Search

The paper presents the experimental results of rigid-lattice proton magnetic resonance spectra and second moments in Pd–H, Pd 0.90 Ag 0.10 –H, Pd 0.80 Ag 0.20 –H, Pd 0.75 Ag 0.25 –H, and Pd 0.65 Ag 0.35 –H alloys at T = 2.4 K in a wide hydrogen concentration range. Free induction decay (FID), solid and inhomogeneous echoes were detected and interpreted. The interpretation based on dipole–dipole interaction of homogeneously distributed proton spin system is satisfactory at high hydrogen concentration, but an inhomogeneous field of paramagnetic origin (most probably coming from Fe impurities) plays the dominant role at small hydrogen content. An unexpected correlation was found between the macroscopic magnetic susceptibility and the second moment coming from the inhomogeneous internal field.

Crystalline Pd 1-x Ag x -H y (x = 0-0.35) alloys were studied as model systems representing a chemically disordered system for hydrogen-storage materials. Extensive 1 H relaxation rate (R 1 , R 1p and R 2 ) measurements were carried out in a wide temperature range of 2.4-400 K. Palladium-silver alloys of silver concentration as high as 35 at.% were studied by NMR spectroscopy for the first time. The relaxation mechanisms magnetic dipolar relaxation (hydrogen diffusion), Korringa relaxation, relaxation by paramagnetic impurity ions and cross-relaxation to quadrupolar metal nuclei were identified and characterized. The presence of cross-relaxation between 1 H and quadrupolar 105 Pd nuclear spins was assumed and demonstrated for the first time in Pd-H and Pd 1-x Ag x -H systems. More than one jump processes of diffusing hydrogen were found to be present in the studied palladium-silver alloys. Only lower estimated limit values of Korringa constants could be obtained since no frequency independent R 1 versus T region was found. These values are significantly higher than the ones reported by others.

In this work, the electrochemical hydrogen doping and withdrawal process of Pd metal was examined with respect to the applicability of Faraday’s law. It was found that after a suitable electrolytic pre-treatment, the normalized resistivity ( R / R 0 ) versus hydrogen concentration ( c F = H/M) curves where c F was determined from Faraday’s law by assuming that the 100% current efficiency is independent of the value of the applied charging current density in a wide range. In the α-Pd(H) phase (up to c (α max ) = 0.015H/M), a good quantitative agreement was found with previous resistivity studies by both electrolytic and gas-phase charging. However, for c > c (α max ) our R / R 0 versus c F data did not agree with the R / R 0 versus c H curve from previously reported electrochemical charging experiments of Pd metal where the doped amount of hydrogen, c H , was determined by direct methods. The two sets of resistivity data can be brought into good agreement by assuming that the current efficiency for the absorption process is η a = 1.55. Similarly, for the electrolytic desorption process, an agreement with the previously reported R / R 0 versus c H curve could be achieved by taking η d = 1.9. These current efficiency values higher than unity would imply that each hydrogen absorbed/desorbed by electrolytic charge/discharge processes is accompanied by the absorption/desorption of some excess H atoms by processes not requiring electric charge transfer (i.e., only neutral H atoms are involved). At present, we do not have a physical/chemical picture about the nature of such processes.

The partition of hydrogen into diffusible and residual parts was realized by pulse NMR spectroscopy, by gas chromatography and by prompt gamma activation analysis (PGAA). The total hydrogen content was determined by the two non-NMR methods and the diffusible (mobile) component by CPMG NMR pulse sequence. Results on amorphous Ni(Cu)–Zr–H systems of different compositions are shown. Partially crystallized samples were also studied as an extension. A method proposed by us directly gives the fractional population of hydrogen atoms in the free (mobile) state on the spin–spin relaxation time scale. On the other hand the least values of the residual hydrogen content correlate surprisingly well with the numbers of filled four Zr-type H-sites calculated by Batalla et al. [NATO ASI Ser. 136 (1985) 203] for 0.21-nm exclusion distance.

The paper presents the results of an experimental investigation of rigid-lattice proton magnetic resonance (PMR) spectra and second moments in Ni 0.67 Zr 0.33 –H, Ni 0.50 Zr 0.50 –H, Ni 0.33 Zr 0.67 –H binary glassy alloy–hydrogen systems in the range of hydrogen-per-metal ( H / M ) ratio of 0.14≤ H / M ≤1.9. The line shape can be described by the empirical Harper-Barnes function with exponent values ranging from 1.45 to 2.95 in the whole hydrogen concentration range. The generally assumed Gaussian form is valid in a narrow H / M interval only. The second moment as a function of H / M can be fitted by a power function of an exponent 3/4. The exponent is independent of the composition of the alloy. We found that a simple lattice gas model based on a homogeneous hydrogen distribution contradicts the experimental results. Instead, a short-range order model has been proposed in which the hydrogen atoms are located in the centres of distorted tetrahedra of metal atoms, that is in the basic building blocks of the amorphous alloys. The Switendick criterion was also taken into consideration. We have shown that few proton first neighbours (0–2) and a small number of second neighbours (1–8), that is, a cluster of hydrogen is sufficient to account for the measured value of the second moment. Ternary Zr 0.5 Ni 0.5− x Cu x –H alloys have been investigated, too. The addition of Cu nuclei did not lead to the second moment enhancement expected for a homogeneous distribution of components.

Results of electrical resistance measurements are given on Pd80 Ag20 alloy during hydrogen uptake from gaseous H2 atmosphere and from liquid electrolyte by cathodic charging. For both cases in the course of a monotonous increase of the hydrogen concentration in the alloy, the electrical resistivity shows anomalous behavior. These kinetic curves reflect the effect of phase changes in the non-equilibrium state. The ‘non-equilibrium’ expression is used in this entire work in the sense that no care was taken in waiting for the stable value of resistivity after any partial filling, but the resistivity was frequently measured while the sample was continuously charged with hydrogen. A model has been proposed for the interpretation of the experimental findings.

The hydrogen charging and discharging processes are investigated in Ni67−xCuxZr33 glassy alloys by in situ resistance measurements in order to clarify the surface and bulk contribution to the activation process. The process has been monitored by conventional DC method using a computer-controlled system, exposing the samples to hydrogen at pressure from 0.1–8.6 MPa at temperatures of 293 K and 393 K. The H-contents (H/M) were determined by weight and by pulse NMR measurements. The NMR data are obtained by the Carr–Purcell–Meiboom–Gill (CPMG) method. The charging is extremely long and the resulting H-distribution is inhomogenous during the first cycle, especially at 293 K. This fact shows the significance of the surface-activation during the H-uptake from the gas phase. The duration of the second saturation decreases significantly, and the resistance–time curves obtained: on two parallel samples are very similar. The saturation value for the normalized resistance (R/R0) remains almost the same during several cycles. Increasing the temperature of the charging (393 K), the equilibrium value of the resistance decreases. The resistance of the samples at the end of the discharging process is always higher than before charging. This indicates the partial irreversibility of the H-absorption at constant temperature. This observation is correlated with the NMR results.

1 H spin–spin relaxation time ( T 2 ) and complementary hydrogen content, PMR spectrum width and spin–lattice relaxation time ( T 1 ) have been measured in Zr y (Ni 1− x Cu x ) 1− y –H ternary amorphous alloys of different hydrogen content at 0≤ x ≤33 at.% Cu and y =33, 50 and 67 at.% Zr concentrations using Carr–Purcell–Meiboom–Gill (CPMG), solid-echo and saturation recovery pulse sequences, respectively. At high hydrogen contents the T 2 measured by the slope of the CPMG echo train depends on both the Zr and Cu content, but is independent of the hydrogen content. The differences in the spin–spin relaxation behaviours can be attributed to the substantial change of correlation time and not to the change of activation energy or local fields. The measurements were made in the “motional narrowing” state, consequently our E a and τ ∞ quantities are averaged to the diffusion motion of protons taking part in this process. At small hydrogen contents T 2 depends on hydrogen content and the T 2 vs. 1/ T curves cannot be fitted by single Arrhenius plots. The hydrogen content measured by echo train has turned out to be systematically smaller than that measured by weight increase, demonstrating that not all the hydrogen takes part in the diffusion process.

A two-exponential spin-lattice ( T 1 ) relaxation process of hydrogen nuclear magnetization was found in a series of hydrogenated amorphous carbon samples prepared under different self-bias conditions from benzene. The temperature dependence of these relaxation times reveals the different dynamic background of the two sub-systems. Low activation energies (depending on self-bias voltages) were measured for the longer relaxation time components; temperature independence was detected for the shorter relaxation time components.

Spin-spin relaxation time ( T 2 ) and complementary hydrogen content (H/M), PMR spectrum width (FWHM) and spin lattice relaxation time ( T 1 ) have been measured in (Ni 1 − x Cu x ) 0.5 Zr 0.5 -H y ternary amorphous alloys at 0 ⩽ x ⩽ 29 at.% concentrations using Carr-Purcell-Meiboom-Gill (CPMG), solid-echo and saturation recovery pulse sequences respectively. The T 2 relaxation time measured by the slope of the CPMG echo train depends on both the hydrogen and Cu content. The dependence can be attributed to the change of correlation time and not to the change of activation energy or local field. The measurements were made in the ‘motional narrowing’ state; consequently our E a and τ ∞ quantities are averaged to the diffusional motion of protons taking part in this process. The extrapolated amplitude of the echo train measured on the alloy-hydrogen system and normalized to the same quantity in water, gives the value of proton magnetization in thermal equilibrium, that is the hydrogen content free from any perturbing effects. The hydrogen content measured by CPMG echo train has turned out to be systematically smaller than that measured by weight increase.

Equilibrium pressure and potential measurements at room temperature as a function of hydrogen concentration are reported for Ni 1− x Zr x amorphous alloys with 0.33 ⩽ x ⩽ 0.45. The full range of charge of the samples was achieved by taking into account the hydrogen gas evolved in the closed electrochemical cell. The change in slope of the equilibrium potential with the state of charge shows that different types of site are filled. On the basis of the pressure-composition (P-C) diagrams, the samples with x = 0.33, 0.36 and 0.39 were selected for further electrochemical investigations.

The aim of this paper is to reinvestigate the proton magnetic resonance spectra in Zr 0.5 Ni 0.5− y Cu y H x metallic glasses. Solid-echo pulse sequences were used to analyze the line shape. The measurements were made in the temperature range 2.2–300 K and at a frequency of 87.6 MHz. The reinvestigation of the line shape mostly at low temperatures was initiated after having realized that the splitting reported earlier on our samples was a consequence of the relatively long recovery time of our spectrometer. The present solid-echo measurements yielded simple spectra describable by a Harper-Barnes line shape close to gaussian.

The proton NMR line shift with respect to water, the bulk susceptibility and in a few cases the Korringa contribution to the spin lattice relaxation time were measured in Zr0.33Ni0.67 Hx amorphous alloys with hydrogen concentration 0.1 ⩽ x ⩽ 0.7and at 343 K. All the parameters, including the hydrogen content and the bulk susceptibility, were measured simultaneously in situ, and in the shift measurements a resolution of ±1 ppm was attained. The residual Knight shift depends on the hydrogen content and is positive relative to water. Preliminary low temperature spin lattice relaxation time measurements indicate that the positive shift is the consequence of the balance of s- and d-contributions.

Proton magnetic resonance spectra of Zr 0.5 Cu y Ni 0.5−y H 1 metallic glass samples with 0 ≤ y ≤ 0.3 were measured in the temperature range 77 K to 400 K and at 8, 28.5, 87.6 and 88.5 MHz frequencies. The data of the free-induction decay (FID) signal were compiled and Fourier transformed by a SMIS NMR spectrometer. The temperature dependence of the line shapes shows three distinctly different features: (i) in the temperature range 400 K to 270 K the spectrum is a motionally narrowed Lorentzian line; (ii) in the temperature range 270 K to 140 K the spectrum consists of a narrow Lorentzian component and a substantially wider Gaussian one; (iii) in the temperature range 140 K to 77 K, the spectrum is a “two-peaked” line. The distance between the two peaks (splitting) was found to be independent of the temperature and the frequency, that means that the low-temperature spectrum is probably the Pake doublet of hydrogen pairs. From the analysis of the splitting of the Pake doublet, a proton-proton distance of 0.167 nm was estimated. This distance corresponds to a transition metal mediated hydrogen pairing.

Results of resistivity measurements during electrolytical hydrogen charging on some Fe, Co and Ni based alloy ribbons in amorphous, nanocrystalline and body-centred cubic state are presented. The peculiarities of the H uptake are significantly different for the different structural modifications of alloys with the same chemical composition. In the limited concentration interval of equality of the transferred charge and the number of actually absorbed hydrogen atoms, the concentration variation of the resistivity increment caused by unit hydrogen concentration was determined.

The proton content in Zr 0.33 Ni 0.67 H x amorphous alloy was monitored by changes in electrical resistivity and with the extraplated spin echo amplitude measured by the Carr-Purcell-Meiboom-Gill (CPMG) nuclear magnetic resonance pulse sequence in the temperature interval 300 to 450 K. Both methods show that the hydrogen desorption processes can be approximately divided into two ranges. The first is a fast process which is at present not interpretable. In the second stage, the process is thermally activated, and the activation energy calcualted from the resistivity changes is equal to 0.32 ± 0.04 eV. The CPMG spin echo measurements also give the spin-spin relaxation time of diffusing protons and the activation energy of this process, 0.34 ± 0.02 eV, is exactly the same as extracted from the resistivity measurements.

The electronic structure of amorphous ZrNiP and ZrNiPH alloys was studied by means of ultraviolet photoemission spectroscopy. It was established that new states were induced around 5.7 eV binding energy by hydrogen and, at the same time, the density of states (DOS) at the Fermi level decreased. The Ni peaks below the Fermi level are shifted to higher binding energies upon hydrogenation whereas the position of the Zr peak remains unchanged but its intensity is reduced due to the presence of hydrogen. The decrease of the DOS at the Fermi level indicates that H atoms interact with Zr atoms only. On the other hand, the clear shift in the Ni d-states shows that the Ni states are affected by hydrogen via changes in the ZrNi bonding.

Conductance noise measurements done on (melt-quenched) glassy metal point contacts are reported. The lower limit of the noise factors obtained was stronger by 3–5 orders of magnitude than that usual in ordinary metals. That result supports well the view about the role of defects in generating the 1/f noise, but there are some doubts, whether the relevant defects are necessarily mobile.