13332
2008, 112, 13332–13335 Published on Web 08/12/2008
Maximized Proton Conductivity of the HPF6 Clathrate Hydrate by Structural Transformation Jong-Ho Cha,† Kyuchul Shin,† Sukjeong Choi,† Sangyong Lee,‡ and Huen Lee*,† Department of Chemical and Biomolecular Engineering, Korea AdVanced Institute of Science and Technology (KAIST), 335 Gwahangno, Yuseong-gu, Daejeon 305-701, Republic of Korea, and Department of Chemical and Natural Gas Engineering, Texas A&M UniVersity, KingsVille, Texas 78363 USA ReceiVed: June 23, 2008; ReVised Manuscript ReceiVed: July 18, 2008
The unique and specific interaction of ionic guests with the surrounding cage framework might play a key role in promoting the electrochemical properties of ionic clathrate hydrates. In this work we focus on addressing (1) structural transformation simply due to an increase of water content, (2) structure-dependent ionic conductivity, (3) the existence of maximum ionic conductivity at a specific hydration number, and (4) proton migration through channels in the crystalline hydrate matrix. The melting temperature and ionic conductivity of hexafluorophosphoric acid hexahydrate (HPF6 · 6.0H2O) are found to be approximately 29.5 °C and 10-1 S · cm-1, respectively, showing that HPF6 · 6.0H2O possesses desirable features as a solid proton conductor. Introduction Nonionic clathrate hydrates are stabilized by van der Waals interaction between guest molecules and the host framework. Owing to the peculiar inclusion behavior, they have been applied to versatile energy resources, gas storage, and carbon dioxide sequestration.1-4 In contrast, ionic clathrate hydrates are generated by an ionic interaction between an ionic guest and a surrounding host water framework, which quite differs from nonionic clathrate hydrates.5,6 Most ionic clathrate hydrates are known to form unique structures, largely depending on the size and valence of the cations or anions as well as the hydration number. In particular, we have found that the water content in ionic clathrate hydrates plays a significant role in inducing structural transformation and thus affects the phase behavior and physical properties.7 Ionic clathrate hydrates have received attention in relation to the development of proton conductors, because both complex polyatomic and simple monatomic ions incorporated in cage lattices can function as proton conductors.8-11 However, despite that ionic clathrate hydrates have been widely studied, relative little work has been carried out to find a favorable structure with respect to obtaining high ionic conductivity. The HPF6 clathrate hydrate has been reported to have two different crystalline hydrate phases according to its hydration number. In these hydrate structures, the PF6- ion is known to occupy the 51262 cages of the structure-I (sI) hydrate and the 4668 cages of structure-VII (sVII) for HPF6 · 7.7H2O and HPF6 · 6.0H2O, respectively.5,12-14 Meanwhile, the countercation, namely proton, exists in the host lattice, and thus the negative charge from the anion is compensated. An interesting feature of HPF6 hydrate is that the structure is transformed from sI to sVII by a small reduction in the amount of H2O. More importantly, the melting point drastically increases from ap* To whom correspondence should be addressed. Fax: 82-42-350-3910. Tel: 82-42-350-3917. E-mail:
[email protected]. † Korea Advanced Institute of Science and Technology. ‡ Texas A&M University.
10.1021/jp805510g CCC: $40.75
proximately -40 °C for sI to +30 °C for sVII,13,14 lowering the chemical potential of the fresh host lattice formation. Considering the low melting temperature of acid clathrate hydrates such as HBF4 · 5.75H2O and HClO4 · 5.5H2O,14 the high melting temperature of HPF6 · 6.0H2O might contribute to its use as a solid proton conductor with high ionic conductivity. Generally, the ionic conductivity is strongly influenced by the microstructure of the conductors, and thus the two different crystallographic structures of HPF6 hydrates could provide different environments for the ionic motions. In spite of the high ionic conductivity of acid clathrate hydrate arising from the high mobility of the protons,15,16 measurement of the ionic conductivity of these two types of HPF6 hydrates has not yet been attempted. In this context, we have found that structural transformation to sVII in HPF6 hydrates strongly influences their ionic conductivities. As a result the conductivity value tends to be proportional to the composition ratio of sVII. Based on this observation, we suggest that the channel patterns created through structural transformation and the weaker interaction between the PF6- ion and proton in sVII could contribute to an increase in ionic transportation. Accordingly, in this study, we provide direct evidence of the effect of structural transformation to sVII on enhancing ionic conductivity, which leads to extremely high conductivity. Experimental Section Reagents. HPF6 aqueous solution (65 wt% of HPF6) was purchased from Fluka. Water of ultrahigh purity was obtained from a Millipore purification unit. Experimental Measurements. HPF6 · 6.0H2O was synthesized by methods reported in the literature.13 HPF6 · 6.6, 7.2, and 7.7H2O having a hydration number higher than 6.0 were prepared by simply adding H2O to liquid state HPF6 · 6.0H2O and were subsequently crystallized at -70 °C for at least 1 day, resulting in the formation of solid phase mixture of sI and sVII. HPF6 · 4.2, 4.7, and 5.3H2O were synthesized by adding H2O 2008 American Chemical Society
Letters
J. Phys. Chem. C, Vol. 112, No. 35, 2008 13333
Figure 1. Temperature dependence of ionic conductivity for HPF6 · 6.0H2O.
to 65 wt% commercial HPF6 aqueous solution with HF impurity.17 The powder X-ray diffraction (PXRD) patterns for the hydrate samples ground to a fine powder (∼200 µm) were recorded on a Rigaku D/MAX-2500 with low temperature equipment. A light source was the graphite-monochromatized Cu KR1 radiation with wavelength of 1.5406 Å at generator voltage of 40 kV and current of 300 mA. The PXRD experiments were carried out in step mode with a fixed time of 2 s and a step size of 0.03° for 2θ ) 5∼60° for each hydrate sample. During the measurements the temperature was maintained to about -180 °C. The ionic conductivities for the HPF6 hydrate samples were obtained by a complex impedance analysis using a Solartron 1260 impedance/gain-phase analyzer and a 1287 electrochemical interface. This apparatus was connected to Teflon-coated cell containing a pair of Pt-coated Teflon electrode (1.1 cm × 0.6 cm), the two being 5.5 mm apart. The electrodes in the cell were immersed in liquid state of the hydrate sample. The electrolyte was saturated with bubbling argon, and then crystallized at -70 °C, which is identical to sample preparation condition of PXRD and differential scanning calorimeter (DSC) analyses. The real and imaginary parts of complex impedance were plotted over frequency range from 1 to 106 Hz and the ionic conductivities could be calculated from the bulk resistance (Rb) found in the complex impedance diagram. The temperaturedependent phase patterns were determined in a sealed gold pan at the cooling and heating rates of 2 °C · min-1 under N2 atmosphere using a NETZSCH DSC 204 F1. Results and Discussion HPF6 · 6.0H2O is known to have a relatively high melting temperature above the ambient temperature unlike other acid clathrate hydrates such as HBF4 · 5.75H2O and HClO4 · 5.5H2O. We thus attempted to extend the ionic conductivity measurements up to 26 °C, and the results are shown in Figure 1. The ionic conductivity values are of an order of magnitude of approximately 10-1 S · cm-1 and tend to increase with temperature (activation energy ) 0.10 eV). Here, note that the ionic conductivities of HPF6 · 6.0H2O are slightly higher than those of tetramethylammonium hydroxide decahydrate (Me4NOH · 10H2O) which is known as a superionic conductor.8 The more important feature for practical application is that HPF6 · 6.0H2O shows a much higher melting temperature than Me4NOH · 10H2O (+29.5 vs -20.0 °C). Subsequently, HPF6 · nH2O samples of several different hydration numbers (n ) 4.2, 4.7, 5.3, 6.0, 6.6, 7.2, and 7.7) were examined to identify
Figure 2. Ionic conductivities of HPF6 hydrates with a variation of hydration number at -78.5 (filled circles), -30.0 (triangles), -1.0 (diamonds), and +26.0 °C (unfilled circle).
the effect of hydration number on the ionic conductivity. The resultant data are shown in Figure 2. In the first trial, we lowered the measurement temperature of ionic conductivity to -78.5 °C. At this isotherm the increase of the hydration number promotes the ionic conductivity, reaching a maximum value of 6.18 × 10-3 S · cm-1 for HPF6 · 6.0H2O. However, above 6.0H2O, the ionic conductivity considerably decreases, reaching the lowest value, 3.92 × 10-4 S · cm-1, for HPF6 · 7.7H2O. The ionic conductivity of HPF6 · 6.0H2O is approximately 20 times higher than that of HPF6 · 7.7H2O and HPF6 · 4.2H2O, respectively. At two other isotherms, -30.0 and -1.0 °C, the maximum was also observed at a hydration number of 6.0. However, the increase of temperature tends to cause the solid hydrate phase to be melted by dissociation of sI phase at n > 6.0 and the existence of excess PF6- ions at n < 6.0, even though the amount of coexisting liquid phase is actually quite small. It should also be noted that it is necessary to be cautious during the measurements, because the liquid appearance might enhance the ionic conductivity, leading to an erroneous value. For polymer and liquid electrolytes, the ionic conductivity tends to increase according to the concentration of charge carriers. However, the HPF6 hydrates fail to follow this pattern, exhibiting a maximum value at HPF6 · 6.0H2O. Interestingly, HPF6 · 6.0H2O coincides with the stoichiometric composition of sVII. At the present stage, it is worth noting that the HPF6 hydrates possess two intrinsic hydrate structures, which are drastically transformed by inclusion/exclusion of a small H2O amount.12-14 With the observation of maximum ionic conductivity it could be conjectured that the most favorable structure for transporting the charge carriers is formed by controlling the H2O content in the hydrates and reorganizing the host water lattices. Through direct ionic conductivity measurements we have observed that the H2O composition in the HPF6 hydrates is a key variable for changing aspects of their physicochemical nature, such as lattice structure and ionic transport. In this context, using the DSC and PXRD, we subsequently evaluated the thermal behavior and structural pattern that could be sensitive to the hydration number. Figure 3 exhibits the DSC curves of the HPF6 hydrates measured at temperature range of -70∼35 °C. HPF6 · 6.0, 6.6, 7.2, and 7.7H2O were found to possess two distinct endothermic peaks at roughly +25 and -45 °C for heating cycle, indicating the existence of two different solid
13334 J. Phys. Chem. C, Vol. 112, No. 35, 2008
Letters
Figure 3. DSC curves of HPF6 · nH2O with a variation of hydration number (n ) 4.2, 4.7, 5.3, 6.0, 6.6, 7.2, and 7.7).
phases in the HPF6 hydrates. The peaks at higher temperature correspond to a sVII phase transition, whereas those at lower temperature indicate the occurrence of the sI phase transition.13,14 From the DSC pattern of HPF6 · 7.7H2O, we observe that the sI phase is dominant over the sVII phase, but note that the sI peak greatly reduces as the hydration number is decreased. On the contrary, at a low hydration number of 4.2, the sVII peak intensity appears to be weak, but continuously increases, reaching a maximum at the idealized sVII stoichiometric hydration number of 6.0. Furthermore, below 6.0H2O, the sI peaks almost disappear and the corresponding sVII peaks become smaller. The peak ratio of sVII to sI roughly matches with the corresponding phase composition ratio. The DSC curves in Figure 3 reveal that this ratio is maximal at HPF6 · 6.0H2O, finally reaching the ideal state of the sVII phase. Here, we note again that the maximum ratio appears at a hydration number of 6.0, where the corresponding ionic conductivity is also maximal. Accordingly, sVII HPF6 · 6.0H2O is verified to be the most favorable structure for the transportation of charge carriers. The structural transformation of the HPF6 hydrates was confirmed through comparison of the PXRD patterns according to the hydration number (Figure 4). The resulting unit cell parameters are listed in Table 1. As also shown in the DSC results, the observed PXRD reflections confirm that two different hydrate structures of sVII and sI coexist for HPF6 · 6.6, 7.2, and 7.7H2O. For HPF6 · 7.7H2O, the sI phase becomes more dominant than the sVII, while the reverse trend appears for HPF6 · 6.6 and 7.2H2O. However, at low hydration numbers below 6.0, sI disappears and sVII prevails in the hydrate phase. HPF6 · 4.7 and 5.3H2O maintain sVII, but HPF6 · 4.2H2O exhibits a complex structural pattern that differs from that of sI and/or sVII (Supporting Information, Figure S1). This unexpected unique structure might arise from the excessive PF6- ions. The present PXRD results clearly indicate that the structural transformation tends to proceed toward sVII up to a hydration number of 6.0. Thus, we arrive at the conclusion that the PXRD patterns almost coincide with the DSC patterns. At this stage, the naturally arising question is why sVII more strongly promotes ionic conduction than sI. The sVII HPF6 hydrate was proven to exhibit relatively high ionic conductivity due to its favorable cage framework for enhancing proton conduction. In this regard, the transport pathway for charge carriers is likely to govern the ionic conductivity.18 The 4668 cage in sVII is highly symmetrical compared to the 51262 cage in sI (Figure 5, panels a and b). Therefore, the hexagonal and
Figure 4. XRD patterns of HPF6 · nH2O (n ) 4.7, 5.3, 6.0, 6.6, 7.2, and 7.7) at -180 °C (* impurity peak).
TABLE 1: Hydrate Crystal Cell Structures Determined by the PXRD Patterns HPF6 · 4.7H2O HPF6 · 5.3H2O HPF6 · 6.0H2O HPF6 · 6.6H2O HPF6 · 7.2H2O HPF6 · 7.7H2O
type
crystal
space group
a (Å)
sVII sVII sVII sVII sI sVII sI sVII sI
cubic cubic cubic cubic cubic cubic cubic cubic cubic
Im3m Im3m Im3m Im3m Pm3n Im3m Pm3n Im3m Pm3n
7.4472 ( 0.0107 7.5111 ( 0.0113 7.5200 ( 0.0083 7.4941 ( 0.0026 11.6633 ( 0.0086 7.5104 ( 0.0025 11.6850 ( 0.0020 7.7318 ( 0.0122 11.5812 ( 0.0019
tetragonal faces of the 4668 cage in sVII serve as windows, creating a continuous channel patterns by stacking 4668 cages (Figure 5c). At this time, the 4668 cage affords four pairs of hexagonal face windows and three pairs of tetragonal face windows. On the contrary, the large 51262 cage in sI provides a single pair of hexagonal face windows for creating the channel pattern (Figure 5d). The PF6- ions enclthrated in sVII 4668 and sI 51262 cages are likely to be regularly arranged through the channel while the proton occupies one of the polyhedral vertices in the channel pattern. The proton diffusivity is known to largely depend on the number of conduction paths. Accordingly, we might suggest that the HPF6 · 6.0H2O structurally possesses much more diffusion paths than the HPF6 · 7.7H2O, thus inducing HPF6 · 6.0H2O to be more favorable for proton conduction. The channel patterns and interconnecting channel structures in the clathrate hydrate matrix can act as proton diffusion paths in the host lattice. Additionally, the larger cavity diameter of the sVII 4668 cage over that of the sI 51262 cage might lead to a weak interaction between the PF6- ion and proton, contributing to higher ionic conduction through water host lattices. Conclusion Ionic hydrates show thermal and physical properties that differ significantly from those of nonionic hydrates. In particular, the noteworthy feature is that sVII structurally provides a much more effective pathway for transporting charge carriers. The
Letters
J. Phys. Chem. C, Vol. 112, No. 35, 2008 13335 Acknowledgment. This work was supported by the Korea Science and Engineering Foundation (KOSEF) through the National Research Laboratory Program funded by the Ministry of Education, Science and Technology (R0A-2005-000-100740(2008)) and partially supported by the Brain Korea 21 Project. Supporting Information Available: The XRD patterns of HPF6 · 4.2H2O. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes
Figure 5. (a) 4668 cage of sVII in HPF6 · 6.0H2O, (b) 512 and 51262 cages of sI in HPF6 · 7.7H2O, (c) Channel patterns composed of 4668 cage stacking in sVII clathrate hydrate: (left) channels with hexagonal face windows; and (right) channels with tetragonal face windows, and (d) Channel patterns composed of 51262 cage stacking in sI clathrate hydrate. Red-oxygen, green-phosphorus, blue-fluorine. Hydrogen atoms are omitted for clarity.
ionic conductivity of the hydrates can be tuned according to the water content (hydration number), and shows the maximum point at specific hydration number. This implies that cages with more optimal shape and size can be formed, and simultaneously channels can be created in the crystalline hydrate matrix, accompanying a structural transformation. More significantly, the present approaches and outcomes drawn from the HPF6 hydrates might be extended to other ionic hydrates that may better meet the requirements of specific chemical and physical properties.
(1) Park, Y.; Kim, D.-Y.; Lee, J.-w.; Huh, D.-G.; Park, K.-P.; Lee, J.; Lee, H. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 12690–12694. (2) Yeon, S.-H.; Seol, J.; Lee, H. J. Am. Chem. Soc. 2006, 128, 12388– 12389. (3) Kim, D.-Y.; Lee, J.-w.; Seo, Y.-T.; Ripmeester, J. A.; Lee, H. Angew. Chem., Int. Ed. 2005, 44, 7749–7752. (4) Lu, H.; Seo, Y.-T.; Lee, J.-w.; Moudrakovski, I.; Ripmeester, J. A.; Ross Chapman, N.; Coffin, R. B.; Gardner, G.; Pohlman, J. Nature 2007, 445, 303–306. (5) Jeffrey, G. A. Inclusion Compounds; Academic Press: London, 1984; Vol. 1. (6) Platteeuw, J. C.; van der Waals, J. H. Mol. Phys. 1958, 1, 91–96. (7) Choi, S.; Shin, K.; Lee, H. J. Phys. Chem. B 2007, 111, 10224– 10230. (8) Borkowska, Z.; Tymosiak, A.; Opallo, M. J. Electroanal. Chem. 1996, 406, 109–117. (9) Borkowska, Z.; Opallo, M.; Tymosiak-Zielinska, A.; Zoltowski, P. Colloids Surf. A: Physicochem. Eng. Aspects 1998, 134, 67–73. (10) Opallo, M.; Tymosiak-Zielinska, A.; Borkowska, Z. Solid State Ionics 1997, 97, 247–252. (11) Prokopowicz, A.; Opallo, M. Solid State Ionics 2001, 145, 407– 413. (12) Bode, V. H.; Teufer, G. Acta Crystallogr. 1955, 8, 611–614. (13) Davidson, D. W.; Garg, S. K. Can. J. Chem. 1972, 50, 3515–3520. (14) Mootz, D.; Oellers, E.-J.; Wiebcke, M. J. Am. Chem. Soc. 1987, 109, 1200–1202. (15) Huang, T.-H.; Davis, R. A.; Frese, U.; Stimming, U. J. Phys. Chem. 1988, 92, 6874–6876. (16) Cappadonia, M.; Kornyshev, A. A.; Krause, S.; Kuznetsov, A. M.; Stimming, U. J. Chem. Phys. 1994, 101, 7672–7682. (17) The HF might replace the H2O in the host structure due to their similar hydrophilicity and size when the HPF6 aqueous solution is crystallized to form the clathrate hydrate. The HF incorporated in the hydrogen bonded lattice tends to be quiet immobile, resulting in its negligible effect on proton conductivity. (18) Cha, J.-H.; Shin, K.; Choi, S.; Lee, H. J. Phys. Chem. C 2008, 112, 10573–10578.
JP805510G
