Yongfa Zhu

Activation energies for the desorption of H2, H− and electron from saline hydrides heated in vacuum

Abstract To study the thermodynamic and thermionic properties of saline hydride (MH n , n =1 or 2; M=Li, Na, Mg, Ca or Sr), powders (ca. 1xa0mg) were deposited on a molybdenum heater, and the currents of H − and electrons (e − ) desorbed directly from the hydrides were measured in vacuum as a function of temperature ( T ) up to 1100xa0K at a constant rate. A dual mass spectrometer system, in which the H 2 desorbed from MH n was converted into H 2 + by electron impact, was used. The temperature-programmed desorption spectra for H − , e − and H 2 were analyzed using a leading edge method. The following results were obtained: (1) each desorption became appreciable at an appearance temperature, which increased in the order of H 2 (ca. 700xa0K) and both e − and H − (above ca. 800xa0K); (2) the activation energies ( E 0 , E e and E − in kJxa0mol −1 ) for the desorption of H 2 , e − and H − were as follows — LiH (95, 596, 660), NaH (60, 221, 173), MgH 2 (91, 335, 318), CaH 2 (87, 428, 612) and SrH 2 (63, 354, 308); (3) as T increased from ca. 900 to 940xa0K in the case of LiH, E e corresponding to the work function increased from ca. 580 to 600xa0kJxa0mol −1 ; (4) such an increase was caused by a decrease in the surface density of these active sites (e.g., Li) which were created by thermal decomposition (e.g., LiH → Li + 1 2 H 2 ) and also which promoted the desorption of e − and H − .

Thermal desorption of H2, H- and electron by temperature-programmed heating of saline hydrides in vacuum

Abstract To clarify the thermochemical and thermionic properties of saline hydrides, a small amount (ca. 1xa0mg) of powdery NaH or LiH deposited on a molybdenum ribbon was heated up to ca. 1000xa0K either stepwise at ca. 10xa0K intervals or continuously at a constant rate (ca. 2–20xa0K/s) in vacuum (ca. 10 −4 xa0Pa), and the desorption rates of H 2 , electron (e − ) and/or H − were measured mass spectrometrically as a function of the sample temperature ( T ), the introduced hydrogen gas pressure ( P H ) or the time ( t ) after a change in T or P H . Theoretical analysis of the data thus achieved yields the following results: (1) In the temperature-programmed desorption spectra observed with NaH, both e − and H − showed a single peak at ca. 800xa0K while a broad peak of H 2 appeared around ca. 750xa0K. (2) The activation energies ( E − and E 0 ) for the desorption of H − and H 2 from NaH were 172xa0±xa018 and 61xa0±xa07xa0kJ/mol, respectively, whilst the work function ( φ ) of NaH at those temperatures corresponding to the leading edge of an electron desorption peak was 261xa0±xa019xa0kJ/mol. (3) In the case of LiH, E − , E 0 and φ were 940xa0±xa089, 97xa0±xa012 and 747xa0±xa041xa0kJ/mol, respectively. (4) By the thermal dissociation such as LiH(solid)xa0→xa0Li(solid)xa0+xa0H 2 (gas)/2, φ was decreased by 20xa0kJ/mol or much more depending upon t or T , but the active spots (mainly Li) thus produced was destroyed by admission of H 2 up to ca. 10 −1 xa0Pa. (5) The deactivation (Lixa0+xa0H 2 /2xa0→xa0LiH) depending upon both T and P H was readily recovered (reactivated) after stopping the admission. (6) Our new method of monitoring e − was very convenient and useful for studying the thermal decomposition of saline hydrides.

Desorption energy of H− from heated saline hydrides and their work function effective for thermal electron emission

Abstract For the thermochemical study of negative hydride-ion desorption from powdery saline hydride (MH n , n = 1 or 2) deposited on a molybdenum ribbon heater, both negative-ionic and electronic desorption currents from MH n were measured simultaneously as a function of sample temperature (ca. 700–800 K) by using a mass spectrometer, thereby yielding the new data that the desorption energy ( E − ) of H − and the work function φ) of the desorbing surface of MH n are, respectively, 536 and 318 kJ mol −1 for LiH, 728 and 492 kJ mol −1 for CaH 2 , and 937 and 702 kJ mol −1 for SrH 2 . Our energy cycle, consisting of the four indirect processes equivalent in reaction energy to the direct desorption of H − from MH n , indicates that the theoretical values of E − — φ are 227, 224 and 227 kJ mol −1 for LiH, CaH 2 and SrH 2 , respectively. They are in good agreement with our respective experimental values of 218, 236 and 235 kJ mol −1 within the experimental error of ca. ±5%. This agreement indicates that the desorption of H − from MH n is explained reasonably well by our simple model based on chemical thermodynamics.

General applicability of our empirical formulae expressing the threshold temperature range for dissociative positive ionization of halide molecules on heated metal surfaces

Abstract The positive ionization efficiency of alkali halide (MX) incident upon Mo or Ta heated in a high vacuum (∼ 0.2 μTorr) always attained an appreciable value (∼ 0.01) at the appearance temperature (T0) and then steeply increased up to unity at the saturation temperature (T1). The data on the threshold temperature range (T0–T1) were analyzed to find the quantitative relation between each boundary temperature and the main factors governing it, thereby yielding the empirical formulae of T 0 = (D 0 + I 0 − φ 0 + ) R 0 k and T 1 = (D 1 + I 1 − φ 1 + ) R 1 k . Here, D, I and φ+ are the dissociation energy of MX, the ionization energy of M and the effective work function for the ionization on each surface, respectively, in eV at T0 or T1 in K, and k is Boltzmanns constant. The empirical constants (R0 = 42.1 ± 1.0 and R1 = 28.2 ± 1.1) thus determined are essentially equal to our previous values (41.7 ± 2.2 and 29.1 ± 1.4 for MX/Re and 41.5 ± 2.0 and 28.7 ± 1.1 for MX/W) and also present ones (39.5 ± 2.9 and 27.8 ± 1.1 for TlX/Re, InI/W, etc.) determined in high vacua (∼ 0.2–20 μTorr). These results indicate that our empirical formulae expressing the range are generally applicable to any diatomic halide molecule/metal surface system and also that the most probable values of R0 and R1 are 41.3 ± 2.2 and 28.4 ± 1.3, respectively.

Optimum temperature range for positive ion production from metal halide molecules incident upon heated metal catalysts

Abstract When the surface temperature (T) of catalytic metal (Re, W, Mo, Ta or Nb) was gradually increased in a high vacuum (∼10−5–10−3 Pa), the positive ionization efficiency (β+) of diatomic halide molecule YX (e.g., LiI, KF or TlCl) impinging with a constant flux (∼1012–1014 molecules cm−2 s−1) upon the surface steeply increased up to unity at the first boundary temperature (T1). Above T1, β+=1 continued until the second boundary temperature (T2), above which β+ decreased as T increased. Theoretical analysis of β+ around the optimum temperature range (T1−T2) yields the empirical formulae of T1=(D1+I1−φ1+)/R1k and T2=(φ2+−I2)/R2k. Here, D, I, and φ+ are the dissociation enthalpy of YX, the ionization enthalpy of Y and the effective work function for the ionization on each metal surface employed, respectively, in eV at T1 or T2 in K,and k is Boltzmanns constant. The empirical constants (R1=28.4±1.3 and R2=5.47±0.30) thus determined always hold for any system, irrespective of the difference in species of both sample and catalyst.

Effective work functions of polycrystalline refractory metals heated for thermal positive-ionic and electronic emissions

Abstract A molecular beam of diatomic metal halide (MX, e.g., LiCl, NaBr, KF, RbBr, CsF, TlCl or InI) was directed onto a polycrystalline surface of refractory metal (e.g., W or Re) heated in high vacua, and the ion current of M + emitted from the surface was measured as a function of either (1) surface temperature ( T ≈ 800–2300 K), (2) the elapsed time ( t ≈ 0–10 3 s) after having made the surface essentially clean, or (3) the introduced gas pressure ( P ≈ 10 −5 −10 −3 Pa) of air or oxygen, while the sample beam flux ( N ) incident upon the surface was kept constant in the range of 10 12 –10 14 molecules cm −2 s −1 . The data thus achieved were analyzed by our theory to determine the effective work function φ + ) for ion emission. In addition, the work function φ e ) for electron emission from each surface was measured under various conditions, thereby making it possible to determine the thermionic contrast ( Δφ∗ ≡ φ + − φ e ). The values of [φ + 0 ,φ e 0 and Δφ∗ 0 in kJ mol −1 ] determined with an essentially clean surface of each metal heated to a high temperature (usually above ca. 1800 K) in a readily attainable high vacuum (2 × 10 −5 Pa) are as follows; Nb [463 ± 10, 388 ± 5 and 75 ± 10], Mo [480 ± 6, 424 ± 4 and 56 ± 6], Ta [493 ± 5, 413 ± 5 and 80 ± 5], W [504 ± 5, 434 ± 6 and 70 ± 6], and Re [530 ± 5, 476 ± 5 and 54 ± 5]. With decreasing temperature (usually from ca. 1800 to 1400 K), both φ + and φ e become higher by up to ca. 100 kJ mol −1 than φ + 0 and φ e 0 , respectively, mainly owing to the adsorption of residual gases (especially of oxygen). However, Δφ∗ itself remained virtually constant at Δφ∗ 0 with little dependence upon T , t , P and N in the above respective ranges.

Thermal positive-ionic and electronic emissions from iridium heated in vacua

To clarify the thermionic property of a polycrystalline iridium filament surface heated in vacua, the emission current (I/sup +/) of positive ion (M/sup +/) produced from alkali halide molecule (MX) impinging upon the filament was measured as a function of surface temperature (T), incident sample beam flux (N) or residual gas pressure (P/sub r/). The current of thermal electron (e/sup -/) was also measured under the same conditions. Theoretical analysis of the experimental data thus obtained yields the following notable results and conclusions. 1) In a low temperature range (T/sub 1/ /spl les/ 1200 K), the ionization efficiency (/spl beta//sup +/) of MX decreases steeply with a decrease in T because the work function (/spl phi//sup +/) effective for the ionization is decreased by adsorption of MX. 2) In a middle temperature range (T/sub 1/-T/sub 2/ 1200-1300 K), /spl beta//sup +/=1 is attained, thereby yielding I/sup +/ /spl ap/ 10/sup -5/ A/cm/sup 2/ when N is 10/sup 14/ molecules/cm/sup 2/ s. 3) In a high temperature range (T/sub 3/ > 1500 K), the surface is kept virtually clean, and /spl phi//sup +/ is constant at 5.73/spl plusmn/0.03 eV while the work function (/spl phi//sup e/) effective for emitting e/sup -/ remains at 5.15/spl plusmn/0.03 eV. 4) As T decreases from T/sub 3/ to T/sub 2/, both /spl phi//sup +/ and /spl phi//sup e/ are increased by up to /spl sim/0.5 eV. 5) As P/sub r/ increases, T/sub 2/ and T/sub 3/ increase while T/sub 1/ decreases, clearly indicating that /spl phi//sup +/ is increased by adsorption of residual gases (especially of oxygen). 6) The thermionic contrast (/spl phi//sup +/-/spl phi//sup e/) is kept constant at 0.57 /spl plusmn/ 0.03 eV without depending upon T, N and P/sub r/. 7) Ir is useful for effectively ionizing those elements whose ionization energy is less than /spl sim/6 eV.

Activation energies for thermal ionic and neutral desorptions from thin films of lithium halides

Abstract To clarify the mechanism of positive-ionic and neutral desorptions from heated lithium halide (LiX, Xxa0=xa0F, Cl, Br or I), a small amount (approx. 10 −12 –10 −7 xa0mol) of LiX was deposited on a platinum plate (ca. 0.03–0.04xa0cm 2 ) to prepare a thin film ( θ 0 xa0=xa010 −1 –10 3 molecular layers at the start), and it was heated up to ca. 1500xa0K at a constant rate ( β xa0=xa00.4–140xa0K/s) in vacuum (approx. 10 −4 xa0Pa) using our dual-ion source system which made it possible to measure simultaneously the desorption rates ( D 0 and D + ) of neutral molecule (LiX 0 ) and ion (Li + ). The temperature-programmed desorption spectra thus obtained were different in pattern from that observed previously with NaX where each of NaX 0 and Na + had only one peak. Namely, the high peaks ( P 1 0 and P 2 + ) of LiX 0 and Li + appeared at a temperature generally lower and higher than the melting point ( T m ) of each LiX, respectively, while low peaks ( P 1 + , P 2 0 , P 3 0 and P 3 + ) appeared usually above T m . Theoretical analysis of the β -dependence of peak appearance temperatures yields the activation energies ( E 1 0+ – E 3 0+ ) for desorption of LiX 0 (or Li + ) giving P 1 0+ − P 3 0+ , respectively, and also the frequency factors ( ν 1 0+ – ν 2 0+ ) corresponding to respective peaks. With respect to LiF ( θ 0 xa0≈xa013 molecular layers), for example, E 1 0 and E 1 + were 220 and 167xa0kJ/mol, respectively, while ν 1 0 and ν 1 + were respectively 1xa0×xa010 17 and 8xa0×xa010 10 /s. In conclusion, (1) each desorption obeys the first-order kinetics, (2) P 1 0 − P 3 0 originate from the desorption from LiX at the state of physical adsorption, crystal or chemisorption, (3) P 1 + − P 3 + are due to the desorption from active sites (high work function sites; e.g., 724xa0kJ/mol for LiC1) on the heterogeneous surface of LiX itself or Pt, and (4) the ionization efficiency ( D + / D 0 ) even at P 1 + − P 3 + is usually less than 1% mainly because the fraction of the active sites is less than 1% of the desorbing surface area.

Experimental research for new materials usable as strong thermal electron emitters

Abstract To find new materials usable as strong electron sources, a powdery sample (30–350 mesh) of metal hydride (NaH, LiH, MgH2, CaH2, SrH2, TiH2, ZrH2 or LiAlH4) was deposited on a molybdenum ribbon, and the thermal electron current (J−) emitted from the sample in vacuum was measured as a function of: (1) the sample temperature (T), (2) the pressure (P) of introduced hydrogen, or (3) the time elapsing after a change in T or P. The amount of H2 desorbed from the sample was measured by a mass spectrometer. Theoretical analysis of the data yields the following results. (1) By ageing of LiH at 705 K for 5 h, for example, J− is increased 103 times, corresponding to a work function change of −0.5 eV. (2) This is due to the formation of active sites (Li) by thermal decomposition of LiH. (3) J− from LiH becomes strongest when the decomposition reaches ∼75%. (4) The sites are destroyed by admission of H2 beyond ∼10–5 Torr, but readily constructed again by stopping the admission. (5) NaH is smallest in work function (ϕ≃2 eV), but thermally too unstable to keep J− constant in time. Finally, (6) CaH2 is high in ϕ (∼5 eV), but best as a thermal electron source material among the eight hydrides because it is stronger (>1 mA cm−2) and more stable ( 100 h) against heating up to ∼1000 K.

Positive-ionic and neutral-molecular desorptions by temperature-programmed heating of a thin film of lithium bromide

Abstract To elucidate the mechanism of thermal desorptions of Li+ and LiBr from a thin film (0.2–2250 molecular layers) of lithium bromide deposited on a platinum surface, both ionic and neutral desorption rates were simultaneously measured as a function of heating rate (0.4–24 K/s). Theoretical analysis of the spectra thus obtained with the desorptions from heterogeneous surfaces with respect to work function shows that (1) desorption of LiBr exhibits three peaks corresponding to activation energies (E) of 139, 344 and 377 kJ/mol owing to desorptions from non-active sites (low work function spots) on the surface of LiBr or Pt, (2) desorption of Li+ shows two peaks (E=375 and 535 kJ/mol) due to desorption from active sites (high work function spots) on the inhomogeneous surface of LiBr itself or Pt, and (3) the fractional surface area of the active sites over each surface is as small as ∼10−2–10−3.