Volume 17, Issue 11, 01 November 1949
Index of content:
17(1949); http://dx.doi.org/10.1063/1.1747104View Description Hide Description
Deuterated diborane has been prepared by isotopic exchange and its infra‐red absorptionspectrum measured. The Teller‐Redlich product rule ratios lend support to the assignment based on the protonated double bond or bridge model. Redetermination of the Raman spectrum of normal diborane indicates several changes from previous assignments of Anderson and Burg and Wagner. The resulting vibrational assignment was used to calculate the thermodynamic properties of diborane in the range 100−1500°K. The ΔH f 0, ΔS f 0, and ΔF f 0 are also given. Appendix I shows that a quadratic potential term arises for the puckering of a planar ring if the ring is strained.
17(1949); http://dx.doi.org/10.1063/1.1747105View Description Hide Description
The infra‐red absorption spectra of the isomeric mono‐deutero propanes have been observed. A difference in the position of the primary and secondary C–C stretching frequencies is reported. This difference places the secondary C—D frequency 45 cm−1±5 cm−1 lower than the corresponding primary C—D vibration.
17(1949); http://dx.doi.org/10.1063/1.1747106View Description Hide Description
A study, by means of a mass spectrometer, has been made of the thermal decomposition products of hydrocarbons. The materials were pyrolyzed in closed ampoules which were subsequently opened to allow the products to be analyzed. A series of runs were made for cetane, n‐tetradecane and tetradecene−1 in which the pyrolysis time and temperature were independently varied. Data were obtained giving the compositions of the product gases as a function of the pyrolysis parameters. In examining a series of aromatic hydrocarbons it was found that the amount of yield depends primarily upon the number of substituted carbon atoms. A molar yield was defined and found to be a regular function of A/N where A is the number of substituted carbon atoms and N is the total number of carbon atoms per molecule. The pyrolysis residues were washed with iso‐octane which was then examined by ultraviolet absorption spectroscopy. It was found that the aromatic skeletons often remained intact throughout the cracking. In addition, it was found that these structures often polymerized or condensed, giving longer wave‐length absorption bands.
17(1949); http://dx.doi.org/10.1063/1.1747107View Description Hide Description
The method of assigning the ionic character of chemical bonds by the use of the dipole moment is examined for the case of hydrogen chloride. Using both the atomic orbital and the LCAO molecular orbital approximation, the ionic character of HCl is calculated. Not only do the values so calculated seem unreasonable from energy considerations, but the calculated rate of change of dipole moment with respect to internuclear distance does not agree with the experimentally observed infra‐red intensity. This discrepancy is due to the large dipole moment associated with the covalent bond. Although the results for a purely covalent compound are in agreement both with the dipole moment, and the infra‐red intensity, if the dipole is assumed to have the orientation Cl+H−, this assumption is discredited by the fact that one would have to assume more ionic character for HBr than HCl. If we make the assumption that s—p hybridization occurs, these difficulties are removed. It is shown that the dipole moment of the purely covalent molecule is a very sensitive function of the amount of hybridization, and the covalent molecule can have almost any reasonable dipole moment. Thus it is shown that there are difficulties which must be surmounted before one can use dipole moments to obtain ionic character.
17(1949); http://dx.doi.org/10.1063/1.1747108View Description Hide Description
A general solution for the secular equation is obtained, giving the force constants and normal coordinates, for any molecule, in terms of the fundamental vibration frequencies and molecular constants. The solution is based on the reduction of the secular determinant to Jacobiancanonical form; it is shown to have a simple physical interpretation. Methods are developed which enable adjustments to be made to the standard solution as required. The theory is fully illustrated by several worked examples.
17(1949); http://dx.doi.org/10.1063/1.1747109View Description Hide Description
Solid solutions in the system, NH4Cl–MnCl2–H2O, exhibit properties characteristic of order‐disorder in alloys. Disordered solidsolutions are indistinguishable in structure from that of ammonium chloride. Order is revealed by superstructure formation. The familiar compositions A 3 B and AB appear as end‐members of the ordered solid solution series. Crystal symmetry alteration from cubic to tetragonal results from ordering. Tetragonality is intimately related to the state of order. Though increase in ``solute'' content produces significant lattice distortion, superstructure unit cell volumes are constant.
Despite their unusual properties, the ordered salt solid solutions exhibit a normal Vegard's law relationship.
The salt system's features resemble those of a low temperature isotherm across Shockley's theoreticalphase diagram for alloy order‐disorder transformation and are regarded as experimental verification of Shockley's qualitative predictions. A concept of isothermal, compositional order‐disorder is used in presenting the results.
Demonstration of a two‐phase region, wherein ordered and disordered solidsolutions coexist in equilibrium, provides a basis for resolving the controversy over whether the phase rule governs order‐disorder transformations.
Identity of the order‐disorder phenomenon in metal and salt systems was further shown by producing the disordered state of salt solutions from the ordered by thermal and cold‐working methods. In both instances, order was spontaneously restored with aging at room temperature.
17(1949); http://dx.doi.org/10.1063/1.1747110View Description Hide Description
The absorption spectra of aluminum,lithium, and sodium borohydride have been obtained in the region from 1–25μ. The aluminum borohydride was investigated as gas and an analysis of the infra‐red and Raman bands fits a bridge‐type structure of symmetry D 3 (probably D 3h ). The spectra of lithium and sodium borohydrides were obtained from finely ground slurries of these materials. Their spectra indicate how the bridge structure of the H2BH2 metal groups passes over to the tetrahedral structure in a completely ionic lattic.
Using certain bond‐distance force‐constant relationships in conjunction with data from the diatomic hydrides and accepting electron and x‐ray diffraction data for bond distances between the heavier atoms, values for the geometrical parameters are calculated for beryllium,aluminum, and sodium borohydride. These seem more plausible than any so far suggested. Observations on the electronic spectrum of aluminum borohydride have also been made down to 1000A. The diffuseness of the absorption obtained at short wave‐lengths prevented any detailed interpretation.
The spectrum of a compound formed between beryllium borohydride and sodium chloride has also been obtained and a consideration of its absorption bands has given information concerning the nature of this material. A spectrum of a slurry of the methyl derivative of lithium borohydride (LiBH3CH3) indicates considerable change in the structure relative to that of lithium borohydride.
A table of frequencies of related bands in the metal and other borohydrides is given which illustrates the changes referred to above. It includes a number of boron‐nitrogen hydrides and shows that the BH frequency, when it is associated with analogous electronic groups, behaves in much the same way as that of CH when it is attached to saturated, olefinic or aromatic groups.
17(1949); http://dx.doi.org/10.1063/1.1747111View Description Hide Description
A procedure has been developed for the determination of molecular structure by electron diffraction which yields accurate intensity data and obviates the necessity for visual examination of the diffraction photographs. The theory for computing radial distribution curves has been extended to permit accurate curves to be obtained from scattering data covering only a restricted range of angle. From this method, it is possible to obtain not only equilibrium distances but also the probability distributions for the vibrational motion between pairs of atoms in a molecule. The procedure has been applied to CCl4 and CO2 and when comparisons may be made with spectroscopic results, satisfactory agreement is obtained.
Thermodynamic Functions for Tritium Deuteride. The Dissociation of Tritium Deuteride. Equilibria among the Isotopic Hydrogen Molecules17(1949); http://dx.doi.org/10.1063/1.1747113View Description Hide Description
Energy Levels and Thermodynamic Functions for Molecules with Internal Rotation. III. Compound Rotation17(1949); http://dx.doi.org/10.1063/1.1747114View Description Hide Description
A method is given for calculating the elements of the kinetic energy matrix for rotation for any molecule. The treatment includes the effects due to any number of linked rotating groups, balanced or unbalanced. In a simple case these equations reduce to the simpler ones of the two previous papers of this series. This rotational matrix is then converted into the matrix of the internal rotations. The reduced moments of inertia that form the latter are then used with the methods of the previous papers of this series to calculate energy levels and thermodynamic functions.
17(1949); http://dx.doi.org/10.1063/1.1747115View Description Hide Description
The characteristics of steady‐state one‐dimensional flames are expressed in terms of a set of first‐order ordinary differential equations suitable for solution by differential analyzers or high speed digital computing devices. Arbitrary systems of chemical kinetics and reaction rates can be investigated. The effect of ambient temperature, pressure, heat transfer from the flame to the flame holder, diffusion of free radicals, thermal conductivity, etc., are easily estimated. The equations which we use are the ordinary hydrodynamic equations of change generalized to include the effect of the chemical reactions. In these the usual expressions for reaction rates are introduced, that is, the rate at which the composition would change in a closed vessel under the local conditions of temperature and density. Equations expressing the diffusion velocities in terms of the composition gradients are given. The flame holder has been idealized in the form of a porous plug through which the fuel can pass freely from left to right, but a semipermeable membrane prevents the product gases moving in the opposite direction. It is found that heat transfer to this flame holder is required to stabilize the position of the flame. The conditions obtained at the hot boundary are expressed parametrically in terms of the roots of a secular equation.
17(1949); http://dx.doi.org/10.1063/1.1747116View Description Hide Description
The self‐diffusion coefficient of argon was measured at five different temperatures between −183°C and 53.5°C by observing the diffusion of radioactive A41 into normal argon. The decrease in the observed diffusion coefficient caused by the larger mass of A41 was deduced and found to be 0.6 percent. The experimental values are compared with several theoretical calculations. In particular, it is shown that use for argon of an inverse power model (V ∝ r −ν), with different values of ν ascribed to different temperature ranges, is not permissible since widely different values of ν at the same temperature can be obtained from different experimental data. The experimental figures agree very well with numerical calculations by Hirschfelder, Bird, and Spotz based on an interaction potential of the form 4E[—(r 0/r)6+(r 0/r)12].
17(1949); http://dx.doi.org/10.1063/1.1747117View Description Hide Description
The evidence pertaining to the heat of sublimation of graphite is summarized, with the conclusion that the value of 170.6 kcal. per mole is the most probable. This figure is consistent with a value of 393 kcal. for the heat of removing the four hydrogen atoms from methane, the assignment of the heats of the successive dissociations being as follows:The limits to the values for the second and third processes are determined on the basis of the kinetics of the reaction between sodium and methylene chloride. The methylene appearing above is in its lowest state, a singlet state in which the bonding to the hydrogen atoms is through p‐orbitals of atomic carbon, the bond containing a certain amount of ionic character. It is suggested that the formation of CH2 in a triplet excited state, in which the bonding to the hydrogen atoms is through sp hybrid orbitals of atomic carbon, corresponds to the following thermochemical values:From this suggestion it follows that the energy of excitation of the methylene ``molecule,'' the lowest state (singlet), to the methylene ``diradical,'' the triplet excited state, is at least 19 kcal. The evidence for the existence of the two states is discussed.
17(1949); http://dx.doi.org/10.1063/1.1747118View Description Hide Description
Hydrogen peroxide was identified as an intermediate product in the steady thermal combination of hydrogen and oxygen near 540°C by an examination of the absorptionspectrum of the reacting mixture. The absorption at a particular wave‐length (2537A) was then used in a quantitative study of the variation of hydrogen peroxide concentration in the reacting mixture as a function of conditions. The results obtained were compared with those obtained by chemical analysis of the gas mixture. The commonly accepted mechanism of the reaction proposed by Lewis and von Elbe is applied to the interpretation of the results obtained.
The Born‐Mayer Model for Ionic Solids and the Heats of Formation and Lattice Spacings of Alkali Halide Solid Solutions17(1949); http://dx.doi.org/10.1063/1.1747119View Description Hide Description
Previous attempts to calculate the heats accompanying formation of alkali halide solid solutions from the component salts have met with only limited success. The difficulties may have originated either in the use of Vegard's law to obtain lattice spacings of the solutions or in the general inadequacy of the model employed in computing the energies of the lattices concerned. Since earlier studies were based on the old Born concept of ionic solids, it was of interest to examine the subject again using as a basis the more refined treatment due to Born and Mayer. An expression for the potential energy of a crystalline solution has been developed in terms of the known interaction parameters of the pure constituents and has been used to compute the equilibrium spacings and heats of formation of three chosen systems. The calculated results are compared with experiment. The comparison suggests that further refinement in the model is perhaps necessary before the small changes in lattice spacings and heat contents which accompany solid solution formation can be completely accounted for.
Rotation‐Vibration Spectra of Diatomic and Simple Polyatomic Molecules with Long Absorbing Paths. II. The Spectrum of Carbon Monoxide below 1.2μ17(1949); http://dx.doi.org/10.1063/1.1747120View Description Hide Description
The third overtone of CO at 1.19μ has been photographed under high dispersion. From this band precise values for the rotational constants of the CO molecule are derived. They agree excellently with the values obtained from the ultraviolet spectra but are considered to be much more accurate. The equilibrium internuclear distance of CO in the ground state is re =1.12819×10−8 cm. An attempt to obtain also the fourth overtone failed even with path lengths of 3000 m atmos. The second overtone was observed under low resolution with a path length of 30 m.
17(1949); http://dx.doi.org/10.1063/1.1747121View Description Hide Description
Assuming the Urey‐Bradley field for the intramolecular potential, we have calculated the normal frequencies of long chain molecules in the extended form. Especially the vibration type of the frequencies remaining almost constant throughout the homologous series has been discussed. The result has been applied to the assignment of Raman lines of N‐paraffins and of the infra‐red absorption lines of polyethylene.
Thermodynamic Studies of Polyvinyl Acetate Solutions in the Dilute and Moderately Concentrated Range17(1949); http://dx.doi.org/10.1063/1.1747122View Description Hide Description
Osmotic pressure measurements on polyvinyl acetate solutions have provided values of the constant μ in methyl ethyl ketone, 1,2,3‐trichloropropane, acetone, dioxane, and dimethyl phthalate. Modifications in techniques have permitted measurements in the first two solvents at different temperatures at concentrations (c) up to 0.07 g/cc. The osmotic pressure can be expressed as a power series in c with three terms; the coefficients of c 2 and c 3 are larger in trichloropropane than in methyl ethyl ketone. The partial molal heats, entropies, and free energies of dilution in these two solvents have been calculated. All are numerically greater for trichloropropane. The heats of dilution are negative; is roughly constant for methyl ethyl ketone but increases somewhat with concentration for trichloropropane. The entropies of dilution are much smaller than those given by the lattice theory.
A Theory of Vapor Pressures of Liquids Based on van der Waals' Equation of State. II. Binary Mixtures17(1949); http://dx.doi.org/10.1063/1.1747123View Description Hide Description
An expression for vapor pressures of binary liquid systems obeying van der Waals' equation of state is developed by generalizing a recent theory of vapor pressures of pure liquids. The resulting partial vapor pressure equations involve composition and temperature, and contain no empirical quantities other than van der Waals' constants including an inter‐species interaction term.
The conditions under which Raoult's law holds for these systems are formulated, together with the conditions for which positive or negative deviations occur. The compositions at which azeotropes form are predicted, and the partial miscibility of binary liquid systems is discussed. A number of examples are calculated to illustrate the theory.