Index of content:
Volume 8, Issue 9, 01 September 1940
Studies of Equilibrium Solid Solutions in Ionic Lattices Systems: KMnO4–KClO4–H2O and NH4Cl–MnCl2–H2O8(1940); http://dx.doi.org/10.1063/1.1750735View Description Hide Description
1. The systems, potassium permanganate‐potassium perchlorate‐water and ammonium chloride‐manganous chloride‐water, have been studied at equilibrium conditions by phase rule and x‐ray diffraction methods. 2. The system potassium permanganate‐potassium perchlorate‐water yields a continuous series of solid solutions having orthorhombic crystal symmetry. For these, the components of which are of similar crystal and chemical structure and of identical valence type, Vegard's additivity law is followed by the a 0 and c 0lattice constants, but the b 0 constant shows a definite deviation. 3. A method is described of obtaining homogeneous crystals with a particle size which gives excellent powder diffraction photograms in cases where heat annealing is not feasible. 4. A method of analyzing for ClO4 — is given. 5. The danger of making serious errors is pointed out in the use of rapid precipitation by chemical reaction or from supersaturated solutions as a method of preparing solid solutions for the study of the relationship between their lattice constants and composition. 6. The system ammonium chloride‐manganous chloride‐water shows three solid solution series; the crystals obtained in the first series have cubic symmetry, while in the others the symmetry is tetragonal. 7. The mechanism of the formation of the ``anomalous'' solid solutions between ammonium chloride and manganous chloride is given. Experimental and other considerations verify it. 8. It is shown that Vegard's law is not followed by the first solid solution series in which the components are of dissimilar chemical and crystal structure and dissimilar valence types. The curve for the relationship between the lattice constants and composition rises to a maximum and then falls off. Reasons to explain why this law does not apply to these solid solutions are given. 9. The existence of a new ``compound,'' 6NH4Cl·MnCl2·2H2O, is demonstrated. This and the known compound, 2NH4Cl·MnCl2·‐2H2O, are considered to be examples of ``compounds of variable composition.'' They are tetragonal with a 0 = 15.256±0.004A, c 0 = 16.008±0.007A and a 0 = 7.5139±0.0005A, c 0 = 8.245±0.003A, respectively. The structure of the latter is that of 2NH4Cl·CuCl2·2H2O which belongs to space group D 4 14 h or P4mnm.
Application of the Theory of Absolute Reaction Rates to Heterogeneous Processes I. The Adsorption and Desorption of Gases8(1940); http://dx.doi.org/10.1063/1.1750736View Description Hide Description
By regarding the process of adsorption as involving a reaction between a molecule of gas and an adsorbing center on the solid surface, it has been possible, with the aid of the theory of absolute reaction rates, to derive for the rates of adsorption and desorption of gases simple equations which can be tested experimentally. Different results are obtained according as the adsorbed gas forms an immobile or a mobile layer on the surface: in the latter circumstance, which does not appear to be common, the rate of adsorption would be given by the Hertz‐Knudsen equation, provided there were no activation energy for adsorption. Combination of the expressions for the rates of adsorption and desorption gives an adsorption isotherm of the same form as that originally derived by Langmuir. Various cases of adsorption accompanied by dissociation of the adsorbed molecule are considered and the appropriate rate equations are deduced. If the molecule undergoes dissociation in the course of adsorption and the atoms remain on neighboring sites, these equations and the isotherm are the same as if there had been no dissociation, but if the molecule is adsorbed as such and then dissociation occurs as the result of an atom jumping from one adsorption center to another, the rate equations and isotherm differ from those obtained previously. The isotherm involves p ½ instead of p, but the form of the rate equation depends on whether the adsorption of the molecule or the jump of the atom is the slow stage. The possibility of interaction between adsorbed atoms or molecules in an immobile film on the surface is considered, and a modified isotherm which makes allowance for such interaction is derived. Equations have also been deduced for adsorption on covered surfaces.
Application of the Theory of Absolute Reaction Rates to Heterogeneous Processes II. Chemical Reactions on Surfaces8(1940); http://dx.doi.org/10.1063/1.1750737View Description Hide Description
The method developed in the preceding paper is used to derive equations for the rates of unimolecular and bimolecular heterogeneous gas reactions under various conditions of surface coverage by either reactants or poisons. The resulting equations are of the same form as those previously deduced by Langmuir and others, but they are more explicit; provided the activation energy is known they can be utilized to calculate rates of surface reactions which are in good agreement with the observed values. In the case of the dissociation of hydrogen on a tungstensurface, the simple assumption that the activated state consists of hydrogen atoms permits the absolute rate of the reaction to be calculated with an accuracy which is at least as good as that attained by direct experimental measurement. The factors responsible for the difference in rates of the same reaction taking place homogeneously or heterogeneously are considered; it is shown that adsorption of the activated complex lowers the over‐all activation energy, and this has the most important influence in favoring the surface reaction. If one of the products, or any other substance acting as a poison, is strongly adsorbed, the effective activation energy is increased, but there is some compensation resulting from the increase of entropy accompanying the desorption of the poison.
8(1940); http://dx.doi.org/10.1063/1.1750738View Description Hide Description
The potential energy surfaces for the reactions of two hydrogen atoms with another hydrogen‐like atom having a single valence electron in a 2p orbital are investigated. The simple Heitler‐London approximation is used in the calculations. The calculated activation energies for the reactions:where X is the atom having the 2p valence electron, are quite low, ∼1 Calorie. This is in good agreement with the experimental results for X = halogen, and the general principle that directed valence lowers activation energies. The symmetrical configurations for the H2X complex are found to have the lowest energy. In fact, a triangular complex was found to be stable with respect to dissociation into either HX+H or H2+X. The mechanism of the reactions between hydrogen and the halogens is discussed.
The Mechanism of Reactions Involving Excited Electronic States The Gaseous Reactions of the Alkali Metals and Halogens8(1940); http://dx.doi.org/10.1063/1.1750739View Description Hide Description
In many of the reactions of the alkali metals with the halogens the reacting system changes from a homopolar to a polar state, or vice versa. The possibility of a restriction on this change is considered, and it is concluded that there will be a serious restriction only for some of the atomic reactions which are of little importance. The crossing points between the polar and homopolar states for the reactions are drawn in so close that the interaction is large enough to insure transition. Due to the threefold degeneracy of the initial state of b (which arises because X is in a P state) this reaction has the possibility of leading directly to the excitation of the M atom which is formed. The nature of the potential surfaces for this process are considered. The process first suggested by Polanyi, the excitation of M atoms by vibrationally excited MX molecules, is discussed. This process is possible because of a crossing of the two lowest polar states for the M2X system. These two surfaces cross for a certain isosceles triangular configuration. Reacting systems thus easily go from the lowest to the second state in this region.
8(1940); http://dx.doi.org/10.1063/1.1750740View Description Hide Description
The Raman spectra of dibromochloro‐, bromo‐dichloro‐, dichlorofluoro‐, chlorodifluoro‐, dibromofluoro‐ and bromochlorofluoro‐methane have been studied. Their fundamental frequencies have been correlated and systematized. Simple graphical means were used to predict the Raman frequencies of CHBrF2 and CHF3. It was found that the C—X bond will show a ``characteristic'' frequency whenever the halogen X is the only and lightest of this type in a halomethane.
8(1940); http://dx.doi.org/10.1063/1.1750741View Description Hide Description
The absorptionspectrum of heavy benzene at 2700–2300A has been analyzed. As in light benzene it represents a forbidden transition 1 A 1g →B 2u 1 made allowed through the interaction of vibrations of type ε g +. Interpretations have been given for the majority of the bands. The measured bands together with the assignments are exhibited in Table I.
8(1940); http://dx.doi.org/10.1063/1.1750742View Description Hide Description
A method is developed for calculating thermodynamic functions for long chain molecules with particular attention to normal paraffins. In this connection the infinite chain approximation method for calculating vibration frequencies is considered. Starting with the results of Kirkwood, a modification is made which improves the agreement with the exact values for the simpler cases. In addition this method of attack is extended to out of plane motions. This vibrational analysis shows that all skeletal frequencies for molecules of the normal paraffin type can be put into two groups, one fairly narrow band near 1000 cm—1, and a broader band extending from 0 to 460 cm—1. The partition function is then set up on the assumption that motions in the low frequency group can be treated classically, and that the high frequency band can be replaced by a suitable number of 1000 cm—1 frequencies. Contributions from hydrogen atom vibrations are added on later. A formula is finally obtained which is quite simple, considering the complexity of the molecules. The calculated entropies can be brought into agreement with experimental values on the basis of very reasonable internal rotation restricting barriers.
8(1940); http://dx.doi.org/10.1063/1.1750807View Description Hide Description
A theory of depolymerization of long chain molecules is developed on a statistical basis. It is assumed that all bonds connecting monomeric elements in the system have the same probability of being broken regardless of their position in a given polymer and regardless of the size of the polymer in which they are found. Expressions are derived for the distribution of molecular sizes in the depolymerized system as a function of the initial chain length and the average number of bonds split per molecule. Also, relationships are established between the average molecular weight of the degraded product and the average number of bonds split per molecule. Experiments on the acetolytic degradation of cellulose acetate are briefly discussed.
8(1940); http://dx.doi.org/10.1063/1.1750808View Description Hide Description
The theory recently developed by Frank‐Kamenetzky for thermal explosions in which heat is removed by conduction only, has been applied to the azomethane, ethyl azide and methyl nitrate explosions. A fairly detailed discussion has been given of the experimental results and their relation to the theory. In the case of azomethane and ethyl azide the conclusion is reached that at low pressuresthermal conduction as contrasted to convection is an important, if not the exclusive, method of removal of heat from the reacting gas, with deviation appearing at high pressures. As was already concluded by Rice and Campbell, the results appear to indicate that the methyl nitrate explosion is not a thermal explosion. Certain criticisms directed by Frank‐Kamenetzky against the determination of the heat of reaction from the induction period, as carried out by Rice and Campbell, have been considered and refuted.
8(1940); http://dx.doi.org/10.1063/1.1750809View Description Hide Description
The reaction of hydrogen atoms, produced by the Wood‐Bonhoeffer method, with butane has been investigated over the temperature range 35° to 250°C. The activation energy is 9±1.5 kcal. The products consist solely of methane at low temperatures. At high temperatures ethane is also formed. It is concluded that the results indicate a mechanism in which a series of ``atomic cracking''reactions play the main role. The main steps in the postulated mechanism are: Primary process Secondary processes at low temperaturesAdditional secondary processes at higher temperatures
8(1940); http://dx.doi.org/10.1063/1.1750810View Description Hide Description
New electron diffraction photographs have been taken of NO2 extending the region previously investigated to include larger angles of scattering. An interference ring was found at (1/λ) sin ½θ = 0.49 followed by another ring appearing at 0.94 as determined by visual measurements. The outer portion of the pattern consists of two rather broad rings and two well‐defined minima. Theoretical intensities of scattering were computed for various nitrogen valence angles, assuming the positions of the two oxygens to be equivalent. The best fit, and probably the correct structure, gives the angle O–N–O = 130±2° with the N–O distance 1.21±.02A. Photographs were obtained from pure nitric acid vapor at 70°—85°C. The interference maxima were measured visually as far out as the eighth maximum at (1/λ) sin ½θ = 1.83; a prominent minimum was seen at 1.54. Theoretical intensities were computed for various likely models, disregarding the scattering by the hydrogen atom. Good agreement was obtained for a planar model having an NO2 group with the same structure found for nitrogen dioxide. The third oxygen atom O′ is located at a distance of 1.41±0.02A from the nitrogen atom and equidistant from the other oxygen atoms. A model having the nitrogen atom slightly out of the plane containing the oxygen atoms also gave good agreement with the experimental results. This model however is considered less probable in view of Raman spectra data which apparently require a planar structure for O′–NO2.