Separation of Adsorbed and Intercalated Hydrazine in Hydrazine

A comparative study of the dehydroxylation process in untreated and hydrazine-deintercalated dickite. F. Franco , M. D. Ruiz Cruz. Journal of Thermal ...
1 downloads 0 Views 76KB Size
1244

Langmuir 2002, 18, 1244-1249

Separation of Adsorbed and Intercalated Hydrazine in Hydrazine-Hydrate Intercalated Kaolinite by Controlled-Rate Thermal Analysis Ja´nos Kristo´f,*,† Ray L. Frost,‡ Wayde N. Martens,‡ and Erzse´bet Horva´th§ University of Veszpre´ m, Department of Analytical Chemistry, H-8201 Veszpre´ m, P.O. Box 158, Hungary, Centre for Instrumental and Developmental Chemistry, Queensland University of Technology, 2 George Street, GPO Box 2434, Brisbane, Queensland 4001, Australia, and Research Group for Analytical Chemistry, Hungarian Academy of Science, H-8201 Veszpre´ m, P.O. Box 158, Hungary Received July 25, 2001. In Final Form: November 16, 2001 The thermal behavior of a low-defect kaolinite fully expanded with hydrazine-hydrate has been investigated in a nitrogen atmosphere at a constant, preset decomposition rate of 0.15 mg/min. Under controlled-rate thermal analysis (CRTA) conditions, it was possible to distinguish between loosely bonded (adsorbed) and strongly bonded (intercalated) reagent. The loosely bonded reagent is connected to the internal and external surfaces of the expanded mineral and is present as a space filler between the sheets of the delaminated mineral. The strongly bonded hydrazine-hydrate is connected to the kaolinite innersurface OH groups by the formation of hydrogen bonds. Based on the thermoanalytical results, three different types of bonded reagent could be distinguished in the complex. Type 1 reagent (approximately 0.20 mol hydrazine-hydrate/mol inner-surface OH) is liberated between approximately 50 and 70 °C. Type 2 reagent is lost between approximately 70 and 85 °C, corresponding to a quantity of 0.12-0.15 mol hydrazine-hydrate/mol inner-surface OH. Type 3 reagent is lost in the 85-130 °C range, amounting to some 0.30 mol hydrazine/mol inner-surface OH. The quantity of this third type of reagent is independent of the conditions of sample pretreatment (drying). The liberation of bonded hydrazine-hydrate can be followed by FT-IR (DRIFT) spectroscopy in the OH and NH stretching ranges as well. When the complex is heated to 70 °C under CRTA conditions, a new reflection appears in the XRD pattern with a d-value of 9.6 Å, in addition to the 10.3 Å reflection. This new reflection disappears in contact with moist air and the complex re-expands to the original d-value of 10.3 Å. The appearance of the 9.6 Å reflection is interpreted as the expansion of kaolinite with hydrazine alone, while the 10.3 Å one is due to expansion with hydrazinehydrate.

Introduction The industrial application of kaolinite, an important industrial raw material, is closely related to its reactivity and surface properties. The reactivity of kaolinite internal surfaces can be tested by intercalation, that is, via the insertion of low molecular weight organic compounds (e.g., potassium acetate, formamide, dimethyl sulfoxide) between the kaolinite layers consisting of the two-dimensional arrangements of tetrahedral (siloxane) and octahedral (gibbsitic) sheets.1,2 The mechanism of intercalation is not known with certainty, but it involves the disruption of the hydrogen bonding between the kaolinite layers and the formation of new hydrogen bonds between the kaolinite inner surfaces and the inserted molecule. Molecules such as formamide and acetamide may connect to kaolinite inner surfaces through the CdO group, the lone pair of nitrogen, or the amide group. It was reported quite early that formamide readily intercalates kaolinite.3 Based on the results of Raman microscopy and Fourier transform infrared (FT-IR) spectroscopy, new models have been proposed for the intercalation of formamide in kaolinites.4 Thermoanalytical studies showed * To whom correspondence should be addressed. Phone/fax: +36 88 421 869. E-mail: [email protected]. † University of Veszpre ´ m. ‡ Queensland University of Technology. § Hungarian Academy of Science. (1) Wada, K. Am. Mineral. 1962, 46, 78. (2) Weiss, A. Angew. Chem. 1962, 73, 736. (3) Ledoux, R. L.; White, J. L. J. Colloid Interface Sci. 1966, 21, 127.

that the removal of intercalated formamide below 300 °C is a complex process and that the liberation of formamide from the complex takes place in two overlapping stages.5 Later on, it was also observed that a higher amount of formamide can be connected to the clay if intercalation is carried out in the presence of water.6 The complexity of the thermal decomposition patterns and the subtleties of the vibrational spectroscopic (FT-IR and Raman spectrometric) data7 required a detailed study of the thermal decomposition mechanism. With the use of controlledrate thermal analysis (CRTA) methodology, it was possible to distinguish between loosely bonded (adsorbed) and strongly bonded (intercalated) formamide.8 Also, the effect of water on the intercalation process could be clarified. The possibility to intercalate hydrazine into kaolinite has been known for a considerable length of time. Ledoux and White reported the expansion of kaolinite from 7.2 to 10.4 Å upon introducing hydrazine in the kaolinite structure.3 Mild heating resulted in deintercalation accompanied with the partial collapse of the structure to 9.4 Å. Johnston and Stone showed the effect of evacuation on the kaolinite-hydrazine complex with the subsequent (4) Frost, R. L.; Lack, D. A.; Paroz, G. N.; Tran, T. H. T. Clays Clay Miner. 1999, 47, 297. (5) Kristo´f, J.; Frost, R. L.; Kloprogge, J. T.; Horva´th, E.; Ga´bor, M. J. Therm. Anal. Calorim. 1999, 56, 885. (6) Frost, R. L.; Kristo´f, J.; Horva´th, E.; Kloprogge, J. T. Spectrochim. Acta, Part A 2000, 56, 1711. (7) Frost, R. L.; Kristo´f, J.; Horva´th, E.; Kloprogge, J. T. Spectrochim. Acta, Part A 2000, 56, 1191. (8) Kristo´f, J.; Horva´th, E.; Frost, R. L.; Kloprogge, J. T. J. Therm. Anal. Calorim. 2001, 63, 279.

10.1021/la011179+ CCC: $22.00 © 2002 American Chemical Society Published on Web 01/12/2002

Separation of Adsorbed and Intercalated Hydrazine

collapse of the structure from 10.4 to 9.6 Å.9 A new model was proposed for hydrazine intercalation based on the insertion of a hydrazine-water unit by Frost et al.10 The thermal behavior of hydrazine-intercalated kaolinite shows a close similarity to that of the formamideintercalated mineral. In addition to the involvement of water in the intercalation process (and in the structure of the complex as well), hydrazine is also liberated from the intercalated clay in two overlapping stages.5 Ruiz Cruz and Franco identified three stages for the loss of water + hydrazine in the 25-200 °C range by TG-DTA measurements.11 In addition, with high-temperature X-ray diffraction measurements, they could follow a structural rearrangement of the complex, which initially caused a contraction of the basal spacing from 10.4 to 9.6 Å. Upon further heating, a third stage of contraction was reported to a basal spacing of 8.5 Å. In this study, a combination of controlled-rate thermal analysis with X-ray diffraction and FT-IR (DRIFT) spectroscopy is used to follow the structural changes in hydrazine-intercalated kaolinite on heating. The objective of this work is to distinguish between differently bonded hydrazine molecules in the complex as well as to reveal the role of water in the intercalation process. Experimental Section Preparation of Intercalates. The kaolinite mineral used in this study is a highly ordered kaolinite from Kira´lyhegy in Hungary. The mineral has been previously characterized with X-ray diffraction and Raman microscopy.10 The intercalate, using the 2-20 µm sized fraction of the clay obtained by sedimentation, was prepared by mixing 500 mg of the kaolinite with 10 cm3 of an analytical grade, 85% hydrazine-hydrate aqueous solution (Reanal Budapest, Hungary) for 80 h at room temperature, magnetically stirred in a closed ampule. Immediately before analysis, the excess solution was decanted and the complex was dried under either a vacuum or flowing nitrogen for a specified period of time. X-ray Diffraction. The X-ray diffraction (XRD) analyses were carried out on a Philips PW 1050/25-type vertical goniometer equipped with a graphite diffracted beam monochromator. The radiation used was Cu KR from a long fine focus Cu tube, operating at 40 kV and 25 mA. Spectra were recorded in stepscan mode with steps of 0.02° 2θ and a counting time of 2 s. Diffuse Reflectance Fourier Transform Infrared Spectroscopic (DRIFT) Analyses. FT-IR (DRIFT) spectra of the samples heated to different temperatures were recorded by means of a Bio-Rad FTS 60A-type spectrometer. Scans (512) were coadded at a resolution of 2 cm-1. Approximately 3 mg of sample was dispersed in 100 mg of oven-dried, spectroscopic grade KBr. Reflected radiation was collected at ∼50% efficiency. Background single-beam KBr spectra were obtained, and the averaged sample single-beam spectra were divided by those of the background. The DRIFT accessory used is of the so-called “praying monk” design accommodating powdery samples mixed with KBr in the sample cap (3 mm deep, 6 mm in diameter). The reflectance spectra expressed as Kubelka-Munk unit versus wavenumber curves are very similar to absorbance spectra and can be evaluated accordingly. Thermal Analysis. Thermal decomposition of the intercalate was carried out in a Derivatograph PC-type thermoanalytical instrument (Hungarian Optical Works, Budapest, Hungary) in a flowing nitrogen atmosphere (250 cm3/min) at a preset, constant decomposition rate of 0.15 mg/min. (Below this threshold value, the samples were heated under dynamic conditions at a uniform rate of 0.5 °C/min.) The samples were heated in an open platinum crucible at a rate of 0.5 °C/min up to 300 °C. With the quasiisothermal, quasi-isobaric heating program of the instrument, (9) Johnston, C. T.; Stone, D. A. Clays Clay Miner. 1990, 38, 121. (10) Frost, R. L.; Kloprogge, J. T.; Kristo´f, J.; Horva´th, E. Clays Clay Miner. 1999, 47, 732. (11) Ruiz Cruz, M. D.; Franco, F. Clays Clay Miner. 2000, 48, 63.

Langmuir, Vol. 18, No. 4, 2002 1245

Figure 1. Mass loss (TG), rate of mass (DTG) and differential thermal analysis (DTA) curves of hydrazine-hydrate intercalated kaolinite under dynamic heating at 5 °C/min in a nitrogen atmosphere (sample mass is 87.97 mg). the furnace temperature was regulated precisely to provide a uniform rate of decomposition in the main decomposition stage.

Results and Discussion The thermal behavior of the hydrazine-hydrate intercalated kaolinite in a nitrogen atmosphere under dynamic heating conditions (5 °C/min heating rate) was investigated previously.5 The intercalating reagent is liberated from the heated sample in two overlapping stages at 100 and 138 °C (Figure 1). A third mass loss step can be identified as a shoulder in the derivative thermogravimetric (DTG) and the differential thermal analysis (DTA) curves at 70 °C. Dehydroxylation of the thermally deintercalated mineral takes place in the 414-650 °C temperature range. Under the conditions of dynamic heating, however, a better resolution of the closely overlapping mass loss stages cannot be obtained. Therefore, controlled-rate thermogravimetric experiments were performed providing time enough for the inherently slow transport of heat between the furnace chamber and the sample and for the establishment of an equilibrium between gas-phase diffusion processes of opposite direction (transporting decomposition products from the sample to the furnace chamber and nitrogen to the sample to replace gas-phase decomposition products). When a constant rate of decomposition (a straight line in the DTG signal) is attained, the state of equilibrium in the opposite gasphase diffusion processes has been reached as well. The thermoanalytical curves of 455.56 mg of wet (decanted) hydrazine-hydrate intercalated kaolinite heated previously in the thermobalance until a mass loss of 100.0 mg has occurred at a preset rate of 0.15 mg/min (to remove the major part of the liquid phase) are given in Figure 2a. On prolonged heating, the temperature increased to 61 °C, and then a spontaneous decrease of the decomposition rate was observed at 69 °C. With further heating, a decomposition stage at 77 °C and another one at 113 °C were observed. At approximately 140 °C, the complex is completely decomposed. The deintercalated kaolinite was heated to 700 °C in order to determine the amount of water released during dehydroxylation of the mineral. The thermal behavior of the kaolinite fully expanded by hydrazine-hydrate (as evidenced by X-ray diffraction measurements) can be interpreted as follows. Adsorbed (surface-bonded) hydrazine-hydrate is lost in a quasiisothermal, equilibrium step until 69 °C. The adsorbed reagent is connected to the outer and inner surfaces of the

1246

Langmuir, Vol. 18, No. 4, 2002

Kristo´ f et al.

Figure 2. Controlled-rate thermal analysis experiment of (a) wet hydrazine-hydrate intercalated kaolinite, (b) hydrazine-hydrate intercalated kaolinite dried under flowing nitrogen before CRTA analysis, (c) hydrazine-hydrate intercalated kaolinite dried under soft vacuum before CRTA analysis, and (d) hydrazine-hydrate intercalated kaolinite dried under strong vacuum before CRTA analysis.

fully expanded mineral. Weakly bonded hydrazine can be connected to the siloxane layer through the hydrogen atoms, can be present as a space filler, or can be inserted into the ditrigonal cavity of the siloxane layer.9 The mass loss stages between 69 and 91 °C as well as between 91 and 140 °C belong to strongly bonded hydrazine lost in equilibrium but nonisothermal processes. These hydrazine molecules form hydrogen bonds with the inner-surface OH groups (intercalated hydrazine). Based on earlier results with FT-IR and Raman spectroscopy, the innersurface OH groups are infrared active and Raman inactive in the intercalation complex. Based on this observation, the intercalation of a hydrazine-water unit (as [NH2NH3]+OH-) was proposed.10 TG-MS results also showed that water and hydrazine are lost simultaneously but with an “offset” from the heated complex (water is released at a lower temperature than hydrazine, resulting in the presence of water-free hydrazine in the last stage of decomposition5). In a parallel experiment, heating was stopped at 70 °C and the quenched sample was immediately subjected to XRD analysis (Figure 3a). The XRD pattern of the partially heated complex showed two reflections with d-values of 10.3 and 9.6 Å. When the sample was exposed to room air for some 15 h, the 9.6 Å reflection almost disappeared and the complex re-expanded to the original d-value of 10.3 Å (Figure 3b). This phenomenon can be explained by the uptake of water. Considering the results of Johnston and Stone with the intercalation of water-free hydrazine resulting in an expansion of kaolinite to 9.6 Å,9 the following conclusion can be drawn. When hydrazinehydrate is intercalated, kaolinite expands to 10.3 Å. When

the clay reacts with pure hydrazine, an expansion of the kaolinite to 9.6 Å occurs. Knowing the amount of dehydroxylation water and supposing that the inner hydroxyls are not accessed by the intercalating reagent, the amount of bonded reagent is 0.09 mol hydrazine-hydrate/mol innersurface OH (77 °C step) and 0.32 mol hydrazine/mol innersurface OH (113 °C step). The thermoanalytical curves of 200.75 mg of intercalate dried in flowing nitrogen for 2 h before heating was started are given in Figure 2b. In this case, with a “powdery” sample, a better resolution of the overlapping decomposition processes could be achieved. In addition to the two types of bonded reagent lost at 77 and 115 °C corresponding to 0.12 mol hydrazine-hydrate/mol inner-surface OH and 0.31 mol hydrazine/inner-surface OH, respectively, another mass loss stage of a less strongly bonded reagent can be identified in the DTG curve at 59 °C. The amount of the intercalation reagent lost in this step (between 51 and 68 °C) is 0.20 mol hydrazine-hydrate/mol inner-surface OH. This means that three types of bonded hydrazine can be identified in a completely expanded kaolinite structure. There are two different explanations for having bonded reagents of different strengths in the complex. Since the inner-surface OH groups completely lose their Raman activity upon forming hydrogen bonds with hydrazine molecules and in the infrared spectrum only one hydrogenbonded OH band appears at 3626 cm-1,10 a uniform structure can be supposed at room temperature. When the FT-IR (DRIFT) spectrum of the complex after heating the sample to different temperatures is recorded, only one type of hydrogen-bonded OH band (at 3626 cm-1) can be identified in the spectra, closely overlapped with the

Separation of Adsorbed and Intercalated Hydrazine

Langmuir, Vol. 18, No. 4, 2002 1247

Figure 3. X-ray diffraction patterns of hydrazine-hydrate intercalated kaolinite: (a) CRTA-treated to 70 °C and (b) CRTA-treated to 70 °C and exposed to air for 15 h.

inner OH band at 3620 cm-1. Since the heated complex cannot be uniform with respect to the location of reagent molecules in the interlayer space as evidenced by the thermal behavior, the conclusion can be drawn that although at least three different bonding patterns exist, they cannot be distinguished by vibrational spectroscopic means based on the differences in the strengths of hydrogen-bonding interactions. Figure 2c shows the result of the CRTA analysis of 310.00 mg of intercalate after drying in a “soft” vacuum (approximately 0.1 bar) for 30 min. Under these conditions of sample preparation, the presence of type 1 bonded reagent cannot be observed (this band is completely overlapped with that of loosely bonded hydrazine-hydrate). The amount of the bonded reagent is 0.14 mol hydrazinehydrate/mol inner-surface OH (type 2) as well as 0.29 mol hydrazine/mol inner-surface OH (type 3). When the same experiment was repeated after drying in a “strong” vacuum (approximately 0.01 bar) for 30 min, a different decomposition pattern was obtained (Figure 2d). The preset decomposition rate was reached in the second step of decomposition only (at 113 °C). In this step, the amount of bonded reagent was 0.32 mol hydrazine/mol innersurface OH. It means that under a stronger vacuum not only the loosely bonded but also part of the hydrogenbonded reagent is lost, with the exception of type 3 reagent the quantity of which is independent of the conditions of drying. DRIFT Spectra of the NH Stretching Region. The NH and OH stretching regions of hydrazine and water are shown in Figure 4, and the results of the band component analyses are shown in Table 1. The figure displays the DRIFT spectra of the NH stretching region of hydrazine-intercalated kaolinites: (a) at 25 °C under dry nitrogen; after controlled rate thermal analysis treatment to (b) 50 °C, (c) 70 °C, and (d) 85 °C. Significant differences may be observed between the DRIFT spectra of these four hydrazine-intercalated kaolinites. Treatment under Dry Nitrogen. When hydrazineintercalated kaolinite is treated under dry nitrogen, the DRIFT spectrum (spectrum a) is characterized by a broad band profile centered around 3270 cm-1 upon which

Figure 4. DRIFT spectra of the N-H stretching region of hydrazine-hydrate intercalated kaolinite at 25 °C in dry nitrogen (a) and CRTA-treated to 50 °C (b), 70 °C (c), and 85 °C (d).

several features are superimposed. These features include a doublet of bands at 3356 and 3362 cm-1 which are assigned to NH stretching vibrations, a low-intensity band at 2887 cm-1, and a second low-intensity band at 3300 cm-1. The two bands observed at 3362 and 3356 cm-1 are attributed to the antisymmetric NH stretching vibrations.

1248

Langmuir, Vol. 18, No. 4, 2002

Kristo´ f et al.

Table 1. Results of the Band Component Analysis of the DRIFT Spectra of the NH Stretching Region of CRTA-Treated Hydrazine-Hydrate Intercalated Kaolinite band characteristics

ν12

ν11

ν10

ν9

ν8

ν7

ν6

ν5

ν4

band position (cm-1) bandwidth (cm-1) % relative area

Hydrazine-Intercalated Kaolinite Dried under Flowing Nitrogen at 25 °C 2887 2976 3058 3183 3273 3300 3356 3362 27.4 56.4 395 59.20.87 298 8.7 11.7 7.8 0.1 0.23 30.7 51.5 0.10 0.63 0.43

band position (cm-1) bandwidth (cm-1) % relative area

2889 41.6 0.44

Hydrazine-Intercalated Kaolinite CRTA-Heated to 51 °C 2959 3036 3178 3244 3301 3359 79.8 193 48.2 245 7.5 10.4 0.93 9.9 0.41 55.2 0.12 2.0

3363 5.7 0.42

band position (cm-1) bandwidth (cm-1) % relative area

2885 43.1 0.26

Hydrazine-Intercalated Kaolinite CRTA-Heated to 70 °C 2921 2957 3037 3185 3266 3300 3354 27.0 25.4 403 50.2 276 8.3 17.2 1.4 0.37 26.7 0.6 45.4 0.07 0.34

3360 10.0 0.71

band position (cm-1) bandwidth (cm-1) % relative area

2889 41.5 0.23

Hydrazine-Intercalated Kaolinite CRTA-Heated to 85 °C 2958 3055 3190 3260 3300 3358 77.1 363 66.9 279 9.6 10.1 0.4 21.5 1.9 47.5 0.13 0.98

3362 6.2 0.58

The observation of two bands means that two different types of NH2 units are present. The slight differences in wavenumbers mean that the two NH2 units are involved with slightly different hydrogen-bonding strengths. Several models are possible: one likely model is that one NH2 unit hydrogen bonds to the oxygen of the siloxane layer through the hydrogen and the second NH2 unit bonds to the inner-surface hydroxyls of the kaolinite through the lone pair of electrons on the nitrogen. If the assumption is made that the 3362 cm-1 band is due to weaker hydrogen bonding, then this may be attributed to the bonding between the lone pair of electrons on the nitrogen to the inner-surface hydroxyl. It is noteworthy to study the variation in relative intensity of these two bands. For the dry nitrogen treated hydrazine-intercalated kaolinite, the ratio of intensities of the 3362 to 3356 cm-1 bands is 2:3. In other words, there are about equal amounts of bonding between the siloxane and hydroxyl surfaces of the kaolinite. Treatment of the Hydrazine-Intercalated Kaolinite Using Controlled-Rate Thermal Analysis Technology to 51 °C. Upon CRTA treatment of the hydrazineintercalated kaolinite to 51 °C (spectrum b), in addition to the features observed in spectrum a, two additional bands are observed at 3481 and 3514 cm-1. These two bands are assigned to the hydroxyl stretching vibrations of water molecules involved in the hydrazine-intercalation complex. The broad band observed at 3270 cm-1 in spectrum a appears to shift to lower wavenumbers and is now observed at 3244 cm-1. The band also is narrower. The two bands observed at 3356 and 3362 cm-1 are shifted to 3359 and 3363 cm-1. Some intensity variation is noted between these two bands with an intensity ratio of ∼2:1. Treatment of the Hydrazine-Intercalated Kaolinite Using Controlled-Rate Thermal Analysis Technology to 70 °C. Upon CRTA treatment of the hydrazineintercalated kaolinite to 70 °C (spectrum c), two additional bands over and above the bands reported for spectra a and b are observed at 2885 and 2921 cm-1. A low-intensity band is also observed at 2969 cm-1. Two sharp bands are observed in spectra b and c at 3514 and 3481 cm-1 which are absent in spectra a and d. When the hydrazineintercalated complex is CRTA-treated to 70 °C, it collapses to the 9.6 Å phase. In the DRIFT spectra of this complex, two additional bands are observed at 2885 and 2921 cm-1. These bands are assigned to N-H vibrations of hydrazine involved in strong hydrogen bonding with the kaolinite surface. These bands are not observed in the CRTA-treated hydrazine-intercalated complex at 85 °C. Some low-

ν3

ν2

3446 85.0 1.13 3419 161 13.6

3417 20.0 0.06

ν1 3532 124.6 2.23

3481 27.1 1.5

3514 20.8 0.57

3481 18.7 0.41

3512 16.5 0.16 3448 232 6.7

intensity broad bands are observed in about these positions for the nitrogen-dried hydrazine-intercalated complex. Johnston et al. observed a broad band centered at 2975 cm-1, which is shifted by some 275 cm-1 from the normal N-H symmetric stretching wavenumber of liquid hydrazine.12 In spectra a-c, we also observe a broad band at 2976 cm-1 for the dry nitrogen treated hydrazineintercalated kaolinite and at 2959 and 2957 cm-1 for the 51 and 70 °C CRTA-treated hydrazine-intercalated kaolinites. When the hydrazine-intercalated kaolinite is CRTA-treated to 50 °C, the ratio of the two bands at 3356 and 3362 cm-1 is now 1:5. Thus, under these conditions the hydrazine is preferentially bonding to the siloxane surface. Upon CRTA treatment to 70 °C, the ratio becomes 2:1. Hence, the hydrazine is forced to hydrogen bond to the hydroxyl surface. This is likely when the hydrazineintercalation complex collapses to the 9.6 Å phase. Upon CRTA treatment to 85 °C, the ratio becomes 2:3, the same as for the dry nitrogen treated hydrazine-intercalated kaolinite. Treatment of the Hydrazine-Intercalated Kaolinite Using Controlled-Rate Thermal Analysis Technology to 85 °C. The spectrum of the 85 °C CRTA-treated hydrazine-intercalated kaolinite (spectrum d) strongly resembles the spectrum of the dry nitrogen treated hydrazine-intercalated kaolinite. The bands at 2885 and 2921 cm-1 are absent when the hydrazine-intercalated kaolinite is either CRTA treated to 85 °C or exposed to dry nitrogen for 1 h. The results reported in this paper are in harmony with the results of Johnston et al. These workers suggest that upon partial collapse of the hydrazineintercalation complex to 9.5 Å, the broad band at 2975 cm-1 is no longer observed but is replaced by a band at 3270 cm-1. We also observe a broad band at 3270 cm-1 for the dry nitrogen treated hydrazine-intercalation complex. The band appears to shift to lower wavenumbers upon CRTA treatment and is observed at 3244, 3266, and 3260 cm-1. However, we observe this broad band for all the CRTA-treated intercalation complexes. Heating the intercalate under CRTA conditions to 70 °C results in the collapse of the complex to 9.6 Å, and at the same time two new bands are observed at 2885 and 2921 cm-1. These two bands are assigned to the NH vibrations of very strongly hydrogen bonded hydrazine molecules. Such increased hydrogen bonding strength would be brought about through the reduction in interlamellar space by 0.70 Å. (12) Johnston, C. T.; Bish, D. L.; Eckert, J.; Brown, L. A. J. Phys. Chem. 2000, 104, 8080.

Separation of Adsorbed and Intercalated Hydrazine

Johnston et al. also reported two bands at 3368 and 3363 cm-1 for the 10.3 Å phase of hydrazine-intercalated kaolinite and the coalescence of these two bands into a single band at 3363 cm-1 for the 9.5 Å phase.9,12 In this work, we find two bands at 3356 and 3362 cm-1, which are present for both the 10.3 and 9.6 Å phases. These two bands are assigned to two different NH2 units, and it has been proposed that one NH2 unit bonds to the siloxane layer and the second to the hydroxyl layer. The intensity of the two bands varies according to the conditions and treatment of the hydrazine-intercalation complex. In addition to the 9.6 Å phase, we find two additional bands at 2885 and 2921 cm-1. In previous work, we therefore proposed a model based on the formation of [NH2-NH3]+[OH]- units as hydrazine functions as a weak monoacid base forming a monohydrate. The interaction of the hydrazine complex occurs between the negative charge on the OH group and the inner-surface hydroxyls. This suggests that the band occurs at 3626 cm-1 because the interaction of the [NH2-NH3]+[OH]- unit and the innersurface hydroxyls of the kaolinite is weak. A second interaction can occur between the hydrogen atoms of the hydrazine and the siloxane layer. Whereas the hydrated part of the hydrazine molecule bonds to the inner-surface hydroxyls, the opposing end of the molecule bonds to the siloxane surface between the amine hydrogen atoms and the oxygen atoms of the siloxane surface. Water is essential to the intercalation of the kaolinite and is intimately involved in the intercalation process. Based on this model, there are two types of NH groups, and hence two sets of bands are observed in the DRIFT spectra. Conclusions 1. With the use of controlled-rate thermal analysis, the presence of three different types of hydrogen-bonded reagent can be identified, and distinguished from the

Langmuir, Vol. 18, No. 4, 2002 1249

loosely bonded (adsorbed) part, in a kaolinite structure fully expanded by hydrazine-hydrate. Type 1 reagent (some 0.20 mol hydrazine-hydrate/mol inner-surface OH) is liberated between approximately 50 and 70 °C. Type 2 reagent is lost between approximately 70 and 85 °C, corresponding to a quantity of 0.12-0.15 mol hydrazinehydrate/mol inner-surface OH. Type 3 reagent is lost in the 85-130 °C range. The quantity of this reagent is fairly constant (0.29-0.32 mol hydrazine/mol inner-surface OH), independently of the conditions of sample preparation (drying). 2. DRIFT spectroscopy of the hydrazine molecule shows two types of NH2 stretching vibrations, thus indicating two types of hydrazine molecules in the intercalation complex. 3. The occurrence of the 9.6 Å band in the partially decomposed complex is due to the presence of hydrazine hydrogen-bonded directly to the inner-surface OH groups. The presence of the 10.3 Å reflection is explained by the connection of hydrazine molecules to the OH groups through water (i.e., in the form of hydrazine-hydrate). 4. Importantly, the combination of CRTA technology combined with X-ray diffraction and DRIFT spectroscopy has enabled the separation of adsorbed hydrazine-hydrate, weakly intercalated hydrazine-hydrate, and strongly intercalated hydrazine. The intercalation of hydrazinehydrate into kaolinite is complex and results from the different types of surface interactions of the hydrazine with the kaolinite surfaces. Acknowledgment. This research was supported by the Hungarian Scientific Research Fund (OTKA T034356). The support of the Queensland University of Technology Centre for Instrumental and Developmental Chemistry is also gratefully acknowledged. LA011179+