- Email: [email protected]
Hydrogen isotope effects in Ti1.0Mn0.9V1.1 and Ti1.0Cr1.5V1.7 alloys
L
Journal of Alloys and Compounds 297 (2000) 253–260
www.elsevier.com / locate / jallcom
Hydrogen isotope effects in Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys a,b b, b b Sung-Wook Cho , Etsuo Akiba *, Yumiko Nakamura , Hirotoshi Enoki a
b
The Japan Steel Works, Ltd., 1 -1, Nikko-cho, Fuchu-shi, Tokyo 183 -8503, Japan National Institute of Materials and Chemical Research, 1 -1 Higashi, Tsukuba, Ibaraki 305 -8565, Japan Received 21 July 1999; accepted 31 August 1999
Abstract The hydrogen isotope effects in Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys have been investigated at 313 and 353 K and at 313 and 338 K, respectively, using protium and deuterium. The crystal structures including the phase abundance and the lattice parameters were determined by the Rietveld method. Using SEM and EDX, the morphology was examined and an elemental analysis was carried out. In both alloys and regardless of temperature, the amounts of deuterium absorption were larger than those of protium. The hydrogen isotope effects of Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 were fairly large in comparison with that of LaNi 5 . The protide of the Ti 1.0 Mn 0.9 V1.1 alloy was more stable than the corresponding deuteride, although at 313 K the equilibrium pressures of deuterium and protium showed slightly opposite behaviour. In the Ti 1.0 Cr 1.5 V1.7 alloy, the deuteride was always more stable than the corresponding protide in the experimental temperature range. The Ti 1.0 Cr 1.5 V1.7 alloy absorbed more deuterium and protium than the Ti 1.0 Mn 0.9 V1.1 alloy. 2000 Elsevier Science S.A. All rights reserved. Keywords: Hydrogen isotope effect; Hydrogen storage alloy; BCC alloy
1. Introduction Metal hydrides which have hydrogen isotope effects can be applicable for the recovery, storage and concentration of deuterium and tritium which form the fuel of a nuclear fusion reactor. Many articles [1–4] have reported the hydrogen isotope effect of metals and / or binary alloys. The majority showed that metal deuterides are less stable than the corresponding metal hydrides. Wiswall and Reilly [5], however, found an inverse behaviour that the replacement of protium by deuterium and tritium results in more stable compounds in some metal hydrides such as VH 2 , NbH 2 (V, Nb)H 2 and LaNi 5 H 6 . Also, titanium has been known to show a small isotope effect. Haag and Shipko [6] found a zero isotope effect in titanium at 5008C. Ueda [7] reported that the deuterides of Ti–V alloys with D/ M51 are more stable than the corresponding protides without regard to the alloy composition by comparing the composition dependence of the activation energy for the diffusion of protium and deuterium. Akiba and Iba reported a new generation alloy, ‘‘Laves phase related BCC solid solution alloy’’, which has a large
*Corresponding author. Tel. / fax: 181-298-54-4541. E-mail address: [email protected] (E. Akiba)
hydrogen capacity of .2 mass% [8]. The Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys, in particular, were found to have a high hydrogenation rate as well as a large hydrogen capacity [8,9]. Thus, it is expected that if these BCC alloys show large hydrogen isotope effects they can be used for the separation of hydrogen isotopes. With this in mind, in this study the hydrogen isotope effects in Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys were investigated at 313 and 353 K, and at 313 and 338 K, respectively, using protium and deuterium.
2. Experimental procedure The alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 were prepared in massive quantities (1 kg lot) in a vacuum highfrequency induction furnace to ensure the reproducibility of the experimental results and were used without heat treatment. Table 1 shows the chemical compositions of the Table 1 Chemical compositions of the alloys analyzed by ICP (at.%) Sample
Ti
Mn(Cr)
V
Chemical formula
Alloy Ti 1.0 Mn 0.9 V1.1 Alloy Ti 1.0 Cr 1.5 V1.7
33.3 24.0
30.7 36.0
36.0 40.0
Ti 1.0 Mn 0.9 V1.1 Ti 1.0 Cr 1.5 V1.7
0925-8388 / 00 / $ – see front matter 2000 Elsevier Science S.A. All rights reserved. PII: S0925-8388( 99 )00585-X
254
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
alloys, which were analyzed using an ICP emission spectrometer. The ingots were crushed mechanically into powder in air. Powders ,500 mm were taken for the experiments. The activation treatment was carried out as follows. In the case of alloy Ti 1.0 Mn 0.9 V1.1 , sample powders of about 10 g were placed in the reactor and evacuated for 1 h by a diffusion pump at 773 K, and then hydrogen gas at a pressure of 5 MPa was introduced into the reactor at room temperature. This was repeated three times. Thereafter, in order to desorb the hydrogen, the reactor was evacuated for 1 h by a diffusion pump at 773 K. In the case of alloy Ti 1.0 Cr 1.5 V1.7 , the reactor was evacuated for 2 h by a rotary pump at 313 K, which was carried out only once. The P–C isotherms of the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 were measured at 313 and 353 K, and at 313 and 338 K, respectively, up to a hydrogen pressure of 5 MPa by a volumetric method. Since the purity and pressure of commercially available deuterium (99.5% min., 3 MPa) are not sufficient, deuterium was absorbed by MmNi 5 to increase the pressure, to purify it, and to recover the used one. The phases, crystal structures, phase abundance, and lattice parameters of the alloys were determined using XRD and the Rietveld analysis method. Sample powders ,30 mm were used for the XRD measurements. We determined the morphology and composition of the phases using a scanning electron microscope (SEM; Hitachi S2350) and an energy dispersive X-ray spectrometer (EDX; Horiba EMAX 2770).
3. Results and discussion Figs. 1 and 2 show micrographs of the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 , respectively, observed by SEM. As can be seen from these figures, these two alloys consist of the matrix (A) and network-shaped precipitates (B). The compositions of each phase obtained by EDX analysis are shown in Tables 2 and 3. From the EDX analysis it was found that the network-shaped precipitate of the alloy Ti 1.0 Mn 0.9 V1.1 , the composition of which is shown in Table 2, consists of a relatively Ti-rich phase and a phase of high Ti concentration. The former Ti-rich phase showed large standard deviations for Ti and V contents. On the other hand, the Mn content was almost the same as in any other part of the sample. Alloy Ti 1.0 Cr 1.5 V1.7 (Table 3) was found to be composed of the matrix and a high Ti concentration phase, as shown in the SEM micrograph. While the Cr content of the matrix was nearly constant, the Ti and V contents showed large standard deviations, which is different from the matrix of the alloy Ti 1.0 Mn 0.9 V1.1 with nearly constant contents of Ti, Mn, and V. The reason why we found differences between the matrix homogeneities is not clear. The reason for the appearance of the Ti-concentrated phase can be deduced from the phase diagram of the Ti–V system [10]; the Ti-concentrated phase is considered to be the a-Ti phase with HCP structure, which is supported by the X-ray Rietveld analysis described below. Fig. 3 shows the results of the EDX mapping for each alloy. RIETAN 97 [11] was used for the Rietveld analysis.
Fig. 1. SEM micrograph of alloy Ti 1.0 Mn 0.9 V1.1 .
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
255
Fig. 2. SEM micrograph of alloy Ti 1.0 Cr 1.5 V1.7 .
Alloy Ti 1.0 Mn 0.9 V1.1 was assumed to be composed of three phases, as derived from the XRD qualitative analysis: the matrix with BCC structure, a phase with FCC structure, and a-Ti with HCP structure. Fig. 4 shows the result of the Rietveld analysis for this alloy. In this figure, the diffraction pattern calculated from the model, which is represented by the full curve, agrees well with the measured data represented by the plus sign. The phase abundances of the matrix, the phase with FCC structure, and a-Ti calculated from the parameters refined by the X-ray Rietveld analysis were 57.1, 39.0 and 3.9%, respectively. The refined parameters are summarized in Table 4. Alloy Ti 1.0 Cr 1.5 V1.7 was assumed from the XRD qualitative analysis to consist of two phases: the matrix with BCC structure and a-Ti with HCP structure. The result of the X-ray Rietveld analysis for this alloy is shown in Fig. 5. As shown in this figure, the diffraction pattern calculated from the model is in good agreement with that measured. The phase abundances of the matrix and the a-Ti calculated from the parameters refined by the Rietveld analysis were 96.1 and 3.9%, respectively. The refined parameters for this alloy are also summarized in Table 4. Figs. 6 and 7 show the desorption isotherms of deuterium and protium for the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 , respectively. It was found that both alloys, except for the data for Ti 1.0 Mn 0.9 V1.1 measured at 313 K, show marked hydrogen isotope effects. As a reference, data for LaNi 5 measured at 313 K are shown in Fig. 8. In the latter material the hydrogen isotope effect was hardly observed. Also, from Figs. 6 and 7, differences in the
absorbed amounts were found between deuterium and protium. In both alloys and regardless of temperature, the absorbed amount of deuterium was larger than that of protium. Alloy Ti 1.0 Cr 1.5 V1.7 absorbed more deuterium and protium than alloy Ti 1.0 Mn 0.9 V1.1 . This is because the content of the BCC phase of alloy Ti 1.0 Cr 1.5 V1.7 was larger than that of alloy Ti 1.0 Mn 0.9 V1.1 , which is shown in Table 4. Fig. 9 shows a plot of ln P vs. 1 /T obtained in this study, where the respective plateau pressure was obtained from the middle point between the two points indicated by the arrows in Figs. 6 and 7, assuming a straight line between the two points. Data for zone-refined vanadium obtained by Wiswall and Reilly [5] are also shown by the dotted line in this figure. Results for LaNi 5 obtained by Biris et al. [4] as well as results obtained in this study are also shown. Alloy Ti 1.0 Cr 1.5 V1.7 exhibited inverse behaviour without regard to the measured temperature, which was different from those for alloy Ti 1.0 Mn 0.9 V1.1 ; the plateau pressures of deuterium were lower than those of protium in alloy Ti 1.0 Cr 1.5 V1.7 and vise versa in alloy Ti 1.0 Mn 0.9 V1.1 . This difference can be explained by the differences in the formation enthalpies and entropies between protide and deuteride in each alloy. Fig. 10 shows a schematic drawing of the van’t Hoff equation. The formation enthalpies and entropies of the hydrides are correlated with the slopes and intercepts, respectively, on the y-axis of the van’t Hoff plot. Also, the temperature of the crossing point (T cross ) can be expressed as T cross 5 (DH Do 2 DH Ho ) /(DS Do 2 DS Ho )
(1)
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
256
Table 2 Compositions of each phase in the alloy Ti 1.0 Mn 0.9 V1.1 obtained by EDX (at.%)
Table 3 Compositions of each phase in alloy Ti 1.0 Cr 1.5 V1.7 obtained by EDX (at.%)
Phase
Area
Ti
Mn
V
Phase
Area
Ti
Cr
V
Total
1 2 3 4 5 Ave SD
33.4 33.4 34.6 34.6 34.8 34.2 0.7
30.3 29.7 29.7 28.5 30.1 29.7 0.7
36.3 36.9 35.7 36.9 35.1 36.2 0.8
Total
1 2 3 4 5 Ave SD
24.3 24.7 25.2 25.2 26.5 25.2 0.8
35.7 35.9 35.0 35.8 35.5 35.6 0.4
40.0 39.4 39.8 39.0 38.0 39.2 0.8
Matrix
1 2 3 4 5 6 7 Ave SD
27.8 28.0 28.5 28.7 28.9 29.0 31.2 28.9 1.1
30.2 30.7 30.1 29.7 30.5 29.6 29.8 30.1 0.4
42.0 41.3 41.4 41.6 40.6 41.4 39.0 41.0 1.0
Matrix
1 2 3 4 5 6 Ave SD
16.6 19.4 20.1 21.3 22.1 26.2 21.0 3.2
37.7 37.7 36.7 36.9 38.0 37.5 37.4 0.5
45.7 42.9 43.2 41.8 39.9 36.3 41.6 3.2
Network-shaped precipitate (Ti-rich phase)
1 2 3 4 5 6 7 8 9 10 11 Ave SD
34.1 36.2 38.0 43.2 45.4 45.7 53.7 55.0 55.3 56.8 59.5 47.5 9.0
28.1 30.1 30.6 32.8 29.2 27.1 28.8 28.6 29.9 30.3 25.5 29.2 1.9
37.8 33.7 31.4 24.0 25.4 27.2 17.5 16.4 14.8 12.9 15.0 23.3 8.6
Network-shaped precipitate (Ti-concentrated phase)
1 2 3 4 5 Ave SD
67.7 78.1 82.3 84.2 86.9 79.8 7.5
15.8 10.1 7.2 7.5 5.5 9.2 4.0
16.5 11.8 10.5 8.3 7.6 10.9 3.5
12 13 14 Ave SD
82.9 83.1 88.4 84.8 3.1
10.8 9.7 4.2 8.2 3.5
6.3 7.2 7.4 7.0 0.6
(Ti-concentrated phase)
where DH oD , DH oH , DS oD , and DS oH are the standard formation enthalpies and entropies of the deuteride and the corresponding protide. The standard formation entropy of the hydride is mainly determined by the loss of the absolute entropy of hydrogen gas (S Do 2 5 144.8, S Ho 2 5 130.5 J / mol H 2 K at 298 K). This means that DS oD and DS oH can be assumed to be constants and do not depend on the nature of each alloy [12]. Therefore, Eq. (1) can be rewritten as T cross 5 (DH Do 2 DH Ho ) /C
(2) o D
o H
where C is a constant (DS 2 DS ). Here, it should be pointed out that the C value is negative because DS Do is more negative. Thus, only when the deuteride’s formation enthalpy is more negative than that of the corresponding protide (DH Do 2 DH Ho , 0) can inverse behaviour occur. In such a case, when the value of (DH Do 2 DH Ho ) becomes more negative, the temperature of the crossing point increases. Fig. 10 shows such a case, where the absolute value of the difference in standard formation enthalpies
between the deuteride and protide shown by dotted lines is larger than that shown by the solid lines. Therefore, T cross of the dotted lines (T 1 ) is higher than that of the solid lines (T 2 ). Consequently, it can be stated that if the difference in standard formation entropies between isotope hydrides is constant, T cross depends only on the difference in DH o . The ‘‘inverse isotope effect’’ [5] can be explained in the same way. Table 5 shows the standard formation enthalpies and entropies of hydrides obtained in this work together with those reported for V [5] and LaNi 5 [4]. The values of DS o in Table 5 show dispersed behaviour; in particular, the absolute values of DS Do measured for the BCC alloys and V are larger than the absolute entropy of deuterium (144.8 J / mol H 2 K at 298 K). There are several reasons for this. One is the difficulty in determining the equilibrium pressures from the the P–C isotherms obtained. It is not easy to define them from sloping isotherms. Figs. 6 and 7 are typical examples. Also, the errors in measuring temperature and pressure are not negligible. Another difficulty is the difference in standard formation entropies between the monohydride and dihydride in BCC alloys. In addition, deuterium in V occupies different interstitial sites than protium. Because the DS oD values obtained for Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 differ from each other, the order of the magnitude of (DH Do 2 DH Ho ) of these alloys does not correlate with that of T cross . However, it is clear that DH o and DS o can describe the hydrogen isotope effect of metal–hydrogen systems. Fig. 11 shows the framed part of Fig. 9. It should be
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
Fig. 3. EDX mapping for the alloys.
Fig. 4. Rietveld refinement for alloy Ti 1.0 Mn 0.9 V1.1 .
257
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
258
Table 4 Parameters for the alloys refined by X-ray Rietveld analysis Sample
Phase
Lattice
RI
˚ a (A)
˚ c (A)
Phase abundance (%)
Ti 1.0 Mn 0.9 V1.1 R wp 5 19.0, S 5 2.47
Matrix Ti-rich Ti-conc.
BCC FCC HCP
4.38 13.9 17.6
3.0183 (1) 11.293 (1) 2.9669 (9)
– – 4.784 (2)
57.1 39.0 3.9
Ti 1.0 Cr 1.5 V1.7 R wp 5 11.0, S 5 1.97
Matrix Ti-conc.
BCC HCP
1.28 10.2
3.0212 (9) 2.9602 (7)
– 4.789 (1)
96.1 3.9
Fig. 5. Rietveld refinement for alloy Ti 1.0 Cr 1.5 V1.7 .
Fig. 6. Desorption isotherms of deuterium and protium for alloy Ti 1.0 Mn 0.9 V1.1 .
Fig. 7. Desorption isotherms of deuterium and protium for alloy Ti 1.0 Cr 1.5 V1.7 .
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
259
Fig. 10. Schematic drawing of the van’t Hoff equation. Table 5 Standard formation enthalpies and entropies of hydrides Sample
H(D)
DH o (kJ / mol H 2 )
DS o (J / mol H 2 K)
Ti 1.0 Mn 0.9 V1.1
H D H D H D H D
242 263 234 252 240 250 231 235
2120 2187 2119 2168 2141 2164 2109 2123
Fig. 8. Desorption isotherms of deuterium and protium for LaNi 5 . Ti 1.0 Cr 1.5 V1.7
noted that the hydrogen isotope effect of LaNi 5 is very small and those of the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 are fairly large. In the case of LaNi 5 , the differences in DH o and DS o between protide and deuteride are small, which makes the hydrogen isotope effect of LaNi 5 small.
Fig. 9. Van’t Hoff plots of the studied alloys: (- - -) V–H [5]; (? ? ?) V–D [5]; ( ) LaNi 5 –H, this work; ( ) LaNi 5 –D, this work; (- - -) LaNi 5 –H [4]; (? ? ?) LaNi 5 –D [4].
Va LaNi 5 b a b
Ref. [5]. Ref. [4].
Fig. 11. Enlarged representation of the framed part shown in Fig. 9. (d) Ti 1.0 Cr 1.5 V1.7 –H; (j) Ti 1.0 Cr 1.5 V1.7 –D; (- - -) LaNi 5 –H [4]; (? ? ?) LaNi 5 –D [4]; ( ) LaNi 5 –H, this work; ( ) LaNi 5 –D, this work.
260
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
4. Conclusions
Acknowledgements
The hydrogen isotope effects in Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys have been investigated at 313 and 353 K, and at 313 and 338 K, respectively, using protium and deuterium. The conclusions of this study are summarized as follows.
The authors would like to thank Mr. Y. Ishido of the National Institute of Materials and Chemical Research for his help in measuring the P–C isotherms of LaNi 5 .
References 1. In both alloys and regardless of temperature, the absorbed amounts of deuterium are larger than those of protium. 2. While the hydrogen isotope effect of LaNi 5 is very small, those of the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 are fairly large. 3. The plateau pressure of deuterium is higher than that of protium in the Ti 1.0 Mn 0.9 V1.1 alloy, although at 313 K it shows the opposite behaviour. In the Ti 1.0 Cr 1.5 V1.7 alloy the plateau pressure of deuterium is always lower than that of protium in the experimental temperature range. 4. The inverse isotope effect can be explained by the differences in standard formation enthalpies and entropies between isotope hydrides.
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]
W.L. Korst, J.C. Warf, Inorg. Chem. 5 (1966) 1719. A.D. McQuillan, A.D. Wallbank, J. Chem. Phys. 51 (1969) 1026. F. Ricca, T. Giorgi, J. Phys. Chem. 74 (1970) 143. A. Biris, R.V. Bucur, P. Ghete, E. Indrea, D. Lupu, J. Less-Common Met. 49 (1976) 477. R.H. Wiswall, J.J. Reilly, Inorg. Chem. 11 (1972) 1691. R.M. Haag, F.J. Shipko, J. Am. Chem. Soc. 78 (1956) 5155. T. Ueda, Doc. Diss., Faculty of Science, Osaka University, 1995. E. Akiba, H. Iba, Intermetallics 6 (1998) 461. H. Iba, E. Akiba, J. Alloys Comp. 253 / 254 (1997) 21. H. Okamoto, J. Phase Equilibria 14 (1993) 266. F. Izumi, in: R.A. Young (Ed.), The Rietveld Method, Oxford University Press, Oxford, 1993, Chapt. 13. Y. Osumi, H. Suzuki, A. Kato, K. Oguro, M. Nakane, Nihon Kagaku Kaishi (1981) 1493.
Journal of Alloys and Compounds 297 (2000) 253–260
www.elsevier.com / locate / jallcom
Hydrogen isotope effects in Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys a,b b, b b Sung-Wook Cho , Etsuo Akiba *, Yumiko Nakamura , Hirotoshi Enoki a
b
The Japan Steel Works, Ltd., 1 -1, Nikko-cho, Fuchu-shi, Tokyo 183 -8503, Japan National Institute of Materials and Chemical Research, 1 -1 Higashi, Tsukuba, Ibaraki 305 -8565, Japan Received 21 July 1999; accepted 31 August 1999
Abstract The hydrogen isotope effects in Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys have been investigated at 313 and 353 K and at 313 and 338 K, respectively, using protium and deuterium. The crystal structures including the phase abundance and the lattice parameters were determined by the Rietveld method. Using SEM and EDX, the morphology was examined and an elemental analysis was carried out. In both alloys and regardless of temperature, the amounts of deuterium absorption were larger than those of protium. The hydrogen isotope effects of Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 were fairly large in comparison with that of LaNi 5 . The protide of the Ti 1.0 Mn 0.9 V1.1 alloy was more stable than the corresponding deuteride, although at 313 K the equilibrium pressures of deuterium and protium showed slightly opposite behaviour. In the Ti 1.0 Cr 1.5 V1.7 alloy, the deuteride was always more stable than the corresponding protide in the experimental temperature range. The Ti 1.0 Cr 1.5 V1.7 alloy absorbed more deuterium and protium than the Ti 1.0 Mn 0.9 V1.1 alloy. 2000 Elsevier Science S.A. All rights reserved. Keywords: Hydrogen isotope effect; Hydrogen storage alloy; BCC alloy
1. Introduction Metal hydrides which have hydrogen isotope effects can be applicable for the recovery, storage and concentration of deuterium and tritium which form the fuel of a nuclear fusion reactor. Many articles [1–4] have reported the hydrogen isotope effect of metals and / or binary alloys. The majority showed that metal deuterides are less stable than the corresponding metal hydrides. Wiswall and Reilly [5], however, found an inverse behaviour that the replacement of protium by deuterium and tritium results in more stable compounds in some metal hydrides such as VH 2 , NbH 2 (V, Nb)H 2 and LaNi 5 H 6 . Also, titanium has been known to show a small isotope effect. Haag and Shipko [6] found a zero isotope effect in titanium at 5008C. Ueda [7] reported that the deuterides of Ti–V alloys with D/ M51 are more stable than the corresponding protides without regard to the alloy composition by comparing the composition dependence of the activation energy for the diffusion of protium and deuterium. Akiba and Iba reported a new generation alloy, ‘‘Laves phase related BCC solid solution alloy’’, which has a large
*Corresponding author. Tel. / fax: 181-298-54-4541. E-mail address: [email protected] (E. Akiba)
hydrogen capacity of .2 mass% [8]. The Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys, in particular, were found to have a high hydrogenation rate as well as a large hydrogen capacity [8,9]. Thus, it is expected that if these BCC alloys show large hydrogen isotope effects they can be used for the separation of hydrogen isotopes. With this in mind, in this study the hydrogen isotope effects in Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys were investigated at 313 and 353 K, and at 313 and 338 K, respectively, using protium and deuterium.
2. Experimental procedure The alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 were prepared in massive quantities (1 kg lot) in a vacuum highfrequency induction furnace to ensure the reproducibility of the experimental results and were used without heat treatment. Table 1 shows the chemical compositions of the Table 1 Chemical compositions of the alloys analyzed by ICP (at.%) Sample
Ti
Mn(Cr)
V
Chemical formula
Alloy Ti 1.0 Mn 0.9 V1.1 Alloy Ti 1.0 Cr 1.5 V1.7
33.3 24.0
30.7 36.0
36.0 40.0
Ti 1.0 Mn 0.9 V1.1 Ti 1.0 Cr 1.5 V1.7
0925-8388 / 00 / $ – see front matter 2000 Elsevier Science S.A. All rights reserved. PII: S0925-8388( 99 )00585-X
254
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
alloys, which were analyzed using an ICP emission spectrometer. The ingots were crushed mechanically into powder in air. Powders ,500 mm were taken for the experiments. The activation treatment was carried out as follows. In the case of alloy Ti 1.0 Mn 0.9 V1.1 , sample powders of about 10 g were placed in the reactor and evacuated for 1 h by a diffusion pump at 773 K, and then hydrogen gas at a pressure of 5 MPa was introduced into the reactor at room temperature. This was repeated three times. Thereafter, in order to desorb the hydrogen, the reactor was evacuated for 1 h by a diffusion pump at 773 K. In the case of alloy Ti 1.0 Cr 1.5 V1.7 , the reactor was evacuated for 2 h by a rotary pump at 313 K, which was carried out only once. The P–C isotherms of the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 were measured at 313 and 353 K, and at 313 and 338 K, respectively, up to a hydrogen pressure of 5 MPa by a volumetric method. Since the purity and pressure of commercially available deuterium (99.5% min., 3 MPa) are not sufficient, deuterium was absorbed by MmNi 5 to increase the pressure, to purify it, and to recover the used one. The phases, crystal structures, phase abundance, and lattice parameters of the alloys were determined using XRD and the Rietveld analysis method. Sample powders ,30 mm were used for the XRD measurements. We determined the morphology and composition of the phases using a scanning electron microscope (SEM; Hitachi S2350) and an energy dispersive X-ray spectrometer (EDX; Horiba EMAX 2770).
3. Results and discussion Figs. 1 and 2 show micrographs of the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 , respectively, observed by SEM. As can be seen from these figures, these two alloys consist of the matrix (A) and network-shaped precipitates (B). The compositions of each phase obtained by EDX analysis are shown in Tables 2 and 3. From the EDX analysis it was found that the network-shaped precipitate of the alloy Ti 1.0 Mn 0.9 V1.1 , the composition of which is shown in Table 2, consists of a relatively Ti-rich phase and a phase of high Ti concentration. The former Ti-rich phase showed large standard deviations for Ti and V contents. On the other hand, the Mn content was almost the same as in any other part of the sample. Alloy Ti 1.0 Cr 1.5 V1.7 (Table 3) was found to be composed of the matrix and a high Ti concentration phase, as shown in the SEM micrograph. While the Cr content of the matrix was nearly constant, the Ti and V contents showed large standard deviations, which is different from the matrix of the alloy Ti 1.0 Mn 0.9 V1.1 with nearly constant contents of Ti, Mn, and V. The reason why we found differences between the matrix homogeneities is not clear. The reason for the appearance of the Ti-concentrated phase can be deduced from the phase diagram of the Ti–V system [10]; the Ti-concentrated phase is considered to be the a-Ti phase with HCP structure, which is supported by the X-ray Rietveld analysis described below. Fig. 3 shows the results of the EDX mapping for each alloy. RIETAN 97 [11] was used for the Rietveld analysis.
Fig. 1. SEM micrograph of alloy Ti 1.0 Mn 0.9 V1.1 .
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
255
Fig. 2. SEM micrograph of alloy Ti 1.0 Cr 1.5 V1.7 .
Alloy Ti 1.0 Mn 0.9 V1.1 was assumed to be composed of three phases, as derived from the XRD qualitative analysis: the matrix with BCC structure, a phase with FCC structure, and a-Ti with HCP structure. Fig. 4 shows the result of the Rietveld analysis for this alloy. In this figure, the diffraction pattern calculated from the model, which is represented by the full curve, agrees well with the measured data represented by the plus sign. The phase abundances of the matrix, the phase with FCC structure, and a-Ti calculated from the parameters refined by the X-ray Rietveld analysis were 57.1, 39.0 and 3.9%, respectively. The refined parameters are summarized in Table 4. Alloy Ti 1.0 Cr 1.5 V1.7 was assumed from the XRD qualitative analysis to consist of two phases: the matrix with BCC structure and a-Ti with HCP structure. The result of the X-ray Rietveld analysis for this alloy is shown in Fig. 5. As shown in this figure, the diffraction pattern calculated from the model is in good agreement with that measured. The phase abundances of the matrix and the a-Ti calculated from the parameters refined by the Rietveld analysis were 96.1 and 3.9%, respectively. The refined parameters for this alloy are also summarized in Table 4. Figs. 6 and 7 show the desorption isotherms of deuterium and protium for the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 , respectively. It was found that both alloys, except for the data for Ti 1.0 Mn 0.9 V1.1 measured at 313 K, show marked hydrogen isotope effects. As a reference, data for LaNi 5 measured at 313 K are shown in Fig. 8. In the latter material the hydrogen isotope effect was hardly observed. Also, from Figs. 6 and 7, differences in the
absorbed amounts were found between deuterium and protium. In both alloys and regardless of temperature, the absorbed amount of deuterium was larger than that of protium. Alloy Ti 1.0 Cr 1.5 V1.7 absorbed more deuterium and protium than alloy Ti 1.0 Mn 0.9 V1.1 . This is because the content of the BCC phase of alloy Ti 1.0 Cr 1.5 V1.7 was larger than that of alloy Ti 1.0 Mn 0.9 V1.1 , which is shown in Table 4. Fig. 9 shows a plot of ln P vs. 1 /T obtained in this study, where the respective plateau pressure was obtained from the middle point between the two points indicated by the arrows in Figs. 6 and 7, assuming a straight line between the two points. Data for zone-refined vanadium obtained by Wiswall and Reilly [5] are also shown by the dotted line in this figure. Results for LaNi 5 obtained by Biris et al. [4] as well as results obtained in this study are also shown. Alloy Ti 1.0 Cr 1.5 V1.7 exhibited inverse behaviour without regard to the measured temperature, which was different from those for alloy Ti 1.0 Mn 0.9 V1.1 ; the plateau pressures of deuterium were lower than those of protium in alloy Ti 1.0 Cr 1.5 V1.7 and vise versa in alloy Ti 1.0 Mn 0.9 V1.1 . This difference can be explained by the differences in the formation enthalpies and entropies between protide and deuteride in each alloy. Fig. 10 shows a schematic drawing of the van’t Hoff equation. The formation enthalpies and entropies of the hydrides are correlated with the slopes and intercepts, respectively, on the y-axis of the van’t Hoff plot. Also, the temperature of the crossing point (T cross ) can be expressed as T cross 5 (DH Do 2 DH Ho ) /(DS Do 2 DS Ho )
(1)
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
256
Table 2 Compositions of each phase in the alloy Ti 1.0 Mn 0.9 V1.1 obtained by EDX (at.%)
Table 3 Compositions of each phase in alloy Ti 1.0 Cr 1.5 V1.7 obtained by EDX (at.%)
Phase
Area
Ti
Mn
V
Phase
Area
Ti
Cr
V
Total
1 2 3 4 5 Ave SD
33.4 33.4 34.6 34.6 34.8 34.2 0.7
30.3 29.7 29.7 28.5 30.1 29.7 0.7
36.3 36.9 35.7 36.9 35.1 36.2 0.8
Total
1 2 3 4 5 Ave SD
24.3 24.7 25.2 25.2 26.5 25.2 0.8
35.7 35.9 35.0 35.8 35.5 35.6 0.4
40.0 39.4 39.8 39.0 38.0 39.2 0.8
Matrix
1 2 3 4 5 6 7 Ave SD
27.8 28.0 28.5 28.7 28.9 29.0 31.2 28.9 1.1
30.2 30.7 30.1 29.7 30.5 29.6 29.8 30.1 0.4
42.0 41.3 41.4 41.6 40.6 41.4 39.0 41.0 1.0
Matrix
1 2 3 4 5 6 Ave SD
16.6 19.4 20.1 21.3 22.1 26.2 21.0 3.2
37.7 37.7 36.7 36.9 38.0 37.5 37.4 0.5
45.7 42.9 43.2 41.8 39.9 36.3 41.6 3.2
Network-shaped precipitate (Ti-rich phase)
1 2 3 4 5 6 7 8 9 10 11 Ave SD
34.1 36.2 38.0 43.2 45.4 45.7 53.7 55.0 55.3 56.8 59.5 47.5 9.0
28.1 30.1 30.6 32.8 29.2 27.1 28.8 28.6 29.9 30.3 25.5 29.2 1.9
37.8 33.7 31.4 24.0 25.4 27.2 17.5 16.4 14.8 12.9 15.0 23.3 8.6
Network-shaped precipitate (Ti-concentrated phase)
1 2 3 4 5 Ave SD
67.7 78.1 82.3 84.2 86.9 79.8 7.5
15.8 10.1 7.2 7.5 5.5 9.2 4.0
16.5 11.8 10.5 8.3 7.6 10.9 3.5
12 13 14 Ave SD
82.9 83.1 88.4 84.8 3.1
10.8 9.7 4.2 8.2 3.5
6.3 7.2 7.4 7.0 0.6
(Ti-concentrated phase)
where DH oD , DH oH , DS oD , and DS oH are the standard formation enthalpies and entropies of the deuteride and the corresponding protide. The standard formation entropy of the hydride is mainly determined by the loss of the absolute entropy of hydrogen gas (S Do 2 5 144.8, S Ho 2 5 130.5 J / mol H 2 K at 298 K). This means that DS oD and DS oH can be assumed to be constants and do not depend on the nature of each alloy [12]. Therefore, Eq. (1) can be rewritten as T cross 5 (DH Do 2 DH Ho ) /C
(2) o D
o H
where C is a constant (DS 2 DS ). Here, it should be pointed out that the C value is negative because DS Do is more negative. Thus, only when the deuteride’s formation enthalpy is more negative than that of the corresponding protide (DH Do 2 DH Ho , 0) can inverse behaviour occur. In such a case, when the value of (DH Do 2 DH Ho ) becomes more negative, the temperature of the crossing point increases. Fig. 10 shows such a case, where the absolute value of the difference in standard formation enthalpies
between the deuteride and protide shown by dotted lines is larger than that shown by the solid lines. Therefore, T cross of the dotted lines (T 1 ) is higher than that of the solid lines (T 2 ). Consequently, it can be stated that if the difference in standard formation entropies between isotope hydrides is constant, T cross depends only on the difference in DH o . The ‘‘inverse isotope effect’’ [5] can be explained in the same way. Table 5 shows the standard formation enthalpies and entropies of hydrides obtained in this work together with those reported for V [5] and LaNi 5 [4]. The values of DS o in Table 5 show dispersed behaviour; in particular, the absolute values of DS Do measured for the BCC alloys and V are larger than the absolute entropy of deuterium (144.8 J / mol H 2 K at 298 K). There are several reasons for this. One is the difficulty in determining the equilibrium pressures from the the P–C isotherms obtained. It is not easy to define them from sloping isotherms. Figs. 6 and 7 are typical examples. Also, the errors in measuring temperature and pressure are not negligible. Another difficulty is the difference in standard formation entropies between the monohydride and dihydride in BCC alloys. In addition, deuterium in V occupies different interstitial sites than protium. Because the DS oD values obtained for Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 differ from each other, the order of the magnitude of (DH Do 2 DH Ho ) of these alloys does not correlate with that of T cross . However, it is clear that DH o and DS o can describe the hydrogen isotope effect of metal–hydrogen systems. Fig. 11 shows the framed part of Fig. 9. It should be
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
Fig. 3. EDX mapping for the alloys.
Fig. 4. Rietveld refinement for alloy Ti 1.0 Mn 0.9 V1.1 .
257
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
258
Table 4 Parameters for the alloys refined by X-ray Rietveld analysis Sample
Phase
Lattice
RI
˚ a (A)
˚ c (A)
Phase abundance (%)
Ti 1.0 Mn 0.9 V1.1 R wp 5 19.0, S 5 2.47
Matrix Ti-rich Ti-conc.
BCC FCC HCP
4.38 13.9 17.6
3.0183 (1) 11.293 (1) 2.9669 (9)
– – 4.784 (2)
57.1 39.0 3.9
Ti 1.0 Cr 1.5 V1.7 R wp 5 11.0, S 5 1.97
Matrix Ti-conc.
BCC HCP
1.28 10.2
3.0212 (9) 2.9602 (7)
– 4.789 (1)
96.1 3.9
Fig. 5. Rietveld refinement for alloy Ti 1.0 Cr 1.5 V1.7 .
Fig. 6. Desorption isotherms of deuterium and protium for alloy Ti 1.0 Mn 0.9 V1.1 .
Fig. 7. Desorption isotherms of deuterium and protium for alloy Ti 1.0 Cr 1.5 V1.7 .
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
259
Fig. 10. Schematic drawing of the van’t Hoff equation. Table 5 Standard formation enthalpies and entropies of hydrides Sample
H(D)
DH o (kJ / mol H 2 )
DS o (J / mol H 2 K)
Ti 1.0 Mn 0.9 V1.1
H D H D H D H D
242 263 234 252 240 250 231 235
2120 2187 2119 2168 2141 2164 2109 2123
Fig. 8. Desorption isotherms of deuterium and protium for LaNi 5 . Ti 1.0 Cr 1.5 V1.7
noted that the hydrogen isotope effect of LaNi 5 is very small and those of the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 are fairly large. In the case of LaNi 5 , the differences in DH o and DS o between protide and deuteride are small, which makes the hydrogen isotope effect of LaNi 5 small.
Fig. 9. Van’t Hoff plots of the studied alloys: (- - -) V–H [5]; (? ? ?) V–D [5]; ( ) LaNi 5 –H, this work; ( ) LaNi 5 –D, this work; (- - -) LaNi 5 –H [4]; (? ? ?) LaNi 5 –D [4].
Va LaNi 5 b a b
Ref. [5]. Ref. [4].
Fig. 11. Enlarged representation of the framed part shown in Fig. 9. (d) Ti 1.0 Cr 1.5 V1.7 –H; (j) Ti 1.0 Cr 1.5 V1.7 –D; (- - -) LaNi 5 –H [4]; (? ? ?) LaNi 5 –D [4]; ( ) LaNi 5 –H, this work; ( ) LaNi 5 –D, this work.
260
S.-W. Cho et al. / Journal of Alloys and Compounds 297 (2000) 253 – 260
4. Conclusions
Acknowledgements
The hydrogen isotope effects in Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 alloys have been investigated at 313 and 353 K, and at 313 and 338 K, respectively, using protium and deuterium. The conclusions of this study are summarized as follows.
The authors would like to thank Mr. Y. Ishido of the National Institute of Materials and Chemical Research for his help in measuring the P–C isotherms of LaNi 5 .
References 1. In both alloys and regardless of temperature, the absorbed amounts of deuterium are larger than those of protium. 2. While the hydrogen isotope effect of LaNi 5 is very small, those of the alloys Ti 1.0 Mn 0.9 V1.1 and Ti 1.0 Cr 1.5 V1.7 are fairly large. 3. The plateau pressure of deuterium is higher than that of protium in the Ti 1.0 Mn 0.9 V1.1 alloy, although at 313 K it shows the opposite behaviour. In the Ti 1.0 Cr 1.5 V1.7 alloy the plateau pressure of deuterium is always lower than that of protium in the experimental temperature range. 4. The inverse isotope effect can be explained by the differences in standard formation enthalpies and entropies between isotope hydrides.
[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12]
W.L. Korst, J.C. Warf, Inorg. Chem. 5 (1966) 1719. A.D. McQuillan, A.D. Wallbank, J. Chem. Phys. 51 (1969) 1026. F. Ricca, T. Giorgi, J. Phys. Chem. 74 (1970) 143. A. Biris, R.V. Bucur, P. Ghete, E. Indrea, D. Lupu, J. Less-Common Met. 49 (1976) 477. R.H. Wiswall, J.J. Reilly, Inorg. Chem. 11 (1972) 1691. R.M. Haag, F.J. Shipko, J. Am. Chem. Soc. 78 (1956) 5155. T. Ueda, Doc. Diss., Faculty of Science, Osaka University, 1995. E. Akiba, H. Iba, Intermetallics 6 (1998) 461. H. Iba, E. Akiba, J. Alloys Comp. 253 / 254 (1997) 21. H. Okamoto, J. Phase Equilibria 14 (1993) 266. F. Izumi, in: R.A. Young (Ed.), The Rietveld Method, Oxford University Press, Oxford, 1993, Chapt. 13. Y. Osumi, H. Suzuki, A. Kato, K. Oguro, M. Nakane, Nihon Kagaku Kaishi (1981) 1493.