I Reaction Kinetics by Differential Thermal Analysis

Wesley W. Wendlandt. Texas Technological College. Lubbock. Differential Thermal Analysis. I A physical chemistry experiment. The determination of the ...
1 downloads 0 Views 3MB Size
Wesley W. Wendlandt Technological College Lubbock

Texas

I

I

Reaction Kinetics by Differential Thermal Analysis A physical chemistry experiment

The determination of the reaction kinetics for homogeneous reactions is a common experiment in the undergraduate physical chemistry course. However, the study of heterogeneous reaction kinetics is a field of chemistry that is virtually untouched by standard physical chemistry courses. Outside of the initial cost of the equipment, the determination of the kinetics of a thermal decomposition reaction IYIII he P R S I I ~ and sin~plycarried out I]? t l ~ us? r of he ~ w t ~ n i rof, uditTwenti~1 ~~ ~ h w ~ nmnlvsis al (LYl'.A,. The methodsof DTA have been amply ilfustrated by Borchardt (I), mainly from the viewpoint of determining the heat of reaction from the dehydration of salt hydrates. Since the reaction kinetics can also be determined by this technique, this experiment is presented to suv~lement the previous heat of reaction -studies (1). There are a number of methods involving DTA that may be used to determine the kinetics of a thermal decomnosition reaction. Most of them involve reactions df the type heat solid A

solid B

+ gas

The first of these methods is that proposed by Kissinger (8, S), which was derived from the work of Murray and White (4) and Vaughn (5). This method is based on the equation.

where 6 is dT/dt, the heating rate of the furnace, T, is the peak temperature maxima, E* is the activation energy, and R the gas constant. Once the above values are known, the frequency factor, A, can he calculated from

The above equations contain no terms relating to the reaction order, n. However, from the asymmetry of the curve (3), described by the shape index, S, the order of reaction, n, was shown to he equal to: Thus, by measuring the peak maxima temperatures a t a number of different furnace heating rates, and by plotting log @/Tm2versus 1/T,, a straight line is obtained of slope -E*/2.3R. A second method, proposed by Borchardt and Daniels (6), was originally derived for the study of homogeneous kinetics but, as shall be shown later, can he applied to heterogeneous kinetics as well.

This method is based upon the equation, derived for the first-order reaction (n = I), k

=

Cn(dAT/dt) K(A - a)

+ KAT

+ C,AT

where k is the reaction rate constant, K, is a constant characteristic of the experimental apparatus, A is the total curve area, C , is the total heat capacity of the reactant or reference material, AT is the differential temperature, t is the time, and a is the area under the curve up to time t. The above equation was later simplified by Borchardt (7) to

where k is the initial reaction rate, AT is the peak height (differential temperature), A is the total peak area, and -d(n/no)/dt is the rate of reaction in fraction ronverted per unit time a t the time (temperature) where the DTA peak height is AT. From a rough triangulation of the curve peak, A = TAT,.,; where T is the has? of the triangle (peak) and AT,., is the height of the peak. Thus, it is seen that

Another approximation of the Borchardt and Daniel's equation (6) has been derived (8, 9). Since C , is normally quite small, the equation simplifiesto:

where the terms have the same notation as before. It. is rather interesting to note that an identical method for determining k was developed by Allison in 1954 (10) from a consideration of t.he area under specific heat-temperature curves. The approximation method in equation 17) yields good results as confirmed by the thermal decomposition of metal oxalates. An excellent correlation was found between the DTA and thermogravirnetric analysis (TGA) reaction kinetics data (8, 9). Also, since this approximation method is simple to use, it is the basis for the determination of the reaction kinetics described in this experiment. Experimental

There are many different types of DTA setups that may be used to determine the thermal decomposition curves. The apparatus described by Borchardt ( I ) may be used, if a small synchronous motor is attached to the Variac (11) so as to give the furnace a fairly linear heating rate. The apparatus described by Wendlandt Volume 38, Number 1 1 , November 1961

/

571

(18) may alfio be used. An inexpensive apparatus, developed over a period of years, is illustrated in Figures 1 and 2. The apparatus consists of a furnace and sample holder, a furnace temperature controller, a lowlevel dc amplifier (Honeywell, type 2HLA-7) and a strip-chart potentiometric recorder (Varian, type G-10). All of the above components are standard commercial items except the furnace and sample holder assembly, which can be easily constructed. Certainly other amplifiers, recorders, and furnaces may be substituted for the above components.

IOTURN HELlPDT

synchronous motor was attached to the powerstat through a 5: 1 vernier radio tuning dial. The advantage of this system is that the tuning dial is inexpensive plus the fact that it acts as a friction clutch to allow the powerstat to be returned manually to its starting position. The CaCZOa.lHn0 used in this experiment was prepared by adding oxalic acid to a dilute calcium nitrate solution, digesting on a steam bath for one hour, and then filtering off the precipit,ate into a sintered porcelain crucible. After washing several times with water, the precipitate was allowed to dry in air for about 48 hours. The NaHC03 used was of reagent grade quality. Procedure for the DTA run consisted of the following: 100 mg weight of sample; 10°C min-I heating rate; previously ignited alumina being used as the reference material. Duplicate runs gave peak maxima temperatures reproducible to ~ 5 % . Results

A

RECORDER

B

Figure 1. A, Schematic illurtration of DTA furnace and sample holder. Schemotic diogrorn of DTA apporatur

8,

The furnace consisted of an 8-in. length of 1-in. i.d. high temperature ceramic tube wound with enough Nichrome wire (1.06 ohms per ft) to give a total resistance of 25 ohms. The furnace windings were insulated with about a 0.5-in. layer of asbestos. The base of the furnace was constructed from 0.25-in. thick aluminum sheet in such a manner that the lower plate could be easily attached to the upper plate by two bolts and wing nuts. Attached to the bottom plate were two 0.25-in. 0.d. diameter ceramic rods, each containing two 0.060-in. holes. The sample chambers consisted of two nickel or ceramic cylinders, 0.25 in. i.d. and 0.75 in. long, which are placed over the ceramic rod and thermocouple junctions. Each of the sample chambers could be easily removed for cleaning out the residue from the previously ignited sample. The furnace temperature controller was constructed from two powerstats, type 116U 0-135v, one of them driven by a 2 rph Cramer synchronous motor. The

The results of the DTA studies on Cacaoa.lHzO and NaHC08 are given in Figure 3. Although only a a static air atmosphere can be maint,ained in the apparatus described herein, the CaCzOa.lHzO was also run under helium in a controlled atmosphere DTA apparatus (IS). The results obtained ITere quite different from those found in a static air atmosphere. The thermal decomposition of CaC2O1.1H2O, as

I

I 100

200

I 400

SO0

I 500

I

I

I

600

700

800

TEMPERITURE. .C

Figure 3.

Differential thermal molysir curves.

shown by thermogravimetric studies on an automatic recording thermobalance (14) and illustrated in Figure 4, takes place according to the following reactions:

-

CaC9O4.lHIO(s)

and CaC204(s)

U

A

B

Figure 2. A, Schematic illustrmtion of motor-driven powerrtot. Wiring diogram for furnace temperature controller kchernotic).

572

/

journal of Chemical Education

8,

CaCnO.(s)

CaCOs(s)

+ H20(g)

+ CO(g)

This thermal decomposition pattern is shown in the DTA curve (in air) by an endothermic peak for the dehydration reaction and a broad exothermic peak for the decomposition reaction. However, under helium, the second peak is an endothermic, rather than an exothermic peak. This difference can be explained in that the exothermic peak in the air DTA curve is the resultant of two competing reactions. One of them is the

kinetics data obtained, E* was found to be 20 kca mole-'. I n this reaction, as well as in the above examples, the reactions were assumed to be first-order. The dehydration reaction of CaCzOa.l H 2 0 has previously been shown to be first-order also but the C~CSO, decomposition reaction was 0.76 order (16). The two examples listed above certainly do not exhaust the number of compounds that can be studied by this method. A series of hydrogen carbonates containing diierent cations may be studied as well as a number of inorganic salt hydrates. Many compounds evolve oxygen when heated as well as other gases such as carbon dioxide, sulfur dioxide, and sulfur trioxide; the list is virtually endless. It is hoped that this experiment will stimulate further studies in this exceedingly interesting area of investigation. It is a pleasure to acknowledge the assistance of Mr. William Robinson and Miss Phyllis J. Kuhn. Figure. 4.

Thermogrovirnetric mndysis curves and a gar evolvtion curve.

above decomposition reaction while the second is the oxidation of carbon monoxide to carbon dioxide, according to the reaction

-

co + '/202 cog which is a highly exothermic process. Thus, the effect of various atmospheres on the thermal decomposition reaction can be readily seen. Another illustration of the decomposition of CaC204.lHzO can be seen in Figure 4 where the gas evolution is plotted as a function of furnace temperature. Since the thermal decomposition was carried out in helium (IS), the first peak is due t o water vapor, and the second peak is due to carbon monoxide evolution. The kinetics of the dehydration reaction of CaCzOg lHzO was determined according to eqnation (7). Four points were chosen a t random on the dehydration peak. The areas were obtained by integration of the peaks by the use of tracing paper and a semimicrobalance. The data obtained were in the followingunits, although other units may be employed: AT, in mm, while A and a were in mg of tracing paper. The value for E* was obtained by use of the Arrhenius equation:

A plot of log k versus 1/(T) gave a straight line, as shown in Figure 5, of slope -E*/2.3R. The value obtained for B* in this study was 27 kcal mole-'. A previous TGA method by Freeman and Carroll (16) gave 22 kcal mole-'. The slight difference between results may be attributed to the effect of particle size, heating rate, furnace atmosphere conditions, and so on. By similar calculations, the kinetics of the decomposition of CaC2O4(under helium) were also evaluated. An E* of 85 kcal mole-' was obtained, compared to a previous value (15) of 74 kcal mole-'. However, the latter value was obtained in an air atmosphere. To illustrate the above method with another compound, the thermal decomposition of NaHC03 was also studied. The decomposition reaction studied was:

-

Na82Ods)

(08

Arrheniur plotfor the dehydration of C a G 0 4 . 1 H1O.

Literature Cited

k = Ae-E*/RT

2NaHCOds)

!+!x Figure 5.

+ H*O(g) + COdg)

(1) BoRcnnnD~,H. J., J. CHEM.E ~ u c33, . 103 (1956). H. E., J . Res. Natl. But. of Stand., 57, 217 (2) KISSINGER, (1956). (3) ~ S S I N G E R ,H. E., Anal. Chem., 29, 1703 (1957). P., AND WHITE,J., T ~ a n sBrit. . Ceram. Soe., 54, (4) MURRAY, 204 (1955). F., Clag Minerals Bull. 2, 265 (1953-1955). (5) VAUGHN, (6) BORCHARDT, H. J., AND DANIEL$F., J . Am. Chem. Soc., 79,41(1957). (7) BGRCEILRDT, H. J., J. Znorg. Nucl. Chem., 12,252 (1960). V. M., S ~ A I Y AS., C., AND SUNDARAM, (8) PADMANABHAN, A. K., J . Inorg. Nucl. Chem., 12, 356 (1960). R. P.. AND NAIK. M. C.. Anal. Chirn. Acta, (9) , . AGARWALA. 24,128(1960). ' 110) AT.T.~RON_ E. B.., Clav " Minerals Bull. 2. 242 (1953-1955). M. M.. RORYTA, D. A,; AND' CAPRIOLA,' G., ~

~~

.

The DTA curve is shown in Figure 3 while the TGA curve is illustrated in Figure 4. From the reaction

.

J . Znwg. ~ u c l~. h k .in, press. (15) FREEMAN, E. S., AND CARROLL, B., J. P h y ~Chem., 62,394 (1958).

+ + + Volume 38, Number 1 1, November 1961

/

573