- Email: [email protected]
Topotactic transformation of Ni-based layered double hydroxide film to layered metal oxide and hydroxide
Applied Clay Science 124-125 (2016) 236–242
Contents lists available at ScienceDirect
Applied Clay Science journal homepage: www.elsevier.com/locate/clay
Research paper
Topotactic transformation of Ni-based layered double hydroxide film to layered metal oxide and hydroxide Takahiro Takei a,⁎, Hiroki Fuse a, Akira Miura b, Nobuhiro Kumada a a b
Center for Crystal Science and Technology, University of Yamanashi, 7-32 Miyamae, Kofu, Yamanashi 400-8511, Japan Faculty of Engineering, Hokkaido University, Kita 13, Nishi 8, Kita-ku, Sapporo 060-8628, Japan
a r t i c l e
i n f o
Article history: Received 10 October 2015 Received in revised form 12 February 2016 Accepted 15 February 2016 Available online 23 February 2016 Keywords: Topotactic transformation Layered double hydroxide Layered metal oxide
a b s t r a c t Topotactic transformation of film-shaped samples comprising a layered double hydroxide (LDH) to a metal oxide phase was performed by a hydrothermal or heat treatment. The degree of transformation was examined by X-ray photoelectron spectroscopy (XPS) and synchrotron X-ray diffraction (SXRD). The LDH film prepared as a raw material by a reflux process with urea was composed of plate-shaped particles that were partially perpendicular to the substrate. For the transformation, the LDH powder and film were treated hydrothermally at 90, 120, and 200 °C in a LiOH aqueous solution or heated to 400, 600, and 800 °C with a LiOH·H2O contact. The film shape and morphology of the LDH particles were retained after the hydrothermal treatment at 120 °C. XPS and SXRD confirmed that two types of LDH with different compositions, Ni:Co = 62:38 and 90:10, were formed in the as-prepared LDH. In the case of the sample treated hydrothermally, metal hydroxide with no intercalated anion within the interlayer space was observed and the metal oxyhydroxide formed with a ratio of approximately 15%. For the heated samples, the transformation to LiNi1-xCoxO2-related materials was carried out by heating to 600 or 800 °C. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Layered double hydroxide (LDH) has a layered structure with anions intercalated between brucite-type hydroxide layers. The hydroxide layer is composed of divalent and trivalent cations. The interlayer anion compensates for the charge discrepancy arising from the occurrence of trivalent cations in the hydroxide layer. Generally, the interlayer anion can be exchanged by ion-exchange or reconstruction processes. For ion exchange, Cl− or NO-3type LDH should be used as raw materials because CO2− is generally most stable within the inter3 layer space. For reconstruction, calcined LDH is immersed into an anion-contained aqueous solution to rehydrate with the existed anion. The structural changes in this process have been examined by X-ray absorption fine structure (XAFS), but the formation is sometimes difficult for transition metal-containing LDH due to facile formation of a spineltype phase (Rives, 2002). For applications of the ion-exchanged LDH, some studies have reported the preparation of functional hybrids (Pinnavaia et al., 1995; Tang et al., 2011; Takei et al., 2014), removal of toxic anions (Nakahira et al., 2007; He et al., 2011; Park et al., 2004), and use of drug delivery systems (Aisawa et al., 2006, 2007). The most common LDH is “hydrotalcite,” which is composed of Mg2+ and Al3+. For the LDH, including hydrotalcite, the divalent and trivalent cations can be replaced with period-four transition metal cations such as Cr, ⁎ Corresponding author. E-mail address: [email protected] (T. Takei).
http://dx.doi.org/10.1016/j.clay.2016.02.018 0169-1317/© 2016 Elsevier B.V. All rights reserved.
Mn, Fe, Co, and Ni, because most period-four transition metals are either di- or trivalent. For the crystal structure of LDH, all metals form M(OH)6 octahedra in which all OH groups are shared. For such polyhedral polycondensation structures, six edges in an octahedron are shared with each other. The structure of the LDH is, actually, quite similar to that of LiCoO2 despite the positive valence of the interlayer ion (Li+). Therefore, some researchers have examined the synthesis of LiCoO2-type cathode materials from LDH as a raw material (Lu et al., 2007; Yushi et al., 2008). For LiCoO2, Co occurs in a trivalent state and the Li cation can be deintercalated from the structure owing to the oxidation of trivalent Co to tetravalent Co during the charge process of a Li-ion battery. When tetravalent Co is reduced to the trivalent state, Li ions can be intercalated again. However, LiCoO2 has a slight risk of decomposing with the generation of O2 gas at relatively low temperatures (above 200 °C). This can be dangerous because the generation can expand the battery cell and sometimes result in an explosion. To improve the decomposition temperature, many researchers have examined some additives, partial substitution or change in composition. LiMn2O4 (spinel structure) and LiNiO2 can decompose above 300 °C and approximately 200 °C, respectively (Biensan et al., 1999; Kitano et al., 2005). The use of LiCoO2 as an electrode in secondary batteries has two disadvantages. One is the low electronic conductivity. As the conductivity of LiCoO2 is approximately 10− 3 S/cm (Kwon et al., 2014), both an adhesive agent and a conduction aid are necessary for the formation of an electrode device, particularly regarding the use of powder-type
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
LiCoO2. The adhesive agent and conduction aid are generally PTFE and carbon, respectively. Because LiCoO2 is a p-type semiconductor, the number of holes will increase with increasing Co4 + concentration. Nobili et al. reported that Mg substitution at the Co site in LiCoO2 results in an increase in electronic conductivity (Nobili et al., 2005). The other disadvantage is the layered structure. The structure can be an advantage because the Li cation can move within the interlayer space; however, it becomes a disadvantage for large LiCoO2 particles. Large plate-like particles sometimes decelerate the intercalation and deintercalation of Li because of the long diffusion path. Because Li diffusion will be predominant, the synthesis of LiCoO2 nanoparticles or the preparation of an oriented structure perpendicular to the substrate will be effective. For these reasons, both the crystal structure and particle orientation for LiCoO2-type compounds are very important. Therefore, the topotactic transformation from LDH is possible. In this study, the preparation of oriented LDH on the substrate and topotactic transformation to Nibased layered metal oxide with a LiCoO2 structure was examined by a hydrothermal and heating process. The crystal structures of the sample were examined by synchrotron-XRD (SXRD). 2. Experimental The transformation was performed for the sample in two forms, powder and film. Both samples were prepared, which were then transformed by heating with LiOH or by a hydrothermal treatment in an aqueous LiOH solution. The typical processes are as follows. 2.1. Preparation of LDH Ni and Co nitrate mixed salt was used as the metal sources at ratios of 1, 2, 4, and 8. Distilled water was added to the salts to achieve a concentration of 5 mmol/L. Urea was placed in the aqueous solution of the mixed salts with a molar ratio of 20. In the case of film preparation, a 10 × 50 × 0.2 mm Ti or Ni plate used as a substrate was soaked perpendicular to the stirring flow in an aqueous solution. The solution was then refluxed with stirring at 120 °C for several hours. The prepared LDH are designated as mNi1Co, where m means the loaded molar ratio of Ni/Co; e.g., 4Ni1Co. 2.2. Transformation treatment of LDH powder and film The obtained LDH underwent a transformation by a heat-treatment or hydrothermal treatment. The heat treatment was performed using LiOH·H2O at 400–800 °C for 3 h in air. The LiOH·H2O to metal cation ratio in LDH was 3:2. The heated samples were washed with distilled water and dried at 50 °C. In the case of the hydrothermal treatment, the LDH samples were soaked in a LiOH aqueous solution with a concentration of 1.0 mol/L. The solution was treated hydrothermally at 90, 120, and 200 °C for 72 h. The treated samples were washed with distilled water and dried. Here, “HT” is defined as a prefix of the hydrothermal temperature for use as the sample name.
237
could not be obtained, which prevented accurate measurements. Therefore, the powder sample was used for these measurements. The chemical composition of the samples was measured by ICP (PS-3500DDII, Hitachi High-Tech Science Corp.) The structures were examined by SXRD using a Debye-Scherrer camera at BL02B2 in SPring-8. A borosilicate capillary tube, 0.2 mm in radius, was filled with the samples. The wavelength was 0.49560 Å, which was determined using CeO2 as the reference crystal. The SXRD patterns were refined using RIETAN-FP (Izumi and Momma, 2007). 3. Results and discussion 3.1. Preparation of LDH film Fig. 1 shows XRD patterns of the Ni-Co LDH powder with different compositions, Ni:Co = 1:1, 2:1, 4:1, and 8:1. The LDH phase can be observed at all compositions. However, the XRD reflection intensity increased gradually depending on the Ni ratio. Generally, because the M2+/M3+ ratio should be 2–4, the chemical composition ranges of 2:1 and 4:1 will be the optimum because the divalent cations tend to be deficient at the higher ratio. In actual experiments, the crystallinity is high and the yield is low for 8Ni1Co. For 2Ni1Co, the yield is good but the crystallinity is somewhat low. Therefore, in this study, a 4:1 composition was adopted for the subsequent experiments. Fig. 2 presents the XRD patterns of the grown LDH films with a Ni:Co ratio of 4:1 on the Ti substrate for different synthesis periods (1.0, 1.5, 2.0, 3.0, and 5.0 h). Their baselines appear to be worse because the substrate Ti metal is pliable and the measured surface is not flat. The halo in the sample at 1.5 and 5.0 h might be from a glass slide. These patterns suggest that the amount of LDH increases depending on the synthesis period. The patterns could not confirm whether the LDH particles were ordered or not. Fig. 3 presents the FE-SEM images of the grown LDH films with a different period, 1.0, 1.5, 2.0, 3.0, and 5.0 h. The top and bottom micrographs show the low and expanded magnifications, respectively. The LDH nuclei begin to form at 1.0 h, and the number of nuclei increases steeply during an additional 1 h. The number of LDH particles perpendicular to the substrate tends to increase over synthesis periods of 3 h or shorter. At 5 h, the particle size appears to become approximately double. However, the excess particles emerge on the perpendicular particles, as shown in the low-magnification image. 3.2. Transformation treatment for LDH The transformation treatment for LDH was carried out by heating and/or hydrothermal treatments. Fig. 4 presents XRD patterns of the
2.3. Characterization The structure of the prepared LDH and the sample produced by the heat or hydrothermal treatments were examined X-ray diffraction (XRD) with monochromated CuKα radiation (RINT-2000, Rigaku). In the case of the film measurement, the sample film with the metal substrate was fixed on a glass slide. The surface morphology of the prepared film was observed by field emission scanning electron microscopy (FE-SEM, JSM-6500F, JEOL) with an acceleration voltage of 15 kV. The chemical states of the composed metal were examined by X-ray photoelectron spectroscopy (XPS, Kratos Axis-Ultima, Shimadzu) using monochromated AlKα radiation. For inductively couple plasma (ICP) analysis and structural refinement, the mass of the film sample
Fig. 1. XRD patterns of the mNi1Co LDH powder prepared by reflux process with urea at 120 °C with different chemical compositions (m = 1, 2, 4 and 8).
238
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
Fig. 2. XRD patterns of the 4Ni1Co LDH films prepared on Ti substrate by reflux process with urea at 120 °C for different synthesis period.
heated Ni-Co LDH powders and films. For the powder samples, heat treatment at 400 °C produced broad diffraction lines, which may be due to a cubic Ni1-xCoxO-like phase (Lu et al., 2011), which is an intermediate solid solution of NaCl-type CoO (ICSD No. 191776) and NiO (ICSD No. 182948). At 600 °C, the LiNi1-xCoxO2 phase (there are some similar phases, e.g., ICSD No. 83303) formed with lower crystallinity and a very low level of impurities, which can be confirmed at 30°–35°. At 800 °C, the LiNi1-xCoxO2 phase showed good crystallinity. In contrast, the intensities of the XRD reflections for the film on the Ti substrate at 600 °C were very small due to the collapse and dropout from the Ti substrate. In addition, LiTiO2 (ICSD No. 28323), an unwanted phase, remained on the substrate. However, the film did not collapse on the Ni substrate at the same temperature. In this sample, the LiNi1-xCoxO2 phase showed relatively good crystallinity and a very small amount of the impurity phase, LixNi2-xO2 (there are some similar phases; e.g., ICSD No. 71422) emerges. When heated up to 800 °C, the XRD pattern (data not shown) could not be measured because both films were broken and only the Ti and Ni substrates were observed. Fig. 5 shows the XRD patterns of the Ni-Co LDH powder treated hydrothermally at 120 and 200 °C, and film at 90, 120, and 200 °C. These patterns confirm that the LDH phase apparently transformed to other phases. The interlayer space decreased rapidly to approximately 0.5 nm after the hydrothermal treatment. However, the total intensity
Fig. 4. XRD patterns of the heated 4N1Co LDH powder at 400, 600 and 800 °C and film deposited on Ti and Ni substrate at 600 °C.
of the reflections was not large, and LiNi1-xCoxO2, Ni1-xCoxOOH, and Ni1-xCox(OH)2 phases (Fujita et al., 1997) were very difficult to distinguish due to their similar XRD patterns. Therefore, SXRD was performed. Fig. 6 presents FE-SEM images of the LDH film, LDH film heated on a Ni substrate at 600 °C and the film treated hydrothermally at 120 °C. From these micrographs, the film heated at 600 °C tended to restructure to form slightly larger particles compared to the original texture. On the other hand, the hydrothermal treatment at 90 and 120 °C can maintain the rose petal-like texture. From the photographs, the thickness of each plate as a petal appeared to decrease by hydrothermal treatment. Such a decrease might be provided by the reduced interlayer space due to the transformation of the LDH phase. In addition, some pinholes, such as a moth hole, were confirmed to generate within the plates, in which the size appears to increase slightly. At 200 °C, the particle shape became thicker without any change to each petal size. For the patterns of the film and powder in Fig. 5, the crystal phase and crystallinity are probably similar, and the powder sample can be regarded as a film sample for further structural analysis. To analyze
Fig. 3. FE-SEM photographs of the 4Ni1Co LDH films prepared on Ti substrate by reflux process with urea at 120 °C for different synthesis period.
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
Fig. 5. XRD patterns of the hydrothermally treated 4Ni1Co LDH powder at 120 and 200 °C and film deposited on Ti substrate at 90, 120 and 200 °C.
the effect of the heating and hydrothermal treatment, the crystal structure of the treated samples were examined by SXRD, XPS, and ICP elemental analyses.
3.3. Structural confirmation 3.3.1. XPS analysis Fig. 7 presents the Co, Ni 2p3/2, and O 1 s XP spectra of the asprepared LDH film, LDH powder treated hydrothermally at 120 °C and LDH powder heated to 800 °C. For the heated samples, the film on the Ni substrate heated at 600 °C is the only sample, in which the Ni 2p3/ 2 cannot be measured due to overlap between the sample and substrate. Therefore, powder sample heated at 800 °C was used.
239
For Co 2p3/2 spectra, broad peaks at approximately 784–786 eV were observed, which were assigned to a shake-up mechanism due to the non-closed shell of the d-electron orbital in the Co cation. For the as-prepared LDH, the main peak was split into two peaks at 781.6 and 780.2 eV, which correspond to the Co(OH)2 and CoOOH(or CoO) phases, respectively. These spectra confirmed that both trivalent and divalent Co2+ cations exist in this sample (Jiang et al., 2015). In the O 1 s spectrum, the peak assigned to the OH− group was observed at 531.3 eV. By the hydrothermal treatment, the peak at 780.2 eV increased slightly due to the increase in trivalent Co3 +. However, because there are no LDH phases and the sample was composed of Ni1-xCox(OH)2 (see Fig. 5), some proton defects might occur in the hydroxide of this sample, whose chemical formula can be expressed as Ni 1-x CoxO 2 H2-δ . This phase was an intermediate state between NiOOH and Ni(OH)2 (Deabate et al., 2006). From the O 1 s spectra of this sample, a new peak emerged at 529.6 eV, which is probably derived from O 2 − (Payne et al., 2012). The existence of O2 − results from both Ni 1-xCoxOOH and/or proton defects for metal hydroxide. For the samples heated to 800 °C, the peak by the shake-up mechanism apparently increases. Such an increase might result from the change in ionic valence. In this spectrum, a peak at 779.9 eV was derived from the LiCoO2-related phase (Dahéron et al., 2009; Moses et al., 2007). For Ni 2p3/2, all spectra showed broad peaks corresponding to a shake-up mechanism at approximately 862 eV, which is the same as the Co 2p3/2 spectra. All spectra were composed of two peaks, around 855.5 and 857.2 eV. The 855.5 eV peaks in the as-prepared LDH and HT 120 °C sample resulted from the existence of Ni(OH)2 (Grosvenor et al., 2006). The 857.2 eV peaks possibly indicate the existence of a NiOOH-like structure (Payne et al., 2012). For the 800 °C heated sample, the 854.7 eV peak may be due to a LiNiO2-related phase (Moses et al., 2007) and the 857.4 eV peak may be assigned to the Ni2O3 phase. Consequently, these XP spectra confirm that the metal oxide found in the 800 °C sample and hydroxide exists mainly in the HT 120 °C sample. In the next section, crystal structure and composition are confirmed by SXRD.
Fig. 6. FE-SEM micrographs of heated 4Ni1Co LDH films on Ni at 600 °C and hydrothermally treated 4Ni1Co LDH films on Ti at 90, 120 and 200 °C.
240
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
Fig. 7. XPS spectra of as prepared and hydrothermally treated 4Ni1Co LDH film on Ti at 120 °C, and heated 4ni1Co LDH powder at 800 °C.
3.3.2. SXRD refinement Fig. 8 shows the SXRD patterns of the as-prepared LDH powder, LDH powder heated at 800 °C and LDH powder treated hydrothermally at 120 °C. The patterns were refined using the Rietveld method (Izumi and Momma, 2007). Table 1 lists the refinement results of the SXRD patterns. Three phases were observed in the as-prepared LDH and HT 120 °C samples, and a single phase was found in the sample heated at 800 °C. The chemical compositions of the phases in these samples were determined by ICP for 800 °C sample. For the LDH and HT 120 °C samples, however, the chemical compositions could not be determined because each sample had three phases. Therefore, these chemical compositions were estimated based on the M–O distance by the ionic radii of 69, 56, 65, 54.5, and 140 pm for low-spin Ni2+, Ni3+, Co2+, Co3+, and O2−, respectively, due to six-fold coordination of the metal cations. For the third phases in the LDH and HT 120 °C samples,
Fig. 8. SXRD patterns of as prepared 4Ni1Co LDH powder, hydrothermally treated and heated 4Ni1Co LDH powder at 120 and 800 °C, respectively.
the amounts were less than several mol% and are too small to estimate accurately. In the case of the as-prepared LDH, there were three phases, two LDH and a very small amount of Ni1-xCoxOOH. The molar ratios of each phase were approximately 0.51, 0.48, and 0.01, respectively. From the FT-IR spectra (data not shown), the absorbance assigned to CO23 − can be observed (Li et al., 2010). Although a very small amount of NO− 3 might also be included in the sample, the amounts of CO23 − and NO− 3 in each sample cannot be determined accurately. Thermogravimetric-differential thermal analysis (TG-DTA) (data not shown) revealed approximately 5%, 21%, and 7% mass losses in below 250 °C, from 250 to 350 °C, and above 350 °C, respectively. These values show relatively good agreement with the intercalated H2O, H2O from OH− group, and CO2 from CO2− estimated by the Rietveld refinement 3 of approximately 7, 17, and 5 mass%, respectively. Therefore, all the intercalated anions were fixed to CO23 − in these phases. In these two LDH phases, the lattice constant, a, is different because of the different Co/Ni ratios in these phases. The lattice constant, c, is also different due to the different amounts of interlayer H2O. The total amounts of the interlayer H2O estimated by TG-DTA was approximately 7 mass%, which is similar to the estimated total amount of H2O by the Rietveld refinement, as mentioned above. In these cases, the smaller Co/Ni phase indicates a larger amount of H2O contained within the interlayer space. Otherwise, a very small amount of Ni1-xCoxOOH phase, in which all metal cations should be trivalent, exists at concentrations greater than 1% in this sample. The total chemical ratio of Co / (Co + Ni) of this sample was calculated to be approximately 0.24 according a Rietveld refinement of the SXRD data. The ratio is larger than that determined by ICP analysis (0.14) due to the broad XRD reflections and the lack of information for each phase in the SXRD pattern, as shown in Fig. 8. Three phases were observed in the sample treated hydrothermally at 120 °C. The majority phase was mixed metal hydroxide without an interlayer anion. This phase might include the Li cation within the intralayer of the hydroxide due to the similar ion radius of 74 pm for Li+ and 69 nm for Ni2+. Because this phase includes no anions within the interlayer space, the OH group should be dissociated, which is designated as O2H2-δ as indicated in Table 1. Consequently, the intercalated anion had been removed from the LDH phase topotactically without valence change of Co3+, as shown in FE-SEM. In contrast, the Ni0.98Co0.02OOH phase composed of trivalent cations grew to approximately 15%. The molar ratio of Ni and Co was estimated to be 98:2 in this phase. In addition, a small amount of LixMO2 phase of approximately 3% emerges. For the total chemical ratios, the Co/(Co + Ni) ratio of these samples were 0.29 according SXRD and 0.13 by ICP. The
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
241
Table 1 shows the compositional phases in the LDH, HT120 and H600 samples. Sample LDH (x ≈ 0.14)
HT120°C (x ≈ 0.13)
Phase (estimated chemical composition⁎)
Lattice constants/Å
Mean M–O distance/Å
Molar ratio⁎⁎
Ni0.62Co0.38(OH)2(CO3)0.19·0.14H2O
R3m (No. 166)
a = 2.997(6), c = 22.122(1)
2.035(4)
0.508
Ni0.90Co0.10(OH)2(CO3)0.05·0.73H2O
R3m (No. 166) P3m1 (No. 156)
a = 3.1077(3), c = 23.606(5)
2.0748(1)
0.484
a = 2.799(6), c = 4.899(3) a = 3.12655(6), c = 4.6068(2)
– 2.07748(4)
0.008 0.817
R3m (No. 166)
a = 2.8110(4), c = 4.9159(6) a = 2.83548(8), c = 14.06(5)
1.960(1) –
0.152 0.031
R3m (No. 166)
a = 2.87600(3), c = 14.2020(1)
1.971(0)
1.000
Ni1-xCoxOOH Li0.02Ni0.62Co0.35O2H2–δ Ni0.98Co0.02OOH Li(Ni1-xCox)O2
800 °C
Space group
(Li0.95Ni0.05)(Li0.09Ni0.76Co0.17)O2
P3m1 (No. 164) P3m1 (No. 156)
⁎ The ratios of Li, Ni and Co in each sample were determined by ICP for 800 °C sample, and were determined from M–O distance by Rietan-FP for as prepared LDH and HT120°C samples. ⁎⁎ The molar ratio of each phases in the sample were determined by Rietan-FP.
difference apparently increased compared to the LDH samples. The reason for such a large discrepancy can be explained by the ionic radii of the divalent cations. For the mixed metal hydroxide phase, most of the metal was comprised of divalent cations, Ni2 + and Co2 +, which have similar ionic radii of around 69 and 65 pm, respectively. Because the reciprocal of the difference indicates the error size, the error must be 3–4 times larger than that of the LDH phase composed of Ni2+ and Co3+ with radii of 69 and 54.5 pm, respectively. Basically, these discrepancies mean that the general ionic radii may make it impossible to express the crystal structure of LDH and the hydroxide phases completely. The result of the structural refinement shows that the hydrothermal treatment with LiOH tends to transform from LDH to a deionized layered hydroxide topotactically. For the sample heated to 800 °C, only the LixMO2 phase forms. The refinement of the SXRD pattern confirmed that several atomic% of Li and Ni were exchanged with each other (Idemoto et al., 2009). The chemical composition can be refined as (Li0.95Ni0.05)(Li0.09Ni0.76Co0.17)O2 by the Rietveld method. From these results, the sample was transformed completely to the phase from LDH. The two processes, hydrothermal and heat treatment with LiOH, provide different crystalline phases. The sample treated hydrothermally was composed of anion-removed metal hydroxide as the majority phase. This removal of the intercalated anion was possibly the result of the temporal formation of Li2CO3, which has a relatively low solubility of 1.33 g per 100 mL H2O. The formation reaction can remove the CO3 anion from the LDH phase, and the Li2CO3 formed can be washed with distilled water. The heated sample showed the formation of a LiMO2 (M is mainly Ni and Co)-type phase. This phase emerges from the oxidation of metal, dehydration from hydroxyl groups and Li insertion within the interlayer space. 4. Conclusion LDH has a layered structure composed of edge-sharing M(OH)6 octahedra and interlayer anion for charge compensation. Such a layered structure is similar to that of Ni(OH)2 in the nickel metal-hydride battery and to LiCoO2 in the lithium-ion battery. Therefore, hydrothermal and heat treatments were carried out to induce a topotactic transformation of LDH to M(OH)2 or LiMO2 structures. LDH was prepared on the Ti or Ni substrate by a reflux process with urea. The precipitant, urea, worked well for sharply defined nucleation on the Ti substrate due to the homogeneous increase in pH by the thermal decomposition of urea. A part of the LDH plates form perpendicularly on the Ti substrate. The transformation of LDH was performed by a hydrothermal or heating process. From the SXRD patterns, the major phases of the starting LDH had two different chemical compositions, Co/Ni ≈ 1/9 and 4/6. The hydrothermal treatment resulted in a slight increase in Co3 +, despite the phase change from LDH to metal hydroxide without anion intercalation. These results suggest that the hydroxyl group decomposes partially and the chemical formula of the hydroxide should be Li0.02Ni0.62Co0.35O2H2-δ, and Ni0.98Co0.02OOH phase emerges. The topotactic transformation appears to occur by
taking the film shape during the treatment. This sample can show good charge-discharge properties for use as an electrode in the NiMH battery. For the heated sample, heating to 800 °C can lead to the formation of a complete LixMO2 (M:Ni, Co) phase with an estimated chemical formula of (Li0.95Ni0.05)(Li0.09Ni0.76Co0.17)O2. This heating sample might be used as an electrode in the Li-ion secondary battery.
Acknowledgments This study was supported financially by the Grant-in-Aid for Scientific Research (C) 24560822. The experiments at Spring-8 were performed with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (proposal No. 2015A1004).
References Aisawa, S., Higashiyama, N., Takahashi, S., Hirahara, H., Ikematsu, D., Kondo, H., Nakayama, H., Narita, E., 2007. Intercalation behavior of l–ascorbic acid into layered double hydroxides. Appl. Clay Sci. 35, 146–154. Aisawa, S., Sasaki, S., Takahashi, S., Hirahara, H., Nakayama, H., Narita, E., 2006. Intercalation of amino acids and oligopeptides into Zn–Al Layered double hydroxide by coprecipitation reaction. J. Phys. Chem. Solids 67, 920–925. Biensan, P., Simon, B., Pérès, J.P., de Guibert, A., Broussely, M., Bodet, J.M., Perton, F., 1999. On safety of lithium–ion cells. J. Power Sources 81–82, 906–912. Dahéron, L., Martinez, H., Dedryvère, R., Baraille, I., Ménétrier, M., Denage, C., Delmas, C., Gonbeau, F., 2009. Surface properties of LiCoO2 investigated by XPS Analyses and theoretical calculations. J. Phys. Chem. C 113, 5843–5852. Deabate, S., Fourgeot, F., Henn, F., 2006. Electrochemical behaviour of the β(II)-Ni(OH)2/ β(III)-NiOOH redox couple upon potentiodynamic cycling conditions. Electrochim. Acta 51, 5430–5437. Fujita, Y., Amine, K., Maruta, J., Yasuda, H., 1997. LiNi1-xCoxO2 prepared at low temperature using β-Ni1-xCoxOOH and either LiNO3 or LiOH. J. Power Sources 68, 126–130. Grosvenor, A.P., Biesinger, M.C., Smart, R.St.C., McIntyre, N.S., 2006. New interpretations of XPS spectra of nickel metal and oxides. Surf. Sci. 600, 1771–1779. He, S., Zhao, Y., Wei, M., Duan, X., 2011. Preparation of oriented layered double hydroxide film using electrophoretic deposition and its application in water treatment. Ind. Eng. Chem. Res. 50, 2800–2806. Idemoto, Y., Takahashi, Y., Kitamura, N., 2009. Dependence of property, Crystal Structure and electrode characteristics on Li Content for LixNi0.8Co0.2O2 as a cathode active material for Li. J. Power Sources 189, 269–278. Izumi, F., Momma, K., 2007. Three-dimensional visualization in powder diffraction. Solid State Phenom. 130, 15–20. Jiang, J., Zhang, A., Li, L., Ai, L., 2015. Nickel-cobalt layered double hydroxide nanosheets as high-performance electrocatalyst for oxygen evolution reaction. J. Power Sources 278, 445–451. Kitano, S., Uebou, Y., Wada, H., Aoki, T., Murata, T., 2005. Thermal behavior of overcharged state of lithium–ion cells using LiCoO2 positive electrode. GS Yuasa Technical Report 2, pp. 18–24. Kwon, N.H., Yinm, H., Brodard, P., Sugnaux, C., Fromm, K.M., 2014. Impact of composite structure and morphology on electronic and ionic conductivity of carbon contained LiCoO2 cathode. Electrochim. Acta 134, 215–221. Li, K., Kumada, N., Yonesaki, Y., Takei, T., Kinomura, N., Wang, H., Wang, C., 2010. The pH effects on the formation of Ni/Al Nitrate form layered double hydroxides (LDHs) by chemical deposition and hydrothermal reaction. Mater. Chem. Phys. 121, 223–229. Lu, Y., Wei, M., Yang, L., Li, C., 2007. Synthesis of layered cathode Material Li[CoxMn1−x]O2 from layered double hydroxides precursors. J. Solid State Chem. 180, 1775–1782. Lu, Z., Xia, G., Templeton, J.D., Li, X., Nie, Z., Yang, Z., Stevenson, J.W., 2011. Development of Ni1-xCoxO as the cathode/interconnect contact for solid oxide fuel cells. Electrochem. Commun. 13, 642–645. Moses, A.W., Flores, H.G.G., Kim, J.-G., Langell, M.A., 2007. Surface properties of LiCoO2, LiNiO2 and LiNi1−xCoxO2. Appl. Surf. Sci. 253, 4782–4791.
242
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
Nakahira, A., Kudo, T., Murase, H., 2007. Synthesis of LDH–type Clay Substituted with Fe and Ni ion for arsenic removal and its application to magnetic separation. IEEE Trans. Magn. 43, 2442–2444. Nobili, F., Dsoke, S., Croce, F., Marassi, R., 2005. An Ac impedance spectroscopic study of Mg–doped LiCoO2 at different temperatures: electronic and ionic transport properties. Electrochim. Acta 50, 2307–2313. Park, M., Lee, C.-I., Choy, J.-H., Kim, J.-E., Choi, J., 2004. Layered double hydroxides as potential Solid Base for beneficial remediation of endosulfan–contaminated soils. J. Phys. Chem. Solids 65, 513–516. Payne, B.P., Biesinger, M.C., McIntyre, N.S., 2012. Use of oxygen/nickel ratios in the XPS Characterisation of oxide phases on nickel metal and nickel alloy surfaces. J. Electron Spectrosc. Relat. Phenom. 185, 159–166. Pinnavaia, T.J., Chibwe, M., Constantino, V.R.L., Yun, S.K., 1995. Organic chemical conversions catalyzed by intercalated layered double hydroxides (LDHs). Appl. Clay Sci. 10, 117–129.
Rives, V., 2002. Characterisation of layered double hydroxides and their decomposition products. Mater. Chem. Phys. 75, 19–25. Takei, T., Miura, A., Kumada, N., 2014. Soft–chemical synthesis and catalytic activity of Ni–Al and Co–Al Layered double hydroxides (LDHs) intercalated with anions with different charge density. J. Asian Ceram. Soc. 2, 289–296. Tang, P., Feng, Y., Li, D., 2011. Fabrication and properties of acid yellow 49 dye–intercalated layered double hydroxides film on an alumina–coated aluminum substrate. Dyes Pigments 91, 120–125. Yushi, H., Li, P., XiaoZhen, L., ZiFeng, M., 2008. Preparation and performance of LiNi0.8Co0.2O2 cathode material based on Co–substituted α–Ni(OH)2 precursor. Chin. Sci. Bull. 53, 1324–1328.
Contents lists available at ScienceDirect
Applied Clay Science journal homepage: www.elsevier.com/locate/clay
Research paper
Topotactic transformation of Ni-based layered double hydroxide film to layered metal oxide and hydroxide Takahiro Takei a,⁎, Hiroki Fuse a, Akira Miura b, Nobuhiro Kumada a a b
Center for Crystal Science and Technology, University of Yamanashi, 7-32 Miyamae, Kofu, Yamanashi 400-8511, Japan Faculty of Engineering, Hokkaido University, Kita 13, Nishi 8, Kita-ku, Sapporo 060-8628, Japan
a r t i c l e
i n f o
Article history: Received 10 October 2015 Received in revised form 12 February 2016 Accepted 15 February 2016 Available online 23 February 2016 Keywords: Topotactic transformation Layered double hydroxide Layered metal oxide
a b s t r a c t Topotactic transformation of film-shaped samples comprising a layered double hydroxide (LDH) to a metal oxide phase was performed by a hydrothermal or heat treatment. The degree of transformation was examined by X-ray photoelectron spectroscopy (XPS) and synchrotron X-ray diffraction (SXRD). The LDH film prepared as a raw material by a reflux process with urea was composed of plate-shaped particles that were partially perpendicular to the substrate. For the transformation, the LDH powder and film were treated hydrothermally at 90, 120, and 200 °C in a LiOH aqueous solution or heated to 400, 600, and 800 °C with a LiOH·H2O contact. The film shape and morphology of the LDH particles were retained after the hydrothermal treatment at 120 °C. XPS and SXRD confirmed that two types of LDH with different compositions, Ni:Co = 62:38 and 90:10, were formed in the as-prepared LDH. In the case of the sample treated hydrothermally, metal hydroxide with no intercalated anion within the interlayer space was observed and the metal oxyhydroxide formed with a ratio of approximately 15%. For the heated samples, the transformation to LiNi1-xCoxO2-related materials was carried out by heating to 600 or 800 °C. © 2016 Elsevier B.V. All rights reserved.
1. Introduction Layered double hydroxide (LDH) has a layered structure with anions intercalated between brucite-type hydroxide layers. The hydroxide layer is composed of divalent and trivalent cations. The interlayer anion compensates for the charge discrepancy arising from the occurrence of trivalent cations in the hydroxide layer. Generally, the interlayer anion can be exchanged by ion-exchange or reconstruction processes. For ion exchange, Cl− or NO-3type LDH should be used as raw materials because CO2− is generally most stable within the inter3 layer space. For reconstruction, calcined LDH is immersed into an anion-contained aqueous solution to rehydrate with the existed anion. The structural changes in this process have been examined by X-ray absorption fine structure (XAFS), but the formation is sometimes difficult for transition metal-containing LDH due to facile formation of a spineltype phase (Rives, 2002). For applications of the ion-exchanged LDH, some studies have reported the preparation of functional hybrids (Pinnavaia et al., 1995; Tang et al., 2011; Takei et al., 2014), removal of toxic anions (Nakahira et al., 2007; He et al., 2011; Park et al., 2004), and use of drug delivery systems (Aisawa et al., 2006, 2007). The most common LDH is “hydrotalcite,” which is composed of Mg2+ and Al3+. For the LDH, including hydrotalcite, the divalent and trivalent cations can be replaced with period-four transition metal cations such as Cr, ⁎ Corresponding author. E-mail address: [email protected] (T. Takei).
http://dx.doi.org/10.1016/j.clay.2016.02.018 0169-1317/© 2016 Elsevier B.V. All rights reserved.
Mn, Fe, Co, and Ni, because most period-four transition metals are either di- or trivalent. For the crystal structure of LDH, all metals form M(OH)6 octahedra in which all OH groups are shared. For such polyhedral polycondensation structures, six edges in an octahedron are shared with each other. The structure of the LDH is, actually, quite similar to that of LiCoO2 despite the positive valence of the interlayer ion (Li+). Therefore, some researchers have examined the synthesis of LiCoO2-type cathode materials from LDH as a raw material (Lu et al., 2007; Yushi et al., 2008). For LiCoO2, Co occurs in a trivalent state and the Li cation can be deintercalated from the structure owing to the oxidation of trivalent Co to tetravalent Co during the charge process of a Li-ion battery. When tetravalent Co is reduced to the trivalent state, Li ions can be intercalated again. However, LiCoO2 has a slight risk of decomposing with the generation of O2 gas at relatively low temperatures (above 200 °C). This can be dangerous because the generation can expand the battery cell and sometimes result in an explosion. To improve the decomposition temperature, many researchers have examined some additives, partial substitution or change in composition. LiMn2O4 (spinel structure) and LiNiO2 can decompose above 300 °C and approximately 200 °C, respectively (Biensan et al., 1999; Kitano et al., 2005). The use of LiCoO2 as an electrode in secondary batteries has two disadvantages. One is the low electronic conductivity. As the conductivity of LiCoO2 is approximately 10− 3 S/cm (Kwon et al., 2014), both an adhesive agent and a conduction aid are necessary for the formation of an electrode device, particularly regarding the use of powder-type
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
LiCoO2. The adhesive agent and conduction aid are generally PTFE and carbon, respectively. Because LiCoO2 is a p-type semiconductor, the number of holes will increase with increasing Co4 + concentration. Nobili et al. reported that Mg substitution at the Co site in LiCoO2 results in an increase in electronic conductivity (Nobili et al., 2005). The other disadvantage is the layered structure. The structure can be an advantage because the Li cation can move within the interlayer space; however, it becomes a disadvantage for large LiCoO2 particles. Large plate-like particles sometimes decelerate the intercalation and deintercalation of Li because of the long diffusion path. Because Li diffusion will be predominant, the synthesis of LiCoO2 nanoparticles or the preparation of an oriented structure perpendicular to the substrate will be effective. For these reasons, both the crystal structure and particle orientation for LiCoO2-type compounds are very important. Therefore, the topotactic transformation from LDH is possible. In this study, the preparation of oriented LDH on the substrate and topotactic transformation to Nibased layered metal oxide with a LiCoO2 structure was examined by a hydrothermal and heating process. The crystal structures of the sample were examined by synchrotron-XRD (SXRD). 2. Experimental The transformation was performed for the sample in two forms, powder and film. Both samples were prepared, which were then transformed by heating with LiOH or by a hydrothermal treatment in an aqueous LiOH solution. The typical processes are as follows. 2.1. Preparation of LDH Ni and Co nitrate mixed salt was used as the metal sources at ratios of 1, 2, 4, and 8. Distilled water was added to the salts to achieve a concentration of 5 mmol/L. Urea was placed in the aqueous solution of the mixed salts with a molar ratio of 20. In the case of film preparation, a 10 × 50 × 0.2 mm Ti or Ni plate used as a substrate was soaked perpendicular to the stirring flow in an aqueous solution. The solution was then refluxed with stirring at 120 °C for several hours. The prepared LDH are designated as mNi1Co, where m means the loaded molar ratio of Ni/Co; e.g., 4Ni1Co. 2.2. Transformation treatment of LDH powder and film The obtained LDH underwent a transformation by a heat-treatment or hydrothermal treatment. The heat treatment was performed using LiOH·H2O at 400–800 °C for 3 h in air. The LiOH·H2O to metal cation ratio in LDH was 3:2. The heated samples were washed with distilled water and dried at 50 °C. In the case of the hydrothermal treatment, the LDH samples were soaked in a LiOH aqueous solution with a concentration of 1.0 mol/L. The solution was treated hydrothermally at 90, 120, and 200 °C for 72 h. The treated samples were washed with distilled water and dried. Here, “HT” is defined as a prefix of the hydrothermal temperature for use as the sample name.
237
could not be obtained, which prevented accurate measurements. Therefore, the powder sample was used for these measurements. The chemical composition of the samples was measured by ICP (PS-3500DDII, Hitachi High-Tech Science Corp.) The structures were examined by SXRD using a Debye-Scherrer camera at BL02B2 in SPring-8. A borosilicate capillary tube, 0.2 mm in radius, was filled with the samples. The wavelength was 0.49560 Å, which was determined using CeO2 as the reference crystal. The SXRD patterns were refined using RIETAN-FP (Izumi and Momma, 2007). 3. Results and discussion 3.1. Preparation of LDH film Fig. 1 shows XRD patterns of the Ni-Co LDH powder with different compositions, Ni:Co = 1:1, 2:1, 4:1, and 8:1. The LDH phase can be observed at all compositions. However, the XRD reflection intensity increased gradually depending on the Ni ratio. Generally, because the M2+/M3+ ratio should be 2–4, the chemical composition ranges of 2:1 and 4:1 will be the optimum because the divalent cations tend to be deficient at the higher ratio. In actual experiments, the crystallinity is high and the yield is low for 8Ni1Co. For 2Ni1Co, the yield is good but the crystallinity is somewhat low. Therefore, in this study, a 4:1 composition was adopted for the subsequent experiments. Fig. 2 presents the XRD patterns of the grown LDH films with a Ni:Co ratio of 4:1 on the Ti substrate for different synthesis periods (1.0, 1.5, 2.0, 3.0, and 5.0 h). Their baselines appear to be worse because the substrate Ti metal is pliable and the measured surface is not flat. The halo in the sample at 1.5 and 5.0 h might be from a glass slide. These patterns suggest that the amount of LDH increases depending on the synthesis period. The patterns could not confirm whether the LDH particles were ordered or not. Fig. 3 presents the FE-SEM images of the grown LDH films with a different period, 1.0, 1.5, 2.0, 3.0, and 5.0 h. The top and bottom micrographs show the low and expanded magnifications, respectively. The LDH nuclei begin to form at 1.0 h, and the number of nuclei increases steeply during an additional 1 h. The number of LDH particles perpendicular to the substrate tends to increase over synthesis periods of 3 h or shorter. At 5 h, the particle size appears to become approximately double. However, the excess particles emerge on the perpendicular particles, as shown in the low-magnification image. 3.2. Transformation treatment for LDH The transformation treatment for LDH was carried out by heating and/or hydrothermal treatments. Fig. 4 presents XRD patterns of the
2.3. Characterization The structure of the prepared LDH and the sample produced by the heat or hydrothermal treatments were examined X-ray diffraction (XRD) with monochromated CuKα radiation (RINT-2000, Rigaku). In the case of the film measurement, the sample film with the metal substrate was fixed on a glass slide. The surface morphology of the prepared film was observed by field emission scanning electron microscopy (FE-SEM, JSM-6500F, JEOL) with an acceleration voltage of 15 kV. The chemical states of the composed metal were examined by X-ray photoelectron spectroscopy (XPS, Kratos Axis-Ultima, Shimadzu) using monochromated AlKα radiation. For inductively couple plasma (ICP) analysis and structural refinement, the mass of the film sample
Fig. 1. XRD patterns of the mNi1Co LDH powder prepared by reflux process with urea at 120 °C with different chemical compositions (m = 1, 2, 4 and 8).
238
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
Fig. 2. XRD patterns of the 4Ni1Co LDH films prepared on Ti substrate by reflux process with urea at 120 °C for different synthesis period.
heated Ni-Co LDH powders and films. For the powder samples, heat treatment at 400 °C produced broad diffraction lines, which may be due to a cubic Ni1-xCoxO-like phase (Lu et al., 2011), which is an intermediate solid solution of NaCl-type CoO (ICSD No. 191776) and NiO (ICSD No. 182948). At 600 °C, the LiNi1-xCoxO2 phase (there are some similar phases, e.g., ICSD No. 83303) formed with lower crystallinity and a very low level of impurities, which can be confirmed at 30°–35°. At 800 °C, the LiNi1-xCoxO2 phase showed good crystallinity. In contrast, the intensities of the XRD reflections for the film on the Ti substrate at 600 °C were very small due to the collapse and dropout from the Ti substrate. In addition, LiTiO2 (ICSD No. 28323), an unwanted phase, remained on the substrate. However, the film did not collapse on the Ni substrate at the same temperature. In this sample, the LiNi1-xCoxO2 phase showed relatively good crystallinity and a very small amount of the impurity phase, LixNi2-xO2 (there are some similar phases; e.g., ICSD No. 71422) emerges. When heated up to 800 °C, the XRD pattern (data not shown) could not be measured because both films were broken and only the Ti and Ni substrates were observed. Fig. 5 shows the XRD patterns of the Ni-Co LDH powder treated hydrothermally at 120 and 200 °C, and film at 90, 120, and 200 °C. These patterns confirm that the LDH phase apparently transformed to other phases. The interlayer space decreased rapidly to approximately 0.5 nm after the hydrothermal treatment. However, the total intensity
Fig. 4. XRD patterns of the heated 4N1Co LDH powder at 400, 600 and 800 °C and film deposited on Ti and Ni substrate at 600 °C.
of the reflections was not large, and LiNi1-xCoxO2, Ni1-xCoxOOH, and Ni1-xCox(OH)2 phases (Fujita et al., 1997) were very difficult to distinguish due to their similar XRD patterns. Therefore, SXRD was performed. Fig. 6 presents FE-SEM images of the LDH film, LDH film heated on a Ni substrate at 600 °C and the film treated hydrothermally at 120 °C. From these micrographs, the film heated at 600 °C tended to restructure to form slightly larger particles compared to the original texture. On the other hand, the hydrothermal treatment at 90 and 120 °C can maintain the rose petal-like texture. From the photographs, the thickness of each plate as a petal appeared to decrease by hydrothermal treatment. Such a decrease might be provided by the reduced interlayer space due to the transformation of the LDH phase. In addition, some pinholes, such as a moth hole, were confirmed to generate within the plates, in which the size appears to increase slightly. At 200 °C, the particle shape became thicker without any change to each petal size. For the patterns of the film and powder in Fig. 5, the crystal phase and crystallinity are probably similar, and the powder sample can be regarded as a film sample for further structural analysis. To analyze
Fig. 3. FE-SEM photographs of the 4Ni1Co LDH films prepared on Ti substrate by reflux process with urea at 120 °C for different synthesis period.
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
Fig. 5. XRD patterns of the hydrothermally treated 4Ni1Co LDH powder at 120 and 200 °C and film deposited on Ti substrate at 90, 120 and 200 °C.
the effect of the heating and hydrothermal treatment, the crystal structure of the treated samples were examined by SXRD, XPS, and ICP elemental analyses.
3.3. Structural confirmation 3.3.1. XPS analysis Fig. 7 presents the Co, Ni 2p3/2, and O 1 s XP spectra of the asprepared LDH film, LDH powder treated hydrothermally at 120 °C and LDH powder heated to 800 °C. For the heated samples, the film on the Ni substrate heated at 600 °C is the only sample, in which the Ni 2p3/ 2 cannot be measured due to overlap between the sample and substrate. Therefore, powder sample heated at 800 °C was used.
239
For Co 2p3/2 spectra, broad peaks at approximately 784–786 eV were observed, which were assigned to a shake-up mechanism due to the non-closed shell of the d-electron orbital in the Co cation. For the as-prepared LDH, the main peak was split into two peaks at 781.6 and 780.2 eV, which correspond to the Co(OH)2 and CoOOH(or CoO) phases, respectively. These spectra confirmed that both trivalent and divalent Co2+ cations exist in this sample (Jiang et al., 2015). In the O 1 s spectrum, the peak assigned to the OH− group was observed at 531.3 eV. By the hydrothermal treatment, the peak at 780.2 eV increased slightly due to the increase in trivalent Co3 +. However, because there are no LDH phases and the sample was composed of Ni1-xCox(OH)2 (see Fig. 5), some proton defects might occur in the hydroxide of this sample, whose chemical formula can be expressed as Ni 1-x CoxO 2 H2-δ . This phase was an intermediate state between NiOOH and Ni(OH)2 (Deabate et al., 2006). From the O 1 s spectra of this sample, a new peak emerged at 529.6 eV, which is probably derived from O 2 − (Payne et al., 2012). The existence of O2 − results from both Ni 1-xCoxOOH and/or proton defects for metal hydroxide. For the samples heated to 800 °C, the peak by the shake-up mechanism apparently increases. Such an increase might result from the change in ionic valence. In this spectrum, a peak at 779.9 eV was derived from the LiCoO2-related phase (Dahéron et al., 2009; Moses et al., 2007). For Ni 2p3/2, all spectra showed broad peaks corresponding to a shake-up mechanism at approximately 862 eV, which is the same as the Co 2p3/2 spectra. All spectra were composed of two peaks, around 855.5 and 857.2 eV. The 855.5 eV peaks in the as-prepared LDH and HT 120 °C sample resulted from the existence of Ni(OH)2 (Grosvenor et al., 2006). The 857.2 eV peaks possibly indicate the existence of a NiOOH-like structure (Payne et al., 2012). For the 800 °C heated sample, the 854.7 eV peak may be due to a LiNiO2-related phase (Moses et al., 2007) and the 857.4 eV peak may be assigned to the Ni2O3 phase. Consequently, these XP spectra confirm that the metal oxide found in the 800 °C sample and hydroxide exists mainly in the HT 120 °C sample. In the next section, crystal structure and composition are confirmed by SXRD.
Fig. 6. FE-SEM micrographs of heated 4Ni1Co LDH films on Ni at 600 °C and hydrothermally treated 4Ni1Co LDH films on Ti at 90, 120 and 200 °C.
240
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
Fig. 7. XPS spectra of as prepared and hydrothermally treated 4Ni1Co LDH film on Ti at 120 °C, and heated 4ni1Co LDH powder at 800 °C.
3.3.2. SXRD refinement Fig. 8 shows the SXRD patterns of the as-prepared LDH powder, LDH powder heated at 800 °C and LDH powder treated hydrothermally at 120 °C. The patterns were refined using the Rietveld method (Izumi and Momma, 2007). Table 1 lists the refinement results of the SXRD patterns. Three phases were observed in the as-prepared LDH and HT 120 °C samples, and a single phase was found in the sample heated at 800 °C. The chemical compositions of the phases in these samples were determined by ICP for 800 °C sample. For the LDH and HT 120 °C samples, however, the chemical compositions could not be determined because each sample had three phases. Therefore, these chemical compositions were estimated based on the M–O distance by the ionic radii of 69, 56, 65, 54.5, and 140 pm for low-spin Ni2+, Ni3+, Co2+, Co3+, and O2−, respectively, due to six-fold coordination of the metal cations. For the third phases in the LDH and HT 120 °C samples,
Fig. 8. SXRD patterns of as prepared 4Ni1Co LDH powder, hydrothermally treated and heated 4Ni1Co LDH powder at 120 and 800 °C, respectively.
the amounts were less than several mol% and are too small to estimate accurately. In the case of the as-prepared LDH, there were three phases, two LDH and a very small amount of Ni1-xCoxOOH. The molar ratios of each phase were approximately 0.51, 0.48, and 0.01, respectively. From the FT-IR spectra (data not shown), the absorbance assigned to CO23 − can be observed (Li et al., 2010). Although a very small amount of NO− 3 might also be included in the sample, the amounts of CO23 − and NO− 3 in each sample cannot be determined accurately. Thermogravimetric-differential thermal analysis (TG-DTA) (data not shown) revealed approximately 5%, 21%, and 7% mass losses in below 250 °C, from 250 to 350 °C, and above 350 °C, respectively. These values show relatively good agreement with the intercalated H2O, H2O from OH− group, and CO2 from CO2− estimated by the Rietveld refinement 3 of approximately 7, 17, and 5 mass%, respectively. Therefore, all the intercalated anions were fixed to CO23 − in these phases. In these two LDH phases, the lattice constant, a, is different because of the different Co/Ni ratios in these phases. The lattice constant, c, is also different due to the different amounts of interlayer H2O. The total amounts of the interlayer H2O estimated by TG-DTA was approximately 7 mass%, which is similar to the estimated total amount of H2O by the Rietveld refinement, as mentioned above. In these cases, the smaller Co/Ni phase indicates a larger amount of H2O contained within the interlayer space. Otherwise, a very small amount of Ni1-xCoxOOH phase, in which all metal cations should be trivalent, exists at concentrations greater than 1% in this sample. The total chemical ratio of Co / (Co + Ni) of this sample was calculated to be approximately 0.24 according a Rietveld refinement of the SXRD data. The ratio is larger than that determined by ICP analysis (0.14) due to the broad XRD reflections and the lack of information for each phase in the SXRD pattern, as shown in Fig. 8. Three phases were observed in the sample treated hydrothermally at 120 °C. The majority phase was mixed metal hydroxide without an interlayer anion. This phase might include the Li cation within the intralayer of the hydroxide due to the similar ion radius of 74 pm for Li+ and 69 nm for Ni2+. Because this phase includes no anions within the interlayer space, the OH group should be dissociated, which is designated as O2H2-δ as indicated in Table 1. Consequently, the intercalated anion had been removed from the LDH phase topotactically without valence change of Co3+, as shown in FE-SEM. In contrast, the Ni0.98Co0.02OOH phase composed of trivalent cations grew to approximately 15%. The molar ratio of Ni and Co was estimated to be 98:2 in this phase. In addition, a small amount of LixMO2 phase of approximately 3% emerges. For the total chemical ratios, the Co/(Co + Ni) ratio of these samples were 0.29 according SXRD and 0.13 by ICP. The
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
241
Table 1 shows the compositional phases in the LDH, HT120 and H600 samples. Sample LDH (x ≈ 0.14)
HT120°C (x ≈ 0.13)
Phase (estimated chemical composition⁎)
Lattice constants/Å
Mean M–O distance/Å
Molar ratio⁎⁎
Ni0.62Co0.38(OH)2(CO3)0.19·0.14H2O
R3m (No. 166)
a = 2.997(6), c = 22.122(1)
2.035(4)
0.508
Ni0.90Co0.10(OH)2(CO3)0.05·0.73H2O
R3m (No. 166) P3m1 (No. 156)
a = 3.1077(3), c = 23.606(5)
2.0748(1)
0.484
a = 2.799(6), c = 4.899(3) a = 3.12655(6), c = 4.6068(2)
– 2.07748(4)
0.008 0.817
R3m (No. 166)
a = 2.8110(4), c = 4.9159(6) a = 2.83548(8), c = 14.06(5)
1.960(1) –
0.152 0.031
R3m (No. 166)
a = 2.87600(3), c = 14.2020(1)
1.971(0)
1.000
Ni1-xCoxOOH Li0.02Ni0.62Co0.35O2H2–δ Ni0.98Co0.02OOH Li(Ni1-xCox)O2
800 °C
Space group
(Li0.95Ni0.05)(Li0.09Ni0.76Co0.17)O2
P3m1 (No. 164) P3m1 (No. 156)
⁎ The ratios of Li, Ni and Co in each sample were determined by ICP for 800 °C sample, and were determined from M–O distance by Rietan-FP for as prepared LDH and HT120°C samples. ⁎⁎ The molar ratio of each phases in the sample were determined by Rietan-FP.
difference apparently increased compared to the LDH samples. The reason for such a large discrepancy can be explained by the ionic radii of the divalent cations. For the mixed metal hydroxide phase, most of the metal was comprised of divalent cations, Ni2 + and Co2 +, which have similar ionic radii of around 69 and 65 pm, respectively. Because the reciprocal of the difference indicates the error size, the error must be 3–4 times larger than that of the LDH phase composed of Ni2+ and Co3+ with radii of 69 and 54.5 pm, respectively. Basically, these discrepancies mean that the general ionic radii may make it impossible to express the crystal structure of LDH and the hydroxide phases completely. The result of the structural refinement shows that the hydrothermal treatment with LiOH tends to transform from LDH to a deionized layered hydroxide topotactically. For the sample heated to 800 °C, only the LixMO2 phase forms. The refinement of the SXRD pattern confirmed that several atomic% of Li and Ni were exchanged with each other (Idemoto et al., 2009). The chemical composition can be refined as (Li0.95Ni0.05)(Li0.09Ni0.76Co0.17)O2 by the Rietveld method. From these results, the sample was transformed completely to the phase from LDH. The two processes, hydrothermal and heat treatment with LiOH, provide different crystalline phases. The sample treated hydrothermally was composed of anion-removed metal hydroxide as the majority phase. This removal of the intercalated anion was possibly the result of the temporal formation of Li2CO3, which has a relatively low solubility of 1.33 g per 100 mL H2O. The formation reaction can remove the CO3 anion from the LDH phase, and the Li2CO3 formed can be washed with distilled water. The heated sample showed the formation of a LiMO2 (M is mainly Ni and Co)-type phase. This phase emerges from the oxidation of metal, dehydration from hydroxyl groups and Li insertion within the interlayer space. 4. Conclusion LDH has a layered structure composed of edge-sharing M(OH)6 octahedra and interlayer anion for charge compensation. Such a layered structure is similar to that of Ni(OH)2 in the nickel metal-hydride battery and to LiCoO2 in the lithium-ion battery. Therefore, hydrothermal and heat treatments were carried out to induce a topotactic transformation of LDH to M(OH)2 or LiMO2 structures. LDH was prepared on the Ti or Ni substrate by a reflux process with urea. The precipitant, urea, worked well for sharply defined nucleation on the Ti substrate due to the homogeneous increase in pH by the thermal decomposition of urea. A part of the LDH plates form perpendicularly on the Ti substrate. The transformation of LDH was performed by a hydrothermal or heating process. From the SXRD patterns, the major phases of the starting LDH had two different chemical compositions, Co/Ni ≈ 1/9 and 4/6. The hydrothermal treatment resulted in a slight increase in Co3 +, despite the phase change from LDH to metal hydroxide without anion intercalation. These results suggest that the hydroxyl group decomposes partially and the chemical formula of the hydroxide should be Li0.02Ni0.62Co0.35O2H2-δ, and Ni0.98Co0.02OOH phase emerges. The topotactic transformation appears to occur by
taking the film shape during the treatment. This sample can show good charge-discharge properties for use as an electrode in the NiMH battery. For the heated sample, heating to 800 °C can lead to the formation of a complete LixMO2 (M:Ni, Co) phase with an estimated chemical formula of (Li0.95Ni0.05)(Li0.09Ni0.76Co0.17)O2. This heating sample might be used as an electrode in the Li-ion secondary battery.
Acknowledgments This study was supported financially by the Grant-in-Aid for Scientific Research (C) 24560822. The experiments at Spring-8 were performed with the approval of the Japan Synchrotron Radiation Research Institute (JASRI) (proposal No. 2015A1004).
References Aisawa, S., Higashiyama, N., Takahashi, S., Hirahara, H., Ikematsu, D., Kondo, H., Nakayama, H., Narita, E., 2007. Intercalation behavior of l–ascorbic acid into layered double hydroxides. Appl. Clay Sci. 35, 146–154. Aisawa, S., Sasaki, S., Takahashi, S., Hirahara, H., Nakayama, H., Narita, E., 2006. Intercalation of amino acids and oligopeptides into Zn–Al Layered double hydroxide by coprecipitation reaction. J. Phys. Chem. Solids 67, 920–925. Biensan, P., Simon, B., Pérès, J.P., de Guibert, A., Broussely, M., Bodet, J.M., Perton, F., 1999. On safety of lithium–ion cells. J. Power Sources 81–82, 906–912. Dahéron, L., Martinez, H., Dedryvère, R., Baraille, I., Ménétrier, M., Denage, C., Delmas, C., Gonbeau, F., 2009. Surface properties of LiCoO2 investigated by XPS Analyses and theoretical calculations. J. Phys. Chem. C 113, 5843–5852. Deabate, S., Fourgeot, F., Henn, F., 2006. Electrochemical behaviour of the β(II)-Ni(OH)2/ β(III)-NiOOH redox couple upon potentiodynamic cycling conditions. Electrochim. Acta 51, 5430–5437. Fujita, Y., Amine, K., Maruta, J., Yasuda, H., 1997. LiNi1-xCoxO2 prepared at low temperature using β-Ni1-xCoxOOH and either LiNO3 or LiOH. J. Power Sources 68, 126–130. Grosvenor, A.P., Biesinger, M.C., Smart, R.St.C., McIntyre, N.S., 2006. New interpretations of XPS spectra of nickel metal and oxides. Surf. Sci. 600, 1771–1779. He, S., Zhao, Y., Wei, M., Duan, X., 2011. Preparation of oriented layered double hydroxide film using electrophoretic deposition and its application in water treatment. Ind. Eng. Chem. Res. 50, 2800–2806. Idemoto, Y., Takahashi, Y., Kitamura, N., 2009. Dependence of property, Crystal Structure and electrode characteristics on Li Content for LixNi0.8Co0.2O2 as a cathode active material for Li. J. Power Sources 189, 269–278. Izumi, F., Momma, K., 2007. Three-dimensional visualization in powder diffraction. Solid State Phenom. 130, 15–20. Jiang, J., Zhang, A., Li, L., Ai, L., 2015. Nickel-cobalt layered double hydroxide nanosheets as high-performance electrocatalyst for oxygen evolution reaction. J. Power Sources 278, 445–451. Kitano, S., Uebou, Y., Wada, H., Aoki, T., Murata, T., 2005. Thermal behavior of overcharged state of lithium–ion cells using LiCoO2 positive electrode. GS Yuasa Technical Report 2, pp. 18–24. Kwon, N.H., Yinm, H., Brodard, P., Sugnaux, C., Fromm, K.M., 2014. Impact of composite structure and morphology on electronic and ionic conductivity of carbon contained LiCoO2 cathode. Electrochim. Acta 134, 215–221. Li, K., Kumada, N., Yonesaki, Y., Takei, T., Kinomura, N., Wang, H., Wang, C., 2010. The pH effects on the formation of Ni/Al Nitrate form layered double hydroxides (LDHs) by chemical deposition and hydrothermal reaction. Mater. Chem. Phys. 121, 223–229. Lu, Y., Wei, M., Yang, L., Li, C., 2007. Synthesis of layered cathode Material Li[CoxMn1−x]O2 from layered double hydroxides precursors. J. Solid State Chem. 180, 1775–1782. Lu, Z., Xia, G., Templeton, J.D., Li, X., Nie, Z., Yang, Z., Stevenson, J.W., 2011. Development of Ni1-xCoxO as the cathode/interconnect contact for solid oxide fuel cells. Electrochem. Commun. 13, 642–645. Moses, A.W., Flores, H.G.G., Kim, J.-G., Langell, M.A., 2007. Surface properties of LiCoO2, LiNiO2 and LiNi1−xCoxO2. Appl. Surf. Sci. 253, 4782–4791.
242
T. Takei et al. / Applied Clay Science 124-125 (2016) 236–242
Nakahira, A., Kudo, T., Murase, H., 2007. Synthesis of LDH–type Clay Substituted with Fe and Ni ion for arsenic removal and its application to magnetic separation. IEEE Trans. Magn. 43, 2442–2444. Nobili, F., Dsoke, S., Croce, F., Marassi, R., 2005. An Ac impedance spectroscopic study of Mg–doped LiCoO2 at different temperatures: electronic and ionic transport properties. Electrochim. Acta 50, 2307–2313. Park, M., Lee, C.-I., Choy, J.-H., Kim, J.-E., Choi, J., 2004. Layered double hydroxides as potential Solid Base for beneficial remediation of endosulfan–contaminated soils. J. Phys. Chem. Solids 65, 513–516. Payne, B.P., Biesinger, M.C., McIntyre, N.S., 2012. Use of oxygen/nickel ratios in the XPS Characterisation of oxide phases on nickel metal and nickel alloy surfaces. J. Electron Spectrosc. Relat. Phenom. 185, 159–166. Pinnavaia, T.J., Chibwe, M., Constantino, V.R.L., Yun, S.K., 1995. Organic chemical conversions catalyzed by intercalated layered double hydroxides (LDHs). Appl. Clay Sci. 10, 117–129.
Rives, V., 2002. Characterisation of layered double hydroxides and their decomposition products. Mater. Chem. Phys. 75, 19–25. Takei, T., Miura, A., Kumada, N., 2014. Soft–chemical synthesis and catalytic activity of Ni–Al and Co–Al Layered double hydroxides (LDHs) intercalated with anions with different charge density. J. Asian Ceram. Soc. 2, 289–296. Tang, P., Feng, Y., Li, D., 2011. Fabrication and properties of acid yellow 49 dye–intercalated layered double hydroxides film on an alumina–coated aluminum substrate. Dyes Pigments 91, 120–125. Yushi, H., Li, P., XiaoZhen, L., ZiFeng, M., 2008. Preparation and performance of LiNi0.8Co0.2O2 cathode material based on Co–substituted α–Ni(OH)2 precursor. Chin. Sci. Bull. 53, 1324–1328.