首页   按字顺浏览 期刊浏览 卷期浏览 Thermal evolution of cobalt hydroxides: a comparative study of their various structural...
Thermal evolution of cobalt hydroxides: a comparative study of their various structural phases

 

作者: Z. P. Xu,  

 

期刊: Journal of Materials Chemistry  (RSC Available online 1998)
卷期: Volume 8, issue 11  

页码: 2499-2506

 

ISSN:0959-9428

 

年代: 1998

 

DOI:10.1039/a804767g

 

出版商: RSC

 

数据来源: RSC

 

摘要:

J O U R N A L O F C H E M I S T R Y Materials Thermal evolution of cobalt hydroxides: a comparative study of their various structural phases Z. P. Xu and H. C. Zeng* Department of Chemical Engineering, Faculty of Engineering, National University of Singapore, 10 Kent Ridge Crescent, Singapore 119260. E-mail: chezhc@nus.edu.sg Received 23rd June 1998, Accepted 11th August 1998 Through an atmosphere-controlled method, a new phase of hydrotalcite-like Co hydroxide with mixed valent states has been synthesised, along with preparations of known a and b phases.Structural and thermal behaviours of all the Co hydroxides have been compared. Three major stages of decomposition are found: (i) 149–164 °C for dehydration of interlayer water, (ii) 185–197 °C and (iii) 219–222 °C for dehydroxylation of hydrotalcite- and brucite-like phases, respectively.Intercalated nitrate anions in hydrotalcite-like phases decompose largely during stage (ii). The oxide Co3O4 starts to form at temperatures as low as 165 °C especially for hydrotalcite-like phases. An intermediate compound, HCoO2, which is formed thermally, decomposes at 258–270 °C. The Co3O4 oxide converts into CoO at 842–858 and 935–948 °C respectively in nitrogen and air, which is much lower than the previously reported range of 1000–1200 °C.Surface areas of calcined samples are found to be proportional to the intercalated anion content. The catalytic activity of the resultant Co3O4 oxides with nitrous oxide is 7.2–8.2 mmol N2O g-1 h-1 at 375 °C, which is comparable to some reported active catalyst systems.conducted in liquid phase under static atmospheric conditions. Introduction Thus, oxidation of divalent transition metal cations by oxygen In recent years, divalent transition metal hydroxides including in static air, if occurring, may not be severe because transport their double hydroxides with trivalent p block (AIII ) or d of oxygen to the liquid phase is limited by the gas–bulk-liquid block metals have received increasing attention owing to their interface.However, if air (and hence oxygen) is constantly unique physico-chemical properties for electrochemical, supplied through bubbling to the liquid phase, the gas–liquid magnetic and catalytic applications.1–8 contact area can be significantly increased. The oxidation of One of the common features for this class of materials is Co2+(3d7) to Co3 +(3d6) has been recently observed in our their layered structure.9,10 A divalent metal cation is located preparation of the LDH compound MgIICoIICoIII-HT using a in the center of the octahedron formed by six hydroxyl groups.dynamic air-flow approach.11 In this connection, without The metal-octahedra then share edges to form two-dimen- involvement of MgII, the possibility of fabricating a CoIICoIII sionally infinite sheets, which is similar to the basic structure hydrotalcite-like phase had been indicated in our previous of brucite [Mg(OH)2].9,10 The brucite-like sheets can stack synthesis of Mg–Co mixed oxide spinels.12 upon each other to build a three-dimensional network accord- Here, we report a systematic investigation on preparation ing to various chemical interactions (mainly hydrogen bond- of Co hydroxide compounds using the dynamic gas-flow of ing) between the sheets.9,10 It has been well known for layered either protective (nitrogen) or oxidative gas (air).In addition double hydroxides (LDH) that when some of the divalent d to formation of a and b phases, the oxidation of Co2+(3d 7) block metal cations are substituted by a trivalent cation, a to Co3+(3d 6) occurs in the dynamic air-flow experiment, positive charge is generated in the brucite-like sheet. To restore leading to generation of a new type of CoIICoIII hydrotalciteoverall charge neutrality of solid, the extra positive charge can like compound. Using the well defined starting compounds (a be balanced by intercalating anion species into inter-brucite- and b phases, along with the newly found CoIICoIII hydrotallike- sheet space, resulting in a hydrotalcite-like phase [HT; cite-like phase), we are in a better position to conduct a named after the mineral compound Mg6Al2(OH)16CO3 comparative investigation on the thermal evolution of all these 4H2O] in most cases.9 Co hydroxides, which has been unclear in the literature owing Nevertheless, the formation and structure of layered mono- to insuYcient materials characterization (such as metal oxitransition metal hydroxides are not studied as explicitly as in dation state and anion content in the interlayer space) and LDH materials.9,10 For example, it has long been known that lack of mechanistic understanding on material formathere are two major types of nickel and cobalt hydroxides (a tion.2,3,6–8,13 The study will also correlate thermal evolution and b phases).1–8 The structure for the latter form (b) has with gaseous chemical constituents and chemical reactivity (an been identified as a brucite-like phase, but the former (a) has important index for catalytic application) of the thermally remained largely unknown regarding its actual formation obtained end product Co3O4 using N2O as a probe molecule.14 mechanism.Among many speculative models proposed for the a phase formation,1–5 two prominent models, ‘hydroxyl Experimental vacancies’ and ‘mixed valent state’ (with mixedM2+ andM3+) appear plausible.2,3 In particular, the ‘hydroxyl vacancies’ Materials preparation model has been experimentally confirmed very recently, which Two series of Co hydroxide samples (N1–N5 and A1–A6) reveals that the nickel and cobalt are strictly in the oxidation with various structural phases (brucite and hydrotalcite-like state of 2+ while the missing OH- groups in brucite-like and their mixtures) were prepared using an atmosphere- sheets are compensated by intercalated anions.8 controlled precipitation method.11 In the sample prepara- We have noted that most of these material syntheses are tion, 20.0 ml of 1.0 M aqueous cobalt nitrate solution [Co(NO3)2·6H2O, >99.0%, Fluka] was added into 100 ml 0.5 M ammoniacal solution in a three-necked round-bottom *Tel: +65 874 2896.Fax: +65 779 1936.J. Mater. Chem., 1998, 8(11), 2499–2506 2499Table 1 Sample nomenclature and preparation conditions further explain the DSC/DTA results, TGA measurements were carried out from 50–950 °C (at two heating rates: Addition Aging Structural 10 °Cmin-1 for 50–900 °C and 2 °Cmin-1 for 900–950 °C) Sample Atmosphere time time phasea with an air flow at 60 ml min-1. The as-prepared precipitates were also heat-treated at 200, N1 N2 15 s 5 min HT(II ) N2 N2 15 s 2 h HT(II ) 300 and 400 °C, respectively, for 2 h with static laboratory air N3 N2 15 s 4 h HT(II )+B(II ) in an electric furnace (Carbolite).Specific surface areas were N4 N2 15 s 6 h B(II ) determined by a multi-point BET method using a Nova-1000 N5 N2 15 s 8 h B(II ) Instrument. Prior to N2 adsorption–desorption measure- A1 Air 4 min 3 min B(II )+HT(II,III ) ment, each sample was degassed with N2 purge for 3 h at a A2 Air 4 min 2 h B(II )+HT(II,III ) temperature lower than its respective heat-treated temperature.A3 Air 4 min 4 h HT(II,III )+B(II ) In catalytic activity tests, 150 mg of 40–60 mesh sample A4 Air 4 min 8 h HT(II,III )+B(II ) calcined at 400 °C (2 h) were used in a tubular quartz reactor A5 Air 4 min 16 h HT(II,III ) (inner diameter=0.4 cm, V=0.088 cm3) in each run.In a A6 Air 4 min 24 h HT(II,III ) typical experiment, N2O gas (1 mol%, balanced with He) was aHT(II )=CoII hydrotalcite-like phase [i.e., a-Co(OH)2]; B(II )=CoII fed at a rate of 30 ml min-1 (F) through the catalyst bed brucite-like phase [i.e., b-Co(OH)2]; HT(II,III )=CoIICoIII hydrotalcite- like phase; order of structural phase appears according to intensity (GHSV=20 500 h-1).The GHSV was also increased to 41 000, of XRD data. 61 500, and 82 000 h-1 to investigate the eVect of feed rate on the catalytic activity. The outlet gases, cooled in a coil, were analyzed by gas chromatography (GC) on a Perkin-Elmer AutoSystem-XL (TCD detector) using a 4 ft Porapak Q flask within a given time (Table 1) under stirring. The atmos- (80/100 mesh) column.The oven temperature of GC was phere for precipitation was controlled by bubbling the solution maintained at 120 °C, and the flow rate of He carrier gas was with either purified nitrogen (Soxal, O2 <2 vpm) or purified 40 ml min-1. The catalytic activity for N2O decomposition air (Soxal, O2=21±1%, H2O<2 vpm, and hydrocarbons was evaluated in terms of conversion percentage, X=(Pi,N2O- <5 vpm) at a rate of 40 ml min-1 at room temperature.It Pf,N2O)/(Pi,N2O+0.5Pi,N2OPf,N2O), where Pi,N2O and Pf,N2O are should be noted that there are three preparative parameters inlet and outlet partial pressures of nitrous oxide.15,16 (atmosphere, addition time, and aging time) varied in these experiments (Table 1).The final pH values of the filtrate were in the range 8.5–8.2 depending on the addition/aging time in Results and discussion the experiment. Preparation of Co hydroxides with designed phases Materials characterisation In this work, three preparative parameters (atmosphere, addition time, and aging time) were varied. As indicated in Crystallographic information on the precipitates was Table 1, two diVerent hydrotalcite-like phases and one pure investigated by X-ray powder diVraction (XRD).The XRD brucite-like phase or their mixtures can be obtained using an patterns with diVraction intensity versus 2h were recorded in appropriate combination of these parameters. Fig. 1 shows a Shimadzu XRD-6000 X-ray diVractometer with Cu-Ka some representative XRD patterns which indicate that the radiation (l=1.5418 A° ) from 8–40° at a scanning speed of synthesis atmosphere gives a greater influence on crystallo- 3° min-1.Chemical bonding of cobalt–oxygen, hydroxyl and graphic structure of the resulting precipitates. some anions (mainly nitrate) was studied by FTIR (Shimadzu In a nitrogen atmosphere, an a phase precipitate can be FTIR-8101) using the potassium bromide (KBr) pellet techobtained upon fast addition and short aging (N1, Table 1).nique. The spectra were measured with a resolution of 2 cm-1 EA investigation reveals that there is no trivalent cobalt but and 100 scans were accumulated. Elemental analysis (EA) for there is anion intercalation in this phase (Table 2). On the nitrogen and carbon contents in the as-precipitated samples was performed using a Perkin Elmer 2400 CHN elemental analyzer.Cobalt content in the precipitate samples was determined by thermogravimetric analysis (TGA, Shimadzu TGA- 50) based on the end products Co3O4 at 600 and CoO at 950 °C. The trivalent cobalt content in some important precipitates (N1, N5, A1 and A6) was determined by a redox titration method in which 10.00 mg of solid sample were dissolved in 20.0 ml 1.0 M HCl solution upon gentle heating.The produced Cl2 was gradually purged with N2 (60 ml min-1) and passed through a 50.0 ml 0.01 M KI solution mixed with starch indicator. The resultant blue mixture was then titrated against a Na2S2O3 solution (0.0100 M) until disappearance of the blue coloration.DiVerential scanning calorimetry (DSC, Netzsch DSC200), diVerential thermal analysis (DTA, Shimadzu DTA-50), and thermogravimetric analysis (TGA, Shimadzu TGA-50) studies were carried out with various gas backgrounds in order to understand the thermal evolution of the prepared cobalt hydroxides. Samples in DSC measurements were heated from 40–400 °C at a rate of 10 °Cmin-1 under nitrogen with a gas flow rate of 15 ml min-1.To diVerentiate thermal processes at various heating stages, FTIR spectra were recorded for samples that were heated to a specified temperature in DSC measurements. In DTA measurements, the heating/cooling rate was Fig. 1 Representative XRD patterns for as-prepared Co hydroxides: kept the same as in DSC, but with air gas-flow at 60 ml min-1 N1, N5, A1, and A6, noting that major diVraction peaks of the samples are located in the reported 2h range. and with nitrogen gas-flow at 100 ml min-1, respectively.To 2500 J. Mater. Chem., 1998, 8(11), 2499–2506Table 2 Elemental analysis and TGA results for some representative samples Sample [Co3+]/[Co]a [NO3-+2CO32-]/[Co]b Structural phasec L1 (%)d L2 (%) L3 (%) N1 0 0.19 HT(II ) 8.9 11.8 4.6 N5 0 0.02 B(II ) 0.8 9.1 4.6 A1 0.04 0.13 B(II )+HT(II,III ) 3.6 13.4 4.0 A6 0.28 0.32 HT(II,III ) 8.0 16.5 4.7 Sample Co (%) NO3- (%) CO32- (%) H2O (%) Chemical formula N1 55.2 6.6 2.1 10.9 CoII(OH)1.81(NO3)0.11(CO3)0.04·0.65H2O N5 63.1 0 0 0.5 CoII(OH)2·0.03H2O A6 52.0 14.0 1.5 5.1 CoII0.72CoIII0.28(OH)1.96(NO3)0.26(CO3)0.03·0.3H2O aMole ratio of trivalent cobalt to total cobalt in the precipitate samples.bMole ratio of total charges of anion species to total cobalt in the precipitate samples; CO32- results from ambient CO2 dissolution. cThe phase notation is the same as in Table 1. dL1, L2 and L3 are weight loss percentages from 40–160 °C, 160–230 °C, and 230–500 °C. basis of XRD/EA/TGA studies, the precipitate (N1) of the hydrotalcite-like phase has an inter-brucite-like-sheet distance of 8.09 A° , and a chemical formula CoII(OH)1.81(NO3)0.11 (CO3)0.04 0.65H2O. In contrast, a sample similarly prepared in nitrogen but with a longer aging time is a pure brucite-like compound (N5, Fig. 1). Again, cobalt is strictly divalent and there are virtually no anions or water intercalated in this compound.The inter-brucite-like-sheet distance determined by XRD is 4.66 A° , which is identical to the literature data for the b phase.17 From N1 to N5 (Table 1) the sequential evolution from CoII(OH)1.81(NO3)0.11(CO3)0.04 0.65H2O to CoII(OH)2 can be conducted by changing the aging time. A twophase material is seen in sample N3 in which both CoII(OH)1.81(NO3)0.11 (CO3)0.04 0.65H2O and CoII(OH)2 are present.The above CoII hydrotalcite-like to CoII brucite-like transformation occurs under a protective atmosphere of nitrogen. However this transformation can be steered in the reverse direction when nitrogen is replaced by air. A predominant brucite-like phase can be prepared within 4 min (A1, Table 1 and 2, Fig. 1). Surprisingly, this newly formed brucite-like phase can be transformed into a pure hydrotalcite-like phase Fig. 2 DSC curves for the sample series prepared under nitrogen upon prolonged aging in air. As demonstrated in the A1–A6 atmosphere: N1–N5. series, bi-phasic mixtures are observed in A1–A4 while hydrotalcite- like compounds are found in A5 and A6. Related to this phase evolution, the Co3+ to total cobalt ratio is increased evolution revealed by XRD, a smooth transition from the from 0.04 to 0.28 from the sample A1 to A6 (Table 2) and hydrotalcite to brucite-like phase can be seen from these about 30% of initial Co2+ has been oxidized to Co3+ in A5 DSC curves.and A6. In response to the increase in positive charge, a DSC scans for air-prepared samples A1–A6 are shown in similar amount of negative anionic species is found in the Fig. 3. As can be seen, thermal events corresponding to interlayer space (i.e., 0.28 vs. 0.32 determined for A6; Table 2). depletion of the interlayer water and dehydroxylation of The chemical formula for both A5 and A6 is thus hydrotalcite-like structure take place at ca. 152–164 °C (first CoII0.72CoIII0.28(OH)1.96(NO3)0.26(CO3)0.03·0.3H2O according peak) and 185–197 °C (second peak) respectively, in all to XRD/EA studies and the inter-brucite-like sheet distance samples.In good agreement with the XRD results, which for this compound is 7.98 A° . show a predominant brucite-like phase and a small hydrotal- In view of the above results, Co hydroxides can apparently cite-like phase in A1 and A2 (Table 1), the DSC investigation be tailor-made with a desired structure and chemical composiin Fig. 3 also indicates the co-existence of the two phases. tion, including electronic configuration for the cations, using Compared to hydrotalcite-like phase, the brucite-like phase is the current approach. thermally more stable. The third endothermic eVect (217–219 °C) due to dehydroxylation of the brucite-like struc- Thermal decomposition of Co hydroxides ture is observed in A1 and A2.The endothermic eVect observed at ca. 258 °C in A5 will be explained below. Fig. 2 shows results of DSC investigations on the nitrogenprecipitated samples N1–N5. Low-temperature endothermic It is noted that thermal behaviours of N1 and A6 considerably diVer (Fig. 2 and 3), although they both have hydro- humps at ca. 130 °C in N1–N3 can be assigned to depletion of surface adsorbed water, while bands at 149–158 °C can be talcite-like structures (Fig. 1). In view of the absence of trivalent cations, the CoII hydrotalcite-like structure in N1 attributed to the release of interlayer water molecules.18–20 Collapse of the hydrotalcite-like structure is observed at ca. (a phase; CoII(OH)1.81(NO3)0.11(CO3)0.04·0.65H2O) can be described well by the ‘hydroxyl group deficiencies’ model.2,8 190–195 °C for N1–N4 (only a small shoulder in N4, i.e., a trace phase of HT).Finally large endothermic bands at By contrast, the CoIICoIII hydrotalcite-like phase in A6 (CoII0.72CoIII0.28(OH)1.96(NO3)0.26(CO3)0.03·0.3H2O) can be 219–222 °C in N4 and N5 can be assigned unambiguously to decomposition of the brucite-like structure, which normally ascribed to formation of common hydrotalcite-like phase that contains both divalent and trivalent cations due to the partial occurs at a higher temperature.18–20 Similar to the structural J.Mater. Chem., 1998, 8(11), 2499–2506 2501839 cm-1: n3 and n2 modes of NO3- with D3h symmetry) to a tilted or vertical position, which is indicated by a newly emerging peak at 1010 cm-1 (n1 mode of NO3-) of C2v symmetry.21,22,26 Although the final calcined products are all Co3O4, thermal evolution paths may diVer considering the substantial diVerences of starting materials in chemical composition, oxidation state, and molecular structure (Table 2).To diVerentiate the decomposition paths, the formation process of Co3O4 was further investigated by DTA using nitrogen and air.Fig. 5 shows DTA results for N1–N5; decomposition processes in nitrogen are quite similar to those in DSC (Fig. 2) in which nitrogen was also used. However, the thermal behaviour changes considerably in air-DTA experiments. For example, exothermic peaks at 172 °C (N5) and 189 °C (N4) can be attributed to Co2+ oxidation to Co3+ in the brucite-like phase.The peaks are still exothermic even considering the decomposition (which is endothermic): 27 3 Co(OH)2+1/2 O2=Co3O4+3 H2O DHro=-25 kJ mol-1 Co3O4 (1) For the A1–A6 series shown in Fig. 6, the endothermic peaks Fig. 3 DSC curves for the sample series prepared under air atmosphere: A1–A6. at 211 °C (A1) and 206 °C (A2) observed in nitrogen atmosphere can be assigned to the decomposition of a mixed brucitelike phase.Nevertheless, this endothermic eVect is cancelled oxidation of Co2+. The speculated ‘mixed valent state’ model3 out by the exothermic oxidation of the brucite-like phase for can be thus confirmed. With respect to the formation mechanthe same samples measured in air (A1 and A2, Fig. 6). ism, the increase in peak intensity and sharpness in A6 (second Following the DSC /XRD /FTIR /DTA results, weight peak, Fig. 3) reflects an increase in interaction between cations losses before 500 °C in TGA can be broadly classified into and intercalated anions due to presence of trivalent cobalt. three major types (L1, L2 and L3; Table 2). The first, between More anionic species are found in the CoIICoIII hydrotalcite- 40 and 160 °C can be ascribed to dehydration of samples like phase than in the CoII HT-like phase (0.32 in A6 vs. 0.19 (surface and/or interlayer water). This assignment is supported in N1; Table 2). It is thus believed that stronger electrostatic by the fact that a negligible loss is observed for the essentially attraction would lead to a more intense endothermic eVect brucite-like phase N5 while larger weight losses are seen for observed during the dehydroxylation. all hydrotalcite-like samples (N1>A6>A1&N5; Table 2).The second stage between 160 and 230 °C is attributed to Thermal evolution to Co3O4 oxides dehydroxylation of Co hydroxides, as all samples exhibit the Owing to structural and chemical diVerences, diVerent thermal largest losses over this temperature range.The final stage evolution behaviours and thermal stability of the Co hydrox- between 230 and 500 °C can be ascribed to a continued ides are expected upon heat-treatment. Fig. 4(a) shows FTIR decomposition process for large-sized particles (mass transport spectra for the stage-calcined N1 sample (see Experimental limiting) and a decomposition of HCoIIIO2. section). As can be seen, fingerprint IR absorptions at 580 According to TGA all samples show a small weight loss and 660 cm-1 for the Co3O4 spinel phase are fully developed between 270 and 274 °C within the total L3.However, this after heat-treatment at 165 °C.12 Upon the lowering of intensit- loss becomes much more pronounced in samples A4 and A5 ies for hydroxyl group and intercalated nitrate ion bands at with a total loss L3 of 7.3% and 7.1%, respectively, compared 3430 cm-1 and 1384 and 839 cm-1,21–23 the metal-oxygen to L3=4.0–4.7% in Table 2.The weight losses at 270–274 °C vibrational absorption modes at 564–580 and 660 cm-1 correspond to the endothermic peaks at 262–270 °C found in increase markedly.9,24 Fig. 4(b) shows in FTIR spectra the the DTA scans of A4 and A5 (Fig. 6; and similarly observed metal–OH vibration of the A1 sample at ca. 494 cm-1,9,24 in the DSC scan of A5, Fig. 3), which can be assigned to after heat-treatment at 165 °C. It is thus confirmed that sample decomposition of cobalt oxide hydroxide [HCoIIIO2]. This A1 is thermally more stable than N1. Furthermore, the trivalent Co compound also has a hexagonal-layered structure, hydroxyl group (3630 cm-1) in the brucite-like phase is still layers of which are bonded to each other by hydrogen bonding very pronounced at this temperature and it is only significantly and decomposes in the same temperature range under vacuum reduced at around 200 °C when dehydroxylation commences.or oxygen atmosphere: 28 It should be noted that, regardless of types of initial precipitates, the final decomposed products of N1 and A1 at higher 12 HCoO2=4 Co3O4+6 H2O+O2 (2) temperatures are exclusively in the spinel form (IR bands at 564 and 660 cm-1 for Co3O4).12 The oxidative formation of It has been established thermogravimetrically that in the temperature range 120–190 °C trivalent Co hydroxide first Co3O4 phase is completed with precipitation of nitrate anion, in the absence of ambient oxygen.In this connection, a small forms HCoO2 before being converting to Co3O4 at 240–300 °C.13,28 Furthermore HCoO2 can also be oxidatively peak at 1270 cm-1 for N1 heated to 190 °C can be assigned to the asymmetric vibration of monodentate nitrate ion [asym- prepared in an ambient atmosphere via oxidation of b- Co(OH)2.28 Since the TGA measurements were conducted metric vibration mode: nas(ONO2)].20,25 Nitrate ion evolution during heating can be seen clearly in A6 [Fig. 4(c)] that with air, divalent cobalt can be oxidised to the trivalent HCoO2 on heating. contains the largest amount of anions (Table 2). In addition to the formation of monodentate species [1310 cm-1 for the The pronounced HCoO2-decomposition for the A4 and A5 samples in TGA/DTA/DSC can be attributed to the presence nas(ONO2) and 1470 cm-1 for the symmetric vibration mode ns(ONO2)] 20,25 at 165 °C, the nitrate anion also undergoes a of trivalent cobalt plus the on-site formed HCoO2.However, HCoO2 decomposition is not observable in DTA/DSC symmetry-lowering from a flat-laying configuration (1384 and 2502 J. Mater. Chem., 1998, 8(11), 2499–2506Fig. 4 FTIR spectra for three as-prepared samples: (a) N1, (b) A1, and (c) A6 after heating from room temperature (unmarked) to the temperatures (marked) of some major thermal events in the DSC scans (10 °Cmin-1 in nitrogen). Fig. 6 DTA curves for A1–A6 sample series measured under nitrogen Fig. 5 DTA curves for N1–N5 sample series measured under nitrogen and air atmospheres.and air atmospheres. measurements for A6 even though it contains trivalent cobalt. the same for all samples including the anion-free N5. This This diVerence can be related to distribution homogeneity of view is further supported by FTIR studies in Fig. 4(a)–(c) trivalent cobalt in the precipitates. Compared to the A5 which show a significant reduction in the nitrate ion absorption sample, the second endothermic peak of the DSC-scan for A6 at 1384 cm-1 over this temperature range.(Fig. 3 and similarly in DTA scans of Fig. 6) is much sharper. This is also observed for other well aged samples (>24 h, not Transformation between Co3O4 and CoO listed in Table 1). As mentioned earlier, the sharpness of DSC peak indicates a well defined phase. The above observations It is known that Co3O4 is a thermodynamically stable form under an oxygen containing atmosphere.The stability of the lead to our belief that a more homogeneous distribution of trivalent cobalt cations among the divalent ions leads to a prepared Co3O4 was examined under inert gas (nitrogen) or an oxygen-containing atmosphere (air). DTA investigation on dimunition of the HCoO2 phase.Based on the TGA data, it can be concluded that the anionic decomposition and restoration of Co3O4 during heating-cooling cycles is shown in Fig. 7 and 8. The endothermic peaks at species decompose almost simultaneously with dehydroxylation of the Co hydrotalcites (N1, A1 and A6; Table 2) in ca. 846–858 °C for samples N1–N5 heated under nitrogen (Fig. 7) can be described by the forward reaction of the stage two, since the losses (L3) at stage three are essentially J.Mater. Chem., 1998, 8(11), 2499–2506 2503Fig. 9 TGA curves for some selected Co hydroxides during Co3O4= 3 CoO+1/2 O2 phase transformation. Fig. 7 DTA heating-cooling cycles for N1–N5 sample series measured under nitrogen and air atmospheres; arrows indicate the heating– cooling directions.The observed similar DTA-heating–cooling cycles for both sets of samples indicate that all Co hydroxides should have similar chemical constituents over the temperature range studied. This point has been further confirmed with the TGA investigation shown in Fig. 9. In good agreement with the DTA results, there is no further weight loss at temperature >500 °C until the Co3O4 phase decomposes.Conversion of Co3O4 to CoO occurs at >910 °C (note: a slow heating rate of 2 °Cmin-1 is used). The detected weight loss is in the range of 6.23–6.46%, very close to the theoretical value of 6.62% according to eqn. (3). On the basis of DTA/TGA results, the Co3O4 to CoO transformation in these hydroxide-derived spinel phases occurs at lower temperatures compared with the reported data of 1000–1200 °C under air.20,21 As revealed in XRD patterns, the crystallinity of the low-temperature formed Co3O4 is low, which may ease oxygen transport during phase conversion and result in the observed low transformation temperature.Catalytic evaluation of the thermally formed Co3O4 Fig. 10 shows specific surface area data for some representative samples calcined at various temperatures for 2 h.As can be seen, brucite-like compounds, N5 and A1 (A1 contains a small amount of hydrotalcite-like phase), show higher specific surface area at low temperature (200 °C). However, they show Fig. 8 DTA heating–cooling cycles for A1–A6 sample series measured under nitrogen and air atmospheres; arrows indicate the heating– cooling directions.following chemical equilibrium: Co3O4=3 CoO+1/2 O2 (3) However, this decomposition occurs at a much higher temperature (941–946 °C) for the same series under air. As it is reversible, the above reaction shifts to the left when the temperature is lowered and there is suYcient oxygen in the gaseous phase [PO2=2.1×104 Pa using air at this experimental setting (total pressure=1 atm)].As indicated by the large exothermic peaks at 804–828 °C (Fig. 7), the reverse reaction of eqn. (3) is exothermic due to oxidation of divalent cobalt.29 A similar observation is seen for the sample series A1–A6 in Fig. 8 over the same heating–cooling range. Since there is no oxygen, the reverse transition is not observed in the DTA Fig. 10 Specific surface area versus calcination temperature for some representative Co hydroxide samples.experiments using nitrogen (Fig. 7 and 8). 2504 J. Mater. Chem., 1998, 8(11), 2499–2506Table 3 Kinetic data of nitrous oxide decomposition from some representative catalysts Structural Ea/ Specific surface Sample phasea kJ mol-1 ln A areab/m2 g-1 N5 B(II) 94 19.5 29 A1 B(II)+HT( II,III) 80 16.7 40 N1 HT(II) 83 17.4 42 A6 HT(II,III) 79 16.8 52 aThe phase notation is the same as in Table 1.bUpon calcination at 400 °C. i.e. nitrous oxide decomposition is largely carried out on the Co3O4 surface phase, rather than CoO. Decomposition of nitrous oxide on a metal oxide surface involves an electron transferring from a low-oxidation state metal cation to an adsorbed nitrous oxide molecule.33,34 For many metal oxide catalysts, charge transfer from a low-valence metal cation to the adsorbed N2O is often considered as a fast surface reaction.Based on this mechanism, a typical Langmuir–Hinshelwood rate equation can be derived.15 In particular, the following simplified equation can be obtained Fig. 11 Conversion versus temperature curves for some representative considering the adsorption step as a controlling step and a Co hydroxides calcined at 400 °C.negligible inhibiting role of O2 in the decomposition reaction:15,35,36 -dPN2O/dt=kPN2O (4) significantly reduced specific surface areas at elevated temperatures. On the other hand, the surface areas of hydrotalcite- After integration, eqn. (4) becomes: like samples (N1 and A6) are systematically higher than those ln{ln(Pi,N2O /Pf,N2O)}=ln A-ln(F/V )-Ea/RT (5) of N5 and A1 at 300 and 400 °C.In particular the surface areas of these high-temperature calcined samples are pro- where V and F have been defined in Experimental section, portional to the anion content in the precursor compounds, and A is Arrhenius pre-exponential factor, and Ea is the i.e., SA6>SN1>SA1>SN5 (see Table 2 for the anion content).apparent activation energy of decomposition.35 Eqn. (5) has Fig. 11 shows the chemical reactivity of the above thermally been employed in the current work to provide a correlation formed samples using nitrous oxide as a probe gas at 400 °C. between the preparative method and catalytic activity of These conversion versus temperature curves were obtained at Co3O4.Since the ln{ln(Pi,N2O/Pf,N2O)} versus 1/T plots are all a GHSV of 20 500 h-1. Although they are prepared from fitted very well to straight lines, decomposition kinetics of various starting Co hydroxides, the 400 °C-formed oxides nitrous oxide on these Co3O4 oxides can be described as first occur strictly as Co3O4 crystallographic phases,30 as revealed order with partial pressure of nitrous oxide with apparent by XRD before and after nitrous oxide decomposition.Owing activation energies of 79–94 kJ mol-1 as listed in Table 3, to the same Co3O4 phase, the observed catalytic activity varies noting that these values are quite similar. only slightly. Although surface area is usually critical in determining catalytic activity, a variation in specific surface Conclusions area by a factor of ca.two has been shown to have little eVect (N5 versus A6, Table 3). Therefore, the subtle diVerences In summary, the synthetic chemistry and thermal evolution of among these curves can be attributed to a combined eVect of one new and two known Co hydroxides have been investigated total variation in specific surface area, number of reaction systematically with a wide range of characterisation and sites per specific surface area, and activity of a reaction site.analytical methods (XRD/EA/DSC/FTIR/DTA/TGA/GC). Overall all Co3O4 oxides derived from these monohydroxides Three major decomposition stages of Co hydroxides have been are highly active, when compared with the literature conversion identified: (i) 149–164 °C for dehydration of interlayer water, data.For example, the activity of the present catalytic oxides (ii) 185–197 °C and (iii) 219–222 °C for dehydroxylation of at 375 °C is in the range 7.2–8.2 mmol N2O g-1 h-1 (GHSV= hydrotalcite and brucite-like phases, respectively. Intercalated 82 000 h-1, N2O=1 mol% balanced with He), compared with nitrate anions decompose largely at the stage (ii).During the 1.6–2.3 mmol N2O g-1 h-1 (GHSV=30 000 h-1, N2O= above decomposition, Co3O4 oxide starts to form at tempera- 0.1 mol% balanced with He) for an active hydrotalcite-like tures as low as 165 °C especially for hydrotalcite-like phases. compound (CoIIMgIIAlIII-HT) derived catalyst operated at The intermediate compound HCoO2, formed during the ther- 350–400 °C.31,32 mal evolution, decomposes at 258–270 °C.It is found that As the decomposition reaction occurs on the surface, the Co3O4 oxide formed freshly from thermal decomposition of actual crystallographic phase of surface Co3O4 should be the hydroxides coverts to CoO at 842–858 °C and 935–948 °C further addressed. Using XPS, it is known that the transform- under nitrogen and air, respectively.Surface areas of calcined ation in eqn. (3) occurs at ca. 347 °C on the surface of Co3O4 samples are found to be proportional to their intercalated in an oxygen atmosphere of 1×10-3 Pa, which is ca. 186 °C anion content. The resultant Co3O4 materials are catalytically lower than for thermodynamic calculation for the bulk phase active and the activity is comparable to some reported active Co3O4 to CoO transition under the same PO2.29 Since the catalyst systems.For example, the activity of Co3O4 oxide oxygen partial pressure generated by the current nitrous operated at 375 °C is in the range 7.2–8.2 mmol N2O g-1 h-1 oxide decomposition is much higher than 1×10-3 Pa (GHSV=82 000 h-1, N2O=1 mol% balanced with He). (PO2=4–16 Pa for the reactions 250 °C and 446–466 Pa at 400 °C reactions, Fig. 11), it is expected that the surface Co3O4 The authors gratefully acknowledge research funding (RP960716) co-supported by the Ministry of Education and to CoO transition should occur at a temperature much higher than 347 °C (for PO2=1×10-3 Pa) found in the XPS study;21 the National Science and Technology Board of Singapore. J. Mater.Chem., 1998, 8(11), 2499–2506 250519 M. A. Ulibarri, M. J. Hernandez and J. Cornejo, J. Mater. Sci., References 1991, 26, 1512. 20 M. Schraml-Marth, A. Wokaun and A. Baiker, J. Catal., 1992, 1 D.L. Bish and A. Livingstone, Miner. Mag., 1981, 44, 339. 138, 306. 2 C. Faure, C. Delmas and M. Fousassier, J. Power Sources, 1991, 21 J. M. Fernandez, C. Barriga, M. A. Ulibarri, F.M. Labajos and 35, 279. V. Rives, J. Mater. Chem., 1994, 4, 1117. 3 P. V. Kamath and N. Y. Vasanthacharya, J. Appl. Electrochem., 22 I. C. Chisem and W. Jones, J. Mater. Chem., 1994, 4, 1737. 1992, 22, 483. 23 S. Kannan and C. S. Swamy, J. Mater. Sci. Lett., 1992, 11, 1585. 4 P. Rabu, S. Angelov, P. Legoll, M. Belaiche and M. J. Drillon, 24 G. Busca, F. Trifiro and A. Vaccari, Langmuir, 1990, 6, 1440.Inorg. Chem., 1993, 32, 2463. 25 N. Zotov, K. Petrov and M. Dimitrova-Pankova, J. Phys. Chem. 5 P. Benard, J. P. AuVredic and D. Louer, Thermochim. Acta, 1994, Solids, 1990, 51, 1199. 232, 65. 26 E. Kruissink, L. L. van Reijen and J. R. H. Ross, J. Chem. Soc., 6 A. Delahaya-Vidal, K. Tekaia Ehlsissen, P. Genin and M. Figlarz, Faraday Trans. 1, 1981, 77, 649. Eur. J. Solid State Inorg. Chem., 1994, 31, 823. 27 Handbook of Chemistry and Physics, ed. D. R. Lide, CRC Press, 7 P. V. Kamath, M. Dixit, L. Indira, A. K. Shukla, V. G. Kumar USA, 78th edn., 1997, p. 5. and N. Munichandraiah, J. Electrochem. Soc., 1994, 141, 2956. 28 K. Shigeharu, N. Uchida, I. Miyashita and T. Wakayama, 8 P. V. Kamath and G. H. A. Therese, J. Solid State Chem., 1997, Colloids Surf., 1989, 37, 39. 128, 38. 29 M. Oku and Y. Sato, Appl. Surf. Sci., 1992, 55, 37. 9 F. Cavani, F. Trifiro and A. Vaccari, Catal. Today, 1991, 11, 173. 30 Powder DiVraction File, Card No. 9-418. Joint Committee on 10 W. T. Reichle, Solid State Ionics, 1986, 22, 135. Powder DiVraction Standards, Swarthmore, PA, 1967. 11 H. C. Zeng, Z. P. Xu and M. Qian, Chem.Mater., 1998, 10, 2277. 31 S. Kannan and C. S. Swamy, Appl. Catal. B, 1994, 3, 109. 12 M. Qian and H. C. Zeng, J. Mater. Chem., 1997, 7, 493. 32 J. N. Armor, T. A. Braymer, T. S. Farris, Y. Li, F. P. Petrocelli, 13 L. V. Pyatnitskii, Analytical Chemistry of Cobalt, IPST (Israel E. L. Weist, S. Kannan and C. S. Swamy, Appl. Catal. B, 1996, Program for Scientific Translations), Jerusalem, 1966, p. 5. 7, 397. 14 F. Kapteijn, J. Rodriguez-Mirasol and J. A. Moulijn, Appl. Catal. 33 S. Akbar and R. W. Joyner, J. Chem. Soc., Faraday Trans. 1, 1981, 77, 803. B, 1996, 9, 25. 34 D. D. Eley, A. H. Klepping and P. B. Moore, J. Chem. Soc., 15 K. Li, X. F. Wang and H. C. Zeng, Chem. Eng. Res. Des. Part A: Faraday Trans. 1, 1985, 81, 2981. Trans. I. Chem. E., 1997, 75, 807. 35 R. Sundararajan and V. Srinivasan, Appl. Catal., 1991, 73, 165. 16 X. F. Wang and H. C. Zeng, Appl. Catal. B, 1998, 17, 89. 36 A. Cimino, Chim. Ind. (Milan), 1974, 56, 27. 17 Powder DiVraction File, Card No. 45-0031. Joint Committee on Powder DiVraction Standards, Swarthmore, PA, 1995. 18 I. E. Grey and R. Ragozzini, J. Solid State Chem., 1991, 94, 244. Paper 8/04767G 2506 J. Mater. Chem., 1998, 8(11), 2499–2506

 

点击下载:  PDF (198KB)



返 回