15, 1946. F. J. EWlNG EI‘AL 2,409,263 DESICCAN'I' Filed April 1:5, 1943 s Sheets-Sheet 1 7Mru4a5vus a a: . \ g ‘ % ‘it/J4 Q 8 kEDEE/CK dE/wnve, R065»? A. La V577; INVENTOR. Q \ R ‘ BY I @Mur : ATTORNEY. ' F. J. EWING ETAL 2,409,263 DESICCANT Filed April 13, 1943 6 Sheets-Sheet 2 35 2G 80 7” MIA/W55 ?soselcz Jib/Na 806EEA . L OVETT INVENTOR. 2. BY ?ogw ATTORN EY. F. J. EWING ' ETAL 2,409,263 DE SIGCANT Filed April 15, 1943 6 Sheets-Sheet 3 TEMPRAUIQ a ,FkEQEQ/CK d'Ew/A/s, @ZOWA Bose-2A. Lo vsrr, INVENTOR. BY m5% ATTORN EY. Oct. '15, 1946. F. J. EWING EI‘AL 2,409,253 DESICCANT Filed April 13, 1943 6 7 v a WM. "/0 9 6 Sheets-Sheet 4 10 ' iésose/cz tl'?i'wmvaa 30652 A’ .50 r577; INVENTbR. [3' 4‘ BY M ATTORN EY. Get. 15, 1946. F. J. EWING ETAL 2,409,263 DESICCANT Filed April 13, 1943 6 Sheets-Sheet 5 M. .zE’asé-eA. Lo VET’; INVENTOR. ' ATTORNEY. Patented Oct. 15, 1946 2,409,263 " UNITED STATES “PATENT OFFICE DESICCANT Frederick J. Ewing, Pasadena, and Roger A. Lov ett, East Los Angeles, Calif., assignors to Filtrol Corporation, Los Angeles, Calif., a corporation of Delaware r 1 Application April 13, 1943, Serial No. 482,926 3 Claims. (Cl. 2.52'—269) This invention relates to adsorbents for vapors, and particularly to desiccants; that is, adsorbents for water vapor, and to methods of producing such adsorbents from natural clays. We have discovered that we may produce a highly e?icient adsorbent for water vapor from montmorillonite clays. We are enabled to pro duce clays which are highly efficient as desic cants, particularly in adsorbing vapor from air of low humidity of about 30% or lower relative humidity, which are superior to known desic cants. We have accomplished this by drying the 2 adsorptive eiiiciency, the more useful is the ad sorbent. . Another important consideration is the hard vness of the desiccant. It should resist abrasion and crushing. If it is not hard, but 'soft ‘and friable, it will dust and contaminate the pack aged material. , We have developed an adsorbent which has a much greater adsorptive ef?ciency at such rela tive humidities. Our material will,'at tempera tures of about 80°, absorb from 18 to about 20.5% of its vweight of water at 40 relative humidity, desiccants at a controlled rate and to a controlled from 17 to about 18.5% of its dry weight of water ?nal water content. from air of 30 relative humidity, from about Desiccants are now employed for dehydration 15 13.5% to about 15% of its dry weight of Water of air and/or other gases such as refrigerant from air of about 20 relative humidity, and from gases. They are employed in dehydration of air about 8.5 to 11% of its dry weight from air of and other gases in closed spaces in order to pre about 10 relative humidity. “Relative humidity" vent corrosion, mold, or mildew. This is par will be hereinafter indicated by the abbreviation ticularly important in shipment of metallic ob '20 “R- H‘!, jects such as machine tools, ordnance or other Due to its greater adsorptive capacity We may metallic objects which are crated for shipment use less of our desiccant to obtain the dehydra or storage. Packaged foods, such as fresh or de tion required to preserve the package at a rela hydrated fruits, vegetables, and meat are also tive humidity of 30 or less. \ subject to spoilage in humid atmospheres. Our desiccant is also characterized by the fact Lumber and wooden objects are subject to warp that it is hard and not friable and substantially ing or so-called “dry rot.” Packaged material non-dusting under ordinary "conditions of han must be protected from the attack of moisture dling, packaging, and shipment. It will show if they are‘ to be subjected, either in shipment about 1% loss by the standard method of deter or storage, to hot or humid climates. 30 mining hardness as given in “Army-Navy Aero In' order to be effective for such'purposes it nautical Speci?cation,” Speci?cation No. AN is desirable that the desiccant have a high ad 13-6. sorptive emciency for water, not only from air In the following discussion we shall refer to of high relative humidity, but also from air of “volatile matter” and “per cent volatile matter,” low relative humidity. Essentially the objective 35 or its abbreviation “V. M.” By this quantity we to be obtained is to maintain the atmosphere en mean the per cent loss as moisture of clay ignited ' closing the object to be protected at a low relative at a temperature of 1700° F., in accordance with humidity. > the method more fully stated hereinbelow. It is therefore desirable to incorporate into the We have found that we can obtain such a des-' enclosed space, such as a car, ship, warehouse, 40 iccant by controlled drying of montmorillonite or the package itself, a desiccant which has a clays of the sub-bentonite type, such as are activ high adsorptive e?iciency capable of drying the atable by acid treatment to produce useful. oil air to a safe humidity. Experience has shown decolorizing clays and cracking catalysts, Such that such desiccant should have the ability to clays are the substantially non-swelling bentonite adsorb a large amount of moisture from air and 45 clays as distinguished from the swelling type ben must particularly have this high capacitywhen tonites whose decolorizing power is not substan it is contacted with air having a relative humid tially improved by acid treatment. We have ‘ ity of 30 or less. found that-such clays, or such clays when activ Useful adsorbents are such that can absorb ated by acid treatment, such as is used'in mak at least‘13% of their weight of water from an 50 ing acid activated montmorillonite clay adsor atmosphere of 30 relative humidity at a tem bents and petroleum ‘oil cracking catalysts, are perature of 75 to 85° F. The adsorptive efficiency converted to highly e?icient desiccants if theyare at lower humidities may usefully be at least 10% dehydrated to a volatile content of from about by weight at 20 relative humidity and 5% by weight at 10 relative humidity. The higher the - ‘5.5 to 8 V. and most e?cient when brought, by proper drying procedures, to a V. M. content‘ 2,409,263 4 3 of about 5.5 or 6 to 7%. This is somewhat greater than the theoretical hydroxyl water content of pure montmorillonite, which is about 5%. It thus appears that if the dehydration is carried out under such a condition as to remove sub stantially all free water, but not to remove the hydroxyl water content of the montmorillonite, . 801-1’ F., according to the United States o?lcial methods speci?ed by the Bureau of Ships ad In terim Speci?cations issued November 1, 1940, No. 51832 (INT), and Army-Navy Aeronautical Speci ?cation AN-D-6, issued November 20, 1942. According to the method speci?ed in this bulle we can obtain a highly e?icient desiccant. tin, ten gram samples of the clay are weighed into an adsorption bulb. The bulb is connected to the a V. M. content of 5.5 to 7% will give the most a substantially constant value, normally 80° F. adsorption apparatus which consists of a train We have found that the rate of heating of the moist subbentonite clays, that is, the distilla 10 of bottles containing the sulphuric acid whose concentration is adjusted so that an air stream tion rate of the moisture in the preparation of bubbled through these bottles will attain the de the desiccant, has a material effect on its re sired humidity. The temperature of the sul sultant adsorptive efficiency. The removal of phuric acid solution and the bulb is controlled to water at a moderate rate in bringing the clay to e?lcient adsorption. Periodically, at intervals of one or two hours, the bulb is weighed, and the process is repeated until two successive weighings, approximately one hour apart, do not show a weight variation exceedi g In order to obtain the most e?icient adsorbents, 20 ten milligrams. The gain in weight divided by the original weight of the material multiplied by particularly for water vapor at low relative hu 100 gives the percentage by weight of water va midities, we desire to cause a relatively slow dis por that the material will hold in equilibrium tillation, particularly during the drying of the with the air at the relative humidity attained in‘ clay from about 8 or 9% V. M. downward, and to reduce the water content of the clay to the re 25 the saturators, and this quantity is hereinafter referred to as “adsorption e?‘lciency.” . gion of the optimum value at which the maxi The particle strength or hardness of the ad mum e?iciency is obtained, and interrupt the sorbent is determined according to'the method heating to secure a clay of such optimum water speci?ed in the above bulletins. This method content. ' The process controls which produce our unique 30 consists essentially of exposing the clay to the at mosphere of a room to permit it to come into desiccant will be further understood from the fol equilibrium with the water content of the room. lowing description taken together with the fol Then 150 grams of sample are introduced into an lowing drawings, Fig. 1 to Fig. 5, inclusive, which 8" #16 sieve backed by a #18 sieve and shaken in show the effect of the various heating variables on the desiccant e?lciency and illustrate the de 35 an apparatus which has a single eccentric, cir cular motion of about 290 R. P. M. and a‘ tapping sirable limits of heating which permit the attain There appears to be an optimum V. M. content ' for the dried clay produced which gives the great est adsorptive efficiency. action of about 150 strokes per minute. The sieves are shaken for 15 to 20 minutes until not over 0.05 gram passes through the #18 sieve in Fig. 1 to Fig. 5, inclusive, are charts showing 40 one minute of continuous sifting. "The fraction plots of the data hereinafter presented. ment of the desiccants which constitute a pre ferred embodiment of our invention. passing through the #16 sieve and onto the #18 sieve is used for the particle strength test. When testing larger sire particles, 9. suitable quantity non-swelling sub-bentonite type produced at shallbe ground to produce the required fraction Cheto, Arizona, and-having the following analysis 45 through the #16 sieve onto the #18 sieve. based on a volatile free basis: In order to determine this strength, 50 gram Raw Cnnro-V. F. BASIS of the sample prepared as above-described are The clay employed in the following examples was a montmorillonite clay of the substantially Silica (SiOz) __________________________ -._ 67.3 Titanium oxide (T102) ________________ __ 0.3 Aluminum oxide (A1203) _______________ __ 19.5 Ferric oxide (F8203) _____~______________ .._ Manganese oxide (MnO) _______________ __ Magnesium oxide (MgO) _______________ .._ Calcium oxide (CaO) __________________ __ 1.78 0.80 6.9 3.2 99.8 placed on an 8", #30 sieve which is backed by a #45 sieve and a retaining pan. Five copper discs 60 of the size and weight of one cent pieces are placed on the #30 sieve, and the sieves are shaken for 15 minutes in the above apparatus. The sample remaining on the #30 sieve and the quan tity which is contained in the retaining pan are 65 weighed. We use the term “hardness" as the per cent retained on the #30 sieve. We use the term “friability” as the per cent falling through the The original clay so processed has a natural #45 sieve and retained on the pan. _ In the following examples, the clay was dried V. M. of about 38 to 42% as mined, and may drop 60 in an oven heated under controlled temperature to 32 to 3'7 % in transit to the mill. conditions. The loss in weight of clay at various In the following examples V. M. or per cent intervals of time was determined by weighing the . V. M. is determined as follows: clay without removing the clay from the oven. Approximately ?ve grams of sample are placed The clay was heated in accordance with the tem in a porcelain crucible and quickly weighed to the nearest one-tenth milligram. After a preliminary 65 perature schedules hereinafter set forth under each example. heating at a low temperature, the crucible is is Example A.--The clay at room temperature nited at a temperature 01' 1700" F. for approxi was introduced into the oven which was main-' mately twenty minutes. After cooling in a desic tained at. 350° F. during the entire drying proc cator, the crucible is again weighed. The loss in weight divided by the original weight of the clay 70 ess. The clay was held in the oven for four hours at this temperature. The clay was then removed multiplied by 100 is the per cent V. M. to a desiccator and cooled. The volatile matter The per cent gain in the weight of the clay content of the cooled clay was 5.66%. when in equilibrium with air at various relative Example B.—The clay at room temperature humidities, herein referred to as “adsorptive e?l ciency," was determined at a temperature of 75 was introduced into the oven maintained at 400° 2,409,203 I 6 F. during the drying process, and the clay was TABLE III held in the oven tor four hours. The clay was removed to a desiccator and cooled. The, volatile content of the cooled clay was 5.56%. ' Percentage water vapor adsorption at indicated ' .Emample C.—The clay at room vtemperature was introduced into the oven maintained at 500° F. during the drying'processt and the clay was held for four hours in the oven. This constituted a shock heating suddenly to 500° F. The clay Example 10 E ___________________ __ F _________________________ -_ G_-_ ______________________ ._ 20 9. 8 l4. 3 9. 8 15.0 9.8 14.1 40 60 80 19. 5 20. 7 18.9 22. 9 23. 9 21.8 26.8 27.8 26.0 100 35. 6 36. 7 34.6 was removed to a desiccator and cooled. The V. M. of the cooled clay was 5.52%. I ' .The drying curves of Examples vE, F, and G are Example D.—The clay at room temperature charted in Fig. I on which is plotted the V, M. was introduced into the oven maintained at 600° content of the clay at various times, as given in F. during the drying process, and the clay held Table II. . In Fig. 3 we have charted the dehydra for four hours in the oven. The clay was removed 15 tion curve of the clay. giving the v. M. content of to a desiccator and cooled. The volatile matter the clay at the various temperatures as reported was 5.07%. I ‘ I The adsorption emciencies of theselclays are given in Table I: ' . > inExamples Table II. H, I. and J.—The clay at room > I tem . TABLII . perature was introduced into the oven maintained , 20 at 350° F.‘ during the drying process. The cold. > clay was introduced, rose to the temperature of Percentage water vapor adso tion the oven, and then was maintained at that tem- _ at indicatedB. H. m Example perature vfor a period of time. The clay was re ' ' moved and cooled. 10 2o 40 60 80 100 as 13.7‘ ‘19.8 214 27.5 35. 8.0 13.1 19.2 23.2 5.6 ‘11.9 17.9. 22.5 27.5 26.6 35. 34. 25.0 34‘. 15 7.8 16.3 20.7 The clay, as in the previous 25 examples, lost some water while being cooled and the terminal V; M. was determined. The follow ing TableIV gives the temperature schedule; 1. ‘e.. Gibon-1 theheating rate and the volatile matter content of the clay at each temperature level and time in-' 30 terval, thus giving the distillation‘ rate. The ter In the above table, as in other like tables‘here inafter given, the adsorption e?iciency at any intermediate value of R. H. may be taken from; curves drawn for the values of adsorption efn- . ciency vs._ R. H. These curves arev'not given. minal V. M. of the clay was: ‘In Example H, 6.85%; in Example I, 7.15%; and in Example J, 7 15.89%. 35., , herein in order not to multiply the number of ?gures. Temp., °F. 7 The‘ values so given. however, may be ' Time, minutes taken as a disclosure of, the adsorption e?iciencies at intermediate values of R..H., such as 30 R. H., by the plotting of such curves, as will be under stood by those skilled in the art. - " l ' TABLE IV Exam ple H ple I Percent volatile matter Exam ple I Exam ple H Exam ple I ' Examples E, F, and G.-The clay- at room tem perature was introduced into the oven at 90° F. and the oven andv clay were gradually raised in temperature. Tlie clay was :then femoiéifat the various times indicated in the following vtable at the last V, M. recorded under‘ eacl'rexample, and cooled. During the cooling period some ad ditional water was lost. The V. M. content of the cooled clay in Example E was 6.72%; in Example F it was 5.65%; and inExample- G. it was. 7.61%. The following Table No. II gives the tempera‘. ture schedule; i, e., the heating rate of the clay in each of the examples is given. Wegive also the volatile content of the clay at each of the 55 The adsorptive‘ ef?ciencies of the. clays of sam ples H, I, and J at the various values of R. H. are temperature levels and times. I given in TableV. TABLE II Temp, °' F. Percent volatile matter . - - ‘ ‘ TABLE V _ Percentage water vapor adsorption at indicated v80 Time, minutes ' } --.. - R. H. > . ' Exam- Exam- Exam- -Exaln- Exam- Exam ple E ple F ple G I ple E 96 138 ‘189 90 158 210 32.0 30. 4! 25. 7 32 30' 26 234 256 - 250 234 18. 7 12. 5 18 12 298 ' ' ple F \ 304 8.7 311 7.8 77 _7_ 6. 5 318 ' 332 ...... __ ple G Ema. 8. 6 i 2o 40 9.5 14.3 9.5 9.9 14.0'19.4 14.9 30.3 60-80 19.7 22.6 22.3 23.6 100 21.0 36.2 26.7 28.2 33.7 35.3 The vdrying data of Table IV are charted in Fig. 2. In Fig. 2 is charted the time V. M. curve. From . _ 701 Fig. 2 it will be observed that the drying rate of 5.8 samples H, I, and J in going from'32% V. M. to G .at the various relative humiditiesare given )in ' '10 6. 1 5. 8 The adsorptive e?lciencies of samples E,-F','_and Table III. . . 7% V. M. has an average value of 1.59% V. M. loss per minute. . - Examples K, L, and M.—The clay at room 75 temperature was introduced into the oven at a 9,409,283 ‘ 7 - Tut: VIII temperature of 550° F. The temperature ofthe ' clay rose rapidly. The clay was removed from "m ._.c u .n. .e m m uanan.En..u! n xm n.xh n.o. mED.Em.mm. m Wm “"1 MNM2 mm.“M.”.m”w?m m.98m7? m .an Ep .an I was. 1“w. u o2..rr. m .._ ml. m m m m m m .. “8TTo.de.e?0Foncm.El.dmne mgmmmmnu.xnnT:1.m.K.“72m.emyeum.eumaHR.wotwDmb22a.Mana.hor_.e1su._4_ayv.v1Ld“..beme1miLe2l.Mv.0,,5aewm Tmmmxetyd.maMammmmwnm1i5a;e?a.t.stcmwé.mo».dmH.mWmV.Lu23e.".wm5tnm?udVamimmmmu.nmam.m3mimMa?alm.nm».2tab.um3ee.1a-f.ot9-.w0 ?mnu.iceMm.ewmlMM m 8“Uy0taem6_wk6%0mr%?b2nmMitaanm1.vwn6am?.wummma.w%w.9Wm menm inmet eetNtu40t01°_. m amm p.m3nmwymnmvpumwfmmsMwowumc.n.Qa3wmMRéimmvratmnd wWe»M?sGMdSWMuK“t5m?-O71tI.%0r“wnmammaamTmmwm.médwetnm1e4mwuewm ma"snhEPKacwm.vLdsuM?m“wmva.nzMc.mmmt%:.mwMenmw m tmgdawtmmasMy:ma .0 .T _..1._e..oa. .l.._.n. .8.._.’ ndm0LmctpmdurIe"dnI"ielaw!mm nama.m.wn6raaI"oec. m%%ed.d.sxow7d"ltm,hmtc.mo .M .m Sam ethem okea ce.fm ._ _. nememvIne2m.moI"V0 M.1on.sdTa,etdw._lmMr M -t V. a ..d D. u .I S.1. mmwim umm m M m w 5"w dc an7 d W an .t3a O fJ .7.81 ed eta1l.mmmalfmesz?ém n?am uma m.Pum.s.vmjwurmnn m cm .._ .n m _ 21 -a-_ WMm annmm mmm m m mm eaM anmesat.m.u“ mm“1mmmwmmmummm m m MmNaWT.Quagai"stnesamcnw 0mR3.90.tmU“awiotmMw72Jl1?uhw? mnammuWemm?wunm1ma9zs.?ac en m u n u an an m m an. 2T0 .H" .. mm.ammui.nan..as?af?.m .ma s 1n.$364.’ corded in the tablefor each example, and the V. M. content and adsorption e?iciency deter The clay in the case of each example was re mined. The drying data of Table VIII are charted in Figs. 1 and 3. In Fig. 1 are charted 60 moved irom the oven at the times indicated for the last value or the V. M. recorded, placed in a the time V. M. curves and in Fig. 3 the tempera desiccator and cooled. ture V. M. curves. In Fig. 1 are charted the The V. M. content of the cooled clay at the values for Examples S, T, U, V. The remaining end 01' the drying process for each of the samples examples fall closely on these curves. The dry ing rate of these examples was closely similar to 65 N to V, incmsive, is given in Tame up that of Examples E, F, and G, giving eil'ect- to the lower initial V. M. content of the clays of Ex amples N to V. The average value of the drying rate for all the Examples E, F, and G, and N to V, inclusive, are substantially the same, and the 70 average value of all the Examples E, F, G, and 0 to T, inclusive, was 0.47% V. M. per minute. Table VIII gives the V. M. content and the tem perature of the clay at various times during the drying operation oi Examples N to V, inclusive. 76 Tuna IX Percent volatile matter 18.73 129 9.15 1. 47 6.1 5-89 5.62 5.57 5.76 2,409,268 9 10 vThe adsorptive e?lciencies of the clays, Ex; at the most rapid rate of 5.9 V. M. per cent per minute. It win be seen that for each rate of amples N to V, inclusive are given in Table X. TABLE X Percentage water vapor adsorption at indicated R. H. Example 10 20 40 60 80 N ................... -_ -2.5 1.5 5.3 8.75 126 3. 3 7.5 9.25 10. 0 10. 1 9. 0 9.5 9.8 9. 1 12.0 14.1 15. 2 l4. 7 13. 75 14.6 15.0 12. 7 16.5 19.2 20. 2 Z). 3 I9. 8 20.0 20.25 16. 2 2125 22.9 24. 3 24. 4 23. 8 24.1 24.5 20. 3 25.0 27.5 29. l 29. 15 28. 25 29.1 29.1 drying observed for the clays here reported, and at every value of relative humidity, there is an optimum V. M. content at which the adsorption e?iciency is at a maximum and that this opti mum resides in the range or 5.5 to 7% and more 100 closely in the range 01' 5.5 to 6.5% V. M. The ef fect of departing from this optimum value of the 24.3 10 V. M. content in either direction results in a de 33. 2 crease In the adsorption e?iciency at each value 35.3 39.7 of relative humidity. as is indicated in Table XI. 39. l 41. 79 38. 7 41.6 39.6 In this table the adsorption ei?ciency at the op timum value is given as the range of adsorption 15 e?lciencies obtained for the range of drying rates The drying of Examples E, F, and G, and N there charted. The adsorption e?iciency at the optimum V. M. is compared with the adsorption e?lclency of Example D which had been dried to to V, inclusive, was conducted under approxi a. V. M. content of 5.07% and Example P which mately equilibrium conditions. In Fig. 3 we have charted the V. M. content of the clay at equi 20 had been dried to a V. M. content of 9.75%. librium at various temperatures. It will be ob TABLE XI served that, giving e?’ect to the varying initial Adsorption emc'iency V. M. content of the clay entering the drying process, the drying curves of all these clay samples are closely similar, showing a plateau at 25 about a V. M. content of 5 to 7%, in which region _ R. H Example D At gpth um Example P the change in V. M. is substantially flat begin ning at about 6% V. M. The knee of the curve as it enters the plateau curves is in the region of 6 to 7%. It is apparent that the loss of water in 30 the region down to about 7% is of a di?’erent nature from that occurring in the region of 6% downward. This type of dehydration curve is well known to physica1 chemists and it would in 10 20 40 60 80 100 3. 15 7. 8 15. 3 20. 7 25 34 9. 2 to 10. 8 14 50 15. 2 19. 6 to 20.3 22. 9 to 24. 3 E. 5 f0 29. 2 36. 2 to 4D 7. 5 12 16. 5 20. 2 25 36. 3 It will be observed that the drying of the clay dicate that the montmorillonite crystal is losing 35 to a V. M. lower than about 5.5 to 6.5% has a sub stantial deleterious e?ect upon the adsorption emciencies, particularly at the lower values of the relative humidity. As was said before, it is be lieved that this results from the fact that in going its hydroxyl water of constitution when it is dried below about 6%. While we do not wish to be bound by any theory of the action of our dry ing process or of the adsorbent thus produced, it is important to note that this region of 6 to 7% 40 to a V. M. value below the optimum, we are enter ing the plateau region of Fig. 3 where losses of corresponds with the region of optimum adsorp water of constitution occur and changes in crystal tion which we have discovered and as will be later structure result in a destruction of surface ac described. It appears that the clay, if it is heat tivity. ‘ed beyond a V. M. content of about 6 to 7 %_, re While the specific values of the optimum here sults in a. modi?cation of the crystal structure 45 set forth may vary somewhat with clays from of the montmorillonite resulting in some loss of various beds, depending, it is believed, on the hydroxyl water. It is" of some interest to note amount of impurities, it will be found that by - _ that pure montmorillonite having the theoretical structure (OH)4A14SiaO2o has a hydroxyl water applying the principles set forth above the limits content-(i. e. water of constitution) of about 5%. The additional water which results in a V. M. content of 6 to 7% may well be 'the'constitutional ' of optimum V. M. content of sub-bentonite type a small change in V. M. in going below about 6 quality of. the clay, and this will occur‘ at the knee of ‘the curve of dehydration, such as is charted in Fig. 3. The exact V. M. percentage for optimum e?‘iciency may also change with acid 60 treatment of varying degree and intensity em of montmorillonite clays for the production of optimum adsorption e?iciencies at each‘ value of relative humidity may be readily determined. It water of the impurities associated with the mont is believed that the optimum will lie in the .region morillonite in the Cheto clay.‘ _ The sharp decrease in e?lcie'ncy obtained by 55 of about 5 to 8%, depending upon the purity and to 7%, as hereinafter set forth, supports the view that the loss of hydroxyl water has a large dele-l terious effect on the surface, activity and adsorp tive properties of the clay. ' Figs. 4 and 4a chart the eilect of V. M. con- , ployed in acid activation of the clay. This value tent produced at the various rates of drying of the optimum ef?ciency may then be deter- - mined by plotting a chart like Fig. 3 or'Fig; 4, by observing the change in water content at various ous values of R. H. In these curves the V., M. is the terminal V. M. of the cooled clay of Ex 65 temperatures and times during drying, and also by observing the per cent adsorption at a number amples A to V, inclusive, and the per cent ad sorption at each R. H. is that here reported for of relative humidities su?icient in number to establish the curves in Fig. 4 for clays of varying such clay. At each R. H. ‘there is. a series of upon the adsorption efficiency of the clay at vari curves. As will be more fully described below, terminal V. M. content. ' they represent the adsorption e?iciency of the 70 As has been indicated above and as will appear I Examples E, F, G, and N to V, inclusive, dried . from Fig.‘ 4, the magnitude of the adsorption at the slowest rate of about'0.47 V. M. per cent ,e?iciency at all values of relative humidity de per minute. The Examples H, I, and J are dried at the intermediate rate of 1.59 V. M. per cent pends on the rate at which the water has been removed as well as upon the V. M. content of the. per minute, and Examples‘ K, L. and M are dried 75 0153? at the terminus or the diving period. As will 2,409,203 11 . 12 as shown in Figs. 4 and 5 from which may be be observed from Fig. 4, at each value of the V. M. determined the specific values of ‘the drying rate content and at each value of relative humidity, necessary to develop the desired adsorption eiii_ the adsorption humidity of they clay is higher the cienoies at the various relative humidities to slower the rate at which the clay has been dried. The e?ect of the drying rate upon the adsorp 5 which the clay is to be exposed, and by the appli cation of these principles to .determine the tion e?iciency at each value of relative humidity optimum conditions for the drying process to is also shown in Fig. 5. In Fig. 5 is charted the obtain the most desirable clay of highest adsorp adsorption e?iciency of the clay at each value of tion e?iciency under the conditions under which ‘the relative humidity taken at optimum V. M. content of the clay at such humidity, to wit, in 10 it is to be employed. It will be found that the slower the drying rate the higher the adsorption the region of 5.5 to 6%% against the rate of loss eillciency at all values of the V. M. content in of water. Thisrate of loss is taken as the quotient the range of about 5 to 8% at all values of rela of the loss of water ‘from the beginning of the tive humidity, and that for each drying rate drying process down to the time when the clay has a V. M. content of 7% divided by the time 15 there is an optimum V. M. content within the range of about 5 to 8% and more particularly interval for such loss. This is termed the average within the range of 5.5 to 7% where the optimum rate of V. M. loss or the average drying rate. The V. M. is to be obtained. By choosing the lowest terminal value of 7% was chosen as representing practicable drying rate and by choosing the the break-point or knee of the drying curves shown in Figs. 1 and 2. At this point there is a 20 optimum V. M. content, the highest emciency ad sorbent and desiccant can be obtained by the large change in the rate of V. M. loss. Some other application of the principles herein set forth. value close to 7% could be chosen as, for instance, In order to obtain the best results in the pro 6 to (ii/2%, with a small change in the results duction of desiccants and adsorbents of highest attained. In the region of 8% V. M. and lower, major damage may vresult by overdrying. We 25 adsorption e?lciency, we prefer to employ drying apparatus which will expose the clay to the desire to control the drying in that region to optimum temperature for production of the obtain a removal of water at a rate suf?ciently optimum V. M., and will attain the optimum low to give an optimum adsorption efficiency of V. M. at the optimum drying rate in such man high value. The average drying rate in this region is also included within the scope of the 30 ner as to subject the clay uniformly to such tem peratures and drying rates. We have found it term "average rate.” ' desirable to avoid contact of the clay with gases It will be observed that the increase in the dry of excessively high temperature or with hot metal ing rate affects deleteriously the adsorption em surfaces under such conditions as might lead to ciency of the clay dried to its optimum V. M. content. As is evidenced by the curves of Fig. 5, 35 partial overdrying of the clay or result in ex posure of the clay to such high temperature as this e?ect is most pronounced at the relative will produce excessive drying rates. In other humidities of 10 and at 60 to ‘100. The clay is words, while the average temperature conditions most sensitive to drying rate at such R. ‘H. values. may nominally be within the range adequate, if The clay is less sensitive in the region of 20 to, 40 R. H. In fact, a more thansthreefold increase 40 the clay be uniformly heated, it is still possible, by reason of local hot spots and non-uniform in going from a drying rate of 0.47% V. M. per distribution of temperature throughout the clay, minute to 1.59% V. M. per minute has but a small for one portion of the clay to be overdried and e?ect on the adsorption emciency at-2O or 40 another portion be underdried. In like manner R. H. However, at 10 R. H. and at 60 and 100 it is possible, under such conditions, that part R. H. major damage is done in going from 0.47% ofthe clay is subjected to an excessive drying V. M. to about 1.59 per minute. The decrease in rate while other parts of the clay are subjected e?lciency in going to the higher drying rate of to much lower drying rates. Such a clay when' 5.97% per minute is only moderate. IIThis in produced will be a mixture of good and poor ad dicates that the degree of control of the heating rate which is necessary for the processing of this 50 sorbent with an average value much lower than that which could be obtained had the clay been clay will depend upon what level of the humidity dried more uniformly at the optimum condi we desire to use for the clay. If we desire to tions. operate at 30 R. H. or in the region of 20 to 40 We have found both rotary kilns and tunnel R. H., we may dry the clay at a drying rate rang- ‘ ing up to about 4% V. M. per minute without 55 type driers satisfactory drying equipment, pro vided that they be of su?icient capacity to permit materially a?ecting the adsorption e?lciency of drying to take place gradually, as described above. the clay when it is dried to the optimum range ' It is desirable that such drying equipment be of V. M. However, if we desire that the clay be of suf?cient size so that the drying load is satis employed in adsorption in the .region of 10 R. H. . or in the region of 60 to 100 R. H.,v we will hold 60 ?ed by using gases at relative low temperature. We have found it useful to carry on our drying the drying rates to a closer control in the region operation in a plurality of stages by employing a under about 2% V. M. per minute, and we will plurality of driers or tunnel type driers in which get a greater enhancement if we hold the rate ,. a plurality of zones of drying at diilercnt tem under about 1% V. M. per minute. While the speci?c value of the drying rate to 65 perature and di?erent humidity conditions can produce the effects herein set forth will change with clays from different beds, the general prin- ' ciples of the eifect of thedrying rate apply. ; Those skilled in the art will know how to repeat _. the tests here set forth to obtain the ‘specific? 70 effects of the various drying rates in producing? clays with varying -V. M. within the range of about 5 to 8% and determine the adsorption emciencies of the clay at each R. H. and V. M. content and thus obtain a series ‘of curves such 76 be maintained. The ?rst stage of such plural stage drying process may be operated at a low temperature level. For. instance, in going down to about 8 or 10% V. M. to 20% V. M., the clay may be heated to a temperature of about 250 to 300° F. at a drying rate of about 4% V. M. per minute down to .5% V. M. or less per minute. In the second stage the clay is dried to about 5.5 to 7% V. M. at a drying rate of about 4% V. M. per minute or less ‘at a temperature 'of 300 to 450' 1!‘. - 2,409,203 13 14 In this second stage the drying rate may be I controlled to a low rate so that the clay while drying down through the region where V. M. ‘ ~ It will be seen that 'by drying to approxi mately 6% V. M. we are able to produce a particle break-down of 1.2 % and less through the 30 mesh screen and .64% and less through the 45 mesh change may cause a substantial damage in ad sorption e?iciency, the drying rate may be closely controlled. By this separation of the drying ' screen. By a proper sizing of the'material we can produce particles having as low as about .72% through the 30 mesh screen and .45% through the 45 mesh screen. Such material, experience process in the two stages, the load imposed on each drier is reduced, thus permitting more ac curate control of the drying rate and tempera has shown, will be non-dusting under ordinary tures. This permits of ya more accurate control 10 mechanical handling and will‘ not sift through of the V. M. and drying rate in the second stage bagging used for containing the material em to produce the optimum V. M. ployed for packages in which it is used for con We have also discovered that by such proce trolling the humidity for the purpose of avoid dure, not only will clay of high adsorption ef? ing damage to the material packaged. A low ciency be attained, but the clay will also have a 15 friability is also of importance in installations high hardness value such that it will not dust requiring the use of the granular desiccant in under ordinary mechanical handling, as when large beds through which the ?uid requiring dry the clay is used as a desiccant in packaging ing is to be passed. Resistance of the granules various materials. Because of its high hardness to crushing and packing permits maintenance of value, the dry clay will not dust and pass through 20 minimum resistance to ‘gas ?ow in such a bed, the meshes of the bags in which the desiccant is and also minimizes mechanical loss of desiccant. contained. As an example of the nature of the‘ It is entirely unexpected that large granules of hardness values obtained when the desiccant is the natural clay, of particle size such as shown produced in accordance with the principles of in types a to d, above, would exhibit so great a our invention as herein set forth, the following 25 degree of physical coherence when dried under may be taken as illustrative examples: conditions dictated by the desire of developing Example -1.—The clay was dried according to maximum adsorptivity or “porosity.” As a mat our process to a ?nal V. M. content of 5.9%. ter of fact, however, the desiccant produced in When subjected to the hardness test previously accordance with the present invention is at once described, .72% of the clay passed through a 30 30 superior to synthetic commercial desiccants such mesh screen and .45% passed through a 45 mesh as silica gel or activated alumina, both in terms of physical hardness and in adsorptivity at low _ relative humidities. screen. Example 2.—The clay was dried to a ?nal V. M. content of 5.8% and its hardness values were 1% It is to be understood that the foregoing de through a 30 mesh screen and .53% through a, 45 scription of embodiments of our invention is for purposes of illustration, and modi?cations may be made therein without departing from the spirit mesh screen. Example 3.—'I'he clay was dried to a ?nal V. M. content of 6% and then ground and graded into of the appended claims. the following grades or types: type a; type b; type c; and type d. The screen analyses of the various types were as follows: i} Type Mesh size (screen number). Type +3=0% a """""" " -—6+8==33. 7% Mesh size (screen number) #6=0. 0% -3+6=65. 9% 45 bentonite clay which comprises reducing the V. M. of said clay to a V. M. of about 5.5 to 7%‘ by -—6+12==62. 1% b__-...____._.. --8+ =0. 5% ‘ We claim: 1. A desiccant consisting essentially of a native montmorillonite acid activatable sub-bentonite clay having a V. M. of about 5.5 to 7%. .2. The method of producing a desiccant from native montmorillonite acid activatable sub —l2+18=37.8% heating to a temperature insu?icient to modify the crystal structure of said clay. the reduction —18+20==0. 1% —20=9. 2% of‘V. M. being e?‘ected at an average rate 29 50 greater than about 1% V. M. per minute. Type Mesh size (screen number) Type 3. The method of producing a desiccant from Mesh size (screen number) native montmorillonite acid activatablev sub bentonite clay which comprises reducing the V. M. +2=0.0% ' +2=-o.o% of said clay by heating to a V. M. of about 8% 55 and further reducing the V. M. by heating to —-45==0.0% -—45=2.5% within the range of about 5.5 to 7% V. M. at a rate of less than about 0.5% V. M. per minute, The index "minus” indicates through the all of said heating being at a temperature insuf screen, and the index “plus" indicates retained on ?cient to modify the crystal structure of said clay, the screen. The several types were subjected to to produce a desiccant having an adsorption e?l the hardness test previously described with the ciency of at least about 17% at 30% relative following results: ' , c __________ . _ +45=99.2% . d ............ . . +45==97.4% - humidity. 65 Particle strength percent falling through: Screen #30 ......................... -. . 70 l. 2 0. 93 0.87 Screen #46 ......................... .. 0.40 0. 47 0. 53 0. 64 FREDERICK J. EWING. ROGER A. LOVE-TI‘.