A Study of the Nature of the Adsorptive Property of Fuller's Earth Citation Stamm, Alfred Joaquin (1921) A Study of the Nature of the Adsorptive Property of Fuller's Earth. Bachelor's thesis, California Institute of Technology. doi:10.7907/azp4-y170. https://resolver.caltech.edu/CaltechTHESIS:07292024-185226060 Abstract No abstract. Item Type: Thesis (Bachelor's thesis) Subject Keywords: (Chemistry) Degree Grantor: California Institute of Technology Division: Chemistry and Chemical Engineering Major Option: Chemistry Thesis Availability: Public (worldwide access) Research Advisor(s): Unknown, Unknown Thesis Committee: Unknown, Unknown Defense Date: 1 January 1921 Record Number: CaltechTHESIS:07292024-185226060 Persistent URL: https://resolver.caltech.edu/CaltechTHESIS:07292024-185226060 DOI: 10.7907/azp4-y170 Default Usage Policy: No commercial reproduction, distribution, display or performance rights in this work are provided. ID Code: 16599 Collection: CaltechTHESIS Deposited By: Benjamin Perez Deposited On: 30 Jul 2024 16:03 Last Modified: 30 Jul 2024 16:03 Thesis Files PDF - Final Version See Usage Policy. 46MB Repository Staff Only: item control page

THESIS BY Alfred J. Stamm A STUDY OF THE NATURE OF THE ADSORPTIVE PROPERTY OF FULLER'S EARTH. Department of Chemistry California Institute of Technology Class 1921. A STUDY OF THE NATUR~ OF THE ADSORPTIVJ-~ < PROPERTY OF F'ULLER I S l:i:ARTH by Alfred J. Stamm Introductio)l• There is a large industrial demand for an inexpensive a.dsorpti ve materia l for clarifying various 1iq:.1ids such as lubricating oils e.nd numerous aqueous solutions. This demand has been met chiefly by charcoal and two classes of earths , namely, fuller's ea rth and diatomaceous earth. Fuller's ea rth is a clay or kaolin of an indefinite composition. It consists largely of a hydrated a,luminum silicate (Al 2 0 3 •2Si0 2 "2H 2 0) with small amounts of' calcium ·oxide, magnesium oxide, titanium oxide, etc. Diatomaceous earth, also known e.s kieselguhr, is on the other hand an almost pure silica with traces of aluminum oxide s.nd others of the B,bove impurities . Its formation results from the deposition of the remains of minute microorganisms. surface. It consequently has r;. very large internal Large deposits of both of these materials are fou.rid in ve.rious parts of the globe. Very little is kno vm as to the nature of the adsorptive power of these materials, and consequently no def'ini te procedure as to use, treat,ment, or revivification is available that has as a basis real scientific facts. The adsorptive power, however, has been shown to be practically independent of the composition of the adsorbent. It has been thought 1. b~ ~~~-~~::° be connected with the "acidity" of. the adsorbent. 1. C. L. Parsons, "Fuller's Earth." Bull. 71, Bureau of Mines. - 2 - w. D. Bancroft, however, attributes the apparent acidity to selective adsorption of bases, and asserts that actual acidity is not an esAential to adsorption. The following simple empirical relation between the amount of substance adsorbed (x) from solution, the equilibrillrrt concentration ( c) of the solution, and the weight of tha adsorbent (w) has been found to hold for relatively large concentration changes for many types of adsorption. x/w i::::e:. = 1 ii d.. o and 1/n a.re constants 1vhich depend upon the nature of the sol vent, adsorbent, and adsorbed material. Adherence to this equation has been tested in a few instances for adsorption by fuller I s earths eto., but no extensive da.ta are availa.ble on the subject. The results of Pelet and Grande, and of Atherton Sidell, who made fJ. few experiments with these earths, will be considered later. Thi& investigation was undertaken in order to determine whether the a,dsorption of dyes f.rom aqueous solution by means of fuller's earth is in accord'3,nca with the above adsorption equation, and if so to determine the constantsd- and 1/n. By obtaining these constants for various ful- ler' s earths, dia tomaceous earths, silicas, etc., under the se.me conditions and ;r:i th the s1:1,me solute and sol vent, a basis of comparison of the earths is afforded, and some light shed upon the nature of the active adsorbing constituent. 3 Adsorbent~ Employed. Varioue adsorbents were employed in this investigation, namely, fuller's earths, diatomaceous earths, artificial sodium aluminum silica.te, silicic !:lCid, charcoal, sand, and alumina. The fuller's earths 1-v-ere from three different sources. One of these was obtained from Midway Florida. in three different meshes, 30-60 mesh, 60-100, and 100 up. Chemics,l Co. Another of 100 mesh was obtained from the C. w. Hill Its source is unknown. obtained e"rom the Standard Oil Co. Two samples of fuller's earth were One s~mple was a natural earth, and the other the earth after it had been subjected to an acid treatment. This tre 1.tment inc,reases to a very great degree the power of this earth 1 to clarify lubrfo1;,ting oils f'or which purpose it is used very ext~nsi vely. These h 0 samples are referred to as the untreFited and trerited Standard 1 Oil Earths. The treated sample does not cause water to become acid to lit,... mus, but a one gram sample of it is equivalent to about 5.6 cc. of .1 N acid v1hen the titration is performed in the presence of the earth and phenolphthalein used as indicator. The following table gives the analysis of a typical batch of the treB,ted Standard Oil Co. earth which they furnished with the sample. Celite, a, diatomaceous earth, wa.13 also used. California, and is used extensively in filtering. and porous. Its analysis is given below. It is mined at Lompoc, It is extremely light -4- Celi~_. Standard Q_il Earth As rec'd. Dry basis (treated) Silica Si0 2 80.05% 89.24% 66.3% 73.9% Alumina Al 2 03 1.34 1.49 21.4 20.4 Iron Oxide FeO :L.54 1.77 1.9 0.7 THB.nium Oxide Ti0 2 0.80 0.88 Lime CaO 0.03 0.03 0.5 0.7 Magnes ia MgO o.oo o.oo 3.6 3.1 4.78 5.33 6.3 8.0 Los f_: on ignition H2 0 Moisture 10.28 l.13 U11determined Total tre.ca --~ 1100.00 1.26 _ ..,._. ,. . .... --- 100.00 100.00 c....J .... -- - 105 .80 The silicic acids employed were from two different sources. One was Merck's pure silicic acid, and the other was prepared in this laboratory ·by adding a slight excess of HCl to a sodium silicate solution. was washed free from acid, and dried at 100-105 degrees c. The gel All of these were ground to 100 mesh. An earth ,,as artificially prepared by adding a concentrated solution of sodium silicate to a concentrated solution of .sodium alumina.te. The precipitate vms filtered and washed until it gave only a faint pink color with phenolphthalein. This slight alkalinity, which could not be removed, was likely due to the hydrolysis of the sodium ~.luminum silio&t-a. ea,rth was not analyzed, but a,ccording to A. s. Cushman 1. 1 This the composition "The Colloid Theory of Plasticity, 11 · Trans. Am. Ceramic Soc. vol. 6. 5 of the earth prepared in this way is Silica. 46.1% Alumina 26.1 Soda. 15.8 Water 12.0 The sand employed wa.s pure sea sand. mesh. It was ground to pass 100 Part became so fine on grinding that it remained 8-UBpended when placed in water, and would not settle out completely even on centrifuging. The material was hence separated by sedimenta,tion a,nd only the heiwier part used for experiments. Dyes Employed. For most of the experiments methylene blue /(f:~)i,,,Y-~-5=0-IY~#J) }i 1 2 ( V-1V= ; ' was employed. It is a deep blue dye of basic character, and a crystalloid whioh forms true solutions. There are a number of reasons for this choice, first the fact that it is appreciably soluble in water, second that it is unaffected by acids and is precipitated only with concentrated alkalies, third that basic dyes are strongly adsorbed, fourth that it is a crystalloid and thus avoids all complications of colloidal B.ction, and fifth that the concentration of methylene blue solutions can be determined very simpl}' by colorimetric means. Auramine f #:J)JY-o_jJ _ o - / Y/C#:J), /IC) dye, was also employed. It another basic ff a yellow dye, fairly soluble in water. The ' concentration of its solutions c~n be readily determined colorimetrioally and in general it possesses the same advantages as methylene blue. Erythrosine employed. 6 I (~-a-,,O•) o /" -f ~ c0/Y4 is an acid dye that was It was used to obtain any relations between the a.dsorption of the _b o distinctive types of dye. It has a bright red color, and si.lso lends its elf readily to c olorimetric ana.lysis. It ha.a, however, the disadva.ntage that it is preci pitated by a slight excess of acid. Colorimetric Analy2 is. The simplest and most convenient method of determining the concentration of the dye solutions was found to be by colorimetric comparisons. A simple colorimeter which served the purpose satisfGctorily is shown diagr&.n,. maticaJly in the fo1Jowing figure. - - v1 A 13 A 8 E It consists of a vmoden case B. with two openings above to admit 100 cc. Nestler tubes A, and slig htly smaller openings below over v,hich the tottom of the tubes rest. source of lig ht. This is supported on a. box open on one side to the An inclined white porcelain plate C served a.s a reflector. A mirror was tried but it proved to g ive too bright an illumination for the concentration of solutions matched. Eis a frosted glass plate; F a tungsten lamp (40 watts). The entire inside of the apparatus was painted black to eliminate counter reflections. The apparatus was at first used - 7 - v;i thout the artificial light, and set so as to face a l a rge window, just outside of the range of the direct rays of the sun. Due to variations in the intensity and nature of the light, and the fact that ana,lysis could not be made at night or on dark cloudy days, it we,s deemed a.dvisable to employ artificial light. The device used does not give a pure white light but a yellowish light which imparts a greenish color to the methylene blue solutions. It was found by experiment thiil,t this did not interfere with the colorimetric comparisons. On the contrary, it was slightly ad- vantageous as a.n increa,se of depth not only varied the intensity of the color, tut also produced a. slight color chang e from green to blue. A black hood was used over the apparatus when em.ploying artificial light to cut out the light from the room. With erythrosine and aura.mine solu- tions, on the 9ther hand, it was found that natural light gave more reproducible and reliable results than artif"icia.l lig ht. The best concentr,.tion for matching the three dyes employed ranges from .0002 to .0006 grams per liter for a height of column of solution of eight inches (8 in. column= 100 cc.). The standard used for the in- "Vestigations with methylene clue contained .0004145 grams per liter. Thie was just one-tenthousandth of the concentration of the stock solution (4.145 g. per 1.) used for ma.king up solutions for adsorption. The sta.n- dards of auramine and erythrosine contained .0004467 and .0006295 grams per liter respectively, and v,ere just one:.fi vethousandth of the concentration of the stock solution. Solutions to be matched were diluted to -8approximately the concentra.tion of the sta.ndards, a.nd then the height of the solution in the Nestler tubes varied until the depth of color was identical. The heights were me£<sured with a scale 32 di visions equal to 8 inches or 100 cc., on which tenths of a di vision could be read ( a.bout .3 of 2., cc.). The error fro m readings did not exceed on half division or fifteen hundredths percent. The follo,,ip..g results on a series of solutions of unknown concentration shovj the degree cf .accuracy of the colorimetric til..ndysis of methylene blue. The concentrations are given in terms of one of ttie solutions as unity. Artificial Light. Natural Light. Obs. cone. Cal. cone. %error. .620 .574 + 7.4 .830 .735 + 12.9 .500 .424 +15. 2 .530 .506 + .495 .473 + 4.4 .770 .683 + 11. 1 .322 .326 J •2 1.020 1 .00 . 308 . t7~J + 9.4 .e3o .f-80 .502 .5 06 o.a .980 , 856 .638 .604 + 5.3 Obs . cone. Cal. cone. %er-ror, + f'..O + 14 . 5 Av. Av. 4.7 8.t:fo% 6.24% In the case of the comparisons using artificia.l light all errors a.re in the same direction, and an error in the standard was suspected. The same series of solutions was therefore referred to a. new sts.ndard with the results given in the following table. - 9 - Artifici e,l Light. % error. Obs. cone. Cal. cone. . 780 . 735 .496 .506 .745 .683 .947 l .013 6.5 . 574 . 580 1.0 .878 .856 Av. 4.55% From these d::i. ta it m&.y be a ssumed that the analyses ::J, re r eliable to a t leo.st seven percent. Experiments showed that glass ware in which tha solutions to be m.s.tched were placed should not be wiped out with tovrels ~s the slightest amount of lint ads orbs sufficient of the dye from solution to ap9 recia bly affect the concentra.tion. It vrn.s found that an error of from ten to t wenty percent could rea dily be ca,used by a scarcaly visible a.mount oJ' lint in the Nes tler comparison tubee . The following d~t t a give th e results of the colorimetric fl.nf,1,lysis of a series of erythrosine solutions o .f unknown concentra tion. The ss.me s e t of solutions a re f:! hol'm in both columns, bu t were ma tched 4ga inst sta.ndards of diff,,rent concentration6. - 10 Obs. cone. Ca.l. cone. % error Obs. cone. Cal. c0nc. :t .765 l.71 + 3.2% .570 .585 2-6% 1. 290 1 .28 + o.s . 711 .7413 4.7 J. .420 1.39 + 2.1 .784 .812 3 .4 1. 995 1.90 + 1.8 J . 1.J 0 1 .14 2.6 :1 • ! 4!) l.1-15 + 2.7 .613 -e50 5.7 :1..600 1.75 _+_ 8.6 . 892 1.02 -lQdL_ Av. 0 % error. Av. 4.9% 3.2% Procedure. The adsorption mixtures wer e made up in 150 cc. g lass Gtoppered 'bottles by 1\ eighing a de.f.'ini te amount of earth into the bottle, and then pipetting in th.ti a.mount of '1,ater and of stock dye solution neca::irrn.ry to make 100 cc. of sol11tion of' t he desired concentration. The bottles containing these mixtures ,tere placed on a rota ry agit a ting device in such a way that they were turned end over end thirty to for ty times per minute. The time of agi tation vp •ied for different sr:i.mples, and vdll be considered under the hetlding of "Equilibrium. 11 In order to determine th;:.) concentration of the .solutions they we:i:-e decant ed into 100 cc. oil sample bottles, and c entrifug ed in ~. milk tester centrifugal for 15-30 minuteE. Thie time w~s :1.de'.luate in most cases for sepa rating the earth from the dye solution. Experiments sho wed tha t when a s olution was once clear, further centrifuging did not r1ffect the conoentrft. :_.ion. Sc.mpleA nere then pipetted off and diluted f or analysis. few ~i,,s es 8olutions h"'.d to be r ecentrifuged after dilutio-;1. In l!I. - 1l - Eg uil l t>_i:_i u~. The graph on th e following page s hows the effect of ti me of agitation on tb.e amount of dye adsorbed. For this ca,se, and cases of lower initial concentra tion, t wo hours was found to be a deque.te. The time for equilib- rium to be established, how ever, increases with an increas e of initial concentra tion, and e.lso varies vii th the adsorbent. To be assured that equi- librium had been reached in every ca se all the samples were a.gi tated for l0-20 hours, and those of hig ner concentra tion were analyzed, rea.gitated :f'or at least eight hours, and again analyzed until the change in concentration wus negligible. The time of agitation is given in the tables of data for each sample. The revers ibility of a.ds<.,rpi.ion i, t:.s tested in a numbe r of instances, giving further indications tha t equilibrium had been established. This is s hown in Table I, Series 4,, Table VII, Series 1., and Table VIII, Series 1. Points No. 3rev. and 4rev., Te.ble I, Series 4 show less adsorption thl:t.n l'!o. 3 and 4, thus indicating that on dilutine the s olution part of the adsorbed dye returned to the solution. If equilibrium conditionE had not been reached the dye would no~ have reverted to the solution. A simi- l a r indica tion, but les s apparent, is given in Table II, Series 1., in which the a mount of a dsorbent 1vas increased in s amples No. 3 and 4. of dye adsorbed per unit of adsorbent decreased. The amount This shows that part of the dye was given up by the original earth, and taken up by that added. - 13 Temper~t_ur~ Effect. As convenient f acilities were not available for agi tat.ing the mixtures in a thermostat, they were rotated i:tt room temperr-..ture, and hence were subjected to a certain temperature variation. It was found, however, tha.t this temperature change had a negligible effect on the equilibrium. The agitating a.pparatus was placed on an exposed porch on a cool night. The air temperature a t 8 a . m. when the run was, completed was 8.5° C and 15° C respectively on two different occasions. The averl'lge temperature throughout the night may be assumed not to have exceeded this final temperature. The agita.ting apparatus wa.s also placed under a. hood in which a tem_peratur.::, of 36 ° was maintained by means of an electric hot plate. Following is the experiment~.l data . F).,i,ch of the two samples v,a.s subjected to the three different temperatures. No. of Sample. Time of Agitation. Temperature. Equilibrium Cone. 1 14 hr. 8.5° C max. . 657 2 14 " ~3;5° G m&.x. 1.018 1 16 " 15° C mas. .650 2 16 " 15° C max. 1.015 1 24 " 36° C av. .664 2 24 36° C a.v. 1.050 The clevia.+Jion in ac;uilibrium concentration is obviously within th e experimente-1 error and the variations in temperature in these experiments iVere much gre&.ter than would be met under the usue.l experimenthl conditions, and hence no temperature rag ula.tion is necessi;.ry. - 14 - Accur~ of Results. It was s ho wn under the heading "Colorimetric Analysis" that the error for the determination of the concentration of methylene blue was less than seven percent, and still less for eryt!lrosine. The ' error arising from the dilution of the samples to the desired concentration for analysis is entirely neg lig ible in compari s on with tha analytica l error. It may hence he .s: ssumed that the ec: uilibrium concentrations were determined with a.n accuracy of at least seven uercent. The amount e.dsorbed per unit of weight of s.dsorbent (x/m) is given by the following relation: x/m = - - - m All of these quanti tieB can be determined very accura.tely with the exception o:E' the equilibrium qdncentra. tion (C) 1vhich is reliable to about seven percent. In order to make the error in the amount adsorbed quite small, ( C) must be small in comparison with (Co ) the initial cone en tra tion. Preliminary experiments were run with the different earths in order to determine the amounts of ee.rth and the ini tia.l concentra tions necessary for obtaining the gre~test degree of accuracy. With the exception of the ds,ta for sand and alumina (with methyline blue), the maximum error in the amount e,dsorbed is seven tenths percent. This was calculated from point No. 1, Table No. II, Series No. 3 (Artificia.l Earth). An A.Verage error in the ~mount adsorbed for a seven percent error in the equilibrium concentra tion is four hundredths percent. It may hence be assumed that the error-' in the amount &dsorbed is neglig ibl •a in most c A.s es in comoarison with - 15 - the error in the equilibrium concentra tion. This means that the points in the logarithmic curves can be shifted only seven percent in a vertical direction, and not at all horizontallyJ and the resulting error in 1/n should not be greater than ten percent. In the Cf~sa of sea sand the degree of accuracy of the data is less than in the other cases, due to the relatively small concentration changes. For an error of seven percent in the equilibrium Mncentration an error of one and one-half 'percent might result in the amount adsorbed, and an error of ten to twenty-five percent in the value of 1/n. The empirical adsorption equation 1/n x/m =~c.,_ is a logarithmic relation which ma.y be written log x/m = loga::.. + 1/n log c. If the loge.rithms of the amount adsorbed per unit weight of adsorbent are plotted against the logarithms of the equilibrium concentra_ tion, a straight line will result if' the relation holds. The constant l/n is represented by the slope of the line, and log.::X::. by the value of log x/m at the intercept with the logarithm of unit concentration. 1'he value of 1/n given in the following tables was calculated from the slope of the most representative line through the logarithmically plotted points, and a.:::.. calculated for the concentration of one gram per liter. - 16 - TABLE I· M~THYU~NF:: BLUI!: AND NATURAL E:ARTHS. Fuller' s Earth (Midway, Floridtt.) 100 mesh. Series 1. 1/n = .0667 No. Hr. rotated. 17 23 17 23 17 23 17 G3 l• 1' 2 2' 3 3' 4 4' Series 2. (g) (g . per I.) l .900 1.900 2 .003 2 .003 2.004 2 .004 2.203 2.203 .00195 .00778 .00925 .0256 .0451 .0940 .135 .136 4 a 13 8 13 Series 3. 2.018 2.528 2 .002 2.002 2.000 2.000 4 Orig . g . dye pres. . 1243 .:l. 347 .1450 .1554 .:J.658 .1760 .2073 . 2060 Vol. (cc.) 1.00 80 (a) 100 95 (a) 100 95 ( a) 120 115 (b) = .0678 .:X:'. = .113 .00156 .00690 .00870 .0167 .0408 .0688 .0724 .0816 .0825 . 0871 .0911 .0947 .1450 .2070 .1658 .1760 -1867 -1968 100 100 100 95 (a) 110 105 (a) Fuller's Earth (Midway, Florida) 30-60 mesh 1/n 1 2 3 (g. perg .of earth) .0654 .0'706 .0720 .0764 .0807 .0833 .0868 -0866 Fuller's Earth (Midway, Florida.) 60-100 mesh ~3 3' 2 2' = .0980 Wt. e11.rth Equilibrium cone. Adsorbed 1/n 5 4 3 ~ 16 16 16 16 1.706 1.631 1.514 1. 455 = .0663 .000617 .00106 .00324 .00693 oL. = .0994 .0608 .0634 .0682 .0708 . 1036 .1036 .1036 .1036 100 100 100 100 The prime numbered solutions were made by adding 5 cc. more dye solution tot.lie same numbered solutions after removing that re~uired for analysis. (b) Removed 10 cc. for analysis from No. 4 and added 5 cc. of water. - 17 Fuller's Ea.rth (Midway, Flori de.) 100 mesh from 60 Series 4. 1/n = .0635 Hr. rotated No. 2.001 4 4 P.lore 10 4 more 10 2 .001 2.000 2 .000 2.100 2.100 • ()919 .0978 .0988 .1002 .1013 .1658 .1865 :2027 .2070 .2211 .2280 Vol. (cc.) 100 100 105 ( d) 100 105 (d) 100 Fuller.1 '3 Earth (C. W. Hill Chem. Co.) 5 5 5 5 29 Series 6. 1.378 .00062 1-322 1.263 1.020 1.151 .00098 ex::- .00213 .00447 .00895 = .17 4 .1050 .1095 .1150 .1213 .1450 .1 253 .1450 .1450 .1450 .1243 Fuller's Earth (Standi:trd Oil Co. ) Treated. 1/n 3 2 5 4 6 l .0827 .00516 .0257 .0688 .0917 .0976 .155 1/n = .0675 6 8 4 5 7 = .115 Oris . g. dye v:t. earth Equilibrium Cone. Adsorbed pres. (g ) (g. 9er 1.) (g. per g. of earth) 4 SerieB 5. ~ 6 7 1.001 1.001 10 .• 932 1.002 .860 6 15 16 l.007 = .0706 .000704 .00123 .00166 .003l4 .00720 .00850 d-- 21.7% water and volatile. = .344 .2069 .2176 .2220 .2277 .2430 .2480 100 100 100 100 100 .2070 .2177 .2070 .2280 .2070 .f.488 100 100 100 100 100 100 .3362 .4145 .4145 .4145 -4145 .2942 .2733 .2733 .2695 . 2695 100 Maximum Adsorption 94 51 daya(e) 71 days 51 days 71 94 hr.. 116 hr. 152 hr. 94 hr. 130 ls 1 1 2 2 2B 48 48 5s 3s - - -· ---- f-:i, \\.l I (e) I - ., . "--11 - - J..004 1.301 1.301 l.360 1.360 1.000 1.002 1.002 1.000 1.000 .650 .576 .578 -448 -405 . 294 .1215 .0928 .0810 .0387 . 2712 .2740 .2740 . 2720 . 2745 .2648 .2611 .2640 .2614 . 2656 lOO 90 100 90 100 100 90 100 90 ---- Removed 45 cc. e.nd replaced with 50 cc. water. Where the time is given in days the samples were shaken by hand several times a. day, and agitated on the agitator when it was not in use. - 18 - Series 7. No . Hr . rota ted. 3 1 2 4 32 32 32 32 Series 8. Fuller's Earth (Standard Oil Co.) Untre11ted. 100 mesh. 1/n = .0595 .x::- = .477 1:Vt. earth Equilibrium cone. Adsorbed Orig. g. dye Vol. pre;'3. (cc. ) (g) (g. per 1.) (g . per g. of earth) .3042 100 .3400 .000530 .1036 .3300 .000865 .3140 .1036 100 . 3(;33 .1036 .3205 .00160 100 . 3342. .3095 .1036 100 .00294 Diatomaceous Earth (Celite) 1/n = .0654 15 20 15 3 6 1 8 11.903 5.495 8.652 6.394 5 .132 6.696 7.006 20 15 15 40 4 5 2 .0174 .0188 .0191 0194 .0201 .0214 .0229 0 60+ 60+ 62.482 48.772 TABLE II. .0151 .0740 ~ = .000625 .000698 3' 6 more 28 6 more 1 40+ 28 5 3 60+ 28 6 (f) .04055 .04145 100 100 .:l.036 ME:'rHYLENE BLUE AND ARTIFICIAL ADSOR.SENTS. 1/n = ■ 143 4' 4 -1450 .1650 100 100 100 100 100 100 100 .000838 Silicic Acid. (Merck) 100 mesh. Series 1. .2070 .1036 .1650 .1243 100 mesh 1/n = .0695 1 2 a::::.. = .0279 .000847 .00265 .00288 .00364 .00535 .0173 .0576 Silica (Sea Sand) Series 9. 10.4% water and volatile. 3.324 3.284 3.203 3.149 3.102 3-082 2.978 .00448 .00523 .00600 .00693 .00870 .00987 .0106 Added more earth to Nos. 3 and 4. 11.8% water. == .0657 .0305 .0310 .0317 .0327 .0329 .0333 .0341 .1020 -1020 .1025 l036 -1032 .1036 .1025 0 90 (f) 100 90 (f) 100 100 100 100 - 19 - Silicic Acid (Dr. s. J. Bates) 100 mesh Series 2. 1/n No. Hr. rote t,ed lx ly 2x 2y 12+ 12+ 12+ 12+ 12+ 12+ 20 20 26 26 3x 3y 4x 4y 5X 5y Series 3. =.151 'Nt. earth Equili~rium cone. Adsorbed. Orig. g. dye Vol. (g) (g. per 1.) (g. per g. of (cc.) pres. P,arth) 2.000 2.222 2.000 2.222 2.000 2.222 2.000 2.222 2.000 2.222 .0002 .000370 .000368 .000560 .00133 .00143 .00246 .00274 .00683 .00752 Artificial Earth 1/n = .159 20 10 more 20 27 10 more 3 3' 1 2 1' Series 4. 2.502 2.502 2.501 2.755 2.501 6 6 4 6 5 6 3 2 "'> t:.,f.., 22 1.328 1.256 1.204 1.102 1.005 Series 5. 40 20 20 40 3 4 4' (h) (i) .0244 .0279 .0312 .0342 -0337 .0621 .0725 -0828 .1036 .0953 100 95 (h) 110 125 105 (h) -1560 .2070 .2070 -2070 .2070 .2070 100 100 100 100 100 20.5% water. ~ = .0502 =.134 0.::::... = .351 .00239 .00353 . 00527 .00829 .0164 .1650 .1715 .1870 .2040 166.49 112.80 91.74 91.74 .0820 .0706 .0772, .0740 aL is indeterminate .000163 .000253 .000296 .000310 .0-114 .0414 .0414 .0414 f .. Added 5 cc. more dye sol. to each successive number of x and y. Added 5 cc. more dye sol. to Nos. 1 and 3. Added more alumina. to No. 3. . .., ...,,_.. _ (~) ,6. 100 100 105 (g) 105 (g) 85 85 80 80 75 75 Alumina. ._,'2;1 - ~....~- .1243 .1450 .1450 .1658 .1658 .1865 .1865 .2070 .2070 .2280 .0725 .0746 .0828 .0838 .0930 .0931 -1031 .1023 .00995 .0266 .0439 .0752 .0872 1/n approaches --- ~~- . - - .0622 .0652 Animal Charcoal (Kahlbaum) 1/n - = .225 ~ 175 (i) 185 185 175 - 20 TABLE III. M8THYLENE BLUE: AND C!\LCIN~:D ADSORBENTS. Fuller'e. Earth (Standard Oil Co.) Tre•.i,t3d. Series 1. 1/n = .0615 No. Hr. rotated 1 2 3 4 \i:t. ea.rth Equilibrium cone. (g) (g. per 1.) 60 60 .552 .530 .00241 .00487 t30 .500 .0112 60 .482 .025? Series 2. 1 32 2 3 4 32 32 5.987 4.563 5.234 4.911 32 Series 3. 2 Adsorbed (g. per g. of earth) Orig. g . dye Vol. pres. (cc.) .1875 .1945 .2050 . 2110 .l036 .1036 .1036 .1036 0..:::..,:: 100 100 100 100 -0173 .1036 . 00224 .0100 .0300 -0179 -0828 • .0196 .0205 .1036 1036 Silicic Acid (Merck) 1.00 mesh. 2.678 2.638 20 20 20 20 .0258 .oorno 1/n == .270 1 :: .270 Diatomaceous Earth (Celite) Calcined at 320° for 8 hr. 1/n = .0621 4 3 ~ Calcined at 320° C for 7 hr. Calcined at 320° C for 16 hr. o.::::- = .150 .00283 .0031J .00317 .00359 2.601 2 .517 0 100 100 100 100 .0308 .0313 .031.7 .0328 .0828 .0828 .0828 .0828 100 100 100 100 Calcined at 330° C for 14 hr. 1/n = .262 1 3 4 2 20 20 20 20 2.770 3 parts 1 part .00360 .()0450 56 60 56 1 44 56 .0298 ♦ 0828 .0316 .0828 . 0828 .0828 2 .600 2 -515 .00490 2 .419 .00617 TABLE IV. MIXTURl!~ HTH ME'I'HYL~~Nl'.~ BLUE. .0027 .0341 100 100 100 100 Silicic Acid (Merck) 100 mesh. Fuller's Earth (Midway, Florida ) 100 mes h from 60 1/n = . 105 4 6 3 2 O<:'.'..- :: • 12 7 2.728 2.500 2.347 2 .200 2.100 .00281 .00442 .00706 .0133 .0192 d = .0733 .0379 .0412 .0438 .0465 .0463 .1036 -1036 . 1036 .1036 .1036 100 100 100 100 100 - 21 - Theoretical Calculation of Mixture. 1/n = .102 ~= . 0743 log c log x/m(F) log x/m(s) x/m(F) x/m(s) x/m(mixtu.re) 4.8 3.5 3.6 3.725 3.85 3.95 2.05 2.15 2.25 2.35 1..30 2 .858 . 903 .910 .918 .926 .932 .938 .945 .951 .957 1.020 2 .361 .0722 -0800 .0813 .0828 -0843 -0855 .0867 .0881 .0393 .0906 .1045 .022~ .0290 -0299 .0311 .0324 .0334 -0344 .()354 -0365 .0376 .0526 -035?, .0418 .0427 .0440 .0454 TABLE V. • 4133 .476 .491 .510 .523 .536 . 549 .562 .575 .721 2 3 4 1 2.547 .621 .630 .644 .657 .667 .677 .687 696 .706 .819 .0464 .0475 .0486 .0497 .()508 .0658 0 EFl<'J<~CT OF HCl AND KCl ON ADSORPTION OF METHYLI-i_;N8 BLUE. Fuller's Earth (C. W. Hill Chem. Co.) Acidity .093 N. with HCl. _, 1/n .0927 oe:.. .160 Hr. rot~ted Wt. earth Equilibrium cone. Adsorbed Orig. g. dye HCl cone. Vol. (g ) (g. per 1.) (g. per g. pres. (normality) (oc.) of earth) 1.oos 22 .000850 .1027 .0\13 ■ 1036 100 22 .970 .00156 . 1066 .1056 .093 100 .957 54 .00282 .1105 .1036 .093 100 34 .885 .00768 .116 2 .093 t1036 100 Serie8 1. No. log x/m (mixture) = = Fuller's Earth (Midvray, Florida) 30-60 mesh. Series 2. KCl cone. KCl added as indicated 2' 5' 22 4' 6 22 40 40 7 8 42 5 22 42 1.631 1.514 1.455 1.450 1.452 1.508 1.513 .0005 -00123 .00218 .00545 .00327 .00465 .00177 TABLE VI. .Q&35 .0685 .0711 -0711 .0712 .0683 .0684 (percent) .1036 .1036 .1036 .l.036 .1036 .1036 .1036 2.5 2.5 2.5 2.5 5.0 u: 100 •1 115 ( J; . , 1:1.5 ( J; 100 100 100 100 ~~RYTHROSINE fuller's Earth (Standard Oil Co.) Treated. 1/n ap (Jroaches 6 7 8 5 .) ,J. ( 40 40 40 40 2.729 2.475 r?. .355 2 .252 .00820 .00800 .00865 .00800 ~is indeterminate. .0284 .0313 .0330 .0344 .0783 .0783 .0785 .0783 100 100 100 100 Points No. 2', 3 1 , and 4' were previously run to equilibrium (Table I Series3) and then the KCl added. - 22 TABLE VII. Full er I s Earth (Standard Oil Co. ) Treated. Series 1 1/n = .408 No. 28 28 28 12 12 16 16 28 40+ 40+ 2 5 6 grev. g (K) 4 8 7 . 499 .543 .505 .410 . 593 .380 . 380 . 724 .343 .297 .000522 .000688 .000740 .000875 .000923 .00101 .00102 .00176 .00124 .00201 1/n = .389 16 16 '.L 3 2 6 Orig. g. dye Pres. .1116 .1233 .1290 .1360 .1420 .1465 .1467 .1540 -1605 .1875 Vol. (cc.) .0558 .0670 .0781 .0558 .0558 .0558 . 0558 1117 .0558 . 0558 100 105 110 100 100 JOO 100 125 100 100 .0781 .0558 .1117 . 0558 110 100 125 100 .0558 .0558 .0558 .0558 .0558 100 100 100 100 100 0 Fuller's Earth (C. w. Hill Chem. Co.) Series 2 16 40+ 1..468 .994 1-831 .852 ~ = .796 .000948 .00110 .00136 .00154 .0533 .0562 .0610 .0647 Fuller's Sarth (Midway, Floridt:l.) 100 mesh Series 3. 1/n = .381 (K) c:::k::'... = 2.44 Hr. rota,ted Wt. ea.rth Equilibrium cone. Adsorbed (g. per 1.) (g. per g. (g) of earth) 3 1 1 4 5 2 3 AURAMINE AND NATURAL EARTHS. 16 16 16 32 32 2.202 2. 043 1.862 1.756 1.684 .000842 .00101 .00125 .00153 .00178 ~= .377 .0254 .0273 .0299 .0317 .0330 Rsmo~ed 50 oo. of solution, and added 50 cc. water. - 23 - Diatomaceous Earth (Celite) Series 4. 1/n = .385 No. Hr. rotated ~ 16 5 16 16 16 4 5 (g) (g. per 1.) (g. per g. of earth) 4,. 793 4-386 4.195 4.000 .00168 .00202 .00246 .00271 .01163 .01270 .01330 .01395 Series 1. 4 3 1rev. 1 ( 1) 5 rev. 5 (1) 6 'I' \ I = .136 Wt. earth Equilibrium cone. Adsorbed TABLE VIII. 2 C<. Orig. g. dye pres. .0558 .0558 . 0558 .0558 Vol. (c ~.) 100 100 100 100 AURAMINE AND ARTIF'ICIAL ADSORBENTS Anima.l Charcoal (Kahlbaum) 16 16 16 .620 .422 1.6 more .512 .512 .349 46 16 more 46 72 • 390 .349 .330 1/n = .161 0-::-= .366 .00132 .00178 . 00281 .00366 .004:l.6 .00502 .00577 .00860 .1257 .1320 .1425 .1510 .1517 .,'l5?5'.· .1585 .1670 Removed 70 cc. of solution, s.dded 50 cc. water. .0781 .0558 .0558 .0788 .0781 .:D5'!34 .0558 .0558 110 100 1_00 90 ll0 · '8 (i: { 100 10(, - 50 - -Discussion. - -From the data and graphs s everal interesting relations may be noted, and conclusions drawn. The result of prime signi:f'ica nce is that the ad- sorption of dyes by fuller's ea rth, diatomaceous earths, silicic acids and charcoal is quite closely in accordance with the adsorption equation. That equilibrium conditions were actually attained is shown by the results of certain reversal experiments in which the adsorbent was caused to give up pa rt of the dye which.i t had previously adsorbed. (Table I Series 4, Table VII Series 1, Table VIII Series 1, and Table II Series 1.) In the first of these, part of the solution which had co me to equili brium was withdrawn and replaced by water, and in the latter the amount of adsorbent was increased by the addition of a second portion of earth. A slight tendency of incomplete r eversion was noted in a few cases, indicating that a true equilibritun had not been reached in the original or reverse solutions. These deviations, however, are almost within the experimental error. Only one variation from the adsorption relation noted. x/m =~ cl/n was In the case of the trea ted Standard Oil earth a point of maximum ads ci r ption or a rather abrupt change in equilibrium was reached. Not all of these solut ions of high concentration were carried to equilibrium, but th e da.ta shows that they tend to approach with increasing time of agitation a possible saturation value. Sample No. 1, however, - the one indicating the highes t adsorption, - dld not change in concentra tion in twenty days. The amount ads orbed at this point is more than one-quarter the weight of the earth, and may represent the state of complete coveri ng of the elementary , 1 adso r ption s paces. 1. L~ngmuir, J. Am. Chem. Soc. 38, 2221 (1916). - 51 A very strikine relation to be noted is th e constancy of the exponent 1/n for all the ea rths of natural origin, namely fuller ' s earth, di a tornaceo us earth, a nd sand with each of the dyes , methylene blue and auramine . 1 This constancy of 1/n is furthe r confirmed by the data of Atherton Si dell for a ful l er' s earth, and of Pelet and Grande 2 for a diatomaceous earth. Methyl ene blue . 1/n 0,::::,. Adsorbents. Auramine. 1/n o-L- Fuller ' s Earth . 0 . 0377 0.381 0 . 0675 0 . 0627 o.796 0.389 0. 0706 2.44 0.408 0 .136 0 .385 " 0 . 0980 0. 113 Ot-0994 0 . 0667 0. 0678 0. 0663 100 from 60-100 0.115 0 . 0655 Hill Chem. Co. " " " ( . 09 3 N with HCl) 0. 174 0. 160 0.344 Standard Oi1 Co . (treated) (treated , calci ned 320° C 7 hr.) 0 . 270 0.477 " (untreat ed) 0 . 0615 0.0595 (data of cAtherton Sidel l , 1.) 0.0685 Midway , Florida, 100 mesh " " 60-100 " " " 30-60 " " 0.266 Diatomaceous Earth 0 .0279 Celite 0 . 0258 " (calcined 320° C, 8 hr. ) Ki eselguhr (da t a of Pelet and 0 . 0159 Grande, 2. ) 0 . 000838 Sea Sand (100 mesh ) Silicic Acid Merck' s 100 mesh Q.0657 " (cal cined q20° c, 16 hr) 0. 150 1. 2. J. Am . Chem . Soc. 40, 312 (1918). !) . 0654 0 . 0621 0. 0690 0. 0695 0.143 0. 270 Thi s author did not apply the adsorption equation to his results , but t he calculat ions of~ and 1/n were made f rom hi s data . Zeit . Chem. u. Ind. Koll . 2, 41, 1907. These a.uthora gave 0 .11 as their value of 1/n for kieselguhr, but recalcul a tion f rom their orig inal data gp.ve 1/n = 0 . 0690. Their other values of 1/n proved w be substantially co r rect. - 52 Adsorbents. • Methylene blue. o.:::.. 1/n Silicic Acid Merck's (calcined 330° C, 14 hr) 0.127 1).225 (prepared in this labcratory) (data of Pelet and Grande, 2.) ( " 11 11 11 11 Auramine. o,:::. 1/n 0.262 0.151 0.0728 0.154 (calcined)) Q.0644 O.140 Arti £'icie,l Earth (Na 2 0 .Al 2 0 5 .3SiO2 .3H 2 C) 0 .0502 () .159 Animal Charccal ( l<aLJ baurn) (data c f ?elet and Grande) 0.351 C).134 0.0797 0.143 0.36S It is of interent to note here that the va.lues of 1/n obtained f o r the natura l adso rb ent P wit h methylene blue are smal1_er than any other values Jisted in tlte 1 iterat 1.1 re. The -f'act t,}rnt t .he valu e of 1/n is t:te sam e f o r these various substanc es i ndica te s thr:,t t lH, nature of the active ads c rting constituent or sur:f'ec P- i s s l 1i:i ·l e.r for a.l l th es e ads orbents even t ho ugh they diffE;r g reatly in ch e,;; ica } com1osi tion and physici:•l st ructure. The artificial adsorbents with methylene bJue falJ into another grou1J with practically a ccnstant value of l/n wHct is a p ,_,roximately twic e tl"iat fer ti1e natural ea.rths. Further confirn,aticn of th es e velues is given by the resul ts of Pelet and Grande (reca.Jculated ) for a s il i cic acid, and ani me ) charcoal (see a.bov e t a.ble). The V8.lue of 1/n for animal charcoal and a ur amin e , howev e r, do es not have t h i s r a ti C> of tv,o to one with th e natlff&l eart!.:s a s L:1 the case of methylene blu e. Time wa s not available for s ecuring the corres1;onding data - 53 for silicic acid aI'-d artificial earth with a uramine, which wc,uld be necesee,ry before a ny relation could be ded,_1.ced. It is nevertheless of interest to_.note tL.a.t e.rtificially ::re1-a.red sodLm aluniinurr silicate, and precipite.ted silica. with . methylene blue should have the same value of 1/n, which is Just tvo:ice tl1.at for fulJ er' s earth and for th e na.tura. l silicas. Another ir::: ortant 1)oint shown by the data is the large variation of~. This va riati on can hardly be accounted for by :nerely considering the extent of surface of the riarticles. It is true thr;t t hu ex l:, ent of s urface is a r a ther indet e r minable ciuanti ty, and does not necesE>e.riJ.y d P,pend on the mesh 'ceca.use of the g reat variations in porosity or internal surfa.ce; but a rough ccmr arison can be made by r.1 eans of rdcroscopic exe.rninaticn of the particles. The Celite and Hill Chemical Co. fu1Jer 1 s ea.rth were thus ex8n,ined. The Celite has 8 very fine shell structure of varying but definite shape. The fuller I s earth 1Jarticles have ar1 indefinite sh8.pe and are c onsiderably mere ol)aque than the Celite. The opaqueness of the particles which is a rough measure of the thickness or porosity, indicates the.t the Celite has the greater surface. This same conclusion can be drawn from the fact tha.t the density of tirn Celite is much smaller than that of the fuller's earth. If then the Celite ho.s tb.e greater surface, it is evident that a,c::::,.. is not alone deIJendent on the surface, as ~ ten times that for the Celite. for the Hill Chemical Co. earth is over The value of,:X:::. may, however, vary with the number of elementary adsorption siiaces (Langmuir's theory), or extent of active Burface v:h.lcl, is different for the various adsorbents irrespective c,f U:e surface. - 54 An attem~,t ;{as rHade to determi ne the v~.lue of""-- for the same earth of different meshes to see if it varied with the extent of surface, or r oughly with the !nesh. Fuller's ea rth (!Jiidway, Florida) of different meshes was used for this l-,ur_r:ose. The coarse samples , however, were found to disintegrate on agitation, and hence such co mparisons could net be made. The 100 :r.esh earth ga v e a i::malJ.er value ofo.:::. than the 60-100, and 30-60 2 esh just the reverse of what would be ex pected. A lack of homogeneity cf the sam;lles as received was sus 1lected, and so the 60-100 mesh sample was gr o·;incl to 100 mes h and the constants redetermined. As was ex>ected, this ga,ve a g reater value of o.:::: than that for the 60-100 mesh. The order of the val u es of~ for t he various earths ls not cha nged by using different dyes of tbe same character. r aring the ratios of the values of with e,urarnine. This can be shown by com- oc for two earths with methylene blue and In order to ma ke this ccmparison of the adsorption power of the various earths, which should be free from w.certai:nties in the values of 1/n, t he actual arr.ount of dye adsorbed at some one equilibriUJn concentration within the experimental range ( 0.00159 g. per 1.) was determined for the vario u.s earths. These values and t h eir ratios are given in the following table. Methylene blue. Celite Fuller's Earth (Midway, Florida) ~ller' s Earth (Hill Chem. Co.) Ful J er' s Earth ( Standard Oil Co.) 0.0183 0.0639 0.113 0.219 Auramine. 0 . 0115 0 . 0324. 0. 0649 0 .174 % deviation. Florida Earth 3.49 2.82 + 23.8 11.95 15.10 - 20.8 Celite - Standard Oil Earth ---·•-· ··-····· . . .. . . . . _________ __. ·-··- • -Celite = - 55 - Auramine (Cont.) Methylene blue (Cont.) Hil l Chem. Co. Earth ·-· ... .... ,__ = Celite %deviation (Cont.) 6.18 5.65 + = 1.94 2.68 - 27.6 = 1.77 2.00 - 11.5 = 3.43 5.37 - 36.2 ~ 9.4 Sta.ndard Oil Earth Hill Chem. Co. Earth Hill Chem. Co. Earth F)orida Earth Standard Oil Earth Florida Earth Though these deviations of the ratio are large, they are not all in the same direction and hence cannot be attributed t o the nature of either o:f the adsorbed substances. Only in the case of the treated Standard Oil earth was erythrosine su~ficiently adsorbed to _obtain quantitative data. Fro m a qualitative standpoint, however, the order of adsorption was the same as that given by methylene blue and aura.mine. It hence a ppears that even such totally different adsorbed substances as acid dyes and basic dyes place the earths in the same order of adsorptive power. The fact that with methylene blue the natural and artificial adsorbents have t wo disti nct values of 1/n suggests that this difference may be due to some physical difference between the adsorbents. Thus Celite and silicic acid have practically the same chemical corn_?osi tion, while the value of 1/n for t he former is 0.0654 and the latter 0.143. An imr,ortant distinction between the natural and the artificial substances may result frorr: the di:fferences i:r, the pressure or temperature to which they may have been subjected. - 56 Severa,l of the adsorbents were hence heated a.t about 320° O for 7-16 hours t o determine whether th e vaJue of 1/n would be affected. In the case of the Celite and Standard Oil earth (treated) 1/n was unchanged by calcining. The value of ex::.. , however, decreased. If the effect of calcining wer e merely to remove the free water, e,n increase in the value of would be expected as more earth would have been usee:k per unit of weight. '%- Loss of Water, % Total% decrease of 7.5 21.5 10.4 21.7 17.9 Decrease of Celite Standard Oil earth 43.2 The calcining :i.n this case may hence be said to reduce the extent of active surface but not to affect its nature. In the case of the silicic acid, however, the values of bothoL and l/n were changed. Instead of being decreased as might be expected, 1/n in- creased i n value, being practically doubled. Further data of Pelet and Grande is not in accordance with these re_§ ults, but rather with those of the Celite and Standard Oil earth in which 1/n was not affected by heating the substance. Pel et and Grande, however, did not give the temperature and time of calcining which would likely affect the results, and perhaps ac count for t h e deviation (see t a ble of va.lues of<>"- and 1/n). Sufficient data is not available to make any generalization concerning this phenomenon, which would undoubtedly repay further investigation. It was at first thought that a.n adsorbent which followed the experimental adsorr,tion equation was necessarily a siraple adsorbent as a mixture of two adsorbing constituents with different values of 1/n would give a - 57 curved line when plotted logarithmically. This line is, however, only slightly curved 0 ver a. considera ble range for mixtures of' adsorbents whose values of 1/n differ by one hundred percent or less. The determination for a mixture of three parts of silicic acid and one part of Florida earth (100 mes h fro n'! 60) gave a straight line with the same value of 1/n as the theoretica'4 but a slightly smaller value of~ . It is evident fro m this that a practi- call y stra.ight logarithmic curve does not prove the presence of a single type of adsorbine; constituent in the earth. The additi on of acids and salts to the solution does not alter the value of 1/n, bu t does affect that of o.:::::. (see table of values of °"and 1/n). Both the addition of acid to the adsorption mixture, and the treat ment of the earth with acids, as in the case of the Standard Oil earth, lower the vaJ.ue of oL- for the basic dyes. selective a.dsor ption. This may be explained by the theory of The earths have a special af:finity for bases, even when neutral, and will adsorb t h e bases fro m a basic dye, leaving the solution acid. When more acid is added it has the effect of reversing the r eac t ion o r in other words di minishing t h e adsorption. It is of interest t o note here that the effect on the adsorptive power, or value of DL of treating an ef>.rth with acid is ouite diff erent in clarifying lubricating oils from the adsorption of basic dyes. In oil purificf:l.tion the acid treatment has the effect of increasing the adsorptive power severa.l f old. This difference may be due either to the nature of the sol vent, the na.ture of the adsorbed substance, or both. It appears as though the earths v,;hich are negatively charged to wards water are positivel y charged to wards oils. - 5g - It was not possible to obtain results from the addition of a base to the adsorption rnixture in the case of methylene blue, as this dye is preci pitated by alkalies, and hence com~·,1ications would arise. Potassium chl oride was us ed to determine the effect of a neutral salt. Like hydro- chloric acid, it decreased the value of ,x::. • Complications entered into the effect, however, because of further disintegration of t he earth. Identi- cal e.dsorpti on mixtures, differing onl y in that potassium chloride was added to sorr,e, were run under parallel conditions. The presence of the salt caused a decreas e in the adsorptive power (Table V Series 2.). Peculiar res ult s wer e obtained with alumina and methylene blue, and the Standard Oil earth ( treated) and erythrosine. In both cases a definite amount of dye was adsor bed irrespective of the amount of adsorbent employed. It, is questionable in both of these cases if the ;,henomena. represent a true adsorption as methylene blue is precipitated with alkalies, and erythrosine with acids. The adsorbents do not contain free soluble base and acid, how- ever, as shown by experiment. If some of the alumina is agitated wit h water and then phenolphthalein add, the water turns pink; but if it is filtered before adding the phenolpht halein ther e is no color change. Similarly, if the treated Standard Oil earth is shaken with water and e. piece of blue litmus paper touched to the earth, it will turn red,; but if the mixture is filtered and the litmus paper dipped into the filtrate there is no color change. plained by the theory of selective adsorption. This may be ex- The alumina which evidently has a greater affinity for acids than bases adsorbs the acid from the phenolphthalein, leaving the solution alkaline. When phenolphthalein is added after - 59 - filt ering off the alumina, there is no chance for adsorption, and as the alumina is insoluble in water, the water remains neutral. In the case of the Standard Oil earth the base is adsorbed, leaving the solution acid. From this standpoint the alumina may adsorb the acid from the methylene blue, leaving the solution alkaline, and then the base of the methylene blue is preci pitated. Conversely the treated earth adsorbs th e base from the erythrosine, leaving the solution acid, and then the acid of the erythrosine is pr ecipitated. This, however, does not account for the peculiar fact that t he adsorption, or at least the amount of dye removed from saturation, is independent of the amount of adsorbent. The constant equilibriurn concentrSr- ti ons may represent the solubility of the dyes in basic and acid solutions res pectively, under conditions such that a slight change in alkalinity or acidity does not affect the solubility. Sum:nary. In conclusion this investigation has shown several im1>ortant points. (1) The adsorption of dyes from aqueous solutions by fuller's earth, diatomaceous ea rth, silicic acid, and animal charcoal is, over t he range 1/n tested, in close e.ccordance Vi i th the adsor ption equation, x/m = ~ o . • (2) The value of 1/n is the same for all natural earths, namely fuller's earth, diatomaceous earth, and sand, both with methylene blue and auramine. (3) The value of 1/n is constant for the artificial adsorbents with methylene blue, and is approximately twice that for natural adsorbents with methylene blue. ( 4) The value of~ varies greatly for the different adsorbents, and is probably not dependent on the extent of surface alone. - 60 - ( 5) The order of the values of .X::.. for the vari ous adsorbents is independent of the dye used. (6) Adherence to the adsorption equation does not indicate a single type of adsor::>ing constituent in the adsorbent. (7) Calcining the natural earths does not affect the value of 1/n, and r educes that of ~. Calcining silicic acid, however, does affect the value of 1/n as well as that of ac:::::: • (8) The addition of acids and potassium chloride to the earths de- creases the value ofa.::: for basic dyes, while that of 1/n is not affected. This is in accordance with the selective adsorption theory.