Losses of Sulphur from the Sulphate Paper Pulp Process Citation Gunderson, Norman Robert (1932) Losses of Sulphur from the Sulphate Paper Pulp Process. Master's thesis, California Institute of Technology. doi:10.7907/pfqn-0q59. https://resolver.caltech.edu/CaltechTHESIS:09172024-215758337 Abstract No abstract. Item Type: Thesis (Master's thesis) Subject Keywords: (Chemical Engineering) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemical Engineering Thesis Availability: Public (worldwide access) Research Advisor(s): Lacey, William Noble Thesis Committee: Unknown, Unknown Defense Date: 1 January 1932 Record Number: CaltechTHESIS:09172024-215758337 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:09172024-215758337 DOI: 10.7907/pfqn-0q59 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16739 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 18 Sep 2024 18:03 Last Modified: 18 Sep 2024 18:20 Thesis Files PDF - Final Version See Usage Policy. 14MB Repository Staff Only: item control page

Loss es of Sulfur from the Sulfate Paper Pulp Process Thes i s by N. R. Gunderson In par tial fu lf illment of the requirements for the degree of Master of Sci ence CAL IFORN I A I NS TITUTE OF TECHN OLOGY Pasadena, California 1932 INTRODUCTION In the manufacture of paper pulp from wood by the kraft process, wood chips are heated or digested with a solution of Na 2 s and NaOH in order to extract the more reactive organic compounds from the less reactive, and therefore more durable and desirable, cellulose. The primary srurce of Na 2 s is Na 2 so 4 which is reduced to Na 2 s by Cat a high temperature, and the primary source of NaOH is Na 2 co 3 which is converted to NaOH by reacting with CaO and settling out the resultant caco3 . The inorganic constituents in the solution con- taining the material extracted from the wood are recovered by evaporating the water and burning the organic matter. The burning is done in a re ducing atmosphere so that carbon formed by the decomposition of the organic matter can reduce make-up Na 2 so 4 and other inorganic sulfur compounds to Na 2 s, and at a high enough temperature so that the Na2S and Na 2 co 3 formed, me lt and run out at the bottom of the furnace. Here they are dis- solved in water, Make-up Na 2 co3 is added, then the proper amount of Cao. The resultant Caco3 is settled out to g ive the solution desired. An attempt to remove sul- fur and sodium compounds from the flue gases is made by scrubbing them e ither with the alkaline liquor from the digestors or with water. 2 The object of this research is to investigate the relative losses of sulfur and sodium compounds from the system with the object of using a mixed double salt. Na 2 C'6 2Na 2 so4 and 2Na 2 co 3 • 3Na 2 so 4 • to replace losses instead of Na2CO3 and N82 so4 separately. This d ruble salt could be used at pres ent -irrwith Na 2 so 4 in making up losses but since the ratio of Na2so 4 to Na 2co3 used at present is about 10/1, only a small proportion of the double salt could be used. If losses of sulfur com- pounds from the system could be decreased a smaller ratio of Na 2 so 4 to Na 2 co could be used in making up 3 losses and a larger P!.Oportion of the double salt could be used. Consequently. the investigation has been one of the sulfur compounds found in the system and their reactions with regard to possible losses. 3 Previous Work The standard system of analysis used by the paper making industry has been developed by Otto Kress of the (1) Forest Products Laboratory at Madison, Wisconsin. · This System of analysis is used with solutions containing no organic matter and no polysulfide. The parts relating to sulfur compounds are given briefly as follows: To determine Se- so; - ½s, 2o; titrate a portion with r 2 in basic solution almost to the end point, acidify with acetic acid, and continue the titration to the starch end point. The reactions are S - ., 2I - 80 3 - I 2 - H20 II 2s 20 3 - I 2 so~ - ,2r- - 2H S40~ - 2I- To determine so; and s 2o3 precipitates= with a.mmoniacal ZnCl2, dilute to a given volume, allow the ZnS to settle, and take an alloquote part for titration. Neutralize to the methyl orange end point, titrate to the sta,rch end point with I 2 , and neutralize again to the methyl orange end point. _so 3 is distinguished from s 2o3 by the fact that it liberates acid when oxidized with r 2 • The reactions are: so--4 - 2I - 2H HSO3 - I2 - H20 . 2S 203 - I2 - - S405 - 2I - To determine S04. acidify a portion with HCl, boil to (1) Paper, 17, No. 24, 30 (1916) 4 expel H2 s and so 2 , and det ermine SO~ by precipitating and weighing as BaS0 4 in the usual way. s= + 2H + - H2S The reactions are: so-=3 + 2H + = S0 2 t Hi) 6H-+ + 2S - · + 4so = 3 S203 + 3H 0 2 3 6H + + 2s= + 'so3= - 3S + 3H 2 0 s2 o; s t - S0 3 G. C. McNaughton of the Forest Products Laboratory tested the following method of determining s =1n the black liquor produced by the action of the Na 2 s - NaOH solution on wood. Titrate the diluted liquor with ammoni- acal ZnC1 2 determining the end point by testing a portion of the solution for S- with ammoniacal NiS04 on a spot plate. Griffin( 2 ) recommends the following procedure for the det ermination of about 125 so 4 in black li quor. Dilute to , acidify, heat to boiling, precipitate the so4 as Baso 4 , filter, and ignite in a platinum crucible. Although a great deal of work has be en done on the effects of varirus liquor variables on the yields and properties of the pulp proo.uced, very little has been done on the products of reaction. The active constitu- ents are NaOH and Na 2 S; Na so 4 and Na 2 co 3 having little, 2 if any ef feet. According to Heuser< 3 ) only a comparatively small percentage of Na 2 s is used up in the pulping of beechwocxi~ - 2-------------------------- -·- < >Technical Me thods of Analysis - Second Ed ition ( 3 )Wochbl. Papie r-Fabr. 1913 p. 2 209 (from Sutermeister) 5 about 20% to 25%. Klasen and Segerf~ld ( 4) found that l of the original sulfide in the liquor 51.8% was combined with lignin, 15% expelled as volatile compounds, 16.4% remained as sulphide, and 16.4% was unaccounted for. 5 According to Falk( ) the following compounds are obtained by condensing the steam produced by releasing the pressure from the digestors at the end of a cook. Oily Portion Aqueous Portion Kgs./ton pulp Kgs/ton pulp Mercaptan .062 .06 Dimethyl sulphide .927 .17 Dimethyl disulphide .103 .05 Turpentine 8.49 .92 Distillation residues .721 Methyl alcohol 5.00 Klason( 6 ) has analyzed the flue gases from the smelting furnac.e and has found 13 mg. methyl mercaptan and 4 mg. hydrogen sulphide per cubic meter. The most reliable work on the reduction of Na 2so 4 7 to Na 2s has been done by Budnikov( ~ In his latest article on this subject he draws the following conclusions: (1) The production of Na2S by the reduction of Na 2s04 rests upon the following three principal reactions: (4) Papier Fabr., 1911 9, 1093-99 (from Sutermeister) (5) Papier Fabr., 1909 7, 469-72 ( from Sutermeister) (6) Papier Ztg., 1908 33, 3619 (from Sutermeister) (7) Zeit fur anorg und allgem Chemie, 170, 225-7 6 + 2C ~ ,. C _.,... 2CO Na 2 so 4 CO 2 Na 2 so 4 Na 2 s + 2C0 2 + 4CO -"' Na 2 s + 4C0 2 (2) The temperature of the beginning of reaction depends on the purity and density of the oarbon and upon its .mixture with sodium sulfate. The reduction occurs first between the solid phases. (3) The velocity of the reaction is great and con- tinues only for fractions of minutes. The lowest temperatures for the beginning of reaction found by him were soo 0 c., 75o 0 c, 880°c, and 750-8oo 0 c for sugar carbon. The following table represents the sort cf data obtained: % Yields We i g hts Na 2 so 4 C %Na 2 s %Na 2 so 3 %N a2 S04 .2 .06 80.0 .6 .2 95.0 ---- .6 .2 92. 2 1 10 .6 .12 61.5 1.15 --- 40 .6 .15 88.6 trace No. Temp. Min. Sec. 1 800 a) --- 2 850 4 3 380 4 4 1000 5 1000-1100 --trace trace 1.22 In this experiment the gases given off were not analyzed but gas analyses from other experiments show mostly co 2 . Air was displaced by nitrogen from the electric furnace used, the temperatures were measured by a platinum platinumrhodium thermocouple, and the Na 2 s was de termined by dissolving and precipitating as Ag 2 s with ammoniacal Agi.'t\J0 3 . Budinkov did not attempt to account for a 11 of the sulfur in the original Na 2 so 4 . However, the gases which he ob .. tained always gave a precipitate with AgN0 3 , which he ascribed to the presence of COS. The presence of a side reaction between sodium sulfate and sulfide was proven as follows: 13g of Na 2 s (prepared by calcining the hydrated salt in a current of hydrogen) mixed with .26g of Na 2 so4 was heated at 850° for 40 min. and .0519 g of so 2 absorbed in an iodine solution. This corresponds to 44% of the so 2 in the Na 2 so 4 . The side reactions suggested by Budnikov as the 011:es which reduce the yield of Na 2 s are: Na 2 so 4 + Na 2 s _... 4Na ,. 2S0 2+ S Na 2 S + H2 0 ~ 2NaOH + H2 S 2NaOH 4 CO 2 --,- Na 2co 3 + H2 0 Na 2 s ~ 2C0 2 __,_ Na 2 co 3 + COS The most complete published information on the operation of the recovery system as a whole is given by Wagner( 8 ) for the opera tion of the Wagner furnace, and inasmuch as this sys tern has be en the one studied a somewhat detailed description will be given. The l iquor is evaporat ed to about 70% solids in triple or quadriple effect evaporators and sprayed into the furnace at A. As it falls to the re ducin g zone, B, it is further evaporated (S)Ind. and Eng. Chem. 22 . 1930 122. 8 and the organic matter partly decomposed by the hot gases arising from the reducing zone. The bed of charred organic. matter, B, is about 5 ft. thick and the pool of molten Na 2 so 4 and Na 2co 3 ,C, is about 1 ft. thick. Air nozzle8 project into the reducing zone and the air injected by them burns excess carbon, generating the heat necessary for the endothermic reduction of Na 2S04 to Na2S. The inorganic salts run down into the pool ,C, and out through a spout into a tank c ontaining water. The gases in the furnace are burned by secondary air which is introduced into the furnace at points D. They pass through the waste heat boiler shovm, and into a scrubber. This scrubber consists of a cyclone separator into which is introduced an exceedingly :ffine spray of scrubber liquor liquor, the particles of spray being a pproximately the same size as the colloidal sized salt particles with ·w hich they unite. Part of the liquor is collected by the cyclone separator and the remaining part is recovered by passing the gases over baffles and through a layer of Hexahe l ix til e . Al l gases produced in the plant g containing odorous organic sulfur compounds are added to the furnace with the secondary air and burned. Scrubber liquor is circulated through the trough ,T, in the boiler in order to carry out any particles which settle into this part. The scrubber liquor is removen continuously from the scrubber system and added to the initial or "cooking" liquor either before or after it has been used to remove non-cellulose matter from the wood. The efficiency of the system in recovering inorganic salts from the black liquor injected into the furnace is primarily dependent upon the efficiency of the scrubber in removing these salts from the flue gases. The amount of salts carried over into the scrubber in the flue gases is a function of the type of spraying done; hence it is necessary to regulate the spray closely. Recently, it has been found that a Cottrell precipitator works well at moderate cost in removing solids from the flue gases. Also bag filters have been found to worK well. The temperatures given for the most efficient operation are as follows: Upper zone 1050 - 1150 C Lower zone 900 - 1050 C Emerging smelt 850 - 1000 C 10 The chemical reactions in the furnace are discussed by Wagner as follows; "The chemical reactions occurring in this process are similar to that in the smelter process, with this exception. In the rotary of the old process the low temperature and excess carbon cause sulfide and other sulfur compounds to change to carbonates and sulfur dioxide, the latter being carried out with vola tile combustibles, including marcaptans and sulfide gases. In our process the liquor is instantaneously passed through a high temperature oxidizing zone in an atmosphere strongly oxidizing , so that time and tempe r ature are not favorable to the formation of sulfur dioxide and sulfur trioxide. Except when solid wet liquor happens to reach the fuel bed, sulfur dioxide is not formed in appreciable amount. Carbon oxysulfide is given off under our conditions, due to excess carbon dispersed in the smelt." "These temperatures are well below t he boiling points of the soda salts and only sulfide and sulfate are readily volatile in air currents. It is not thought that the vapor pressures developed can lead to the dissociati6n of carbonates or volatilization of caustic. Sulfides will be converted to sulfates before passing out with the gases. Sulfates a re readily collec ted . 11 The efficiency of the Wagner furnace in recovering 11 inorganic salts from the liquor fed to it is not given for the "kraft" or sulphate process. However, the efficiency is indicated as about 98% for the "so da process" in which NaOH alone is used as the active extracting agent. Since almost the same type of furnace and scrubber are used in t he two processes t h e efficiencies may be assumed to be approximately equal. A brief study has also been made of ano t he r type of recovery system: ·rwhich is outlined as follo ws . The liquor obtained after the extraction of organic matter from wood is first evaporated in triple or quadruple effect evaporators and then further evap orated in "disk evaporators". These contain many circular metal disks supporte d on a common shaft whi ch is r ota ted slowly. The lowe r port i on of each disk dips into the liquor to be evaporated and t he UJlper por t ion is exposed to flue gases fro m the furnace. This evaporator also serves as a scrubber in r emovinc s ol i ds and gas es from the flue gases. After concentrat i on in the disk evaporators the liquor is dried in a ro tar y drier by passing it countercur r en t to t he flue gases is su i ng from the furnace. Final ly it i s a dde d to t he furn ace and burned in a reducing a t mos phere so t hat mak:e - up Na 2S04, added to the furnace, and other i norgan ic sulfur compounds are r educ ed t o Na2S and wi thdr avm from the bottom of the furnace in a molten comdition 12 togethe r with Na 2co 3 .These malten salts are dissolved directly i n water to produce a solution which becomes the "cooking liquor" after treatment with make-up Na 2co 3 and Cao. A certain amount of secondary air is admitted to the rotary drier through the opening through which is discharged the black ash or dried liquor. This serves as secondary air i n burning combustible gases from the furnace. This type of recovery system is not very efficient. Data obtained from a plant of the Crown Wi llamette Paper Co. located at Camas Wash. indicate that the liquor loses 100% of its sulfur content and 30% of its Na2 content during each cycle. 13 Experimental.! The problem resolved itself into finding where the large 100% loss occurred in the old type re covery system and where the small 2% loss occurred in the Wagner ,system. In attacking the problem it was thought to be desirable to obtain analyses of the liquors throughout the plan·ts in order to obtain a more comprehensive idea of the process as a whole and to locate the most probable sources of loss of sulfur compounds. The step involving the loss coul d then be studied in detail on a small scale. The study of a sample of dried liquor containing erganic matter or "black ash" showed t hat t he reacti ons of the various compounds in the system are comparatively simple. The S reacts wi th organic matter in the digestion process to form organic sul fur compounds, possibly mercaptans or sulfides . In the recovery furnace these organic sulfur compounds are given off as gases and more or less comple t ely burned to 80 2 . This 80 2 is more or less completely absorbed in the scrubber system. The so 3 and so 4 are more or less inert in the digestion process and are reduced to 8 in the recovery furn~ce. The s2o3 is probably produced by r eaction of Sand S03 in solution according to the equation, 6H - 28 - 4S03 3S203 - 14 3H o. The s o may or may not react with organic matter 2 2 in the digestion proces s , but a ny Na 2S203 fed to the 3 recovery furnace might decompose to sulfate and poysulf.i de ( 9 ~ a c cording to the equation, 4Na 2S203 3N§ SO 4 + Na2S5. 1'hus onl y half of the sul fur in t he 2 s2o would be available for direct conversion to s; 3 the remainin~ portion formi ng pol ysul i'ide sul fur which would possibly react with organic matter, be converted to so 2 , absorb ed, and cpnverted to s:- I"t .was thought that polysulfide might act as a source of COS in the .. . .• Wagner furnace by r eacting with CO in ·. the recluc.ing zone. Thomaa anQRule(l O) found that H2 reacts rapidly with polysulfides at about 500 C with the formation of H2s. Hence it was thought desirable to know the amount of s 203 in the liquor as it entered the fu:i:nac.e. I 3 Also, in order to trace the reactions of s2o in the system , it was desirable t. o have a method f o:c determining polysulfides. Soon after the work was begun it was found that the existing systems of analysis did not give as accu~ate results as were required. Most of the work done so far and here reported has been in develonirnr metbod,s (9) Ephraim (10) Thomas and Rule J.c.s. 111 , 1063-85 15 of analysis for the solids and liqu ids found in t he s ystem. The mo st i mportant sources of error found in t he me t hods of ana l ysis given by Otto Kress of the Forest Products Lab oratory are as follows: (1) In the oxidation of sulfur compounds with I2 in basic solution par t of the S2 03 is oxidized to S04, thus requi ring more I 2 than for ox id at ion to S406~ (2) It is extremely difficult to neutralize the solution containing S03 and S203 to . the methyl orange end point due to the brown coloring matter in the solution. (3) Due to the oxidation of so 3 by 02 dissolved in the water an error is made in the determination of S03. The experimental data upon which these assertions are based is as follows : (1) To a 250 cc beaker w~s added 10cc of standard .1 N Na2S203 solution, an ap proximate volurne of .l N NaOH and enough water to make up to 100cc. This solution was titrated with an excess of standard I2 solution, acidified with HCl, and back titrated to the starch end point with standard Na2S2 03 solution. vol. lN ccI2 NaOH .oo .oo 9.87 9.86 13.13 5 5 13.07 10 16.19 16.52 10 ccNH40H 47.30 5 equivalent ccI2 theoretical fo difference from theoretical .oo .oo .12 .12 .07 .06 9.87 9.86 13.01 12.95 16.12 16.46 9.86 9.86 9.86 9.86 9.86 9.86 .oo .oo .56 46.75 ccNa2s203 31.9 31.3 63.5 66.9 374. 16 When greater concentrations of NH40H or NaOH are used the %error increases even further. Addition of BaCl2 solu- tion to these solutions after titration, produced a precipitate of BaS04 in the ones in which the I2 titration had been made in basic solution, but not in the others. The amounts of the precipitates were roughly proportional to the %s difference given in the table. If all the S203- were oxidized to S04= ace at' ding to the equation, S203 .+ 4I2 + 10 OH=2S04+, 8I - rf; 5H20, 8 times as much I2 would be I::. consumed as in oxidizing the same amount of s203 to S405 according to the eq_ua tion, S203 -t- I :L ½ S405= ·+ -.'- 2r: Hence the possibilities for error are large in the basic titration. Enough C.P. Na2S03 was dissolved in freshly boiled distilled water to produce an approximately .o5Tu1 solution. Ten cc. of this solution was diluted to about 200cc, methyl orange added and the solution titrated to the methyl orange end point with HCl. Starch was added and the solution titratedtlxo the starch end point with standard I2 solution. After decolorizing with Na2S203 solution the solution was again titrated to the methyl orange end point with standard NaOH solution. The S04 in the solution was determined by precipi ts.ting and weighing as BaS04. The quantity of S03 taken for titration was small in order minimumize possible errors due to the loss of S02 from the acid solution. conditions. This procedure was repeated using different Trial 1 was done as indicated above. In trial 1'7 2, 10cc of 1M NaHC03 was added to the solution with the Na2S03 before titration in order to aid the escape of any S02. In trial 3 the acid produced by the reaction was not measured; the Na2S03 solution was added to 200cc of solution containing equal concentrations of NH4Ac and HAc, andthe solution titrated with I2. In trial 4 the Na2S03 was added to a slight excess of I2 solution in 200cc of water and the solution back titrated to the starch end point with standard Na2S203 solution. In trial 5 the total sulfur content of the 10cc sample of Na2S03 solution was determined by oxidizing with bromine water. Trial ccI2 ccNaOH Equiv. NaOH Equiv. I2 wt BaS04 1 2 3 7.75 7.30 9.50 9.55 11.80 11.12 1.500 1.505 .1229 .1228 ..1230 .1233 .1239 4 5 The results given in this table indicate that in all the trials all of the original S03 was converted to so 4 , and that in the cases in which the acid produced by the reaction wasmeasured it corresponded to the reaction HS03 and I2 and H20 S04 and 2I and 2H. The experiment was repeated exactly except that a more concentrated Na2S03 solution was used. The results are given below. Trial ccI2 ccNaOH 1 2 3 4 28.84 27.57 30.53 30.56 43.90 41.85 5 Equiv. NaOH Equiv. I2 1.501 1.49'7 wt BaS04 .387'7 .3846 .3900 .3898 .3906 18 The results given in this table are much the same as those given in the previous table. In order to study the amount of acid produced by oxidation of so; by I2 in comparison to the amount of I2 used, another experiment was made. performed as follows: Trials 1 and 2 were A 10cc portion of Na 2 so3 solution was added to 200cc of water and neutralized to the methyl orange end point with standard HC1• The solution was ti- trated to the starch end point with standard I2 solution and to the methyl orange end point with standard Na0H solution. Trials 3 and 4 were performed as follows: The Na2S03 solution was added to a solution containing almost enough I2 for oxidation, titratedto the starch end point and then to the methyl orange end point with standard Na0H. Ration, 1, is the ration of equivalents of Na0H used in titrating the solution to the methyl orange end point after titrating with I2 to the equivalents of I2 added. Ration, 2, is the ratio of the total number of equivalents of Na0H added to the solution to the total number of equivalents of I2 adde d . The following table gives the results: Trial cciz ccHCl ccNa0H Ratio 1 Ratio 2 1 2 24.70 24.83 12.20 37.60 27.27 27.30 .983 .984 3 4 37.70 27.64 27.68 1.502 12.13 1.499 1.001 1.000 These results show that the amount of acid liberated by oxidation of HS03 by I2 in acid solution, corres ponclng to ratio 1, is given by the equation, HS03 ,4- _, I2 '·+"'- i-I20 = S04 '+ d 2I - s.-¼,-,-1 3H+. - It also shows that the amount of acid produced in the oxidation of S03 by I2, corres ponding to 19 • ~:i. 2, is given by the equation, S03 ,+ ,. I 2 -H ·r:l H20 = S04 r+ -. 2I- r+ 2H~. The deviation of ratio 2 from the theoretical value 1, may be explained by loss of S02• In order to determine the effect of dissolved oxygen . upon the amount of I2 used in the reaction under various conditions, the previous experiment was repeated in part, using freshly boiled distille d water to dilute the solutions. In trials 1 and 2 the 10cc sample of Na2S03 in oxygen free water was diluted to 200cc with oxygen free water, the solution titrated to the methyl orange end point with standard HCl, and then to the starch end point with standard I2• In trials 3 and 4 the Na2S03 solution was added directly to 200cc of oxygen free water containing almost the proper amount of I2 for oxidation, and the solution titrated to the starch end point. Conditions in the previous experiment were duplicated as closely as possib le except that freshly boiled distilled water was used to dilute the solutions. The results are given in the following table. Trial ccI2 1 29.33 29.37 29.85 29.83 2 3 4 These results show that the exclusion of part of the dissolved oxygen from the solutions tends to reduce the differences in the amount of I2 used in the oxidation of Na2S03 under various conditions. The effect of the complete exclusion of oxygen was not determined due to the difficulty of removing it from the I2 solution. 20 From these experiments it was concluded that the oxidation of S03 by I2 in slightly acid solution is quantitative and is given b y the equation, HSOi :-+ ':.. I2 + .:r" H20 -= S04- ,+ _, _ 2I r + 3H.1. The variations in t he amounts of I2 used in the oxidation of S03 under different conditions cannot be explained by assuming the loss of S03 as S02 or by assuming incomplete oxidation of the S03. The explanation sug gested is the oxidation of part of the so 3 by 02 dissolved in the water according to the reaction, S03 °·+ . ½02 =-S04-==-. In view of the fact that the oxidation of S03in basic solution or of S02 in acid solution by 02 dissolved in the solution is only about 4% per hour this explanation does not seem probable. However, it is the only one whic h explains the facts found. Vol hard tll.) s ta tes t:iat the anomalous re sults are due to the reduction of part of the sulfurous acid to sulfur by hydriodic acid according to the equation, HS03 '+-·. 3H + +,. 4I = 2I2 :-+ :· _ 2H20 '+ •. S. The I2 thus produced oxidizes rema ini ng HS03 to S04 according to the equation, HS03 '+ • I2 r+ ,_ H20 ~-S04""' ' +: 2I - , -{:,:: 3H .i • The resultant of these t wo equations may be written as 3HS03 = 2S04 _4- ,_, S :- + ~ H20 r- + . H"'· If tbis equation were correct the ratio equiv. Na OH equiv. I2 , would not equal 1.5 as it actually does. Also, although a precipitate of sulfur always forms during the titration of S03 with I 2 it is so extremely minute that it could not account for the large variations in the amount of r 2 used. (ll) Ann. Chem. - . , Pharm,. , 242 94 21 Raschig believes that the error which results when iodine is added to sulfurous acid solution is due to loss of so 2 • <12 ) Z. Angew Chem. 1904 580 22 SYSTEM OF ANALYSIS for solutions containing No Organic Matter. Due to the fact that sulfur compounds interfere with each other in anal y sis, and to the fact that the solutions contain brown coloring matter which makes the use of indicators difficult it is desirable to separate as many of the constituents as possible from the solutions as solids and determine the solids. Determination of s= s= is se parated from the solution as ZnS and titrated with standard I2 in sli ghtly acid solution to the starch end point. The reactions are: ZnS t 2H+ = :,.. ·H ' -\- zn-t + H2S, :. 2H. + 2I -t S, and ZnS -+ I2 = Zn-\- 2I --t S. The ZnS preci pitate is tr ansferred to a beaker, standard I2 solution is added, and a s mall amount of HAc is added to hasten the solution of ZnS. After an excess of I2 solution has been added, a small amount of HCl is poured in to complete the solution of ZnS. The excess standard Na2S20 3 solution. r 2 is back-titrated with The colloidal precipitate of sulfur produced sometimes interferes with the starc h and point, producing a pink colo r , and causing further deepening of the starch I2 color upon standing . This effect may be removed by adding a few drops excess of standard Na2S203 solution in the back - titration , filte r ing , adding starch, and titrating to the end point with standard I2 solution. Also, the Z"nS seems to be very slo w in precipi- tating if a large excess of ammoniacal Z:nCl2 is present. Th is effect ma y be removed by using diluted ammoniacal ZnCl2• 23 Procedure: To a 400cc beaker add 5cc of lM ammoniacal ZnCl2 and enough water to make up to about 150cc. Add the sample to .be analyzed for ·:s; and swirl the solution so as to obtain proper mixing. Allow the ZnS to settle while preparing the filter. Place a filter paper in a medium sized Hirsh funnel and after wetting it pour on a suspension of Supercel. -1:- This forms a coating which prevents ZnS from coming in contact with the filter paper and thus makes it unnecessary to add the filter paper to the solution titrated with I2. Decant most of the solution containing the ZnS through the filter, mix some Supercel with the remaining solution containing the ZnS and filter. Wash the beaker and precipitate thoroughly with water. Wash the precipitate back into the beaker with water, add about 5cc of standard .lN I2 solution, 5cc of HAc, and then I2 solution until it ceases to disappear. Add 2cc of HCl, swirl the solution for a few seconds, and back-titrate with standard .lN Na2S2O3 solution until a few drops of it are present in excess. Filter, add starch, and titrate to the starch end point with standard I2 solution. If an accuracy not greater than 5% is desired the filtering may be dispensed with. Experimental results: A solution approximately .lMin K2S and .2N in KOH was made by dissolving H2S in KOH solution. 10cc portions of this solution were analyzed for s= by the methods given above. In trials 1 and 2 no other reagents were present but in trials 3 and 4 10cc of approximately .lM Na2SO4, Na2SO3, Na2S2O3, Na2CO3, and NaCl were mixed -1:- A filter aid produced by Johns-Manville Company 24 with the 10cc portion of K2S solution before determing ::-: s ~ in it. Also the ::s =-in the K2S solution was determined by precipitating with ammoniacal A~No 3 , filtering off the Ag2S, dissolving it in cone. HN03, and determining the Ag+ dissolved by titrating with N}¼CNS. All the precipitates were made as nearly as possible at the s·ame time in order to reduce errors due to the oxidation of ~s: The results are given in the following tables: Trial 1 2 3 4 ccNa2S203 equiv. ccI2 ccI2 solution solution solution equiv. I2 .43 .45 .38 .003643 .003648 .003645 .003646 35.46 35.52 35.43 35.64 .58 35.03 35.07 35.05 35.06 I2 = .1040 N Na2S203 = .1025N Trial ccAgN03 solution ccNB-4CNS solution Equiv cc Nli4CNS Equiv. Nfl4CNS 1 .29 .09 36.02 35.94 35.73 35.75 .003649 .003650 2 NH4CNS = .1021 AgN03 • .1025 Determination of S2= s 2: is precipitated as ZnS2 with ammoniac al Zn012. In the treatment with acid and I2 it undergoes the following reactions: S + 2H+ -+ ZnS2 + zH+ = 2I-, and ZnS2 + I2 T-\- Zn - ~ H2S + S, H2S + I2 = Zn-H- -\- 2 S + 2I - . Hence , polysuth de may be calculated from the equivalents of I2 used in oxidizing the precipitate to the starch end point and from the total sulfur in the ZnS precipitate. The equiva- 25 lents of I2 gives s= s= +-1/2s=2 . s 2 = andthe total sulfur gives The total S in the precipitate is made by oxidation with hot acid bromine water and determination of the S04 produced by precipitating and weighing as BaS04. Procedure: Obtain the ZnS precipitate as indicated in the previous procedure, transfer precipi ta.te and filter paper back into the beaker in which the precipitation was made, add about 100cc of hot bromine water, then about 5cc of HGl, and boil until all the Br 2 is expelled. and determine given for Filter S04 in the filtrate according to the method =so[. Experimental results: The amount of s2= found by the above method was compared with the amount found by acidifying a solution containing s2= and determining the H2S + s. S liberated according to the reaction, s=+2H""" -=- A solution containing S2= ands= was prepared as follows. A . lN KOH solution was sa.tura ted with H2S, and enough sulfur added to convert about half of the s= to s2=• The solution was boiled until most of the sulfur dissolved, cooled and filtered. Enough KOH was added to produce a solution about .2N in KOH. Four 10cc portions were taken for precipitation with ammoniacal ZnGl2. Two of these precipitates were titrated with I2, and the total sulfur in the other two determined by oxidizing with Br 2 . At the same time 2 10cc portions were acidified with HGl, the solutions allowed to stand for several minutes. The sulfur precipitates were filtered off and oxidized with Na202. The So4= produced was determined by precipitating and weighing as BaS04. The following table gives the revults: 26 .57 .63 20.78 20.86 equiv. ccI2 equiv. I2 20.21 20.23 .002103 .002105 wt. Baso4 .4172 .4176 Average m s=2 a .000315 For the direct determination of s2• wt Baso 4 M s2= • 0'730 .0735 .00313 .00315 Determination of : so3 + : S04~ Due to the fact that :-: so3 is oxidized to ,:3 04 in solution and during analysis and to the fact that it is difficult to separate '"S03 from the solutions as S02 without producing troublesome sulfur precipitates , ~so 3 is oxidized to :~ so4 with r 2 and determined with " S04': This is also more desirable in consideration of the fact that Na2S04 and Na2S03 undergo sintllar reactions upon reduction; hence their sum is more significant than either one alone. The sa is separated from the solution by pre- cipitating with amrnoniacal ZnCl2 and filtering. The :-so3- and ;:;3203- in this solution are oxidized by an acid I2 solu tion and the excess of I2 back-titrated to the starch end . point with standard Na 2 S203 solution. The ratio of acid to oxidizing agent in the acid I2 solution should be slightly greater than the ratio of base to reducing solution in the solution being analyzed. Upon mixing these 27 two solutions in ·any proportions I 2 is always in acid solution, being used up by the reducing solution before the solution becomes basic. The reducing solution is al- ways basic until immediately before oxidation with I2, and portions of the solution which have reacted are nearly neutral so that they act merely as diluents for the remaining portion. In this way no So3• is lost due to the acidification of a solution containing much So3=, yet all of the S203= is oxidized to S405= because the oxidation with I2 is carried out in acid solution. The base in the solution may be estinIBted from the amount of NH40H added as ammoniacal ZnGl2 and from the base in the sample taken; the reducing agent must be estimated by trial. After the titration Si02 is removed by heating and filtering, and so 4 = is determined by precipitating and weighing as BaS04. This method of removing Si02 is allowable because the amount present is extremely small. In the precipitation of BaS04, the BaG1 2 is added slowly thus allowing the crystals to grow slowly and to a larger size. This method is more rapid ~in the production of an easily filterable precipitate than t h e addition of a large excess of BaCl2 all at once because in this case the colloidal sized crystals produced must dissolve in a solution in which they are but slightly soluble in order to recrystalize to larger ones. Addition of Ba012 at a rate slower than l/2cc per minute of 25% solution is of no advantage because the crystals settle to the bottom and thus do not become larger. Also, in the 28 ignition of this crystalline precipitate it was found that subsequent treatment with concentrated H2S04 does not increase the weight appreciably. Of the ten precipitates tried the gain in weight was not over .lmg, and this could be fully accounted for by the conversion of chlorides or oxides in the filter paper ash to sulfates. As in other cases colloidal sized particles have markedly different properties than larger sized particles. Procedure: Add the sample to 5cc of l M ammoniacal ZnCl2 in 150cc of water and obtain the approximate volume of HCl required to neutralize to the methyl orange end point. Repeat the precipitation of ZnS, and filter. fer the filterate to a 600cc beaker. Trans- To a 250cc beaker containing about 2o% in excess of the volume of HCl required add some standard I2 solution. Pour t his solution into the one containing So3= and S203• and titrate to the starch end point with standard Na2S203 or standard I2 solution. Repeat until the ratio of acid to oxidizing agent is sli ghtly greater than the ratio of base to reducing agent, and until an e xcess of I2 solution is used. Record the equivalent numtier of cc. if I2 solution used and use this in calculating s03= after s 2 03= has been determined. After the solution ha s been titrated, neutralize to the methyl orange end point, add 2cc of H~ \ , and heat. Filter out Si02, heat to boiling, and add 25% BaCl2 solution at the rate of 1/2 cc per minute for five minutes, then at the rate of 3cc per minute for fifteen minutes. Stirring action i s secured by allowing the solution to boil gently. 29 Filter when convenient using a small Hirsh funnel and Whatman No. 4 filter paper coated with a thin layer of paper pulp. Ignite in a weighed crucible, cool, and weigh. Subtract .7mg. from the difference in weights, due to the ash in the fi]Br paper. Experimental: The ratio of Na2S20 3 to I2 obtained by titrating according to the above method was compared to the ratio obtained by titrating in neutral solution. In trials 1 and 2 the titrations were made in neutral solution. In trials 3 and 3 the Na 2S203 was added to 25cc of NH40H in 50 cc of H20• To this was added a slight excess of I2 in 30cc of HCl and 50cc of H20• The resultant solution was titrated to the starch end point. Trial ccNa 2 s 2 o3 ccI2 Ratio 1 32.07 29.72 30.21 30.75 32.23 29.88 30.38 30.91 1.005 1.005 1.006 1.005 2 3 4 The volume of standard r 2 solution used in titrating a sample of Na2S0 3 by the above method was compared with the volume used in titrating in neutral solution. Enough C.P. _Na2S03 was dissolved in freshly boiled distilled water to produce an approximately .05:M solution. In trials 1 and 2 the Na2S03 was added to an excess of standard I2 in neutral solution and the solution back-titrated with standar•d Na2S203 solution. In trials 3 and 4 the Na2so 3 was added to 25cc NH40H in 100c H20. To this was added a slight ex- cess of standard I2 in 30cc HCl and 50cc H2 o. The result- ing solution was titrated to the starch end point with 30 Trial ccI2 ccNa 2s 2 03 equiv. cc I2 1 27.35 27.-52 27.43 27.60 .42 .61 .54 .73 26.93 26.91 26.89 26.87 2 3 4 The decreasing amount of I2 used was probably due to the oxidation of So3= by 02• S20 3 ~ is deterrr~ned by precipitating as P..g2S203 in nitric acid solution after the s= has been removed with ammoniacal Zn(N03)2, and allowing the precipitate to decompose according to the equ a tion, Ag2S203 -+ H20 = R2s0 4 . Ag2S + The nitric acid catalyzes the decomposition and causes it to be complete within one minute. Enough NH40H is then added to dissolve other silver salts which may have precipitB.ted, and the Ag2S is filtered off. The Ag2S is determined by oxidizing in concentrated HN03 and titrating with standard NH4CNS. A large concentration of s03= in the solution containing the S20 3 • sometimes introduces an error due to !l'ecipi tation of Ag2S03 and its inclusion by the Ag2s 2 o3 precipitate. However, this error may be removed by precipitating under conditions which produce no precipitate of Ag2S03, that is in strongly acid solution. Decomposition of S20 3 = by the acid is prevented by adding acid and A,gN0 3 together. In- clusion of AgCl by the Ag2S203 precipitate introduces no error because the AgCl is not dissolved upon treating with Procedure: Add the sample to 5cc lM ammonia.cal Zn(N03)2 in 150cc H20, and filter. Transfer the filtrate 31 to a beaker. Acidify with HN03 while stirring and add quickJ;sr 3cc l M AgN03 and 15cc HN0 3 in 100cc H20. stand for at least one minute. Allow to Add an excess of NH40H and filter out the Ag2S, using a small Hirsh funnel. Wash the beaker and preci pitate thoroughly with water and transfer the precipitate bac·k into the beci.ker. Add 5cc of concentra- ted HN03, cover the beaker with a watch glass, and boil until most of the nitrous oxide fumes have been expelled. This fills the beaker with HN03 vapor which rapidly attacks any particles of Ag2S on the siqes. As soon as the solution of the Ag 2 s is complete, add 100cc H2 o, cool, and ti tr ate the Ag+ with standard NH 4 CNS solution using ferric alum as an indicator. Expe rimental: The S203= found by the above method was compared with that found by oxidizing to so 4 = with Br2 water and determining So4•. In the determination of S20 3 = by the above method the S203 was mixed with about 70cc of .llVJ Na2S03 25cc of NH40H and a few dro ps of HCl before the determination. The results are given in the following table: Na2S203: AgN03 = NH4CNS = .10301✓1 .1025N .1021N ccNa2s 2 o3 acNR4 CNS ccAgN03 equiv. ccNH4 CNS equiv. NH4CNS equiv. Na2S203 10 10 10 10 20.52 20.66 20.43 20.57 .32 .45 .23 .38 20.20 20.21 20.20 20.19 1.001 1.002 1.001 1.000 32 Determination of Na2 Na 2 is determined by converting the inorganic ss.lts to NaCl, NaBr, and Na2S04, and determining each. The in- organic sulfur compounds are oxidized to S04= by acid Br2 water and the remaining inorganic salts are converted to bromides by acidifying with HBr and evaporating to dryness. The salts are dissolved and the solution diluted to a definite volume. A portion is taken for the determination of 01- and Br- by titrating with standard AgN03 solution until an excess is present, filtering, and back titrating with NH4CNS using ferric alum as an indicator. Also, a portion is taken for the determination of so4=. Bromides are used instead of chlorides because they are more easily determined with AgN03. As AgN03 is added with sirring to a hot solution containing Br-, coa gulation of thepr>ecipitate takes place very rapidly as soon as a few drops of AgN03 solution are present in excess. The solution may be filtered immediately without danger of losing AgN 03, and back-titrated with standard NH4CNS solution. Procedure: Add the sample to an excess of Br2 water, acidify with Ii""Br and evaporate to dryness. Heat with a smoky flame until all of the HBr has been driven off'. dissolve in H2 o and filter. dilute to the mark. Cool, Add to a volumetric flasl{ and Pipet a portion into a beaker, heat to boiling and titrate with standard AgN03 solution until the precipitate begins to coagulate rapidly. Filter and back- titrate with standard NH4 CNS solution using ferric alum as an indicator. 33 SYSTEM OF ANALYSIS For Liquor Containing Organic Ma tter Due to the black color of these solutions colored indicators cannot be used to determine end points; the constituepts must be separated from the solution as solids or gases. Upon acidifying most of the organic matter is precip:li:B.ted as a jelly like precipitate which is very difficult to hand le. A 3 gram sample of the dried liquor containing about enough so 4 = t- So3= to produce a .3 gram precipitate of Baso 4 produces about 30cc of a jelly like precipitate upon dissolving and acidifying. A vacuum can- not be used as an aid to filtering if the filtrate is desir•ed because of the foaming properties of the solution, a few cc. of the liquor being sufficient to fill even the largest filtering flask with foam. In some cases the pr esence of rosin in the s elution also increases the difficulty of filtering. It has been found that filtering may be done easily if about 50# / sq. in. pressure is used and the precipitate is suspended in a large volume of SupePcel to prevent its clogg ing the filter paper. An airtight container was made for a medium sized Buchner funnel, having an cpening for the stem of the funnel and a lid which could be taken off and put on easily. Upon acidifytng and filtering the black solution a clear li ght brown solution is obtained which may be titrated with I2 solution to the starch end point. Howe ver, since 34 I 2 oxidizes mercaptans to sulfides and sometimes f orms addition compounds with sulfides, (g) and since s03• may be oxidize d by dissolved o2 during the determination, this titration is not used in determining any of the constituents. Even if it were accurate it would be of no value unless a -separate determination could be made of s 2 o3 = or s03=. The method of anal yzing for thiosulfate by precipitating with AgN03 in acid solution is interfered with by the presence of some organic compound. Ifs= is removed by precipitating with amn1oniacal Zn(N03) 2 and filtering, the solution acidified with HN03 and the organic matter filtered off a clear brown solution is obtained. Addition of AgN03 produces a white precipitate which ra pidl y becomes black as in the case of S203=. However, upon the addition of an excess of NH40H a colloidal precipitate is produced which is almost impossible to filter. Such a precipitate is not produced when S203= alone is present. De termination of s= s= is determined in exactly the same way as given for a solution containing no organic matter. Determination of So4= S03= is determined in the same way as given for a solution containi ng no organic matter except that the I2 titration is not taken and the precipitate of organic matter obtained upon acidifying is filtered off before the determination of so 4 =. 35 The main sources of error in the determi nation of so 4 = are the precipitation of other barium salts than Baso 4 and the occlusion of organic matter by BaS04. If the determination of Baso 4 is made in the usual way an almost black precipitate is obtained upon ignition, and diserepencies are often as high as m%. The black color of this precipitate is not removed by treatment with H2S04 or with 02 at red heat. However, if the precipitate is heate d to a high temperature in an oxidizing flame on a platinum foil it suddenly looses its blaclmess and at the same time the flame assumes the green barium color. Like other organic barium salts this salt seems to be much more soluble in acid solution and at high temperatures. Its precipitation may be prevented by precipitating and filtering the BaS04 hot and in a solution of dorrect acidity. Occlusion of organic matter by BaS04 is reduced by the formation of larger crystals of BaS04. Procedure: Add the sample to 5cc of l M ammoniacal ZnCl2 in 150cc H20. Filter with pressure into a 600cc be<!l.ker and wash the precipitate and beaker thoPoughly with water. Acidify with acetic acid, add 100cc of Supercel, and filter with pressure. end point. Titrate with r 2 to the starch Repeat the precipitation and filtration of ZnS. To the filtrate add a solution containing about 20% excess I2 solution and 30% HCl to react with the reducing agent and base. Heat to about 40-50°c and allow the precipitate to coagulate. Filter, using pressure, a nd wash the precipitate thoroughly with very dilute HCl. Neutralize 36 to the methyl orange end point, add 2cc of HCl and determine so 4 =, filtering the solution while still hot. Determination of total sulfur. The total sulfur content of the liquor is determined by oxidizing with Br 2 for several hours , filterin g off the precipitate , and oxidizing it with Na 2 o2 . All the S04• is added together and determined by precipitating and weighing ( as Baso 4 . The liquid bromine reacts rapidly with the or- ganic matter producing a well coagulated rust colored precipitate which may be filtered easily. Inorganic sulfur compounds are oxidized by Br 2 to S04=. Also mercaptans and organic sulfides are slowly oxidized to sulfates. Procedure: To the sample in a 250cc flas k add about lee Br2 and cover with a watch glass. Allow to react for about 3 hours. Dilute to 100cc and filter, using a small Hirsh funnel. Transfer the precipitate to a nickel crucible add Na202 and heat. As soon as the fusion is complete pour most of the molten salt on to a sheet of nickel and allow to cool. Set the crucible in a 600cc beaker, dissolve the salts with dilute • HCl, and wash the crucible with water. Add the main portion of salt from the sheet of nickel and dissolve by acidifying. Filter into the filtrate previously obtained and return the total filtrate to the beaker. Determine S04 =. Determination of Na2 The determination of Na2 is carried out in the same way as given for solutions containing no organic matter. 37 Sample calculation of an analysis: The following data sheet gives a brief outline of the calculations involved in the analysis of a cooking liquor obtained from a plant of the Crown Willamette Paper Co. located at Port Townsend, Washington. s= and s2= were precipitated with ammonia.cal ZnCl2 from four 5cc portions of liquor. Two of the precipitates were titrated with I 2 , and two were oxidized with hot Br 2 water and so 4 = determined. equiv. cc I2 1.17 1. 59 25.29 26.26 average wt BaSO 24.61 24.65 .2957 .2959 24.63 .2958 r2 • .1021N Na2S203 • .1030N mols s= + mols s2= • equiv. I2 - (.02463)(.1021) - .001258 2 2 - mols S + 2 mols S2~ • mols BaS04 •• 2~58 • .001267 ~4 mols s2=/liter • (.001267-.001258) lOOOcc/liter • .0002 5cc in sample mols s=/liter = (.10258) 1000 ~ .0002 -s- :::: .R.4-96 s= and s2 = were precipitated and filtere d from two 10cc portions of liquor, the filtrates oxidized with acid I2 and S04= determined in the solution .85 .05 Average 7.55 6.75 equiv. cc I2 wt Baso 4 6.69 6.70 .3849 .3852 6.69 .3850 38 so 3 =+M s2 o3 =) per liter= (equiv.I2) 1000 cc per liter (2 mols 10 cc i n sample : {6.69)(.1021) a .0673 10 (mols so 3 -r mols s 04=) per liter= mols BaS04 1000 cc per liter IO cc in sample = 3850 (1000) = .1650 2'3':3:"4 ~ s= and s2 = were p reci pitated and filtered fro m two 25cc portions of solution with ammoniacal Zn(N03)2, the solutions acidified with HN0 3 and acid AgN0 3 added. NH40H was added, and the Ag2S determined. ccAgN0 3 ccNH 4CNS .oo 18.43 18.51 .06 NH4CNS : equiv. cc NH4CNS 18.43 18.45 18.44 • Avera ge .1021 mols S203 • per liter• (equiv. NH4CNS) 100 0 cc per liter 25 cc in sample = ( 18. 44 2 25 ( .1021) = . 03763 M so 3 = per liter= (.0673 - .0376) = .0148 The total S in t wo 5cc samples was determined by oxiding with Br2 water and determining s 04= . wt Aver. Ba S04 .5578 .5574 .5576 mols S/li ter • (mols BaS04) : 1000cc per liter 5cc in sample .5576 (1000) = .4775 ~4 5 39 Two 5cc samples were treated with Br 2 water and then with HBr and evaporated to dryness. 01- and Br- were determined by titrating with AgN03 and Nfl4CNS ccAgN0 3 equiv. ccAgN03 150.45 144.91 11.48 15.34 139.25 139.70 139.4 = Average AgN03 - .1025N NH4CNS - .1021N (mols Br- and 01-) per liter= (m AgN03) 1000 cc per liter 5 cc in sample = (139 .4) .1025 5 : 2.856 m Na2 per liter= (m S04=+ l/2m(Br-+Cl-)) per liter : s = s2= s2 o~so3_S04Total s= .4775 + 2.856: 1.905 2 m/liter m/m Na 2 .0002 .2498 .0376 .0148 .1502 .4775 .0001 .1310 .0198 .0078 .0789 .2510 2 s2 ·..,. s=-+ 2S203 + S03 =-+ S04 Total s - = .4904 .4775 %difference = ( . 4 9 04- . 4 7 7 5 ) 100 = 2. 7ft .4775 ;;: m/liter m/liter 40 Analysis of Liquors from Plant Using Wagner Fu.rnace The methods previously outlined were used in analyzing liquors obtained from a plant of the Crown iiillamette Paper Co., located at Port Townsend, Wash. Wagner furnace. This plant uses a The samples of liquor were collected at about the' same time so as to be comparable to one another. The liquors are as follows: I Scrubber liquor II Liquor entering digestors III Liquor leaving digestors IV Concentrated liquor entering furnace V Concentrated liquor after decomposing at about 5oo0 c, in the laboratory. The data are given in mols per mol of Na2 the different analyses are comparable. so that The analyses are as follows: I II III IV V .0000 .1115 .0351 .1008 .0001 .4310 .0198 .0000 .0001 .0000 .0000 .0865 .0477 .2354 .0000 .0000 .0000 .0048 .0745 .0745 .1565 .3315 mols Na2/liter .5339 .0078 .0789 .2510 1.905 .2481 .8045 Discussion of results: The analyses of the scrubber liquor and liquop entering digestors probably do not represent their composition at the time the samples were taken with an accuracy less than 5%. However, the analyses of the liquors containing organic matter probably represent their composition at the time the samples were taken with an accuracy of less than 1% due to the absence of s=. 41 The inorganic sulfur compounds in the liquor entering the digestors ares=: s2 o5 , so 3 , and s04=• reactions in the system are as follows: Their general All of the s= in the original cooking liquor reacts with organic matter in the digestion process to form organic sulfur compounds. Upon drying and heating the liquor to 5oo0 c, a temperature below the temerature of reduction of Na 2S04 By C, these organic sulfur compounds are given off as gases and liquids, their odor suggesting mercaptans or sulfides. In the furnace these organic sulfur compounds are burned to so 2 , and are absorbed in the scrubber. The scrubber liquor is therefore the nain source of inorganic sulfur compounds capable of being reduced to s= in the furnace. The reactions of s2 o3 = are unknown. Although Na2S 2 03 deco.mposes upon heating to about 3oo0 c according to the equation 4 Na2S203 =- 3 Na2S04 + Na2S5, the fact that no s= ors~= was found in the liquor after drying and heating to 5oo0 c does not indicate that no S20 3 = wss present in the liquor before heating because the sx= produced may have reacted with organic matter to form volatile organic sulfur compounds. The so 4 • and so 3 = are comparatively next in the digestion process and are reduced to s= in the furnace. Although the rati o of volume of scrubber li quor taken from scrubber to volume of li quor entering digestors per unit time is not known it is assumed that t h is ratio is approximately one. The reasons for t his assumption ar.e as 42 follows: The sulfur content of the scrubber liquor with- drawn from the scrubber per unit time should be at least half of the sulfur content of the liquor entering the digestors per unit time because the s= in the latter is converted to organic sulfur compounds, burned to s02 and absorbed Jn the scrubber. The scrubber liquor contains s= available for use in the digesting process; therefore the scrubber liquor is added to the cooking liquor before entering the digestors instead of after leaving. Also, if the scrubber liquor were added to the cooking liquor as it left the digestors additional evaporation would be necessary to concentrate the liquor. It is desirable to main- tain the concentration of s= in the scrubber liquor as low as possible in order to minimurnize losses of H2S; therefore water is added to the system in the scrubber. The scrubber liquor is then used to dissolve the molten salts from the furnace to produce the cooking liquor or liquor entering digestors. However, water is added to the cooking liquor in washing the precipitate of CaC03 obtained upon causticizing with Gao. The comparatively large amounts of Na2 ands= in the scrubber liquor cannot be explained by vaporization of inorganic salts in the furnace. The explanation offered is the entrainment of liquor sprayed into the furnace in the gases pas sing to the scrubber. The facts upon whict. this view is based are as follows: The scrubber liquor contains finely divided carbon particles amounting to about 5% of 43 its volume. Als o, the liquor entering the furnace upon heating forms a porous mass of about 5 times the volume of the liquor, as does sugar. The explanation offered for the s= on the scrubber li quor is the reduction of Na 2 S04 by C in the entrained particles carried to the scrubber. The facts , upon which this view is based are the rapidity of the reduction of Na2S04 by C andthe high temperature to which the particle s are exposed. It is thought that sub- sequent oxidation of Na2S to Na2S 04 by 02 is slower than the oxidation of C by 02. ber liquor. Carbon ' is present in the scrub- Also, the fact that gases from the :furnace are cooled down quickly upon passing through a boiler would decrease the oxidation of Na2S to Na2S 04. Discussion of possible losses of sulfur compounds. Losses of chemicals from the system can take place only in the rejected wash water used in washing the pulp and in the flue gases passing from the scrubber. assumed that the former is the smaller loss. S to Nt1.2 It is The ratio of in the rejected wash water is • __., given by the analysis of the liquor leaving the digestors. The ratio of Na2S04 to Na2C 03 added to the system is about 1/10, corresponding to a value of .931 for the ratio of S to Na2 The ratio of S to Na2 in the solids in the flue gases is about as given by the analysis of the liquor after decomposing at about 5oo 0 c. Consequently, the main loss of sulfur compounds is the loss of gaseous sulfur compounds in the flue gas passing from the scrubber. 44 The most probable losses of gaseous sulfur compour.ds a re the loss of H2S d~~ to hydrolysis of s= in the scrubber liquor, loss of unburned organic sulfur compounds, and loss of S02 due to incomplete absorption in the scrubber. Experiments on the combustion of the organic matter given off upon heating have met with complete failure possibly because the combustions were performed at only 70o0 -soo0 c and because enough time was not allowed for combustion. These sulfur compounds are given off as colloidal sized liquid particles which are almost impossible to absorb and which burn incompletely at 7oo 0 -soo 0 c to give c olloidal smoke particles. Fur•t her ex periments will be made to de termine the rate of combustion at about 1000°c to1100°c, the temperatures employed in the Wagner furnace. I mprovements in the recovery of sulfur• comp ounds should be directed toward the removal of Na2S from the scrubber, more complete combustion of organic matter, and more complete absorbtion of S02 in the scrubber. Removal of Na2S could be accomplished by the removal of solids from the flue gases before passing to the scrubber using a Cohrell preci pitator or bag filters. It would not be desirable to remove all of the :nids fro m the flue gases before passing to the scrubb er because Na2C03 in the scrubber li quor helps in the absorbtion of so2 . More complete combustion of organic sulfur compounds could be secured by allowing a longer time for combustion by increasing the length of the combustion zone. More complete absorbtion of S02 fro m the flue gases could be accomplished by c r~nging the desi gn of the scrubber. 45 It has been found that it is undesirable to add the double salt of Na2S04 and Na 2 C03, burkeite, to the furnace in the usual way by blowing it in ~dth air because this salt is so fine that a great deal of it is carried into the scrubber by the flue gases. This is undesirable be- cause in the evaporation of the resultant liquor Na 2 so4 is more liable to precipitate on the evaporator tubes. The alternative point of addition is to the concentrated liquor entering the furnace. Wagner states that addition of Na2S04 to this liquor causes solid precipitation with consequent clogging of the spraying nozzles. Experiments showed that the addition of Na2S04 or Burkeite introduced no solid other than the one added. However, the effect of this solid upon the spraying nozzle was not determined. However, the chemical effects of addition of Na 2S04 or burkeite to the liquor can be predicted. A large proportion of the Na2S04 would pass into the scrubber, either as Na2S04 or Na2S by entrainment of the liquor in the flue gases. Either condition would be undesirable. If Na2S were produced due to the reduction of Na 2S04 by C the losses of s= from the scrubber due to hydrolysis would increase. This would reduce the ratio of Burkeite to Na2so 4 added to the system. If the Na2S04 were not reduced it would increase the liability of precipitation of Na 2 S04 during evaporation. Analysis of Black Ash From Plant Not Using Wagner Fui-•nace: The methods previously outlined were used in analyzing a sample of black ash from a plant of the Crown Wilamette 46 Paper Co., located at Camas, Was h. This plant uses the system of recovery outlined in the introduction to this pa per. The black ash is li ouor containing organic matter which h as been dried in a rotary kiln by flue gases from the reduction fu1"I1ace at a fairly low temperature. The analyses ~re as follows, the data being given in mols per mol of Na2 I Black II Black Ash after decomposing at about 50o 0 c in the laboratory. s= so 3 = S04 S03 Total s Ash I II .ooo .0000 .0742 .1290 .1290 .1226 .2681 From these analyses and fro m the difficulty of burning the organic sulfur compoundsgiven off it was concluded that the large loss of sulfur compounds in this plant was due to incomplete combustion of organic sulfur compounds and absorbtion as so 2 . Since any improvement in this recovery system would mean a radical change in the design of the furnace it was thought advisable to discontinue work on this S%em. In order to make up for the large ratio of S t o Na2 lost from the system this plant increases the ratio added by adding NaHS04 with the Na2S04. of S This NaHS04 isbecoming increasingly difficult to obtain. A purely chemical means of making the addition of NaHS04 to the system unnecessary was devised which depends upon the separation of Na2C03 from the system. This method is given 47 as follows: The molten salts from the reducing furnace containing mostly Na 2C03 and Na 2 s are dissolved in just enough water to form a saturated solution of Na 2s. 9 H20 Under these conditions excess Na2C03 precipitates as Na2C03 • 10 H20. This precipitate may be removed from the system by filtering and washing. In order to determine whether or not this system would work the solubility was roughly determined of Na2S 9H20 and Na2C03. 10 H20 in a solution saturated with each. Enough C.P. Na2S and Na2C03 were dissolbed in hot distilled water to correspond to the solubilities of Na2S. 9 H20 and Na 2co 3 10 H20 in water. 'rhe solution was cooled, seeded with Na2S 9H20 and Na 2C03 10 H20, and kept in a thermostat at 25°c for 3 days. The solution was allowed to settle, a portion was pipetted off and diluted to a definite volume. s= was determined by precipitating and weighing as Ag 2 s, ands= co 3 = were determined by titration to the methyl orange end point with standard HCl. T,he results are given as follows: Solubility of Na2S 9H20 in g. Na2S/lOO g. H20 = 8.15 Solubility of Na2C03 10 H2 o in g. Na2C0 3 /100 g H20• 20.3 Ratio of s= to Na2 • .353 The solubilities in water are as follows: Solubility of Na2S 9 H20 in g. Na2s/100 g. H20 = 17.1 Solubility of Na2C03 10 H20 in g. Na2C03/lOO g. H20 • 29.6 Since the ratio~ of total S to Na2 used at the plant at Port Townsend is only .251 this method of separating Na2C03 from the system is theoretically possible. 48 SID.l[llJlARY A system of analysis has been developed for the - inorganic constituents, S=0 .. s= , s20=3. S0-3, and s0·4 as well as total sulfur in the liquors used in the manufacture of paper pulp by the kraft process. This system of analysis has been used in analyzing samples of liquor obtained from a paper pulp plant using a Wagner furnace in order to determine the most probable sources of loss of sulfur compounds from the system. The losses of sulfur compounds from the system have been found to occur mainly in the flue gases from the scrubber system. The possible sulfur compounds lost were found to be H2S, S0 2 , and organic sulfur compounds. Methods of increasing the efficiency of recovery of sulfur compounds have been suggested. A possible method for the removal of Na.,co 3 from the system has been discovered. ACKNOWLEDGEMENT Funds for this research were supplied by the American Potash and Chemical Corporation, and their cooperation in numerous ways is greatly appreciated. The Crown Willamette Paper Co. has been very kind in allowing the writer to visit two of their 49 plants and in supplying samples used in this research. This work has been carried out under the general direction of Professor W. N. Lacey at the California Institute of Technology. Professor E . H. Swift has kindly offered helpful suggestions.