1948
Ind. Eng. Chem. Res. 1997, 36, 1948
Interaction of Iron Oxide with Barium Peroxide and Hydroxide during the Decomposition of Sodium Chlorate Yunchang Zhang,* John E. Ellison, and James C. Cannon Nellcor Puritan Bennett, Inc., 10800 Pflumm Road, Lenexa, Kansas 66215
Decomposition of sodium chlorate catalyzed by iron oxide in the presence of barium peroxide and hydroxide is studied by thermogravimetric analysis. Barium peroxide and hydroxide are moderately active catalysts by themselves but are inhibitors when used together with iron oxide. A loading of 1% barium peroxide or hydroxide can significantly reduce the activity of the iron oxide and raise the decomposition temperature of sodium chlorate. The inhibiting effect decreases with increased loading of the barium compounds due to the formation of barium ferrate. The catalytic activity change of the iron oxide in the presence of barium peroxide and hydroxide is discussed based on the electron configuration change of the iron ions. Introduction Sodium chlorate is used as an oxygen source in chemical oxygen generators. Chemical oxygen generators are used in the majority of passenger airplanes as an emergency oxygen source. They are also used in submarines, diving, mountain climbing, mine rescue, and space flight. A chemical oxygen generator has a chemical core consisting of sodium chlorate as an oxygen source, a transition-metal oxide as a catalyst, a metal powder as a fuel, and an alkaline compound as a chlorine adsorber and a reaction modifier. Once initiated, sodium chlorate in the chemical core decomposes to sodium chloride and oxygen. A chemical oxygen generator can supply oxygen continuously for 10-30 min depending on application. Iron oxide has been reported as a very active catalyst for the decomposition of sodium chlorate by Rudloff and Freeman (1970) and Zhang et al. (1993a). Barium peroxide is usually used to modify the decomposition of the chlorate and to suppress chlorine gas formed through side reactions, as reported by Pappenheimer (1946). Therefore, it is important to understand the interaction between iron oxide and barium peroxide during the decomposition of sodium chlorate. A few percent water is usually added to the chemical mix to facilitate the formation of the chemical oxygen generating cores, and the cores are then dried to remove the water added. Some barium peroxide may be converted to barium hydroxide during the process. Therefore, it is also beneficial to understand the interaction between iron oxide and barium hydroxide in the presence of sodium chlorate. It is reported that there is a strong interaction between cobalt oxide and barium peroxide and hydroxide (Zhang et al., 1993b). When used separately, cobalt oxide is a strong catalyst and barium peroxide and hydroxide are moderately active catalysts. In the presence of barium peroxide or hydroxide, however, the activity of cobalt oxide is greatly suppressed. Iron oxide and cobalt oxide have many similarities chemically even though they have slightly different catalytic activity for the decomposition of sodium chlorate. The purpose of this study is to find out whether barium peroxide and hydroxide interact with iron oxide and inhibit its * To whom all correspondence should be addressed. Telephone: (913) 338-9856. FAX: (913) 338-7353. E-mail:
[email protected]. S0888-5885(96)00675-6 CCC: $14.00
catalytic activity and whether the interaction is parallel to their interaction with cobalt oxide. Experimental Section Iron oxide was prepared by decomposing iron oxalate hydrate (FeC2O4‚2H2O) at 350 °C. The product was analyzed as Fe2O3 by X-ray diffraction analysis. Sodium chlorate was used in the form of crystalline particles. Barium peroxide (BaO2) and barium hydroxide hydrate (Ba(OH)2‚H2O) were used as purchased from Aldrich Chemical. The ingredients of the samples for thermal analysis were intimately mixed by grinding in an agate pestle and mortar for about 5 min for each sample. The compositions of the samples were in weight ratio of the ingredients and reported as weight percentage. The total weight of each mixture was approximately 2 g. The X-ray powder diffraction analyses were conducted using a Rigaku D/max II diffractometer and monochromated high-intensity Cu KR1 radiation. The diffraction spectra were taken in the range 12° e 2θ e 80° using step-scan. The step size was 0.01° 2θ, and the scan rate was 2° 2θ/min. Surface areas were measured using a Micromeritics Gemini II 2370 surface area analyzer. The samples were heated at 150 °C for 30 min in vacuum to drive off any adsorbed moisture and other gases prior to measurements. The specific surface area of BaO2 was measured as 1.0 m2/g using the multipoint measurement technique with krypton as the adsorbate gas, and the surface area of iron oxide was measured as 59 m2/g using the multipoint technique with nitrogen as the adsorbate gas. Thermogravimetric analysis (TGA) was carried out using a Netzsch thermal analysis unit Model 409. The sample was heated at 20 °C/min in an oxygen stream of 150 mL/min. The sample size was approximately 100 mg for all samples. The experimental error for the temperature measurement is about 2 °C. Results and Discussion During the operation of a chemical oxygen generator, sodium chlorate decomposes to sodium chloride and oxygen in a pure oxygen environment according to the following equation:
2NaClO3 ) 2NaCl + 3O2 © 1997 American Chemical Society
Ind. Eng. Chem. Res., Vol. 36, No. 5, 1997 1949
Figure 1. TGA plots of the thermal decomposition of sodium chlorate catalyzed by iron oxide and barium peroxide: (1) 4% Fe2O3, (2) 4% Fe2O3 and 20% BaO2, (3) 4% Fe2O3 and 8% BaO2, (4) 4% Fe2O3 and 4% BaO2, (5) 4% Fe2O3 and 2% BaO2, (6) 4% Fe2O3 and 1% BaO2, (7) 8% BaO2, (8) 4% BaO2, (9) 2% BaO2, (10) 1% BaO2, and (11) No Fe2O3 or BaO2.
Therefore, all the thermal analyses were carried out in a pure oxygen stream even though sodium chlorate would decompose more readily in an inert environment such as in argon or nitrogen or in vacuum. The oxygen produced from the decomposition of sodium chlorate would disturb the oxygen partial pressure in the environment of the sample, which can make the results less repeatable in an inert environment than in pure oxygen. The iron oxide prepared by decomposing iron oxalate is Fe2O3 as verified by X-ray diffraction analysis. It has a surface area of 59 m2/g and has high catalytic activity toward the decomposition of sodium chlorate. However, its catalytic activity can be partially suppressed by barium peroxide and hydroxide, which are moderately active catalysts themselves. When sodium chlorate is mixed with 4% iron oxide by weight, its decomposition temperature is reduced from 566 to 313 °C as shown in Figure 1, as measured by the temperature at which 50% of the sodium chlorate has decomposed. Since the onset decomposition temperature (DT) depends on the particle size and is somewhat difficult to determine, the temperature at which 50% of the chlorate has decomposed is used as an indicator of the catalytic activity. This is referred to as 50% DT. For example, decomposition of pure sodium chlorate to sodium chloride has a weight loss of 45.1%. Therefore, the temperature at which sodium chlorate has lost 22.55% of its original weight is the 50% DT. A short-dashed line is used to indicate the 50% DT on each of the thermogravimetric analysis traces. Since Fe2O3 does not lose weight, the sample of NaClO3 with 4% Fe2O3 has a smaller weight loss than the weight loss of NaClO3 by itself. Similarly, samples containing different amounts of barium peroxide or hydroxide also have slightly different weight losses because barium peroxide and hydroxide hydrate lose less weight than sodium chlorate does. It has been reported previously that barium peroxide and hydroxide are moderately active catalysts for the decomposition of sodium chlorate (Zhang et al., 1993b). Loading of a few percent barium peroxide can lower the decomposition temperature considerably. As a comparison, the thermogravimetric curves of the samples consisting of sodium chlorate and 1, 2, 4, and 8% barium peroxide by weight are also given in Figure 1. The 50% DTs of the samples of sodium chlorate with 1, 2, 4, and 8% barium peroxide are 520, 511, 505, and 495 °C,
respectively, compared to the 50% DT of 566 °C for sodium chlorate by itself. Loading of as low as 1% barium peroxide reduced the decomposition temperature significantly. Therefore, barium peroxide by itself is a catalyst for the decomposition of sodium chlorate. A further increase in barium peroxide loading reduces the decomposition temperature only slightly. In other words, the catalytic effect of barium peroxide tends to saturate. This is another indication that barium peroxide is a catalyst rather than a reactant. The effect of a reactant is usually proportional to its quantity. While barium peroxide by itself is also a catalyst, it suppresses the decomposition of sodium chlorate when used together with iron oxide and behaves as an inhibitor. When 1% barium peroxide and 4% iron oxide is mixed with sodium chlorate at the same time, the catalytic activity of iron oxide is suppressed. The 50% DT is 455 °C for the sample of sodium chlorate containing 4% iron oxide and 1% barium peroxide compared to 313 °C for the sample of sodium chlorate with 4% iron oxide and without barium peroxide. Loading of 1% BaO2 raises the decomposition temperature by 142 °C as measured by the 50% DT. That is, barium peroxide suppresses the catalytic activity of the iron oxide. This result is parallel to the effect barium peroxide has on cobalt oxide, reported by Zhang et al. (1993b). Increasing the loading of barium peroxide, however, does not further suppress the activity of iron oxide as one would predict from the interaction of barium peroxide and cobalt oxide. When the loading of cobalt oxide is fixed, the decomposition temperature of sodium chlorate increases with increased loading of barium peroxide (Zhang et al., 1993b). The quantity of barium peroxide loading also changes the inhibiting effect on iron oxide, but the change is in the opposite direction. That is, the decomposition temperature of sodium chlorate does not increase further when more barium peroxide is loaded when the iron oxide loading is fixed at 4%. Instead, the decomposition temperature decreases progressively as barium peroxide loading is increased from 1% to 2% and 4%. The 50% DTs for the samples of sodium chlorate containing 4% iron oxide with 2% and 4% barium peroxide are 424 and 396 °C, respectively, compared to 455 °C when 1% barium peroxide is used with 4% iron oxide. The effect of barium peroxide loading on the decomposition temperature of sodium chlorate catalyzed by iron oxide tends to saturate at 4% barium peroxide. A further increase of barium peroxide loading further reduces the decomposition temperature of sodium chlorate only very slightly. The sample of sodium chlorate with 4% iron oxide and 8% barium peroxide has a 50% DT at 388 °C, and the sample with 4% iron oxide and 20% barium peroxide has a 50% DT at 391 °C, only very slightly lower than the 50% DT of sodium chlorate with 4% iron oxide and 4% barium peroxide. Barium hydroxide is also a moderately active catalyst for the decomposition of sodium chlorate. As shown in Figure 2, when 1, 2, 4, and 8% barium hydroxide by weight is loaded, the 50% DTs of the samples are 544, 513, 477, and 450 °C, respectively, compared to the 50% DT of 566 °C for the decomposition of sodium chlorate by itself. This result agrees with the result previously reported (Zhang et al., 1993b). It is clear from Figures 1 and 2 that the decomposition temperature of sodium chlorate is more dependent on the amount of barium hydroxide loading than on the amount of barium peroxide loading. This difference is likely due to the difference in the melting temperature of the two barium compounds. The peroxide is a solid.
1950 Ind. Eng. Chem. Res., Vol. 36, No. 5, 1997
Figure 2. TGA plots of the thermal decomposition of sodium chlorate catalyzed by iron oxide and barium hydroxide: (1) 4% Fe2O3, (2) 4% Fe2O3 and 8% Ba(OH)2‚H2O, (3) 4% Fe2O3 and 9.5% Ba(OH)2‚H2O, (4) 4% Fe2O3 and 20% Ba(OH)2‚H2O, (5) 4% Fe2O3 and 4% Ba(OH)2‚H2O, (6) 4% Fe2O3 and 2% Ba(OH)2‚H2O, (7) 4% Fe2O3 and 1% Ba(OH)2‚H2O, (8) 8% Ba(OH)2‚H2O, (9) 4% Ba(OH)2‚H2O, (10) 2% Ba(OH)2‚H2O, (11) 1% Ba(OH)2‚H2O, and (12) no Fe2O3 or Ba(OH)2‚H2O.
It is difficult for the molten sodium chlorate to diffuse into barium peroxide particles. Therefore, higher loading of barium peroxide only reduces the decomposition temperature slightly. Barium hydroxide, on the other hand, starts melting at about 370 °C. The molten sodium chlorate and barium hydroxide can diffuse into one another. Therefore, increased loading of barium hydroxide has more effect in reducing the decomposition temperature of the chlorate. The effect of barium hydroxide on the decomposition of sodium chlorate catalyzed by 4% iron oxide is similar to that of barium peroxide. When used together with iron oxide, barium hydroxide also becomes an inhibitor. The sample of sodium chlorate with 4% iron oxide and 1% Ba(OH)2‚H2O has a 50% DT at 431 °C compared to 313 °C for the sample of sodium chlorate with 4% iron oxide and without barium hydroxide. The inhibiting effect decreases with increased loading of barium hydroxide up to about 8% by weight. The 50% DTs of sodium chlorate with 4% iron oxide and 1, 2, 4, and 8% barium hydroxide hydrate are 431, 417, 407, and 392 °C, respectively. The effect of barium hydroxide loading on the decomposition temperature appears to saturate at 8% loading of the hydroxide. The samples of sodium chlorate containing 4% iron oxide with 9.5% and 20% barium hydroxide have 50% DTs even slightly higher than the 50% DT of the sample with 8% barium hydroxide. This slightly higher decomposition temperature is probably because of the diluting effect at higher loading of barium hydroxide. Increased barium hydroxide loading reduces the extent of contact between the chlorate and the iron oxide. The weight losses between 100 and 150 °C for some samples shown in Figure 2 are due to the dehydration of the barium hydroxide hydrate. The hydrate Ba(OH)2‚H2O loses H2O to form anhydrous hydroxide Ba(OH)2 in this temperature range as previously reported (Zhang et al., 1993b). The effect that barium peroxide and hydroxide have on iron oxide is similar to the effects they have on cobalt oxide (Zhang et al., 1993b) in the sense that they inhibit the catalytic activity of both of the oxides toward the decomposition of sodium chlorate. However, changes in loading of barium peroxide and hydroxide influence the extent of the inhibiting effect in the opposite way.
The inhibiting effect of barium peroxide and hydroxide on cobalt oxide increases with their loading, even though the inhibiting effect tends to saturate at high loading, whereas the inhibiting effect that barium peroxide and hydroxide have on iron oxide decreases with increased loading. This difference is probably because of the fact that Fe3+ has a different electron configuration from that of Co3+. As previously reported, the catalytic activities of the metal oxides are attributed to the metal cations (Zhang et al., 1993a). A metal cation can attract an unshared electron pair from an oxygen atom in the chlorate group to form a coordination bond. Sodium chlorate starts melting at about 250 °C. The molten sodium chlorate can have good contact with the metal oxide and interact with the metal oxide surface to form surface coordination bonds. Formation of the Mn+-O weakens the Cl-O bond in the chlorate and causes it to rupture, which results in the decomposition of the chlorate. Co3+ ion has a partially filled 3d shell, and the shielding effect of the partially filled 3d shell is not very effective. The unshared electron pair from the oxygen in the chlorate is more readily attracted by the positive charges of the Co3+ nucleus. In addition, Co3+ is in the low spin state, and there are empty 3d orbitals available to accommodate extra electron pairs. Both of these factors make it easy for Co3+ to form a coordination bond with the oxygen atom in the chlorate. Therefore, cobalt oxide is a very active catalyst. Fe3+ in the iron oxide has five 3d electrons and is in the high spin state (Kittel, 1986). The five 3d electrons occupy five 3d orbitals, and all five 3d orbitals are half-filled. According to Hund’s law, ions with half-filled 3d orbitals are more stable and the shielding effect is higher. Therefore, the nuclear positive charge of the Fe3+ is more effectively shielded and thus has a lower tendency to attract electron pairs. In addition, in order to form a surface complex, the unshared electron pair that the Fe3+ attracts will have to go into an outer orbital that has higher energy levels, which is less preferred. Therefore, iron oxide has a lower catalytic activity than that of cobalt oxide (Rudloff and Freeman, 1970; Zhang et al., 1993a). The catalytic process during the decomposition of sodium chlorate is dependent on the formation of a surface coordination bond between the metal ions and the oxygen atoms of the chlorate. The metal cations are Lewis acids, and oxygen atoms in the chlorate act as Lewis bases. Alkaline compounds such as barium peroxide and hydroxide are stronger bases than the oxygen atoms in the chlorate group and compete with the chlorate group ClO3- for access to the transitionmetal ions. This competition reduces the chances for the formation of coordination bonds between the metal ions, Fe3+ or Co3+, and the oxygen atom from the chlorate. Therefore, the presence of an alkaline compound reduces the decomposition rate of the chlorate. On the other hand, in the presence of alkaline-earth elements and an oxidizer, particularly under basic conditions, iron oxide can be oxidized to form ferrates in which iron ions have oxidation states higher than +3, such as Fe4+ in barium ferrate BaFeO3. It has been demonstrated previously that metal ions with partially filled d orbitals have a higher catalytic activity when the d orbitals are not exactly half-filled (Zhang et al., 1993a). Fe4+ has only four 3d electrons, and its 3d orbitals are less than half-filled. Therefore, the catalytic activity is increased when Fe3+ is converted to Fe4+. It is reported that alkaline-earth metal ferrates are active catalysts for the decomposition of sodium chlorate
Ind. Eng. Chem. Res., Vol. 36, No. 5, 1997 1951
(Greer, 1991), which supports the idea that Fe4+ has high catalytic activity. At low loading barium peroxide or hydroxide compete with the chlorate for the active sites on the surface of the iron oxide and inhibit the activity of the iron oxide. Even though some iron oxide may be converted to the ferrate, the inhibiting effect is dominant. When more barium peroxide or hydroxide is loaded, more iron oxide is converted to barium ferrate that has higher activity. The increased activity resulting from the formation of the barium ferrate exceeds the increased inhibiting effect from the increased barium compounds loading and the decomposition occurs at lower temperatures. A further increase in the loading of barium peroxide or hydroxide, however, does not result in more iron oxide being converted to the ferrate and therefore cannot increase the catalytic activity further. The extra barium peroxide and hydroxide loaded is a diluent and prevents the contact between the iron oxide, or ferrate, and sodium chlorate and thus even reduces the activity slightly in the case of barium hydroxide. This may be the reason why the samples containing 4% iron oxide with 9.5% and 20% barium hydroxide have slightly higher 50% DTs than that of the sample with 4% iron oxide and 8% barium hydroxide. The reaction residues of the samples with 4% iron oxide and 4% or more barium peroxide or hydroxide are all black, whereas the unreacted mixtures are orange, the color of iron oxide Fe2O3. The color change from orange to black indicates that the iron oxide has been converted to a different compound, probably barium ferrate. The reaction of barium hydroxide and peroxide with iron oxide is likely controlled by kinetics rather than driven by stoichiometry. The molar ratios of BaO2/ FeO1.5 or Ba/Fe in the samples containing 4% iron oxide and 1, 2, 4, 8, and 20% barium peroxide are 0.12, 0.24, 0.47, 0.94, and 2.36, respectively, whereas the Ba/Fe ratio in BaFeO3 is 1. If the reaction is driven by stoichiometry, 4% Fe2O3 by weight would react with about 8.5% BaO2 by weight and the catalytic activity would plausibly increase with BaO2 loading up to 8.5% by weight. The molar ratios of the samples containing 4% Fe2O3 and 1, 2, 4, 8, 9.5, and 20% Ba(OH)2‚H2O are 0.11, 0.21, 0.42, 0.84, 1.00, and 2.11, respectively. The maximum catalytic activity occurs at a Ba/Fe ratio of 0.84, not 1.00. This is consistent with the hypothesis that the catalytic activity change with barium hydroxide is driven by kinetics not stoichiometry. Barium hydroxide starts melting at about 370 °C, which is lower than the decomposition temperature of the sodium chlorate. Molten barium hydroxide has a higher ability to make contact with the iron oxide. Therefore, the amount of barium hydroxide has more effect than that of barium peroxide, since the latter compound does not melt in the temperature range of interest. To find out whether the hypothetical reaction between iron oxide and barium compounds can actually occur in the presence of sodium chlorate, iron oxide, barium hydroxide, and sodium chlorate were mixed at a molar ratio of 1:1:4 corresponding to 27.2% Ba(OH)2‚H2O and 11.5% Fe2O3 by weight. The mixture was heated in an electric furnace at 10 °C/min to 500 °C and held at 500 °C for 10 min. The reaction residue is black, compared to the orange color before heating. X-ray diffraction analysis indicates that barium hydroxide does react with iron oxide as discussed below. The X-ray diffraction patterns of the reaction products for the sample with Ba(OH)2H2O/Fe2O3/NaClO3 at
Figure 3. X-ray diffraction spectra: (a) Ba(OH)2‚H2O/FeO1.5/ NaClO3 at molar ratio 1:1:4, heated at 10 °C/min to 500 °C; (b) high surface area iron oxide as prepared; (c) data from JCPDS file 5-628 for NaCl; (d) data from JCPDS file 23-1024 for BaFeO3-x.
Figure 4. X-ray diffraction spectra of the reaction products of Ba(OH)2‚H2O/Fe2O3 at a molar ratio of 1: (a) heated at 350 °C for 1 h; (b) heated at 10 °C/min to 500 °C and held for 10 min; (c) 800 °C for 10 h; (d) data from JCPDS file 25-68 for BaFeO2.9; (e) data from JCPDS file 23-1024 for BaFeO3-x.
molar ratio 1:1:4 are presented in Figure 3a. Figure 3b is the X-ray diffraction pattern of the iron oxide we used. Parts c and d of Figure 3 are X-ray diffraction data from JCPDS files for NaCl and BaFeO3-x. The lines of the original iron oxide diminished, and the lines of barium ferrates BaFeO3-x can be observed. This is evidence that the original iron oxide has reacted, most likely with barium hydroxide. As shown in parts c and d of Figure 3, the sodium chloride lines superpose on the lines of the BaFeO3-x, which makes it difficult to see the barium ferrate lines. Therefore, it would be beneficial to see the X-ray pattern of the reaction product of barium hydroxide and iron oxide in the absence of sodium chlorate. Samples of barium hydroxide and iron oxide were mixed at a molar ratio of 1:1, heated at 10 °C/min, and held at 350 °C for 1 h, 500 °C for 10 min, and 800 °C for 10 h, respectively. The X-ray diffraction spectra of the samples heated at 350, 500, and 800 °C are presented in parts a-c of Figure 4, respectively. The X-ray diffraction data from the JCPDS files for BaFeO2.9 and BaFeO3-x are given in parts d and e of Figure 4 as a comparison. Figure 4 clearly indicates that the iron oxide reacts with barium hydroxide to form barium ferrate even in the absence of sodium chlorate. After being heated at 350 °C for 1 h, the iron oxide peaks diminished almost completely. The barium ferrate with formula BaFeO2.9 is readily identifiable and becomes the major phase of
1952 Ind. Eng. Chem. Res., Vol. 36, No. 5, 1997
Figure 5. TGA plots of the thermal decomposition of sodium chlorate catalyzed by iron oxide and barium ferrate: (1) 13.5% barium ferrate; (2) 4% Fe2O3.
the sample. The average oxidation state of the iron atoms in BaFeO2.9 is 3.8. 80% of the iron ions in BaFeO2.9 are Fe4+, and the rest are Fe3+. Fe3+ has been oxidized to Fe4+ by oxygen from the air. This clearly indicates that Fe4+ can be formed from the reaction of iron oxide and barium hydroxide even in the absence of the strong oxidizer sodium chlorate. Figure 2 indicates that the samples of sodium chlorate containing both iron oxide and barium hydroxide show little or no decomposition at 350 °C. Therefore, barium ferrate, or Fe4+, can be formed before decomposition of the sodium chlorate. It should be pointed out that sodium chlorate is a strong oxidizer. The oxidation of Fe3+ to Fe4+ may occur at lower temperatures and proceed faster in the presence of sodium chlorate. The sample heated at 800 °C for 10 h is primarily barium ferrate BaFeO3-x with some impurities. BaFeO3-x is a nonstoichiometric barium ferrate. Therefore, one can expect that, during the course of the decomposition of sodium chlorate, BaFeO2.9 or BaFeO3-x or some phases in between the two barium ferrate phases are present and at least some of the iron ions are present in the form of Fe4+ during decomposition of the chlorate containing both iron oxide and barium hydroxide. Barium peroxide should undergo similar reaction with iron oxide to form barium ferrate. Barium hydroxide melts at about 370 °C and some of its ions should be mobile at 350 °C, while barium peroxide does not melt below the temperature range of interest. Typically, a solid-solid reaction is slower than a solid-liquid reaction. Therefore, the reaction between barium peroxide and iron oxide is probably slower. To compare the catalytic activity, the iron oxide was also heated at 800 °C for 10 h. The surface area of the iron oxide heated at 800 °C for 10 h is 2.7 m2/g compared to 1.1 m2/g for the barium ferrate prepared at 800 °C. A sample of sodium chlorate mixed with 4% by weight of the iron oxide heated at 800 °C and a sample of sodium chlorate mixed with 13.5% barium ferrate by weight, corresponding to 4% iron oxide by weight, were made and studied by thermogravimetric analysis. As shown in Figure 5, the thermogravimetric analysis clearly indicates that barium ferrate does have a higher catalytic activity than iron oxide. The 50% DT of the sample containing barium ferrate is at 443 °C, which is 86 °C lower than the 50% DT of 526 °C for the sample containing the iron oxide, even though the surface area of the iron oxide is slightly higher than that of the
barium ferrate. This result supports the idea that formation of barium ferrate increases the catalytic activity and makes sodium chlorate decompose at lower temperatures. The sample of sodium chlorate with the low surface area iron oxide shows a two-step decomposition. The first step is the decomposition of sodium chlorate, and the second step is likely the decomposition of sodium perchlorate formed through a disproportionation reaction. When no catalyst or low activity catalyst is used, sodium chlorate tends to disproportionate to form the perchlorate during decomposition (Solymosi, 1977). Cobalt ion Co3+ in cobalt oxide already has empty 3d orbitals and thus has a higher activity than iron oxide does. When barium peroxide or hydroxide is loaded, the peroxide and hydroxide group compete with the chlorate for the Co3+ on the surface of the cobalt oxide and thus reduce the catalytic activity of cobalt oxide. It is not possible for cobalt oxide to be converted to a catalytically more active compound. Therefore, increased loading of barium peroxide and hydroxide simply results in more sites being occupied by the peroxide or hydroxide, and the inhibiting effect increases with increased loading of the barium compounds. Conclusions Barium peroxide and hydroxide inhibit the catalytic activity of iron oxide for the decomposition of sodium chlorate. At higher loading of barium peroxide or hydroxide and fixed loading of iron oxide, sodium chlorate decomposes at lower temperatures because iron oxide is partially converted to barium ferrate which contains Fe4+. The barium ferrate has a higher catalytic activity because the 3d shell of Fe4+ in the ferrate is not half-filled and has a lower shielding effect and there is an empty orbital in the 3d shell to accommodate the unshared electron pairs from the oxygen atom in the chlorate. Literature Cited Greer, J. S. Oxygen Generating Candles. U.S. Patent 5,049,306, 1991. Kittel, C. Introduction to Solid State Physics, 6th ed., John Wiley & Sons: New York, 1986. Pappenheimer, J. R. Development of Oxygen Candle Apparatus for Use in Aircraft. Report to the Committee on Medical Research of the Office of the Scientific Research and Development, June 1946. Rudloff, W. K.; Freeman, E. S. The Catalytic Effects of Metal Oxides on Thermal Decomposition Reactions. J. Phys. Chem. 1970, 74, 3317. Solymosi, F. Structure and Stability of Salts of Halogen Oxyacids in the Solid State; John Wiley & Sons: London, 1977. Zhang, Y.; Kshirsagar, G.; Ellison, J. E.; Cannon, J. C. Catalytic Effects of Metal Oxides on the Thermal Decomposition of Sodium Chlorate. Thermochim. Acta 1993a, 228, 147. Zhang, Y.; Kshirsagar, G.; Cannon, J. C. Function of Barium Peroxide in Sodium Chlorate Chemical Oxygen Generators. Ind. Eng. Chem. Res. 1993b, 32, 966.
Received for review October 21, 1996 Revised manuscript received February 7, 1997 Accepted February 10, 1997X IE9606754
X Abstract published in Advance ACS Abstracts, March 15, 1997.