Index of content:
Volume 40, Issue 1, 01 January 1964
40(1964); http://dx.doi.org/10.1063/1.1724829View Description Hide Description
The instrumental factors that determine the resolution of an atomic beam apparatus in the measurement of total scattering cross sections are discussed. A definition of resolution as the minimum scattering angle for which the efficiency of detection of scattering is 50% is proposed.
40(1964); http://dx.doi.org/10.1063/1.1724893View Description Hide Description
Ion—molecule reactions in methane, in methane—methane‐d 4 mixture, and in ethane have been studied up to a pressure of 0.08 Torr in the ion source of the mass spectrometer.
Ionization‐efficiency curves, appearance potentials, and pressure‐dependence studies have allowed the parent—daughter ion relationship and the cross sections of the pertinent reactions to be determined.
40(1964); http://dx.doi.org/10.1063/1.1724847View Description Hide Description
Longitudinal ultrasonic absorption and velocity measurements were made in molten zinc chloride in the frequency range 5–95 Mc/sec. Shear relaxation measurements were obtained in the frequency range 43–118 Mc/sec. A single relaxation time was observed for the absorption of longitudinal waves at each temperature investigated. The absorption data demonstrated the existence of a volume viscosity which was found attributable to a structural single‐relaxation mechanism. Shear and structuralrelaxations were both present with the same single relaxation time. The activation energies for the shear and compressional flows are approximately equal. The variation of the relaxation time with temperature is in accordance with the rate theory. The shear modulus and the relaxational part of the compressional modulus increase exponentially with decreasing temperature. The temperature dependence of the former is predicted by the two‐state model of the hole theory. From numerous physical properties of molten zinc chloride and from the known crystalline state, a structural model is proposed for the melt. Two points are emphasized: (1) that there exist large ion complexes and (2) that there is a lack of network character in liquid ZnCl2. Based on this model, an explanation of the existence of distribution of relaxation times, observed in the glass‐forming network liquids, is offered which attributes the presence of a whole spectrum of relaxation times in these liquids to the cooperative nature of their flow process. The measured single relaxation time of molten zinc chloride owes its origin to its nonnetwork character.
40(1964); http://dx.doi.org/10.1063/1.1724888View Description Hide Description
Examination of transference number experiments in fused salt systems reveals that there is a nonspecific interaction with the walls of the containing cell. Attention is directed to the boundary layer on one side of which the electrolyte meets laminar boundary conditions and on the other side of which the bulk flow pattern prevails. A formal description of the phenomenon is obtained in terms of partial viscosities of each species. In fused salt systems the dimensions of the transition layer are estimated as being of the order of magnitude of the ionic radius so that the partial viscosities should be viewed as a device for treating the transition layer and not as being vested with hydrodynamic significance. In other electrolytes the boundary layer may be considerably larger.
40(1964); http://dx.doi.org/10.1063/1.1724889View Description Hide Description
Paramagnetic resonance data is presented for the systems MnCl2 in KCl, in LiCl, and in the LiCl/KCl eutectic over a broad temperature range (25° to 900°C), and a composition range from pure MnCl2 to 0.005M solutions. In the molten state hyperfine structure is observed at dilutions of 0.06M and below, whereas at higher concentrations only a single broad line is seen. The transition concentration (from hyperfine to no hyperfine splitting visible) is almost two orders of magnitude lower than the corresponding transition for aqueous manganese solutions. A preliminary discussion is given in terms of Anderson's theory.
40(1964); http://dx.doi.org/10.1063/1.1724890View Description Hide Description
In this work the relative photosensitivities H = | Δφ/F | i=const of a germaniumanode and a germaniumcathode were determined as a function of quantum energy (Δφ=electrode potential change caused by light, F = quantum flux, measured in quanta/cm2 sec, and i = current density of electrode). The first increase in the photosensitivity H (E ν) of a n‐Ge anode occurs at a quantum energy E ν≈0.93 eV, the second increase at E ν≈1.10 eV, and one relative sensitivity minimum at E ν≈1.27 eV. The sensitivity was approximately constant at larger quantum energies.
According to earlier experimental results, the slow phase in a dissolution reaction of Ge at a Geanode is the breaking of the bond. The form of the sensitivity spectrum can therefore be explained on this basis. The holes, which have an excess energy for activation, break the bond of a Ge atom to the lattice immediately and lead to a fast excess reaction and to a strong photosensitivity.
The photosensitivity spectrum of a p‐Ge cathode is of the same shape as that of a n‐Ge anode. One may therefore make a conclusion, that the photoeffect of a Gecathode is due to the fast dissolution reaction caused by holes with excess energy. The author proposes that this reaction is the dissolution of germanium as germane.
40(1964); http://dx.doi.org/10.1063/1.1724891View Description Hide Description
Theoretical potential curves, spectroscopic constants, and expectation values for some one‐electron operators are given for the 3Π u (1s2pΠ) (metastable) and 1Π u (1s2pΠ) states of the hydrogen molecule. Total energies of 0.7145 and 0.7334 a.u. are obtained for 1Π u and 3Π u , respectively. The experimental total energies are 0.7186 and 0.7381 a.u. for 1Π u and 3Π u , respectively. The maximum in the potential curve of the 1Π u state is found to be much smaller and to occur at a larger internuclear separation (0.02 eV at R≈8.5 a.u.) than estimated by previous theoretical calculations (0.1 eV at R≈4.5 a.u.). Our 1Π u curve is reasonably consistent with the observed data of Herzberg and Monfils.
40(1964); http://dx.doi.org/10.1063/1.1724892View Description Hide Description
The enthalpies of mixing in all the binary‐liquid alloys formed among the alkali metalssodium,potassium,rubidium, and cesium have been measured by reactioncalorimetry at 111°±2°C. All systems except rubidium—cesium show positive enthalpies of mixing. There is a tendency for the mixing enthalpies to become more positive and also more unsymmetrical as the two metals become more separated in the periodic table.
A comparison of the new enthalpy data with reported phase‐diagram information indicates that the three liquid alloys which contain sodium all exhibit nearly ideal entropies of mixing. For potassium—rubidium, potassium—cesium, and rubidium—cesium, the new heat data, combined with the available phase diagrams, permit estimates to be made of the excess free energies of mixing in the solid solutions.
40(1964); http://dx.doi.org/10.1063/1.1724894View Description Hide Description
The heat capacity of α‐NiSO4·6H2O has been measured over the temperature range between 1 and 20°K. The heat capacity passes through a rounded maximum of 1.51 cal deg—1 mole—1 at 2.58°K. The magnetic contribution to the heat capacity is well fitted by that calculated for the three spin states associated with the singly degenerate ground orbital level of the Ni++. From the heat‐capacity data the energy spacing between the lowest and middle spin states is calculated to be 4.48±0.07 cm—1 and between the lowest and highest spin states 5.05±0.07 cm—1. The magnetic susceptibilities in the directions of the three principal axes of a Ni++ ion are calculated and, by comparison with literature data on the magnetic anisotropy and single‐crystal magnetic susceptibilities, the orientations of these axes relative to the crystal axes are deduced. Data of other investigators on the magneto‐optical rotation and high field magnetization are interpreted.
Absolute Quantum Yield Measurements of the NO–O Reaction and Its Use as a Standard for Chemiluminescent Reactions40(1964); http://dx.doi.org/10.1063/1.1724895View Description Hide Description
The spectral distribution of the chemiluminescent reactions O+NO→NO2+hv has been determined over the wavelength region 3875–14 000 Å. The absolute rate constant was determined by a method of chemical actinometry which required only relative intensity measurements and which is virtually free from geometry errors. The rate constant over this spectral region was found to be 6.4×10—17 cm3 molecule—1 sec—1 within an accuracy of 30%. This reaction can be used as a standard from which the rate constants for other chemiluminescent reactions can be readily obtained without requiring detector calibration or geometry corrections.
40(1964); http://dx.doi.org/10.1063/1.1724896View Description Hide Description
The compressibility of pyrolytic graphite has been dynamically measured at pressures up to 0.5 megabar by utilizing explosive‐generated shock waves. Shock and free‐surface velocities normal and parallel to the original plane of graphite deposition were measured optically. The Rankine—Hugoniot shock‐wave equations were used to determine the compression points corresponding to the measuredvelocities. The measurements failed to distinguish statistically between the two major crystal directions in regard to dynamic compressibility characteristics. The measured pressure—volume data for pyrolytic graphite and the results of Alder and Christian on natural graphite differ significantly at pressures above 100 kilobars. Within the range of the present experiments no transition of pyrolytic graphite to diamond was found.
40(1964); http://dx.doi.org/10.1063/1.1724898View Description Hide Description
Pariser—Parr calculations were done for the singlet states of benzene using singly and doubly excited configurations. Good agreement is obtained with the experimental locations of the 1 B 2u , 1 B 1u , and 1 E 1u states as well as the oscillator strength of the 1 A 1g →1 E 1u transition. The 1 E 2g state is calculated to be 0.5 eV above the 1 E 1u state and its photodissociation can be deduced from its bond orders. A possible assignment of the 1 A 1g →1 E 2g transition is discussed. On the basis of this simple model, electro‐optical effects are small, such as the Stark shift for the 1 E 1u and 1 B 1u states.
40(1964); http://dx.doi.org/10.1063/1.1724899View Description Hide Description
Pair function equations have been derived exactly for a restricted class of trial functions containing only closed‐orbital pair correlations. In a matrix representation, it is shown that they reduce to a homogeneous pseudo‐eigenvalue equation. The results are equally applicable to any set of unitarily transformed set of Hartree—Fock orbitals. Comparisons are made with the Sinanoğlu scheme, and computational problems are discussed.
40(1964); http://dx.doi.org/10.1063/1.1724900View Description Hide Description
The electric conductance of methyl‐substituted benzenes has been determined between 20° and 65°C both in the presence and in the absence of oxygen over a range of electric fields (50 to 1200 V/cm). Five of the six compounds studied resembled benzene in their electric behavior. Although the conduction level decreased with increasing number of substituent methyl groups, the thermal activation energy, within experimental error, was the same and was similar to that of benzene. The behavior of 1,3,5‐trimethylbenzene was found to differ in all respects from the other members of this series. Its temperature dependence and over‐all conduction level were, however, closely similar to that of saturated hydrocarbons, such as cyclohexane or n‐hexane. The unusual behavior of this compound suggests that the available π electrons cannot participate in the conduction process. The steric arrangement of the methyl groups appears to significantly influence the conduction process. These findings have been interpreted to suggest the existence of an electron transfer mechanism involving sterically favorable molecular collisions.
Electric Conduction in Liquid Hydrocarbons. III. Comparison of Saturated and Unsaturated Hydrocarbons40(1964); http://dx.doi.org/10.1063/1.1724901View Description Hide Description
The electric conductance of aliphatic and olefinic type hydrocarbons has been determined between 25° and 65°C. both in the presence and absence of oxygen using electric fields of 50 to 1200 V/cm. Based on the observed potential distribution, the thermal activation energies, and the conductivities, it was concluded that conduction processes of the two types of compounds are distinctly different. The conduction in aliphatic hydrocarbons appears to be attributable to the presence of very low concentrations of polar impurities. Olefinic‐type hydrocarbons were found to have higher conductivities as structurally similar aliphatic compounds. Also, their thermal activation energies were consistently higher than those obtained for aliphatic compounds. This difference in behavior has been attributed to the presence of π electrons in the olefinic hydrocarbons. It is suggested that the conduction in unsaturated hydrocarbons involves an electron jumping process which is independent of the presence of trace impurities and is unavailable to saturated substances.
Influence of Atomic Masses on the Coriolis Coupling Coefficients in Some Symmetrical Molecules. I. Tetrahedral XY4‐Type Molecules40(1964); http://dx.doi.org/10.1063/1.1724902View Description Hide Description
Coriolis coupling coefficients (ζ values of tetrahedral XY4 molecules are studied in relation to the same quantities of the Z4 tetrahedron. The existence of upper and lower limits is proved for the ζ values of XY4. In particular, one has |ζ24|→1/2 and ζ44→1/2 as the upper limits with increasing atomic mass ratio m Y/m X. The same values, which are independent of the force constants, are found in the Z4 model, where the central atom is absent. Lower limits are also derived for |ζ24| and ζ44, but appear to be functions of the force constants.
Calculated ζ24 and ζ44 values for 36 molecules of the tetrahedral XY4 type are reported, and are studied as function of the atomic mass ratio. The regularities that have been observed are presented graphically.
Carbon tetrahalides, methanes, and silanes are the molecules which have been studied most extensively, and for which the applied force constants are reported.
40(1964); http://dx.doi.org/10.1063/1.1729851View Description Hide Description
It is proposed that the pyrolysis of ethylene oxide (EO) proceeds via an internal isomerization (similar to that in isoelectronic cyclopropane) to form an excited (CH3CHO)* molecule. It is estimated that the excess energy of this latter is about 85 kcal, which is about 7 kcal in excess of the C–C bond dissociation energy in CH3CHO. The lifetime of (CH3CHO)* for bond fission is estimated to be 10—8.5 sec so that its fate is predominantly quenching at P≥200 mm Hg. Estimates of the isomerization rate give, in good agreement with the inhibited reaction:where θ=4.575T kcal/mole.Analysis of the chain mechanism indicates a chain length between 1 and 2 with the production of a ``hot'' CH2CHO radical formed by exothermic ring opening of the EO radical. This is probably quenched but may decompose spontaneously to CH2CO+H. Published data on the high temperature reactions (1100°K) give absolute rates in reasonable accord with the 700°K reaction. The lower values of E act found for the higher temperatures may not be reliable.More meagre data previously published on the pyrolysis of propylene oxide are in good agreement with expectations drawn from the EO system both qualitatively and quantitatively. Other ``hot'' molecule reactions are discussed briefly.
40(1964); http://dx.doi.org/10.1063/1.1724843View Description Hide Description
The macroscopic state of a large, closed system is specified by the ensemble averages α j of a set of even dynamical variables Aj , together with the averages vj of the quantities iLAj , obtained by operating on the Aj with the self‐adjoint Liouville operator L. Phenomenological equations for the time rates of change, and v̇j , are derived by a technique due to Zwanzig, in which one operates on Liouville's equation with a projection operator which projects out the part which is ``relevant'' to the phenomenological description employed. These phenomenological equations are shown to exhibit the Onsager—Casimir reciprocity relations, including antisymmetry relations whose derivation is found to require a slight modification of Zwanzig's mathematical assumptions. Since , these equations also show that irreversible thermodynamics can be extended to the case where second‐order time derivatives appear representing memory effects, as well as nonlinear terms in the α j , provided the equations are still required to be linear in the vj . Furthermore, Onsager's equations are obtained while allowing the phenomenological matrices to be functions of the variables α j , and not merely of constants of the motion. This serves to generalize and extend Zwanzig's earlier treatment.
40(1964); http://dx.doi.org/10.1063/1.1724844View Description Hide Description
Atomic nitrogen, produced from very pure nitrogen in a microwavedischarge and detected by means of electron paramagnetic resonance, was increased from 2.0×1013 atoms/cm3 to 1.2×1015 atoms/cm3 by the addition of ≈5×1012 molecules/cm3 of SF6 to the gas before the discharge; the pink glow intensity also is increased several orders of magnitude and its time duration extended by ≈10. Nitric oxide and oxygen increase the number of atoms leaving the discharge to the same extent as SF6, but much larger amounts must be added. When nitric oxide is added between the discharge and the subsequent pink glow, comparable increases in atomic nitrogen concentration (≈20 times) are observed without the production of oxygen atoms.
40(1964); http://dx.doi.org/10.1063/1.1724845View Description Hide Description
The volatile products of the reaction between a molecular beam of room‐temperature chlorine and a polycrystallinenickelsurface have been identified; their formation rates have been measured by a mass‐spectrometric technique at surface temperatures between 1000° and 1600°K and beam fluxes equivalent to pressures of 10—7 to 10—5 mm Hg. The surface coverage is estimated to be less than 0.1% under these conditions. NiCl and NiCl2 are formed on the nickelsurface; the former is observed above 1450°, the latter below 1100°; and both are found at intermediate temperatures. About 70% of the incident chlorine is consumed by the interaction with nickel over the pressure and temperature ranges covered. The removal of chlorine and the production of NiCl and NiCl2 are first order in chlorine pressure. A mechanism which accounts satisfactorily for the results involves dissociative adsorption of chlorine to form two NiCl's on the surface; these either evaporate, or disproportionate to form NiCl2 which evaporates.