Vow. 16, 1930
`
`CHEMISTR Y: HILDEBRA ND A ND CARTER
`
`285
`
`'28.
`
`"Natural Radioactivity and the Origin of
`
`19 Olson, A. R., and Lewis, G. N.
`Nature, 121, 673-674.
`Species."
`'28. "The Rate of Induced Mutation in Relation to Dormancy,
`20 Stadler, L. J.
`Anat. Rec., 41, 97.
`(Abstr.)
`Temperature, and Dosage."
`Phys. Rev., 28, S II, 438-443.
`21 Terrill, H. M. '26. "The Energy of X-Rays."
`"Der Stand der Erzeugung von Genovaria-
`22 Timofeef-Ressovsky, N. W. '29.
`J. Psych. Neurol., 39, 432-437.
`tionen durch Rontgenbestrahlung."
`'29. "On the Concentration of Radium in Living Organisms."
`23 Vernadsky, V. I.
`C. R. de l'A cad. d. Sci. d. U. S. S. R., A2, 33-34.
`(Russian.)
`
`THE INFLUENCE ON THE IDEAL SOLUTION LAWS OF THE
`DISTRIBUTION OF POLARITY WITHIN THE MOLECULE
`BY J. H. HILDUBRAND AND J. M. CARTER
`DZPARTMZNT OF CHUMISTRY, UNIVURSITY OF CALIFORNIA
`Communicated March 5, 1930
`
`It has long been known that differences in the degree of polarity be-
`tween two molecular species tend to produce, in their solutions, deviations
`Recent progress in clarifying our ideas
`from the ideal solution laws.1
`concerning polarity, which we owe principally to Debye,2 has made it
`possible to measure the electric moments of molecules and to determine
`how they are influenced by the nature and arrangement of their parts,
`and it is now possible to test an opinion expressed earlier by the senior
`author' that to understand the effect of polarity on the solution laws it
`is often necessary to take account of the polarity of the parts of a molecule,
`and not simply the polarity of the molecule as a whole.
`There are only a few systems for which all the necessary data are at
`hand, and the best for our purpose appear to be the solutions of benzene
`with nitrobenzene, with the three dinitrobenzenes and with 1-3-5-trinitro-
`According to Williams4 the electric moment of benzene is very
`benzene.
`small, <0.1 X 10-18 e.s.u., that of nitrobenzene is 3.90 X 10-18, for the
`ortho-, meta- and para-dinitrobenzenes the moments are 6.05 X 10-18, 3.81
`X 10-18 and 0.32 X 10-18, respectively; for trinitrobenzene the moment is
`Solutions of nitrobenzene and benzene show deviations from
`1.08 X 10-18.
`Raoult's law (vide infra) in harmony with the difference between their
`electric moments; if, then, deviations from Raoult's law depend upon
`the electric moments of the molecule as a whole, we would expect very
`little deviation for solutions of benzene with para-dinitrobenzene, more
`for benzene with meta-dinitrobenzene, and most with benzene and ortho-
`If, on the other hand, the forces between two molecules
`dinitrobenzene.
`are determined chiefly by the polarity of the parts which are nearest
`
`NOVARTIS EXHIBIT 2404
`Roxane v. Novartis, IPR 2016-01102
`Page 1 of 4
`
`

`
`286
`
`CHEMISTRY: HILDEBRAND AND CARTER
`
`PRoc. N. A. S.
`
`to each other, there should be much less difference in the behavior of the
`three dinitrobenzenes toward benzene.
`We have measurements of solubility for the system benzene-nitrobenzene
`by Dahms,5 for benzene with the three dinitrobenzenes by Kremann.
`Since these substances have different melting points and heats of fusion
`we cannot directly compare their solubility curves, but we can make a
`comparison sufficiently accurate for our purpose by means of the equation
`used by the senior author7 to define a "regular" solution,
`
`(1)
`
`F2-IF,= bN
`where r2 is the partial molal free energy of component X2 in the regular
`solution, r, its partial molal free energy in an ideal solution, b, a constant
`which expresses the deviation from Raoult's law and Ni the mole fraction
`For regular solutions Eq. (1) is considered
`of the other component, Xi.
`to be independent of the temperature; for the solutions here under con-
`sideration this may not to be strictly true, but it may, nevertheless, be
`regarded as sufficiently approximate to serve for the rather rough com-
`parison we wish to make, and the data themselves indicate that this is
`the case. The free energy of transfer of a mole of X2 from the ideal to
`the actual solution is given by
`
`= RT ln(N'/N2);
`(2)
`02 -
`the ideal solubility, Ni, is calculated from the heat of fusion, Aii,~ and the
`absolute melting point, Tm, by the familiar equation:
`
`log
`
`-NHf (1
`2
`2.3R
`T
`
`(3)
`
`L )
`T
`Values for the heats of fusion and melting points of the several nitro-
`It is hardly
`benzenes were taken from Andrews, Lynn and Johnston.8
`necessary to tabulate the figures used and the results of the calculations,
`since the conclusion is sufficiently well indicated in figure 1 where r2 -
`is plotted against N . The number and positions of the nitro-groups
`have been indicated in a manner that is self-explanatory.
`It will be seen that a single straight line serves for all three dinitro-
`benzenes, which shows not only that their behavior is satisfactorily ex-
`pressed by Eq. (1) but that there is no significant difference between their
`behaviors toward benzene in spite of the great differences in their electric
`moments.
`The point for nitrobenzene represents the eutectic point and is doubtless
`sufficient to give the slope with satisfactory accuracy even though higher
`values of Ni are thereby excluded.
`It is a very striking fact that the slope
`
`NOVARTIS EXHIBIT 2404
`Roxane v. Novartis, IPR 2016-01102
`Page 2 of 4
`
`

`
`Voi. 16, 1930
`
`CHEMISTRY: HILDEBRAND AND CARTER
`
`287
`
`of this line is approximately half the slope of the line for the dinitrobenzenes.
`It will be remembered that this slope represents b in Eq. (1) and expresses
`the deviation of the solution from the ideal. We may say, accordingly,
`that the substitution of the nitro-group in the benzene ring causes a devia-
`tion from Raoult's law expressed by b = 320 cals., while the substitution
`of two nitro-groups gives rise to nearly twice this deviation, making b =
`580 cals. One must not, of course, attach too great significance to the
`apparent simplicity of this relationship unless it is later confirmed for
`other systems.
`It is to be hoped that data will be obtained with this end
`in view.
`There are data obtained by Desvergnes9 for the solubility in benzene
`of symmetrical trinitrobenzene.
`Unfortunately, we have no figure for
`its heat of fusion, but since its solubilities10 in ortho-dinitrobenzene, in
`
`1200
`
`C
`
`!
`
`Bllio~:REo
`
`A
`
`N12.
`FIGURE 1
`Deviations from Raoult's Law shown by solutions of benzene with
`various nitrobenzenes.
`meta-dinitrobenzene and in 2-4-6-dinitrotoluene are all practically iden-
`tical they may safely be assumed to be approximately ideal.
`Fixing the
`ideal solubility in this way the solubilities in benzene show a departure
`from Raoult's law, indicated in figure 1, which is approximately twice that
`for the dinitrobenzenes, corresponding to b = 1200 cals. as compared with
`580 cals. for the dinitrobenzenes and 320 cals. for nitrobenzene. The
`addition of the third nitro-group in the symmetrical arrangement ap-
`parently leaves so small a non-polar portion exposed that the unlikeness
`to benzene is more than proportionately increased, although the electric
`moment of the molecule is thereby much reduced.
`We may mention, further, that the three dinitrobenzenes, in spite of
`the differences in their electric moments, show mutual solubilities agreeing
`closely with Raoult's law.
`
`NOVARTIS EXHIBIT 2404
`Roxane v. Novartis, IPR 2016-01102
`Page 3 of 4
`
`

`
`288
`
`CHEMISTRY: BANCROFT AND BARNETT
`
`PNoc. N. A. S.
`
`The combined weight of this evidence indicates that it is the number and
`polarity of the substituent groups rather than the electric moment of the whole
`molecule which determines deviations from Raoult's law.
`We must not overlook, however, the fact that when the field of a polar
`bond is sufficiently buried within the molecule its influence upon solu-
`Thus, in spite of the polarity of the bond
`bility tends to disappear.
`between the carbon atom and the nitro-group, as shown in nitrobenzene
`and nitromethane, we might expect tetranitromethane to behave in its
`Again, stannic chloride is non-
`solutions as a substance of low polarity.
`polar but stannic fluoride, apparently on account of the smaller halogen
`atoms, is so polar as to form a high-melting solid.
`1 Cf. Rothmund, Z. physik. Chem., 26, 433 (1898).
`2 Cf. Debye, Polar Molecules, Chem. Catalog. Co., 1929.
`'Hildebrand, Solubility, Chem. Catalog. Co., 1924, Chap. 8.
`4Williams and Schwingel, J. Am. Chem. Soc., 50, 362 (1928).
`5 Dahms, Ann. d. Physik. u. Chemie, 54, 486 (1895).
`6 Kremann, Sitz. Akad. Wiss. Wien, 117, HIb, 569 (1908).
`7Hildebrand, Proc. Nat. Acad. Sci., 13,267 (1927); J. Am. Chem. Soc., 51, 66 (1929).
`8 Andrews, Lynn and Johnston, J. Am. Chem. Soc., 48, 1274 (1926).
`9 Desvergnes, Mon. Sci., 15, 149 (1925).
`10 Cf. Tables ann. int. des const. et donnes numeriques, 5, 136 (1922).
`
`PENTA VA LENT NITROGEN IN ORGANIC COMPOUNDS*
`By WILDIR D. BANCROFT AND C. E. BARNBTT
`BAKR LABORATORY OF CHZMISTRY, CORNELL, UNIVZRSITY
`Communicated February 24, 1930.
`In connection with some work that we are doing on the constitution
`of the proteins, it became essential to know in what cases nitrogen in
`organic compounds will or will not add on hydrogen chloride gas stoichio-
`metrically at room temperature and atmospheric pressure to give what is
`ordinarily called pentavalent nitrogen.
`The following generalizations seem to cover most cases:
`I. The tendency for a nitrogenous compound to react stoichio-
`metrically with hydrogen chloride is increased when hydrogen is replaced
`by an alkyl group and decreased when hydrogen is replaced by a phenyl
`group.
`Introduction of so-called negative radicals such as 0, C1, Br,
`II.
`NO2, etc., decreases the tendency of the nitrogenous compounds to react
`These radicals have most
`stoichiometrically with hydrogen chloride.
`effect when attached directly to the nitrogen.
`
`NOVARTIS EXHIBIT 2404
`Roxane v. Novartis, IPR 2016-01102
`Page 4 of 4