The Activation of Sawdust Carbon for Decolorization Citation House, Harvey W. (1926) The Activation of Sawdust Carbon for Decolorization. Master's thesis, California Institute of Technology. doi:10.7907/HYWP-K634. https://resolver.caltech.edu/CaltechETD:etd-11112004-111421 Abstract No abstract. Item Type: Thesis (Master's thesis) Subject Keywords: Chemical engineering ; Chemistry Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemical Engineering Minor Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1925 Record Number: CaltechETD:etd-11112004-111421 Persistent URL: https://resolver.caltech.edu/CaltechETD:etd-11112004-111421 DOI: 10.7907/HYWP-K634 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 4506 Collection: CaltechTHESIS Deposited By: Imported from ETD-db Deposited On: 11 Nov 2004 Last Modified: 03 Oct 2019 23:35 Thesis Files Preview PDF - Final Version See Usage Policy. 3MB Repository Staff Only: item control page

THE ACTIVATION OF SAWDUST CARBON FOR DECOLORIZATION Thesis by Harvey W. House In partial fulfillment of the requirements for the Degree of Master of Science California Institute of Technology Pasadena, California 1925 THE ACTIVATION OF SAWDUST CARBON FOR DECOLORIZATION, Many methods have been suggested for preparing activated carbons for use in decolorization of sclutions. and this re- search was started with a view to testing certain of the methods suggested, with special reference to the action of the product upon unbleached vegetable oils and brown sugar solutions. One of the methods suggested was the following. "A decolorization carbon is prepared by adding to 100 parts of hard wood sawdust, 100 parts of a 75% solution of niter cake in water, allowing the mass to stand for two hours or somewhat longer, drying and charring the mass by heating it in the open air to 300-500 degrees C while stirred in thin layers for $ to 1 hour, cooling, adding sufficient water to dissolve out the sulfate salts, boiling, filtering, drying, grinding. then heating in a closed retort to 800-1000 degrees C, cooling to below the ignition point, and regrinding. Ray! sSuumarized the types of processes used for produce ing activated carbons as follows: 1. Processes depending upon action of inorganic chem- ical compounds either naturally present or added to prevent the formation of the adslorption complex during carbonization er to cause the breaking down and elimination of the adsorbed hydrocarbons during the succeeding calcination. 8. Processes depending upon solvents to eliminate the hydrocarbons. lpay. Chem, and Het... 28, 977 (1933). ~pu 3. Processes depending solely on long continued calcin- ation to eliminate the hydrocarbons. 4. Processes depending upon selective oxidation to beeak down Gnd remove adsorbed hydrocarbons and to alter the porosity of the carbon. He also gives the following comments regarding each of the four types: 1. Most processes in this group are of no value for gas adsorption. Kelp and rice hulls contain at the start a certain amount of the proper materials. 2. Use of solvents neglects removal of solvent itself, and control of porosity of the carbon. 3. This activation is due to fortuitous presence of oxygen in pores. Activation by heat alone cannot be done, for heat eracks the hydrocarbons, depositing inactive car- bon. "Active" carbon can be very easily "deactivated"® by such treatment. 4, Selective oxidation is by far the most effective for producing highly activated carbons of particular struct- ural characteristics. Only these processes now make possible production of commercial gas and vapor absorbent. The Cheney process, based upon this last method, depends upon three fundamentals which must be apprehended. 1. There are two kinds of carbon, the active and inactive kind. 3. The active form must be liberated at low temperatures (below 500-600 degrees). = Ba 3. Active carbon is available only when adsorbed hydro- carbons (adsorbed during distillation of the carbonacious material) are removed. All types of primary carbons can be activated by this process topreduce many types of carbons suitable for different purposes." Since method 4 that of selective oxidation. was so effect- ive for producing highly active carbons of particular structural characteristics, it was thought that 1t would be best to attempt to activate samples by various modifications of this method, in order to determine whether or not carbons possessing oil and sugar decolcrizing power cculd be produced by it. and if so, to find out the combination of conditions which would produce the best results. The information is also given in the above report that superheated steam at about 1000 degrees can be used as well as air, and that it possesses the advantage of taking part in an endothermic reaction which can be more readily controlled than the direct oxidation by gases containing free oxygen. Carbonization No. 1 (Oregon Pine Sawdust.) This was a trial carbonization, carried on in a five inch hemispherical iron crucible, open to the air. The charge was heated over a bunsen fleme and stirred with a thermometer which was kept below 360 degrees. Caxrbonization No. &. For this purpose 2 flat bottom iron pot was used, about 1l inches diameter, 5 inches high with a cover. The charge was Stirred with a glass rod, using @ thermometer occasionally to determine the temperature. The complete run is as follows: Min. Temp. C 0 Begin heating. Cover off. 18 280 Smoking slightly. 19 300 " " 22 360 n " B83 " Xl a4 Put cover on, slightly tilted, took off to stir occasionally. Turned ges down a little. Slight smoke comes out uncer cover. 40 ‘Turned gas still lower. 69 Gas lower. Sparks of carbon appear on surface of charge when cover is off. Left the cover on. 80 Shut off gas. and allowed charge to cool, covered. wee em Carbonization No. 3. Same apparatus as No. &. Sawdust, 925 ems. C obtained 304 gms. Min. 0 Start heating. ‘Stirred continuously until smoking was nearly over. 8 First smoke. 30 Very rapid smoking. 40 Smoking nearly over. Lowered flame and covered loosely. 42, Shut off gas, and allowed charge to cool, covered. Activation of Carbon by Superheated team Carbon to be activated was placed in a silica tube (7/8" inside diameter, 124" high), which in turn was placed in a vertical position in an electric furnace, The tube was filled with chips of porous plate up to about the mid-point of the furnace and the charcoal filled the space from that point to the upper end. The lower part of the tube, containing porous Plate chips extended several inches out of the furnace, and was heated by a gas flame to cause supetheating of the steam, which entered the lower end of the apparatus. The steam wa's obtained by bolling water in a flask, and the amount used could therefore be measured by the amount cf water evaporated. Dur- ing the first series of activations the gases escaping from the tube were passed through a condenser and the condensed water measured which amount subtracted from the total water evapor— ated in the flask, would give a measure of the steam actually consumed during activation. The amount of charcoal used and obtained was also determined in each case, Most of these measurements, however, were inaccurate, due to such causes as accidental steam leakages, and occasional loss of charcoal puffing over into the condenser. The principle decided upon, however, was to pass such a Large excess of steam through the charcoal in order that (1) the actual amount passed through would be of far less import- ance than the time and temperature of passing, and (3) the amount of steam supplied to the charcoal in the far end of the tube would be about the same as that supi:lisd near the steam entrance. Temperature control was secured by means of an autotrans— former, on a 280=—volt circuit, Temperature msasurement was secured by means of an "Advance" chromel thermocouple, calibrated by comparison with an ordinary thermometer in water, up to boiling, and in an air bath up to about 200 degrees C, and by the boiling point of sulfur (444 degrees), and by obtaining the cooling curve of sodium chloride, the flat portion of the curve corresponding to 801 degrees C, the solidification point of NaCl. The ther~ mocouple was placed in a small silica tube, and inserted into the top end of the activation tube, so that the end of the -Jo couple was in the middle part of the charcoal column. In most eases the procedure of activation was as follows: Heated the furnace and tube up to the temperature desired, then passed steam through, superheating as described. Temperae ture was then controlled both by varying the voltage on the furnace and by varying slightly the amount of steam passed, During the latter runs, after the temperature had become about constant at the point desired, it was thereafter kept so en~ tirely by varying the amount of steam, leaving the furnace voltage constant. Considerable trouble was met with in the apparatus as first set up, due to condensation of moisture in the portions of the lower end of the tube not easily reached by the gas fleme, and by consequent clogging and cccasional puffing out of charcoal into the condenser tube on sudden liberation of pressure, This trouble was entirely eliminated thereafter by inclining the furnace and tube at an angle of about 45 degrees, and passing steam in at the upper instead of the lower end, In this case the steam was superheated before entering the tubs, oy passing. through several inches of steel tubing heated to dull redness by a fish-tail bunsen burner. The activation tube was placed in the furnace so that the top end was near the mouth of the furnace. The porous chips were placed in the lower end 2s before, but only up to a level corresponding to the lower end of the furnace. The charcoal then occupied the lower half instead of the upper half of the furnace, so the ~ Que steam was further superheated before reaching the charcoal by passing down through the tube in the upper half. The thermocouple was stuck into the upper end as before, but was still able to reach down into the middle part of the charcoal. The vapors escaped, of course, at the lower end, to which the condenser was fitted. The condenser was later @liminated, it being decided that further measurements of condensed water were not needed, and in its place simply a glass tube inserted to lead the gases away a few inches from the furnace. It was decided first to run a series of activations, passing steam for various lengths of time at each of several different temperatures between 600 and 900 degrees C. Since the procedure followed was practically the same in all Cases, and corresponded to the description already given, only an outline of the various activations need be reported, The following series of graphs give the relationship be- tween the temperature and the time in each of the activations, respectively, together with the average temperature of each activation as determined by integration of the corresponding temperature-time curve, and the yield of active charcoal. 12 grams of charcoal were taken at the beginning in each case, so the difference between this and the yield gives the loss. (/9 £50 os, 4 ares, SY ede oh ED Summary of Activations Charcoal Temp. Time (min.) Co. water Tube Gms. ¢ evaporated obtained 2c 725 1 47 8,4 ad 735 lle Vv 7.8 28 725 15 89 Vv 8 af 732 19 19.5 Vv 8.5 2g 7238 30 50 Vv 8 an 775 9 Vv 9 Bi - ~ ~- =~» - a4 735 45 87 I= 6.5 ak 780 15 30 I 6.6 21 774 5 8 I 9.9 arm 773 10 ae I 8.9 2n 847 #15 34 I 6.5 26 840 10 17 I 6.5 2p 828 5 11 I 7.8 2G 893 15 56 I 8.5 ar $0? 10 I 5.9 28 886 5 I 7.3 Bu 692 55 90 T 7.67 av 692 75 I 6.58 3a 738 45 64 I 5.85 3b 736 45 64 I 5. 4, 3c 736 43 98 I 5.8> 3a 736 45 81, I 5.86 Bay signifies vertical activation tube. Br signifies inclined tube. osteam not preheated as much as in other oases. The large amount of water required in this case shows that the pre- heating affects this factor considerably, although it does not affect the final yleld very much. The voltage on the furnace has not been recorded here, but 1t was the same for da, b, oc, and d. It was decided at this point to eliminate preheating after this experiment, except enough to insure the steam being dry at entrance, inasmuch as the wpper part of the activation tube itself would accomplish this. md, Qa The variables in connection with these activations will be given in tabular form in connection with the reports on the decolorizing properties of the various charcoals produced. In order to test the decolorizing power of the different charcoals, three types of solution were used. 1. Methylene Blue. - A stock solution was made up con. taining 0.0829 gram per liter of solution. This was called "golution a" and used in all decolcrization tests. The same solution was used 21so0 as a-.standard in color determination, and its color concentration arbitrarily called 100 color units per ce, (100 cu/ec). Lighter standards were made from solu- tion-a by dilution, so as to give the following color concen~ trations (assuming for this work that color concentration is proportional to solution concentration). a. 0.0829 gms./liter 100 cu/ec, b. 0004145 gms. /liter 0.5 cu/ec, c. .00889 gms./liter 10 cu/ce. 2. Sugar Solution. - Made each time by dissolving 50 grame of dark brown sugar in water and diluting to 100 cc, This was called "sugar~a" and was used in all sugar decolor- ization tests. The same solution was used as a standard for color determination, and its color concentration arbitrarily Called 100 color units per co. (cu/ce). Lighter standards were made from solution-a by dilution, so as to give the following color concentrations (assuming as in the case of -Ll- methylene blue that color concentration is proportional to solution concentrations ). ae 100 cu/ce, Dd. 50 " C. 85 " d. 18.5 % 3. Prime Summer Yellow Cotton Seed O11 from Los Angbles Soap Company. This oil, called "oil-~a" was used in all oil decolorization tests, and also as a standard for samples which were but slichtly decolorized. For samples highly desolorized which had a different tint from the original, it was usually more satisfactory to use bleached, deodorized oil furnished by the same company, which had a tint more nearly like that of the charcoal decolorized oil. It was of course impossible to evaluate the color of this bleached oil accurately in terms of the color of oil-a, but a value of 18.4 cu/ce was taken as the nearest approximation, While the use of this bleached oil as standard, with this somewhat arbitrary value of color concentration. failed to make the accuracy of the color deter. minations any greater than they would heave been if clle® hud head been used as standard, yet the color concentration values so obtained were accurately comparable with one another, and it was this comparison of the lighter samples of each acti~ vation with one another which was of greatest importance, wl Be Methylene Elue Decolorizations 85 co. methylene blue solution-a& was mixed in a 100 cc, Erlenmeyer flask vith 0.2 gm. of each charcoal to be tested, and allowed to stand several hours at room temperature with occasional shaking. The color of each solution was then de- termined by comparison with the standard nearest it in color, 10 co. being pipetted out for the purpose, The samples used in color determination were then poured back into the mixture from whence they came, and the color of these determined again after about 24 hours. Finally, after several days standing, the solutions were all verfectly colorless, including even the solution in contact with charcoal 38h, which was the poorest of them all. The results were as foliows: Charcoal Standard Cu/ec. Standard Cu/co. (a few hrs.) Gl day) 20 c 11.3 c 0.86 ad Cc 12.7 c Be 48 2e c 17.1 c 3.89 af c 11.2 c 3.83 ag c 8.4 ; an a 17.3 c 15.06* ai e 0.838 ai & 2.5 ak a 1.8 Al a& 12.5 re] 2.3 2m a 8.4 an a 3.8 20 a 7,0 2p & 2.9 2q & 1.0 er a 64 2s a 2.3 *This eventually became completely decolorized, but took longer than others, Blank speces indicete that sample has been almost or entirely decolorized. m1 gw The fact that these solutions all became colorless shows that the first set of color determinations made was only a measure of the rate at which color was being removed rather than an indication of the final color removal power of the carbons. The experiment also shows that such a large amount of cclor is adsorbed by a small amount of carbon more concen trated solutions and less carbon would have to be used, in order to get comparative results as the mixtures approached equilibrium (color saturation of the acrbon). Since it was especially desired to study these charcoals with respect to sugar Solutions and oll, the time was not taken for continue ing the experiments with methylene blue. Sugar Decolorizations. {Each decolorization experiment will be given a Roman numeral as a designation, and may include tests on one or more charcoals.) Sugar decolorization I. Charcoals gc to 28, and an unectivated charcoal designated as 2%, were tested in the following manner: 1 gm, of chercoal and 15 cc. of sugar solution-a ware placed in a 100 cc, Erlenmeyer flask, shaken, and allowed to stand overe night. Each flask was shaken ccecasionsylly during the afternoon, and on the following morning. Contents were then warmed and filtered. Results were as follows: ~14- Charcoal Standard Gu/cc. Charcoal Standard Cu/eo. 20 I 10.7 21 i 32.8 2a d 11.8 2m d 7.0 Be a 11.6 an a 7.9 af d 7.0 Be d 10.2 2¢ d 7.2 8p a 13.0 Sh b 37.2 24 a 12.4 24 - - ax d 13.7 24 a 4.8 28 d 12.1 2k ad 6.7 2% b 54,8 These results when grouped according to the conditions under which the charcoal wes activated, showed smooth relationships between the conditions of activation and the decolorizing power, showing thet although equilibriwn had not been reached (as in the case of methylene blue), and there was a probability of unequal shaking of the different samples, yet their bee havior as shown by these tests actually did have a close relationship to the conditions of activation. Grouping of results of sugar decolcrizations I, III, and IV according to conditions of activation of charcoal: Samples activated at about 725 degrees C. Charcoal Temp. Time Tube Gms. C Loss % Loss Cu/ece. Oil (min. ) obtained Be 725 11.5 v 8.4 3.6 30 «10.7 Ze 725 15 #V 8.0 4.0 33 11.6 91.6 2¢ 728 30 v 8.0 4.0 33 47.2 658.3 Samples activated at about 735 degrees C. 24 735 11.5 V. 7.8 4.2 35 11.8 10 af 732 419.0 WV 8.5 3.5 29 7.0 34 735 45.0 re 6.5 5.6 46 4,8 Samples ehiveted at same temperature as above. but de- ior a ‘ehorter period of time. From sugar decolorizad onl colorized over IV, placed here to how that although conditions of decoloriza~ tion very greatly at fact the result, a certain process of ectivation may be repeated and the samples so obtained will have similar decolorizing ‘properties when treated in the same way. (These samples should ‘ke treated along with some of the former carbons at the same time, to give results comparable to the action of the former charooshe. ) 3a Te sa Sc 3b 136 #45 #241 4 66 55 283.6 , ; 30 736 42 I 5.85, 6.2 528 283. 2y - vertical tube. & . dnolined tube. (See notes on apparatus) c ; “Less preheating of steam. ~-16~ Samples activeted at about 775 degrees C, Charcoal Temp. Time Tube Gms. © Loss Gloss Cu/eo. Oil (min. ) obtained | 8k 780 15 I 6.6 5.6 47 6.7 75.6 2am 773 «10 I 8.0 4.0 33 7.0 84.2 21 174 5 I 2.0 3.0 85 38.5 49.7 2h 795 9 Vv 9.0 3.0 85 37.2 116, * Samples activated at about 840 degrees C. 3p 828 5 I 7.8 4.4 37 13.0 91, 20. 840 10 I 6.5 5.5 46 10.2 63.8 en 847 15 I 6.5 5.5 46 7.7 45.7 Samples aetivated at about 900 degrees C. q 893 «15 I 2.5 9.5 79 18,4** ar 907-10 I 5.0 7.0 58 13.7 35.6 28 886 5 I 7.3 4.7 39 12.1 56.0 Samples activated at about 692 degrees O (Decol. III). 2u segs I 7.7 4.3 86 13.5 104.47 av 6eas75 I 6.6 5.4 45 9.7 94.6 2m | 10.3 ak 6.7 *Show effect of vertical arrangement, with trouble during activation. tkNote very Little difference between long and short runs at this temperature. tSemples of charcoal 2k and 2m were decolorized with these under the same conditions, and the resulting solution from charcoal 2k used as a standard, assigning to the latter the arbitrary value of 6.7 cu/cec to make it correspond to the solution from 8k in decol. I. The value obtained from 2m when so determined should have been 7,0 to check with decol.I The actual value obtained being 10.3, shows that this method is not dependable. The comparison between 2u and 3v decolor- ized solutions, however, is just as accurate as in the other cases. , a ae Sugar deccolorization ITI. Charcoels 2u and 8v, 1 gm. each, with 15 cc, sugar solution, as in decolorization I. These samples were accidental- ly put in oven with oil samples and heated to 108 degrees along with oil decolorization II. They each gave dark-ecolored solutions, as follows: Bu, 85.8 cu/ec. av, 30.1 " Sugar decolorization III. The procedure in this case was the same as for sugar decolorization I, the purpose being to determine the decolor- izing power of samples from activations 8u and av in terms which could be compared to the decoloriging power of the other charcoals as determined in sugar decolorization I. (See also the grouping of results according to conditions to activa- tion of the charcoals. ) Charcoel Temp. Time Tube Gms. C Cu/ec. (Min.) . obtained au 692 55 I 7.7 13.5 av 69875 I 6.6 9.7 an 10.3 ak 6.7 Sugar decolorization IV. Procedure same as for decolorization I (1 gm. char- coal, 15 ce. solution-a in100 cc. Erlenmeyer flask, shaken occasionally, stood overnicht.) 3a (1) 732 45 I 5.8 21.5 - th} aie5 Av. - 22.1 So (1) 736 45 I 5.4 23.3 sy, — 23,8 (2) 23.9 3c fh Po} 736 42 Sugar decolorization V. was all there was left of 3a). night with occaslionel shaking, as in other decolorizations, and filtered, Charcoal Temp. Time Tube Gms. © Cu/se. 3a 3b 3G od 5.8 2 4 a ~18~ 8.4 8.3 Av. - 38,3 0.83 gm. charcoal (this amount used because that 7338 736 736 736 (min. ) obtained 45 Tt 5.8 18.6 45 I 5.4 o15,5 43 I 5.8 017.3 45 T 5.9 019.1 Repeated the series offol~ or readings: 16,1 13.2 15.7 18.4 25 ec, sclutionwa, stood over= Decol, TV. 22.1 83.6 28.3 The second column of color readings was made using shorter Columns of Licuid in the colorimeter than in the case of the first column. It will be seen that the values are #11 unliform- ly lower than those obtained in the first column. The results of decolorization IX are placed here also for comparison, As will be seen, they do not vary in the same way as do the sam Ples of decolorization V. This discrepancy may be due partly to irregularities in the conditions of adsorbtion in the dif- forent samples, and partly to difficulty in matching due to cloudiness, especially in the standard solution-a, which could not be filtered out with filter paper, and which could not be treated with charcoal without changing the color. Summary of Results of Sugar Decolorization, up to and Including Sugar Decolorization V. The best sugar decolorization carbon so far obtained was produced from sawdust charred at a temperature probably below 400 deerees C, and activated by passing superheated steam through it at 735 degrees for 45 minutes. Other charooals produced within this general region were alse active. For any given temperature of activation, she decolorizing power of the char increases with time of activation. This of course also causes a proportionate increase in charcoal loss due to oxidation. But for temperature above the region of 735 degrees, a given loss of charcoal will correspond to a smaller and smaller amount of activation of the remaining charcoal as the temperature increases, until at about 900 degrees there is very little difference between the decolor- izing power of charcoal produced by passing steam only a short time, with small loss; and that produced by passing steam a longer time with larger loss. At temperatures below 700 degrees (at 692 degrees bY experiment), the rate of oxidation decreases markedly, it requiring nearly twice as long to produce a given loss in weight of carbon as at 735 degrees. It appears also that the decolorizing power of the remaining charcoal is less than that of a charcoal activated at 735 degrees with the same loss tn weight. Difficulty in color comparison, particularly with the standard, and uncertainty as to the effect of slight non-euni- formities in conditions upon the various decolorization ex- periments, makes it desirable to obtain better standard of color comparison and to work out a method of decolorization capable of giving more uniform results with a given carbon than have been obtained so far. Oil Decolorizations O11 decolorization I. Samples of the same charcoals as used in sugar de=~ colorization I were tested with oil by the following method. 25 ce. of oll-a pipetted into a 100 cc. Erlenmeyer Plask, to which 1 gm. of the charcoal was also added. These samples were shaken up, and stood in the open for two days, stoppered, They were then placed on a ‘tray and heated in the oven at about 108 degrees C for 5 hours, and then filtered while hot, The most noticeable results of this test were: Although some of the samples removed about 2/3 of the color from the oil, certain others made the oil even darker, The darkest oil so produced was from the sample which had been activated at 775 degrees for only 9 minutes in the first apparatus set up, and which hed lost but 25% of ite weight, as compared to a loss of about 50% in the case of some of the better charcoals. This phenomenon of darkening (taking on a reddish color) is a complicated one, and will be referred to later. ~21- Oil decolorization ITI, Charcoals 2h and 3j were again tested by the same method, together with samples of 2u and 2v. The oven was kept at about 108 degrees as before. This time oil with 3h was not dark as it had been before, whereas oil with 8j was not as light. That is, both were changed in the same direction as formerly, but not as much, indicating that some unnoticed factor in the decolorizing had had a critical influence, Charcoals 2u and 8v did not decolorize as well as 84, To determine the effect of time upon this type of decolor- ization, four samples of 2c were mixed with equal quantities of ofl (1 gm., 25 cc.), and heated for varying lengths of time in the oven at 108 degrees, All the samples were given @ reddish color which made accurate comparison with oil-a as as standsrd impossible, but such egomparison as Was nade indicated that this sample changed very little after the first two hours, although one of the 5 hour-20 minute samples seemed to become darker, (See Table, O11 decolorization II.) Oil decolorization IV, _ (This is reported here instead of oil decclorization III, because of similarity of method and purpose to oil de- colorization I and II). Since so much trouble was had on account of the oils becoming reddened, it was decided to repeat the method of of1 decolorization I, using samples of the same charcoals and the Same amounts of oil, but running at a lower temperature. Results of this test showed that the oils (heated at 6S degrees for about 6 hours) did not in general decolorize as much as the oils of decolorization I, but the charcoals which produced very dark oils in decolorization I did not produce such dark olls at this lower temperature. It seemed as though the power of the charcoals to "put in red" and "to take out yellow" were both decreased at the lower tempera ture. (Sse Table, O11 decolorization IV.) O41 decolorization III (Charcoal 2f, 5 gms., 150 cc. oil-a) In order to study the effect of agitation and also of time upon the decolorizing action of a charcoal upon oll-a the following experiment was conducted, 5 gms, of charcoal 2f and 150 oc. oll-a were placed ina wide-mouthed, short necked, 250 cco, round bottom flask, fitsed with a thermometer and high speed stirrer, The mixture in — the flask was then stirred briskly and kept at about 90 degrees by a bunsen burner. During the first half hour, however, the temperature went up once to 103 degrees for a short time. 10 cco. samples of the mixture were pipbtted out about every hour, and filtered. During the test, however, the wixture was allowed to stand. twice overnight at room temperature, This test sheiconsiderable light upon the formation of the red color, but did not explain it. It was... found that during the first two hours of agitation the oil became darker and darker, taking on a reddish tint as it did so. Then on standing overnight the o1l became still darker, Agitating —23— and heating, however, for two hours more after that caused the red color to drop enormously, until the oil had lost its reddish tint completely and was only about 3/3 as yellow as the original oll-a. Further agitation caused the oil to lighten still further, but with the taking on of a greyish tint and the exhibition of the Tyndall effect. On standing overnight at this period, the oil became slightly lighter, instead of darker as during the early stage of the run. The same samples which had been taken from this run were tested again as to color after standing clear in their res pective test tubes for two days. It was then found that the samples which had previously been much darker than the original oil were now considerably lighter than the original, and the decrease from their darkest color had been enormous-in the most extreme case from about 196 cu/cc to about 59 cu/ec. The lighter samples had also decreased in color, but not to S0 great an extent. The same samples were tested again after about 10 days total standing. In ell cases there was @ decrease in color from the previous reading, except wherte the two readings were practically the same, in which case the difference between the two was greater than the Limit of accuracy of the color determination. The amount of the cclor decrease was, howeveg, very small compared to that which took place during the first amo days of standing. (See Table and curves, oil decoloriza- tion III.) tk oad O11 decolorization V. (Charcoal 2g; 4 ens. , 120 oo, oil-a) Since there was such a great increase in color during the first part of oi1 decolorization III, it was thought that this darkening might have been due to the accidental increase of the temperature up to 103 degrees during this period in the case of o11 decolorization III. Another run was made, therefore, and at the lower temperature of 80 degrees, using a water bath to increase the stability. Standing overnight was also avoided during the early part of the run. As will be seen from the curves, shown later, even at this Llowsr temperature there was an increase in color (with reddish tint) during the first stage £ the run, but this color began to fall off before the first five hours of stirring had been completed. Thereafter, on standing within the flask in contact with the carbon, the oil became Lighter, thus exhibiting the same effect inside the | flask in contact with the carbon as had been shown by the samples in decolorization III on standing after being filtered. Also, the semples taken during the earliest part of the run became darker after standing two days, and then lighter again efter standing four days, whereas the samples taken during the latter part of the run became lighter and lighter. The final oils (after 4 days), however, up to and including the sample taken after four hours! stirring, were all darker than the or- iginal, whereas stirring for two hours more caused a very great decrease in color, the final sample becoming ultimately only about 16% as deeply colored as the original oll-a. These latter samples, however, as in the case of oil deccolorization III, pos~ stessed @ slightly grey color and Tyndall effect, Bie 2a - 3.3 gms. Oil decolorization VI. (Charcoal go ~ 0,7 ems, ), 200 cc. of 4.0 ems, & Oil decoloriz&tion V showed that the aprearance of the red color in the oil was not due to heating over 100 degrees, and that the oil heated only to 80 degrees also took it on. Therefore it was decided to carry on the runs thsreafter in a, bath of boiling water, to insure uniformity of temperature and olso a sufficiently high temperature to get fairly rapid decolorization, In order to study the action of another charcoal under similar conditions, a sample of 2d (with a small amannt of 20, to make up 4 gms.), and 200 cc, oil-a were treated in a menner similar to that of o11 decolcrization V. The curve of the color of the samples taken immediately after removing the sample was similar to that obtained in o11 decolorization V, except that the oils did not become nearly so dark, Stirring was also carried out longer than in oil decolorization V, so the oils became lighter than that of the last sample of de- colorization V, observed immediately after withdrawal from the flask. (Readings on standing have not yet been taken.) O41 decolorization VII. This decolorization was carried cut in a manner sin- 4lar to thet used in the others, except that a 200-cc, Erlen- meyer flask was used, fitted with a large stirrer, capable of moving almost the whole mass of oil and charcoal within the flask. This stirrer was then operated very slowly (one revolution in about 2 seconds), which was just enough to keep the mass moving and most of the charcoal off of the botton, or moving over the bottom. Using 3 gms. of charcoal 2%, with 200 cc, cil-a, and run- ning with the above stirrer at about 95 degrees in a water bath gave samples about 50% as dark as oil-a after 5 hours stirring, but the color did not decrease very much more after 13 hours stirring, the last sample being about 30% as dark as oil-a. The grey color and Tyndall effect, however, were absent, so fear as could be observed. This would indicate that using a better charcoel undar these conditions might decolorize the oil to a greater extent snd still avoid the grey color and Tyndall effect. Comparison of Suvar and Oil Decoloizations. The charcoals from 8c to 8s can ve listed in the order of their sugar decolorizing power according to sugar deccloriza- tion I. If then the results of o11 decolcorization IT and IV be Listed opposite the corresponding carbons , it will be seen that except in the case of the best and the poorest carbons, there is no relationship between the decolorizing power of a carbon with respect to sugar and that of the same carbon with respect to oil. However, if the results of the oll decoloriza— tions are grouped according to conditions of activation of the charcoals involved, as was done in the case of the sugar de- colorizations, it is found that in general for any given tem- perature. as the time cf activation increases, the decolerizing power also increases, as in the case of sugar decolorization, but the difference between high temperature and low temperature wi oe activation does not appear to be as great as in the case of the sugar solutions.* It was intended to carry on this research from this point as follows: Prepare @ standard sclution, either of sugar or nonesugar materials, which will be olear, of the same color as partly decolorized sugar solutions, and reproducible, Work out a standard method of decolorizing sugar solutions, which will not be subject to criticsl factors difficult to con~ trol, as has apparently been the case with some of the sugar decolorizations so far carried on. Increase the rate of stirring of the oil decolorizations to see if this can be done up to a certain point without producing the grey color and Tyndall effect. Use the rate so decided upon as a standard for other decolorizations. Adopt a uniform procedure of sampling and color determinstion with respect to time, so that all results can be compared to one another. Carry out a series of activations similar to the first series, and use the charcoals produced from these in standard tests on sugar solution and olla. *These values are placed on the chart of sugar decolorizations, grouped according to conditions of activation. PART IT Following the proposals made at the end of Part I, a larger furnace was first made for the purpose of producing larger samples. Description of Furnace A silica tube, 36" Long by about 8" inside diameter was placed vertically within an iron flue which in turn was heat insulated. Gas was fed in at the bottom of the blue, and both air and gas fed in seperately at other points higher up. <A charge of 110 ems. of charcoal was held within the tube by @ perforated porcelain disk supported on a nichrome tripod which rested on a cork containing a glass outlet. tube. This charge filled the silica tube to a depth of about 30". Steam was fed in from the top, thereby cecoming preheated by passing down the unfilled portion of the silica tube. Three thermocouples ("Advance"” and chromel) were placed within one pyrex tube, so as to register temperature at the bottom, middle, and top of the charge. This tube was kept as near to the center as practicable. Another similar thermo- couple placed near the edge of the charge, and at the same height as the central thermocouple, indicated that temperatures near the edge and at the center of the charge were practically the same, for which reason the outer thermocouple was not used after the first few runs. The chief difficulty involved in the operations of this furnace was the control of temperature, the latter being very difficult to maintain both constant and uniform. In fact this was not found possible to accomplish, so the best that could be done was to keep a continuous record of the temperatures at the three thermocouples, and to determine the average of these continually changing temperatures, calling this latter the temperature of ‘the run. This was not theoretically sound practice. A momentary high temperature would not be offset by @& Similar period of low temperature running, since the former action is very much more rapid. However, it was assumed for this purpose that the variation in temperature was sufficiently small to allow the averaze temperature to be used as represent— ine the temperature of the run. 4. Another disadvantage of the gas fired furnace used, be~ Side that of temperature control durine the run was the var- lable length of time necessary to get the furnace and contents to the temperature desired for passing steam. There was always some distillation of residual hydrocarbons during this period, as shown by the efflux of a luminous burning gas at the furnace outlet, and where the period of pre-heating exceeded the time hecessary for this distillation, which was usually the case (varying from $ to 1 hr.) there was doubtless some activation due to airor water vapor present before the steam was regularly pessed, Excess stesm was always used, but some variation in lts amount between runs may have influenced the results, The sawdust was carbonized in the same manner as previously. Three different lots were made durinz this series of experi- ments, each being a mixture of several individual batches. Since each was made in the same manner from the same material, and consisted in a mixture of several batches, it was thought that the average mixtures so obtained would be practically the seme as though the entire amount used had been made up at first and mixed. It was observed, however, that the sawdust in the bottom of the sack, used in the last lot (no. 5) was somewhat lumpy, due to its having absorbed rain water. The lumps alse had a slightiy darker appearance than fresh Sawdust indicating that the water may have been somewhat muddy. This was used just as it was in preference to fresh sawdust, partly out of curiosity to see if such poor material would yicld good char, although it was thought that these new factors would not affect the result aporeciably. The charcoal so produced was found to be considerably less dense than that formerly produced, as shown by the fact thet the beaker formerly used to hold the weighed sample would not hold the weight needed (occupied about 20% more volume). The activated chars from this charcoal were also lighter, blacker, and more active than the others. Unfortunately, since both face tors were Changed at once, it is not definitely known whether these latter facts were due to the character of the sawdust or to the different conditions of activation, but since the in- erease in activation was in the direction which is indicated ~21- Oil decolorization ITI, Charcoals 2h and 2] were again tested by the same method, together with samples of 8u and 2v. The oven was kept at about 108 degrees as before. This time o11 with Sh was not dark as it had been before, whereas oil with 3j was not as light. That is, both were changed in the same direction as formerly, but not as much, indicating that some unnoticed factor in the decolorizing had had a critical influence, Charcoals 8u and 8v did not decolorize as well as 2). To determine the effect of time upon this type of decolor— ization, four samples of 2c were mixed with equal quantities of of1 (1 gm., 85 cco.), and heated for varying lengths of time in the oven at 108 degrees. All the samples were given @ reddish color which made accurate comparison with ollea as as standsrd impossible, but such comparison as was made indicated that this sample changed very little after the first two hours, although one of the 5 hour-26 minute samples seemed +o become darker, (See Table, 011 decolorization II, ) Oil decolorization IV, (This 1s reported here instead of oll decolorization III, because of similarity of method and purpose to oil dew colorization I and II). Since so much trouble was had on account of the oils becoming reddened. it was decided to repeat the method of o11 decolorization I, using samples of the same charcoals and the same amounts of oil, but running at a lower temperature. Results of this test showed that the oils (heated at 6S degrees for about 6 hours) did not in general decolorize as much as the oils of decolorization I, but the charcoals which produced very dark. oils in decolorization I did not produce such dark oils at this lower temperature. It seemed as though the power of the charcoals to "put in red" and "to take out yellow" were both decreased at the lower tempera~ ture. (See Table, O11 decolorization IV.) O41 decolorization III (Charcoal 2f, 5 gms., 150 cc. oils) In order to study the effect of agitation and also of time upon the decolorizing action of a charcoal upon oil-a the following experiment was conducted. 5 gms. of chercoal 2f and 150 cc. olla were placed in 4 wide-mouthed, short necked, 250 cc, round bottom flask, fisted with a thermometer and high speed stirrer, The mixture in the flask was then stirred briskly and kept at about 90 degrees by a bunsen burner. During the first half hour, however, the temperature went up once to 103 degrees for a short time. 10 ce. samples of the mixture were pipetted out about every hour, and filtered. During the test, however, the mixture was allowed to stand twice overnight at room tempereture,. This test sheiconsiderable light upon the formation of the red color, but did not explain it. It was - found that during the first two hours of agitation the oil became darker and darker, taking on a reddish tint as it did so. Then on standing overnight the oil became still darker. Agitating and heating, however, for two hours more after that caused the red color to drop enormously, until the oil had lost its reddish tint completely and was only about 2/3 as yellow as the original oil-a. Further agitation caused the oil to lighten still further, but with the taking on of a greyish tint and the exhibition of the Tyndall effect. On standing overnight at this period, the oil became slightly lighter, instead of darker as durine the early stage of the run. The same samples which had been taken from this run were tested again as to color after standing clear in their res pective test tubes for two days. It was then found that the samples which had previously been much darker than the original oil were now considerably lighter than the original, and the decrease from their darkest color had been enormous-in the most extreme case from about 196 cu/ce to about 59 cu/cc, The lighter samples had also decreased in color, but not to so great an extent. The same samples were tested again after about 10 days total standing. In ell cases there was @ decrease in color from the previous reading, except wherle the two readings were practically the same, in which case the difference between the two was greater than the limit of accuracy of the color determination. The amount of the cclor decrease was, howeveg, very small compared to that which took place during the first tvo days of standing. (See Table and curves, oil decoloriza- tion III.) aS dow O11 decolorization V. (Charcoal 2g; 4 gus., 120 co, oil-a) Since there was such a great increase in color during the first part cf oil decolorization III, it was thought that this darkening might have been due to the accidental increase of the temperature uv to 103 degrees during this period in the ease of oil decolorization III. Another run was made, therefore, and at the lower temperature of 80 degrees, using a water bath to increase the stability. Standing overnight was also avoided during the early part of the run. As will be seen from the curves, shown later, even at this lower teupereture there was an increase in color (with reddish tint) during the first stage of the run, but this color began to fall off before the first five hours of stirring had been completed. Thereafter, on standing within the flask in contact with the carbon, the oil became Lighter, thus exhibiting the same effect inside the flask in contact with the carbon as had been shown by the samples in decolorization III on standing after being filtered. Also, the samples taken during the earliest part of the run became darker after standing two days, and then lighter again after standing four days. whereas the semples teken during the latter part of the run became lighter and lighter. The final oils (after 4 days), however, up to and including the sample teken after four hours’ stirring, were all darker than the or- iginal , whereas stirring for two hours more caused @ very great decrease in color, the final sample becoming ultimately only about 18% as deeply colored as the original oll-a. These latter samples, however, as in the case of oil decolorization III, pos- aTessed a slightly grey color and Tyndall effect. Prd Oil decolorization VI. (Charcoal oo ~- 3 ” 017 ems. ), 800 cc. of 4,0 gms, a Oil decclorizeétion V showed that the appearance of the red cclor in the oil was not due to heating over 100 degrees, and that the oil heated only to 80 degrees also took it on. Therefore it was decided to carry on the runs thereafter in a. bath of boiling water, to insure uniformity of temperature and also a sufficiently high temperature to get fairly rapid decolorization, In order to study the action of another charcoal under ‘gimilar conditions, a sample of 2d (with a small amount of 20, to make up 4 gms.), and 200 cc. oil-a were treated in a manner similar to that of o11 decolorization V. The curve of the color of the samples taken immediately after removing the sample was similar to that obtained in oi] decolorization V, except that the oils did not become nearly so dark. Stirring was also carried out longer than in oil decolorization V, so the oils became lighter than that of the last sample of de- colorization V, observed immediately after withdrawal from the flask, (Readings on standing have not yet been taken.) O11 decolorization VII. This decolorization was carried cut in @ manner sinm- Jlar to thet used in the others, except that a 200-cc, Erlen~ meyer flask was used, fitted with a large stirrer, capable of moving almost the whole mass of oil and charcoal within the flask, This stirrer was then operated very slowly (one revolution in about 3 seconds), which was just enough to keep the mass moving and most of the charcoal off of the botton, or moving over the bottom. Using 3 gms. of charcoal 2f, with 800 cc, cil-a, and run- ning with the above stirrer at about 95 degrees in a water bath gave samples about 50% as dark as oil-a after 5 hours stirring, but the color did not decrease very much more after 13 hours stirring, the last sample being about 30% as dark as oil-<a. The grey color and Tyndall effect, however, were absent, so far as could te observed. This would indicate that using a better charcoal under these conditions micht decolorize the ofl to a greater extent end still avoid the grey color and Tyndall effect. Comparison of Sugar and Oil Decoloizations. The charcoals from 2c to 2s can ve listed in the order of their sugar decolorizing power according to sugar decoloriza- tion I. If then the results of o11 decolorization II and IV be listed opposite the corresponding carbons , it will be seen that except in the case of the best and the poorest carbons, there is no relationship between the decolorizing power of a carbon with respect to sugar and that of the same carbon with respect to oil. However, if the results of the oil decoloriza— tions are grouped according to conditions of activation of the charcoals involved, as was done in the case of the sugar de~ colorizationg, it is found that in general for any given tem- erature. as the time of activation increases, the decolorizing power also increases, as in the case of sugar decolorization, but the difference between high temperature and low temperature ars po activation does not appear to be as great as in the case of the sugar solutions, * It was intended to carry on this research from this point as follows: Prepare a standard sclution, either of sugar or non-sugar materials, which will be olear, of the same color as partly decolorized sugar solutions, and reproducible. Work out a standard method of decolorizing sugar sclutions, which will not be subject to criticel factors difficult to con- trol, as has apparently been the case with some of the sugar decolorizations so far carried on. Inerease the rate of stirring of the oil decclorizations to see if this can be done up to a certain point without producing the grey color and Tyndall effect. Use the rate so decided upon as a standard for other decolorizations. Adopt a uniform procedure of sampling and cclor determination with respect to time, so that all results can be compared to one another. Carry out a series of activations similar to the first series, and use the charcoals produced from these in standard tests on sugar solution and oll-a. *These values are placed on the chart of sugar decolorizations, grouped according to conditions of activation. PART IT Following the proposals made at the end of Part I, a larger furnace was first wade for the purpose of producing lerger samples. Description of Furmace A silica tube, 36" long by about 3" inside diameter was placed vertically within an iron flue which in turn was heat insulated. Gas was fed in at the bottom of the blue, and both air and gas fed in separately at other points higher up. <A charge of 110 ems. of charcoal was held within the tube by & perforated porcelain disk supported on a nichrome tripod which rested on a cork containing a glass outlet. tube. This charge filled the silica tube to a depth of about ZO". Steam was fed in from the top, thereby becoming preheated by passing down the unfilled portion of the silica tube. Three thermocouples ("Advance" and chromel) were placed within one pyrex tube, so as to register temperature at the bottom, middle, and top of the charge, This tube was kept as near to the center as practicable. Another similar thermo- couple placed near the ede of the charge, and at the same height as the central thermocouple, indicated that temperatures near the edge and at the center of the cherge were practically the same, for which reason the outer thermocouple was not used after the first few runs. The chisf difficulty involved in the operations of this furnace was. the control of temperature, the latter being very difficult to maintain both constant and uniform. In fact this was not found possible to accomplish, so the best that could be done was to keep a continuous record of the temperatures at the three thermocouples, and to determine the average of these continually changing temperatures, calling this latter the temperature of the run. This was not theoretically sound practice. <A momentary high temperature would net be offset by a Similar period of low temperature running, since the former action is very much more rapid. However, it was assumed for this purpose that the variation in temperature was sufficiently small to allow the average temperature to be used as represent— ing the temperature of the run, Another disadvantage of the gas fired furnace used, be- Side that of temperature control during the run was the var- lable length of time necessary to get the furnace and contents to the temperature desired for passing steam, There was always some distillation of residual hydrocarbons during this period, as shown by the efflux of a luminous burning gas at the furnace outlet, and where the period of pre-heating exceeded the time hecessary for this distilletion, tthich was usually the case (varying from $ tol hr.) there was doubtless some activation due to atror water vapor present before the steam was regularly passed, Excess steam was always used, but some variation in its amount between runs may have influenced the results, a0 The sawdust was carbonized in the same manner as previously. Three different lots were made durinz this series of experi- ments, each being a mixture of several individual batches. Since each was made in the same manner from the same material, and consisted in a mixture of several batches, it was thought that the average mixtures so obtained would be practically the same as though the entire amount used hed been made up at first and mixed. It was opserved, however, that the sawdust in the bottom of the sack, used in the last lot (no. 5) was somewhat lumpy, due to its having absorbed rain water. The lumps also had a slightly darker appearance than fresh Sawdust indicating that the water may have been somewhat muddy. This was used just as it was in preference to fresh sawdust, partly out of curiosity to see if such poor material would yicld good char, although it was thought that these new factors would not affect the result apereciably. The charcoal so produced was found to be considerably less dense than that formerly produced, as shown by the fact that the beaker formerly used to hold the weighed sample would not hold the weight needed (occupied about 20% more volume). The activated chars from this charcoal were also lighter, blacker, and more active than the others. Unfortunately, since both face tors were changed at once, it is not definitely known whether these latter facts were due to the character of the sawdust or to the different conditions of activation, but since the in- erease in activation was in the direction which is indicated ~ OL by the conditions used in forming the less active charcoals, and since the differences among the various chars of this same carbonization are in uniform relationship to the conditions of activation, it may be assumed that the latter factors were of at least greater influence than differences in the sawdust used, A sample activation is here recorded to show how the temper- aturesvaried and how the temperature of the run was calculated , Necolorization Tests upon Standard Solution of Blackstrap Molasses, It was decided to use a 3% solution of blackstrap molasses with H ton concentration of 0.007 N, as usad by Coates* in the evaluation of decolorizing carbons obtained from bagosse. It was found, however, that if the solution were brought down to this acidity by the addition of NaOH, after the solution head been filtered with kieselgubr that it would again become cloudy, and unfit for use as @ standard. Addition of acid would redissolve the doudiness. It was also found that the color of the solution itself was effected by the acid concentration, »e~ coming lighter with more acid. Finally the standard 3% solu- tion was prepared as follows: 240 ems. blackstrap molasses were dissolved in water, and & solution containing 3 gms. NaOH was added, this being known from a previous trial to be nearly the amount of Na0H reyuired to neutralize the acid in the molasses used. The mixture was then diluted to 6000 cc,, 120 gms. kieselguhr: added, boiled, and filtered. Acidity was then tested by titration with standard NaOH and phosphoric acid solutions, using phenolphthalein *Jour. Ind. Eng. Chem., 14, 395 (1982). as an indicator, Considerable practice was required to be able to recognize the end-point on account of the reddish color. of the solution itself, and the gradual change in color of the solution, independent of the phenolphthalein, which took place on adding acid. However, a definite cherry red end point was discovered which gave consistent results . By this method the solution was first made neutral, and finally the calculated amount cf water and acid added to produce 8 1i. of 0,007 acid N. solution. Subsequent checks showed that the acid concentration actually lay between 0.9075 and 0.0064 N., this being about the limit of accuracy obtainable by the method used, The color standard consisted in one part of the above solution and one part water, but in calculation of color depths the value of this solution was set at 50 "color units" per oc, The standard method of decolorizing adopted was the same as that used by Coates (loc. cit.), except that smaller portions were used, The method was as follows. 3 ems. charcoal and 100 cc. molasses sclution were placed in a 200 cc, Erlemmeyer flask, heated to boiling, stoppered, and Left in a bolling water bath for 10 minutes with occasion- al shaking. It was attempted to secure curves under approximately the following conditions, but the actual conditions as cal- culated from the record, as above indicated, were taken rather than the conditions arrived at. degrees C passing steam. degrees C passine steam. 600 20 725 10 30 20 40 30 50 40 700 20 30 39 750 LO 40 BY 50 39 80 40 Tests wore first run on semples of charcoal in the same condition es obtained from the activation, Following these, tests were made which showed that powdering considerably in- ereasecd the decolorization as obtained by the standard method. There was, however, difficulty experienced in filtering the povdsred samples, and it is probable thet some of the dif. ferences between duplicate runs may be traceable to this. The effect of time of hesting upon the decoloriazing action of powdered charcoal was also tested with the result that there was apparently no decrease in color of solution after the first 10 minutes, but rather an increase if anything. It is thought, however, that these variations. since irregular, were due tc dissimilsrity in the efficiency of filtration of the different samples. = Gd Enaréeal Swi Min. Sample Stendard Color units heating reading reading per oc, 1 10 ~—s-38.8 7.6 10.3 2 15 28.8 7.6 13.3 3 20 29.3 7.6 13.9 4 35 25.9 7.6 14.7 5 30 24.9 7.6 15.3 Similar tests carried out with unpowdered samples, however, (which were easily filtered) showed no definite increase in decolorization with the time of heating. Awd, 1 10 22.8 7.6 16.7 (gray color) 2 15 48,9 6.1 6.5 3 20 50.2 6.8 6.8 4 325 41.8 6.8 8,1 5 30 51.2 5.2 5.1 Since, therefore, neither of these tests indicated sub stantial decolorization after the first ten minutes, the letter time was taken es standard, as sugcested by Coates (loc. oit.). The results of testing the decolorigzing power of all the samples, both in powdered and unpowdered form are given in the following table. Powd | mol- aeses decol. ou/ee. —“~ Unpowd, = Oe Aver. Time % Temp. min. yield July 9 July 7 Powd, Unpowd, Sf 6704 = = 80 55.2 44.8 62.3 21.6 54,1 3g 718 41 17.6 30.1 Sh 86-55 48 59,0 82.6 74,5 3i B70 51 70,5 99.0 69.3 3355S 52 66.6 104.1 69.3 3k 4672 45 66.7 31.1 20.0 ; SL «GA? 50 54.3 42.4 51.3 85.8 Sn | «B77 50 54.3 47.7 52.8 50.9 3n 690 30 58.0 58.7 63.4 300 G78 63 47.6 28.0 36.3 Sp 686 4 47 6 35.3 53.0 55.6 42, 661 33 63.8 73.8 70.0 , 4b 607 50 65.7 61.5 60.8 63.8 4o 600 40 64.7 64, 78.5 64.1 4d 547 30 68.6 94.4 73.4 4e 583 20 73.3 97.5 76.6 4f 935 40 44,8 16.3 5a =—s-749 30 47.6 14,1 42.0 Sb 743 20 57.1 37.9 54.0 13.8 58.7 Se 770 11 66.6 87.5 81.3 5d 710 51 43,9 11.0 8.0 Se 700 30 49,5 12.8 39.7 Sf 710 40 57.1 32.3 55.5 Se 692 30 60.0 50.2 65.0 Untreated charcoal 80.9 ~~ 3Gu Physical Characteristics of the Charcogls. When the samples used in the foregoing tests were powdered, it was noticed that they ranged from greyish to very black in appearance. Also, the vlackest samples were softest (the quality of flour), and had the lower apparent density. Out of curicsity the semples were classified acccerding te blackness, with the following result: Very black Medium Grey 4g Se, 1, m,n, 3f,h, i j,k ©, P 4a, b, c, d, @ 5d, c, ? Sa, b, Bg Se It is interesting to note thet the blackest samples exhibited the greatest decolerizing power, the grey the lesst, and the medium, in general, between these two, Discussion of Results of Sugar and Molasses NDeoclorization Experiments. The most important results obtained are indicated by the contours drawn in plates I and II. In the first place they show that the electric furnace produced better decolor~ izing charcoals for given apparent conditions than did the gas fired furnace. However, both sets of contours indicate that the region of greatest activation is in the upper right hand part of the diagram, or simply that region in which both the temperature and time of activation is high. This is what might have been expected from the suprosed nature of activa- tion, which consists at least in part of increasing the micropores by oxidation. A less obvious probability, however, is shown by the fact that above 775 degrees the contours are practically vertical, cr even sloping upwards to the right, which means that at these higher temperatures even though the rate of oxidation is higher the increase in activation per unit of time of passing steam is no greater, and seems to be even less than at or below 775 degrees. The red contour of plate IT illustrates this probability even more forcibly, showing that at 900 degrees even though most of the charcoal be oxidized (84% loss) the product is not as good as thet produced around 750 degrees with losses of less than 50%. Both plates taken together also indicate that the most rapid increase in activation per unit of time of passing steam or per unit of loss in weight take place at about 750 devrees. The gas fired samples of plate II also showsup toa loss of about 35%, the temperature at which the loss took place makes very little difference upon the decolorizing power of the charcoal, tut for higher losses, the higher the temperature, up to about 750 degrees, the better. At hich temperatures the ef’ect produced hy a given loss in weight becomes less and less (now shown by the electric furnace data), until, as previously stated, at 700 degrees @ very Large loss produces no better charcoal than a simall loss. The final cenelusion to be derived from all these data taken together is that the best charcoals for either sugar or molasses decolcrization will be produced by activation at about 750 degrees, and that at this temperature the best charcoals may be produced by oxidizing to a loss of about 50%. The best individual sample of all, 23, produced in the region of 750 degrees and 50% loss in the electrics furnace, and the low contour of the gas fired furnace which centered about this same region, supports this probability, although it is possible that slichtly better charcoals might be produced by continuing the oxidation at this temperature to still higher losses, Little Light is thrown upon the problem by plates III and IV, except that they show the effect of unpowdered samples to be similar to that of powdered samples, although they are on the whole either less active or slowez in their action. Oil Decolorization Experiments In the latter part of Part I it was shown that there were a number of peculiar phenomena exhibited in oil decolorization by charcoal, such as: 1. The charcoals "put in" a red coler during the first few hours of stirring with the oil, and later took this out again together with some cf the natural yellow color, but added erey after many hours when red and yellow were almost completely removed. 8. On allowing the mixture to stand without stirring, it would become darker (in red) if the standing took place near the vecinning cf the run, (See plot of of] decoloriza~ tion III), o lighter if the stending took place near the end (III and IV). All the samples, however, on standing after filtration for about two days, became much Lighter, and the very darkest samples became nearly as licht as the others. Further standing seemed only tc increase the same change, in smaller proportion. The two new experiments, results of which are plotted in Plates V and VI, were conducted for the purpese of further studying the phenomena of charcoal decolorization of cils in order that a standard method might be worked out for testing the decolorizing effect of different charcoals in oll. In these experiments 400 ca. of prime summer yellow fe- fined cotton seed oil were mixed with six srams of activated “chareoel in a S@co, Erlenmeyer flask. The latter was kept in @é belling water bath. <A stirrer was placed in the mixture and retated by a motor. The two plates given represent two new experiments, the first that of a mixture of charcoel and oil very highly agitated by & small propellor rotating at hich speed. 10 co, samples were pipetted out of the mixture at fairly regular intervals, and filtered as soon as taken. The apparatus was shut off at night, and sometimes left inactive for as long as two days. The abscissse of the graphs, however, represent actual stirring time, while the length of time of the rest we el ee pericd preceding the taking of each sample is indicated at the point representing that semple's color reading. After all the samples had been taken and the stirring part of the exporiment completed, they were each compared to either the original oil, or a bleached sample as a standard, calling the original o11 100 cclor units per ce. The ratio between the color of the original oil and the standard bleached oil was found in the first place, so that the final result was theoret. ically the same as though all the samples had been compared to the original. The method used, however, had practical ad- vantages, Considerable difficulty was often met with in come paring colors, especially due to grayness which appeared in the sample taken towards the end of the stirring period, These difficulties were reduced, however, by practice in preliminary experiments and differential illumination to such an extent thet readings checked satisfactorily. The curve of Plate V labelled "Readings made June 30th" shows how in this case the oil became lighter and lighter with stirring, until a certain minimum point was reached, after whieh it became darker again, although the latter darkness was in gray. The long rest (43 hours, 55 min.) between the taking of samples 3 and 4, with the noticeable increase in color instead of decrease, shows that between these samples the darkening due to standing more than offset the lightening which takes place on stirring. If the "standing" factor could have been made constant or practically eliminated by running ~4). the experiment continuously from beginning to end, the points would probably have fallen on a smoothsr curve, When the whole series of filtered samples wes allowed to stand eight days, they all became lighter, with middle drawn samples showing the least lightening and the letter samples the greatest. The most striking phenomena of all, however, was shown when Color readittes were again taken after five days more had @Clapsed. The first two and last two samples had become still lighter, but those in between had started to darken again (Plate V). It thus appears that a standard method of testing the de- Colorizing power of charcoals in oils, if the method involves heating and stirring, would have to include not only a prescrib= ed temperature, rate of stirring and time of stirring, but also a standard length of time of standing after filtration of samples, This standard time would have to be at least more than two weeks, and as much lenger as is necessary to bring the oil to @ constant color, The only difference observed between the results obtained in the letter of these two experiments (platted on Plate VI), and those of the former experiment was that in the second case the latter samples were not darkened as much as in the corres= ponding samples of the first experiment. The curve of readings taken after a two weeks' standing was in this case similar to that after a weeks' standing in the former experiment, but no more readings were ta ken after that. None of the samples, however, showed as much lightening in color with the slow stirring &s with the rapid, for the amount of time allowed. = 48 The conclusion of these experiments on oil decolorization indicate that at the present writing it ssems as though the only way to form an adequate estimate of the decclorization value of a charcoal on an oil would be to conduct 4 complete experiment Similar to the ones described, plot the results, and evaluate the chercosl in accordance with the appearance of the entire plot. Explenation of Plates Referring to Sugar and Molasses Decolorizations. Plate I. Red points and contours refer to decolorization of sugar solutions by activated charcoals produced by electric furnace with smsll capacity, while black points and contours refer to decolorization of molasses soluticns by charcoals activated in the larger gas fired furnaces, The curves represent contours of equal activation, but the tw sets do not fit each other, since the activation apparatus , solution decolorized, and standard practice of dsecolorization differed in the two cases. To the left of each point is written the number and Letter which identify the corresponding charcoal sample, and to the right of each point is indicated the depth of color in the colorized solution. The contours were drawn as nearly @65 pos~ sible to represent the regions of activation conditions which produce approximately equel results in decolorizing. This method of lebelling points applied to all of the first four plates. Plate II. This represents the same experiments as Plate I, and is in every respect the same as Plate I except thet the abscissa: is in terms of percent loss in weight of charcoal during activation instead of time of passing steam through the charcoal. 4 Ae Plate III. This is a plot of the results obtained by decolorizing solutions with unpowdered samples of charcoal, in-= stead of powdered, as in the case of the data plotted from the gas furnace charcoals shown in Plates I and II. Plate IV. This bears the same relation to Plate IIT as Plate II did to Plate I, in that the abscissa is in percent loss in weight of charcoal instead of in time of activation. Oven temperature 180 degrees. lL em. Carbon, 25 cc. Oil Sample h (») j (bd) (c) ula) v(a) e(b) (c) (a) (e) u(b) v (b) Table, Oil Decolorization IT Time in oven (hrs.) 5-1/3 " ' " % 2-1/3 3a1/3 5-1/3 " 67.0 31.7 36.9 44,3 43.3 34.7 39.0 54.4 33.0 25.8 30.1 a4 5— Color units per cc. = 4 Table, Oil Decolorization III 5 gms.. Charcoal, 150 cc, Oil, Rapid agitation. Sample, af. (Time of stirring and other factors are shown in the plot, but are not given here. ) Color units per ce, Sample Readings taken after 3 days after 190 days immediately _ 1 133 73 88 2 142 65 48 3 167 59 58 4 40 36 22 5 25 24 21 6 28 29 1? 4 18 18 15 8 18 17 1? Q 19 L? = 47 = Table, Oil Decolorization IV Oven temperature, 68 degrees. ® gms. Carbon, 85 cc, oil. Sample Color units per ce, Sample Color units per oc, h 116 n 45.7 a 58.3 ts) 63.5 e 91.6 p 21.0 a 110 r 35.6 k 75.6 s 36.0 1 49.7 u(a) 109.4 m 24.3 v (a) 94.6 u(%) 118 v(b) 56.8 Table, Oil Decolorization V Same. proportions as Decolorization ITI. Same speed of agitation. Temperature 80 dsgrees. Sample Readings taken Color units per cc, immediately (aie 2 days Aver. after 4 days 1 103.3 132 155 142 110 F 126 143 128 135 121.4 3 170 128 135 131 121.0 4 213 128 126 8 137 126.0 on) 173 79.5 el 80 59.9 6 90 47.1 7 21.4 8 16.0 Plot of O11 Decolorization HT avo 3 190 Q & AY We ‘S K 3 aw 170 5% x Sy w* 8 YQ RON 100 & \ Ey S¥ Ife Ky t 140 nh \ 30 zo ho «ao } Color read wninechately aller rernoving . do ‘ ¥ sample from ywuxture, After 2 days. vA y. 8a vo After. 1? WNys é 60 8 S g NX xy OF 40 9 8 s 8 nw ae a7 8 ao oN, e 2 3 4 Ss oe Hes. Heating arid Stirring, FO°C ae 200 (Ge 180 0 {eo iso Be wo vo to oe 3s ao 30 ae to 2 aay s Plot of Oi Decolorization X + -irst readings ty Stood 2 days i Mask with Carbar at Room Jeng. oe Second readings Standings JA. teadng, 4 days —-- Manang 7 2 > i = Hours Heating ana Stirring 80°C. 900 850 925 890 oe of Activation, C. ky FS Ry 8 re lémpetatiuré iad zy 625 600 / Havel ee _ a | : = a3? | Man pers 0. circles. nteale — | ! aap f : | : color ants per Le. MZ | | ) Reg the j | MACE TH te. aid. 7 YE) WIR? | j ! | ayy 7? x YUWIAS roan Whe? He. va snes We? IIA : nore ae : a feu contours, fESALTO [FA17 : : i Gell Pie PIM VI ALE. LGM DADE DAM, SE BN IDE GES ff. tb 6 MAL c. fo 2h 30280 oo Loe Spe : : os ~ | : : : Lah BR : ! S —. 7 Teme | ne (o} 20 10.2 1 i j | | ‘ i ‘ | I i i i i Te — (SINE. p 024 Plates, i i | } { } | i | | | | \ j H iH ' : : \ : i 4 t i | : i i | | | | | —_ i t : Hl } " | wo | @ a noovwaered Janyoles , Plate LIZ @ 50 23 Pf 62.3 Joo 7 700 363 a “Be 700 ol 5d 8 iN y 25 iy : N 9b 9 602 Naty 72.5 N They 4S ny oy ©6973 | Feo 76-6 hy Jt roy 69.3 no Curves HE COMOUIS 4! 073A Of apron, equal Aetof- OV IZING wer, (olor wnrrs per cL, decolorized 0lfr.) as 00 r . _ , TS ; : r 1 5p bo Jo 8 2D 3e 4o line & Athivation, ye, LYOOWAEEA JaYyfyHes Plate Le 5 By als @ ro da, 42.0 Vo as 2 Sly 28.0 700 5g IAT Py 530 WW 30 6363 FAEFAV S 4a Ss Se N S 4602 Neco Moi2.5 Curves are Confeurs Wegihs SF agyrox, egual aecolor- v 3/a69.3 1211 Power, 40% (Olar wits PPlr Ce. Ae “Ob Aecolo7zta stl lire) Bl @ 693 4ad yi3A 2a 20 ag 3e 3s 5s fer Cort 1056 v7) herght fPANIG GAPIOLJ = a of ° i $e 4° OSSD A dtfua (LE) (GOPMOY) YL IA] 1p) UAMOYS O50 Sf PHIUXILM Ua UBYL SOM YMMOE YY Lo. Hag tA) hig paporpl! sydiuos lisanr{ad G004aa {$34 (CUOISO7IGQ) A 2ILOlS 9,00 Yfeg AYHOM ty patil{s Zilog “ BSL) ‘Yaitios [10 eIAAY© fypidoy oo SOMLOPA AYO). 4 xr re buriwzpe peryoo nf O€ az or 9 — LA PAPAS 9d 9 fey cr rer Pers SOUP ORT 2 © e Ok euMnsy Wayos, 62 pesy * GS Of YMUCE [yo fOBLIE AMY 2)