Volume 39, Issue 4, 15 August 1963
Index of content:
39(1963); http://dx.doi.org/10.1063/1.1734381View Description Hide Description
Methanephotolysis with light of wavelength down to 1050 Å has shown that product propane does not result from reaction of the ethane that is produced. The distribution of products is very similar to the distribution found in the radiolysis of methane. This result indicates that both photolysis and radiolysis of methane proceeds mainly by a free‐radical process and/or by reactions of excited molecules, rather than by reactions of ions. A mechanism involving insertion of the ethylidene radical is proposed to account for the results.
39(1963); http://dx.doi.org/10.1063/1.1734382View Description Hide Description
A theory is put forth to account for the dark conduction in cationic dyes, which assumes thermal excitation from ground state to conductiveexcited state, followed by trapping of the hole at the anion. It is applicable in principle to any case in which charge carriers are separated by an intrinsic thermal process which is followed by deep trapping of the minority carrier. Where such processes take place, measurements of thermally excited conduction will in general lead to one of three results when interpreted by simple equilibrium theory: the correct activation energy will be found, but the mobility inferred will be impossibly high; or, the mobility will be of the correct order of magnitude, but the activation energy will be too low; or both will be incorrect. The theory is applied to data for dark conduction in eight triphenylmethane dye halides; it is shown to account for the observed thermal activation energies in terms of optical energy levels, and to lead to mobilities which are similar and of acceptable magnitude.
39(1963); http://dx.doi.org/10.1063/1.1734383View Description Hide Description
By using the Hellmann—Feynman theorem in conjunction with an operator formulation of perturbation theory it is shown that, while for a general Hamiltonian an energy derivative is determined only to order n by an nth‐order wavefunction (Salem and Wilson), for a restricted class of systems it is determined to order n+1, and by restricting the class somewhat further one can obtain energy derivatives to order 2n+1. The example of the force between two atoms interacting through dispersion type forces is briefly discussed.
39(1963); http://dx.doi.org/10.1063/1.1734384View Description Hide Description
When hydrogen is dissolved in palladium, the α phase forms first; the β phase nucleates and grows in this. Because of the large volume change which accompanies the α‐to‐β transition, plastic deformation of the α phase occurs as it is stretched beyond its elastic limit. Hence, absorption is accompanied by growth of the β phase under compressive stress of the α phase, but desorption take place from the β phase of the plastically deformed solid. Since this is not a thermodynamically reversible process, hysteresis results. It is shown that the absorption loop of the isotherm can be calculated from the desorption loop and the yield strength of the metal. Some reinterpretation of earlier theories is presented.
39(1963); http://dx.doi.org/10.1063/1.1734385View Description Hide Description
The thermoelectric effect in single‐crystal cuprous oxide, measured at temperatures of 500° to 1130°C and oxygen pressuresP of 160 to 2×10—6 Torr, was found always to be positive, indicating that the material is a p‐type semiconductor in these ranges of conditions.
Plots of the thermoelectric power Q vs 1/T at different constant oxygen pressures yield parallel straight lines with a slope of 0.75 V. Plots of Q vs logP are mostly straight lines with slopes approximating (—2.3k) / (8q), in which k is Boltzmann's constant and q is the electronic charge.
These results can be explained fairly satisfactorily in terms of a modified Wagner model for the defect structure of Cu2O and existing theories of the thermoelectric effect in semiconductors. It is assumed that the concentration of acceptors is a function of both the oxygen pressure and the temperature and that a gradient of acceptors exists along the temperature gradient. Thus, the model shows that the factor 8, originally obtained from conductivitymeasurements, should also be contained in the slope of Q vs logP. The slope of 0.75 V for the plot of Q vs 1/T can be related to the energy of creating an acceptor, i.e., a copper atom vacancy, plus the energy of ionizing this acceptor.
39(1963); http://dx.doi.org/10.1063/1.1734386View Description Hide Description
A wavefunction similar to the Kinoshita function has been analyzed into natural orbitals. As it was for the hydrogen molecule, the first natural orbital for helium is very nearly the SCF function. The first three natural orbitals describe the atom quite well. These three terms give an energy of —2.8974 hartrees and an overlap with the exact wavefunction of 0.99979. Beyond these three terms, the natural expansion converges slowly with many terms required to even approximately fit the cusp in the correlation hole.
39(1963); http://dx.doi.org/10.1063/1.1734387View Description Hide Description
Detailed absorption spectra of crystals containing cobalt (II) in octahedral or slightly distorted octahedral sites are reported. Prime consideration has been given to the observation of the doublet states and the assignments are discussed in relation to a four‐parameter crystal field theory (Dq, ζ, B, and C).
39(1963); http://dx.doi.org/10.1063/1.1734388View Description Hide Description
The spectra of eight crystals containing trivalent chromium have been measured, and the energy levels fitted to a four‐parameter crystal‐field theory. The values of the parameters derived from the spectra are discussed, and some mechanisms responsible for absorption intensities are evaluated.
39(1963); http://dx.doi.org/10.1063/1.1734389View Description Hide Description
Absolute quantum yields of phosphorescence of biacetyl have been obtained in fluid and rigid solutions. In fluid solutions, the measurements were made relative to the quinine bisulfate standard and in rigid media, relative to benzophenone. The values obtained have been taken together with the known triplet‐state lifetimes of biacetyl in these media to calculate the radiative lifetime of triplet biacetyl. These results compare favorably with the radiative lifetime of biacetyl triplet in the gas phase, calculated in a similar manner using available data on the phosphorescence yield and the actual mean life in the vapor.
39(1963); http://dx.doi.org/10.1063/1.1734390View Description Hide Description
Absorption coefficients of xenon and argon have been measured at wavelengths between the ionization threshold (Xe, 1023 Å; Ar, 787 Å) and 600 Å. Similar measurements have recently been reported by us for krypton [Appl. Opt., September 1963]. A photoelectric scanning technique with a Hopfield helium continuum source as background was used at a bandwidth of 0.5 Å. The oscillator strengths of the auto‐ionized lines which belong to the two overlapping series p 6(1 S 0) —p 5(2 P ½)ns′ and p 6(1 S 0) —p 5(2 P ½)nd′ and which occur between the and 2 P ½ionization limits were calculated from the measured absorption coefficient curve. The total oscillator strength was 0.60±25% for the first six Xe peaks and 0.011±25% for the first four Ar peaks. For xenon, the k value is 2100 cm—1±10% at the 2 P ½ threshold, and it drops smoothly to 1050 cm—1 at 600 Å. For argon, the k value is 930±10% at the 2 P ½ threshold, and the curve rises smoothly to 1200 cm—1 at 600 Å. In both cases, no lines or thresholds were observed at wavelengths shorter than the 2 P ½ threshold. Comparison is made with ionization efficiency data.
39(1963); http://dx.doi.org/10.1063/1.1734391View Description Hide Description
Absorption coefficients of molecular nitrogen have been measured with a photoelectric scanning technique using an intense, stabilized Hopfield helium‐continuum light source as background. These measurements differ from previous investigations where line sources were used as background. The measurements have been done with an instrumental bandwidth of 0.5 Å over a pressure range of 0.040 to 0.17 mm Hg using an absorption cell placed behind the exit slit of a vacuum monochromator. Corrections were made for the pressure gradient in the flow‐absorption cell, for the scattered light, and for fluorescence. From 1000 to 796 Å, the first ionization threshold (N2 +, X 2Σ g +), a number of sharp, intense bands are present with no evidence of a dissociation continuum. Below 796 Å, the ionization continuum appears to rise slowly to 400 cm—1 at 742 Å (N2 +, A 2II u ), but this region is also crowded with many intense bands. Most of these appear to be strongly preionized. Between 742 Å and the threshold of the B 2Σ u + state at 661 Å, the ionization continuum increases from roughly 400 to 780 cm—1, and the pre‐ionized Hopfield Rydberg series is the main feature. At 661 Å, the absorption coefficient rises abruptly to 980 cm—1 and continues at approximately this constant value to 580 Å. The present results are compared whenever possible with previous measurements and discussed in relation to photo‐ionization measurements. Application of this data to calculation of absorption processes in the upper atmosphere is discussed.
39(1963); http://dx.doi.org/10.1063/1.1734392View Description Hide Description
The exchange of energy between a tungsten filament heated at 1710° to 2500°K and cold helium, neon, argon, nitrogen, and carbon dioxide was studied at gas pressures from 0.1 to 10 μ Hg. The exchange of energy was found to be directly proportional to the gas pressure in the 1‐ to 10‐μ range but essentially independent of the filament temperature. The results are interpreted in terms of a process in which the impinging gas molecules are first physically adsorbed, then heated until they gain sufficient thermal energy to overcome the attractive forces of the tungsten surface. On the average, the gas molecules then desorb at a critical temperature that is independent of the filament temperature. For the gases studied, the critical temperatures lie between 580° and 1070°K. Reaction efficiencies of oxygen and carbon dioxide were calculated from the irreversible change in the rate of power consumption of the filament produced by introduction of these gases.
39(1963); http://dx.doi.org/10.1063/1.1734393View Description Hide Description
The emission of a charge‐transfer complex excited by the two‐photon absorption of a laser beam has been investigated. The emission spectra and ratio of phosphorescence to fluorescence have been compared for laser and ultraviolet excitation and are the same indicating that all of the observed processes are the result of the excitation of the charge‐tranfer state.
39(1963); http://dx.doi.org/10.1063/1.1734394View Description Hide Description
Radicals formed by vapor pyrolysis of β, β′ dinaphthyl and diphenyldisulfide have been studied by electron spin resonance. In each case one of the radicals, identified as the arylsulfenyl radical, shows a large g‐value anisotropy in the spectrum. The observed anisotropies are interpreted in terms of the spin—orbit coupling which mixes two states of different orbital angular momentum. The spin densities in one of the states are calculated using a perturbation treatment, according to McLachlan, to allow for the difference in orbitals occupied by different sets of electrons in a nonclosed shell system. The energy differences of the two states are estimated using a molecular orbital approximation. The results indicate that in the ground state of the radical the unpaired electron is localized in a nonbonding orbital of the sulfur.
Correlation of Magnetothermodynamic Data with a Theoretical Model. Powdered CoSO4·7H2O over the Range 0 to 100 kG above 1.2°K39(1963); http://dx.doi.org/10.1063/1.1734395View Description Hide Description
Equations for calculating the magnetothermodynamic properties, under varied conditions of field and temperature, are presented and discussed for a model based on a theory of paramagnetism in the solid state. An averaging technique, making use of a digital computer, is derived for use in evaluating the theoreticalequations for randomly oriented crystals of paramagnetic material. These equations are applied to the magnetic system of randomly oriented crystals of CoSO4·7H2O. It was shown previously that after subtracting a small temperature independent term from the measured isothermal differential magnetic susceptibility and magnetization, the data followed an H/T function, where H is the magnetic field within the crystals. (The experimental magnetic entropy measurements were made in the liquid‐helium range of temperature. Magnetic induction measurements were made in the liquid‐helium and liquid‐hydrogentemperature ranges, and at the boiling point of liquid nitrogen.) Since CoSO4·7H2O has two different Co++ ions per unit cell, seven parameters are theoretically necessary to fit the experimental magnetothermodynamic data. These parameters are: the ``fictitious'' spin of the ground state of the Co++ ions, S′0, and the principal g values for the tensors for each of the two Co++ ions per unit cell. Lacking definitive paramagnetic resonance data for the principal values of the tensors, the assumption was used that (1) both Co++ ions per unit cell have the same tensor, and (2) two of the principal g values for the tensor are equal. Using S′0=½ and deriving the principal g values of g ⊥=1.754 and g ‖=7.528, all of the measuredtemperature‐dependent magnetothermodynamic properties could be fitted to the solid state model over the H/T range 0 to 60 000 G/deg K, the limit of the measurements, to within the estimated experimental accuracy.
39(1963); http://dx.doi.org/10.1063/1.1734396View Description Hide Description
Quadrupole splitting of deuteron magnetic resonance was measured for a single crystal of Ba(ClO3)2·D2O. Quadrupole coupling, asymmetry parameter, and principal axis directions of the field gradient tensor were determined at 20° and at —130°C. eqQ/h=121.5±0.4 kc/sec, η=0.976±0.007 at 20°C, and eqQ/h=243.5±1.5 kc/sec, η=0.074±0.002 at —130°C. The field gradient at room temperature shows a rapid reorientation of the water molecule in a crystal. At room temperature, the spectral line exhibits fine structure due to coupling of two spins, which was utilized to determine the sign of eqQ in the present case.
The line broadening effect in the intermediate temperature region (—45°∼—121°C) was measured and was used to obtain the reorientation rate at various temperatures. The barrier to reorientation of 6.6 kcal/mole was obtained from the comparison of the experimentally determined reorientation rate with that calculated from a tunneling process through a hindering barrier, or that of 6.5 kcal/mole from classical process. The temperature dependence of quadrupole splitting was also measured and the oscillation frequencies were derived for the torsional motion of D2O. It was found that the quadrupole coupling for a vibrationless molecule is about 12% less than that of a free molecule.
The observed eqQ, principal axis directions of field‐gradient tensor are discussed in connection with the disposition of atoms in the crystal. A qualitative explanation is given for the change of eqQ from a free molecule to that in the crystal as due to the effect of the hydrogen‐bonded oxygen atom.
Photo‐Ionization Efficiency Curves. Measurement of Ionization Potentials and Interpretation of Fine Structure39(1963); http://dx.doi.org/10.1063/1.1734397View Description Hide Description
A method is described for measuring photo‐ionization efficiency (PE) curves over the energy range 9–14 eV. The method is experimentally easier than measuring the photo‐ionization cross section and gives essentially the same result. PE curves are given for Xe, O2, and NO and first ionization potentials are determined; Xe, 12.129±0.002 eV; O2, 12.065±0.003 eV; and NO, 9.250±0.005 eV. It has been established that steps in the PE curve of NO give a quantitative measure of the energies of the vibrational levels in the NO+ ion and that peaks in the curves of all three gases correspond to neutral levels that auto‐ionize. The probability laws for ionization and auto‐ionization are discussed.
Electrophilic Character of the Hg 6(3 P 1) Atom and the Mechanism of Electronic Energy Transfer in Mercury Photosensitization39(1963); http://dx.doi.org/10.1063/1.1734398View Description Hide Description
It is shown that in mercury‐6(3 P 1)‐photosensitized reactions the Hg 6(3 P 1) atom behaves as a distinctly electrophilic reactant in most of its reactions and that the H/D isotope effect exhibited by some paraffin molecules with respect to the quenching reaction can be satisfactorily estimated from transition‐state theory. This interpretation of the data, taken in conjunction with the results of some recent primary process studies, point up the chemical nature of electronic energy‐transfer processes in these systems in that the transfer reaction proceeds through the formation of a well‐defined transition complex involving both substrate molecule and sensitizer atom.
39(1963); http://dx.doi.org/10.1063/1.1734399View Description Hide Description
The cross sections for charge transfer between He+ions and molecules of nitrogen and oxygen have been determined within the energy range 100–7700 eV. A mass spectrometer was used to analyze the secondary ions, and the dissociative and nondissociative charge‐transfer cross sections were separately determined. The observations show that in both reactions the major contribution to the total charge‐transfer cross section comes from collisions involving dissociation.
39(1963); http://dx.doi.org/10.1063/1.1734400View Description Hide Description
The formation of negative ions in SbCl3 by low‐energy electrons has been investigated using a modification of a total ionization tube which allows the ion species to be determined and prevents contact potentials from affecting the electron energy in the collision region. The identification of the specific reaction taking place was obtained by modifying the total ionization tube to operate as a magnetron beyond cutoff. The mass of the ion species was then determined from an analysis of the ion current to the collector as a function of the applied axial magnetic field and the ion collector bias. The electron‐energy distribution had an effective width at half‐height of approximately 0.07 eV which was obtained using the retarding potential difference method. With the apparatus the most probable initial ion reaction was SbCl3+e→SbCl2+Cl—. The value of the dissociative capture cross section at the point of ion apparance was (1.8±0.3) × 10—16 cm2, and the minimum energy of the electron for the appearance of ions was approximately 0.01 eV at a gas temperature range 25°—50°C. The dissociation energy for the reaction SbCl3→SbCl2+Cl was 3.69±0.15 eV.