J. Phys. Chem. B 2010, 114, 6117–6125
6117
Thermodynamics of Hydrogen in fcc Pd-Au Alloys S. Luo, D. Wang, and Ted B. Flanagan* Material Science Program and Department of Chemistry, UniVersity of Vermont, Burlington Vermont 05405 ReceiVed: January 28, 2010; ReVised Manuscript ReceiVed: April 1, 2010
Hydrogen isotherms have been measured for a series of solid solution Pd-Au alloys in the temperature range from 393 to 523 K. Standard partial thermodynamic parameters at infinite dilution of H, ∆H°, H and ∆S°, H have been determined from these equilibrium data; both standard values for H2 absorption become more negative with increase of atom fraction Au, XAu. An interesting result is that the dilute phase isotherms at 423 and 523 K are all very similar for alloys with XAu ) 0.15 to about 0.30 although their ∆H°H and ∆S°H values differ. This is due to a compensating effect of the two thermodynamic parameters leading to (∂∆GH/∂r) ) RT(∂ ln p1/2/∂r) ≈ constant for the alloys from XAu ≈ 0.15 to 0.30 at low r where r ) H-to-metal atom ratio. Calorimetric enthalpies and isotherms at 303 K have been determined for a series of Pd-Au alloys over a range of H contents including, for some of the low Au content alloys, the plateau regions. These calorimetric data are the most complete reported for the Pd-Au-H system. Introduction There have been many investigations of the solubility and thermodynamics of H2 in Pd alloys and some general patterns of behavior have emerged.1 Solid solution Pd alloys with expanded lattices compared to Pd such as Pd-Ag dissolve H2 more exothermically than Pd and, consequently, their dilute phase solubilities are greater at a given pH2 in these alloys than in Pd. Contracted alloys, for example, Pd-Ni, Pd-Rh, dissolve H2 more endothermically than Pd and H2 is less soluble in the dilute phase in these at a given pH2 than in Pd. Since Pd-Au alloys have expanded lattices2 compared to Pd, the enthalpies of solution should be more exothermic than in Pd and they should dissolve more H at a given low pH2 than Pd. Recent density functional theory calculations that have been carried out on Pd-Au alloy-H systems3 indicate that there should be a maximum H2 solubility at atom fraction Au, XAu ≈ 0.20 at temperatures from 456 to 1095 K. The H2 pressures for these solubility predictions were not given and it will be shown below that the solubilities in the various Pd-Au alloys depend on pH2. Hydrogen isotherms have been previously measured electrochemically for Pd-Au alloys over the range 273-323 K from which thermodynamic parameters were derived.2,4 A phase diagram at 298 K was determined from the lattice parameters of the coexisting dilute and hydride phases;2 the disappearance of the coexisting phases at XAu ≈ 0.17 indicates the critical composition for this temperature. Shamsuddin and Kleppa5 determined the enthalpies of H2 absorption calorimetrically at small H contents at 555 and 700 K for the Pd0.90Au0.10, Pd0.75Au0.25, and Pd0.60Au0.40 alloys. The enthalpies increased in exothermicity with Au content and the partial entropies also became more negative and these trends agree with those found by Allard et al.4 In the present work, enthalpies of reaction of H2 as a function of r, where r ) H-tometal atom ratio, will be determined calorimetrically (303 K) for Pd-Au alloys as in the work of Shamsuddin and Kleppa,5 but in contrast to their high temperature investigation the present calorimetry will be carried out at a lower temperature, 303 K, * To whom correspondence should be addressed.
where greater H contents can be obtained and where hydride phases form in some of the alloys. Isotherms will also be reported for these alloys at 303 K. A series of isotherms have also been measured at 423-523 K from which higher temperature thermodynamic data will be obtained. It is generally acknowledged that the Pd-Ag-H system is the most amenable to fundamental interpretation of binary Pd alloy-H systems because Ag is adjacent to Pd in the periodic table and in an early interpretation of the Pd-Agsystem it was proposed that Ag, like H, donates one electron to the d-band of Pd.6 It is of interest that the first Pd-alloy-H system investigated was the Pd-Ag-H system by Graham in 1866.7 The interaction of H2 with Pd-Au alloys is also of interest because it is a part of the series, Pd-Cu,8–12 Pd-Ag,10,13–17 and Pd-Au. Experimental Section Alloys were obtained from Engelhardt Inc. and they were generally remelted and annealed in vacuo for 3 days at 1133 K, rolled into foil, and then reannealed for 2 days at 1133 K. Judging from the linearity of the dependence of the X-ray lattice parameter on XAu2 and the agreement of the parameters with other investigations,18 the compositional uncertainty is about XAu ( 0.005. These alloys form solid solutions over the range to be investigated and H-induced ordering was avoided by not heating the alloys >523 K in high pH2.19 The equilibrium pressure-composition-temperature (p-c-T) measurements were carried out in an all-metal Sieverts’ type apparatus. The enthalpies were determined calorimetrically from the heats evolved or absorbed, q, when increments of H2 are absorbed or desorbed. If the increments are relatively small, then (∂q∂nH) ≈ (∂∆H/∂nH) ) ∆HH. The calorimeter has been described elsewhere20 and recent improvements have increased its sensitivity. Results and Discussion Hydrogen Isotherms and Thermodynamic Parameters from Equilibrium p-c-T Measurements (423-523 K). Figures 1-3 show isotherms at 423, 473, and 523 K for a series of Pd-Au alloys, and at these temperatures there are no
10.1021/jp100858r 2010 American Chemical Society Published on Web 04/21/2010
6118
J. Phys. Chem. B, Vol. 114, No. 18, 2010
Figure 1. Isotherms for a series of Pd-Au alloys at 423 K. The numbers on the curves indicate XAu.
Luo et al. ∆HH values for hydrogen solution are obtained from slopes of plots of ln p1/2 against 1/T at constant r and ∆SH is obtained from the intercepts of these ln p1/2 - 1/T plots.21 Table 1 gives ∆HH values as a function of the H concentration for various Pd-Au alloys where the values, ∆H°, H at infinite dilution have been obtained by extrapolation of the ∆HH values to r ) 0. It can be seen that ∆HH values becomes increasingly negative with increase of H concentration for alloys with XAu < 0.265, for XAu ) 0.265 ∆HH values appear to be independent of H concentration, and for XAu > 0.265 ∆HH increases with H concentration. ∆S°,′ H values (Table 2) appear to decrease slightly with H concentration up to about XAu ) 0.21 and then they increase at almost the same Au content at which ∆HH changes from decreasing to increasing with H concentration. ∆SH°,′ indicates values at finite H concentrations from eq 1 using β ) 1.0, that is
∆SHo, ) ∆SH + R ln(r/(β - r))
(1)
where β is the limiting H-metal ratio which, if all the octahedral interstices can be occupied, is 1.0; for the , calculation of ∆SH° ′ shown in the table, β is taken as 1.0. It can be seen that for β ) 1.0, ∆S°H clearly decreases with Au content where ∆S°H is the value at infinite dilution obtained from extrapolation of ∆SH°,′ to r ) 0. For these Pd-Au alloys it is likely that β * 1.0 and eq 2 can be employed to estimate realistic values of β that assumes that the nonconfigurational contributions to ∆S°H are the same for the alloys as for Pd.
Rlnβ ) ∆SHo (alloy) - ∆SHo (Pd) Figure 2. Isotherms for a series of Pd-Au alloys at 473 K. The numbers on the curves indicate XAu.
Figure 3. Isotherms for a series of Pd-Au alloys at 523 K. 0, XAu ) 0.09; ∆, 0.12; 2, 0.15; O, 0.17; ], 0.21; b, 0.265; 9, 0.30; ×, 0.35.
indications of any plateaux for these alloys except possibly for the XAu ) 0.09 alloy at 423 K (Figure 1). These figures show that the relative solubilities depend on pH2. For example, at 473 K and pH2 ) 1 bar the solubilities are similar for alloys with atom fractions Au, XAu ) 0.12-0.30 whereas at 2.5 bar the solubility is greatest at XAu ) 0.12-0.15 (Figure 2). This demonstrates that calculations showing that maximum H2 solubility occurs XAu ) 0.20 are ambiguous without specification of the pH2.3
(2)
) -51.0 where, for an average temperature of 473 K, ∆S°(Pd) H J/K mol H, which is the average of values found by La¨sser,16 Blaurock22 and that given by Kuji et al.23 It is seen from eq 2 that if β ) 1.0, ∆S°(alloy) ) ∆S°(Pd). H H The principal contribution to the nonconfigurational entropy is the H vibration. The only data available for the vibrational frequency dependence of H on the substituted metal content, XM, is for the XAg ) 0.20 alloy, which is found from inelastic neutron scattering to have the same vibrational frequency and therefore the same entropic contribution as H in Pd24 and it is assumed here that the same is true for Pd-Au alloys. Values of β estimated from eq 2 are shown in Figure 4 as a function of XM including a similar plot for Pd-Ag alloys.17 It can be seen that the β values for the Pd-Au alloys are consistently smaller than those for the Pd-Ag alloys. β values are also shown calculated by assuming that only octahedral interstices surrounded by Pd atom nearest neighbors are occupied and, in addition, values of β are shown assuming that interstices surrounded by either 5 or 6 nearest neighbor Pd atoms can both be occupied; for these calculations the alloys are assumed to be completely random solid solutions. It can be seen in the figure that the Pd-Ag alloys are closer to the latter and the Pd-Au alloys closer to the former assumption. Wagner and co-workers25 have shown from Mo¨ssbauer spectroscopy that H atoms are repelled by Au atoms and therefore do not occupy interstices adjacent to Au atoms, which is consistent with the results shown for these alloys in Figure 4. An unusual result for the dilute phase solubilities is that from r ) 0 ∼ 0.03 the pH2 - r isotherms are all rather similar for alloys with Au contents from XAu ) 0.12-0.30 at the three
Thermodynamics of Hydrogen in fcc Pd-Au Alloys
J. Phys. Chem. B, Vol. 114, No. 18, 2010 6119
TABLE 1: ∆HH for H2 Solution (Absorption) in Homogeneous fcc Pd-Au Alloys (393 to 523 K) where Enthalpies Are in Units of kJ/mol (1/2)H2 XAu
r ) 0(extrap)
r ) 0.01
r ) 0.015
r ) 0.020
r ) 0.025
r ) 0.030
r ) 0.035
0 0.09 0.12 0.15 0.188 0.21 0.265 0.30 0.35 0.40
-9.5 -11.1 -12.5 -13.3 -14.2 -15.5 -16.8 -18.7 -18.0 -20.3
-10.3 -11.9 -13.1 -14.0 -14.6 -15.6 -16.4 -18.5 -17.9 -18.9
-10.2 -12.0 -13.3
-10.3 -12.2 -13.4 -14.0 -15.0 -15.8 -16.9 -18.4 -17.2 -17.9
-11.0 -12.4 -13.7
-11.0
-11.6
-13.9 -14.4 -15.3 -15.9 -16.8 -18.2
-14.0 -14.6 -15.4 -15.9
-14.7 -15.7 -18.5 -17.3 -18.5
-15.1 -15.8 -18.2 -17.0 -17.0
TABLE 2: ∆S°H′ for H2 Solution (Absorption) in Homogeneous fcc Pd-Au Alloys (393 to 523 K) where Entropies are in Units of J/K mol (1/2)H2 ,
XAu
r ) 0(extrap)
r ) 0.01
r ) 0.015
r ) 0.020
r ) 0.025
r ) 0.030
r ) 0.035
0 0.09 0.12 0.15 0.188 0.21 0.265 0.30 0.35 0.40
-52 -54.1 -56.1 -56.5 -58.7 -61.3 -62.9 -66.4 -65 -72
-53.7 -54.9 -56.7 -57.7 -59.0 -60.9 -62.8 -65.9 -64.9 -69.1
-53.2 -54.7 -56.6
-53.4 -54.7 -56.7 -57.5 -59.3 -60.8 -62.6 -65.8 -63.9 -67.6
-54.1 -54.8 -57.0
-52.6
-54.6
-57.0
-59.5 -60.7
-59.6 -60.6 -62.5 -65.3
-57.1 -57.6 -59.6 -60.4
-59.0 -60.9 -65.9 -64.1 -68.6
temperatures (Figure 1-3) and this is especially apparent at 523 K (Figure 3). The nearly universal solubility plots for alloys with contents from XAu ) 0.12-0.30 suggests that the ∆HH and -T∆SH terms compensate, that is, even though both change with XAu, their changes cancel. ∆H°H and ∆S°H are shown in Figures 5 and 6 where they both become more negative with XAu and there is an almost linear dependence of each on XAu excluding the limiting value at XAu ) 0, pure Pd. Plots of ∆H°H and -T∆S°H (473 K) against XAu are shown in Figure 7 along with the average of the two. This illustrates how the two contributions nearly cancel leading to similar, dilute phase solubilities over a range of XAu values. ∆S°H becomes more negative with XAu because of the blocking of interstices by Au while the increase of lattice constant with XAu causes an increase of exothermicity of H2 solution. A
Figure 4. O, β for Pd-Ag alloys estimated from the ∆S°H values as described for the Pd-Au alloys;17 b, β calculated assuming that octahedral interstices surrounded by both 5 and 6 Pd atoms nearest neighbors can be occupied. ∆, β estimated for Pd-Au alloys from ∆S°H values; and 2, calculated assuming that only the octahedral interstices surrounded by 6 nearest neighbor Pd atoms can be occupied.
-65.4 -63.5 -65.9
decrease in ∆S°H with atom fraction alloying element is also observed for Pd-Ag alloys17 and most other Pd alloy-H systems.26 Plots of RT ln p1/2(β - r)/r against r according to eq 3 are frequently made for Pd-H where β ) 1.0.21,27 The slope of such plots are equal to g1 and the intercepts, ∆µ°()∆H °H H T∆S°), H
RT lnp1/2(β - r)/r ) ∆µHo + g1r
(3)
where g1 is the first order term of a polynomial expansion of the excess H chemical potential in r. In the configurational contribution of eq 3 β *1.0 for most Pd alloy-H systems. Exceptions appear to be the Pd-Ni28 and Pd-Cu systems.11 The choice of different β values in eq 3 for Pd alloys has not been explored using for plots of RT ln p1/2(β - r)/r against r. This will be done here for an Pd-Au alloy that has a significant fraction of alloying element, that is, XAu ) 0.21. Results are
Figure 5. ∆H°H determined by extrapolation of ∆HH to r ) 0 as a function of XAu.
6120
J. Phys. Chem. B, Vol. 114, No. 18, 2010
Luo et al. TABLE 3: Values of g1 for H2 Solution in Homogeneous fcc Pd-Au Alloys at 473 K where g1 Is Evaluated from the Slopes of Plots of RT ln p1/2(β - r)/r against r where β is Chosen from the Experimental Values Shown in Figure 4 and g1 Values are in Units of kJ/mol (1/2)H2
Figure 6. ∆S°H as a function of XAu where ∆S°H has been determined by extrapolations of ∆SH°,′ to r ) 0.
Figure 7. O, ∆H°H and ∆, -T∆S°H as a function of XAu. ---, average of ∆H°H and -T∆S°H for each XAu.
Figure 8. Plots of RT ln p1/2((β - r)/r) vs r (473 K) for the Pd0.79Au0.21 alloy using different β values. The β values of 1.0, 0.29, and 0.243 are for all octahedral interstices available, for only those with 6 nearest neighbor Pd atoms assuming a completely random, disordered alloy and for the experimental value of β from eq 2.
seen in Figure 8 (473 K) where the plots are seen to be very linear to r ≈ 0.062. The β values chosen are 1.0, 0.29, and 0.243, which correspond to (a) all octahedral sites available, (b) experimental values of β from eq 2, and (c) β calculated using the fraction of interstices with only Pd nearest neighbors. It is seen in Figure 8 that the various plots are all very linear, R2 > 0.99, but the slopes and intercepts differ for the different β values. When the intercepts, ∆µ°, H divided by T are plotted
XAu
β ) 1.0
β ) fraction Pd6 n.n.
exp β from eq 2
0 0.09 0.12 0.15 0.19 0.21 0.265 0.30
-33.8 -30.6 -30.7 -27.3 -22.1 -18.3 -6.5 +6.1
-33.8 -33.7 -35.6 -34.2 -32.8 -32.0 -31.0 -30.5
-33.8 -30.6 -32.6 -30.3 -28.1 -29.4 -20.8 -19.5
against 1/T, however, the slopes give the same values of ∆H°H for the different β values shown in Figure 8 and, as would be expected, the intercepts, ∆S°, H differ. Table 3 shows values of g1 calculated from the three different values of β. For (a), β ) 1.0, g1 falls in magnitude with increase of XAu to become positive at 0.30. For the “experimental” values of β, (b), g1 falls in magnitude but does not become positive at XAu ) 0.30; and for (c) β ) fraction of interstices with only nearest neighbors, it is remarkably constant and similar to that for Pd-H. It seems that the H-H interaction, which is largely due to an elastic effect,29,21 should be similar for Pd and the alloys since the lattice expansion due to H is similar but other factors which contribute to apparent H-H interactions, such as an electronic effect due to “band-filling” would contribute to an increase of g1 with XAu. Isotherms at 303 K and Enthalpies Determined from Reaction Calorimetry. Isotherms. Since the solubility data differ significantly between the higher temperature isotherms (Figures 1-3) and isotherms at 303 K, the latter will be shown for the Pd-Au alloys some compositions of which differ from those shown above. Dilute phase absorption isotherms (303 K) are shown in Figure 9 where, in contrast to the higher temperature isotherms (Figures 1-3), hydride phases form for the lower content Au alloys. At low pH2, the solubility increases with increasing Au content for XAu < 0.264, but this trend is reversed at greater Au concentrations (Figure 9). The alloys studied here which form hydride phases, XAu ) 0.05 and 0.10, exhibit gradual transitions to the hydride phase in contrast to Pd(annealed)-H. Gradual transitions to the hydride phase are often found for other Pd alloys especially at higher added metal contents. This has not been pointed out explicitly before and may be due to small compositional variations such that the hydride phase forms at slightly different alloy compositions that would lead to phase transitions at slightly different pH2. This would also lead to more sloping plateaux than for Pd-H which is often a feature of Pd-alloys. Figures 10 and 11 show absorption and desorption isotherms (303 K) over the entire range of H contents accessible up to pH2 ) 0.36 MPa; plateaus are seen for the XAu ) 0.05 and 0.10 alloys. The ordinate in Figure 10 extends over a large pH2 scale, whereas that in Figure 11 extends over a smaller scale showing more detailed behavior. The average plateau pH2 values, that is, (pf + pd)/2 for both the 0.05 and 0.10 alloys are close to the average value of Pd-H (Figure 10) where pf and pd are the hydride formation and decomposition plateaus, respectively. The absorption and desorption plateaux all lie within each other, that is, pf and pd for the XAu ) 0.05 lie within the plateau pressures for Pd-H and similarly pf and pd for the 0.10 alloy lie within those of the 0.05 alloy. Hysteresis decreases with XAu because the extent of the two-phase regions decrease and
Thermodynamics of Hydrogen in fcc Pd-Au Alloys
Figure 9. Dilute phase absorption isotherms (303 K) for a series of Pd-Au alloys. The numbers on the curves represent XAu.
J. Phys. Chem. B, Vol. 114, No. 18, 2010 6121
Figure 11. Isotherms (303 K) on an expanded p1/2 scale up to ∼6.4 kPa for a series of Pd-Au alloys. The numbers on the curves represent XAu. O, absorption, b, desorption.
to ∼0.17 at 298 K2 but it is possible that the somewhat higher temperature of 303 K causes the difference. It can be seen in Figure 10 that there is a decrease of solubility with XAu in the pH2 regions corresponding to H contents greater than the plateaux or, if there is no plateau, in the regions where pH2 increases steeply with r (Figure 10). The solubility at pH2 ) 1 bar (303 K) as a function of XAu is shown in Figure 12 where it can be seen that the decrease is not quite linear. Calorimetrically Determined ∆HH and ∆SH and the Plateau Values (303 K). Calorimetric values of |∆HH| and |∆Hplat| as a function of the H concentration are shown in Figure 13 for the dilute regions and for the transition to the hydride phases. There is a discontinuous increase in the enthalpy magnitudes in passing from the dilute to the hydride phases; such behavior is expected30 and has been reported for Pd-H.31 |∆SH| and |∆Splat| values are shown in Figure 14. The ∆SH values refer to the reaction of (1/2)H2 (g) at 1 bar with the alloy-H system and have been determined from the calorimetrically measured |∆HH| values and the equilibrium pH2 using
∆SH ) ∆HH /T + R ln pH1/22
Figure 10. Isotherms (303 K) up to ∼0.36 MPa for a series of Pd-Au alloys. The numbers on the curves represent XAu. O, absorption, b, desorption.
therefore the magnitude of the abrupt volume expansions decrease. The XAu ) 0.15 alloy does not exhibit hysteresis and does not form two phases at 303 K as will be shown below. This differs somewhat from an earlier finding that it persisted
(4)
where the first term on the right-hand-side is the entropy change for reaction with (1/2)H2 at pH2(equil) and the second term allows for the change from pH2(equil) to 1 bar. Because of hysteresis, pH1/22 in eq 4 must be replaced by (pf × pd)1/4,32 that is
∆Splat ) ∆Hplat /T + R ln(pf × pd)1/4
(5)
It can be seen in Figure 14 that there is a discontinuous increase from the |∆SH| values at the end of the dilute phase to the plateau value, |∆Splat|, similarly to the behavior of the enthalpies (Figure 13).
6122
J. Phys. Chem. B, Vol. 114, No. 18, 2010
Figure 12. H2 solubility at 1 bar and 303 K for a series of Pd-Au alloys.
Luo et al.
Figure 14. ∆SH as a function of r (303 K). The atom fractions of Au are indicated on the graph.
TABLE 4: Plateau Enthalpy and Entropy Values for Absorption where the Enthalpy and Hysteresis Are in Units of kJ/mol (1/2)H2 and the Entropy in Units of kJ/K mol (1/2)H2a XAu
∆Hplat
∆Splat
hysteresis, (RT ln(pf/pd)1/2
0 0.05 0.10
-19.2 -20.3 -21.2
-46.1 -49.7 -52.7
0.96 0.63 0.38
a
where pf and pd are the plateau pressures for hydride formation and decomposition, respectively, and pplat is taken as (pf × pd)1/2.
Figure 13. ∆HH as a function of r (303 K). The atom fractions of Au are indicated on the graph.
Enthalpies and entropies for the plateau reaction, hydride formation from the dilute phase, are shown in Table 4 where both become more negative with increase of Au content. Of the Pd-Au alloys investigated here, only the XAu ) 0.05 and 0.10 alloys exhibit plateaus at 303 K. It has been shown elsewhere that calorimetrically measured enthalpies for the plateau reaction are unaffected by hysteresis20 and Table 4 gives the values that were experimentally determined for hydride
formation (absorption). (pf × pd)1/2 is nearly constant with Au content for these alloys but ∆Hplat becomes more negative with increase of XAu (Table 4) and therefore ∆Splat must also become more negative to compensate which is a similar compensation that leads to the nearly universal dilute phase solubilities (Figures 2 and 3). Values of pf and pd have been evaluated at the center of the plateaus. For Pd-rich alloys, ∆Splat for absorption is found to be nearly constant with XM at -46 ( 2.0 J/K mol H.33 Pd-Au alloys are obviously an exception to this because |∆Splat| increases with XM (Table 4). The isotherm for the XAu ) 0.15 alloy suggests a plateau region extending from about r ) 0.225 to 0.27 (Figure 15); however, the lack of hysteresis indicates that there is no twophase coexistence region, that is, a plateau. A plateau requires that ∆HH and ∆SH be constant over a range of r values where the two phases coexist. Results from reaction calorimetry for the XAu ) 0.15 alloy as a function of r are shown in Figures 16 and 17 where both absorption and desorption data are shown and they agree quite well. Because of the absence of a region where either of these partial thermodynamic parameters are constant, it can be concluded that there is no two phase, plateau region. There is a maximum in |∆HH| at about r ) 0.25 (Figure 16) and also one for |∆SH| (Figure 17), although it is not as pronounced for the latter. The plateau-like appearance in the
Thermodynamics of Hydrogen in fcc Pd-Au Alloys
J. Phys. Chem. B, Vol. 114, No. 18, 2010 6123
Figure 15. Isotherm for the XAu ) 0.15 alloy (303 K) where O, 4, absorption, and 2, desorption.
Figure 17. ∆SH-r relationship for the XAu ) 0.15 alloy (303 K) where O, 4, absorption, and b, 2, desorption.
Figure 16. ∆HH - r relationship for the XAu ) 0.15 alloy (303 K) where O, 4, absorption, and b, 2, desorption.
isotherm (Figure 15) results from a cancelation as shown in Figure 18 where ∆HH and -T∆SH ordinates are plotted with the same energy sub-division. Their sum is shown by the dashed line which gives rise to a pseudoplateau region where pH2 is rather constant. Calorimetrically measured ∆HH-r plots for absorption over the range of H contents up to contents corresponding to pH2 ≈ 3.6 bar (303 K) are shown for alloys that do not form hydride phases (303 K) in Figure 19. It can be seen that, similarly to the XAu ) 0.15 alloy, there is a steady decrease of |∆HH| with r after maxima for these alloys except for the XAu ) 0.35 alloy, which does not have a maximum. |∆HH|-r plots (303 K) are shown for all of the alloys in Figure 20 including plateau values. The slopes of |∆HH| with r, after the plateau regions, are all similar. Continuous increases of |∆HH| before the decreases at the end of the plateau (Figure 20) are artifacts because there
Figure 18. ∆HH and -T∆SH-r relationships for the XAu ) 0.15 alloy (303 K) where O, ∆HH and b, -T∆SH.
should be a discontinuous increase but, because of some overlap effects due to hysteresis, the transitions appear to be continuous as discussed for Pd-H.20 |∆SH|-r plots are seen in Figure 21 where the increase of |∆SH| in the dilute phase region is similar for all the alloys. There is a decrease of |∆SH| at the end of the plateau for the XAu ) 0.05 and 0.10 alloys that is also present for pure Pd.20 For the XAu ) 0.19 alloy, which does not have a plateau, there is a clear maximum at a similar r value as its ∆HH-r relation (Figure 19). It is of interest to compare isotherms for the Pd-Au alloys with those of other Pd-1B alloys and, for this purpose, alloys of composition Pd0.90M0.10 alloys are compared in Figure 22 (303 K) where the plateaus for the Pd0.90Cu0.10 and Pd0.90Ag0.10 alloys are greater and smaller, than that for the Pd0.90Au0.10 alloy and for Pd that is very similar to the latter. The lattice parameter increases of Pd with Ag or Au substitution are very similar so
6124
J. Phys. Chem. B, Vol. 114, No. 18, 2010
Luo et al.
Figure 19. Calorimetrically determined ∆HH-r relationships for Pd-Au alloys that do not form a hydride phase at 303 K.
Figure 21. ∆SH-r relationships for Pd-Au alloys (303 K).
Figure 22. Absorption isotherms (303 K) for Pd and for Pd0.90M0.10 alloys where M ) Au, Ag,17 Cu.12 Figure 20. ∆HH-r relationships for Pd-Au alloys (303 K), some of which form hydride phases. 18
that their thermodynamic differences are surprising. Enthalpies and entropies for H2 absorption by these alloys are shown in Table 5. It can be seen from Table 5 that the Pd-Cu alloys, which are contracted alloys, have less exothermic enthalpies of H2 solution as the Cu content increases which is expected.
Conclusions Thermodynamic data have been determined for disordered, fcc Pd-Au alloys both from p-c-T measurements (423-523 K) and from calorimetry (303 K). Several unusual features emerge. (1) There is nearly universal dilute phase H2 solubility relationship for alloys from XAu ≈ 0.12 to 0.30 (423-523 K), (2) the plateau pressures are similar for the XAu ) 0, 0.05, and 0.10 alloys despite that fact that ∆Hplat becomes increasingly
Thermodynamics of Hydrogen in fcc Pd-Au Alloys
J. Phys. Chem. B, Vol. 114, No. 18, 2010 6125
TABLE 5: Plateau Enthalpy and Entropy Values for Absorption of (1/2)H2 (303 K) where the Former Is in Units of kJ/mol (1/2)H2 and the Latter in Units of J/K mol (1/2)H2 alloy
∆H°H
∆Hplat
∆S°H
∆Splat
Pd Pd0.95Cu0.05 Pd0.90Cu0.10 Pd0.95Ag0.05 Pd0.90Ag0.10 Pd0.95Au0.05 Pd0.90Au0.10
-10.2 -9.9 -9.6 -12.0 -13.3 -11.5 -13.3
-19.2 -17.6 -16.3 -20.3 -21.2 -20.3 -21.2
-54 -54.3 -54.6 -56.0 -58.4 -55.1 -54.1
-46.1 -44.311 - 43.511 -48.517 - 47.817 -49.7 -52.7
exothermic with Au, (3). ∆Splat is not constant as found for most Pd-alloys, -46 ( 2.0 J/K mol H,33 but becomes more negative with XAu. References and Notes (1) Sakamoto, Y.; Chen, F.; Ura, M.; Flanagan, T. Ber. Bunsen-Ges. Phys. Chem. 1995, 99, 807. (2) Maeland, A.; Flanagan, T. J. Phys. Chem. 1965, 69, 3575. (3) Sonwane, C.; Wilcox, J.; Ma, Y. J. Chem. Phys. 2006, 125, 184714. (4) Allard, K.; Maeland, S.; Simons, J.; Flanagan, T. J. Phys. Chem. 1968, 72, 136. (5) Shamsuddin, M.; Kleppa, O. J. Chem. Phys. 1979, 71, 5154. (6) Mott, N.; Jones, H. The Theory of Meals and Alloys; Oxford University Press: Oxford, 1936. (7) Graham, T. Philos. Trans. R. Soc. London 1866, 156, 415. (8) Burch, R.; Buss, R. J. Chem. Soc. Faraday Trans.I 1975, 71, 913. (9) Fisher, D.; Chisdes, D.; Flanagan, T. J. Solid State Chem. 1991, 2, 149. (10) Salomons, E.; Hemmes, H.; Griessen, R. J. Phys.: Condens. Matter 1990, 2, 817.
(11) Sakamoto, Y.; Ishimaru, N.; Makai, Y. Ber. Bunsen-Ges. Phys. Chem. 1991, 95, 680. (12) Flanagan, T.; Luo, W.; Clewley, J. J. Alloys Compd. 2003, 356357, 13. (13) Brodowsky, H.; Poeschel, E. Z. Phys. Chem. 1965, 44, 143. (14) Holleck, G. J. Phys. Chem. 1970, 74, 503. (15) Picard, C.; Kleppa, O.; Boureau, G. J. Chem. Phys. 1979, 70, 2710. (16) La¨sser, R. Tritium and Helium-3 in Metals; Springer-Verlag: Berlin, 1989. (17) Flanagan, T.; Wang, D.; Luo, S. J. Phys. Chem. B 2007, 111, 10723. (18) Sakamoto, Y.; Hirata, S.; Nishikawa, H. J. Less-Common Mets. 1982, 88, 387. (19) Lee, S.-M.; Noh, H.; Flanagan, T.; Luo, S. J. Phys. Cond. Matter. 2007, 19, 326222. (20) Flanagan, T.; Luo, W.; Clewley, J. J. Less-Common Met. 1991, 172-174, 42. (21) Flanagan, T.; Oates, W. Annu. ReV. Mater. Sci. 1991, 21, 269. (22) Blaurock, J. Ph.D. Dissertation, Universita¨t Mu¨nster, Mu¨nster, Germany, 1985. (23) Kuji, T.; Oates, W.; Bowerman, B.; Flanagan, T. J. Phys. F: Met. Phys. 1983, 13, 1785. (24) Chowdhury, M.; Ross, D. Solid State Commun. 1973, 13, 229. (25) Karger, M. Pro¨bst, F. Schu¨ttler, B. Wagner, F. Metal Hydrogen Systems, Pergamon Press, Oxford, 1982; p 187. (26) Flanagan, T.; Wang, D. J. Alloys Compd. 2010, 488, 72. (27) Wicke, E.; Nernst, G. Ber. Bunsen-Ges. Phys. Chem. 1964, 68, 224. (28) Sakamoto, Y.; Matsuo, T.; Sakai, H.; Flanagan, T. Z. Phys. Chem. N.F. 1989, 162, 83. (29) Alefeld, G. Ber. Bunsen-Ges. Phys. Chem. 1972, 76, 746. (30) Luo, W.; Flanagan, T. J. Phase Equilib. Diffus. 1994, 15, 20. (31) Luo, S.; Flanagan, T. Scr. Mater. 2005, 83, 1269. (32) Flanagan, T.; Oates, W.; Park, C.-N. Prog. Solid State Chem. 1995, 23, 291. (33) Zhang, W.; Luo, S.; Flanagan, T. J. Alloys Compd. 1999, 293295, 1.
JP100858R
