Langmuir 1991, 7, 1660-1674
1660
Air Oxidation of Hydrazine. 1. Reaction Kinetics on Natural Kaolinites, Halloysites, and Model Substituent Layers with Varying Iron and Titanium Oxide and 0Center Contents L. Coyne'lt Department of Chemistry, San Jose State University, San Jose, California 95192
R. Mariner NASA-Ames Research Center, Mail Stop 239-4, Moffett Field, California 94035
A. Rice English China Clays International, P.O.Box 471, Sandersville, Georgia 31082 Received May 4,1990. In Final Form: February 26,1991
Air oxidationof hydrazinewas studied by using a group of kaolinites, halloysites,and substituent oxides as models for the tetrahedral and octahedral sheets. The rate was found to be linear with oxygen. The stoichiometryshowed that oxygen was the primary oxidant and that dinitrogen was the only important nitrogen-containingproduct. The rates on kaolinites were strongly inhibited by water. Those on threedimensional silica and gibbsite appeared not to be. That on a supposedly layered silica formed from a natural kaolinite by acid leaching showed transitional behavior-slowed relative to that expected from a second-order reaction relative to that on the gibbsite and silica but faster than those on the kaolinites. The most striking result of the reaction was the marked increase in the rate of reaction of a constant amount of hydrazineas the amount of clay was increased. This increase was apparent (in spite of the water inhibition at high conversions) over a 2 order of magnitude variation of the clay weight. The weight dependence was taken to indicate that the role of the clay is very important, that the number of reactive centers is very small, or that they may be deactivated over the course of the reaction. In contrast to the strong dependence on overall amount of clay, the variation of amounts of putative oxidizing centers, such as structural Fe(III), admixed Ti02 or Fe203, or 0- centers, did not result in alteration of the rate commensurate with the degree of variation of the entity in question. Surface iron does play some role, however,as samplesthat were pretreated with a reducing agent were less active as catalysts than the parent material. These results were taken to indicate either that the various centers interact to such a degree that they cannot be considered independently or that the reaction might proceed by way of surface complexation, rather than single electron transfers.
Introduction A. Clay-Catalyzed Surface Reactions. Clays are known to produce a variety of reactions mediated by a variety of catalytic sites.l+ Clay-mediated reactions are of voluminously documented importance to geological, agricultural, and environmental chemistry and fossil fuel generation and of frequently postulated importance to the origin of terrestrial life. Mechanisms of surface reactions on clay minerals are intricate because of the presence of multiple types of catalytic sites that can act independently or synergistically. Whether oxidation, reduction, polymerization, or other surface reactions predominate in any given system is a complex function of the reactant, the clay, and the conditions. Clays possess t To whom correspondenceshould be addressed at NASA-Ames Research Center, Mail Stop 239-4, Moffett Field, CA 94035. (1) Solomon,D.H.ClayMineralsaaElectronAcceptorsand/orElectron Donom in Organic Reactions Clays Clay Miner. 1968,16,31-39. (2)Theng, B. K. G. Interactions with Uncharged Polar Organic Compounds in Complex Formation with Some Defined Classes of Compounds. In The Chemistryof Clay 0rganicReactions;AdamHilger, Ltd.: London, 1974; Chapter 3.4. (3) Solomon, D. H.; Hawthorne, D. G. Chemistry of Pigment8 and Fillers: John Wiley and Sons: New York, 1983. (4) Ru rt, J. P.;Oranquiet,W. T.; Pinnavaia,T.J. CatalyticProperties of Clay In Chemistry of Clays and Clay Minerals; Newman, A. C. D., Ed.;Wiley-Interscience and Mineralogical Society: New York, 1987; Chapter 6, pp 275-319.
dkmls.
both electron- and proton-donatinglaccepting centers, both on the edges and on the interlayer faces, each center type having distinctive characteristics,.6 These centers are associated either directly with the crystal structure and defects in it or indirectly with the defect structure, by virtue of the properties of the charge-balancingexchangeable cations. B. Hydrazine Oxidation on Clay Surfaces. Interaction of amino compounds with clay surfaces is a multifaceted process that has been widely studied from a number of points of view112including the formation and reactions of surface complexes or other reactions proceeding via them. Hydrazine is special among the amines in that the complex formed is sufficiently strong as to produce intercalation of the kaolinite 1ayers.B-e Also, its (5) Ramell-Colom, J. A.; Serratoee, J. M. Reactions of Clays with Organic Substances. Chemistry of Clays and Clay Minerals; New", A. C. D., Ed.;Wiley-Interscienceand MineralogicalSociety: New York, 1987; Chapter 8, pp 371-423. (6) van Olphen, H. An Introduction t o Clay Colloid Chemistry, 2nd ed.; John Wiley and Sons: New York, 1963; pp 316. (7) Wek,A.;Thielpape,W.;Goring,R.;Ritter,W.;Schafer,H.KaolinitEinlagerungeVerbmdungen. Roc. Int. Clay Conf. Stockholm Vol I; Rosenqvist,Th., Graff-Petereon,P., Eds.;Pergamon Press: Oxford, 1963; pp 287-306. (8) Ledoux, R. L.; White, J.L. Infrared Studiw of Hydrogen Bonding Interaction Between Kaolinite Surfaces and Intercalated Potawium Acetate, Hydrazine, Fonnamide,and Urea. J. ColloidInterface Sci. 1968, 21, 127-152.
0743-7463/91/2407-1660$02.50/0 0 1991 American Chemical Society
Langmuir, Vol. 7, No. 8, 1991 1661
Kaolinite-Catalyzed Hydrazine Oxidation
addition to kaolinite in air produces observable frothing (simultaneouschemicalreaction) accordingto the reaction
N,H, + 0,
-
2H20 + N, (1) Many of the numerous catalytic centers of clays, most particularly, structural oxidizing entities, might be exp e e d to interact with hydrazine. Since hydrazine intercalates kaolinite, even normally occluded interlayer sites are available as catalysts for the oxidation. Via its effect on the available adsorptive area, the presence of hydrazine may thus increase reactivity of kaolinite for hydrazine, whether special centers or the overall aluminosilicate structure itself mediates the reaction. 1. Transition-Metal Oxides-Structural and Surface. Oxidized iron is a ubiquitous and abundant surface, as well as structural contaminant of both metamorphic and sedimentary kaolinites (1:l clays). The oxidizing power of montmorillonites (21 clays) for other reactions is generally attributed to octahedral Fe(IIII3based on the fact that in typical clays only this form of iron is of measurable intensity by Miissbauer spectros~opy.~~ However, oxidation of hydrazine has been studied on a variety of nontronite surfacesllJ2and was found to be accompanied by reduction of both tetrahedral and octahedral iron in the crystal structure of the clay. The reduced iron is subsequently reoxidized in air. These results show that clays can serve a stoichiometric, rather than a catalytic function for this reaction, with iron as the oxidant. However, it is reasonable to suspect that the iron centers may also serve as catalysts for air oxidation, operating via single, reversible electron-transfer steps. Potential contributions from two extraneous metal oxides must also be delimited in studying catalysis of oxidation reactions on natural kaolinite surfaces. Titanium oxide is an abundant admixed contaminant of sedimentary kaolinites, like those from Georgia. Contamination by iron oxides is more or less universal. 2. Trapped Hole Centers. In clay minerals there exist other structural oxidizing entities, trapped holes, often called 0-centers, in addition to Fe(II1) centers, which are potentially comparable to them in number. 0-centers are trapped electron vacancies (holes) which, in clays, typically are located near isomorphicallysubstituted sites in which Mg or Fe(I1) are substituted for A1 in the octahedral sheet, or Al, or possibly Fe(III), for Si in the tetrahedral sheet of the clay.13 0- centers in clays can be readily produced by ionizing radiation. Thus they can be considered to represent a form of electronicenergy storage, as they are the positive moiety of the trapped separated charge pair produced by electronic excitation. They are commonly referred to as hole centers. C. Scope of Reported Measurements. The purpose of the work described here was to determine the products of the decomposition reaction of hydrazine on kaolinite and the oxidant and perhaps identify some active surface centers for the catalysis. In order to determine whether ~~~
(9) Thompson,J. G.Interpretationof Solid State W and lsgi Nuclear Magnetic Resonance Spectraof KaoliniteIntercalates. Clays Clay Miner. 1886,33, 173-180. (10) Malathi,N.; Puri, 5.P.; Seraswat,I. P. Moesbauer Studiesof Iron in Illite and Montmorillonite. J. Phys. SOC.Jpn. 1969,26,680-683. (11) Ruseell, J. D.; Goodman, B. A.; Fraeer, A. R. Infrared and MOSSbauer Studies of Reduced Nontronite. Clays Clay Miner. 1979,27,6371. (12) Stucki,J. W.; k a r , P. R. Variable Oxidation States of Iron in the Crystal Structure of Smectite Minerals. In Spectroscopic Characterization of Minerals and Their Surfaces; ACS Symposium Series 415; Coyne, L. M., McKeever, 5.W. S., Blake, D., Eds.; American Chemical Society- Wan n, DC,1989; Chapter 17, pp 330-360. (13) Angel,% Jones, J. P. E.; Hall, P. L. ElectronSpin Resonance Studies of Doped Synthetic Kaolinites. Clay Miner. 1974,10,247-256.
CHEMICAL BLEACH
BLEACHED
,
MAGNETIC SEPARATION
SEPARATE
SEPARATE
BLEACHED NONMAGNETIC SEPARATE
BLEACHED MAGNETIC SEPARATE
Figure 1. Process used to prepare catalytic kaolinites. The numbers 1-6 indicate final products for which the reactivity was tested.
the activity of the surface for hydrazine oxidation could be related simply to the overall population of electronaccepting centers, we have varied the concentration of several oxidizing entities in the clay by using a number of natural kaolinites, halloysites, and model substituent oxides having different amounts of structural and surface iron and hole centers. These studies serve as preparation for studies in which the hole center content can be altered independently of other structural constituents by y irradiation. Various possible complicating effects of intercalation are controlled by a period of preintercalation of the clay before initiating the reaction. The effects of moisture are minimized, at least for the initial stage of the reaction, by drying both the clay surface and the hydrazine prior to the reaction. Edge sites were not blocked in these experiments, but a single control experiment with phosphate-treated clays showed these treated clays to have comparable activity to the clays studied here. We describe here the methods and initial data from our studies of kaolinite-catalyzed air oxidation of hydrazine. The approach outlined gives simultaneous consideration to multiple important factors, allowing some separation of their influences to the degree to which they may, in fact, be separable. Relative initial rate data are given, which reveal a striking and complex role for the clay and an unexpectedly subsidiary role for iron in promoting this reaction.
Experimental Section A. Materials. 1. Clays. A number of natural kaolinites
and metahalloysites varying in the amount of structural iron and energy stored as hole centers was selected for these studies.The kaolinites were achemicdy untreatedCornish material,reference number RLO 2101, for which the properties are summarized in ref 14, and two Georgia kaolinites, from different locations in Washington County, all provided by English China Clays International. The Georgia kaolinites were further treated by magnetic separation in hopes of removing admixed iron and titanium oxides and by chemical bleaching to reduce the surface ferric iron to the ferrous oxidation state. The pretreatment method for the Georgia materiale examined is shown schematically in Figure 1. The samples were selected, in part, for (14) Coyne, L.;Pollock, G.; Kloepping, R. Room Temprature Luminescence from Kaolin Induced by Organic Amines. Clays Clay Miner. 1984, 32, 58-67.
Coyne et al.
1662 Langmuir, Vol. 7, No. 8, 1991
Table I. Extraneous Oxide Phases in Selected and Pretreated Catalytic Kaolinites % Fe 0
material Comieh kaolinite, ECC, RLO 2101,untreated Georgia low iron wries, AAC lot 1, 51-03 untreated eeparated/bleached, 6203 lot 2 untreated, 6378 separated only, 6379 separted/bleached 6380 magnetic residue, 6381 and bleached magnetic residue Georgia high iron wries, AAC lot 1, 51-05 untreated wparated/bleached, 6204 lot 2 untreated, 6382 separated only,6379 separated/ bleached, 6384 magnetic residue, 6385 and bleached magnetic residue
% Fez02
0.408
% Ti02 0.18
0.40f 0.03 1.50 f 0.1 0.36 f 0.03 1.4 f 0.1 0.34f 0.03 0.34f 0.03 0.32 f 0.02 0.69 f 0.04
1.5 f 0.1 1.2 f 0.08 1.30 i 0.09 2.8 f 0.2
+ % 'h& 0.59
1.9 1.8 1.8 1.5 1.6 3.4
0.97 f 0.7 1.8 f 0.1 0.97 f 0.07 1.4 f 0.1
2.8 2.4
1.07 f 0.08 1.05 0.08 1.03 f 0.07 1.4 f 0.1
2.6 2.5 2.4 3.0
*
1.5 f 0.08 1.4i 0.05 1.4f 0.1 1.6 A 0.1
Table 11. Proportion of Ferrous Iron in Several Kaolinites Examined total iron, Fe(I1) as wt % clay of total iron as wt % Fez03 RLO 2101 21.36 0.436 f 0.065 Ga 6378 8.20 0.346 f 0.010 Ga 6380 8.92 0.303 f 0.016 Ga 6382 5.92 1.064 f 0.019 Ga 6384 7.31 1.064 A 0.019 variability in the hole center content, but the prime selection criterion for the two Georgia clays waa to maximally vary iron while holding the Ti02 content constant, so as to reduce the number of independent compositional variables by the greatest possible degree. The total iron contents were determined by using an X Met 820 X-ray fluorescence (XRF) analyzer with 5 g of clay in S/4 in. x 4 mm planchettes. The samples were irradiated with a W i source with an activity of 30 mCi for 75 a. The measured values can be trusted to *7 % of the reported value with a confidence level of 99.7%. The iron/titanium analysis data for the kaolinites are summarized in Table I. In the case of the Georgia materials, data were taken on samples from two lots of the same materials. For five of the materials, RLO 2101, 6378, 6380, 6382, and 6384, an independent analysis was done by acid digestion, complexation, and spectrophotometry using methods outlined in ref 12. Not only was total iron determined for these samples but also the ferrous/ferric ratio. The results of the spectrophotometric analyses are summarized in Table 11. Given the low amounts of iron in these materials, and that the value for RLO 2101 in Table I was performed in yet a third laboratory (also by XRF), the agreement between these values is impressive. It implies that the 18%difference in the chemical analyses for the two lob of low-iron Georgia kaolinite is attributable to variability in these lots of natural material, rather than to either analytical error or variability between replicate samples. Throughout the series, the overall iron content varies from 0.32 to 1.4% or by a factor of 4.4-fold. The titanium oxide content differs between the Cornish and Georgia materials by 16-fold. Clearly, from Tables I and 11,for these low levels of iron and titanium oxides, the magnetic separation method is of marginal utility, only the magnetic residue showing large alterations in the amounts of iron and titanium. The chemical reduction process (bleaching) however, does make a long-term alteration in the ferrous/ferric ratio, as shown in Table 11. The kinetic measurements were performed long after the preparation, and the chemical analysis was performed well after the kinetic measurements. The analysis should thus be seen as providing a lower limit on the amount of ferrous iron present during the
measurement of reaction rates. The results of the bleaching on individual species of iron and hole centers will be discussed in the EPR section. The surface areas of the three parent materials have been measured to be 9.8 m2/g for the Cornish, 18.8 m2/g for the high iron Georgia, and 16.5 m2/g for the low iron Georgia. These differences are small, but perhaps not negligible between the unintercalatsd materials in terms of degree of surface coverage at various clay weights. However the effects of these differences in unintercalated surface area are negligible for an intercalated material. They also are negligible in comparison to the differences among the iron, titanium, and the hole center contents. The selected metahalloysites were of two morphologies, Matauri Bay (long tubes) and Te Puke (short tubes), with some platelike portion, and Opotiki (spherical). All three of these materials are from New Zealand and were provided and characterized by B. K. G. Theng of the Soil Bureau in Lower Hutt, New Zealand. A complete description of these halloysites can be found in ref 15. The iron oxide contents as % Fez08 of the untreated materials are 0.25% (Matauri Bay, M.B.), 3.4% (Te Puke), and 4.18% (Opotiki), thus placing the iron content of M.B. under the lowest of the kaolinites, and those of Te Puke and Opotiki above the highest. The titanium oxide contents of the untreated materials are 0.097,0.37, and 0.28%, respectively, placing that of M.B. comparable to the Cornish clay and those of the other two low with respect to the Georgia Materials. In addition, these materials were subjected to a surface deferration process and separated into two particle size fractions. The iron contenta are 0.125 % for the C0.5 pm fraction of Matauri Bay and 2.08% for the
