ABSTRACT THE MEASUREMENT OF HEATS OP REACTIONS OF SOIL COMPONENTS WITH A DIFFERENTIAL THERMAL ANALYSIS APPARATUS by Boyd Gene Ellis. The differential thermal analysis apparatus was used to measure heats of desorption of NH3, methylamine, and ethylamine from clay minerals. A theoretical discussion of the nature of heat flow within a differential thermal analysis sample holder was presented. The conclusions drawn from the theoretical discussion were related to the calibration curve of the heat of reaction versus area under the differential thermal curve made using nine different salts of known heats of inversion or fusion. Two methods of determination of heats of reaction by use of differential thermal analysis were compared. The heats of decom- position of magnesite and dehydroxylation of kaolinite determined. by use of the Clausius-Clapeyron equation and the standard curve prepared by using pure salts deviated less than four per cent. It was shown that the exothermic reaction observed upon release of N83 from bentonite clay is due to oxidation of the NH3 catalyzed by the platinum thermocouples. A method was perfected using nitrogen gas to eliminate oxidation during analysis. Losses of both water and NH3, methylamine, or ethylamine occurred between 200 and 550 degrees centigrade. It was postulated that the water loss between 200 and 450 degrees centigrade evolves from decomposition Boyd Gene Ellis of Al(0H)3. The average heat of desorption for the simultaneous loss of one mole of NH3 and one mole of water between the temper- atures of 200 and 450 degrees centigrade was 35.3 kilocalories. The average heat of desorption for methylamine under similar conditions was found to be 40.9 kilocalories. The desorption of ethylamine was accompanied by losses of large quantities of water; consequently, no heats of desorption could be directly connected with loss of ethylamine. Studies conducted with formic acid, acetic acid, methanol, and ethanol saturated bentonite indicated that small quantities are retained by the bentonite, but that the bonding energies are not measurable by differential thermal analysis. Accurate quantitative studies were not possible with the vermiculite used because the thermal properties of the sample changed upon heating. It was shown that the method of eliminating oxidation with nitrogen gas during differential thermal analysis could be applied to soils high in organic matter yielding differential thermal analyses curves characteristic of the mineral components of the soils. THE MEASUREMENT OF HEATS OF REACTIONS OF SOIL COMPONENTS WITH A DIFFERENTIAL THERMAL ANALYSIS APPARATUS By Boyd Gene Ellis A THESIS Submitted to Michigan State university in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Soil Science 1961 L- /-; ,‘u ’4’) 1":‘:’/;J‘. ‘. ' I , I // / a) I' ,r » _.u fr ACKNOWLEDGMENTS This author wishes to express his sincere gratitude to Dr. Max M. Mortland for his guidance, suggestions, and especially for his willingness to discuss any and all points of interest which arose during the research conducted in this thesis and other research work encountered during the course of study at Michigan State university. His interest and enthusiasm in fundamental research has been an inspiration. 4 Special thanks goes to my wife, Julia, for the many hours she spent tracing differential thermal analyses curves and the tedious task of cutting out the areas enclosed by the curves. This author would also like to thank the members of his guidance committee for their helpful suggestions concerning the course work program and research program. ii TABLE OF CONTENTS CHAPTER I. II. III. IV. V. VI. VII. INTRODUCTION . . . . . . . . . . . . . . . . . . . . MEASUREMENT OF ENTHALFY BY DIFFERENTIAL THERMAL ANALYSIS . . . . . . . . . . . . . . . . . . . . . Application of Clausius-Clapeyron Equation . . . . CLAY CHARACTERIZATION AND PREPARATION. . . . . . . . wyoming Bentonite. . . . . . . . . . . . . . . . . Vermiculite. . . . . . . . . . . . . . . . . . . . DIFFERENTIAL THERMAL STUDIES OF WYOMING BENTONITE. . Importance of Ammonia Reactions in Soils . . . . . Differential Thermal Curves of Ammonia Saturated Montmorillonite. . . . . . . . . . . . . . . . . Differential Thermal Analysis of Methylamine and Ethylamine Saturated Bentonite . . . . . . . . . Differential Thermal Analysis of Organic Acid and Alcohol Saturated Bentonite. General Discussion of Heat of Reaction Data. . . . DIFFERENTIAL THERMAL STUDIES OF VERMICULITE. . . . . DIFFERENTIAL THERMAL ANALYSIS OF SOILS HIGH IN ORGANICMATTER................... SUMMARY AND CONCLUSIONS. . . . . . . . . . . . . . . LITERATURE CITED. 0 O O O I O O O O O O O O O O O 0 O 0 I ~ 0 PAGE 16 26 26 35 39 39 39 S9 65 68 70 81 87 9O LIST OF TABLES TABLE PAGE I. Data for Materials Used in Calibration of Differential Thermal Analysis Apparatus . . . . . . . 14 II. The Oxidation of NH3 Released from Bentonite Clay . . 43 III. The Heat of Reaction of NH3 Released from NHAQBentoniteoooooooooooooooooooo 48 IV. The Total Weight, Water, and N Lost from Bentonite which Had Been Hydrogen Saturated by a Resin Column and Then Neutralized with NHQOH . . . . . . . . . . . 49 V. A Comparison of the Weight Loss during Dehydroxylation of a Na-Bentonite and an NH4-Bentonite. . . . . . . . 53 VI. The Total Weight, Water, and NH3 Lost from Bentonite which Had Been Hydrogen Saturated by Electrodialysis and Then Neutralized with NHQOH . . . . . . . . . . . 54 VII. The Total Weight, Water, and NH3 Lost from Aluminum Saturated Bentonite which Had Been Neutralized with NMOH O O O O O O O O O O O O O O I O O O O O O O O O 55 VIII. The Heats of Reaction of Methylamine and Ethylamine Released from Bentonite . . . . . . . . . . . . . . . 59 IX. The Total Weight, water, and Methylamine Lost from Methylamine Saturated Bentonite . . . . . . . . . . . 63 X. The Total Weight, water, and Ethylamine Lost from Ethylamine Saturated Bentonite. . . . . . . . . . . . 64 XI. Estimation of Errors in Determination of Heats of Reaction. . . . . . . . . . . . . . . . . . . . . . . 69 XII. The Heat of Reaction of NH3 Released from NH3 Saturated Vermiculite . . . . . . . . . . . . . . . . 73 XIII. The Total Weight, water, and NH3 Lost from NH3 Saturated vermiculite . . . . . . . . . . . . . . . . 74 XIV. The Heat of Reaction of Methylamine Released from Methylamine Saturated Vermiculite . . . . . . . . . . 77 iv LIST OF TABLES (continued) TABLE PAGE XV. The Heat of Reaction of Ethylamine Released from Ethylamine Saturated Vermiculite . . . . . . . . 78 XVI. The Total Weight, Water, and Methylamine Lost from Methylamine Saturated Vermiculite . . . . . . . . . . 79 XVII. The Total Weight, water, and Ethylamine Lost from Ethylamine Saturated Vermiculite. . . . . . . . . . . 80 FIGURE 10. ll. 12. l3. 14. 15. LIST OF FIGURES Diagram of the differential thermal analysis sample holder; twice actual size . . . . . . . . . Cross sectional diagram of tools used for packing the sample in the sample holder. . . . . . . . . . Differential thermal analysis curves of materials used for heat of reaction calibration. . . . . . Calibration curve for heat of reaction versus area under the differential thermal analysis curve. . . Differential thermal analyses curves of magnesite and kaolinite. . . . . . . . . . . . . . . . . . Relationship between decomposition pressure and temperature of magnesite and kaolinite . . . . . . X-ray diffraction patterns for bentonite from Upton, wyoming . . . . . . . . . . . . . . . . . . Titration curves of bentonite hydrogen saturated by use of a hydrogen saturated resin . . . . . . . Titration curves of bentonite hydrogen saturated by electrodialysis . . . . . . . . . . . . . . . . Titration curves of aluminum saturated bentonite . X-ray diffraction patterns for less than two micron vermiculite . . . . . . . . . . . . . . . . Titration curves of hydrogen saturated vermiculite . The effect of various treatments on the oxidation of NH3 during differential thermal analysis; one- -sixteenth actual size. . . . . . . . Differential thermal analyses curves of NH saturated bentonite; one-sixteenth actual size . . Comparison of the loss of NH3 and water from bentonite during heating . . . . . . . . . . . . . vi PAGE ' 14 17 18 21 23 28 3O 31 33 36 38 41 47 SO FIGURE 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. LIST OF FIGURES (continued) Differential thermal analyses curves of Na-, NH4-, and Hvbentonite; one-sixteenth actual size . . Differential thermal analyses curves of Al+++and NHa-Al saturated bentonite; one~sixteenth actual size. Differential thermal analyses curves of methylamine saturated bentonite; one-sixteenth actual size . . . . Differential thermal analyses curves of ethylamine saturated bentonite; one-sixteenth actual size . . . . Differential thermal analyses curves of methanol and ethanol saturated bentonite; one-sixteenth actual Size 0 0 O 0 O 0 9 O 0 0 O O O O O O O O O 9 C O 0 O 0 Differential thermal analyses curves of formic and acetic acid saturated bentonite; one-sixteenth aCtLAal. Size. 0 O O O O C O O O O O O O O O O O O 0 0 0 Differential thermal analyses curves of NH3 saturated vermiculite; one~sixteenth actual size . . . . . . . . Differential thermal analyses curves of methylamine saturated vermiculite; one-sixteenth actual size . . . Differential thermal analyses curves of ethylamine saturated vermiculite; one-sixteenth actual size . . . Differential thermal analyses curves of Blue Lake sand 0 O O O O O O O O O O O O O O O O O 0 0 O 0 O O 0 Differential thermal analyses curves of the less than two micron fraction of Blue Lake sand. . . . . . . . . Differential thermal analyses curves of the less than two micron fraction of Blue Lake sand after treatment With H202. o o o o o o a o a o o o a a a o o o o o o 0 vii PAGE 52 58 6O 61 66 67 72 75 76 83 84 85 CHAPTER I INTRODUCTION One of the major aims of science in fundamental research is to isolate, identify, and understand the individual steps occurring in a reaction. Soil science presents a special problem in conduct- ing fundamental research. Soils, as they occur in nature, are very complex. Thus, a variety of soil components may be selected for study. The more inert components are generally found in the sand and silt fractions of the soil. As the particle size is reduced, the specific surface becomes much greater, thereby yielding more highly reactive particles. Mortland and Kemper (20) have suggested that such important properties as water retention and cation exchange capacity are highly correlated with surface area of soils. Since the more reactive portion of the soil is found in the clay and organic fractions, it appears that these should be of primary interest to the soil scientist. HDwever, in the study of these fractions many complex situations arise. Soil organic matter which is resistant to further decay, called humus, is far from consistent in chemical composition or reaction. For example, McGeorge (l7) and Turner (34) have reported values for the cation exchange capacity of soil organic matter ranging from 150 to 250 milliequivalents per 100 grams of humus while Olson and Bray (23) working with other 1 2 American soils have reported values ranging from 30 to 280 millie equivalents per 100 grams of humus. According to Van der Marel (35) even kaolinite from pure, well known deposits have heats of tranSw formation which vary from 100 to 176 and from 23 to 43 calories per gram for their endothermal (6000 C) and exotherma. reactions (9800 C) respectively. Thus, to study such a complex system as the soil one must draw on special research tools and many times revert to pure systems to gain insight to the fundamental reactions occurring in various soils. The research tool utilized in this study was differential thermal analysis. Differential thermal analysis has been utilized by investi~ gators studying soil components as early as 1887 when Le Chatelier (l4) employed a simple apparatus to measure thermal reactions taking place when clay materials were heated. In 1899 RobertsaAusten (24) initiated the forerunner of the modern day differential thermal apparatus in which the temperature differences are measured between a material being studied and a reference material. An examination of the theory underlying the electromotive force arising from a thermocouple shows that a net electromotive force will only be developed when the reference material is at a different temperature than the material being studied. Thus, any change in state during heating of the material which is accompanied by a change in heat content will produce an electromotive force. Since many soil components have characteristic reactions during heating, differential thermal analysis has been utilized for many years in qualitative 3 estimation of soil components. An excellent review of the liters ature concerning identification of clay minerals by differential thermal analysis is given by Grim (10). Since the magnitude of the electromotive force produced is proportional to the net difference in temperature between sample and reference material, differential thermal analysis should be a quantitative measure of the heat changes during a run. Following this trend of thought several investigators, Norton (22), Kiyoura and Sata (13), Morray, t al, (21), and Grimshaw and Roberts (11), have attempted to utilize differential thermal analysis as a rapid, inexpensive, and accurate method for quantitative determinations of clay minerals. Considerable difference of Opinion has been expressed about the reliability of differential thermal analysis for quanti~ tative determinations. Speil (28) found deviations of about 30 per cent in heats of transformation of kaolinite from various locations. The limitations for quantitative determinations of clay minerals by differential thermal analysis has been thoroughly discussed by van der Marel (35). From his discussion it appears that the basic limitation is with the variations in the naturally occurring clay minerals and not with the instrument. However, since the measure- ments made by differential thermal analysis are of heats of reaction occurring when a sample is heated at a constant rate, several factors inherent with the instrument tend to limit its use as a quantitative instrument. According to Barshad (3) the peak area is affected by rate of heating, nature of the sample holder, size of the holes in the sample holder, nature of the thermocouple, and sensitivity of the galvanometer. 4 Two methods of measuringllH, the change in enthalpy, by use of differential thermal analysis have been suggested in the liter: ature. The dependence of the peak area in differential thermal analyses curves on the heat of reaction of the sample has been derived theoretically by several investigators (28), (29), (12), (7), (8), (9). This method has been applied by Barshad (3) in measurement of heat of inversion and melting of several pure compounds and also to dehydration of certain clay minerals. On the other hand, Stone (31), utilizing a variable pressure apparatus, has applied the Clausius~C1apeyron equation for determination of heats of dehydration and decomposition of clay minerals and magnesite. Q The object of this thesis is to (l) examine the differential thermal analysis apparatus as an instrument for the determinations of heat of reactioninvolving clay minerals, (2) compare two methods of determining the heat of reaction utilizing differential thermal analysis, and (3) evaluate the heats of reaction of ammonia and simple organic materials in combination with pure clay minerals. It is hOped that a study of pure systems will yield a greater insight to studies made on soil systems with differential thermal analysis and to bonding energies of various clay-ammonia and clay-organic combinations. CHAPTER II MEASUREMENT OF ENTHALPY BY DIFFERENTIAL THERMAL ANALYSIS For a complete understanding of measurement and interpretation of any thermodynamic quantity the basic principles underlying its origin must be examined. Enthalpy, sometimes called heat content, is defined as H = E + PV (1) where E is internal energy, P is pressure, and V is the volume of the system. It is worthy to note that enthalpy is a property of thefl state of the system since each of the right hand terms of the defining equation is a state function. Since by classical thermo- dynamics, E, the internal energy of a system, may not be measured in absolute magnitude, enthalpy also may not be measured in absolute magnitude and only measurements of changes in enthalpy, generally referred to as AH, may be made. The relationship between change in enthalpy and heat evolved during a reaction may be obtained as given below. By the definition given in equation a),H.= E 4- PV. Differentiation yields dH=dE+Pdv+Vd P. ‘2’ But from the first law of thermodynamics dE = HQ - 6W; therefore, dH=JQ-Jw+Pd\/ +VdP. ‘3) 6 Equation (3) may be simplified greatly if measurements are restricted to systems at constant applied pressure in which there is only pressureevolume work being accomplished. In this restricted system VdP 2 O and dW PdV, thus yielding (d H)p (JQ)P- (4) Therefore, a measure of the heat changes involved in a reaction ll occurring at constant pressure with only pressureavolume work will yield the change in heat content directly. With a differential thermal analysis apparatus the measurement of the heat changes during a reaction is complicated by the fact that heat is being added to the system and the temperature of the system is being increased during a determination. The problem of obtaining the heat of reaction from differential thermal analysis data was approached from a theoretical point of view by Speil (28). Kerr and. Kulp (12) made certain modifications and corrections of the original derivation by Speil. Eriksson (7) suggested that both the derivation and modifications are too simplified to be useful. One approach utilized by Eriksson (9) has been adapted to the boundary conditions of the particular differential thermal apparatus used in this study and is given below. Considering the reference space, see figure 1, the differ- ential equation for heat flow within that space may be given as —’ ~ K 00;. at (s) where A(t) is the external heat being added which is a function of follows: A (t) + VZTr = J— (YT-Y ' 2 time, R is the heat conductivity,‘7 is the Laplacian operator, Tr is Z. the temperature of the reference material,6(y,is the diffusivity, and Porous disk A1203 Sample Differential thermocouple Inlets for dynamic gas Cross Sectional View Temperature thermocouple W \ 0 LG kj/ Top View \ Pt(902), Rd(101) l/ Pt ,3 Figure 1. Diagram of the differential thermal analysis sample holder; twice actual size. 8 t is time. The differential equation for the sample space will be similar except a term must be added to account for the production or absorption of heat by the sample. Therefore, the following differ- ential equation is written for the sample space a A(t) B T 9T l._}€__ -+ .E?1:A 5) g):]+ ‘57 U.F; 0(2; :SREZT (6) is the heat produced or absorbed by the sample. If equation (5) is subtracted from equation (6) and W is substituted for T8 - Tr, the result is ,r V... T31?) i253} __ lair- VZKW‘L Ms” 1 14:31; .6,- at - <7> By multiplying both sides of equation (7) by dt and integrating between the limits of t1 and t2, where t1 is some time before any reaction has occurred in the sample space and t2 is a time after a reaction has occurred in the sample space and T8 again equals Tr, the equation obtained is 't 9.33 a? [tn/at r T. I f (it + [ta Ldt + + %:éigfoz-:) {3(8) ex; 32‘ But the bounded area, 5, recorded is S-—-- K’E Wdtj (9) therefore, equation (9) may be rewritten to yield 1- -T- 'tz / 7— g S r J S K .05— 4‘»? t3; 0 (10) In many cases the diffusivity of the sample may approach the diffu- sivity of the reference material. For example, if the sample to be run is diluted with the reference material and similar packing L utilized in both the reference and the sample space, C(E should 2 approach 0(y. . Evidence will be given later which indicates that 9 2. Z. L_ Z. with bentonite and A1203 cK$.approaches 0(r. If Gnau momzflmcm amEuosu flofiucouomwwa .m snowflm com com 00m 000 com coo com com ooH uo.anH _ _ _ _ _ _ _ _ _ macaw emu. Nmmom «Hog ll ‘11 _ — )P‘IllllL/\ILI“ 1— ||— ) 0H mamas «mm. .ea.- .Hose ::::: _ _ 111 .N L( T masts emu. .Nm.~m .eOmNM 1--- _ _ _ _ 1111111 .m w . a a _ P — _ — 1} . mama mmw NooH uuuoso 11:11 a QEQHW “MN. ANNodq .mozx I|1 — — |I_1—l ‘/ on mass» omN. .NN.H~ .mozmz 11 .o waste new. .No.mm .mozw< -- - .N macaw woN. .NN.~H .Hoqxz 111, _ .11I/ .m mamcm mmfi. .Nm.mm .ocoucoeouowewe45 :- .s 18 2.0 1.81. 1.. s/ l. Calories -o-——«®-— Temp. of reaction.(ZOOOC _{}_.__49_~ Temp. of reaction )20006 1» 1 1 1 1 1 1 1 0 20 40 60 80 100 Weight of Area (mg) Figure 4. Calibration curve for heat of reaction versus area under the differential thermal analysis curve. 19 of the sample, thereby producing a flowing or dynamic gas. Thus, it is pzssible to maintain the gas surrounding the sample at a cone stant compasition during the run despite the decomposition products produced by the sample during the run. A complete description of this arrangement is given by Stone (32). Stone (30) has suggested that this instrument is very useful for determining1fiH of decom- position by use of the Clausius-Clapeyron equation. Before using the Clausius-Clapeyron equation an examination should be made of its origin to determine its applicability and restrictions. The following relationship from the Maxwell equations may be utilized: 011:3...Ssdt—1-VsdP (16) where FS is free energy of the solid or liquid, SS is entropy of the solid or liquid, t is temperature, VS is volume of solid or liquid, and P is pressure; aFaz —83 all} +V33P (17) where F8 is free energy of the gas, S8 is entropy of the gas, and V8 is equal to the volume of the gas. It must be noted, however, that equations (16) and (17) are restricted to a system to which this Maxwell equation applies; namely, Closed system of fixed mass, Single phase, No reaction, and Pressure-volume work only, J-‘UNH or 1. Closed system, 2. Pressure-volume work only, and 3. Reversible reaction. For a reversible phase transition it is required that dF 3 ng. Therefore, 'Sfidt +V33P=‘Ssdt +V‘Sdp' (1.8) 20 Assuming an ideal gas where V8 3 3%1, assuming that V8>>}Vs, and rearranging equation (18) yields (89"33)3t‘ ”RTC‘P- ' (19) -1AH For a reversible phase transitionAS '-' -—.. Therefore, A H £l__ T__ cl P 20 h R TL P ' ( ) Equation (20) may be integrated to obtain fly. P= " ”A‘S'E‘fi—L] 4- C, (21) In summary it may be said that application of the Clausius- Clapeyron equation must be to a system which will meet the following requirements: Closed system, Pressure-volume work only, Reversible phase transition, Ideal gas for the gas phase, and Volume of the gas phase much larger than the volume of the solid or liquid phase. UIJ-‘wNH To apply this to differential thermal analysis the decomposition pro- ducts must be accurately known, the pressure and composition of the gas surrounding the sample carefully controlled, and the temperature at the onset of the reaction determined precisely. Magnesite and kaolinite were used to compare the use ofra calibration curve with the Clausius-Clapeyron equation method of determiningzsfl. These materials were selected because studies had been made on each by Stone using his variable pressure apparatus, but no comparisons were made with other methods. Differential curves for these materials are shown in figure 5. A resistance of 400 ohms was used in the differential circuit in both curves shown in figure 5 to permit all of the endothermic peaks to be recorded. The kaolinite curve was not carried to a sufficiently high temperature to observe the exothermic reaction. 21 com com on: r 636.309. one 9.30:me mo mot/.30 mommaocm. due—Hon... Hmwucouowwa .m shaman com com cos com . oo~ ooh uo .as.a _ e. _ _ _ _ 3833: .H _ _ ouoaaoome .N 22 The decomposition of magnesite under variable 002 pressures was studied by Stone (31), after which he preposed that the reaction was a straightforward decomposition as given below: Mg C03"> M30 ‘1" C02... (22) He found that the plot of lnP versus l--gave two straight lines. C02 T He attributed deviation in the low pressure region to insensitivity of the recorder. He further stated that there appeared to be a break in the curve at 1810 millimeters of mercury, but his deter- minations were not carried out at pressures greater than 2200 millimeters of mercury to determine if this deviation is real. Based on his best straight line he calculated ale of decomposition of 10.1 kilocalories per mole of MgC03. It should be noted that the differential thermal curves shown in Stone's work indicated the presence of an impurity which he attributed to CaCO3. The plot of lnPco versus %.is given in figure 6 for magnesite decomposition obtained with the instrument in this work where the pressure was varied from 104 to 3100 millimeters of mercury. Partial pressures below 100 millimeters of mercury were not attempted because the exact determination of the differential pressure is not possible with the instrument used. In this region differential pressure becomes large in comparision with the total pressure and introduces rather large errors in measurement of pressures below 100 milli- meters of mercury. As can be seen from figure 6 the data very nearlyfits a linear plot even in the low pressure regions. The data deviates from a linear plot in the high pressure regions in a similar manner to data given by Stone. This was not investigated further because examination of the Clausius-Clapeyron equation 23 3 03-~_\O 13r————fiD——- Kaolinite 2,000_ -®-——O— Magnesite ® ‘23 :: €1,000— EL r a: H h- 0 (V ._ c: t) w h— o ‘1”. 3 500... U) m o H n‘ 1—' 100 1 I I I l l 1 I 110 120 130 140 150 l/T x 103 Figure 6. Relationship between decomposition pressure and temperature of magnesite and kaolinite. 24 predicts that at high pressures where a gas will not behave as an ideal gas and at pressures sufficiently high to cause the volume of the solid or liquid phase to become appreciable in comparison to the gas phase the curve will deviate from a linear plot in the direction determined experimentally. Utilizing the slope from the plot given in figure 6 a value of 27.0 kilocalories per mole of MgCO3 was obtained for itslflH of decomposition. This value is 16.9 kilocalories per mole greater than that obtained by Stone. The theoreticalzflH may be computed by the following relationship: AH: ZAHProduo+s “-ZAHveactants. (23) This yields a value of 27.8 kilocalories per mole. Theng.of decomposition was also determined by use of the calibration curve discussed previously. The value obtained was 28.3 kilocalories per mole assuming that the material was 100 per cent MgC03. Therefore, the two methods compare quite favorably and also are in good agree- ment with the value calculated theoretically. The rather large disagreement with thesz determined by Stone may be due to impurities in his sample. However, it should be pointed out that the determin- ation utilizing the Clausius-Clapeyron equation is independent of weight of sample used; therefore, impurities should not affect the results unless two reactions occur simultaneously. The second material used for comparison purposes was kaoliw nite from Lamar, NOrth Carolina. It was chosen because kaolinite has been studied before and also presents a decomposition which may not be reversible. The reaction for the dehydroxylation has been written by Stone and Rowland (33) as follows: A1, 81, 05 (011)4701 1,03 +310LCMFI..)+ZH.0. (2.) 25 Values for the [iH of dehydroxylation vary considerably from sample to sample. Van der Marel (35) has reported values from 100 to 176 calories per gram and Stone found values from 140 to 170 calories per gram of clay. The plot of lnPHZO versus % for kaolinite is shown in figure 6. It can be seen that considerable scatter of points was obtained with kaolinite. The exact reason for this scatter of points is not known; however, some difficulty may be expected in maintaining 100 per cent water vapor for the dynamic gas. The best fit for the points was determined by the method of least squares which yielded the following equation: jvu PHLO :_ —10,2x103(—i’;) 4—10.13 (25) Utilizing the lepe from this equation a value of 157 calories per gram was obtained as thezLH of dehydroxylation of kaolinite. A determination of AH by use of the calibration curve gave a value of 154 calories per gram of material. Again the two methods compare favorably. It may be noted that the values obtained are in the expected range as reported by other investigators. From the two materials tested it would appear that either method is satisfactory. However, each system must be examined care~ - fully before selecting the method to be used. For example, to apply the Clausius-Clapeyron equation the reaction must be well defined, produce a gas phase of known composition which may be supplied from an external source, and yield a single,reversib1e reaction. 0n the other hand, the method of area determination may be applied to .any system in which the heat capacities and thermal conductivities change little during a reaction. However, this measurement yields only the net heat evolved or absorbed and may be very difficult to interpret if more than one reaction occurs simultaneously. CHAPTER III CLAY CHARACTERIZATION AND PREPARATION Two clay minerals, bentonite and vermiculite, were used for this study. They were selected because the bentonite will hold large quantities of exchangeable cations with very little fixation, whereas vermiculite will produce considerable fixation of potassium and NH4 ions. Wyoming Bentonite Pure bentonite from Hole 25, Upton, Wyoming, was used as a representative of the montmorillonite clay mineral group. The chemical analysis of this particular sample as recorded in Reference Clay Minerals, American Petroleum Institute Project 49 is given below. Chemical Analysis SiOZ ------------- 57.49% A1203 ------------- 20.27 Fe203 ---------- 2°92 “80 ------------- 3.18 CaO ------------- 0.23 NaZO ------------- 1.32 K20 ------------- 0 28 “20+ ............. 6 85 “20" ------------- 7 63 T102 --------- __Q;L; Total 100.48% 27 The symbols H20“ and H20+ represent water loss below 105 degrees centigrade and above 105 degrees centigrade respectively The formula suggested for this montmorillonite is (A11.55Fe.15”533) (A1.08313.92)°1o(°H32(Na.17£§:.03)- The montmo- rillonite family of clay minerals is characterized by high cation exchange capacity arising principally from substitution of magne~ sium for aluminum in the octhedral layer, high water holding capacity, and swelling properties. X-ray analysis was conducted by depositing a thin layer of calcium saturated, glycerol solvated clay upon a porous ceramic plate so as to orient the clay particle with respect to the basal plane (001). The oriented sample was then rotated with respect to an X-ray beam produced by cOpper radiation. The deflections were recorded with a scanning goniometer, utilizing a GeigereMuller counter tube in conjunction with a scalerwratio meter with an auto- matic recorder. In each case a second and third pattern was made of the same sample potassium saturated and heated to 105 and 550 degrees centigrade respectively. The results of the X-ray analysis for the bentonite sample are shown in figure 7. The sample is nearly pure as evidenced by the very intense 17.7 angstrom peak in the pattern of the calcium saturated, glycerol solvated bentonite. Little evidence of impurities is shown except for a small trace of quartz. Differential thermal analysis also indicated a pure sample, but since Chapter IV will deal with differential thermal analysis in detail, no data will be presented at this time. Both conductimetric and potentiometric titration curves were run as a means of characterizing the clay materials. Since it is a 28 C ass son moi 117A Potassium Saturated, Hound at 550' C Potassium Saturated, Hound at IOS' C “A _‘ w v w' Calcium Saturated, Glycerol Solvotsd 1 1 l l l l J 30 25 20 I5 IO 5 2 Degrees 2 6 Figure 7. X-ray diffraction patterns for bentonite from Upton, Wyoming. 29 well established fact that most clay minerals may contain large quantities of exchangeable aluminum, the bentonite clay was prepared by three different methods to yield different degrees of aluminum saturation. The minimum amount of aluminum saturation was obtained by passing a bentonite clay suspension through hydroxyl saturated Amberlite IR-45 anion exchange resin and then through hydrogen saturated Amberlite IR-120 cation exchange resin. The titration of the clay was started approximately fifteen minutes after the first bentonite had passed through the IR-120 resin and was finished approximately one hour later. The results of titration curves made using both NaOH and NH40H are given in figure 8. Both the NaOH.and the NH40H curves indicate a cation exchange capacity of approximately 90 milliequivalents per 100 grams clay of which approximately 15 milliequivalents per 100 grams clay is exchangeable aluminum. To obtain a bentonite clay which was moderate in the amount of exchangeable aluminum the clay was electrodialyzed to a pH of 3.3. Titration curves for this material are shown in figure 9. The cation exchange capacity of this sample appears to be 74 milliequivalents per 100 grams of clay of which 31 milliequivalents are exchangeable aluminum. Two factors may account for the apparent lowering of the cation exchange capacity. First, electrodialysis may not success- fully remove all exchangeable bases, and secondly, the treatment undoubtedly results in some breakdown of the mineral lattice. An aluminum saturated clay was prepared by passing the clay material through an aluminum saturated IRPIZO exchange resin. The resin was aluminum saturated by leaching with A1C13 solution and then washing with distilled water until the leachate was free from chlorides as pH 30 --o—- - - -3..- Sp . Conductance —@——o—PH \ \ b. ”guano- .31“ a... \ 'o 20 40 60 __ so NH40H added m.e.llOO grams clay _ “0" - — --o-- Sp . Conductance. # '4D-—--1 U ”4 40 ,3 U U a “U 20 g L) 0 0 20 40 6O 80 100 NH40H added m.e.llOO grams clay 10.. ,’ , 140 -@ ----- ®--Sp. Conductance , H" I 'm 8— —0———o—PH ' ‘ A 100 E I O I v I £5 __ I a. I x I rs 6__ I —1 60 3: L / a S *- \ \ I —' 40 1%“ 1° ‘5: ®\ ‘0‘ II o 4_ ‘3 ,5" —-J 20 \ ’ @‘ _®— CL.‘9‘.6PJ9 3 1 l 1 l 1 I 1 l 1 o 0 20 40 60 80 100 NaOH added m.e./100 grams clay Figure 9. Titration curves of bentonite hydrogen saturated by electrodialysis. 32 indicated by AgN03. One difficulty was encountered in that the clay material flocculated upon contact with the exchange resin and was then very difficult to extract from the resin column. The saturation was accomplished by using short columns, frequent recharging with A1Cl3, and slight suction from a water aspirator to draw the clay suspension through the resin column. The titration curves of this material are shown in figure 10. It is evident that the clay material is almost one hundred per cent aluminum saturated. The titration curves show poorer defined breaks in this case, but at pH 7.0 the material should still contain nearly 75 milliequivalents of NH3 per 100 grams of clay. To prepare NH3 saturated clays for differential thermal analysis the clay materials were hydrogen saturated and/or aluminum saturated as described above and then immediately titrated to pH 7.0 with 1N NH40H, stirred overnight with a magnetic stirrer, and then readjusted to pH 7.0. This process was repeated until the equilibrium pH was 7.0. At this time the material was dried under an infrared light, ground in a mortar and pestle, thoroughly mixed, and stored in a desiccator until differential thermal analyses were run. Methylamine saturated samples were prepared by adding 10 milliliters of 40 per cent methylamine in water to the freshly prepared hydrogen saturated clay and shaking for 48 hours. The material was then dried under an infrared light, ground with a mor- tar and pestle, and dried at 60 degrees centigrade under vacuum. During the drying the excess methylamine was volatilized leaving the clay methylamine saturated. 9 _ 33 l 130 8 _ fl- - _ ®-- Sp. Conductance .1 120 _ llO ._ lOO 3 1 l 1 l 1 l 1 I 1 70 0 20 40 60 80 100 10— ,—140 8 _ — ®— — — —®- Sp. Conductance 120 -(}-—--<>- I I 7.. , -110 a. G' / 6 _ , — 100 / 6) 5 / —‘ 90 R. . / \ @ ’ @- rd 4 ,- ’, —-’ _ 80 \ ..0'—49"49"6> ®' <3 @— ”0‘ 3 . l 1 1 1 I 1 1 1 7° NaOH.added m.e.llOO grams clay Figure_10. Titration curves of aluminum saturated bentonite. Conductivity x 106 (ohms°1) Conductivity x 106 (ohms'l) 34 The bentonite clay was saturated with ethylamine by two different procedures. The procedure utilized in the first replication was to hydrogen saturate the bentonite clay with a hydrogen resin column and then to treat the hydrogen saturated clay with ethylamine gas. The ethylamine gas was introduced into a glass tube containing approximately 300 milliliters of the clay suspension through a one-holed rubber stopper fitted with a glass tube and stopcock. Approximately 20 milliliters of liquid anhy- drous ethylamine were vaporized and passed into the clay material. The clay suspension was then transferred to a flask, stOppered, and stirred for 48 hours with a magnetic stirrer. After stirring, the material was dried under an infrared lamp, ground with a mortar and pestle, and further dried at 60 degrees centigrade under vacuum. The second replication of ethylamine saturated bentonite was prepared by adding 20 milliliters of ethylamine previously cooled to zero degrees centigrade to a suspension containing 10 grams bentonite. The mix- ture was stirred for 48 hours with a magnetic stirrer and dried similarly to the first replication. Although it would be expected that organic materials possessing reactive carboxyl or hydroxyl groups would not be held as exchange- able ions to any large extent by clay minerals, it was felt that quantities detectable by differential thermal analysis might be adsorbed. Thus, two organic acids, formic and acetic acid, and two alcohols, methanol and ethanol, were selected for study. To prepare the clay-acid materials, a large excess of formic or acetic acid was added to a suspension of bentonite clay and shaken on a shaker for 48 hours. The excess acid was then removed by dialysis. The 3S clay-alcohol samples were prepared by placing dry bentonite in abSOm lute methanol or ethanol and shaking on a shaker for 48 hours. The excess alcohol was then removed by evaporation. Both the acid» and alcoholaclay samples were dried under an infrared lamp. Vermiculite Zonolite, a vermiculite which may be obtained commercially, was used as a representative from this group of clay minerals. The material was ball-milled for 24 hours to reduce the particle size without destroying the crystal structure. Samples of zono- lite typically contain considerable interlayer potassium which must be removed to obtain a uniform, homogeneous vermiculite sample. The sample of zonolite utilized in this study was leached with approximately 0.5 N NaCl at a slow leaching rate for a period of two months to remove this interlayer potassium. The total potassium remaining at this time was 2.99 per cent. After this leaching pro- cess, the sodium saturated vermiculite was dispersed in a water system and a separation of the less than two micron material made utilizing Stoke's law. Perhaps the best single method of identifying and character- izing vermiculite is_X-ray analysis. The results of this analysis carried out in a similar manner as for bentonite are given in figuregll. The 14.8 angstrom peak, characteristic of a calcium saturated vermiculite, is very intense. However, some indication of interstratification of the mineral is shown by the broadening of the peak as it returns to the base line. The potassium saturated sample heated to 105 degrees centigrade collapsed to 10 angstroms 36 3r‘ moi ltB‘ Poloulum Solwolod, Hosted at 550' C Pdosdum Solwohd, Hum .11 105’ c ' l Calcium Sduolod, Glycerol Solvolod L 1 l l 1 4 5 30 25 20 I5 IO 5 2 Figure 11. micron vermiculite. Door“ 2 0 X-ray diffraction patterns for less than two 37 as would be expected. Although heating to 550 degrees centigrade had little effect on the basal spacing, it did sharpen the peak considerably. Titration curves using NaOH are shown in figure 12. How- ever, for vermiculite this proved to be a poor method of character~ izing the clay mineral. No sharp endpoints are indicated in the curve. Because the titration curves did not yield a good measure of the cation exchange capacity, it was determined by the procedure described by Ellis and Mbrtland (6). The cation exchange capacity as determined by this method is 114 milliequivalents per 100 grams clay. It should be noted that cation exchange capacities deter- mined by this means may be slightly high due to removing sodium ions from the mineral lattice. Samples of vermiculite were prepared saturated with NH3, methylamine, and ethylamine. The method of sample preparation was the same as for bentonite except for ethylamine. For ethylamine the procedure followed for the second replication of bentonite was utilized. It was necessary to pass the vermiculite suspension through the hydrogen resin column many times to obtain a hydrogen saturated clay. Therefore, the vermiculite undoubtedly contained considerable exchangeable aluminum.' 38 -.®. .. __ -9- _ Sp . Conductance I —- 49—0— — a pH 0 9b 4 L I 1 l 1 I I I 0 20 40 6O 80 NaOH added m.e./100 grams clay Figure 12. Titration curves of hydrogen saturated vermiculite. 180 180 160 140 120 100 80 60 40 20 Conductivity x 106 (ohms'l) CHAPTER IV DIFFERENTIAL THERMAL STUDIES OF WYOMING BENTONITE Importance of Ammonia Reactions in Soils Anhydrous NH3 has been used as a nitrogen fertilizer during the past several years. Retention of NH3 by soils has been largely attributed to the formation of NH4 ions and subsequent attraction of the NH4 ions by the cation exchange complex. McDowell and Smith (16) reported that soil texture had a pronounced effect on NH3 movement and retention in soils. They found that the greatest movement occurred in sand and silt loam soils and the least movement in clay soils. Sohn and Peech (27) attributed the capacity of a soil to sorb NH3 to neutralization of exchangeable hydronium and aluminum ions as well as to the formation of organic nitrogen compounds upon autoxidation of soil organic matter and simultaneous NH3 fixation. An extensive review of the importance of NH3 and its reactions in the soil has been given by Mortland (l9). Differential Thermal Curves of Ammonia Saturated Montmorillonite In the past various attempts have been made to relate the temperature at which NH3 is lost from clay minerals with its adsorption site. Cornet (5) suggested that NH4 ions on broken 39 40 bond surfaces were decomposed at temperatures below 125 degrees centigrade while rapid loss of N34 ions from 275 to 400 degrees centigrade was from NH4 ions on planar surfaces. Mortland (19) suggested that ammonia chemically sorbed by clay minerals probably does not exist at discrete energy levels, but that a wide range exists in the energy of chemical sorption of ammonia by clays. t al. (25) found differential thermal analysis curves of Scott, NHZ-bentonite exhibited exothermic peaks near 550 degrees centigrade. They drew the following conclusions: The exothermic peak in the bentonite differential thermal curves occurred in all cases where fixed NH was present. If the NH4 ions were blocked from the fixing sites by prior fixation of K, the curve exhibited no exothermic effects from NHa ions on the bentonite. Apparently any exothermic effect due to the loss of NH3 from the exchange- able NH4 ions is counteracted by a concurrent loss of water. 0n the other hand, the fixed NH4 ions are decomposed and NH lost without a concurrent loss of water, since an exothermic peak occurs. This indicates that the exchange- able NH“ ions in bentonite are closely associated with water molecules while the fixed NRA ions are not. This suggested a system in which the heat of desorption of ammonia could be measured by differential thermal analysis to yield valuable information in differentiating between exchangeable and fixed NH3. Preliminary studies yielded a differential thermal analysis curve of NH3 saturated bentonite similar to that shown as curve 1, figure 13. The appearance of the exothermic peak near 550 degrees centi- grade is very similar to that noted by Scott. However, in trying to predict the reaction which is occurring, great difficulty is encountered. The breakdown of NH“ ions to yield NH3 and subsequent 1A. D. Scott, J. J. Hanway, and C. Stanford, "Thermal Studies of Ammonium Fixation and Release in Certain Clay Minerals," American Mineralogist, 47:718, 1956. 41 .ouam Hoouom nucoouxwmnoco “mamhflono Hosponu Hofluaouowman magnum mzz mo cowuovwxo on» no mucosuoouu moowuo> mo uoowmo one .ma ouowfim oma own . own one o£m owe own oma OQm oo .asme ago—SEE lIllIlllIl ll IF 1_1\/_/l .H .ou«:0ucon1qmz was ow can .oamsoooaponu osu o>ono census ouacouaoa-amz Eoooo> was Nz nouns. can oudcouconuozz ouamuona nova: Nz saga mason 0H coumouuoua ouaaouconuqmz omsmuoua nova: Nz as“: “so: a vouoouuoun ouqaouconuoxz 42 evolution of NH3 should, in fact, yield an endothermic reaction. An estimate of the heat of reaction obtained from heats of formation data is 20.1 kilocalories per mole of NH3 evolved. No plausible explanation could be found for a simple breaking of the clay-NH3 bond or simple dissociation of NH“ ions which would yield an exo- thermic reaction. Barshad (3), however, found similar difficulties with NHaNO used in temperature calibration of the differential 3 thermal analysis apparatus. He noted that large exotherms appeared when NH4N03 was used and suggested that these were due to the oxidation of NH3 catalyzed by the platinum thermocouples. He suggested that this oxidation could be prevented by layering the NH4N03 above the thermocouple instead of below or surrounding the thermocouple. To verify Barshad's suggestion a simple semiquantitative experiment was designed to ascertain if NH. was oxidized by heat 3 and if this reaction is enhanced by a platinum catalyst. Samples of NH3 saturated bentonite were placed in a test tube equipped with a two-holed stOpper arranged so that glass tubing from one hole led into a boric acid trap and glass tubing from the other hole allowed one to blow into the test tube and flush the component gases through the boric acid trap. The test tube containing the NH4- bentonite was heated with a Bunsen burner to a maximum heat and then the gases flushed into the boric acid. The boric acid was subse- quently titrated with standard HCl to determine the quantity of NH3 evolved. This process was repeated with A1203 mixed with the NHa- clay, and with A1203 and a piece of platinum wire present in the test tube. Results of this experiment are given in Table II. 43 Although the results given in Table II are semiquantitative in nature, they effectively show that NH3 is eliminated and presumedly oxidized by heat when a platinum catalyst is present. It would thus appear that the exothermic peak which Scott has attributed to release of fixed NH“ ions is nothing more than the catalytic oxidation of NH3 in a reaction which follows the dissociation of the NH3 from the clay mineral and is in no way related to its position on the clay mineral. TABLE II THE OXIDATION OF NH3 RELEASED FROM BENTONITE CLAY NH3 Recovered NH3 Treatment m.e.llOO gms Clay Recovered {1) none 61 64.0 A1203 Added 65 69.0 A1203 +pc Wire 10 10.8 The problem then reverted to one of eliminating oxidation of NH3 during the heating of the sample in differential thermal analysis. Theoretical considerations would completely rule out application of Barshad's suggestion to layer the sample containing NH3 above the thermocouple since this could produce a nonsymetrical orientation of the sample with respect to the thermocouple and produce a vertical heatflow gradient. Nevertheless, this method was attemped during the course of investigation of methods to eliminate the oxidation of the NH3. The resulting curve is shown as curve 2 in figure 13. It should be noted that the sensitivity of the instrument was increased greatly in order to record the smaller heat changes 44 registered by the thermocouple. The results still showed the exo~ thermic reaction. Since the instrument used in this study is equipped to run a sample under a controlled atmosphere at elevated pressures, atmospheric pressure, or under vacuum, it appeared that the oxidation might be prevented by use of an inert gas such as nitrogen during the run. The first procedure investigated involved evacuation of the system and subsequent introduction of nitrogen gas through the sample and into the chamber. This was repeated three times and the differential thermal analysis made with the sample under vacuum. ’The resulting curve is shown as curve 3 in figure 13. Again the exothermic peak persists. The failure of this method may be attributed to slight leakage in the manifold of the differential thermal apparatus which introduced oxygen into the system and sample while the evacuation process was being utilized. The final method tested, and the one which proved to be effective, was that of evacuating the system and then introducing nitrogen gas under a slight pressure. The nitrogen gas was puri- fied by passing it through a pyrogallic acid trap and a H2804.trap. This procedure was repeated three times and then the sample was allowed to pretreat with nitrogen passing through it at a pressure of slightly greater than one atmosphere. The flowing gas was stopped and the pressure reduced to atmospheric pressure prior to making a differential thermal analysis. Thus, the determination was made under static conditions with no vertical heat flow gradients produced by a flowing gas. various pretreatment times were tested. Two curves of the same sample, but one pretreated for ten hours and the other pretreated for one hour, are shown as curves 4 and 5 respectively in figure 13. No difference could be 45 noted between the two curves; therefore, a one hour pretreatment was utilized for preventing oxidation in the remaining studies con» ducted in this work. Examination of the differential thermal analysis curve, which resulted from an NH3 saturated bentonite clay once oxidation had been eliminated, presented rather large difficulties to quantitative heat of reaction measurements. It was apparent that the loss of NH3 was not homogeneous in nature. Indeed, once the differential thermal curve deviated from the base line near 250 degrees centi- grade due to the loss of NH3, it did not return to the base line until after the completion of dehydroxylation of the mineral near 750 degrees centigrade. Thus, the lack of homogeneity ruled out the application of the Clausius-Clapeyron equation to this system and the intermixing of two reactions coupled with the long departure of the curve from the base line seriously limited the applicability and accuracy of the method of area determinations. It was noted, however, that the curve did return to the base line once dehy- droxylation was complete which would theoretically justify the use of area determinations as a quantitative measure of the heat of reaction. Due to the complexity of the reactions involved in loss of NH3 from bentonite, it was necessary to determine the change in area of the differential thermal analysis curve after a portion of the NH3 had been removed by heating to obtain an estimate of the heat of reaction when NH3 is evolved. 'This was accomplished by placing samples of NH3 saturated bentonite in platinum crucibles, weighing them, and then heating the samples for a period of time 46 to remove a portion of the NH3. The samples were then reweighed to determine the total loss in weight upon heating. Differential thermal analysis was conducted on subsamples before and after heating. Examples of these curves are shown in figure 14. Subsamples were taken and NH determined by a modified semimmicro Kjeidahl process. 3 The determination was accomplished by digesting a 0.1 gram sample of the clay material with concentrated sulfuric acid and a salt mixture of K2804 and CuSOQ. The mixture was digested over a Bunsen burner until the solution became a bright green and then digested for an additional hour. The flask was allowed to cool, a small amount of distilled water added, and then the flask cooled in a water bath. After the flask and contents were cool, a sufficient quantity of solid NaOH was added to make the final digestion mixture basic as evidenced by a dark brown color, and then the NH3 was voltalized by steam distillation. The NH3 was trapped in 40 milliliters of two per cent boric acid and titrated with 0.0196 N HCl to a gray end point with a methyl red, methylene blue indicator. The heats of reaction calculated for the loss of NH3 from bentonite for the temperature range 200 to 450 degrees centigrade are given in Table 111., Two points in Table III are of considerable interest. First, all values obtained for the heats of reaction are considerably higher than the theoretical 20.1 kilocalories per mole for the formation of NH3 from NHa ions. Second, the heat of reaction is larger for the bentonite which had been hydrogen satu- rated by electrodialysis. At this point it must be recalled that the differential thermal analysis apparatus records only the net! heat change occurring during a reaction and thus, for proper 47 .ouwm Hosuom nucoouxamaoco moua:0u:on woumusumm mmz mo mo>u=o mommamam nguonu aowuaouommfia .¢H ou:w«m Gem oom oo~ com com coo com com oo~ oo .QEoH III _ L _ _ _ _ w _ _ haw aw any Juacoucmaoamz l 1 l l I ll I II Ir _ .H Nz ca can ouacoucmn-qxz was :a can .u oo0m ou voodoo unaccuconnomz ~z so can .0 000m Ou oouoon ouwc0ucon1¢=z Mam n« as» .o oo0m ou commoo ouacOuconaomz IllllII.n Nz as .5udo8m 11111! _ _ lb _ cu wouoo: IllH/Il..o unaccuconuomz 48 interpretation of heat of reaction data we must know the reactions which are occurring. TABLE III THE HEAT OF REACTION OF NH3 RELEASED FROM NHa-BENTONITE* Hydrogen Saturated . Hydrogen Saturated Temperature by Resin, by Electrodialysis, Range NH3 Saturated NH3 Saturated** ,.o. c AH (K Cal/mag)“ AH (K Cal/mole) 2004300 35 .3 41 .8 300-400 37.0 41.0 250-350 35.2 52.7 350-450 33.6 59.1 , , *Values for temperature ranges 200-300 and 300-400 are from one sample while values for the temperature ranges 250-350 and 350'- 450 are from another sample preparation. **Expressed as kilocalories per mole of NH3 evolved. Further information concerning the reaction or reactions occurring in this system may be obtained from the weight loss data given in Table IV. It is apparent that the entire weight loss may not be attributed to loss of NH3. The most probable constituent to be lost other than NH3 is water. It is also evident that this loss, presumably water, is continuous. The dry weight of the clay free of water and NH3 used in Table IV was obtained by assuming that NH3 lost above 400 degrees centigrade had the same amount of water associated with it as that NH3 lost between 200 and 400 degrees centigrade. For purposes of illustration the data for NH3 and water lost from bentonite are summarized in figure 15. It should be emphasised that each point above 250 degrees centigrade 49 I: II“..I : NH.0 3.? I. . fig .ouacoucon ooum mmz .oouw nouns mo macaw 00H nod macaw no commokmxms No.0 mm.0 Nm.0 NN.0 mm.H 0a 0mm one mm.0 mw.0 o~.0 No.0 qm.0 mm.0 0H Ono 0mm no.0 mm.0 00.0 No.0 HN.H nw.0 0H 0mm OmN mH.0 N¢.H 00.0 0¢.0 m0.a mm.0 0H 0mm 00 00.0 mn.~ 00.0 00.0 m0.N 00.0 11 Basom> w 00 0H.0 00.0 00.N MH.0 00.0 0N.N OH 000 00m wm.0 0H.0 mm.0 «0.0 ma.0 m€.H 0H 00m 00¢ ¢¢.0 No.0 00.0 ~m.0 nn.0 mm.0 0H 00¢ 00m em.0 ©0.~ 00.0 no.0 0N.H Hw.0 0H 00m 00m 00.0 N¢.H 00.0 0¢.0 mn.H dn.0 0H 00N 00 H0.01 0m.# 00.0 ¢<.« m~.~ m¢.~ 0H Eaaom> w 00 00.09 mo.a 00.0 mm.m m0.m On.m m.~ Bazoo> w 00 No.0 0¢.H 00.0 m©.N No.5 m0.N m.H Easum> Q 00 00.0 Nq.H 00.0 00.0 m0.0H 00.0 11 _ 11 no samewooflmmEmv sAmamooH\mEmvsroHummNm smmswmmfi\mamm sfimwwommwmemv sAmeooH\mawv A.wu;v oo .oth co .man L122 We... M221“. newml..1..mmfi....2 .mlfi .82 .2 2.... .5 :2. 3:22 306:2 :HHS QmNHMH mAmuoo wommedm Mosuonu Hmwucouowmaa .oH ouswwm com com ooh coo oom . ooa oom cow OOH oo .ae.e _ u e u _ _ _ _ _ oufiGOuconumz IIIIII _ _ III I» IIHIHIII .H _ q .IIIIIIIIIIILUI L\//LF/ UufifiOucwnl m2 .N 53 temperature range of 450 to 750 degrees centigrade while dehydroxyo lation occurs, the water loss is identical from both sodium and NBA ion saturated bentonite. This data does not favor MacKenzie’s postulate. TABLE V A COMPARISON OF THE WEIGHT LOSS DURING DEHYDROXYLATION OF A NaoBENTONITE AND AN NHa-BENTONITE Initial Final Nawfientonite NH4sBentonite Temp. Temp. Time Water Loss NH3 Loss H20 Loss (° C) (° 0) (hrs.) (gms/lOOgms) (gms/lQOgms) (gms/lOOgms)~ 200 300 24 0.28 0.45 0.62 350 450 24 0.23 0.38 0.77 450 750 18 3.94 0.59 3.90 Barshad (2), studying vermiculite, suggested that the NH“ ion is closely associated with one water molecule even while on the surface of the clay mineral. He suggested that the energy binding the water and NH4 ions must be large. If this is true, the ratio of water loss to NH3 loss should always remain constant.‘ Tables V1 and VII give the loss of water and NH3 from the intermediate aluminum saturated and the aluminum saturated clay respectively. The ratio of water loss to NH3 loss is greater, approaching two to one in these samples. Thus, it would appear that the water loss may be independent of the NH3 loss. A One possible explanation is that the water evolves from decomposition of Al(OH)3. The titration curves showed that consider- able exchangeable aluminum was present on the clay mineral. When the suspension is neutralized, the aluminum ions may be precipitated 54 .oowcoucoo oouw mmz .mouw noum3 mo macaw ooH you macaw mo commouawms I: IIHH'M I.“ “INIIHIIVHNNHF. L...r I “LIIIIIIIIL1 .iIIIIIF II..I 1 III?! ”I” .1 1Lul J III 1 .1F.I 1 . 1‘ II..HI"I 1.11IIIII.I.II.hI .lIIIIII-II..1d.."1-.I..1..1.I.I.I.J ’II.I1IIIN LII. 1 4 11.1...1I. fiI1IL. ..Nr.I1 IIIIII .I..A.1.1|Ia..'l I I . NREIE 1 .II I. III... .1 1 .IlrIi I mm.0 HH.0 m~.0 no.0 NH.0 mo.H 0H 0mm omo mN.o «0.0 00.0 mm.o 00.0 00.0 CH Ono 0mm mH.0 mm.o 00.0 um.o m¢.H om.0 0H 0mm 0mm 00.0 -.~ 00.0 mo.0 ow.~ mm.0 0H omw 00 00.0 0N.H 00.0 00.0 mm.~ 00.0 11 11 11 mm.o mo.0 mH.N no.0 00.0 oo.m 0H com com mm.o no.0 o~.o o¢.o mm.o ao.~ 0H 00m 00¢ mN.o 00.0 00.0 no.0 m¢.0 N~.0 0H 00o oom mm.o oo.a 00.0 om.o Nq.a mm.o 0H 00m ooN o~.0 on.# 00.0 oo.o- mo.a Ho.o 0H com 00 00.0 o¢.~ 00.0 00.0 mm.H 00.0 11 11 In 3:;I1I'I III IIII'I'IIIFIIrIIE.u1i5NU Ill-".1.- 1N1iIIIII..Hl1l.I.l.I.Lu'1.d.I..-.llhu. a.Amswoc _\n2mc *AmeooH\meva ceathWX «2meooa\nEmvs AmEmooH\mEmv sAmewooH\mewv “.munv oo.anH oo .dEoH .m0:1..msz ucoocoo mmz anoucmcmo umou noom3. acou.cou uocm3 umoa .u3 Hmuoa mafia andan HmauwcH 11'1“ NIIII’NIII .1 .11... .312I: II .1 I N1I1.IVI.'I1I..I 1 II IIIL. IIII.Iw I1 1 lHI. I.-1.1 IIIIIIIIA... ...Il..1I 1H1lII|I|I1l II. .IIJIII.II'II.ILIIE1I IIIHEN. ’ENIIIIT :n '10” IV... .a:. iril I!” 1Iriu51huv§t3yl I... ' '11 5!? mos; EH3 amaaaécfiz 29$ 92 mamifioofiomam 5 8.3.3:... 250mm»: zoom 3.: was: 350.2% so: .53 £2 92 amass 68:5 .289 a: H> mam<8 __ 55 .ou.:ou:mn omuw mmz .mmuw “mama mo msmuw do. awn mswum mm cummmumxm* qo.o No.0 mm.“ oo.o oo.o me.“ o. coo , .oom mm.o ...o no.0 90.0 00.0 q... 0. com co. mm.o «o.o oo.o eu.o N..o .... o. co. com .m.o mm.o oo.o mm.o m... Nm.o o. com com No.o om.. oo.o mo.. mo.~ mo.. o. co. co No.0- mm.. oo.o om.m oo.m mm.m m.. co co .o.o .m.. oo.o m~.m om.m m~.m m.. co co 00.0 N... oo.o oo.o «o... oo.o . - - - .mmmmoo.\mamw1gflmmmwo.mmawv.mowmmmwx .Amewoo.\.swm-mfimeoo.\m5mv I..swoo.\mswv ..musv co .ae.a oo .nawm 1.11wmmm-mmz1- HHHMm-wimmmI-hmmmmmp .3... “mumrfiflwwhwwg 30.. £3 .30... ms... .3... WWW-Hm m0¢zz saw: QMNHA mAmuso mommamsm HmEuosu Hmwucouumman .NH ouamam IT++ com com 00m coo com 00¢ com com ooH 00 .anH _ _ _ _ _ h _ h _ Ham cw can lllll I.|.l .I.I.I .H .ouHGOuaona~< .muaaouaun-a< .2 a - ...... : Hariflry _ > . .muaCOucon|Hu=o mommfimcm Hmsuonu «maucouomwan .wH ouswfih 000 000 00m 000 00m 00¢ 00m 00m 00H 00 .QEmH _ _ p . _ . \u _ . who :« can .H ...sa.s.s..z Nz a“ can .ocfiamamsumz .N ~z 5 as» .o ooom ou nouns: _ _ _ _ 0583.3qu lllllllllllll. .m N2 5 E: o ooom _ I s _ _. P _ _ cu nouns: 1.1: I. . .d madamamsumz 61 .msaw Hwauum nucoouxamnoao.mou«=oucon umuouaumm unwamHmnuo mo mo>uso mommamcm amsuosu Hawuaouowwun .mH ousmam 00m 000 005 000 com 00¢ 00m 00m 00H 00 .0608 _ was ca any .ocusuasnum «2 m« ::u .ocusuasnum «2 as 65H .0 Doom ou vouaos ocssuasasm Nz as as“ .u coon cu nouns: oshsaasaum 62 The heats of reaction of methylamine are very similar to those ob:aiced for release of NH3 from bentonite. Examination of the weight loss data in Table IX shows that the loss of one mole of methylamine is again accompanied by the loss of one mole of water. It should also be noted that a methylamine saturated bentonite dried at 60 degrees centigrade under vacuum contained 4.15 grams or 129 milliequivalents methylamine per 100 grams bentonite. This is approximately 40 milliequivalents greater than the cation exchange capacity. Twenty milliequivalents of the methylamine were lost upon heating the sample to 200 degrees centigrade and the sample heated to 250 degrees centigrade contained 91 milliequivalents of methylamine, approximately the cation exchange capacity of the bentonite. The rapid loss of this excess methylamine suggests that it was held by weak physical forces, although the concurrent loss of an unknown quantity of water in this region makes it impossible to estimate the energy associated with these physical forces. The average.£mifor the loss of methylamine from 200 to 450 degrees centigrade is 40.9 kilocalories per mole lost as compared to 35.3 kilocalories per mole of NH3 lost during the same temperature region from the bentonite containing the least exchangeable aluminum. The heats of reaction obtained for loss of ethylamine from bentonite given in Table VIII are extremely large except for the 350 to 450 degree centigrade range. This very low value is attri- buted to experimental error. It may also be noted in Table X that the loss of extremely large quantities of water occurred up to 400 degrees centigrade. For example, in the range 300 to 400 degrees centigrade for each mole of ethylamine lost approximately seventeen 63 .ooacoucon ovum n 120.cou0 poems mo macaw 00H you macaw mo commouaxm unfillllllileflfllflIfimilIIllliIIInCILilifllIBlF.filfifififlwn Ijflu ;, % mm.o am.o ~m.m ~m.o oo.o ~e.q oa can one NH.H ma.o No.o oo.o um.o m¢.~ om one omm Na.o em.a oo.o oa.o mm.o no.a ca omm omm om.o ma.~ co.o an.m mo.a ~m.o as own as oo.o m~.e oo.o oo.o o~.k oo.o I- assou> a co ca.o oo.o cm.a mo.o oo.o mn.~ as com com m~.a ea.o oa.o «c.o mo.o om.~ as con cos Ho.o se.a oo.o mm.o Hw.o em.o om cos com m¢.~ mo.~ oo.o mm.o «H.H mm.~ om com com am.o on.m oo.o mo.a No.~ mo.a oH com oe oo.o to.e oo.o co.o oH.n oo.o .. assom> a co I~mawmu4umem¢IIomswooH\as v masseuse “waw00a\aemo ~mamooa\mswo Aaewooa\mswo A.ssno oo .aEma co .aEwa umoq nrzu occucoo mmzu Iouohzoa unoq noon: ucoucoo nouoz umoq .u3 Houoa oEHH Assam amauacH ..IIUIE I'llu' wn'r 'I Ibis-II I...1.Ih.:.“ I III Fl Ini FIEL’EEulFlhIIiIIHII‘llr. [[14 “II'. PIE I.I lflhfi"; .|.. ”Ill-tr. I‘LL-I"! I.. .F‘NIIH I .l E MHHZOHme 0m9<¢09 0 00 00.N 00.0 05.¢ H0.0 00.0 00.5 0a 000 00¢ m~.0 50.N 00.0 50.H H0.0 0~.N 0H 00¢ 00m mm.~ 0m.~ 00.0 0H.m 05.N H5.0 0H 00m 00m ¢~.m m¢.¢ 00.0 00.0 00.5 00.0 0H 00m 00 00.0 M5.5 00.0 00.0 «0.0 00.0 II Esaom> 0 00 «Amewoofi\msm0 smmswooa\mewviwo«umamx «Ame00~\memV «Amewooa\mewv tnmaw00H\mBmv . Annunv cc .0809 00 .0509 0004 5mz~0 uc0ucoo 5mz~o Iouvms00 smog H0003 0:00:00 00003 0004 .03 #0008 0808 Honda HofiuficH mHHZOHZNm 0m9<¢0fiuoo m0mh~000 H0800nu H000c0u0wmwa .0N 000000 com com oo5 com com cos com com 000 no .0508 _ _ _ _ _ _ _ . _ 000 :0 0:0 IIIIII -IIIF .Hocanumz .0 .6. _ z :0 0:0 I I.I I lllll _ _ fl . Hon—0:002 _ LII . N 000 00 :30 .0oameum N2 =0 :50 .0osa000 67 .0000 000000 500000000I0oo M000CO0000 000005000 0000 000000 000 0080om mo m0>0=0 000>0000 0080000 000000000000 .HN 00=w0m oom .oom oo5 com com ooe com com 000 00 .0800 _ _ _ P _ _ _ _ _ .0 000 :0 0:0 .000< 005000 .2... _II _ _ t\/_{ .0000 008000 I I. IIIII II I. .N Madcap—3H ..IIIIII IDII'IJ llllllll II .M .0000 000004 «z :0 :00 .0000 o0umu< 68 However, once again when oxidation was eliminated, little evidence of bonding energy is shown. It is interesting to note that the endotherm is smaller for curves 2 and 4, figure 21, for the organic acid saturated bentonite than for the corresponding curves of alcohol saturated bentonite shown in figure 20. This again gives evidence that the presence of hydrogen ions results in a lower energy dehydroxylation. General Discussion of Heat of Reaction Data No discussion of heat of reaction data would be complete without generalizations as to its applicability and some estimate of the errors involved in its measurements. To illustrate the errors involved in estimation of the heats of reaction of NH3 desorption by differential thermal analysis, an example will be used. An estimate of the errors in determinations for the NH4~ bentonite with least exchangeable aluminum heated from 200 to 300 degrees centigrade is given in Table XI. Two sources of error not listed in Table XI exist which are difficult to estimate. First, the reliability of the standard curve is Open to some question. However, if the base line is accurately established, it appears that this error should be less than two per cent. The second error comes from the effect of a temperatugf on the heat of reaction. It is well 0 known that AH '-' AHo-I- JAdet. Since the heat of reaction used in the calibration curvefi‘were all for 298 degrees centigrade, an error undoubtedly occurs because of the effect of temperature on heat capacity. However, it would appear that both of the above errors are small in comparison to the error in determining the change in 69 nitrogen content and the change in area. This example readily points our that when measurements are based on small differences, rather large errors may result. The nitrogen determinations were made in dupli;éfe and the areas were determined in duplicate which should reduce the probability of attaining the maximum error. TABLE XI ESTIMATION OF ERRORS IN EETERMHNATION OF HEATS OF REACTION _ __ _— — — _— ‘.=_ - ..n'ru _‘ Estimated Estimated Determination Quantity Error ___ Z Error 1. Total Wt. Loss 0.7449 gms. ism-002 gms. 0.03 2.(a) Total Nitrogen 81.6 m.e.llOOg.th.8 m.e.llCOg. 1.0 2000 C (b) Total Nitrogen 60.6 m.e./lOOg.‘tO.6 m.e./lOCg. 1.0 300° c (c) Nitrogen Lost 21.0 m.e./l00g. tl.4 m.e./lOOg. 7.0 3. Water Lost 0.45 gm/lOOg. It0.02 gms. 4.0 4.(a) Wei ht of Area 0.728 gms. i:3.007 gms. 1.0 200 C (b) Weight of Area 0.642 gms. i;0.006 gms. l.0 300° C (c) Change in Area 0.086 gms. $0.013 gms. 15.0 While the heats measured are from the heat absorbed when a bond is broken, it must not be assumed that this is the energy involved in the release of the ion in the soil. In a natural soil system the energy involved is that of exchange of one ion for another, for example, the exchange of hydrogen ions for NH4 ions. The net energy change for cation exchange in soils would not be the same as the energy change when NH“ ions decompose to form NH3 and hydrogen ions, CHAPTER V DIFFERENTIAL THERMAL STUDIES OF VERMICULITE Differential thermal curves of vermiculite tend to be quite variable. Barshad (2) found differential thermal curves which range from vermiculite with little or no heat changes occur» ring until 800 degrees centigrade to material which exhibited rather strong dehydroxylation peaks near 500 degrees centigrade. In studying the nature of an NH3 saturated vermiculite he obtained differential thermal curves which gave no sharp peaks, but produced small, broad exothermic peaks. He studied the loss of NH3 and water from vermiculite and drew the following conclusions: (a) The loss in weight up to a temperature of 255° C. is of water alone, (b) the loss between 255° 0. - 550° c. is relatively small and consists of equal amounts of NH3 and H20 in terms of equivalents, (c) the loss between 5500 C. - 6000 C. is considerable and also consists of equivalent amounts of NH3 and H20, and (d) between 6000 C. - 8000 C. the loss consists of water alone and is equal to crystal lattice water, i.e. OH' of the octahedral layer.1 Scott, g;_al (25) found that exchangeable NRA ions were decomposed by heating at 400 degrees centigrade while NRA ions 1Issac Barshad, "Vermiculite and Its Relation to Biotite," American Mineralogist, 333669, 1948. 70 71 fixed on vermiculite were stable until nearly 400 degrees centi- grade and then required heat at 600 degrees centigrade for 2h hours for complete removal of fixed NH4 ions. The experimental methods used for studying the heats of desorption of NH3, methylamine, and ethylamine from vermiculite were the same as those described in Chapter IV for bentonite. Differential thermal analyses curves for NHa-bentonite are given in figure 22. As can be seen from curve 2, the loss of NH3 did not yield sharp, well defined peaks. It should also be pointed out that the base line was not the same as that obtained by comparing A1203 against A1203. This drift of the base line in the differential thermal analyses curves is associated with change ing thermal prOperties, i.e. thermal diffusivity and/or heat capacity of the sample, upon heating. From a theoretical point of view a simple relationship between peak area and heat of reaction would no longer be expected and the use of a calibration curve to obtain heats of reaction is very questionable. Furthermore, the drift of a base line makes it impossible to establish the new base line with any degree of accuracy. The heats of reaction estimated for desorption of NH3 from vermiculite are given in Table XII. It can be seen that the values calculated are quite variable. Part of the variability would appear to be accounted for by the simultaneous loss of water. The weight loss data given in Table XIII shows that below 350 degrees centigrade the entire loss may be attributed to water. It may also be seen that the larger4AH,of desorption values are asso- ciated with large water losses. However, there appears to be more 72 .ouam “sauce sucoouxwmaoco mouwHoquuo> voumpouwm mmz mo mo>uso mommamco HmEuonu Hoaucouowwfin .NN ouswwm com com 00m com com coo oom ocm 00H 00 .anH _ _ _ _ _ _ _ r _ man a« can .ouaasuaauo>u¢:z Nz ca can MI I. lllllll HHW llhllllrl_\|l, _ llllfillllllllhllll .N .ouzoowshoeaaqmz (Hos: omen Nz so can .0 000m Ou vouoou ouaH30HEuo>a mz m N z :a eat .0 ooow cu mouse: IIIIII .u I I _ — _ lllllrlr _ _ ouaaaowauo>uq=z .o 73 variability than can be accounted for by water loss. It is felt that large experimental errors have resulted from difficulties in correctly establishing the base line. TABLE XII THE HEAT 0F REACTION OF NH3 RELEASED FROM NH3 SATURATED VERMICULITE* Initial ' Final lgn " Temp. 00 Temp. °C K Cal/mole NH:E* 300 400 96.4 400 500 120.9 500 600 18.3 350 450 29.8 450 550 55.2 550 650 91.5 *Values for temperature ranges 300-400, 400-500, and 500-600 are from one sample while values for the temperature ranges 350-450, 4500550, and 5506650 are from another sample preparation. **Expressed as kilocalories per mole of NH3 evolved. Differential thermal analyses curves of methylamine and ethylamine saturated vermiculite are shown in figures 23 and 24 respectively. The curves obtained by making analyses under nitro- gen gas are very similar to those obtained from NH3 saturated vermi- culite. The base line again drifted during the determination. The estimated heats of desorption for methylamine and ethylamine from vermiculite are given in Tables XIV and XV. Again a large variability is found. The weight loss data given in Table XVI 74 ..-.mw..o .if. None : mm.o wH.o mn.o no.0 oo.o: o~.~ Mo.o qH.H 00.0 mH.H mm.o “0.0 mm.o mm.o m~.o mm.o mo.o mm.o mo.o cm.o $0.0 mo.H a... is- m... a V. I n}. s .. \ Nazi cow mzz No.0 we.” n~.o mm.m ma.o 00.0 OH.O mm.~ ou.o Om.~ om.q 00.0 t : ,..«DHIIILIF.I.HI. iii—Fail. urn inmfimoow\mam 203mm Ema.” g “HIM-Ml. q ..llIHIlul‘rlurfilll 555:2“; amasseé $2 $2 92 £8.43 $2sz .289 a; HHHX MAm scum noom3 oouu mmz no mamum cos you mamuw no v.0mmouaxme (”brain Ll immWMuu. ,- -.Wa.o oa one;- can No.0 mm.a ca 0mm one me.~ om.o ca one can nn.a mm.m ea 0mm can ee.m ea.e oa own as n~.aa 00.0 -- esser> e as 00.0 me.o oH ooe oom ea.o wn.a ca cow ooe me.a He.o 0e cos com nm.~ mm.a oH oom ooN nH.m mm.e ea oo~ oe ne.n oo.o -- assee> e co Anemoo~\mawe eheswon\nemv A.enrv uo .eEee oe .eEee uemwwmmwmwwmmnuunnnnme.mm“my eewmwvul Heeemnauaummwwmmw 75 .onau Honuom ounoouwflmuoco mouawsowawo> menopause codenamnuoa mo moanoo mon>Hmcm Hmauofiu Hoaucouommgn .mN ouswwh com com Ooh coo con coo com com God 00 .anH was ea c .ouwaoowauo>unmzo Nz cw can iluflfli nouHHDUHEpo>ummzo /\ avocaa omen «z a“ can . _ 0 08m 8 serene l? «I. IIIII _ Ill—Ii _ _ _ .n ouwguaekoerummzo é How: soon N 2 so can Rm, _ .o coon cu mouse: In _ _ _ pi _ _ _ .o ouaasoHauo>ammzo mafia moon 76 .ouam amauuo nucooux«mooco mouaaoowsuo> pouououom onwawamnuo mo mo>uao mommfiocm Hospocu adducouommfla .ou enough ooe com can ooe com ooe com oo~ cos oo .naea p F _ _ _ _ _ _ _ new a“ any .oueasunsue>onmz~o I’llll Ill l|'l- NZ aw Ga _ _\’-_{IIL(—u\llr .ougfiaogapo>nn=z~o I I N 2 CH can .0 000m 3 e33: LII- lllllll _ F C Ll .m mnemooH8uo>ummz~u muocaa omen II. Nz a“ any _ \HI _ _ .o coon 3 .033; ilk» lllll _ l lrll .e ouaasowaho>inmz~u.11. /fl\ Aflooaa moon 77 shows that relatively little loss of methylamine occurs below 200 degrees centigrade. However, the loss of methylamine begins at a lower temperature than did the loss of NH3. It may be seen in Table XVII that more than twice as many equivalents of ethyle amine are held by the vermiculite than either methylamine or NH3. However, some of the ethylamine is driven off by heating the sample to 200 degrees centigrade and much more is removed by heating the sample to 300 degrees centigrade. However, the largest losses occur between 250 and 450 degrees centigrade. TABLE XIV THE HEAT OF REACTION OF METHYLAMnNE RELEASED RROM METHYLAMINE SATURATED VERMECULITE* =fililil‘ FiSZi QKE‘ Temp. 0C Temp. 0C K Cal/mole CNH5*“ 200 300 266 300 400 150 400 500 47 250 350 230 350 450 22 450 550 52 I ,.n *Values for the temperature ranges 200-300, 300a400, and 400~500 are from one sample while values for the temperature ranges 250-350, 350°450, and 450-550 are from another sample preparation. **Expressed as kilocalories per mole of CNHS evolved. 78 It should again be pointed out that the estimated heats of reaction involving vermiculite cannot be considered valid because of the change in thermal prOperties of the sample upon heating. TABLE XV THE HEAT OF REACTION OF ETHYLAMINE RELEASED FROM ETHYLAMINE SATURATED VERMICULITE* =Initia1 Final All Temp. 0C Temp. °C K Cal/mole C2NH7** 200 300 62 300 400 32 400 500 58 500 600 21 250 450 27 450 550 49 *Values for the temperature ranges 200-300, 300~400, 400e500, and 500e600 are from one sample while values for the temperature ranges 250-450 and 450-550 are from another sample preparation. **Expressed as kilocalories per mole of CZNH7 evolved. .II) .I .I.... .mufleceafiuo> menu Loom3 .moum oceans votes mo macaw 09H you mamum mm nonwouaxma II. ..EUIJ...I.M ”.1I.InII|I. HIIIHII‘“ PU. ION"... h IIIIIIIIIIIHIlhl . I...» IIHIIII. II.II... IIIIII. .INr. Igg.nu.”fi IIHIIII. “IUINHIIII .II I. I |.. E II .I. I... IonulI I. l1l.l\ll In. IlihflI I‘M I . I «IIIIAWI I..,..N .II I. II. II. I. ILII I.II...|..r.II I11 . I Ilr IIIIII. ,.I.....unp I II IIIiI III. IIIF. In. I IIIKMII . .II ..I I I .II .III II I. I. No.0 00.9 No.0 om.c mm.o m@.o 90.0 Fl C; O Pd om.H om.“ no.~ oo.o no.0 m¢.~ w©,H Hm.~ mm.~ raUEMC a\mFm% L Lerner, nmzc nICH sotoz I... I . ..III I IN” I. I NNIIHIIHi I , I.III.IIII .rIIIhHHIfiu III. . .I,I. IIIIIII .IIlIIHIHI....II.II.IIu PIIIIIIpIuHJIH'HIIIhII I II .I II. .14.“. .I. I III. III IIII I. \l em.o oo.e LM.L es one one om.o em.e mm.o 0L ems omm ee.o se.o em.o ea own on“ na.m ee.e ea.m oL own oe 00.0 me.w assure e 0e so 0. eo.o ee.o oL ooe com Na.o- no.0- ee.o oL eon ooe ee.o ML.o- LL.o oL ooe 00m om.~ mm.o mm.H OH oom cow C“ (V ‘¢ (“ \D P“! am.¢ OH 00m om 00.0 om.m 00.0 II Eosoo> o oo .IflrnntflIJIII. . .II..Il..Il ..IILI. II.II.§ I ....I.HI .I INCH“ ...I"...fil mc e\IEn , nsmoc. mEMV .Lwewood\mywv mawwsv Do .MHWB UHW;GCJ Hmsm3 mwOJ .w3 dmuQH USHH MWfiHW .IIIII I IVII.| W. in I. I.II.. I ..I., ... II I. III . .n'IHV I! I'll-Ill ..I» IIIIIII I..I.III. oo .mEoH Hmauaca MHEML Hme> mmemmsemhmz Zcmm Hmcg mzsz1em2 m2< mmL¢3 EIQLmz AX m4m<9 80 .II IIFIIIIII I 1!“. IHIIHIHJIJII .I.}...HIIWI E..III..IIIIIIRIII1IHI.... . . I II ..I | II..II.I.|IIIIH4IIIIIII..III.I. ...IIIIIIIHI.II. "I HIEHIII-“II'InIIHIII II‘EI oouawduwfium> mmyu umum3 amouw mawawflmgum mo meum COL you wawuw mu @mwmwumxmw IIII IIIVII'll.“.IHIPINHII.|n15.“.l.ulllIII...IIuI'II“.HI“IHIIDIIEIIIIHI'IIIFIIII'IIIII'IHIPHHIIII|"H..IIIII'IIII.IIIIIIIN..-..I,u.. InIIIuIL Ingl I.I.....L -MMIM. .IIIImme. .mmIO oooo mm.H ON omo 0mm chow mmoo No.0 mmoo mm.d 0H own omq qoog ¢H9~ mqgo no.0 mo.~ OH 0m¢ 0mm «cud wLIM MOIH oqoa mo.~ 0H 0mm 0mm LHON o¢.m Hmon m¢.~ wo.o~ OH omm ow 00.0 mm.m 00.0 ¢m.o~ oo.o II Essow> a 00 m~.o mo.o o~.o 00.0 mo.~ OH com com ma.o ww.o no.0: o~.o om.o 0H oon ooq mm.m Hm.H omoa mw.o mm.m OH ooq oom Hm.o mo.m mN.N mm.fl wo.m OH com com mm.o mq.o mo.q LmIm ~m.¢ 0H CON om oo_o wmom 00.0 cm.n 00.0 In Esauw> ¢ oo ....w............. awacavaIHIaflL.. 2.32:... .3323- ...... 0.. .....m 00 ...... -..... ....LNWI..WML.-...WMMII....Ig- :5 IE. .2: 3:2 message, amegfifi szEEE :05 $3 §Ex mgmuao momhaocm HmEuonu Hoaucouomwaa com com ooh coo com cos com com _ _ _ _ _ _ _ t .aonauon 5m Ill, uwm cw can icouwuon um Nz :a can le_ 1::upnrr _ r .lrlr, _ _ .cosauon om on“ unawam ooH oo .aama _ .H 84 .vamm mxmq mafia mo coauoouw condos oau ammo mmofl man mo mo>uso mommfiwco Hoauosu Hmaucouowwao .ou ouswah com com 005 com com coo can com ooH 00 .Qan _ _ _ . _ _ b _ _ Had a“ any .coufiuo: H< .H Nz ca can .oouwuo: H< .N “an ca :3» _ .nouauon am _ .m Nz a“ can .aoufiuon an 0“. new a“ can III\\\HIII # _ .couquo: um .m Nz a“ can .couauon um 85 .NONm soak womauoouu nouns mama oxog moan mo cowuomum aouowe can :wnu mama mnu mo mo>uao mommamco Hmfiuosu Hwfiunouowwfin .NN ouawwm com com 005 com com coo com com ooH oo .oEoH _ _ _ . _ a _ _ _ yaw cw can .couauon H< .H Nz cw can .couguon H< .N was :a can _ F! F .couauos an .m Nz :« can 1? .couwuon cm .o _ r r poo ca can N2 ow can .aouwuoz on .o 86 the sample. Thus, not enough clay was present to be observed with the sensitivity used. It is not in the sc0pe of this thesis to investigate the interpretations of differential thermal analyses curves of soil clays when oxidation has been eliminated. Rather, based on the results of one soil profile, it is suggested that this method may be useful for future studies of soils which are high in organic matter. CHAPTER VII SUMMARY.AND CONCLUSIONS A theoretical discussion of the nature of heat flow within a differential thermal analysis sample holder was presented. It was shown that a simple relationship exists between the area of the differential thermal peak recorded and the heat change in the sample during a reaction if the differential thermocouple is exactly centered, the heat conductivity of the sample holder much greater than that of the sample, and there is no vertical heat flow. The theoretical discussion was in agreement with the calibration curve made using nine different salts with known heats of inversion or fusion. A comparison was made of the heat of reactions for decome position of magnesite and dehydroxylation of kaolinite as deterw mined using the Clausius-Clapeyron equation and the standard curve prepared using the pure salts. The deviation was less than four per cent for both materials. The average value obtained by the two methods for decomposition of magnesite was 27.6 kilocalories per mole. This is much higher than the value of 10.1 reported in the literature but very close to the theoretical 27.8 kilocalories per mole. The average heat of dehydroxylation of kaolinite as determined by the two methods was 156 kilocalories per gram of material. It was concluded that either method was satisfactory for 87 88 determination of heats of reaction and that each system must be evaluated to determine which method would be most satisfactory. A Study was made of the heats of desorption of NH3, methylo amine, and ethylamine from bentonite clay. Evidence obtained showed that the exothermic peak near 500 degrees centigrade reported by other workers is due to oxidation of NH3 catalyzed by the platinum thermocouples. A method was developed to eliminate oxidation without using flowing gas during differential thermal analysis. The method consisted of evacuating the system and then introducing nitrogen gas, free of oxygen, into the system under slight pressure. This procedure was repeated three times and the sample allowed to pretreat for one hour with nitrogen gas passing through the sample under a slight pressure. The flow of gas was stopped and the pressure reduced to atmospheric pressure prior to making a differ~ ential thermal analysis. The following conclusions are drawn from the heats of reaction data and data concerning weight loss and nitrogen loss upon heating: 1. Below 200 degrees centigrade the entire weight loss may be attributed to water. From 200 to 400 degrees centigrade nearly equal weights of water and NH3 are lost. Above 400 degrees centigrade increasing amounts of a weight loss which is attributed to dehydroxylation occurs. 2. The average heat of desorption of NH3 between 200 and 450 degrees centigrade is 35.3 kilocalories per mole of NH3 lost. Only part of this energy may be attributed to the desorption of NH3 since water is lost at the same time. 3. The water loss between 200 and 450 degrees centigrade is not due to low temperature dehydroxylation. It is postulated that this water loss arises from decomposition of Al(OH)3 to yield A1203 and water. 89 4. The average heat of desorption of methylamine between 200 and 450 degrees is 40.9 kilocalories per mole of methylamine lost. As in the case of NH3, a portion of the heat absorbed mat be attributed to loss of equal amounts of water. Slightly more heat appears to be absorbed by loss of methylamine than NH3. 5. Reactions involving ethylamine are accompanied by large water losses; consequently, no value can be associated with desorption of ethylamine. Differential thermal analyses were made of formic acid, acetic acid, methanol, and ethanol saturated bentonite. In all cases, some of the compound was held by the bentonite,but the bonding energy was not measurable. Studies were made of the heats of desorption of NH3, methyl~ amine, and ethylamine from vermiculite; however, the base line drifted during differential thermal analysis indicating changes in the heat capacity and/or thermal diffusivity of the sample upon heating. Consequently, the values obtained are not expected to be valid. It was shown that the method of eliminating oxidation with nitrogen gas during differential thermal analysis could be applied to soils high in organic matter yielding differential thermal analyses curves characteristic of the mineral components of the soils. 10. 11. 12. LITERATURE CITED Arens, F, L. A study of the differential thermal analysis of clays and clay minerals. Thesis, Wageningen, Netherlands. 1951. Barshad, Isaac. Vermiculite and its relation to biotite as revealed by base exchange reactions, Xoray analyses, differential thermal curves, and water content. Am. Mineral. 33:655s678. l948. Barshad, Isaac. Temperature and heat of reaction calibration of the differential therm'l analysis apparatus. Am. Mineral. 373667m6940 19529 Coleman, N. T., and Craig, Doris. The spontaneous alteration of hydrogen clay. Soil Sci. 91:14~18. 1961. Cornet, l. Sorption of NH on montmorillonitic clay. J. Chem. Phys. 11:217m226. 1943. Ellis, B. G., and Mbrtland, M. M. Rate of potassium release from fixed and native forms. Soil Sci. Soc. Amer. Proc. 23:451-453. 1958. Eriksson, Erik. Problems of heat flow in differential thermal analysis. Kungl. Lantbrukshogskolans Annaler. 19:127~l43. 1952. Eriksson, Erik. Problems of heat flow in differential thermal analysis. Part II. Kungl. Lantbrukshogskolans Annaler. 20:117-123. 1953. Eriksson, Erik. Problems of heat flow in differential thermal analysis. Part III. Kungl. Lantbrukshhgskolans Annaler. 21:189-196. 1954. Grim, R. E. Clay Mineralogy. McGraw-Hill Book Co., Inc., New York. 1953. Grimshaw, R. W., and Roberts, A. L. The quantitative deter= mination of some minerals in ceramic materials by thermal means. Trans. Brit. Ceram. Soc. 52:50~6l. 1953. Kerr, P. F., and Kulp, J. L. Multiple differential thermal analysis. Am. Mineral. 33:387-419. 1948. 90 13. 14. 15. 16. 170 18. 19. 20. 21. 22. 23. 24. 25. 26. 91 Kiyoura, R., and Sata, T. The quantitative analysis of the CaCO3 - Ca(0H)2 - Mg(OH)2 system by differential thermal analysis. J. Ceram. Assoc. Japan. 5823-6. 1950. Le Cthelier, H. De l'action de la chaleur sur l'argiles. Bull. Soc. Franc. Mineral. 10:204m211. 1887. MacKenzie, R. C., and Bishui, B. M. The montmorillonite differential thermal curve. 11. Effect of exchangeable cations on the dehydroxylation of normal montmorillonite. Clay Min. Bull. Vol. 3, No. 20. 1958. McDowell, L. L., and Smith, G. E. The retention and reactions of anhydrous ammonia on different soil types. Soil Sci. SOC. Amer. Proc. 22:38:42. 19580 McGeorge, W. T. The base exchange prOperties of organic matter in soils. Arizona Agric. Expt. Sta. Tech. Bull. 30. 1930. Mitchell, B. D., and MacKenzie, R. C. An apparatus for differential thermal analysis under controlled-atmosphere conditions. Clay Min. Bull. Vol. 4, No. 21. 1959. Mortland, M. M. Reactions of ammonia in soils. Advances in Agronomy. Vol. X2325-348. 1958. Mortland, M. M., and Kemper, W. D. Specific surface. Methods of Soil Analysis. Am. Soc. of Agron. Monographs. In press. Murray, J. A., Fischer, H. C., and Shade, R. W. Thermal analysis of limestone. Proc. 49th Annual Conv. Natl. Lime Assoc. Washington. 1951. Norton, F. H. Analysis of high alumina clays by the thermal method. J. Amer. Ceram. Soc. 23:281-282. 1940. Olson, L. C., and Bray, R. H. Determination of organic base-exchange capacity of soils. Soil Sci. 45:483-496. 1938. Roberts-Austen, W. C. Fifth report to the alloys research committee. Proc. Inst. Mech. Eng. 1:35-102. 1899. Scott, A. D., Hanway, J. J., and Stanford, G. Thermal studies of ammonium fixation and release in certain clay minerals. Am. Mineral. 47:701-721. 1956. Smothers, W. J., and Chiang, Yao. Differential Thermal Analysis: Theory and Practice. Chemical Publishing Co., Inc., New York. 1958. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 92 Sohn, Jerome 8., and Peech, Michael. Retention and fixation of ammonia by soils. Soil Sci. 8531-9. 1958. Speil, S. Application of thermal analysis to clays and alumineus materials. U. S. Dept. of Int., Bureau of Mines, R. I. 3764. 1944. Speil, S., Berkhamer, L. H., Pask, J. A., and Davies, B. Differential thermal analysis. U. 8. Dept. of Int., Bureau of Mines, Technical Paper 664. 1945. Stone, R. L. Preliminary study of the effects of water vapor pressure on thermograms of kaolinitic soils. Proc. of Second Natl. Conf. on Clays and Clay Minerals. 1954. Stone, R. L. Thermal analysis of magnesite at C02 pressures up to six atmospheres. J. Amer. Ceram. Soc. 37:46-47. 1954. 13": an. Inn ”.3 i. . j Stone, R. L. Differential thermal analysis by the dynamic gas technique. Anl. Chem. 12:1582-1588. 1960. Stone, R. L., and Rowland, R. A. DTA of kaolinite and montmorillonite under water vapor pressures up to six atmospheres. Proc. of Third Natl. Conf. on Clays and Clay Minerals. 1955. Turner, P. E. An analysis of factors contributing to the determination of saturation capacity in some trapical soil types. Journ. of Agr. Sci. 22:72-91. 1932. Van der Marel, H. H. Quantitative differential thermal analysis of clay and other minerals. Am. Mineral. 41:222-244. 1956. Wittels, Mark. The differential thermal analyzer as a micro-calorimeter. Am. Mineral. 36:615-621. 1951. Wurman, E., Whiteside, E. P., and Mortland, M. M. PrOperties and genesis of finer textured subsoil bands in some sandy Michigan soils. Soil Sci. Soc. Amer. Proc. 23:135-143. 1959. ' ‘ 1293 03058 123 R" L" Yl| " I“ H A“ H