As a matter of fact, when the gas was contacted with the gel-supported catalyst at 600°C. and a space velocity of 505, the yields were greater than those obtained at 7 0 0 ° and 337 space vel­ ocity with the unsupported catalyst. "In all cases In which the catalyst was supported on the alumina gel, the efficiency was approximately 80% or better and the catalyst maintained its activity over the entire period of the run without showing any loss in activity. ” "As has previously been said, catalysts prepared In accordance with my invention retain their activity for longer periods of time than catalysts which have hitherto been tried. In the case of chromium, copper and copper tungstate in combination Yfith alumina, the activity of the catalyst actually increases gradually for a time vfhen in use. "The structure of the metal or oxide deposited on the alumina is immaterial. No effort Is made to deposit the metal or oxide in any particular form so long as it is evenly distributed over the alumina support. The methods * o i -4) used to deposit the metals or oxides on the alumina are not conductive to gel formation thereof. "The catalysts can he easily reactivated’ after their activity has fallen off, hy heating in air, and reducing with hydrogen in the case of metallic substances such as copper. A catalyst prepared from chromic acid and activ­ ated alumina in the proportion of 200 AI2O3: 1 Or. was com­ pletely restored by heating in air for 16 hours at 4 2 0 ° C . and at 550°C. for 2 hours. tr Some interest has been shown in the partial oxida­ tion of ethane to ethylene and acetylene by means of oxygen or an oxygen containing gas. This method appears not to be widely used,” probably because of difficulty in controlling the conditions,vhence also the nature of the products form­ ed. Two brief abstracts vfill serve to illustrate the type of reaction that is carried on.

Hans Klein, Ferdinand Hanbach, and Wilhelm Hofeditz, (vested in alien property custodian) U.S. Pat. 2 ,301,727, Hot. 1 0 , 1942, C.A. J37 2013 (1943)

''Hydrocarbon materials such as C2H6 or C3H6 is heat­ ed (suitably to about 600°) in the presence of such a pro­ portion of oxygen that at the prevailing pressure (which may be about atmospheric pressure) there is no formation of a , the heated mixture being passed through the first part of a reaction zone at a high speed of flow and tiirou^i the following part at a lower speed of flow effected by pro riding a larger cross-section of the apparatus used through which the material flows."

M. E. Spaght, Petroleum Processing 1 135 (1946) C.A. L2 9139 (1948)

was obtained by I.G-. at Leuna in 70$ yield (based on ethane charged) by partial oxidation of ethane. Three volumes of G2H6 and one volume oxygen were separate­ ly preheated and burned under slight vacuum at 1470-1560°E. in a special burner. The product consists of CH2GH2 4 8 .1 , C2H6 30.5 , CO l6 .0 , CO2 0.6, O2 1.6, propylene and higher olefins 2 .4 , C2H '2 0 .8 . Recycling recovered increases ultimate G^H.^ yield to 74$. Purification was by fractional distillation after removal of CO2 and C2H2. "A second process employs straight thermal cracking at atmospheric pressure and 1550°E. After quenching, G2H4 is concentrated by selective solvent extraction. Yield. * 30$ per pass." Literature on the reactions of light saturated hydrocarbons with sulfur is not extensive. Two interest­ ing references to the reaction of hydrocarbons with carbon disulfide are worthy of mention. The first is described by Angus. 1

Louts Angus, French Pat. 648,340, Issued Aug. 1 3 , 1928

"The transformation of methane Into ethylene re­ quiring a deposit of carbon as well as heat, it is advan­ tageous to make use of a composition of carbon of end.o- thermic formation. The inventor has recognized that carbon disulfide is particularly adapted to this reaction. "The present process consists of making use of carbon disulfide and methane. "A method of realizing this consists of causing methane to bubble into a receiver containing carbon di­ sulfide which can be heated to some degree below its boiling point. The mixture thus obtained is directed Into the reaction zone which is heated. The reaction zone con­ tains preferably porous materials, such as pumice stone, , etc. "The action of catalysts can equally be considered and permits lovirering considerably the reaction temperature. "This reaction can also be carried out at high pressure. "In order to increase the yield, the reaction can be carried out in the presence of materials capable of fixing the sulfur, such as, for example, copper, zinc, or others.-- A very similar process is described by Spindle r.

Henri Spindler, French Pat. 649,687 Issued Sept. 4 , 1928

"The object of this invention is to obtain, starting with saturated hydrocarbons such as kerosene, etc., the components of the ethylene series. "The procedure consists of causing the saturated hydrocarbons to react with carbon disulfide at a tempera­ ture of 400 to 800° C . depending on the molecular weight of the hydrocarbon treated. "To this end a mixture of carbon disulfide and the saturated hydrocarbons is passed into a tube heated to the most favorable temperature. "The tube- may preferably contain materials capable of transferring heat, such as pumice, stone, bauxite, or others. Catalysts may also be added to facilitate the reaction. "The products leaving the reaction zone may be cool­ ed in order to condense liquid prod_ucts formed.. The gas consists principally of ethylene, propy3.ene, butylene, etc. "In this process a phenomenon can be obtained sim­ ilar to cracking so that from a saturated hydrocarbon with a certain number of carbon atoms, there can result ethyleni products with fewer carbon atoins. r o V_ i - 3 "Through. the intermediate use of hydrocarbons of the ethylene series obtained according to the present pro­ cess, it is possible to prepare from saturated hydrocarbons a multitude of organic chemical products.” DISCUSSION OF THE LITERATURE As indicated previously, little direct reference vfas found to the reaction of sulfur with light hydro­ carbons in the literature. It Is well known that sulfur reacts with heavy hydi’ocarbons to form hydrogen sulfide and olefinic linkages in the hydrocarbon. It is also known that hydrocarbons, such as ethane, can be cracked thermally, catalytically or non-catalytically, to olefins and acetylenes, with the formation of hydrogen. It is reasonable to speculate that any material that would re­ act with the byproduct hydrogen would result in a greater yield of the unsaturates,. Oxygen has been used somewhat for this purpose, and ultimately may find wider use in this reaction if a relatively simple control means can be devised. Unless very efficient methods of mixing are used, und.esirable oxidation products, up to and including carbon dioxide, could readily be formed. Sulfur, on the other hand, is a much milder oxidizing (dehydrogenatIng) agent, and so would exert a much milder.influence. The very short reaction times referred to in the literature are a second indication of the careful reaction control required in cracking processes. This is necessary since at the temperatures rea.uired for dehydrogenation there is a strong tendency to crack entirely to hydrogen and car­ bon. With the use of sulfur the dehydrogenation temperature may be reduced and, if dehydrogenation can be limited to that obtained by reaction with sulfur, subsequent.carboni­ zation may be avoided. This would eliminate the need for very short reaction time and the critical heat transfer operation would be avoided. THERMODYNAMIC CALCULATIONS

The probable effect; of sulfur on the dehydrogena­ tion of ethane can be obtained by thermodynamic calcula­ tion of the equilibria involved in the system. For this purpose the terminology of Lewis and Randall will be used.

G. N. Lewis and Merle Rand.all, "Thermodynamics and the Free Energy of Chemical Substances” , First Ed., McGraw- Hill Book Company, Inc., New York (1923)

For a detailed account of the deriviation of the equations the reader is referred to the above text. The fundamental relationship between free energy and the composition of a.reaction mixture at equilibrium is given by the-familiar equation ^ F ° ‘ = -RT in Kp where AF° is the free energy change involved when an equation "weight of reactants in the standard state react to form products also in the standard state. For the reaction aA / bB / -- — *». rR / sS / -- K is defined as ^ Y - ( R ) — " (A)a (B)b --- where the bracketed quantities are concentrations at equi­ librium. In the case of gaseous reactions, partial press­ ures may be substituted for concentrations to yield the ft'— £",■! expression !... i ;

Kp = P3— ‘ *1 PB— Thus, If values of AF° are known, and a! given ratio of reacting materials is chosen, the concentrations at equilibrium may he calculated, as will he illustrated below. Values of may be obtained readily from pub­ lished data. Since A'iJ'° Is an extensive property depending only on quantities involved and on the condition of the

initial and final states of the 'System, a F° for any reac­ tion may be obtained from the relation AF° = r / S (^Fo0)g / --

-a (a f 0 °)a - b U F 0°)B ------where 1'0° is the standard free energy of formation. Valu© of' F0° may inturn be evaluated from the heat of formation of the compound involved by use of the relation ( Af / t ) Z A H ( A T )p T 2 The value of H may be obtained from the expression H = J CpdT

Tor the calculations presented here, values of a F° (andof AH and Cp where required) were obtained from the following sources: 5 r ~ ? L. - Data on sulfur and hydrogen sulfide were obtained from Kelley,

K. K. Kelley, "Conti’ibutions to the date on Theoretical Metallurgy, VII The Thermodynamic Properties of Sulfur and its Inorganic Compounds” , U. S. Dent, of Interior, Bureau of Mines Bulletin A06, U. S.

George S. Parks, and Hugh M. Huffman, ”The Free Energies cf §ome Organic Compounds” . The Chemical Catalog Company, Inc New York, (1932) with reference as required to Rougen and Watson

0. A. Hougen and K. M. Watson, "Industrial Chemical Cal­ culations”, Second Ed., John Wiley and Sons, Inc. New York (1936) and Perry.

John li. Perry, "Chemical Engineers Handbook,11 Second Ed.., McGraw-iiill Book Co., Inc., New York ^1941^

From these sources, the following free energies of forma­ tion were determined: C2H 6 : A F 0° - -19,190 / 15.7 TlnT -0.0046 T2 - 59.6t C 2H 4 : A F 0° - 10,010 / 9.2 TlnT -0.0026 T2 - 43-7 T

C2H 2 : AF0° ~ 54,900 - 13.6 T CH4 : A F 0° =-15,450 / 11.1 TlnT -0.0081 T 2 / 0.6klO“6T3- 50. IT

h 2 / h S2(g) -> h2s ^ F 0° = -19,405 / 7.71 TlogT - 1 .033x10-3t2 - 12.50T From the above, the free energy changes for the fol­ lowing reactions were calculated: (1) C2H 6 -> H 2 / c2h 4 A F° = 29,200 -6.5TlnT / 0.002 T 2 / 15.86 T,

(2) C2H4 - > Ii2 / C2H 2 A F° = 44,890 - 9.2TlnT / 0.0026 T 2 / 30.1 T

(3) H 2S -> H 2 / 5 S2 ' , A F ° = 19,405 - 3.345 TlnT / 1.033x10"^T2 / 12.51 T (4) 2CH4 -> C2H 6 / H 2 AF° = 11,710 -6 . 5Tlnt / 0.0116T2 -1 .2x10“6t 3 / 40.6 T Using the above equations, values of Kp were calcu­ lated over the range of conditions likely to be encountered in this study. The calculated values are summarized in Table I. TAB IF I. VALUES 'OF Kp Temp. °K Temp. °F. Km Kp? Kp? 500 441 2 .46xl0”S9 .6xl0"15 1 .62x10"“ ^ 4 .4x10 750 891 1.20x 10“3 1.6x10-7 1.91xl0“;+ 2 .14x10“ 3 1000 1341 0.331 8.33x10 -4 6 .90x10“ 3 1 .26x10“^ 1200 1701 6.10 0.0804 0.0440 —

1250 1791 11.30 0.160 0.0663 2 .88xL 0“^

1500, 2241 3 .67x10"^ 5.20 0.274 3.75x10“^ In addition to the above reactions, reactions in­ volving the formation of carbon disulfide and of elemental carbon were also considered and free energy changes cal­ culated. Two thing’s were immediately apparent: (1) In the case where carbon was considered, in the entire useful temperature range practically complete de­ composition to carbon and hydrogen (or hydrogen sulfide) would be obtained at equilibrium. This is probably the reason that very short contact times are required in thermal cracking of light saturated hydrocarbons. (2) In the case where it was assumed that no carbon would be formed, decomposition to ethane and carbon di­ sulfide was found to occur at equilibrium. This is a con­ dition that must be regarded as artificial, however, since there would probably be no feasible way to carry out de­ gradation to these products without also forming elemental carbon. In order to obtain values at equilibrium that would be useful in predicting"the effect of sulfur on the con­ version of ethane to unsatura$es, therefore, it was assumed that operation would be carried out under such conditions that carbon-carbon linkages would not be ruptured (cf. G-rosse and Ipatieff, loc. cit.}. This would imply for ex­ ample , that an Intermediate addition product of ethane and sulfur might be formed by hydrqgen bonding, for example, with, subsequent decomposition into hydrogen sulfide and ethylene* The calculation of equilibrium conditions was a very tedious operation. A relatively simple example will be presented here to illustrate the method of calculation used, and the results of all calculations presented in. summary form. For the purpose of this calculation, a basis of 100 volume (molar) units of ethane was used.. Equilibrium conditions Yfere then calculated when 100 units of ethane Yfas mixed with various quantities of sulfur, at different temperatures. Consider the case where 100 units of ethane and 50 units of sulfur (as monatomic sulfur, Sq) are mixed at 1250°K. We will calculate equilibrium in the system involving C 2H 6 , C 2H^, C 2H 2 , H 2 » H 2S> and S2 (the stable vapor form of sulfur at this temperature is the diatomic gas S2 ). Rupture of the carbon-carbon bond, and dehydro­ genation to elemental carbon will be assumed not to occur. (Without these assumptions equilibrium calculations v?ere found to be of no value in this system. Experimental work bears out the validity of these assumptions). Simplification is afforded by using abbreviations for the quantities of the above gases at equilibrium. To this end the following terminology will be used: Units G2h6 a"k equilibrium = y C2H4 a e C2H 2 = a H2 = ii H2S - x S2 = s For a starting mixture of C2H 6 a 100 s 2 = 25 The following material balances may he written. Y / 3 / a - 100 3y / 2e / a / h / x = 300 x / 2 s = 50 This yields three equations in six unknowns. Thi'ee addi­ tional equations are necessary to permit solution for the gas composition. These are obtained from, the reactions cited above: (X) C;jH6 -> h2 / c 2H Zl. (2 ) C 2H 4 -■> H2 / C 2H 2 (3) h 2s -*• h2 / is2 From Table X values of Zp for the above at1250 °Z are fouzd to be 11.30, 0.160, and O.O663 respectively. For reaction (1), the equilibrium equation is Kp = 11.30 = pC2H k 1^2 P°2 h 6 \ The pressure of any gas in a mixture is equal to the total pressure of the mixture times the mole fraction of the gas. In this case pC2H£, = * _§ _ ' T / e / a / h / x / ? •' S_

PH 2 = h /±. etc. Then, substituting above there is obtained Kn = 11.30 = . y£. Similarly for reactions (2 ) and (3) 0.160 •= fra e i

0.0663 =‘ h H

The three material balance equations and the three equilibrium equations■permit the calculation of the con­ ditions at equilibrium. It might be pointed out that any three independent chemical equations (and corresponding equilibrium constants) relating all the chemical compounds being considered could be used for this purpose. The three equations used were chosen as being the simplest that could be obtained, and do not necessarily represent the mechanisn of the reaction involved. The same results at equilibrium would be calculated with any other independent set of re- For simplicity in following the calculations in­ volved, the equations will be regrouped and numbered: y / e / a = 100 (l) 3y / 2e/a/h/x= 300 (2) x / 2s = 50 (3)

■ylT = 1:L‘30

-§~ = 0.160 (5)

= 0.0663 } 6)

These equations are not readily adapted to analytical solution, so a trial and error method will be used. Assume x - 49, b = 76, ^ = 225 From (6), /s'" = 49211.5 x 0.0663 = 0.611, s = 0.374 76 Assume e . - 70 From (4 ) y = 70x76 = 2.1 11.30 x 225 From (5) a = 70 x 225 x 0.160 = 33.2 ■76 e / y / a - 70 / 2.1 / 33.2 a 105.3 It is apparent that e:y:a = 70:2.1:33.2 From (1), e / y / a = 100, it follows that e - ?0 x 1QQ «• 66.4 105.3 y = = 2.0 a » = 3 1 . 5 These values satisfy equations (1), (4), (5), and (6), from which they were calculated. C A Substituting in (2) gives 132.8 / 6 / 31.5 / 49 / 76 / = 300 295.3 = 300 Tims the hydrogen content is too low or the degree of un saturation is too high. To correct this, a higher value for hydrogen (h) should be assumed. From (3) h / 2s = 50 49 / 0.74 = 50 49.74 = 50 Using the above values, j£-is calculated to be 225.3. Assume that h = 79, x = 49.3, ^ - 225 49.3 x 15 x O.O663 = 0.592, s = 0.350 79 Assume e .= 67 ' y = 67 x 79 = 2.08 11.30x225 a = 67 1 225 x 0.160 = 30.55 79 e/y/a=67/ 2 .OS / 30.55 = 99.63 Then e = 67 x 100 =67.2 99.55 y = 2.1 a = 30.7 Substituting in (2) gives

49.3 / 79 / 134.4 / 6.3 / 30.7 = 300 299.7 = 300 This is sufficiently close for the present purpose v- . 3 Using the above method, of calculation, the equili­ brium conditions were calculated over the useful range of temperature and composition. The results are summarized in Table II. It is apparent from the data in Table II that the presence of sulfur vapor effects unsaturation of the hydro­ carbons present. At lovf temperature the effect is to in­ crease the concentration of ethylene; at higher tempera­ tures acetylene appears at the expense of the ethylene, so that the ethylene concentration at equilibrium actually d.e- creases as the sulfur content is increased. The overall unsatu^ation however, increases with increasing sulfur content over the entire range considered. An interesting observation may be made in'the con­ centration of sulfur vapor at equilibrium. At both sulfur compositions, the sulfur concentration at equilibrium goes through a minimum as the temperature increases, then be­ gins to increase. The decrease in sulfur vapor as the temperature increases in the low range is the result of the increasing dehydrogenation of the hydrocarbons treated. At higher temperatures, the dehydrogenation effect is lost as-the sulfur is largely converted to hydrogen sulfide. Thermal dissociation of hydrogen sulfide occurs in this range, and as a result the sulfur vapor content increases. c - ft CO O LTNlTN mON K \f t- t f \ 3 0 CM m • * • • « LT\0_itK>K> CM I I I I I CM O O O O m f t CO ft m i LiN o I n o f t CD O CM ft'ft K ftCDNO ft ft CO CM O 4 C 0 -ft- UN ♦ ♦ • •P ft t I I « NNO CMp±gn p] » I • CM •H O N^CVftCD » CM 'f t t — O O f t l A O CO _ftO ft ft • m m I ft ft ft n o (M m m CM CM M © CO I I I 1 I • • • • « CM CM m o N O o no n onco c M p j-m O M CM C— ONCJNON H CMpj'LfVj3* 0 , NO > ft m NO 1 o f t im I m ooo cO -P i i m O 1 o jH m co H fl CM 1! o o CM f t o m m K o mftoo H CO © M f t f t K ftftfC M D— J t • • • • f t CM CM M X m It NO |xj • • • I! • o mft m ra O CO NO NO f t CM o o H pqJ-H Orinoco m 3 •H *•111 CM • * f t m CM 8 m •V ON ft O f t CM m f t NO CO £ o f t CM tt 0 o NO •V •\ ft o f t ftNQ CTNPt o _d-o o CM o NO_j-mCM CM W O ftfC O • • o • • * • o t • « < • © CM n * • • f t m f f t O ON ft CO o - f t coco o c - m ft ?H O o m o \c o o - CM-d-IN-e-NO cm C'- onno m H cM NO -ft’ 11 II ft W m conq > ft ra NO CM • • • NO NO o fACO f t 'NO m m m t — o H f t M O o no o no m w • • • • • H » • • • • ft •H CM o c o m CM ONOCO mcM CM ft ft NO ft ft H o f t O c - uncm o C--CM G?ft ft I H H ft ft ft ft ft ft ft ft ft ft ft ft ft ft H ^Crvd-O on ftCTvij-O ON _c±ON_d-o on •=4 “rGQ m r—n- 3co me—^ 3

To . Thermocouple Hood

Sample Sample P oint Point Reactor Furnace

Sulfur Vaporizer 1 —1 furnace tube by means of standard taper joints ground on the porcelain tube. These joints were tightly assembled, without a sealing fluid, and with use became so tightly sealed that disassembly was virtually impossible unless the glass were broken. All heating circuits were attached to variable voltage Yaritran transformers (500 watt capacity) except the exit winding used in early work. This vras' designed for and installed in a 110 V. circuit. Temperatures were measured by means of thermo­ couples attached to a student type potentiometer through a selector switch. An iron-constantin couple was used in the well in the sulfur vaporizer and c.hrome-alumel couples in the reactor furnace and tube. All couples had cold jun­ ctions immersed in an ice-water bath. Operation of the apparatus described was performed as follows: all circuits were turned on to bring the apparatus up to temperature. Initially each circuit was individually checked with an ammeter in the line to deter­ mine the relation of current.flow to voltage setting on the transformer. The maximum allowable operating voltage was clearly marked on each transformer for future reference. Sulfur vaporizer and furnaces temperatures were determined by the thermocouples; the Intermediate line heaters were adjusted to maintain a slightly higher current than FV supplied to tiie vaporizer. L ^^ • When the vaporizer was above the xaelting point of sulfur (23S°F.), solid sulfur was added to the funnel section in the addition line. The addition line heater current was increased to speed, the melting of , the sulfur added. As indicated above, a piece of asbestos board was held at the drain line exit to permit the drain line to become plugged with solid, sulfur. A low gas flow was established while the sulfur was being added. The control stopcock and needle valve were shut off, and the ethane tank valve opened. The needle valve was then adjusted to give a moderate rate of bubbling of ethane through the escape bubbler. The three way stopcock was turned, to connect the ethane . exit line from the flowmeter with the back pressure manometer. The control stopcock was then opened to establish a low flow of ethane through the system. So effort was made to meas­ ure this flow, but it was usually of the order of 20-50 cc/ min. Sulfur addition was continued until the desired lev­ el was obtained, as determined by the back pressure on the ethane. When the sulfur addition was complete, the current to the sulfur addition line was adjusted to keep the sulf 12? just above the and. the sulfur vaporizer tan- \ perature was gradually brought up to the desired value by [•i- > - p increasing the heating current. As the temperature in­ creased, a very curious phenomenon occurred. In the temperature range of very viscous sulfur, the back pres­ sure of the vaporizer decreased markedly; with a further increase in temperature the original back pressure was re­ stored. As the operating temperature for the experiment was approached, the ethane flow was adjusted to the desired value, 200 cc/min, (STF) in practically all experiments. Final temperature adjustments were made and carefully maintained. Prior to taking a gas sample, the temperatures, flow rate, and sulfur back pressure were carefully main­ tained at constant values for about thirty minutes. Sul­ fur back pressure was maintained by frequent addition of small quantities of sulfur to the vaporizer through the addition line. These additions were made as solid sulfur. Samples of the exit g&s; were then taken for analysis and, in certain cases inlet samples were taken between the vaporizer and reactor. The techniques of analysis are described later. To shut down the reactor, the current to all circuits was decreased to a low. standb2/: voltage and the ethane flow was stopped by shutting off the control stopcock, the needle valve and the tank valve. The three way stopcock between the ethane flowmeter and the manometer was turned ■bo admit air to the ethane line. This prevented sulfur creeping toward the flowmeter by slow solution of or re­ action with the ethane. Usually the sulfur was permitted to remain in the vaporizer,^ at a temperature just above its melting point. At least once a week however, the sulfur was drained and discarded. This sulfur was usually very dark in color and was discarded to avoid excessive carbon buildup in the vaporizer. No inert gas purge was used either before or after a run, and no difficulty was experienced with burning of ethane or sulfur in the equipment. There was no draft through the equipment, since (1) the apparatus was sub­ stantially horizontal, and (2) the sulfur vaporiser when * ' ~ \ filled served as a seal to prevent gas movement through the system. In early work it was found virtually impossible to maintain a rate of ethane flow from the gas cylinder through the needle valve. From a given setting, the flow would decrease to such an extent that bubbling would cease and the flowmeter reading drop off, or the flow would in­ crease so much that the liquid in the bubbler would be blown out. The needle valve became very cold, and so a heater was installed around the valve. This did not materially improve the control obtained. A pressure gauge was attached to the tank; the ga!'s was found to be at approx:- imately the critical temperature and pressure (ca. 600 psi) . A second gas cylinder was obtained, and attached, to the supply tank through.-a high pressure manifold fitted with a pressure gauge and the needle valve. The second tank was i filled to a pressure of about 400 psi from the supply tank. The supply tank was then closed off and the ethane fed from the lower pressure tank. This resulted in very satisfactory operation. When the feed tank pressure dropped to about 200 psi, it was repressured from the supply tank. CALIBRATIONS Only two flow calibrations were necessary in this study: The ethane orifice flowmeter and the sulfur level in the vaporizer. Both of these were very simple, and so will be dealt with only briefly. It was decided initially that the sulfur vaporizer should be operated with a high sulfur level, and that the sulfur content of the exit gas would be controlled by ad­ justing the heat input rather than changing appreciably the sulfur level. This would impose a reasonably constant back pressure on the ethane flowmeter. In order to cali­ brate the flowmeter under actual conditions of use, the following method was employed. The line between the flow-

V - , meter and the sulfur bubbler was disconnected and a Pre­ cision Scientific Wet Test Meter was interposed in the line. This meter was of construction with a displacement of 3 liters per revolution. The sulfur vaporizer was filled using the method described above, except that in this case the temperatureof the sulfur was maintained just above the melting point. The gas flow was adjusted to the desired rate, and the time for a given volume of gas to pass through the Wet Test Meter was noted. From the temperature and pressure of the gas as indicated on the Wet Test Meter, the volume of gas passing per minute (corrected to'760 mm and 0°C) was cal- V- P 0 culated. This was then plotted . the flowmeter differ­ ential, The data obtained are presented in Table I and in Figures 2 , 3, and 4 * The flowmeter readings presented are in half-inch units. < . - TABLE I FLOWMETER CALIBRATION DATA. W.T.M. Press, Flow Flow Rate Volume Temp. Inches Meter cc/min. C Liters) °C. Hg. Reading Time (STP) Tip #1 2.000 37*6 I.I4-OO 16,2. 0.600 28.5 0.90 20.7 2.56 210 3*650 , 32.05 2.900 “ £ 2.45 07750 25 O.89 9.60 5.51 12l|. 1 .1+50 . 33*50 0.800 21.00 1.150 25.I1. • O.89 12.50 7.01 11+9.5 3.350 31+.65 2.150 19._85 1.200 2 5 .1+ O.89 l4.80 6.51 I67.8 1.750 36.25 0.750 18.30 1.000 25.4 0.89 17-95 4.73 192.8 Tip #2 1.500 5O.IO 0.500 2.70 1.000 27.2 0 47.40 1.19 736 3.500 • 41.55 2 .500 11.25 1.000 27.2 0 30.30 1.52 576 V- . W.T.M. Press. Plow Plow Rate Volixme Temp. Indies Meter oc/mln. (Liters) ° C. H r '. Rdg. Time (STP) Tip #2 ' : 1.600 _ 38 -i+o 0.600 14-25 1.000 27.6 0 24.05 1.83 477 3.900 34-95 2.900 i _7_.8o 1.000 27.6 0 17.15 2.24 390 2.200 32.60 1.400 20.20 0.800 27.6 0 12.40 2.20 318 j Tip #5 1.550 31.05 0.050 21.70 1.500 ■ 25 0 9.35 2.76 482 11.100 - 32.70 2.600 20.10 17500 25 '0 12.60 2.32 574 3.500 36.50 2.000 .. 16.20 1.500 25.2 0 20.30 1.81 735 4.100 40.00 1.600 12.75 2.500 29.0 0 27.25 2.54 854 4.900 43.15 1.900 9 .65 3.000 29.0 0 33-50 2.70 964 3.100 48.95 /, ; 0.100 3.85 37000 29.0 0 45.10 2.32 1120 ,

3.700 49 *00 0.700 3.80 3.000 29.0 0 45v20 2.26 1150 OQ 1. I The figures in the last column of Table I, ex­ pressing the flow rate in ce./min. (STP)wer© calculated in the following manner: (Data from the first calibra­ tion point on Tip #1 are used to illustrate the method of s calculation) Yol. flow - 600 cc. Y.P. Ii20 @ 26.5°©- - 29.5 rrnn. Hg. Static Pressure -0 .90” Hg. = 23 mm. Pressure CpH6 - 760 /23 -29.5 = 753.5 Yol. G2H6 XSTP) - 600 x 753.5 x 273 = 539 cc. 760 301.5 539 Plow rate - 2.56 ='210 cc./min. (STP) The above data were plotted in working plots of the type illustrated in Figures 2y 3 , and L. In practice, a flow rate was selected for an experiment and the flowmeter setting determined from the calibration curve. The flow- meter was then accurately set to this flow and the fluid levels in the manometer was indicated by means of index markers. This permitted a quick visual check of the flow­ meter to be made at frequent intervals without the necess­ ity of reading the scales and subtracting to determine the flow rate. In addition to the above, the sulfur vaporiser was also calibrated to obtain a relation between the sulfur contained in the vaporizer and the back pressure across the vaporizer. This was done by establishing a low flow of ethane through the vaporizer^, adding small incremental quantities of sulfur, and noting the back pressure pro- Flow Rate 190 200 170 180 210 Differential Figure Figure FI glare Ohio State Univ. State Ohio Dept. Engr.Chem. cc/min, C2Hk of rate Flow ferential (STP) July July Eng.Chem. TipNo. Eugene Mornings Mornings Eugene tar 2 12 s v 2 , 1 . . TET 19 Flowmeter Dif*- Flowmeter

95 52

^ Figure 3* Flow rate of C2H£ cc/min (STP) vs. flowmeter diff erential Tip No. 2 Chem. Engr. Dept. Ohib State Univ. Chem. Eng. 9$0 July 12, 1952 Eugene Mornings tar

16 20 Differential ^ Figure 3 01 1200

1100

1000

000

Poo

700 Figure it. Flow rate of ethane cc/min ( STP ) vs. flowmeter differ­ ential . Tip No. 3 Chem. Engr. Oeot.. «56oo Ohio State Univ. BS Chem. Eng. 9 5° o July 12, 1952 r-H fX4 Ergen e Mo m in gs tar

500

l+oo 50 ho 50 1Q Differencial Figure h d.uced. T.he sulfur was added from a fared, container, the q u a n t i t y added "being obtained by difference. The data observed are presented in Table II and. Figure 5. The back pressure readings, as in the case of the flowmeter differentials, are in half'-inch .units. TABLE II CALIBRATION OF HOLDING CAPACITY OF THE SULFUR VAPORIZER

Rinfirp Tara Sulfur Added Back Pares sure

664*3 g. O O 632.0 32.8 g. 1.63 units 600.0 64.3 2.85 569.2 9 4 .6 4.10 541.1 123.7 5.21 517.3 147.0 6.19 484.0 180.8 7.53 447.1 217.7 8 . 8 6 414.6 250.2 9.93 387.6 277.2 10.81 364.2 300.6 11.52 345.1 ’ 319.7 12.00 327.2 337.6 12.37 298.5 366.3 12.98 264.7 4 0 0 . 1 13.57 227.0 437.8 14.30

These data are presented in graphical form in Fig. 5 Sulfur in Vaporizer Uoc U 5 Q. Ohio State Univ. State Ohio uee Mornings Eugene tar iue . | Eng.Chem. Dept. Engr.Chem. 5.Figure Sulfur in vaporizer ( vaporizer in Sulfur gm. j ) July vs. back pressure (in.E^O) j (in.E^O) pressure vs. back 12 , 1952 950

Back Pressure Back iue 5 Figure 10 B & *■. . * * , ANALYTICAL METHODS It was necessary to develop’ a method of analysis for the gas mixture expected, as well as to determine the products that must he analyzed. The following list of> products of reaction was considered to include the likely materials that would he obtained. In the system C2H£-S2«

C2H 6 c2h 4

c 2h 2

CHll h2 h2s c s2 '

C2H 5SH o2h 5-s-g2h 5

c2h '5-s -s -c 2h ^

S2 The first analytical study carried but was the analysis for the hydro carhons and hydrogen expected. To this and a mixture was prepared containing approximately 22# C2H£, 22# C2H^, 22# C2H2 , 12# H2 and 22# Ng. This mixture was prepared and stored over water, hut solubility effects altered the final analysis of the gas. Various absorbents were ased to separate the components. A meth­ od of analysis was finally evolved as described below. 1 . Analysis for acetylene: In order to distinguish between acetylene and ethylene j, the acetylene was dissolved in iodo- ' mercurates' solution as described in the UOP Manual.

UOP Laboratory Test Methods for Petroleum and its Pro­ ducts, Union Oil Products Co., 19^ 0 .

The adsorbing solution was prepared as follows: 50g. of potassium Iodide crystals, 20g. of mercuric chloride, and Ip g. of potassium hydroxide were dissolved in 7&S* of water. Various me thods of solution prepara­ tion were tried, but the concurrent dissolving of all materials was found to be as effective as any.. When all the solids were added., an orange colored precipitate, presumably of mercuric oxide, was formed. With continued agitation, this precipitate slowly dissolved to form a clear yellowish solution. The potassium iodomercurate solution was used In an autobubbler pipet in a portable type Burrell Gas analyzer. Absorption of the acetylene in this solution was found to be very slow under these conditions. Var- ious methods were used to facilitate the absorption, in­ cluding the use of contact pipets and the use of glass packing in the autobubbler pipet. This did not materially increase the absorption rate. The UOP method of absorp­ tion specified that the gas sample be transferred to the potassium iodomercurate pipet, then the pipet should be disconnected from the apparatus and be shaken for three minutes, then reinstalled and the residual gas volume mea­ sured. Two such absorptions were specified, although in the ncrmal gas analysis routine of absorbing to a con­ stant volume more than two absorption periods would have been required. The possibility of gas leaks during re­ moval and re installation of the pipet from the gas analysis apparatus was recognized, hence this procedure was not used. Instead, the slower bubbling absorption was used. The absorbing solution was changed regularly to reiain as high activity as possible. The use of cuprous chloride was tried in early work for acetylene absorption; in this case ethylene was found to be incompletely dissolved In the cuprous chloride, making-quantitative separation Im­ possible. 2 . Analysis for ethylene: Two alternate procedures were considered: solution in strong sulfuric acid or In bromine water solution. The ,r*o ... • J latter adsorbent was used on the basis that very rapid and complete absorption was obtained. Xn the normal analysis using bromine water as an absorbent, the gas is bubbled repeatedly through the bromine water to saturate the gas with bromine vapor. When an aqueous confining liquid is used in the gas buret, as was the case in this study, the bromine in the gas dissolves, in the confining solution. To avoid subsequent evolution it is necessary to add a to the confining liquid to reduce the bromine. However, this liberates HBr ( an, acid confining solution was used) which had to be scrubbed from the gas after bromine treatment,. Thus, even though in the normal course of analysis the bromine absorption step is followed by adsorption of the residual bromine in sodium solution, bromine sub­ sequently desorbed from the confining liquid, and the presence in certain cases of HBr, render the results using this method somewhat variable. A much more precise and convenient method of mani­ pulation was devised for bromine absorption. Ten cc. of liquid bromine was added to an autobubbler pipet on the Burrell gas analyzer. The pipet was filled with water to which a few crystals of potassium bromide were added. This pipet was installed on the analyzer adjacent to a W sodium sulfite absorber. The surge reservoir on the bro­ mine pipet was connected to a second reservoir filled with saturated salt solution. Ihe second reservoir was fitted with a levelling bottle. A stopcock was installed in the glass line between the pipet surge reservoir and the auxiliary reservoir so that the confined air volume could be adjusted. In use, the gas sample is transferred from the measuring buret in the ]gas analysis apparatus to the *. bromine pipet in the usual manner. As the gas sample is transferred, the air confined between the reservoirs is forced into the auxiliary reservoir. After the sample is completely transferred to the pipet, the stopcock is closed. Ihe gas is permitted to stand about a minute to permit the bromine and ethylene to react. The stopcocks are then turned to connect the bromine pipet with the. sodium sulfite pipet, and the gas sample bubbled through the sulfite solution to absorb the residual bromine. The levelling bottle on the auxiliary reservoir is used to force the gas into the sulfite pipet. The gas sample is then transferred from the sodium sulfite pipet to the measuring buret and the gas volume noted. This procedure is repeated 'ontil constant volume is obtained. In prac%i cally every case, two absorptions were sufficient. Some difficulty was observed in early work due to liberation o f HBr from the sodium sulfate pipet by the reaction Na2SO^ / Brz / H2° *"> Wa2S0j+ / 2 HBr This was prevented in later work by making up the sulfite solution with two moles of per mole of sulfite. This resulted in a rapid and reliable method of analysis. 3. Analysis for hydrogen: Hydrogen was determined by oxidation to water by means of copper oxide at 285°G. in a Burrell combustion unit. Loss in volume of the sample on oxidation re­ presented the hydrogen present. The copper oxide was re- oxidi^ed by permitting it to stand overnight in contact with oxygen at 285°G. 4. Analysis for ethane and methane: The final step is the analysis for saturated hydro carbons. Oxygen is added to the gas and the mixture is passed over a special Burrell slow combustion catalyst to effect the reactions °2h6 / 3ir 02 - > 2. co2 / 3 h2o CH^ / 2 02 -> C02 / 2 H20 The volume after combustion is measured. The 002 formed is then absorbed in caustic and. the volume determined again. Prom the two volume changes, the volumes of ethane and methane are calculated. The usual method of calcula­ tion involves the solution of two simultaneous equations. However for this study a quick method of calculation was devised. Let A = volume change on combustion ! B = volume change on KOH absorption m - volume of methane in sample y = volume of ethane in sample. A —”2 m / 2 . 5 y B = m / 2 y Solution of these equations for any observed values of A and. B willyield the volumes of methane and ethane in the sample. For pure methane, it will be observed that A/B equals 2 , and for pure ethane, I.25. For mixtures the ratio is given by A/B - 2m / 2.5 y m / 2. y = 1.25 m / 2.5 2T / OF75_m m / 2 y “ m / 2y 1.25 / 0.0075 c— 8 ) m / 2y fT r*n / .. Hero ,rkP.Q ,.*** represents the percent of the 0Go formed on m / 2y 2 combustion which was foimed from methane. Plotting the percent of CO^ formed from methane vs. the ratio A/B results in a straight- line through the points (1.25,0) and (2 ,100). Such a plot is shown in .Figure 6. An ex­ ample of use of this chart is as follows: for A - 33*1 cc* B - 2lj..5, A/B =-1 .562. From the chart, at a ratio of 1.562, the % GO2 from methane is found to be 1)4.9/. Bence: CH]^ in sample = 2i|.5-x O.IJ4.9 “ 5-6 cc

g2H5 - (2)4.3 - 5.6) x 1/2 B 10.35

This chart was found to be v e r y convenient in the routine calculations invplved in this study.

Al second and much more extended study was devoted to determining the sulfur gases present and a suitable method of analysis. In early work hydrogen sulfide, assumed to be the principal sulfur gas, was determined by iodine titration in a Tutweiler buret. The Tutweiler buret Is a gas holder fitted with a small iodine buret to permit titration with­ in the gas holder. The iodine buret is connected to the gas holder through «■ three way stopcock on what might be termed the top of the Tutweiler. The other arm of the - i" stopcock Is used to draw in a gas sample. The bottom of the Tutweiler carries a two way stopcock to which may be U&-.ft* ^t--.- ^ attached, a levelling bottle. In use the iodine buret and the corresponding barrel In the stopcock plug Is filled with standard N/lO iodine solution. The gas hold­ er Is flushed to remove any Iodine In the gas space, then it is filled with s tarch solution. A sample of gas Is drawn into the gas buret by starch solution displacement until a volume in excess of 100 cc* is obtained. The upper stopcock Is closed, and the gas In the buret com­ pressed to exactly 100 cc. The lower stopcock Is then closed. (The Tutweiler has a 100 cc. calibration mark in a small neck near the bottom). The upper stopcock Is then opened to the atmosphere momentarily to allow the \ excess pressure inside the buret to be relieved and the stopcock Is closed. The. bottom stopcock Is opened and the levelling bottle lowered to draw a partial vacuum Inside the buret. The, stopcock Is closed and the connecting rubber tubing Is removed. ; Iodine is added In small Increments to the gas through the top stopcock, and the Tutweiler is vigorous­ ly shaken. Titration to a permanent blue end point gives the hydrogen sulfide present. The end point is preceded by a red coloration which slowly changes to blue with sub­ sequent addition. This makes determination of the end point more troublesome than in the ordinary iodine titration, "but very reproducible results are obtained* In early work the hydrogen sulfide was titrated with iodine in the above manner, and the residpSil gas !; was transferred to the gas measuring buret of the Burrell gas analyzer. It was found that the residual volume of the gas was less than that calculated by subtracting the titrated volume of H2S from the total sample ■volume. This unaccounted difference appeared to be related to the vol-, ume of H2S in the gas sample analyzed. The observations of several runs are shown in Table III. COMPARISON OF UNACCOUNTED GAS LOSS TO HYDROGEN SULFIDE CONTENT OF SAMPLE

Run No. Vol. H2S - Vol. Loss Loss/^S

1 9*1 1*8 0.20 2 2 8.4 2.5 0.30 3 2.7 Q.3 0.11 4 5*1 1*1 0.22 4-i 10.0 2.7 - 0 . 2 7 5 12.2 2.1 0.17 6 23.5 5.7 0 . 2 4 6-1 11.1 3.6 6.32 Average O.25 / 0.04

The reasonably cons tent value of los S / H 2 S sug­ gested the possibility of mercaptan formation in\the reaction, since mercaptan reacts with iodine as follows: 2 RSH / I2 — RSSR / 2 HI (cf. H2S / I£ — S / 2 HI) It is seen that one mole of iodine Is equivalent to one mole of hydrogen sulfide, but to two moles of mercaptan. Thus If the gas titrated w^Lth iodine were actually all mercaptan the calculation as hydrogen sulfide would yield a calculated volume only half the actual gas Volume ti­ trated. This suggests the presence of mercaptan in .the gas samples described above. In order to distinguish between hydrogen sulfide and other sulfur bearing gases, a new technique was de- - iL - ' vised. In this case a sample of gas was taken in the Burrell gas analyser and analysis for hydrogen sulfide was accomplished by absorption in acidified 10^ solution. Other sulfur gases (mercaptan, car­ bon disulfide, etc.) were then dissolved in strong caustic. Subsequent analysis for hydrocarbons was. made using the techniques described above. The results obtained using the fiadmium chloride-caustic method of analysis are given in Table IV. TABLE IV Analysis for Sulfur Gas Components Run Ho. CdCl2 Absorption Subsequent KOH Abs. ’v — \ 13 4-2 0.1 il|. ‘ q.. 5 o « o • 15 10.6 0.0 16 65.I1. 0.1 17 35-5 0.1+ 16-1 30.9 0.1 Prom the data presented In Table IV, it appears that the only sulfur gas present In the sample studied was hydrogen sulfide. Other gases, such as mercaptah or carbon disulfide, which are insoluble in acidified cad­ mium. chloride and soluble in caustic, were present in very small quantities. , During the course of the above study it was ob­ served that absorption of the gas In cadmium chloride solution was rapid Initially, but that a small final volume of gas was only slowly absorbed. This suggested that a gas might be dissolving physically in the . Analysis of the observations made indicated that acetylene might~be dissolving in the cadmium chloride solution and that in early experiments acetylene solubil­ ity in the titrating solution in the Tutweiler was re­ sponsible for the loss observed. . The observed correla­ tion of loss with hydrogen sulfide shown in Table III appeared to be due to two factors: (1) as the hydrogen sulfide content of the gas iricreased, the acetylene con­ tent increased, and (2) as the hydrogen sulfide increased, the liquid present in the Tutweiler increased, permitting greater solubility. ' As a result of the above tests, a series of analyses was made in comparison of methods of hydrogen sulfide analysis. The methods studied were the Tutweiler analysis, acidified cadmium chloride absorption, and absorption in 50$ potassium hydroxide. The results of this comparison are given in the’'following- summary. Time of Sampling Method of Analys is %B.2 S

11:55 -6dOl?* 39-7 11:51 KOH-' 43.6 11:58 GdClp* , 45-1 12:00 12 ' 4 2 . 0 *-In this s tudy, only- the initial rapid absorption of HpS was used to calculate the percent HpS in the sample. Subsequent slower absorption presumably of acetylene was not used in the calculation.

The variation in the percent hydrogen sulfide was attributed to normal gas variation rather than to dif­ ferences in method of analysis.' This gas variation is best shown by the two cadmium chloride analyses. Xt appe ars then, that hydrogen sulfide analysis can\be satisfactorily performed using any of the above methods.

Because of the low solubility of acetylene in 50% potassium hydroxide, this reagent was used for hydrogen sulfide sinalysis. On the basis of the above tests and observations on known and comparative samples, a complete method of analysis was devised. The complete method is very briefly summarized, below: Apparatus Used: - Burrell portable gas analyzer equipped with four autobubbler pipets, and with cc^mbus tion units for hydrogen and hydrocarbons. The apparatus was all glass -f fixri h: ..p •with three-way stopcock manifolds of the flushing type. The manifold was extended beyond the heaters, and oxygen, i ,, and vacuum (rubber aspirator bulb}, were con­ nected to the manifold. The oxygen and nitrogen con­ nections were made to glass reservoirs of the two gases. These reservoirs were equipped with head reservoirs to keep the contained gases under a positive hydrostatic pressure of about three feet of water. The reservoirs w were filled periodically from compressed gas cylinders. Two closed stopcocks were used in each line to assure that the compressed gases would not leak into the gas analyzer during an analysis . The apparatus was also equipped with an. extra reservoir and levelling bottle on the bromine pipet to transfer the gas from the bro­ mine to the sulfite pipet directly. This was described in greater detail earlier under "Analysis for Ethylene". Prom right to left on the gas analyzer (away from the gas buret) the pipets were filled with the following reagents respectively. 1 . 50^ KOH 2. K 2 HgIi,. 3. Bromine Water i].. l\ra2S05.2Na0H v Saturated sodium chloride solution, containing about hydrochloric acid and a few drops of* methyl orange indi­ cator, was used as a confining liquid in the gas buret. All stopcocks were lubricated with silicone lubri­ cant except the stopcock on the gas buret, in which vaseline was used. All the pipets were filled with ab­ sorbents and the levels, brought up to a point -just below the rubber connection below the'Tmanifolds. The level of the confining solutions in the gas reservoirs below the combustion units was brought up to the stopcocks. Nitrogen was drawn into the gas buret and bubbled through each reagent in order, then was passed through the copper oxide combustion tube. The nitrogen was then rejected* A sample of gas was purged through the rubber con­ nection between the sample point and the Burrell. The tube was then attached to the Burrell, and a sample of gas in excess of 100 cc. was slowly drawhninto the gas buret. The stopcock was closed, and the gas sample was compressed to exactly 100 .cc, The excess pressure was relieved by momentarily opening the stopcocks on the gas buret to the atmosphere. The stopcock was closed, and the volume of the sample observed. Analysis was made for H£S, C2H2, H2 , C2H 6 C H i}, in ^ ^ t order by the absorption and combustidn techniques described 1 ' ' 5 above. After hydro gen analysis over copper oxide, oxygen was admitted first to the copper oxide tube and reservoir, then to the gas sample being analyzed. Thus, the the copper oxide was regenerated after each analysis. The calculations made in this study were quite simple;

» sample calculations are presented to illustrate the meth­ ods employed. 1 . Calculations used for Tutweiler analysis for H^S, Burrell analysis for hydrocarbons, Run #2 . N.B. In this run diluent nitrogen was added to the furnace exit gas to dilute the sample being analyzed. This practice was later abandoned as being of no value. 2 . Calculations used for caustic absorption for H^S analysis• In the final method of analysis, using absorption In caustic rather than titration with iodine, the cal­ culations were much simplified. In this case the gas values were no t corrected to s tandard conditions, as long as the temperature of the gas sample did not change during analysis. Where appreciable temperature change occurred, a cdrrectlan was applied to correct all the volumes to the same temperature • The calculations for Run #2 lj. are pre- sen ted as an example• ICS CALCULATIONS FOR RUN #2 . (For Experimental Data see Table VIII)

Sample Vol. - 100 cc ( aat’d. with. H20 ) @ 29°C. Ig Used 7*9 cc. x O.O953N Residual Gas 87-4 co- KOH Absorption 87-4 K2HgI^ 83.7 Br2 81.2 OuO combustion 79-6

Reject Portion of Sample: Sample Remaining 50.75 cc. Vol • Shrinkage COp Oxygen Added to 99.2 Combustion . 62.2 37*0 \ KOH Abs. 35-8 2 6 . 4 Comb. 3b - h 1.4 KOH Abs. 31 3.0 O

*4 H Comb „ 31.3 • KOH 3 1 . 0 0 . 3 To tals 38.5 2 9 . 7 v.p. H20 @ 2 9 0 0 a 3 0 m . Vol. Sample (STP) - 100 x x 15Q-3 .Q = 8 6 . 7 HpS - 7.9 x 11.2 x O.O953 - 8.4 Residue s 8 7 . 4 x O.867 = 75.8 Sample Vol. H-S = ,86.7 - 8-4 * 78.3 Unaccounted Gas Lo ss ( cf*. Table III) - 78.3 - 75-8 = 2.5 c 2H2 = 3 - f x O . 8 6 7 = 3 . 2 CpHR = 2.5 x O . 8 6 7 = 2.2 h2 = 1.6 x 0.867 = 1 . 4 \ " t " p From above, Comb. Shrinkage - 38.5, C02 = 2 9 . 7

Ratio - * 1*297 Percent C02 from CH^ = 6 *5$ (From Figi. 6)

CH[|_ in total sample » 29.7 x 0.065 x 7.^ x 0.867 » 31.2 W 2 by difference - 3 3 . 7

Summary: HpS8 . If. cc Equiv. Ho r 8.4 c | h 2 3.2 3.2 C^Hf" . .2.2 Ij. .if. Hp 1 •!+ 1.4 CH)..1 q..4*1 x o8.2 .

H->S s 8.4 2c '^?o: o = 4l cc./min. CpH2 16,V Hp ^ 6 CSLl 20 c 2h 6 150 i " 3 CALCULATIONS FOR RUN #2lj..

(For Experimental Data see Table VIII) J

H2S s 100.0 - 69.6' r 30 .Ip c 2h 2 = 69.6-55.6 » 16.0 C 2H ^ r 55.6 - 50.8 S 2.8 h 2 - 50.8 - 17.6 - 33.2 From Table VIII Combustion shrinkage = 32.6, C02 = 16.2 Ratio sr ^g jjr “ 2.0 j QHj| as 10Q/& x C02

Summary: H„S - -^O-.lf. cc Equiv. H2 = 3 0 .I4. CpHn 1 6 . 0 16.0 C2Hj, 2.8 5.6 h „ ^ . 33.2 33.2 \ CHJ, 16.2 32.k . ' Total Equiv. H2 = 117. 6 On the basis of a hydrogen balance, and with an inlet ethane flow of 200 cc/min., the following output flows are calculated. H2S = 5 0 . J4. x / = 1)4.8 cc/min. C 2 H 2 1 1 7 . 6 78 c 2 h , : 13 h2 ^ 1 6 2 c h ^ 79 These methods of calculation are typical of all calculations used in this study.

I 100

90

80

?o

Figure 6. Combustion analysis Calculator. Percent of* COp from CHi vs. shrinkage: C0'5 ratio., Chem. Engr. Dep t. Ohio State Univ. Chem . En gr. 9 50 July 12 , 19 52 Eugen e M o m in gs tar o

20

10

1.7 Shrinkage sC02 J* -HI , • EXPERIMENTAL : "

In general the experimental work in this' study was designed to cover the useful range of* temperatures and sulfur vapor concentration, with a limited amount of study devoted to space velocity effects and to the use of catalysts. In all these experiments, tank ethane was used. The ethane was obtained from the Ohio Chemical and Manufacturing Co., Cleveland, Ohio, The analysis of this gas is given below. One series of experiments was also performed in which ethylene was reacted with sulfur, in an attempt to study the partial reaction. In this series also, * the useful range of temperatures and sulfur vapor con­ centrations was studied. ' ■ I f ii E F F E C T OF TEMPERATURE AND SULFUR CORCEN IRATION: This series of studies was the most extensive conducted. This was due at He ast In part to the fact that during this series analytical methods were "being continually studied and developed, so that more exten­ sive rechecking was done. In addition, it was neces­ sary to become familiar with the operation of the apparatus, and to determine by experiment the trans- ^former settings required to obtain the desired tem­ peratures throughout the apparatus. A constant ethane flow was used in .this series of experiments. A flow rate of 200 cc./min. (STP) of ethane was chosen for this study. ¥fith a reactor volume of 180 cc (see Section on Construction and Operation of Equipment) an average temperature of llpOOOF. , hind a 100% increase in volume with reaction, the contact time In the reactor would be 7*2 seconds. This Is considera­ bly longer than the contact time used in commercial thermal cracking units. The longer contact time was chosen to avoid the necessity for very rapid heating and the uncertainty in the maximum gas temperature act­ ually obtained. It was one of the objects of this study, of course, to arrive at a process that would not require the close control of conditions, of temperature and contact time used in commercial cracking units. The range of reactor temperature used in this first series of experiments extended from 1030°F. to l8lO°F. with studies in the range from 1250°F to 1700°F. At the lowest temperatures used, little reaction occurred; at l8lO°F. on the other hand, so much carbonization occurr­ ed that the narrow inlet end of the combustion tube be­ came plugged with carbon. The range of sulfur vapor concentration, expressed as outlet flow of hydrogen sulfide formed, extended from no sulfur to an equivalent of one mole of hydrogen sul­ fide formed per mole of input ethane gas. In certain cases slightly higher concentrations were used. In one experiment in which almost two moles of hydrogen sulfide were formed per mole of Inlet ethane, excessive carbon­ ization also occurred. As a corollary to the above study, gas samples were taken at the exit of the sulfur vaporizer to determine the reactions occurring in the vaporizer. Analysis of these samples were made using the techniques described above. Analyses were also made of the input ethane gas. The data obtained, in these studies are summarized in Tables V and VI. ■“f ■'f J ~ ...... TABLE V INPUT GAS ANALYSIS

Tank No, 1 TankNo, 2 Average C2H6 88.0# 88.3# 88.2# CHjtf. 6.2 6.3 6.2

C2H* 5.0 3.6 4.3

H2 0.8 1.8 1.3

Moles G p e r 100 moles gas - 1 9 1 .2

Moles H2 p e r 100 moles gas — 2 8 6 .9 TABLE VI OBSERVED ANALYTICAL DATA, VAPORIZER EXIT GAS Rian Nos, 4-1 6-1 8-1 9-1 16-1 39-1 Al-I Input Flow (cc/mln) 200 200 200 200 ;200 400 1150 Sample Volume 100*0 100.0 100.0 100.0 97.3 100.0 100.2 Sample Temp* °F 39 89 89 89 92 83 *2 9.4 10.4 18.3 29.5 — — . — Residue 85.2 82.3 72.0 47.3 — (86.5s) (87) CdCl2 —— ■ — — 69.1 * — KOH 85.2 82.2 71.9 47.0 69.0 88.2 97.8 KaHgl* 80.5 75.5 67.1 40.3 59.0 82.6 98.0 Br2 77.4 75.0 61.6 38.5 56.5 80.0 94-7 CuO 76.0 73.6 59.3 37.8 55.5 80.0 93.3 Vol.after rejection 25.3 22.45 26.2 17.12 17.85 18.3 18.5 After Cg. Addition 95.3 99.5 98.1 99.5 101.1 97.6 98.8 Combustion 52.2 46.4 55.1 59.4 60.9 51.7 54.8 KOH 20.2 8.2 16.0 29.9 30.7 17.6 22.6 Og Addition 49.1 38.5 51.7 — —— — Combustion 30.0 36.0 31.0 26.2 25.2 16.4 19.4 KOH 15.9 34.0 21.9 22.3 20.4 14.0 17.0 02 Addition 50.3 — — — — — — Combustion .48.5 33.6 . 19.8 21.9 19.1 13.7 16.8 KOH 47.0 33.4 18.0 21.2 18.0 13.3 16.2 Combustion 46.6 — 17.0 20.8 17.6 — 16.1 KOH 46.4 — 16.5 20.6 17.3 — 15.9 Combustion — — 16.2 ———— KOH - 15.9 — - - .. - i ■ -** Er '... .1.. 3 Prom the data In Table VI, based on the Input flow rate and the exit gas analysis, the flow rates of the gases leaving the vaporiser were calculated. In addition, the flow rate of sulfur was calculated from the hydrogen sul­ fide flow in the flirnace exit gas. These data were then used to calculate the s ulfur/e than e ratio In the inlet gas, and the percent conversion of ethane to ethylene and acetylene. These dat^ind calculations are summarized in Table VII, and presented graphically in Figure 7* Percent Conversion S:C Eugene Mornings tar Mornings Eugene ethylene to ethane f o conversion Percent Figure hm Eg. pt, ho tt Univ, State Ohio 950 Eng. t., Chem. ep D Engr. Chem, n aeyee n aoie vs SCH rato. , . 1952 tio a r S:CpH^ 12, s. v July vaporizer in acetylene and 2 O T Hg.Ratio 7 .

Figure Figure 7

TABLE VII

VAPORIZER OUTLET FLOW RATES AND CONVERSION OF ETHANE TO

ETHILENE AND ACET2LENE IN VAPORIZER « Input Run Flow Rate __ Output Flows (cc./min.) Ratio No, (cc/min.) C2H4 CaHg CH* S(vap) 0 2 % CaHg M M l 1 J 6 l M i l Ik M t. S:C2H6, 4-1 200 29 152 7 12 44 3 5 0.19 4.0 6.8

6-1 200 33 140 1 20' 63 3 143 1.00 0,6 11.3 8—1 200 57 138 15 13 28 4 4 0.35 8.5 7.4

9-1 200 125 117 6 24 29 2 35 0.91 3.4 13.6

16-1 200 74 139 14 26 20 3 331 2.30 7.8 14.7

39-1 400 52 315 8 27 29 0 346 1.13 2.3 7.7 a - i 1150 24 1009 37 _ 83 11 954 0.96 3.7 ..j It may be observed, from the data presented in Table VII that some dehydrogenation of the ethane oc­ curred in the vaporizer, the extent was not great, how- . e v® r • The gas from the vaporizer exit was passed through the reaction zone, and the final product was analyzed as described. The analytical data observed are s^^mmar— ized in Table VIII, and the exit flow rates and percent conversions to ethylene and acetylene are presented In Table IX. In addition, the conversions to ethylene and acetylene are presented graphically in Figures 8-1 1 . In Figures .8 and 9 ’^rle conversions of ethane in the in­ let gas to ethylene and acetylene respectively are pre­ sented as a function of sulfur to ethane ratio in the inlet gas with temperature as a parameter. In Figures 10 and 11 the conversions to ethylene and acetylene respectively are shown as a function of temperature with the sulfur ethane ratio as parameter. Figures 8 and 9 were prepared using the experimental points given in Table IX. Points from the curves of Figures 8 and 9 were used to prepare the curves of Figures 10 and 11, since data at constant sulfur-ethane ratios are not readily available from Table IX* T A B L E V I

OBSERVED ANALYTICAL DATA, FURNACE EXIT GAS, TEMPERATURE AND SULFUR CONCENTRATION STUDY

Run No, 1 2 5 [j. 6______l 7-A 8 9

Sample Vol. 100.0 100.0 100.0 100,0 100,0 100.0 100.0 100.0 100.0 100.0 Sample Temp. °C. 29 35.5 30 29 51.5 31 92°f . 91.5 89 87 b 8.8 7*9 2.3 Ip.8 11.4 22,0 0.9 0.9 10.9 24.3 Residue 87.l1 87.4 93.9 93.7 83.4 65.7 98.0 85.7 63.8 KOH Abs. 8 7 4 87.V 93.cL 93,6 83.1 65.5 97.O 97.8 85.0 63.5 K2Hgl4Abs. 83.I 83.7 88.9 89.O 78.9 ^6 4 97.0 97.6 81.4 56.3 Br2 A d s . ■ 83.2 8l .2 87.8 7 L 2 62.0 [4.2 94.0 94.8 77.1 52.0 CuO Comb. 80.0 79*6 54*4 50.2 54.8 394 92.2 92.7 75.6 50.6 Vol. after Rej. 29.8 36.8 29.7 294 27.8 39.4 28.5 28.1 29.3 26.5 Oxygne Addition 99.0 99.2 95.1 96.5 96.8 99.1 99.2 9 7 4 99 *7 9? 4 Combustion 6k.6 62.2 70.2 80.4 77*0 78.8 68.3 67.4 69.1 74.0 KOH Abs. 39.O 39.8 57.8 7-1.2 68.6 70.4 45*8 45*3 46.2 55*8 Combustion 37*2 34*4 85*9 68.6 65.8 67.6 1|4 • 6 |4*3 |4*4 54*2 •KOH Abs. 3ii.0 31.lL 5I1.2 66.8 64.7 66.1 42.8 42.6 42.5 52.5 Combustion 35*8 31*3 54*0 65.7 63.6 64*9 42.8 42.2 42.3 52.2 KOH Abs. 33.2 31,0 53.8 65.3 63.2 6Il4 42.6 42.2 42*0 52.2 Combustion 68.0 62.8 63.9 KOH Abs. 64.9 62.8 63,9 TABLE VIII (Continued)

Run No. 10 11 12 lk 16 15 1? ' 18 1?

Sample Vol. 100.0 100.0 100.0 100.0 99 *9 100.0 200.3 100.0 100.1 100,0 Samnle Tenro 88.5°P. 9'Q 5 87 Q2 87 Qk 86 93 94.5 s ♦ 73 12 4.7 8.0 28.1 W* mm Residue - 93.il 89.1 58.2. - ■a CdClp Abs. mt \ 99-4- 95.5 89.4 3,4*6 64*5 KOH Abs. 93-4 57 A 99.2 95-3 89 -4 34*5 64*1 55.4 100.0 K2HgIji Abs. 91.2 45.2 98.7 9 4 4 88.4 29.3 57*0 97.6 Brp Abs. 7k ’2 36.0 71.8 72.5 72.0 17.6 40.9 ’7 01.0 CuO Comb. W 5k.k 30.5 45-6 33*0 56.5 15.0 34*0 32.0 41.3 Vol. after Re3. JO,2 50.5 19.0 19.0 l6.8 15.O 22.6 18.2 19.8 Oxygen Addition 101.2 lOO.ij. 98.9 97.8 97*1 99*2 94*8 93.2 Combustion 8i|.8 89. A 59-3 55*6 56.0 72.9 82.8 59*0 61.3 KOH Abs. 7)1.6 83.O 29.9 23.9 27.7 56.0 l64 32.3 44 *o Oxygen Addition 58.3 Combustion 72*6 63.I1 80.2 28.3 22.6 26.6 51.1 30.8 26.2 43.1 KOH Abs. 71.i|. 62.0 79.0 2 6 4 20.6 24.7 48.0 26.6 21.3 41.6 Combustion 71*0 61.0 78.7 2 6 4 20.4 24.0 43*4 23.6 20.0 41.5 KOH Abs. 70.7 60.7 78.6 2 6 4 20.1 24.3 1l2. 8 24.8 19.0 4 i *4 Combustion 60.4 42.2 244 18.5 KOH Abs. 60.2 k2.0 24.2 18. 3

I, 39 o TAEl E (Continued)

Run No, 20 21 22 25 24 25 26 27 _28___ 2g_

Sample Vol. 100.0 100.0 100% 0 100.0 100.2 100.0 100.0 100.0 100.0 KOH Alas. 98.6 97*0 90.° 75.2 1+6.7 100.0 85.7 •1*7 K^Hgl), At s. 97*5 86.5 33.0 97*0 79*8 t t l .2.1i Erg A d s . 90.5 m 79*6 8:4 32.2 75*6 59*0 I+8.0 32.6 CuO Comb. 40.5 3 9 -k 3 ? + 24.2 7.6 39*7 25.3 19.2 il+ .a Vol. after Rej, 23*7 z k .3 26.4- J.6 19.8 25*3 19.2 1I+.0 Oxygen Addition 9 7 4 98.9 100.ii 45*3 9.9*1 98.0 97.0 54*5 Combustion 514 5k.8 58.3 31.1 68.3 55*2 70.6 33.8 KOH Abs. 31*3 33-9 ■ 38.3 26.2 51.9 32.9 56.0 23.1 Combustion 30.3 32.6 38.0 42 .' 24.3 50.0 29.8 47.0f . n 0 19.2 KOH Abs. 28.9 31.8 36.il 23.I 48.4 27.2 45*9 1.6.8 Combustion 2910 31.8 37-5 22.8 Ii8.ii. 26.2 15.3 ^ • 8 KOH A b s . 29.0 31.6 - 22.5 48.2 25*5 L1.8 144 Combus tion 25.0 f 1 - 5 13.9 KOH A b s . 2ij..8 1+0.6 13*5

hi* TABLE IX

REACTOR EXIT FLOW RATES AND OVERALL CONVERSION OF ETHANE TO ETHYLENE AND ACETYLENE

INFUT PLOW Ratio Reactor 'Percent , ROT RATE Output Flows Temp. Conversion to NO. (cc/min) H2S c 2h 6 c 2h ^ c 2h 2 ? ® j % J k °F- c2h1j C2H2

1 200 kk 166 0 18 0 13 0.25 1030 0 10.2 2 200I4I 150 10 16 20 6 0.23 1235 5.7 9-1 3 200 "11- 8 ki *138 182 0.11 1810 L5 23.2 k 200 4 19 86 26 75 138 O.I9 1615 1l8.8 4 . 7 5 200 65 0 76 16 135 71 O.37 1565 li-3 -1 9.1 6 200 0 58 1.00 17,6 7? ■k8 1365 4 . 8 -32.9 7 200 4 5 4 0 8 0.02 1280 8.0 0 c h i 7-A. 200 132 13 1 66 8 0.03 1280 7.ii 0.6 8 200 61 137 20 16 20 7 O.35 1270 11.3 9.1 9 200 160 80 39 ko 7 0.91 1265 13.0 22.1 10 200 33 3k 98 12 57 ill O.I9 1595 59.8 6.8 11 200 51 26' 90 23 72 98 0.29 1605 51.0 13.0 12 200 23f 6 84 38 38 1.33 ■ 1350 33.7 ltf.6 200 2 96 78 1 18 76 0.01 1J485 4 . 2 0.6 4 200 11 105 1 52 0.06 life 32.3 0.6 15 200 26 101 h i 12 5k 39' 0.13 1H40 23.2 6.8 16 200 ko6 11 29 12 28 6 2.30 ll|i|0 16.i(. 6.8 200 115 23 21 22 0.63 1)435 29.0 13.0 " 16 200 1I4.8 87 21 20 1330 27.8 11.9 H 9 0.8k 200 0 26 200 19 3 ? 1?1 0.00 1690 lli.7 9.5 M 20 200 6 0 29 k i&9 206 0.03 1700 16.I4 2.3 21 200 12 0 29 6 ik7 202 0.06 1690 16.il 3.if ^ 22 200 k 3 3 30 16 13k 175 0.2k 1680 17.0 9.1. TABLE IX (Continued)

Input Plow Ratio Reactor Percent Run Rate Output Plows______S:C2H6 Temp, Conversion to Wo* (cc/min) HpS C?H| OpHp CH), H2 (■•Input) 92*1 °p. C pH|, CpHp

200 113 0 26 35 99 171 0.6k 1690 14.7 19*9 4 200 II4.8 0 78 . 79 162 0.8k 1705 7-4 1)4.2 25 200 288 0 k 74 % 134 1.65 1700 2.5 1+2.0 26 200 0 18 82 11 102 138 0.00 1560 46.5 '6.2 27 200 56 ■ 11 82 23 81 133 O.32 : 1+6.5 13.0 10k 28 200 0 78 31 80 120 O.59 1555 ljJi.2 17.6 29 155 171+ 6 36 33 1+0 67 1.27 1560 26.1). 21+.2 Percent; Conversion 10 20 50 30 0 0.2 ufrEhn ratio a r Sulfur-Ethane 80 F. 18X0° - iue 8 Figure Percent conversion of ethane to ethylene vs. vs. ethylene to ethane of conversion Percent hm Eg. et, ho tt Univ. State Ohio Deot., Engr. Chem. Chem. Eng. 950, July 12, 12, Mornlngstar July Eugene 950, Eng. Chem. • . S Figure sulfur-ethane ra tio at various temperatures. temperatures. various at tio ra sulfur-ethane 1952 . iVS o — — -- Figure 9*. Percent conversion of ethane to acetylene vs. sulfur-ethane ratio at various temperatures Chem, Engr. Dept. Ohio State Univ. Chem. .Eng. 950, July 12, 1952 Eugene Mornings tar

Sulfur-Ethane ratio Figure9 Percent Conversion 1200 Eugene Mornings tar Mornings Eugene Univ. State Ohio to ethane of conversion Percent July July Engr. Chem. Dept. Engr. Chem. Figure tyee s tmeaue ( temperature °F.) vs. ethylene at various sulfur-ethane ratios ratios sulfur-ethane various at 2 1 1300 0 1 , 195 . 95 ^ ° lLj.00 eprtr °F, Temperature Figure Figure 15017 10 nstrcr S:C2H6-0 1700 Percent Conversion 70 60 1200 Percent conversion of of conversion Percent Eugene Mornings Mornings tar Eugene Univ. State Ohio Chem. Eng. Eng. Chem. Figure Figure July July Dept. Engr. Chem. sulfur-ethane ratios. ratios. sulfur-ethane vs acetylene to ethane temperature at various various at temperature 2 1 1 1 1300 , 19 .

9 52

50

U0 1500 1U00 Temperature Temperature Figure Figure 11 f o . 1600 S:C2H6 j * o 1700 DEH3EPROGENATIOM OF ETKYUENE BY SULFUR: Because an appreciable yield of acetylene was obtained from, e thane (see Fig. 9 sond 1 1 ), it was de­ cided to study tbe reaction between ethylene and sulfur. Commercial compressed e thylene was used for these ex­ periments, Using tbe same techniques and apparatus described above. Tbe same input flow rate (200 cc./ min. ) was used for ethylene as for ethane as given in Table IX. The experiments were planned to cover the same ranges of temperature and sulfur concentration as used in the ethane study. The observed analytical data using input ethylene are summarised in Table X, and the calculated exit flows, conversion to acetylene, and the percent unreacted ethylene in the exit gas are presented in Table XI. The percent conversion to acetylene is also presented graphically in Figures 12 and 15« Figure 12 shows the effect of the sulfur-ethylene ratio on the conversion, with temperature as a parameter. The effect of temperature is shown in Figure 13, with the sulfur— ethylene ratio as a parameter. TABLE X

ANALYTICAL DATA - REACTION OP ETHYLENE MID SULFUR

Run No, Inlet 30 31 32 33 34 35 36 37 38

Sample Vol. 99.8 100.0 100.0 100.0 100.0 100.1 199*2 100.0 200.0 100.0 KOH Abs. 99.8 99.1| 67.0 37.2 9 9 4 60.0 N-7.0 100.0 83.I 73.2 KoHglk Abs.. 92.I1 984 58.5 22.8 99-0 53*8 26.6 93.0 59.3 63.3 BRp Abs. 7.6 8 3 4 4 . 6 2 0 4 69.I 384 22.2 18.5 26.6 4 4 CuO Comb. 7.0 504 23.7' 7.8 42.8 20.5 9*5 9*6 26.6 18.7 Vol. after Rej. 7*0 25*3 23*7 7*8 26.2 20*5 9*5 9*6 26*6 18.7 02 Addition 58.9 90.8 9 8 .0 3 0 4 96.2 90.I kl.l 60.8 IOI.I4. 9 9 .1 Combustion 43.6 6k. 2 60 .k 20.1 52.8 69.8 22.6 37.2 4*1 664 KOH Abs. 32.6 kb.k 1*1.6 15-7 30.6 58.5 15.1- 25.6 1±3.0 [j.5.2 Combustion 31.2 43*4 37*6 13*2 2 7 .1 81.0 10.2 2I4..3 35.8 37.8 KOH Abs. 29.8 J|l*9 34*9 11*8 24.2 46.9 7 4 22.4 31.6 33.7 O2 Addition ------2I-.3 - Combustion 29.2 , 41.6 33*7 11.0 22.0 42.3 16.0 22.4 27.8.32.3 KOH Absorption 2 9.0 kl.O 33.0 10.4 20*7 kO.l 13.6 21.6 25.9 31.1 Combustion - 32.8 10.2 20.1 37.6• 11.8 - 23.9 30.8 KOH Abs. - - 32.6 10.0 19.6 36.3 10.9 - 22.9 30.k Combustion - - 34.8 10.1 - 22.1 KOH Abs. - - - - - 34*2 9*8 ** 21.6 Combustion - - - - - 53.I4 KOH Abs. - - 32.9 Combustion - - 32.6 KOH - - 32.3

l.

3 TABLE H

REACTOR EXIT FLOW RATES, ETHYLENE INPUT GAS

Input C2H4 Ratio Reactor $ Unre-* Run Gas Flow Output Flows S:C2H* Temp, acted $ ** No, (cc/min,) M CaHfi. C3H4 C2H2 CHft J 2. (Input) ♦F. Ethylene Acetyl Cylinder Gas 200 0 10,6 170,0 14.8 4 .0 1*2 _ 100 8.0

30 200 2 11 37 3 93 99 0.01 1575 21.8 1 .6

31 200 101 4 30 26 61 76 0.59 1560 17.6 14.1

32 200 230 0 9 53 24 47 1.35 1560 5.3 28.7

33 200 1 10 71 1 80 62 0.01 1505 441.8 0.5

34 200 119 6 46 18 48 53 0,70 1495 27.0 9.8

35 200 283 0 8 38 24 18 1.66 1485 4.7 20.6 36 200 0 6 167 16 21 9 0,00 1330 98.2 8.7

37 200 64 19 93 23 26 14 0.38 1335 54.7 12.5 38 200 178 6 50 36 % 0 1,05 1325 29.4 19.5

* Based on input C2H4 * 170 cc/min. ■ -

\ r - \ ** Based on input C2H4 + O2H2 * 184,8 cc/min. f 3 Percent Conversion 10 20 50 0 0 Percent -conversion of ethylene ethylene of -conversion Percent uee Morningstar Eugene peratures. hm Eg. et, ho tt Univ. State Ohio Dept., Chem. Engr. Chem. Eng, tyee ai a vros tem­ various at ratio ethylene input to acetylene vs. sulfur- sulfur- vs. acetylene to input S:C0Hi Ratio 0.2 ^ 4 iue 12 Figure 950 Jl 1, 1952 12, July , 0.6 0.8 1.0 1.2 i.k 1560 1.6 1500 O" Percent Conversion 0 3 bo 70 60 20 50 10 0 Percent Conversion of ethylene ethylene of Conversion Percent iue 13. Figure Ohio State Univ. State Ohio hm Eng. Chem. Eugene hm Eg. Dept. Engr. Chem. ratios. July July tempera­ vs. acetylene to Input ture at various sulfur-ethylene sulfur-ethylene various at ture 1 , Mornings Mornings tar 1300 1952 950

lto r Tljtro eprtr °F. Temperature Figure Figure 13 1700 EFFECT OF SPACE VELOCITY ON REACTION OF ETHANE WITH SULFUR A brief study was made of the effect of space velo­ city on the reaction of sulfur and ethane. As indicated above, this reaction study was designed to obtain a de­ hydrogenation reaction which would not depend on careful space velocity control to assure a high yield of unsat- urates. For this reason the major portion of this study was conducted at an average contact time in the reactor of 7 seconds (see page M f )• This is a much longer con­ tact time than is reported for optimum yields in thermal processes, were subsequent carbonization will occur in the reactor unless the reaction mixture is rapidly re­ moved and quenched. In this space velocity study longer contact times were not investigated; the range from 7 seconds to about 1 second contact time was studied. High­ er flows (shorter contact times) were not readily obtain­ able in the apparatus used. The observed analytical data in this study are summarized in Table XII, and the calculated flow rates and conversions are presented in Table Xllland Figure 1I4.. All these da.ta were taken at reasonably cons tan t sulfur- ethane ratios and reactor temperatures, as shown in Table XIII. TABLE XII ANALYTICAL DATA - SPACE VELOCITY STUDY OF ETHANE - SULFUR REACTION

Run Nos* 40 41 Sample Vol. 150.0 150.0 150.0 KOH Abs. 65.4 67.4 85.9 KjjHgl* Abs. 61.£ 63.0 66.3 Br2 Abs* 40.4 34.9 34.6 CuO Comb* 19.0 19.4 25.4 Oz Addition 100.2 101.1 100,0 Combustion 69.2 74.6 55.6 KOH Abs. 52.3 56.8 23.6 02 Addition —— 46.8 Combustion 50.2 51.7 42.4 KOH Absorption 4B.1 49.5 37.0 Combustion 47.6 47.9 35.2 KOH Abs. 47.4 46.8 34.0 Combustion — 46.7 33.2 KOH Abs. — 46*2 32.8 Combustion — - 32.4 KOH Abs. —— 32.2 TABLE XIII

OUTPUT PLOWS AND CONVERSION OF ETHANE TO ETHYLENE MID ACETYLENE - SPACE VELOCITY STUDY .

Input Flow ______Plows______SiGoK, Reactor R i m _ _ N o ( . p . q / m i n HpS G2H6 G2H4 C2H2 H2 , (Input) Temp.

6 '20c .176 0 79 58 67 1|8 1.00 1565 39 l+oo 398 23 131 15.6 75 131 1*13 1600 1+0 700 643 70 288 252 83 161 1.05 1600 1+1 1150 972 252 480 294 96 138 0.96 1600

_ Percent Conversion to Total Output Plow Contact Time Run No. G2% g2h2 (cc/min., STP) (Sec.) 6 44*8 32.0 428 6.15 39 37.2 4 i .4 9o4 • 2.85 40 46-7 4o-9 i499 1.72 lil [17.4 29.0 2232 . 1.15 • \ C2H2

Figure li(.. Percent Conversion of ethane to ethylene and acetylene vs. Contact Time in reactor ( seconds). Temperature l600° F. S; C2H. ~ 1 Chem. Engr. Dent. Ohio State Univ., Chem Engr. 950 Eugene Momingstar

2 5 ” t 5 7 Contact Time in Reactor (seconds) Figure llj. J* 7 USE OF CATALYSTS R> R THE REACTION OF ETHANE AND SULFUR: Three solids were used as catalysts for the reaction of ethane and sulfur. These were chcs en from the many catalysts described in the literature. The catalysts used in this study were as follows: 1 . -ir" - 3/ 8" Impure Activated Alumina. This mater­ ial, sold by the Porocel Corporation is used as a de­ hydration agent and cracking catalyst. It Is activated alumina containing an appreciable quantity of iron oxide. A description of this catalyst Is given by LaLande et alh

LaLande, W. A., Jr., McCarter, W. S. W. and Sanborn, J.B.

Ind. Eng. Chem. 3 6 * 9 9 “109 (19^4-)

2 . Chromic oxide on activated alumina. This cata­ lyst was prepared using substan tially the method des­ cribed by Thacker,

Thacker, Carlisle M* (loc • cit)

In his example 2 , with the follow­ ing exceptions: (1 ) iq-Q mesh, rather than 8-liq mesh activated alumina was used* (2 ) preheating of the acti­ vated alumina was not carried outj and (3) reduction by means of methanol was not used. Instead, the actual -I— v ,'J hydrocarbon gas used for tills study was used as a re­ ducing agent for tlie chromium oxide, 3* Lump Pumice, This catalyst was indicated as a useful heat transfer mass for the reaction of carbon disulfide with saturated hydrocarbons to form olefins. Pumice was suggested by both Angus

Angus, Louis, loc. cit.

and Spindier.

SpindlBr, Henri, loc. cit*

Lump pumice was crushed to approximately l/I}. inch size and the fines rejected. The I/I4. inch particles were used as catalyst without any additives, In all the catalyst studies, the furnace was cooled, the reactor tube removed, and the catalyst added, to the reactor tube. In this case a straight reactor tube was used Instead of the reduced end tube used In previous studies. A 3/ 8" O.D. Porcelain thermocouple well was introduced Into the furnace from the exit end, extending the entire length of the furnace, and the catalyst was packed Into the annular space between the thermocouple well and the furnace tube. A long thermocouple was in­ stalled in this well in such, a manner that it ccmld be positioned at any point in the length of the reactor zone. This permitted a temperature traverse to be made of the reactor 2one. The exit line on the apparatus was rebuilt when the nevj furnace tube was installed. Tbe new exit line was made of glass, and was not heated or insulated. No sul­ fur condensation was observed in the exit line during any study, indicating the complete reaction of the sulfur during all the experiments made. The data observed in this series of experiments using catalysts are summarized in Tables XIV and XV. Table XIV presents the observed analytical data and Table XV, a summary of the calculated output flow rates and conversions. In order to permit an evaluation of the effectiveness of the various catalysts, the conver­ sion as a function of sulfur-ethane ratio is presented in Figures 15 and 16 for each of the three catalysts, and for the non-catalytic process. Figure 15 presents the data for ethylene and Figure 16 for acetylene. In each case the data at lL^O0^* £re shown • TABLE XIV ANALYTICAL DATA - CATALYST STODIES IN ETHANE - SU! RJR REACTION

Catalyst POROCEL PUMICE Run No • 42 1x5 44 45 46 1 T ~ "58 49 50 Sample Vol. 150.0 100.0 100.1 100.0 100.0 200.0 100.0 100.0 200.0 KOH Abs. 87.O 76.5 94.0 98.6 63.6 75.2 100.0 78.2 '81.L. K 2HglK Abs. 70,0 : 67.9 90.5 97*2 38.0 58.0 95.0 67.5 51.6 Br? Abs, 6l.l , 57.6 76.5 86.6 45*2 66.9 55.5 32.5 Cuu Comb, 20.0 21.2 27.5 39*7 1 : 5 19.0 32.7 21.6 13.0 Vol. after Rej. 20.0 21.2 27.5 • 21.0 24*4 19.0 20.2 21.6 13.0 O2 Addition 98.8 98.8 99-6 99.6 99*4 99-7 96.5 98.5 93.5 Combustion 75-k 60.4 47.2 65.4 57*4 65.5 5^.0 37.0 70.7 KOH Abs. 66.7 39 *5 19.0 43*8 29.8 45*2 29.6 31.2 37.5 0 2 Addition - ■■ - 55.4 W M Combus tion 6g.6 38.6 52.8 4l.6 25*3 52.0 27.5 26.0 30.2 KOH Abs. ■57 -U . 37-0 30.0 39*7 22.0 39.6 2k.6 22.2 k&'k Combus tion 53-6 37*o 49*7 39*5 21.2 39 . 3 2iU 20.6 ijli.l KOH Abs. 51.6 36.8 49*3 39*2 20.5 3 9 . 1 2 4 . 1 19.5 42.8 Combus tion 20,4 19.1 lj.2.6 KOH A b s . 46.0 20.2 18.8 42.3 Combustion 47.0 KOH Abs. • Combustion . a KOH Abs. 4 . 2 Combustion 44-9 KOH Abs. 44-9

F>a. i '**) TABLE XIV (C ontinued)

C a ta ly st PUMICE Chromium Oxide-Aluminum Oxide Run No. 51 52 53 ' ~5T" 55 56 57 58 Sample Vol. 100.0 100.0 150.0 100.0 200.0 150.0 100.1 150.0 KOH Abs. 98.6 81+.1 90.4 100.0 71.1 81.6 99.6 89.2 KgHglj, Abs. 9 6 4 74*7 8i *7 99*9 45.8 65.5 QQ77 •eL ? 76.8 BPp Abs. 76.6 58.8 67.5 97.0 43*4 53.0 90.0 61.2 CuO Comb, 51.0 22.9 25.6 60.5 27.5 37.2 42.7 34.8 Vol. after Rej. 51.0 22.9 25.6 21.4 27.3 22.2 22.4 23.8 Oxygen Addition 98.2 98.6 100.4 100.2 101.2 99.0 98.0 100.8 Combustion 45*8 59.6 56.8 r55.5 62.4 53*6 52.8 53.6 KOH Abs. 17.6 59.4 36.5 1 23.2 39*7 24.8 26.2 30.4 Oxygen A ddition 35* P Combustion 27.4 51.1 31.5 20.0 35.0 23.O 23.2 28.2 KOH Abs. 2^.5 26.2 28.0 17.0 32.0 20.7 20.6 26.4 Combus tion 22.6 25.4 27.4 16.3 31.6 20.5 20.2 26.0 KOH Abs. 22.1 24.6 27.O 15.8 31,2 20.2 19.6 25.6 Combustion - 2L<3 KOH Abs. 24.2 TABLE XV

REACTOR OUTLET FLOW RATES CATALYTIC STUDIES IN SULFUR-ETHANE SYSTEM

Input Gas ERatlo Run Flow Output Flows S:CaH6 Reactor No. Catalyst (cc/min.) M CgHfi CftH4 C2H2. CH* Ik- (Input) Temperature 42 Forocel 200 206 0 . , 29 55 59 135 1.17 1600 43 n 200 103 18 46 38 65 160 0,58 1600 44 11 200 25 20 55 13 84 195 0.14 1600 45 11 200 5 60 42 5 58 184 0.03 1450 46 11 200 151 50 38 23 35 102 0*86 1450 47 ti 200 304 9 31 42 35 64 1.72 1450 48 Pumice 200 0 57 92 ■17 42 113 0.00 1450 49 11 200 81 33 82 40 49 89 0.46 1450 50 ti 200 284 7 46 72 30 42 1.61 1455 51 11 200 5 11 74 8 101 173 0.03 1600 52 ti 200 64 9 64 38 84 145 0.36 1600 53 n 200 183 0 .44 24 75 128 1.04 160$ 54 C1O3 200 0 91 8 0 89 104 0.00 1450 55 11 200 334 15 6 66 39 42 1.89 1450 56 H 200 171 48 34 45 50 44 0.97 1450 57 It 200 2 48 31 1 99 163 0.01 1600 58 II 200 171 3 44 34 99 74 0.97 1600

' ■

i TABLE XV (Continued)

Bun, Percent Conversion to i 8®t_ C,Hfc (SB,

42 16*5 31.0 43 26*0 21.5 44 31.0 7.5 45 24.0 3.0 46 21.5 13.0 . 47 17.5 24.0 48 52.0 9.5 49 46.5 22.5 50 26.0 *a. 0 51 42.0 4.5 52 36.0 21.5 53 25.0 13.5 54 4.5 0,0 55 3.5 37.5 56 19.5 25.5 57 17.5 0.5 • 58 25.0 19.5

&

I Percent Conversion 0 -o- 0.2 : C S: 2 0.4 R 5 Ratio 0.6 0.8 Figure Figure 15 1.0 - at lij50°F. vs. sulfur-ethane sulfur-ethane vs. lij50°F. at - j Conversion of ethane to ethylene ethylene to ethane of Conversion j j j Eugene Mornings Mornings tar Eugene Univ. State Engr. Chem. Dept.,Ohio Engr. Chem. ra tio , using various cataly sts. sts. cataly various using , tio ra Figure Figure 15 1.2 . 950 Porocel , July 12, 12, July , "-s. Chromium 14 Pumice Oxide 1952 1.6

Percent Conversion 10

0 nv} hm Eg 95^ Eng. Chem. Univ.} ctln a l vs* * O l^ at at Acetylene Acetylene Chem. Engr. Dept. Ohio State State Ohio Dept. Engr. Chem. Conversion of Ethane to to Ethane of Conversion uy 2 15, uee Momingstar Eugene 1952, 12, July aiu ctl sts. cataly various sulfur-ethane ra tio , using using , tio ra sulfur-ethane iue l Figure Sulfur-ethane ratio Sulfur-ethane 6

o Catalyst No Figure Figure 16 Pumice Oxides" Chromium Porocel 0 . — J TEMPERA TORE GRADIEN T THROUGH REACTO'R; As indicated above in a discussion of the appara­ tus used in this study, the reactor section consisted of a procelain tube extending through, a Sentry high, tem­ perature combustion furnace with Globar heating elements. The length of tube within the heating section was lip inches. In addition, the intermediate line from the vaporizer to the reac.tipr furnace was electrically heated and insulated, and in early work the furnace exit line was also heated. It was recognized that a temperature gradient would exist in the reactor tube under these conditions. This gradien t was determined at lip50°F. and l600°P. during the^ catalyst studies. For this purpose, a closed end porcelain thermo­ couple well was installed, extending entirely through the reactor zone and protruding into the inlet line. A movable chromel-alumel thermocouple was installed in the well. An index was placed on the outside end of the thermocouple and a scale was positioned so that at zero reading the thermocouple was located at the exit end of the heating zone. Thus when the thermocouple was at the 7" position, it was in the center of the heated zone, and at the lip'1 , at the inlet end of the heated zone. Greater distances than ll+" represented extension into the inlet line. Gradients at the two temperatures are presented in Table XVI, and are shown graphically in Figure 1 7 * TABLE XVI TEMPERATURE GRADIM T ACROSS REACTOR Distance from Temperature Outside Tube Exit End______l 60QQF. ll+50°F.

0” b k o I+05 1” 715 670 2" 1070 1000 ?" 1370 1250 h" 156O 11+55 5" 1610 11+55 6" 1605 l k 6o 7 n 1590 if{l+5 0» 1570 1U30 9" 1560 1U15 10" 1530 1I+00 n " l45° 1335 12" 1285 1210 1070 1085 i h " 960 1115 15" 965 1300 16" 895 1260 17" 710 985 o Temperature 900 IOOO 150c* lUoc 150c 160c 170c lie 3.2C hm Eng. Chem. Univ. State Ohio Eugene Mornings Mornings tarEugene July July Dept. Engr. Chem. 17. Figure n fHae Zone Heated of end Inlet from Distance vs. Temperature Inside Reactor Reactor Inside Temperature 6 3 Distance (inches) (inches) Distance 2 1 Figure 17 Figure , 19

9 52

50

td Zone-r ated 1 , dieated ep. \ Temp . Temp . dated

\ DISCUSSION OF THE DATA: Because of the many aspects of this study, a point by point discussion of several items is presented. Ob­ viously some duplication may occur; however, this may be justified If greater clarity is obtained. 1 . Precision of observed data. Because of the relatively., slow method of analysis used, and the wide range of conditions to be investigated, duplicate gas analyses' were not made, except in cases where methods of analysis were being compared. In cases where duplicate analyses (or partial analyses) were per­ formed, satisfactory agreement was obtained. In addition, the smooth curves' generally obtained indi.cate that in­ herently the precision of the analysis was satisfactory, hence the method of analysis was deemed acceptable. Prom the method df calculation of flow rates based on a hydrogen balance, it Is obvious that an error in any part of the analysis will have its effect on the calculated results. In general, the variations thus introduced are not great, and would not In general account for the vari­ ations occasionally observed (see Figure 9 )* A more significant type of variation In analysis is demonstrated by the data presented previously.(page . ®her© the analysis of the reactor exit gas for hydrogen sulfide was presented as a function of time. It is ob­ served that a relatively wide variation in analysis oc­ curred in a short period of time. The exact reason for this variation is not known, but is believed that the degree of saturation of the -gas in the vaporizer may have been erratic, or that a variable quantity of entrained liquid sulfur carryover might have occurred. In any case, this variation would have the effect of producing varia­ ble calculated resul ts . In later experiments particu­ larly the method of sampling was modified in that samples were drawn slowly over a period of about two minutes. This had the effect of averaging out the composition, and led to more nearly reproducible results. 2 . Reactions In the Vaporizer. As indicated earlier, the reaction between sulfur liquid or vapor and the ethane stream passing through the vaporizer was Investigated briefly. The same method of analysis was used as had been developed for the furnace exit gas stream; this may not have been entirely justified, and it must be recognized that there was a possibility that this gas stream might have carried such compounds as me reap tans or sulfides, which were shown to be absent in the furnace exit gas. Since the i main emphasis was on the conversion to ethylene and acetylene, the analysis of this gas was not pursued further. It was observed that no appreciable dehydro­ genation occurred in the vaporizer, and that such reaction as did occur at the highest sulfur content (also 'highest vaporizer temperature) h d primarily to acetylene. There was also some carbonization in the vaporizer, as indicat­ ed by a definite discoloring of the sulfur used. It had been the original intent to determine the sulfur partial pressure in the vaporizer exit from the temperature of the sulfur in the vaporizer. This was found to be unreliable, the actual sulfur content de­ pending not only on the sulfur temperature as measured in the thermocouple well but also on the current through, the circuit on the top of the vaporizer. At low top temperatures refluxing occurred, and at high tempera­ tures In the top sulfur droplets appeared to vaporize completely In the top and lead to higher sulfur partial pressures than was calculated from partial pressure curves. For this reason all correlations of the sulfur- ethane ratio were based on the furnace exit gas analysis rather than on the vaporizer conditions. 5. Conversion to Ethylene, A number of very interesting observations were v made in this study Relative to the formation of ethylene. Perhaps the most unexpected phenomenon was the relative­ ly hi, yields of ethylene obtained by non-catalytic thermal cracking of the ethane to ethylene. From the information in the literature it would have appeared that at a reaction time of the order of 5-7 seconds, little or no ethylene would have been formed, whereas cracking yields approaching 50% were obtained (see Figure 1 0 ) in the absence of sulfur vapor. It is of course apparent from Figure 17 that although no quenching apparatus was used, the temperature drop in the gases leaving the reaction zone was very rapid and might be considered equivalent to a quench. Hence it may be that the apparatus and its use were not as far removed from commercial cracking practice- as originally considered-. It is of course apparent that contact time is not a unique factor in thermal cracking, and that time-temper- ature relations are important. Thus at a lower tempera­ ture and a higher contact time appreciable yields of ethylene might be obtained. At an input flow rate of 200 cc./min. ( STP) of ethane gas, the maximum conversion -n o i v ,J> to ethylene was obtained at l6lO°F. In the absence of sulfur vapor. At higher flows higher temperatures would probably be required, and the conversion would in gen­ eral be different. At all sulfur concentrations, the conversion of ethane to ethylene went through a maximum with tempera­ ture at aborit l600°P. It is interesting to note that a similar effect was noted in the equilibrium calculations on tiiis system (see Table II) except in the latter case a much wider temperature range was considered. The de­ crease at the higher temperatures was not due to a lack of reaction, but rather to subsequent dehydrogenation to acetylene and carbon. The effect of sulfur vapor on the conversion to ethylene was not great. In the lower temperature range, the conversion to ethylene increased slightly with an Increase in sulfur content (see Pig. 8, curve at 1270°P). This was also observed In the thermodynamic calculations, although In this temperature range the system was so far from equilibrium that the effect was almost certainly a kinetic effect rather than one of chemical equilibria. At higher temperatures, the effect was reversed, the conversion to ethylene decreasing with increasing sulfur content. This was also observed in the thermodynamic studies, the lower conversion being due to further dehydrogenation of the ethylene* In many experiments performed In this work, carbonization occurred to some extent, particularly in the high temperature and high sulfur concentration ranges. The limits of both variables were determined largely by the extent of this carbonization, 1750°F. being the practical upper temperature limit, and 1.ip the uppe r sulfur-ethane ratio limit used* Ip. Conversion to acetylene. ■In the case of conversion of ethane to ethylene, it was noted above that the conversion was sensitiveto temperature and. largely independent of the sulfur content' of the gas. Conversion to acetylene , on the other hand, appeared to depend largely on the sulfur-ethane ratio, and was much less temperature-dependent. Referring to Figure 9 > seer that the conversion to acetylene - Increased almost linearly with increasing sitlfur-ethane ratio In the range studied, and tha t all the data could be reasonably represented by a single line. This in Itself indicates the lack of an appreciable temperature effect; the -plot of _con vers ion vs. temperature in Fig. 11 demonstrates even more clearly this independence with temperature. The rmodynamic considerations indicated that the conversion to acetylene should increase both with in­ creasing sulfUr content and increasing temperature. The first effect was also noted experimentally. In the case of the thermodynamic calculations, an. assumption was made that no carbon ization occurred. Thus the lowest saturation state of the carbon recog­ nized. In the calculations was acetylene, whereas in practice further dehydrogenation to carbon also occurred. This would have the effect of lowering the conversion to acetylene at higher temperatures, thus nullifying the calculated conversion Increase. It is Interesting to notice that at lower temperatures and sulfur concentra­ tions, the conversion did increase somewhat with temper­ ature, but at high temperatures and sulfur concentrations where relatively more carbonization occurred, the con­ version to acetylene decreased with increasing tempera- ' I". ture • It should be pointed out that no emphasis has been given to the degree of attainment of equilibrium in this system;. In calculating the equilibrium conditions, It was found that literature values for the required thermo­ dynamic functions for the hydrocarbons varied widely • This appeared to be due largely to difficulty In measuring -!-• .J the heat capacity of the hydrocarbons involved in the temperature range used, because of the instability of all the hydrocarbons under these conditions. The thermodynamic equations used in this calcu­ lation were chosen from the standpoint of consistency and simplicity rather than accuracy, with the reserva­ tion that the calculated conversions might not be exact, but would serve to indicate the trends that might-be .v expected with changes in the major variables. For this reason the actual calculated conversions are not con­ sidered to be an accurate representation of the actual equilibrium conditions. It is also Important to point out again that carbonization and rupture of carbon-carbon bonds .vere both assumed not to occur in making the thermodynamic calculations, since either of these conditions would lead to the calculation of an equilibrium state with lit­ tle or no C£ hydrocarbons. In practice, conditions of operation were chosen to keep both these reactions to a minimum, with good agreement being observed between the experimental data and the trends predicted thermo­ dynamically, except for the effect of temperature on the conversion to .acetylene as noted above. 5 » Reaction of e thylene and siilfur* In the early stages of this study, ethylene was considered to be an intermediate in the formation of acetylene from ethane. For this reason the formation of acetylene from ethylene was studied, with the object that ethylene, might be formed by thermal cracking in one step, then optionally purified, and caused to react with sulfur in a second lower temperature step. On the basis of the curves in Figures 8-11 , it was speculated that such a process might give a higher yield of acetylene than obtained in a one-step process. A series of experiments was performed In which commercial compressed ethylene was reacted with sulfur vapor In the above reactor, except that a slightly lower temperature range was used. In this case it was again observed that the conversion to acetylene Increased linearly with increasing sulfur content, and with little temperature sensitivity. (See Figure 1 2 .) A somewhat different temperature effect was noted in this series than, in the ethane studies, as seen in Figure 13. At all sulfur-ethylene ratios the conversion went through s a minimum with temperature. Ihe reason for this Is not known• In the entire range of con ditions studied, the conversion to acetylene was less than that observed using an ethane input, and the percent ethylene in the exit gas was less Li.sing relatively pure ethylene as a feed gas, than using an ethane feed* ( cf. Figure 8 and Table XI) In the case of the ethylene input, relatively more carbcn ization occurred. This at least suggests that the sulfur present might tend to react in the ratio of two atoms of sulfur (one molecule of S2) to one molecule of hydrocarbon. Thus there might be a ten­ dency for the following reactions to occur: C2H 6 / s2 — C2H2 / 2 H2S (1) C2Hl / S2 — 2 C / 2 H2S (2 ) Ethylene was found to be formed at a rate substantially independent of the sulfur present, presumably by thermal non-catsilytIc cracking. C2H6 — 02% / H2 (3) The hydrogen evolved might then be subsequently reacted with the sulfur present to form hydrogen sulfide. If the ethylene thu.3 formed was substantially less than the equilibrium value, the reverse reaction (to form ethane) would be less significant^, and so the removal of hydrogen f 1 J "by reaction with, sulfur would not appreciably affec-t the - conversion to ethylene. It must he recognized that cracking type reactions, leading progressively to ethy­ lene, acetylene, and carbon and direct reaction of sul­ fur with the hydrocarbons as in Reactions (1) and (2 ) above occur simultaneously in the reactor, so that clear cut effects are not always apparent. A comprehensive study of the reaction mechanisms involved was beyond the scope of the present work. 6. Effect of Space Velocity. The conversion of ethane to ethylene and acety­ lene was studied over a five-fold range of space velo­ cities. As indicated In Figure lij., no really significant trends were observed. It was found that over the range studied, the conversion to ethylene went through a min­ imum and to acetylene, through a maximum in the range studied. The conditions used were chosen to give a high conversion to both ethylene and acetylene. This high conversion was observed over the entire range studied. 7. Effect of Catalysts. Catalysts were found to be of little value In promoting the reaction of ethane and sulfur. Two temper­ atures were used, li+^OCF. and l600°F. At the higher tem- f s .1 perature, no beneficial effect was noted with any cata­ lyst using any sulfur-ethane ratio. The results at ll4.50°F. were somewhat more promising. These results are presented graphically in Figures 15 and 16. It is ob­ served that the only catalyst that materially increased the conversion to both ethylene and acetylene was lump pumice. This was observed even in the case where no sulfur was used, the conversion of ethane to ethylene in the ab­ sence of sulfur Increased from about 28% for the empty tube to 52% for a tube packed with I/I4- inch pumice. This ef­ fect was lost at higher temperatures, however. Even at the lower temperature, porocel and chromium oxide on activated alumina1did not offer any advantage over the use of an empty tube reactor. Furthermore, except for the advantage of a lower reaction temperature, the use of pumice at lij.50oF. did not give an appreciably great­ er conversion than was obtained at l600°F. in the absence of catalyst. The lower temperature is a distinct advan­ tage and should* not be overlooked, however, since the process involved is endothermic and heat must be supplied to the reactor. 8. Temperature Gradient through Reactor. The temperature gradient through the reactor is presented here In order that the results of the invest!- fl

g&tion be interpreted in the light of the actual gas temperature throughout the reactor* Observed gradients are presented at two temperatures, llj.50° F and l600°F. (See Figure 1 7 )* 1h.e peak In the curves at indicated negative distances (outside the furnace) is attributed to non-uniform heating in the connecting line to the reactor. During the first four inches of travel the gas was rapidly heated to near* the furnace temperature. The next six inches might be regarded as the hot reactor f zone, in which the temperature rose farther, bxit only slowly. Beyond that the temperature dropped■! rapidly to the exit end. In. calculating contact time— throughout this • paper, the entire length of heated zone was used. It might be argued with considerable justification that only the six Inch length at the high temperature should have been used in this calculation, rather than the entire length. It might also be preferred in certain cases not to use the Indicated furnace temperature as shown (measured at a point just outside of the tube mid­ way along its length), but to use an average through the heated zone, either six inches of the entire lip Inches. It must be recognized, however, that these are conven­ tions arbitrarily chosen and that the actual effect of i. *

. j time (reactor length) and temperature on the overall reaction Is an integrated effect, in which neither time nor t emp e ra tur e , but rather the reaction occurring, is integrated. Since the effect of temperature on the differential rate of reactions occurring is very complex, only an empirical relation between time and temperature, and the extent of reaction, is generally used. Since the relation is emperical, any measure of the time and temperature may be used in the correlation, as long as the basis is stated*

In using such data to predict the reactions under a different set of conditions (in the case of a larger unit, for example) an uncertainty is necessarily intro­ duced. This would be inevitable in any case, however, wince the rate of heating to the reaction temperature would In general be different, hence the re actions oc­ curring during this heating up period would be different. Thus the conditions required to obtain a maximum conver­ sion In a different unit might be different from those presented here, and provision would need to be made In the design of such a unit to permit some variations in the operating conditions. For this reason, the use of the indicated temperature and space velocity values given in this paper is considered valid, although they are cer­ tainly not the only ones that could be selected. CONCLUSIONS: 1.^3 A brief but comprehensive study of the reactions occurring between ethane and sulfur was made. In this investigation a wide range of variables was studied, and the conversion of ethane to ethylene and acetylene was determined. The thermal non-catalytic cracking of ethane to ethylene was found to be less critical in nature th.an had been expected on the basis of Information In the literature Thus with a contact time of about six sec­ onds an d a temperature of l600°P., a conversion of 50% based on input ethane was calculated. This conversion was not materially changed by the addition of sulfur vapor to the reac'ting mixture, except that at high sulfur concentrations (S:G2H£>1) a decrease in conversion was obtained, presumably the result of further dehydrogena- tion to acetylene or carbon. The formation of acetylene from ethane was found to be a function of the concentration of the sulfur vapor In the gas, and to be largely independent of tem­ perature in the range studied.- It was also observed, that under similar conditions considerably less acetylene was formed in the reaction of sulfur and commercial ethylene. This strongly suggests that In the dehydrogenation of ethane "by sulfur, acetylene may he formed directly from the ethane rather than through ethylene as an in termed- late. This might occur, for example, through the forma-

j tion of a cyclic type intermediate comples of sulfur and ethane of the type |

H

0 H — S in which the sulfur might be oriented into the above configuration by hydrogen bonding with the available electron pairs on the sulfur. Subsequent rupture of this molecular complex would lead to the formation of two moles of hydrogen sulfide and one of acetylene. The formation of such a complex would suggest a first order reaction with respect to sulfur vapor concentration. This would lead to the observation noted, that the conversion to acetylene was a linear function of the sulfur vapor concentration. On the other hand, the conversion of ethane to ethylene appeared to be independent of the sulfur vapor content of the reacting mixture bver a fairly wide range of composition and temperature. The formation of ethylene appeared to he the result of thermal cracking of ethane to foim ethylene and hydrogen. A portion of the hydrogen thus formed subsequently reacted with the sulfur to form hydrogen sulfide; this was indicated by a decrease in the hydrogen flow in the reactor exit as the sulfur content of the gas Increased (See Table IX). It appears then that the reaction of ethane and sulfur to form ethylene C2H b / h S2 -3> C2H^ / H 2 S was really stepwise in the following sense; C2H 6 C-jBjj / H 2 h 2 / * s2 h 2s Hie first reaction appeared to be controlling in this case, so that the conversion to ethylene was relatively independent of the sulfur vapor present. In cases where a high sulfur vapor concentration was used ( S ;C2 H£,;> 1 ) , some decrease in the conversion to ethylene was noted. This was due to two factors; (1 ) At the higher concentrations of sulfur vapor, relatively more conversion to acetylene occurred, so that less ethane was available for cracking to ethylene. (2 ) Subsequent deh.yd.ro gen a ti on of ethylene occur­ red, leading primarily to carbon. This was noted par­ ticularly at the higher temperatures. The effect of space velocity was studied at the conditions giving maximum conversion of the ethane to ethylene and acetylene. A temperature of l600°P. was used to assure high ethylene conversion, and a sulfur- ethane ratio of approximately 1.0 to give a good conver­ sion to acetylene.. Over the five fold range studied no significant trends were observed with change in space velocity. 'The ethylene conversion wen t through a minimum and the acetylene through;a maximum. The total conversion to the two unsaturates weint, through a maximum at a contact time in the reactor of -two seconds. *\ V v V Catalysts wqfp found to be of little value in the v •> V\ dehydrogenation of ethane to either ethylene or acety-*- lene. The only condition\ where a catalyst had an ap­ preciable effect was in the case of lump pumice at lif.^O0 ^1. in which case higher conversions to both acetylene and ethylene occurred. The higher conversion m th pumice was no better than that obtained Without pumice at l600°P., so that the only advantage of the catalyst would be to permit use of a lower reactor temperature. This ought to be outweighed, however, "by the increased possibility of plugging the reactor bed with carbon. Neither poro- cel nor chromium oxide on activated alumina offered any advantages over the non-catalytic reaction. Ihe trends predicted by thermodynamic equilibrium calculations were in general borne out by experiment, except that thermodynamically it appeared that the con­ version to acetylene should increase with temperature, whereas actual observation indicated almost no tempera­ ture effect. This was attributed to the assumption in the equilibrium calculations that carbonization would not occur; some carbonisation did occur in practice. No attempt was made to calculate the degree of attainment of equilibrium beeause of the inherent inaccuracies in the thermodynamic data on hydrocarbons at the tempera­ tures used. -f AU TO BIO C-RAPHY J-

I, Ralph. Eugene Mornings tar, was born in Aitch, Pennsylvania, on April 20, 1922. I completed my secondary education at Huntingdon Higfr School, in Huntingdon, Pa. Immediately after high school I enrolled at Juniata Col­ lege from which I received the degree Bachelor of Science in 19i|-2 « I entered the Ohio State University in June, 19i|2 to do graduate study in the Department of Chemistry. While pursuing graduate study, I was also engaged on an industrially sponsored research project in the Research Foundation. 'This work aroused an Interest in Engineering training, resulting in my transfer to the Chemical Eng­ ineering Department, in which department I completed the requirements for the degree Doctor of Philosophy. . I was actively engaged on the project work in the Research Foundation during all my graduate s tudy.