Article pubs.acs.org/JPCC
Simultaneous Quantification of Hydride Ions and Electrons Incorporated in 12CaO·7Al2O3 Cages by Deuterium-Labeled Volumetric Analysis Toshihiro Yoshizumi,*,† Yoji Kobayashi,‡ Hiroshi Kageyama,‡ and Katsuro Hayashi† †
Secure Materials Center, Materials and Structures Laboratory, Tokyo Institute of Technology, 4259 Nagatsuta, Yokohama 226-8503, Japan ‡ Department of Energy and Hydrocarbon Chemistry, Graduate School of Engineering, Kyoto University, Nishikyo-ku, Kyoto 615-8510, Japan ABSTRACT: An extended volumetric method, combined with quadrupole mass spectroscopy (QMS), is proposed. This method enables us to distinguish and simultaneously quantify hydride (H−) ions and electrons (e−) incorporated in cages of 12CaO·7Al2O3 (C12A7), which is accomplished upon annealing with CaH2. When a sample is dissolved in a deuterium chloride solution, most of the H− ions and electrons released from cages react to form HD and D2 molecules, respectively. These isotope-labeled molecules are then detected by QMS. We have used this method to follow the concentrations of H− ions and electrons in C12A7-treated CaH2 over the thermal treatment time. We found that during the initial seven days of treatment, both concentrations increased. Thereafter, the electron concentration begins to decrease, while the H− ion concentration continues to increase toward the theoretical maximum. This diverging behavior is due to differences in their diffusion ratios and thermodynamic stabilities. (e−).15 The total anion concentration is equivalent to the positive charge of the lattice, and the theoretical maximum concentration for the monovalent anions is 2.3 × 1021 cm−3. The stoichiometric unit cell comprises two O2− ions (corresponding to the concentration of 1.2 × 1021 cm−3), which are responsible for the fast O2− ion conductivity at high temperature.16 The incorporation of electrons into the cages causes an electrical conductivity ranging from less than 10−10 to ∼103 S cm−1 at room temperature, depending on the concentrations of electrons and other anions within the cages. C12A7, fully doped with electrons, is also known as “electride”, which exhibits a metallic conductivity17 and a superconducting transition at 30 mK.18 Electron doping can be achieved by either the chemical reduction or ultraviolet irradiation of H−doped C12A7 at room temperature. In each process, H− ions release two electrons via its conversion to H+ ions.14 The electron-doped C12A7 exhibits an excellent electron emission with a low work function of 2.4 eV,19 and acts as a reductant in organic reactions that even work in aqueous solution.20 These results reflect the wide range of properties achievable with C12A7 upon anion incorporation. However, no quantification methods can simultaneously cover the possible concentration ranges and differentiate between the various anionic species. For example, electron paramagnetic resonance (EPR) cannot be applied to metallic C12A7, as the reliable
1. INTRODUCTION Volumetric analysis is one of the simplest and most accurate methods for the determination of nonstoichiometry in metal hydrides.1 When a metallic lanthanide hydride, for example, is dissolved in HCl solution, the resulting reaction can be described as LnHx + 3HCl(aq) → Ln(III)Cl3(aq) + 1/2(3 + x)H 2↑ (1)
Thus, the hydride (H−) ion content can be readily calculated from the sample weight and the volume of evolved H2 gas. Hydride compounds have recently been extended to mixed anion hydrides, in which the H− ions and other anions, such as halide,2,3 oxide,4,5 and pnictide6 ions, occupy the anionic sites. The determination of H− ion content in a mixed hydride is generally more complex, because the redox species responsible for the additional hydrogen evolution involves altervalent cations and electron-occupied anion defects. Therefore, the effects of the H− ions and other redox species need to be separated. 12CaO·7Al2O3 (C12A7) is a metal oxide that incorporates both H− ions and electron-like F+ centers (a single electron in an oxide ion vacancy). This study examines the simultaneous determination of H− ion and electron concentrations in C12A7 cages using extended volumetric analysis. The unit cell, [Ca24Al28O64]4+, is composed of 12 cages with an inner diameter of ∼0.4 nm. These cages incorporate various anionic species to compensate for the positive charge. The anionic species include O2−, OH−,7,8 F−,9 Cl−,9 O−,10 O2−,10,11 O22− 12 and S2− 13 ions, as well as H− 14 ions and electrons © 2012 American Chemical Society
Received: November 7, 2011 Revised: March 23, 2012 Published: March 23, 2012 8747
dx.doi.org/10.1021/jp2106742 | J. Phys. Chem. C 2012, 116, 8747−8752
The Journal of Physical Chemistry C
Article
Table 1. Sample Annealing Conditions and Concentrations for Volumetry/QMS, 1H-NMR, and Iodometry volumetry and QMS sample 1: C12A7:H− 2: C12A7:H− 3: C12A7:H− 4: C12A7:H− 5: C12A7: e−
annealing temperature (°C)
annealing time (days)
800
2
800
4
800
7
800
14
900
1
H − NMR H− (cm−3)
1
H− (cm−3)
e− (cm−3)
1.11(+0.19, −0.18) × 1021 1.39(±0.17) × 1021 1.53(±0.18) × 1021 2.22(+0.08, −0.12) × 1021 2.3(±0.4) × 1020
5.1(+0.9, −0.8) × 1020 6.7(±0.8) × 1020 7.5(±0.9) × 1020 5 × 1019
iodometry e− (cm−3)
calculated value from results of volumetry O2− (cm−3)
4 × 1020
3.4(+1.2, −1.3) × 1020
1.2 × 1021
1.2(+1.3, −1.2) × 1020
1.2 × 1021
1(+13, −1) × 1019
1.7 × 1021
1(+6, −1) × 1019
1.10(±0.20) × 1021
6.9(±2.0) × 1020
4.8(±1.2) × 1020
10−5 Pa. The background mass spectra within the chamber was measured and subtracted from each measured spectrum. Before each measurement, the subchamber was cleaned via multiple purges with pure Ar gas. The collected gas was then sampled with a syringe and injected into the subchamber. The sample gas was introduced via a variable leak valve to balance the pressure in the main chamber at 10−4−10−3 Pa. The signal intensities with mass-charge ratios (m/e) of 2, 3, and 4 were assigned to the H2, HD, and D2, respectively. The sensitivities of 2 and 4 m/e were calibrated using pure H2 and D2. An interpolated sensitivity value was used for m/e = 3 (i.e., HD). The absolute concentration of each species in the collected mixed gas was calculated from the total gas volume and the intensity ratio obtained by QMS. The quantified concentrations of each species may be influenced by various experimental factors, such as material type, the particle size of the material, dissolving temperature, DCl concentration, and agitation. Care was taken to maintain consistent experimental conditions to ensure the same branching ratio for other samples. To validate the H− content, as determined by the volumetry/ QMS method, 1H magic angle spinning nuclear magnetic resonance (1H-MAS NMR) spectra were measured with a Bruker Biospin DSX-400 spectrometer. The pulse sequence comprised a single pulse with a length of 2.5 μs and an interval of 175 s. The rotation frequency was 15 kHz. Chloroform (+7.25 ppm with respect to tetramethylsilane, TMS) was used as secondary reference for the chemical shift. A fully OH−incorporated C12A7 ([OH−] = 2.3 × 1021 cm−3), as prepared by annealing in a wet N2 atmosphere, was used as a quantification standard.23 The concentration of H− ions was evaluated by comparing the H− signal area (from +10 to +2.5 ppm) with the standard sample. Photoabsorption was measured with a Hitachi U-4000 spectrophotometer to check the electron concentration determined using the volumetry/QMS method. Bulk samples were polished to a thickness of 50 μm and a mirrored surface. The absorption coefficient at ∼2.1 eV, located in the tail region of the F+ band, which typically peaks at ∼2.8 eV,15 was used to measure the electron concentration.
evaluation limit in terms of electron concentration is approximately 20% of the theoretical maximum due to the metallic conductivity. An alternative method is iodometry, which exploits the redox process for I−/I2 in aqueous solution.21 However, the presence of H− ions hinders an accurate determination of the electron concentration, since both species can reduce I2. This can also be the case for conventional volumetry using acid; the presence of electrons (in electrides) can reduce the accuracy in H− quantification. This is because both species have lower redox potentials than that for H2/H+ to evolve hydrogen from water. To solve this problem, the volumetric method is extended by using a solution of DCl in D2O in conjunction with isotope analysis. When the H− ions and electrons reduce this solution, they evolve into, in principle, two different gaseous species, HD and D2, respectively. These species can be distinguished via mass spectrometry, which enables us to simultaneously quantify the H− ions and electrons.
2. EXPERIMENTAL SECTION Both H− ions and electrons were incorporated into the samples using the same process, as described in ref 22. A Czochralskigrown C12A7 single crystal was cut into slices with dimensions of 0.5 × 7 × 12 mm. These slices were wrapped in platinum foil (thickness of 50 μm) with CaH2 powder, and sealed in evacuated silica-glass tubes. All chemical reagents were supplied by Wako Pure Chemical Industries, Ltd., Japan. The glass capsules were annealed at 800 °C for 2, 4, 7, and 14 days, respectively. A sample doped with electrons was prepared using the process described in ref 17. Slices of a single crystal, with a thickness of 1 mm, were sealed with titanium shot in an evacuated silica-glass tube and annealed at 900 °C for 1 day. After annealing, the reacted surface layer was removed by grinding. The annealing conditions for all samples are summarized in Table 1. For the volumetry measurement, a solution of 0.1 M DCl in D2O (DCl-D2O) was prepared in an Einhorn fermentation tube with a maximum gas collection volume of 5 mL and a total liquid reservoir volume of ∼15 mL. Bulk samples with a weight of 20−30 mg were dissolved in the DCl-D2O solution at 18 °C, and the total volumes of evolved gas were measured. The samples were not ground, to minimize the effects of surface degradation and overly rapid gas evolution upon dissolution in DCl-D2O. The relative ratios of the gaseous species in the collected gas were measured with a quadrupole mass spectrometer (QMS), installed in a high vacuum chamber and equipped with a small gas injection subchamber. The background pressure in the main chamber was lower than 1 ×
3. RESULTS After treatment with CaH2, the samples turned color from green to black, suggesting the incorporation of electrons into the cages. The total volume of gas, V, as generated by the volumetry procedure, ranged from 0.2 to 1.7 mL. The total molar amount, n, of collected gas is described as 8748
dx.doi.org/10.1021/jp2106742 | J. Phys. Chem. C 2012, 116, 8747−8752
The Journal of Physical Chemistry C n=
Article
spectra in Figure 1. The signal intensity is weaker than that for a typical sample over multiple injections. Upon comparing the spectra for 2 and 14 days, the signal at 2 m/e increases together with the increasing signal at 3 m/e. These observations suggest that the signal at 2 m/e can be ascribed to H2 gas being generated from H− ions during sample dissolution. This is because only H− ions from C12A7 cages are available as a source for H2 gas, except for any contaminant hydrogen species. The presence of H2 may be due to isotopic impurities in the D2O. Other than this, we speculate that the following reactions are possible:
V (pat − pwl − pD O ) 2
(2) RT where R is the gas constant, T is the temperature, pat is atmospheric pressure (1.01 × 105 Pa), pwl is the decrease in pressure, caused by the difference in water surface height between the collected gas chamber and the water reservoir, and pD2O (1.7 × 103 Pa at 18 °C) is the vapor pressure of D2O.24 Possible errors include reading of the Einhorn fermentation tube scale, which is estimated to be ±0.05 mL; dissolution of H2, HD and D2 molecules in the DCl-D2O solution, which correspond to ∼0.3 mL from available solubility data;25 and fluctuations in the temperature within ±1 °C. In total, typical error for the volumetry method was ∼4−13% depending on the experiment run. The mass spectra for the gases evolved from samples annealed for 2 and 14 days are shown in Figure 1. In the 2-day
H− → 1/2H 2 ↑ +e−
(5)
e− + D3O+ → 1/2D2 ↑ +D2 O
(6)
which are overall described as H− + D3O+ → 1/2H 2 ↑ +1/2D2 ↑ +D2 O
(7)
It is not possible to verify the mechanisms above (eqs 5 and 6) from the QMS data alone, and thus this is only a phenomenological assumption. Equation 7 assumes that the H− ions are never transferred to the solution, namely, they do not form HDO, H2O, HxD3−xO+ (x = 1, 2, and 3) or OH−, and they are all evolved as gaseous species. In any event, combining eqs 3, 4, and 7, we find that the ratio of H− ions to electrons within a sample is given by [H−]: [e−] = I3 + 2I2: 2I4 − 2I2
where I is the calibrated relative intensity of the QMS signal, and the subscript is the m/e ratio. Assuming this equation, the contributions of H− ions and electrons to the mass spectrum signals in Figure 1 have been shown with red and blue, respectively. Other possible side reactions will be discussed in section 4. All of the results obtained using volumetry and QMS are listed in Table 1. The total error for the quantified values was estimated to be ∼4−17%. Figure 2 plots the H− ion and electron concentrations, together with O2− ion concentration, against the annealing time. The single crystal, before annealing with CaH2, is stoichiometric (i.e., [Ca24Al28O64]4+(O2−)2), and hence O2− ions are the only species present within cages at that point. After annealing with CaH2, according to the thermodynamic
Figure 1. QMS spectra for the evolved gases obtained from samples annealed with CaH2 for 2 and 14 days. Red and blue areas represent the H− ions and electrons, respectively. See text for details.
spectrum, two main peaks at m/e = 3, 4 and a weak peak at 2 were observed. The intensity of that at 4 m/e in the 14-day spectrum was weaker than that of the 2-day sample. The main peaks at 3 and 4 are assigned to HD and D2, respectively. They are principally formed via the following reactions: H− + D3O+ → HD ↑ +D2 O
(3)
e− + D3O+ → 1/2D2 ↑ +D2 O
(4)
(8)
We see the formation of HD (m/e = 3) gas and changes in the formation ratio of HD and D2, with different annealing times. The HD gas is evolved from H− ions during the destruction of the host lattice (eq 3). It makes sense that the electrons released from the C12A7 cages reduce D3O+ to D2 gas via the reaction pathway in eq 4, as the dissolution of C12A7 electrides in hydrochloric acid reduce the potential of the solution with respect to the hydrogen evolution potential.21 Thus, according to reactions 3 and 4, these two species can be distinguished. There are two possible origins for the weak signal at 2 m/e: a H2+ ion generated by the single ionization of a H2 molecule upon electron-impact ionization in the QMS measurement, or a D+ ion generated through the fragmentation of HD or D2 molecules. The fragmentation ratio, D+/D2+, in our apparatus was estimated to be 0.6% from calibration measurements using pure D2 gas. The contribution to the signal at 2 m/e is very small compared to that from the H2 molecule itself. The background spectrum for the vacuum chamber also exhibits a weak signal at 2 m/e, which has been subtracted from the
Figure 2. Annealing time dependencies for the H− ion, electron, and O2− ion concentrations determined by volumetry/QMS. Lines are drawn for a guide to the eye. 8749
dx.doi.org/10.1021/jp2106742 | J. Phys. Chem. C 2012, 116, 8747−8752
The Journal of Physical Chemistry C
Article
discussion in ref 26, it is unfavorable for OH− ions and electrons to coexist in the cages. We have confirmed this by 1HMAS NMR measurements, as described below. Thus, the only possible species present in the cages are H− ions, electrons, and O2− ions. The concentration of O2− ions is calculated with the restriction of total charges for the anionic species in the cages: 2[O2 −] = 2.3 × 1021 − [H−] − [e−](cm−3)
(9)
2−
Over the initial 7 days, the O ion concentration rapidly decreased and was mostly replaced with electrons and H− ions. Interestingly, after 7 days, the electron concentration decreases, while the H− ion continues to increase. The concentration of H− ions increased monotonically, and nearly reached the theoretical maximum after 14 days. These exchange behaviors for the anionic species within the cages are plausible, considering the differences in the diffusion rates and thermodynamic stabilities. This will be further discussed in section 4. The quantification measurements for the H− ion concentrations using the present method were validated using 1HMAS NMR. The NMR spectra are shown in Figure 3. The
Figure 4. Correlation between the H− ion concentrations obtained by volumetry/QMS and NMR.
Figure 5. Relationship between electron concentrations obtained by volumetry/QMS and optical absorption measurements. The inset shows the absorption spectra. Relative electron concentrations were estimated from the absorption coefficient at 2.1 eV.
Figure 3. 1H-MAS NMR spectra measured for the samples annealed with CaH2 for 2, 4, 7, and 14 days. C12A7, fully incorporated with OH− ions, was used as a standard sample, which is indicated by the blue line. Peaks at 5.1 and −0.8 ppm originate from the H− and OH− ions, respectively.
at 2.1 eV was used as a measure for the relative electron concentration. Unfortunately, the signal-to-noise ratio in the spectra was insufficient to determine the absolute concentration. Nevertheless, a linear correlation was observed between the values obtained for the two methods. Additionally, a C12A7 electride sample (No. 6 in Table 1) was prepared by annealing with a titanium getter. The electron concentration was measured by iodometry21 and volumetry/ QMS. The electron concentrations obtained were 7 × 1020 and 1.1 × 1021 cm−3, respectively. This reasonable agreement again supports the validity of our method.
signal at 5.1 ppm is assigned to the H− ions in cages, while that in the reference sample at −0.8 ppm represents OH− in the cages. All CaH2-annealed samples contain negligible levels of OH−. The broad signals between ∼2.5 to −5 ppm are attributed to adsorbed surface water, or hydroxide compounds formed upon surface degradation. The concentrations of H− ion were estimated by comparing the signal areas for H− and OH− and fixing the total anion concentration at 2.3 × 1021 cm−3. Figure 4 compares the H− ion concentrations obtained by NMR to those from the volumetry/QMS method. The data exhibits a linear correlation, which implies that volumetry can effectively measure H− ion concentrations. The error bars for the NMR results are larger than those from the volumetric measurements, so the data scatter is probably due more to inaccuracies associated with the quantification by NMR than otherwise. Optical absorption measurements were used to check the measured electron concentrations. Figure 5 shows relative electron concentrations obtained via absorption measurements plotted against those obtained using the volumetry/QMS method. The inset presents the raw spectra. Since high electron concentrations cause deep coloration, the absorption coefficient
4. DISCUSSION We have demonstrated that a volumetric analysis, combined with isotope labeling, enables us to quantify the H− ion and electron concentrations incorporated in C12A7 cages. This section will discuss reaction pathways providing the reasons why the H− ion is not collected as HD molecule. This has significant implications on the H− ion-to-electron ratio determination. We will also examine the diverging incorporation behavior for the H− ions and electrons in C12A7 cages as a function of time 4.1. H− Not Forming HD. The QMS results for the sampled gases show that the H− ions in cages are converted to 8750
dx.doi.org/10.1021/jp2106742 | J. Phys. Chem. C 2012, 116, 8747−8752
The Journal of Physical Chemistry C
Article
This is expected, as having an H− ion in the cage is much more stable than having an electron in the cage when a hydrogen source is present:26
not only HD molecules, but also H2 molecules. A similar behavior has been observed for metal hydrides dissolved in D2O by Wender et al.27 For both reactions (eq 3 and 7), all the H− ions in cages are assumed to transfer one electron to the solution and be evolved as gaseous species. Other possible reaction pathways, which have not yet been considered, include the transfer of not only electrons, but also nuclei to the acid solution, which would be accompanied with a two-electron reduction of the acid (or, in other terms, an acid−base reaction involving proton/deuteron scrambling). One representative reaction is H− + D3O+ → D2 ↑ +HDO
2e−(c) + H 2(g) → 2H−(c),
(12)
where (c) indicates a species in the cage. Unfortunately, no data for the H− ion diffusion rate in C12A7 is available. According to the potential diagram for the hydrogen species in C12A7, as discussed in ref 29, the rate-determining elemental step for H− ion diffusion is likely ascribed to H−(c) → H 0(c) + e−(c),
(10)
ΔHH 0 = 3.5 eV
(13)
The electron in a cage diffuses rapidly with the intercage hopping process, and migration of the neutral H0 is easier than the ionic H−. However the reaction of eq 13 requires an enthalpy change up to 3.5 eV, which suggests that the thermal formation of H0 is very limited, and thus the apparent H− ion diffusion rate in C12A7 is very low. Thus, it is possible that the incorporation behavior for the H− ion is determined not from the surface reactions at the C12A7−CaO−CaH2 interfaces but from the bulk diffusion for the H− ion in C12A7. In conclusion, the observed behaviors in Figure 2 originate from (1) the large differences in diffusion rate for the O2− and H− ions and (2) the higher thermodynamic stability of the H− ion compared to the electron.
Products from the two electron reduction reactions include H2O, HxD3−xO+ (x = 1, 2, and 3), and OH− ions, as well as HDO. All of these reactions would form the same number of moles of D2 as the H− ions. Here, it is assumed that the D2 molecules formed by such reactions, as represented by eq 10, share the QMS signal at 4 (I4′). Equation 8 then becomes [H−]: [e−] = I3 + 2I2 + I4′: 2(I4 − I4′) − 2I2
ΔHH− = −4.5eV
(11)
This suggests that eq 8 underestimates the actual ratio of [H−]: [e−] and gives the lower limit. Unfortunately, the value of I4′ itself cannot be measured. However, the upper limit for I4′ can be estimated from the data for the 14-day sample. This assumes that all the signals at 2, 3, and 4 originate from H− ions. The ratio for I2:I3:I4 was measured as 0.075:1:0.087; from this, the value for I4 − I2 was calculated to be 0.012, giving the upper limit for I4′. This value corresponds to approximately 1% of H− ions reacting via the two-electron transfer reaction represented by eq 10. However, the 14-day sample actually contains electrons on the order of 1020 cm−3, and hence the actual ratio for the two-electron pathway may be reduced to 0.5%. Considering this value, a reasonable error in eq 7, caused by the two-electron transfer path, may be +1% for [H−] and −2% for [e−]. 4.2. Incorporation Behaviors for H− and Electrons. The incorporation behaviors for the H− ions and electrons in the C12A7 cages are probed from the results in Figure 2. Before the annealing, the O2− ions are the only species occupying the cages. The rapid decrease in O2− ion concentration, in the initial 7 days, suggest that oxygen extraction occurs preferentially, leaving electrons in the cages. At a glance, this behavior is convincing because C12A7 is a fast O2− ion conductor, with an activation energy of 0.81 eV.16 However, the rate of oxygen extraction is most likely limited by the surface reaction rather than the bulk diffusion of the O2− ion. This is because the diffusion rate for the O2− ion at 800 °C (3.5 × 10−6 − 1.1 × 10−7 cm−2·s−1)16,28 in the stoichiometric C12A7 is enough for the penetration of O2− ion into a disk sample with a thickness of 0.5 mm within only 1 day. The apparent bulk diffusion constant for the O2− ion, as estimated from the result in Figure 2, is lower by 2−3 orders of magnitude than that expected from the diffusion constant at 800 °C. Throughout the annealing process, an interface layer of CaO is grown between C12A7 and CaH2 with a reaction of the CaH2 and the O2− ion diffused out of the C12A7. The chemical diffusion of oxygen in the CaO layer may determine the overall oxygen extraction rate in C12A7 cages. The H− ions in cages are formed via the incorporation of hydrogen in the electronrich C12A7. Despite the H− ion concentration increasing sluggishly after 7 days, the concentration continues to increase and reaches the theoretical maximum upon further annealing.
5. CONCLUSIONS The concentrations of H− ions and electrons in C12A7 cages have been simultaneously quantified using volumetry in conjunction with deuterium labeling in a deuterium chloride solution (DCl-D2O). The majority of H− ions and electrons are converted to HD and D2 molecules, respectively, during the dissolution of the C12A7 in DCl-D2O. This allows us to distinguish between the two species. A minor proportion of the H− ions form H2 and D2, but we found this to have only a minor influence on the H−/e− ratio determination. The observed incorporation behaviors for the H− ions and electrons in C12A7 cages are explained by the faster diffusion rate for the O2− ion compared to that of the H− ion, and the higher thermodynamic stability for the H− ion than the electron. The present method is applicable to the characterization of oxyhydrides, with either altervalent cations or electron trapping anion defects.
■
AUTHOR INFORMATION
Corresponding Author
*Address: Secure Materials Center, Materials and Structures Laboratory, Tokyo Institute of Technology, R3-34, 4259 Nagatsuta, Yokohama 226-8503, Japan. Tel: +81-45-9245375; Fax: +81-45-924-5365; E-mail:
[email protected]. titech.ac.jp. Notes
The authors declare no competing financial interest.
■
ACKNOWLEDGMENTS This work was supported by a Grant-in-Aid for Elements Science and Technology (No. 08055013), and a Grant-in-Aid for Young Researchers A (No. 19685019) from the MEXT, Japanese Government.
■
REFERENCES
(1) Conder, K.; Kaldis, E. J. Less-Common Met. 1989, 146, 205−211.
8751
dx.doi.org/10.1021/jp2106742 | J. Phys. Chem. C 2012, 116, 8747−8752
The Journal of Physical Chemistry C
Article
(2) Schaefer, G. J. Phys. Chem. Solids 1960, 12, 233−244. (3) Pohl, R. W. Proc. Phys. Soc. 1937, 49, 3−31. (4) Hayward, M. A.; Cussen, E. J.; Claride, J. B.; Bieringer, M.; Rosseinsky, M. J.; Kiely., M. J.; Blundell, S. J.; Marshall, I. M.; Pratt, I. M. Science 2002, 295, 1882−1884. (5) Huang, B.; Corbett, J. D Inorg. Chem. 1998, 37, 1892−1899. (6) Hanna, T.; Muraba, Y.; Matsuishi, S.; Igawa, N.; Kodama, K.; Shamoto, S.; Hosono, H. Phys. Rev. B 2011, 84, 024521. (7) Imlach, J. A.; Glasser, L. S. D.; Glasser, P. F. Cem. Concr. Res. 1971, 1, 57−61. (8) Lee, D. K.; Kogel, L.; Ebbinghaus, S. G.; Valov., I.; Wiemhoefer, H. D.; Lerch, M.; Janek, J. Phys. Chem. Chem. Phys. 2009, 11, 3105− 3114. (9) Jeevaratnam, J.; Glasser, F. P.; Glasser, L. S. D. J. Am. Ceram. Soc. 1964, 47, 105−106. (10) Hayashi, K.; Hirano, M.; Matsuishi, S.; Hosono, H. J. Am. Chem. Soc. 2002, 124, 738−739. (11) Hosono, H.; Abe, Y. Inorg. Chem. 1987, 26, 1192−1195. (12) Hayashi, K.; Hirano, M.; Hosono, H. Chem. Lett. 2005, 34, 586−587. Kajihara, K.; Matsuishi, S.; Hayashi, K.; Hirano, M.; Hosono, H. J. Phys. Chem. C 2007, 111, 14855−14861. (13) Zhmoidin, G. I.; Chatterjee, A. K. Cem. Concr. Res. 1984, 14, 386−396. (14) Hayashi, K.; Matsuishi, S.; Kamiya, T.; Hirano, M.; Hosono, H. Nature 2002, 124, 738−739. (15) Matsuishi, S.; Toda, Y.; Miyakawa, M.; Hayashi, K.; Kamiya, T.; Hirano, M.; Tanaka, I.; Hosono, H. Science 2003, 301, 626−629. (16) Lacerda, M.; Irvine, J. T. S.; Glasser, F. P.; West, A. R. Nature 1988, 332, 525−526. (17) Kim, S. W.; Matsuishi, S.; Nomura, T.; Kubota, Y.; Takata, M.; Hayashi, K.; Kamiya, T.; Hirano, M.; Hosono, H. Nano Lett. 2007, 7, 1138−1143. (18) Miyakawa, M.; Kim, S. W.; Hirano, M.; Kohama, Y.; Kawaji, H.; Atake, T.; Ikegami, H.; Kono, K.; Hosono, H. J. Am. Chem. Soc. 2007, 129, 7270. (19) Toda, Y.; Yanagi, H.; Ikenaga, E.; Kim, J. J.; Kobata, M.; Ueda, S.; Kamiya, T.; Hirano, M.; Kobayashi, K.; Hosono, H. Adv. Mater. 2007, 19, 3564−3569. (20) Haritha, B.; Toda, Y.; Hirano, M.; Hosono, H.; Takeuchi, D.; Osakada, K. Org. Lett. 2007, 9, 4287−4289. (21) Yoshizumi, T.; Matsuishi, S.; Kim, S. W.; Hosono, H.; Hayashi, K. J. Phys. Chem. C 2010, 114, 15354−15357. (22) Hayashi, K. J. Solid State Chem. 2011, 184, 1428−1432. (23) Hayashi, K. J. Phys. Chem. C 2011, 115, 11003−11009. (24) Besley, L.; Bottomley, G. A. J. Chem. Thermodyn. 1973, 5, 397− 410. (25) Young, C. L., Solubility Data Series; Pergamon Press: Oxford, 1981; Vol. 5/6, p 4. (26) Hayashi, K.; Hirano, M.; Hosono, H. Bull. Chem. Soc. Jpn. 2007, 80, 872−884. (27) Wender, I.; Friedel, R. A.; Orchin, M. J. Am. Chem. Soc. 1949, 71, 1140. (28) Hayashi, K.; Matsuishi, S.; Hirano, M.; Hosono, H. J. Phys. Chem. B 2004, 108, 8920−8925. (29) Sushko, P. V.; Shluger, A. L.; Hayahi, K.; Hirano, M.; Hosono, H. Phys. Rev. B 2006, 73, 045120.
8752
dx.doi.org/10.1021/jp2106742 | J. Phys. Chem. C 2012, 116, 8747−8752
