close

Вход

Забыли?

вход по аккаунту

?

ASYMMETRIC MOLECULES OF SIMPLE STRUCTURE

код для вставкиСкачать
This dissertation has been
microfilmed exactly as received
69-18,267
B E R R Y , Kenneth Lester, 1913A S Y M M E T R I C M O L E C U L E S O F SIMPLE STRUCTURE.
Yale University, PluD., 1940
Chemistry, general
—
U n iversity M icrofilm s, Inc., A n n Arbor, M ichigan
Copyright by
KENNETH LESTER BERRY
1969
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
ASYMMETRIC MOIECUEES OF SIMPLE STRUCTURE
By
r
i'*
&
Kenneth L. Berry
B.S., Juniata College, 1935
M.S., Rutgers University, 1937
A Dissertation Presented
to the Faculty of the Graduate School
of Yale University
in Candidacy for the Degree of Doctor of Philosophy
1940
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
ASYMMETRIC MOLECULES OF SIMPLE STRUCTORE
By
Kenneth I,, Berry
B.S., Juniata College, 1935
U.S., Rutgers University, 1937.
Efforts were made to prepare in an optically active state molecules con­
sisting of one asymmetric carbon atom having attached to it either monatomic
or cylindrically symmetrical groups, in order to provide experimental tests
of recent theories of optical rotatory power, particularly that of Kirkwood.
a-Bromopropionitrile was prepared in an optically active state.
Its
specific rotation was found to be higher than would be predicted by Kirk­
wood’s theory in its simplest form.
The melting point, boiling point, density, and near-ultraviolet absorp­
tion spectrum of this substance were determined, along with the refractive
and rotatory dispersions in the visible spectrum at several temperatures.
A procedure was developed for the preparation, in good yield, and puri­
fication of bromochlorofluoromethane.
Its melting point, boiling point,
density and refractive dispersion at several temperatures were determined.
The atomic refractivity of fluorine was calculated.
Attempts to resolve bromochlorofluoromethane were based upon addition
compound formation, vapor pressure difference of enantiomers due tc asymmetry
of solvent action, and optically selective adsorption on an active adsorbent.
Bromochlorofluoromethane was found to form a digitonide with which a partial
resolution of the former was effected.
Exploratory experiments were carried
out on the two other methods of resolution.
Several promising, optically
active adsorbents for bromochlorofluoromethane were found.
R e p ro d u c e d with perm ission of th e copyright owner. F urther reproduction prohibited without permission.
The author wishes to acknowledge
his appreciation of the aid and suggestions
given by Dr. Julian M. Sturtevant who di­
rected this research.
i
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
TABIE OF CONTENTS
Page
Introduction
Chapter I.
Chapter II .
Chapter III.
Theories of Optical Rotatory Power.
Bibliography for Chapter I.
The Preparation and Properties of Optically
Active a-Bromopropionitrile.
The Preparation of Optically Active
a-Bromopropionitrile.
The Melting Point of a-Bromopropionitrile.
The Density of a-Bromopropionitrile.
The Refractive Dispersion of a-Bromopropi onit rile.
The Rotatory Dispersion of a-Bromopropionitrile.
The Absorption Spectrum of a-Bromopropionitrile in the Near Ultraviolet.
Bibliography for Chapter II.
The Preparation, Properties and Attempted
Resolution of Bromochlorofluoromethane.
The Preparation of Bromochlorofluoro­
methane .
The Melting Point and Boiling Point of
Bromochlorofluoromethane.
The Density of Bromochlorofluoro­
methane .
The Refractive Dispersion of Bromo­
chlorofluoromethane .
The Attempted Resolution of Bromo­
chlorofluoromethane .
Bibliography for Chapter III.
Summary
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
1
37
40
41
52
64
67
76
86
88
90
92
99
100
103
106
125
129
INTRODUCTION
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
I N T R O D U C T IO N
In recent years considerable attention has been fo­
cused on the problem of finding a theoretical basis for
the phenomenon of optical activity.
Most theories lead
to conclusions amenable to experimental test.
Although a great number of optically active com­
pounds have been prepared and their rotations measured,
the majority are too complex in structure to be con­
sidered by modern theories at their present stage of
development.
Then too, in cases where simple molecules
have been prepared, the importance of getting complete
dispersion data was frequently not recognized, so that
there is a need for such data.
The object of this work was to attempt to prepare
optically active compounds possessing the simplest
possible structures and to measure their dispersion, in
order to test the validity of and compare the conclusions
derived by recent theories of optical rotatory power.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I .
THEORIES OF OPTICAL ROTATORY POWER
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
N O TA TIO N
/°
=
Density
M
=
Molecular weight
a
*
Angle of rotation
[a]
=
Specific rotation.
[aj
=
For pure compounds
— ~—
where
1
1
= length of tube in
decimeters
For compounds in solution,
T 1
Ia I
=
100
a
r n
la
L
or
1 Gcjo
where
G
=
100 a
— — —
1C
is the grams of solute per 100 g. of solution
of density
and
C is the grams of solute per 100
cc. of solution.
=
Specific rotation at wavelength
=
A
and t°.
[m ]J
=
Molecular rotation
* 1X50
<p
=
Ellipticity of an elliptically polarizedray
tan -1
where
=
“ al
ar
al
ar , a^ = amplitudes of right and left circularly
polarized components of an elliptically polarized ray.
In the special case where
ar = a^
the ray is plane
polarized.
=
Frequency
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
A
=
Wave length
c
=
Telocity of light
n
=
Index of refraction
=
Index of refraction for a right circularly polarized ray
nl
=
Index of refraction for a left circularly polarized ray
N
=
Avogadro* s number
=
Number of molecules per cubic centimeter
eo
=
Charge of electron
mo
=
Mass of electron
nr
N1
R e p ro d u c e d with perm ission of the copyright owner. F urther reproduction prohibited without permission.
THEORIES OF OPTICAL ROTATORY POWER
Many substances, including solids, liquids and gases, exhibit the
property of rotating the plane of polarization of a beam of plane polar­
ized light passing through them.
Such substances are said to be op­
tically active, or to possess optical rotatory power.
It has been
long recognized that this phenomenon is to be attributed to a dis­
symmetry of the arrangement, either of atcms or molecules in a crystal
lattice, as in the case of many optically active solids, or of atoms
within individual molecules, as in the case of all substances showing
optical rotatory power in the liquid or vapor phases.
(Throughout
this discussion the absence of an external electric or magnetic field
is assumed.)
This connection between optical rotatory power and
crystal or molecular dissymmetry is the cause of the tremendous
amount of knowledge which has been accumulated in this field, and of
the constantly recurring attempts to develop a sound theoretical
understanding of the subject.
This Chapter constitutes a brief re­
view of the theoretical aspects of optical activity.
In 1811 the French astronomer Arago^" discovered optical rotation
in quartz but did not clearly appreciate the nature of the phenomenon.
.2 ,3.4.5.6 .
ciou
- - - in a series of experiments carried out from 1812 to
1838 esteblished the simple laws of optical rotation and discovered
the phenomena of rotatory dispersion and optical superposition.
He
examined materials other than quartz and discovered optical rotatory
power in organic substances, nanely, in oils of turpentine, laurel
7
and lemon, and in alcoholic solutions of camphor . Another result
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without perm ission.
of one of his many classical experiments was the important discovery of
4
the inversion of cane sugar by acid hydrolysis .
Fresnel
8
advanced the universally accepted interpretation of the
phenomenon of optical rotation as being due to circular double re­
fraction.
A beam of plane polarized light can be regarded as the
vector sum of two circular vibrations of opposite sense and equal
amplitude, the phase relation being such that the circular vectors
meet on the direction of polarization.
In an optically active medium
the beam can be imagined to divide into its two circularly polarized
components which will travel with different velocities, the one falling
behind the other to an extent proportional to the length of the medium.
The emergent resultant of the two beams is plane polarized light but
has its plane of polarization differing from that of the incident beam.
Fresnel's equation is
a
where
1
=
IT 1
-■
.
(n-L - nr )
(l)
= length of the medium.
That an exceedingly small disturbance of ordinary refraction re­
sults in the production of optically activity of considerable degree
can be realized from the fact that specific rotations of about
100 °
are caused by a difference in refractive indices of the medium for
dextro and levo circularly polarized light of the order of magnitude
of
10"6 .
Just as optical activity is produced as an indirect effect of the
difference in velocity of propogation of right and left circularly po­
larized waves there is an indirect effect arising fran differential
R e p ro d u c e d with perm ission of the copyright owner. F urther reproduction prohibited without permission.
absorption of the two kinds of waves.
The influence of the medium in
causing the velocity differential is exerted on all light waves.
For
light of wave length falling within a region where the optically active
molecule absorbs, not only the velocities of the two circularly po­
larized rays but also, in certain instances, their respective absorp­
tion coefficients are different.
In the latter case the emergent ray
is net plane but elliptically polarized.
The phenomenon is known as
9
circular dichroism and was discovered by Cotton in 1895 .
Similar to
the Fresnel equation for rotation the ellipticity is given by
<t>
where
and
Xy
“
(K-l - Ky)
(2 )
are the absorption indices of the two circularly
polarized rays.
Cotton experimentally determined the course of rotatory dispersion
in a region of absorption and in certain cases found the rotation to
increase to a maximum near the absorption band, become zero, and then
increase in the opposite sense to a second maximum.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
6
+5
-
1
- 2.
yooo
£ooo
X
Sooo
A
Figure
1.
He correlated this anomaly with the circular dichroism observed in the
same wave length interval and found that the ellipticity was a maximum
at that wave length which corresponds to the point of inversion of the
rotatory dispersion curve.
This is known as the Cotton effect and is
represented in Figure 1.
The optical rotatory power of active molecules is the sum of
partial contributions each of which originates in a definite absorp­
tion band.
It often happens that the Cotton effect in a band is
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
partially masked by other rotatory contributions and that the observed
rotatory dispersion in the band doesn’t represent the true character
of the anomaly.
Accurate information concerning the rotatory contri­
bution of a band can be obtained through a study of its circular dichroism since some of the theories of rotatory power have produced
formulas xvhich are successful in quantitatively correlating dichroism
and rotation.
An absorption band may have no rotatory contribution;
the Cotton effect cannot be detected in such bands.
Following Fresnelfs postulation of the nature of optical activity,
Pasteur‘S
formulated his general principle of dissymmetry and demon­
strated it in his experiments on tartaric acid and the tartrates.
In
expressing his views as to the structure of tartaric acid Pasteur put
forward the idea of the irregular tetrahedron as the simplest illus­
tration of molecular dissymmetry.
His speculation also foreshadowed
the helical or spiral configuration which was later found in the
structure of crystalline tartaric acid11.
Pasteur’s theory was based
upon the fundamental postulate that molecules are three dimensional
figures, since molecular dissymmetry is impossible in two dimensional
figures.
V/hen the quadrivalency of carbon was clearly established,
the tetrahedral model was adopted almost immediately as a correct
representation of the carbon atom in space.
employed as a basis for explaining isomerism.
The model was quickly
The concept of the
asymmetric carbon atom as a necessary condition for optical isomerism
was introduced almost simultaneously in 1874 by le Bel
12
and van’t
Hoff13.
The theories of Crum Brown1^ and of Guye13 were early attempts
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
to obtain a quantitative relation between the optical activity of a
molecule containing a single asymmetric carbon atcm. and some parameters
of the molecule.
They proposed that the degree of asymmetry or the
magnitude of the rotation should be proportional to the product
(a - b)(b - c)(c - d)(a - c)(a - d)(b - d)
in which the quantities
a, b, c, d, are the masses of the radicals.
Thus when any two substituents are equal the rotation is zero and if
any one of the differences changes sign, the sign of rotation is
changed.
Some limited success was achieved in correlating empirical
data but the theories were quickly discredited when Walden
16 showed
that molecules containing two isomeric radicals are optically active
in spite of the equal masses of the two groups.
After the development of our present conceptions of molecular
structure, and of the connection between optical activity and molecular
asymmetry by Pasteur, le Bel, and van’t Hoff; and after the breakdown
of Crum Brown's and Guye's attempts quantitatively to relate activity
to structure, it was recognized that any complete theory of optical
rotatory power would have to be obtained by treating the subject from
an optical viewpoint.
Accordingly all theories which have superseded
those of Brown and Guye have their origin in the electromagnetic
theory of light.
One of the first to contribute to the application of the electromagnetic theory to the problem of optical activity was Drude 17 who
showed just what form of electromagnetic equations must hold in an
active medium in order to produce rotation.
However, a detailed
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
9.
molecular model had to be developed to show why the asymmetric molecule
should give rise to these particular equations.
Drude’s model from
which he obtained his equations pictures the medium as one containing
charged particles or ions which possess natural periods of vibration.
These particles, in order to respond differently to a dextro and a levo
circularly polarized ray, have to vibrate along a helicoidal path.
They are set into more or less violent vibration according as their
natural periods agree more or less closely with the periods of the
light vibrations which fall upon them.
This conception was used first
by Maxwell as the basis of a theory of anomalous dispersion in ab­
sorbing media, but was afterwards developed independently, as the
basis of a general theory of dispersion, by Sellmeier, von Helmholtz
18
and Ketteler. Drude’s formulas
for rotatory dispersion as modified
TQ
by Natanson-" are similar in form to those of Sellmeier, Helmholtz
and Ketteler for ordinary dispersion.
a
iH? I
2
C
=
* ■ >
v
2 --- L _2--- 1 2
^
■
O
t
-
'
O
p. _ ^ 2
P..}2+
P ^
.'.P—
(^ 2
2 ^ t ’p
2
According to Drude:*
„
D
i>
<t
r
^
_
-
V
TrV o -7 2
V-\T
p
v
c2
{^2_ ^ 2 }2+ V
(3)
where
D
is a constant,
^
head of an absorption band, and
dimensions of frequency.
is the frequency corresponding to the
'V
is a damping factor having the
More than one band must usually be taken in­
to account, the rotation of the simplest molecule thus being expressed
*
In this discussion of the theories of optical rotatory power we shall
in general restrict ourselves to a description of the physical pic­
ture involved and a presentation of the results obtained, without
going into the mathematical derivation of the results.
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
*2
as a sum of t e m s of the above type, each with its own
D,
v
•
and
values. Outside an absorption band the expression fox the total
o
rotation of the molecule can be simplified to
B o m 20 observed that Crude’s mathematical treatment does not correspond
to his model, and Kuhn
21
showed that Drude had neglected a factor which
would reduce the rotatory power of the model to zero.
However, Drude*s
equations, considered from an empirical point of view, are successful
in expressing the rotatory dispersion of active molecules, and are
used for that purpose.
In 1915 Parson*^ introduced the magneton concept of atomic struc­
ture.
He preferred to view the electron not as a charge of negative
electricity concentrated at a point but distributed uniformly around
a ring which rotates about its axis with high speed and therefore be­
haves as a small magnet.
This he called a magneton.
Allen 23 pro­
posed to replace Drude*s electron moving in a spiral path by a
magneton vibrating along a straight line.
On this basis he arrived
at an equation identical with that of Drude.
An attempt on the problem of the physical basis of optical
rotatory power was made by Stark in 191424. He applied his own
theory of valency in describing the mechanism of rotation of the
plane of polarization as follows.
The displacement of a valency
electron by the electric vector of a light wave is opposed by a
restitutional force directed, in an isotropic field, towards the
R e p ro d u c e d with perm ission of th e copyright owner. F urther reproduction prohibited without permission.
position of rest of the electron.
However, the field is not isotropic
since the restoring force acts along the direction of the chemical bond.
The combined action of the forces is not parallel to the electric
vector of the light wave, and the polarization of the molecule is
therefore slightly rotated from the direction of the electric field.
The average rotation for all random positions of the molecule is non­
vanishing only when there are at least four dissymmetrically arranged
electrons.
In this case, if two of the electrons are interchanged
the direction of rotation is reversed, and if two are identical the
molecule acquires a plane of symmetry and the rotation vanishes.
Stark’s theory was the first attempt to give an account of op­
tical activi+y on the basis of the chemist’s picture of molecular
structure.
It was entirely descriptive and no mathematical develop­
ment was ever attempted.
The mechanism is not found in more modern
theories because it is no longer regarded as essential that the re­
stitutions! forces acting on the electrons shall be anisotropic.
The influence of one moving electron on the motion of the others was
not considered by Stark.
In 1915 Born
25
, Oseen
26
, and Gray
27
independently and almost
simultaneously pointed out that the calculation of a molecular rota­
tory parameter had to take into account the finite ratio of molecular
dimensions to the wave length of the incident light, a factor neg­
lected in theories of ordinary dispersion; and the coupling of differ­
ent resonators in the molecule.
scribed as follows.
The latter phenomenon may be de­
YJhen the electric vector of a light wave inter­
acts with a molecule it produces a polarization in each of its com­
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
ponent units.
Each, of these polarized units then produces a field of
force which acts upon each of the other units.
The resultant polari­
zation of each is the sum of that produced by the external field and
the force fields of each of the other units.
Most of the recent theories are based upon coupled oscillator
models but have been developed in two different ways, as regards
treatment of the polarizability of the separate units of the molecule.
In those of Gray
27
, de Mallemann
28
, and Boys
29
no inquiry is made
into the nature or form of the function of polarization.
These
theories aim at producing a formula in which the rotation is an ex­
plicit function of the polarizabilities (obtained frcm molecular refractivities) of the component units of the molecule.
They are there­
fore directly applicable to available experimental data.
In Gray's treatment of the complicated interactions of the force
fields resulting frcm polarization of each of the groups of the asym­
metric molecule by the electric vector of the light wave the problem
reduced to the calculation of the net polarization in an incident
harmonic plane polarized field in which the molecules have a random
orientation.
The calculations were too complicated to lead to an
explicit formula for rotation in terms of the geometrical relations
of the molecule, but Gray was able to deduce the general, experi­
mentally observable facts of optical activity.
de Mallemann's model was that of an asymmetric carbon atom con­
taining four radicals each of which could be represented by an
ellipsoid with three principal axes of polarization at right angles
to one another.
The electrons of these radicals were assumed to be
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
held in their equilibrium position
by quasi-elastic forces like those
by which the radicals are attached to the central carbon at cm.
de
Mallemann’s simplifying assumptions were (a) that the four dissimilar
atoms, or groups in seme cases, can be treated as isotropic spheres,
of which the mean polarizability can be obtained from the measurable
atomic refractions of the atoms or groups; and, (b) that the edges of
the irregular tetrahedral model are determined by the dimensions of
the substituent radicals or atoms.
Kis treatment took into account
the direct mutual coupling of the four radicals.
On this basis he
found that the specific rotatory power of his model was given by the
formula
3
[a]
=
*
— (n2 + 2 )2
If
—
27
In this expression
A^, Ag, Ag, A^_
A^AgAgA^
f (a,b,c,p,q,r), --
(5)
are the reduced refractivities of
the four atoms as deduced from the molecular refractivities of gases
containing the atoms, by means of Newton*s formula
A
f (a,b,c,p,q,r)
=
(n2 - 1)
—
is a complicated function of the dimensions of the
molecule, of which
a, b, c
are coordinates, showing the distances
of the centers of the groups, taken three at a time, froaa the origin
of the rectilinear axes, and
p, q, r
clined edges of the tetrahedra.
are the lengths of the in­
de Mallemann calculated, by means
of this formula, the specific rotation of the then hypothetical com­
pound, CHBrClI
and obtained
=* - 3.2“ .
This value is probably
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
of the right order of magnitude hut has not yet been checked experimentally.
The compound was unknov/n until 1926 when Garino and Teofili
claimed to have succeeded in preparing it.
30
No attempt to resolve it
has ever been made, and indeed no compound containing only five atoms
has yet been resolved into optically active forms.
Boys* theory resembles that of de Mallemann in that optical rotatory
power is attributed to a mutual interaction between the radicals con­
tained on the asymmetric carbon atom.
His model consists of an irreg­
ular tetrahedron of four isotropic particles which may be atoms or
radicals.
All the tetrahedra are distributed at random so that the
medium as a whole is isotropic.
Under the action of the electric field of a light wave each par­
ticle becomes an oscillating electric doublet.
Inside the molecule
the electric field of the wave is altered by the fields of the doublets
themselves.
The sum of the actions of these induced force fields on
the electric vector of the light wave changes the plane of polarization
of the original wave, if the model is asymmetric.
Boys attempted to
calculate the magnitude of the polarizations of the individual mole­
cules, and of the medium composed of these molecules randomly oriented,
when the latter is expressed in terms of the electrical fields, it is
possible to find the velocity of a light wave in this medium by use of
Maxwell's equations.
The difference in velocities of dextro and levo
circularly polarized light can be calculated, leading directly to
specific rotations by the use of Fresnel’s equation (l).
Boys' ex­
pression for specific rotation is
[° J
’
i T n ^ T * » h3
W
o
W
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
<*>
R^, Rg, R^, Rp
are the refractivities of the radicals, deduced by means
of the Lorenz and Lorentz formula for molecular refractivities of liquids.
B
-
n-g- I
n2 + 2
M
/°
I is a complicated function of the radii of the spheres.
This formula
differs frcm that of de Mallemann (a) in the form of the model on which
it is based,
(b) in the use of refractivities derived from the formula
of Lorenz and Lorentz for liquids, and (c) in that magnetic as well as
electrostatic interactions are postulated, as a result of which the
factor (n^ + 2 )2 is replaced by (n^ + 2) (n2 + 5).
In order to compare calculated rotations with those observed experi­
mentally, Boys found it necessary to make two additional simplifying
assumptions.
He assumed that radicals such as CH, NHg*
C2^5»
as well as atcans such as the halogens can be treated as isotropic spheres
of known refractive index and radius, in close-packed contact with one
another.
This assumption may be justified, in the case of the simpler
radicals, CH, NHg, CHj, by the fact that they are isosteric with an atcm
of fluorine.
It is invalid in the case of more complex radicals which
cannot be either spherical or isotropic.
Boys also distributed the
volume and refractivity of the central asymmetric carbon atom equally
among the four surrounding groups.
Boys deduced the following values for the specific rotations of
four of the simplest asymmetric molecules.
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
Table 1.
[a]D (cal.)
act-Amyl aialne
C2H5 ^
CH3
C gH g
/ H
5.86°
4.0°
5.90°
7.4°
7.44*
H
^ c h 2ch
C gH g
H
°h 3
nh 2
C2H5
H
g h 3^
^(H
sec-Butylamine
sec-Butyl alcohol
3.6®
CHgNHg
act-Amyl alcohol
z&z
[a]D (obs.)
9.3®
13.9*
The factor I in. Boys formula contains the expression
(a-b)(a-c)(a-d)(b-c)(b-d)(c-d)
(a+b+c+d)^
where the terms are the radii of the four radicals.
This expression
allows the determination of the absolute configuration of the molecule
from its sign of rotation.
The convention is adopted of viewing the
molecule with the group A toward the observer and designating the re­
maining groups B, C, D in clockwise order.
When the size sequence
is
(a-b) (a-c), etc., have a
A ^ B ^ C > D
all the six differences
positive value, and the rotation is consequently dextro.
may be expressed as follows:
The rule
A compound is dextrorotatory when,
viewed with the largest group towards the observer, the arrangement
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
of the three other groups appears clockwise in the order of diminishing
volume.
The de Mallemann and Boys molecular theories of optical rotatory
power are best tested by using their formulae to predict the course of
rotatory dispersion curves of molecules which have been experimentally
investigated.
This test has the advantage of eliminating any errors
inherent in the hypotheses concerning the shape and size of the molecu­
lar models, since these are not functions of the wave length of the
incident light.
The portions of the formulae indicating the dispersion
are:
(n2 •+• 2)2 AjAgAgAi
--------------------- —
2
,,
>
Ide Mallemann)
(n2 + 2)(n2 + 5) RaRbRcRd
-----------------------------
(Boys)
X2
The values for this term are calculated for several wave lengths and
the ratios of the products are compared with the ratios of the observed
rotations for the same wave lengths.
The formulae are found to account
to a first approximation for the rotatory dispersion of those simple
molecules whose absorption bands are most remote from the visible
spectrum, outside the range of experimental observations.
Their in­
adequacy becomes apparent when applied to dispersion in a region of
absorption.
In the region of absorption, the rotatory dispersion is dominated
almost completely by the optically active absorption band.
As the
absorption band is penetrated, the anomaly in the curve of rotatory
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
dispersion is therefore often extremely large, and in some cases a
Cotton effect is observed, that is, the rotation passes through a zero
value near the middle of the band.
A zero rotation is deducible frcm
the formula of de Mallemann or of Boys only if one of the refraotivities
also passes through a zero value.
This phenomenon has never been ob­
served in organic groups and indeed, is impossible in chromophoric
radicals, such as
=
0,
since all the outer electrons of the
group contribute to its refraction.
Thus in the case of the carbonyl
group the major contribution to refraction appears to be made by elecO
trons with natural frequencies in the Schumann region at about 1200 A,
including probably the two unshared pairs of the oxygen atom and the
shared electrons of the two single bonds to the carbon atom.
The
shared electrons of the double bond, which are responsible for the
ketonic absorption band in,the ordinary ultra violet and hence the
major contribution to optical activity in the visible spectrum, con­
tribute only a part of the refractivity of the group and have no in­
fluence tending to produce zero refractivity.
These theories thus
place emphasis on the wrong absorption bands, and focus attention on
the fact that electrons, particularly those which produce the opti­
cally active absorption bands, must be considered rather than atoms
or groups.
In the electronic theories of Born
into the nature of the polarization.
25
°6
and 0seen~ inquiry is made
The polarizability is a definite
function of frequency, and a complete mathematical analysis of the
problem v/as undertaken on this basis.
The final formulae account
qualitatively for the known relations between optical activity and
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
molecular dissymmetry.
The rotation does not depend explicitly upon
the electric susceptibilities of the individual units but upon a large
number of constants, characteristic of the molecule, which are diffi­
cultly derived from known data.
Mathematical analyses of Born’s model were also made by L a n d e ^
and by Gans*^ .
Born himself attempted in 1935, to simplify his
original formula^.
Born’s newer model still consists of a system of four isotropic
vibrators but is simplified by assuming that these are situated at the
corners of a tetrahedron formed frcm four congruent triangles with
identical lengths of sides, a, b, and c
as in Figure
2.
/C
X
Figure 2.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
The two pairs of vibrators are perpendicular to each other and the model
has a binary axes of symmetry.
An active molecule of this type is di­
amino spiroheptane I.
H
HH,
C
CH /
\
2
CH /
2
/ CH?
X
/
CH.
By application of perturbation methods, Born obtained a third order
approximation to the interaction of the two pairs of resonators, as
well as the interactions of the two resonators in each pair.
Bom's
formula for such simplified models is
a = 3.49 x 10 "11 P *n
* 2)
M
t? ft
Xl A
2.3
<K-K'
'8
x - K
x 2- x
(7 )
* V
(?)
where
r ( g g - T
<?>
-
i6
d
=
=
a
=
d
>2
? ■
A
distance between the two pairs of vibrators in A.
distance between the vibrators in A.
rotation in degrees per decimeter.
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
X v
x
2
=
wave length in A corresponding to the natural vibrations
of the two different resonators (H and NHg in I).
f]_, fg
=
the intensity factors of the two vibrations, given by
e2
o
e2
o
o
m.o
(8 )
where
e-j_, eg, m-^, and mg
are the charges and masses of the two vibra­
tors.
Born calculated for diaminospiroheptane the rotation,
30° in general agreement with the observed rotatory power,
M
M
5402. =
5452
=
18° of its salts.
34
Oke
in 1935 showed that B o m ’s model containing four identical
mutually interacting vibrators could display optical activity if one
pair of vibrators is tv.lsted around the line of centers.
tem was shown to give five characteristic frequencies.
Such a sys­
The angle of
optical rotation was calculated in terms of these frequencies, the
edge of the tetrahedron, and the angle of twist of the pair of vibrators.
The fundamental principles of J. J. Thomson’s theory
to those of Bora’s theory.
are similar
A coupling between resonators is assumed,
and it is further postulated that the optically active molecule con­
tains two dissymmetric systems, one produced by a rigid tetrahedron of
the four groups attached to the asymmetric carbon atom, the other cor­
responding to the valence electrons of the four bonds.
Such a system
is shown to give rise to a rotation of the right order of magnitude
and to account also for the occurrence of rotatory dispersion.
A revievr of the facts of the phenomenon of optical activity which
must be explained by any valid theory was presented by Kuhn3® in 1929.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
They are as follows:
1.
Tresnel’s interpretation of the phenomenon must be ac­
counted for, that is, it must be demonstrated how the medium can ex­
hibit a difference of the order of
10“®
in its refractive indices for
right and left circularly polarized light rays.
2.
The refractive index for ordinary light is governed al­
most entirely by the very intense bands in the Schumann region.
The
weak ultraviolet bands which contribute so little to the refractive
index contribute still less to the circular-double refraction.
The
contribution of the bands in the nearer ultraviolet to the latter is
only a very small fraction of their contribution to the usual re­
fractive index.
3.
It follows from 1 and 2, that if the contributions of the
intense Schumann region bands to the circular double refraction were all
of the same sign and proportional to their intensities, the rotatory
power of a compound in the visible region would be several thousand
times larger than those observed.
Therefore ’’the relative difference
in the behavior towards right and left circularly polarized light in
the strong bands lying on the outer ultraviolet must often change sign
in such a way that the rotation at a great distance frcm these bands
disappears to a first approximation."
Observations show that the con­
tribution of an optically active absorption band to rotatory power is
not proportional to the intensity of the
for strong bands than for weak bands.
band but is usually -smaller
The rotatory power in the
visible spectrum is therefore often governed principally by the first
weak absorption band.
This accounts for the observation that, on
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
approaching shorter wave lengths, the rotation in general increases up
to the point of passing through the first, usually weak, absorption band.
Kuhn sought a theoretical basis for these facts deduced from ob­
servations by working out a special, simple case of Born’s theory of
coupled electronic vibrators.
His model contains two anisotropic
rectilinear oscillators; particle 1 (Figure 3) being bound elastically
m
Figure 3.
to its position of rest in such a way that it is able to oscillate only
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
in tlie direction
measured along
Ox.
Particle 2, at a distance
d
from particle
1
Oz, is similarly bound, and is able to oscillate only
in the direction
Oy.
A coupling force between the particles must be
postulated in order for the model to show optical activity.
Kuhn demonstrated that, if a plane polarized vibration is incident
on the coupled system from the negative
Oz
direction,
the right circular component will accelerate both particles, whereas
the left circular component would accelerate one and retard the other.
Vfhen the frequency of the incident vibi'ation differs from the two fre­
quencies of the system, this gives rise to a difference in the re­
fractive indices of the two circular components, and therefore to op­
tical activity.
When the frequency of the incident distrubance is the
same as one of the frequencies of the system it corresponds to a dif­
ference in the absorptive power for the two circular components and
thus to circular dichroism.
This common origin of rotatory power and dichroism suggests a
quantitative relation between the two effects.
The quantitative discussion of Kuhn’s model by classical methods,
closely similar to that followed in the theory of ordinary dispersion,
gives the result that the contribution to the optical rotation due to
an absorption bond
£
, whose frequency is
*= ~ —
Aq
and
whose half width, commonly called the damping factor, is
will be:
1
(nq - Hj.)
(9)
1
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
Expressed in wave lengths and angle of rotation, this equation simpli­
fied for large distances from the absorption band (i.e. in regions of
transparency) becomes
(a)
-
X?
-
where
x*
is a constant characteristic of the absorption band.
If
there are several active absorption bands, each band will give its con­
tribution and the observed rotation will be a sum:
2
y
a4
^
^
>?- X ?
Xl
no)
Formula (1 0 ) above is identical in form with Drude's equation (5) if
the constant
a^
is included in the Drude constant
D.
Kuhn has shown that the constant can be expressed as a linear
function of an "anisotropy factor"
(g) ,
-
where
gj_
defined as
v
<si> f.
5 i
g
(11)
K i
is the value of
g
^
for
j, and
^
1
and
r
are
the absorption coefficients of the two circular components of the vi­
bration.
For the constant
=
D
he arrives at the expression
N1
eo c
2 TT mQ
_
Si
5 1
*V I
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
where
f.,
is the intensity factor of an absorption band (for gases)
S l
defined by the Ketteler-Helmholtz type of equation
TTe§
/•
fydV
=
NX f
(IB)
m Qc
representing the absorption coefficient.
The integral is taken
over the absorption band in question.
Comparing Kuhn’s rotatory dispersion formulas with those of ordi­
nary refraction and absorption it is seen that the anisotropy factor
is the coefficient by which the contribution of one band to the ordi­
nary refraction has to be multiplied in order to obtain the rotatory
contribution of the same absorption band.
It can be deduced that
.
_
-v1!
f f 2 ga
v
2
or, in general, that
81
0
Therefore the sum of the Drude type numerators determining the con­
tributions to the rotatory power of the various absorption bands
taken over all visible and ultraviolet bands is zero.
The molecular
rotation therefore vanishes toward both ends of the spectrum.
The equations obtained so far for rotatory dispersion are based
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
upon the assumption that absorption curves can be represented by equa­
tions of the type (12).
They hold therefore only in regions of trans­
parency since Bielecki and H e n r i ^ have shown that equation (12) does
not represent accurately the shape of absorption curves of organic com­
pounds.
They proposed an exponential equation based upon a Maxwellian
distribution, which more satisfactorily expresses experimental results:
€
=
£
e
^
6
'
(13)
max
where
V *
=■ -— — —
1 «OOO
0
.
Based upon this Kuhn derived the following
equation for the molecular partial rotation of the band at
V
:
2
e A dx _
9
o
(14)
where
*-s
maximum, value of £ MJ , which, is reached for
- 0.9 0. At a distance from the absorption band the integral reduces to
6
— — 5----—
2 (v 0 - V
p- , and the equation takes the form of a simple
)
Drude term.
The most important contribution resulting frcci Kuhn's theoretical
considerations was the development of the first formulas satisfactory
in expressing the rotatory dispersion and circular dichroism inside an
absorption band.
Kuhn attempted to determine the absolute configuration of second33
ary butyl alcohol^ •
He assumed that (a) the CH3 and CgHg radicals,
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
as well as the asymmetric carbon aton, can be considered as isotropic
resonators;
(b) the QH group is anisotropic, for, as a second approxi­
mation, its anisotropy is a necessary condition for the appearance of
a rotatory contribution;
(c) there is a restricted rotation of the C2I
group around the C-0 axis; (d) the partial contribution of the first
absorption band is predominant.
On this basis, Eiuhn came to the conclusion that the carbinol xvith
the following configuration is levorotatory:
CH
CH j
A
i
f
I
I
H
Figure 4.
The dotted lines indicate the groups located below the plane of the
paper#
Measurements on a series of secondary carbinols have shown that
the sign of the first rotatory contribution of secondary carbinols of
the above type is opposite to the sense of the rotation observed in
the visible part of the spectrum; hence the above carbinol should be
dextrorotatory, a conclusion arrived at by the theory of Boys.
The attack by the methods of quantum mechanics upon the problems
of optical rotatory power commenced along two different lines of ad­
vance.
These two approaches are occupied (a) with the nature of op­
tically active molecules and (b) with the actual process of the ro-
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
tation of the plane of polarization.
The structure of optically active molecules was discussed in a
general, qualitative manner, especially with reference to the character
of their spectra, by Hund
studied by Rosenfeld
40
.
39
.
The process of optical rotation was
These two lines of advance did not unite,
however, until 1937 when Kirkwood^- and Condon, Altar, and Eyring^2,
almost simultaneously, brought out theories of optical rotatory power
employing quantum mechanical methods.
Both have as their starting
point the quantum-mechanical theory of dispersion as developed by
Rosenfeld.
His treatment includes a theory of the molecular rotatory
parameter, which is left to be explained by consideration of detailed
molecular models.
The one-electron theory of rotatory power of Condon, Altar and
Eyring is in striking contrast to the coupled oscillator theories of
Born, Oseen, Gray, Thomson, de Mallemann, Boys and Kuhn.
The approach
to the problem is gained by studying some simple potential field, in
which a single electron moves, of such a character as to give activity.
The force field must be such that it has no planes of symmetry and no
center of symmetry.
V
=
A field satisfying this requirement is given by
(kj^ x2 -t- kg y 2 f kjj z“")
+
A x y z
(15)
The cross-product term, is the one which produces the necessary dis­
symmetry.
An eauipotential surface of equation (15) is qualitatively
like what one would get if he took an ellipsoid of three unequal axes
and subjected it to a torsional stress.
It is shorn how fields of this type may be adapted to the descrip-
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
tion of the field of a molecule in which the chomophoric electron moves.
The constants
kp, kg, kg
and
A
are largely determined by the average
charge on the different atoms of the molecule as found from additivity
of dipole moments due to the bonds in the molecule.
The potential
energy associated with the movement of the chromophoric electron through
this field is calculated by the Hartree approximation.
The detailed
coupling of the electron with those producing the field, the effect
considered to the complete exclusion of other mechanisms in the coupled
oscillator theories, may be taken into account in higher approximations.
Both mechanisms undoubtedly contribute to the total effect.
In general, Condon, Altar and Eyring are able to show by considera­
tion of the one-electron theory,
(a) that the model shows optical ac­
tivity both classically and quantum mechanically; (b) that it accounts
for the order of magnitude of observed rotations; (c) that it leads to
a general quantum mechanical theory of circular dichroism; and (d) that
optical activity is quantitati- ely related to total dipole moment, this
last conclusion leads to a treatment of the solvent effect on optical
rotatory power.
The one-electron theory is illustrated in its application by Gorin,
Walter and Eyring 43 to the calculation of the optical activity and ab­
solute configuration of secondary butyl alcohol.
The problem was sim-
plified from the outset by accepting Kuhn’s44 conclusion from a survey
of the experimental rotatory dispersion data that in the case of
secondary butyl alcohol the sign and magnitude of the rotation in the
visible is controlled by the transitions (absorption bands) nearest
the visible which in this case are associated with the hydroxyl group.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
Attention is therefore centered on transitions in the hydroxyl group.
For a rigorous treatment a knowledge of the exact eigenfunctions for
secondary butyl alcohol in its normal and excited states is required.
These are not available so approximate models were used.
The chromophoric electron is considered to be one of the non­
bonding electrons of the oxygen atom since such electrons require the
least excitation energy for transitions to result.
The procedure consists in obtaining a rough solution to the wave
equation for the non-bonding oxygen electron moving in the Hartree
field of the other nuclei and electrons considered as charge distribu­
tions.
The fields of the other electrons are obtained with slight
modification frcsn the Slater type eigenfunctions for the various atoms.
The follov/ing absolute configuration is considered:
HO
C 2H5
Figure 5,
T ito different orientations of groups in this configuration are treated:
(a) the 0-H bond of the hydroxyl group is not coplanar with the plane
of the C* - 0 and C* - H bonds; and (b) the three bonds are coplanar.
In addition the case of free rotation of the 0-H group was considered.
For orientation (a) a theoretical value is obtained for the molecular
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
rotatory parameter, £ = - 2.70 x 10~*^, representing the contribution
of the hydroxyl group to rotatory power at the sodium D line wave
length.
The experimental value calculated from the observed rotatory
povrer by the expression
16
TT 3 N
3
is - 1.40 x 10
for levo-rotatory secondary butyl alcohol.
orientation (b) the theoretical result is
For
P = - 9.67 x 10”^ .
To
the configuration shown in Figure 5 these calculations assign the
levo form.
This happens to agree with Kuhn *s result, and disagrees
with that of Boys.
On the basis of free rotation of the hydroxyl the
calculations would assign the dextro form, to Figure 5.
At present the
evidence seems, however, to favor the postulation of restricted rotation.
Kirkwood
theory
44
41
started with the quantum, mechanical form of B o m ’s
and simplified it with the aid of certain approximations.
It leads to an expression for the rotatory parameter of an active
molecule in terms of the geometrical configuration and the polarizability tensors of its groups.
The latter describe the interaction
by pairs of the optically anisotropic groups.
The formula which
Kirkwood obtains is the quantum mechanical analog of Kuhn’s special­
ization of Born’s classical oscillator formula, equation (14).
How­
ever, anisotropic polarizable groups replace the hypothetical dis­
persion oscillators of Kahn and the coupling parameter is explicitly
calculated from the electrostatic interaction of the groups.
Kirk­
wood’s formula for the molecular rotatory parameter is
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
33.
N+l
g (o)
+
g (l)
+
gk
(16)
k=l
It should be noted that Kirkwood1s rotatory parameter
that of Condon, Altar and Eyring by the factor
gk
in simple molecules.
g^°^
.
TT
is the "internal gyration parameter” of group
g
k.
differs from,
In equation (16),
The
g^
vanish
is a sum over pairs of groups involving the
magnetic moments of the electronic centers of gravity of each group
relative to the molecular center of gravity,
g^1 ^
is a similar sum
involving the magnetic moments of each group relative to its own cen­
ter of gravity,
g^1 ^
vanishes if the retardation of the electro­
magnetic wave over all parts of each group is approximated by its
value at the center of the group.
rotatory parameter then reduces to
This approximation is made.
g^°^
The
which is a function of the
polarizability of the subsidiary groups and the geometry of the
molecule.
Like previous theorists Kirkwood chooses secondary butyl alcohol
as perhaps the simplest asymmetric molecule upon which to test his
theoretical deductions.
It is supposed that the groups II, QH and CKg
have optical axes of symmetry directed along their bonds to the
asymmetric carbon atom.
The optical axis of the ethyl group could be
assigned a direction along the C-C bond of the group but it is probable
that the remainder of the molecule tilts it toward the C*.
A reason­
able assignment for it is in the plane of the C* and C-C bond of the
ethyl group and inclined at one-half the tetrahedral angle to the
latter.
It is further assumed that two orientations of the ethyl
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
group are of equal probability:
(a) the C-C bond coplanar with C* - CH;
or (b) coplanar with C* - CH 3
The orientation of the C-C bond under the C* - H bond is of negligible
probability.
This doesn't mean that there is not free rotation of the
ethyl group through these positions in the usual sense of the terra
since the potential barriers hindering such rotation probably corres­
pond to low activation energies.
ing term in
(b)
from
g^°^
CgHg
In position (a) the only non-vanish­
arises from the
and
CH.
CgHg
and
CH 3
groups, in position
The mean rotatory parameters for these two
positions are calculated for two different directions of the optical
axis of
CoHn;.
For the orientation of this axis along the G-C bond in
the group the value of the rotatory parameter gives the following
specific rotation of secondary butyl alcohol for the sodium
length at
D
wave
20 °:
r
W
t
20°
d
=
2 1 -9°
For the orientation in which the optical axis of
CgHg
is assumed to
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
"be tilted 54° 4 4’ to the C-C bond
w
r
The observed value is 13.9®.
-
».=•
Since the -values are positive the con­
figuration shown in Figure 6 is dextro.
clusion of Boys.
This agrees with the con­
Probably the most serious defect in this calculation
lies in the treatment of the hydroxyl group as axially symmetric rela­
tive to the C-0 bond, since the hydroxyl hydrogen may produce same
anisotropy in the plane perpendicular to this bond.
The form of
g^°^
in equation (16) is such that this term van­
ishes for any pair of groups attached to the asymmetric carbon atom
which have optical axes of symmetry parallel to the bonds linking them
to the asymmetric carbon atom.
Such symmetry would result if (a) there
xvere free rotation of polyatomic groups about the bonds linking them
to the asymmetric carbon atom;
(b) the groups were monatomic; or (c)
they had cylindrical symmetry about an axis parallel to the bond
joining them to the asymmetric center.
In cases of such symmetry it
would be necessary to seek the rotatory power of the group in a higher
approximation of the polarizability theory which, Kirkwood says, is
usually quite small.
Thus any active molecule consisting of an
asymmetric carbon atom holding any four of the following groups
might be expected, on the basis of this theory, to have a vanishingly
small rotation.
H,
CH3 ,
Cl,
Br,
I,
F,
CN,
C=CH
Up to the time of the present work no molecule containing any combina­
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
tion of four of the above simple groups had been prepared in an opti­
cally active condition.
It was the object of the present research to
prepare such molecules and study their refractive and rotatory dis­
persion not only for the immediate purpose of testing Kirkwood's de­
ductions but to provide such data on ccmpounds with relatively simple,
fixed structure for further theoretical considerations.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
BIBLIOGRAIHY FOR CHAPTER I.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
1.
Arago,
Mem. Inst., I_, 93-134 (1811).
2 . Bi ot,
Ibid., I, 1-372 (1812).
3.
Biot,
Mem. Acad. Sci., _2, 41-136 (1817).
4.
Biot,
Ibid., 13, 39-175 (1835).
5.
Biot,
Ibid., 15, 93-279 (1838).
.
6
Biot,
Ibid., 16, 229-396 (1838).
7.
8
.
9.
10 .
Biot,
Bull. Soc. Philomat., 190-192 (1815).
Fresnel,
Ann. Chim. Phys.Cotton
Ibid.,
[7], £3, 347 (1896).
Pasteur,
Alembic Club Reprints, 14,. 26 (1860).
11. Astbury,
Proc. Roy. Soc.
12 . le Bel,
Bull. Soc. Chim.
13.
[2], 28. 147 (1825).
van*t Hoff,
Ibid.
A 102. 506 (1923).
[2], 22, 337 (1874).
[2], 25. 295 (1875).
14.
Crum Brovm,
Proc. Roy. Soc. Edinburgh, 1 7 , 181 (1890).
15.
Guye,
Compt. rend., 110, 714 (1890).
16.
Walden,
Zeit. phys. Chem., 17, 245 (189:
17.
Drude,
Gottinger Kachrichten, 366 (1892).
18.
Drude,
The Theory of Optics, Leipzig (1900).
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
19.
20
.
Hatanson,
J. phys.
[4], Q, 321 (1909).
Born,
Private communication to T. M. Lowry;
Power, p. 373 (1935).
Optical Rotatory
21 . Kuhn,
Zeit. phy-. Chem., B 20. 325 (1933).
22.
23.
Parson,
Smithsonian Collection, No. 2371,
Allen,
Phil. Mag.
(1915).
Te], 40. 426 (1920).
24.
Stark,
Jahrb. Radioakt., 11, 194 (1914).
25.
Born,
Physik. Z., .16, 251 (1915).
26.
Oseen,
Ann. Physik., 4 8 , 1 (1915).
27.
Gray,
Phys. Rev., 7., 472 (1916).
28.
de Uallemann,
Coiapt. rend., 181, 298 (1925).
29.
Boys,
Proc. Roy. Soc. (London), A 144, 655 (1934).
30.
Garino and Teofili,
Gazz. chim. ital., .56, 847 (1926).
31.
Lande,
Ann. der Physik, 56. 225 (1918).
*
Gans,
Z. Physik, 17, 353 (1923); 27, 164 (1924).
Ann. der Physik, 79., 548 (1926).
33.
Born,
Proc. Roy. Soc. (London), A 150, 84 (1935).
34.
Oke,
Ibid., A 153, 339 (1935).
35.
Thomson,
Phil. Mag.
[e], 40, 713 (1920).
Z. physik. Chem., B 4 , 14 (1929).
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
36.
Kuhn,
Trans. Faraday Soc., _26, 293 (1930).
37.
Bielecki and Henri,
Physik. Z., 14, 516 (1913).
30.
Kuhn,
Z. physik. Chen., B 31, 23 (1936).
39.
Hund,
Z. Physik, 43, 805 (1927).
40.
Rosenfeld,
Ibid., 52, 161 (1928).
41.
Kirkwood,
J. Chem. Phys., j5, 479 (1937).
42.
Condon, Altar and Eyring,
Ibid., 5,, 753 (1937).
43.
Gorin, Walter and Eyring,
Ibid., 6., 824 (1938).
44.
Kuhn,
Naturwiss. _19, 289 (1938).
45.
B o m and Jordan,
Elementare Q,uantenmechanik, p. 250 (1930).
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
CHAPTER I I .
THE PREPARATION AND PROPERTIES OF OPTICADLY
ACTTVE Op-BRCMOPROPI (2JITRXLE
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
THE PREPARATION AND PROPERTIES OF OPTICALLY
ACTIVE a-BROMOPROPIONITRILE
It was pointed out in Chapter I that Kirkwood’s theory of optical
activity leads to a relation expressing optical rotatory power as the
sum of three types of terms.
Kirkwood states that only one of these
types is important, one of the others vanishing for geometrical rea­
sons, while the third is assumed to be relatively insignificant.
For
any molecule consisting of four different groups attached to a central
carbon atom, and having optical axes of symmetry parallel to their
valence bonds to the central carbon, even the type of term retained
by Kirkwood in his calculations should vanish; the prediction would
thus be made that compounds fulfilling this structural requirement
would have very small rotatory power.
No such compounds have ever
been prepared in optically active foim, so that it appeared important
to submit the validity of Kirkwood’s assumption to experimental test.
It appears to be definitely established that each of the groups
in a-bromopropionitrile,
CH3
H
I
C—
CN
Br
has geometrical cylindrical symmetry about an axis coincident with its
bond to the central carbon atom, and it is therefore reasonable to
suppose that each group has suitable optical symmetry for this compound
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
■to serve as an experimental test of Kirkwood’s theory.Strong evidence
in favor of this assumption in the case of the nitrile
by Brockway\ who demonstrated the linearity of
group is given
C-C=N
in methyl cyanide
ft
O
by the electron diffraction method, and by Weissberger and Sangewald ,
who found a zero dipole moment for p-dicyanobenzene.
The result of
3
Bretscher
attributing a relatively large moment to 4,4*-dicyanodiphenyl
has been considered anomalous by Le Fevre and Vine^, and as not proving
any lack of axial symmetry in this molecule.
THE PREPARATION OF o-BROMOPROPIONITRILE
The starting point in the preparation of active a-bromopropionitrile
vms optically active a-bromopropionic acid.
the acid chloride, thence to the amide.
The acid was converted to
The amide was dehydrated to
yield the active nitrile.
H
H
i
I
1 - CHr>— C — COCH
I
Br
--- -
1— CHrr — C — C0C1
3 I
Br
H
l
--- 1 — CH«— C — CCWH«
|
Br
V
H
I
1 - C H g — C — CN
Br
The optically active forms of a-bromopropionic acid were first ob5
tained by Ramberg , who partially resolved the dl-acid by fractional
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
crystallization of the diastereoisomeric cinchonine salts.
had a specific rotation of only
-7.55°.
His acid
In an attempt to get opti­
cally pure forms of the acid, Fischer and Warburg6 carried Ramberg’s
procedure through 15-20 stages of the fractional crystallization to
obtain a 10% yield of the acid having
* -26.7°.
Ramberg7 re­
peated Fischer and Warburg’s work and further purified the l-GHgGHBrCOCK
by repeated freezing and centrifuging.
tion obtainable was
= -28.5°
or
had the following physical properties:
25 = 1.692.
1.700;
By this method the maximum rota­
£aj^° = -29.0*.
This material
M.P. = -0.3* to -0.5*;
20
=
It underwent very slow autoracemization, the
specific rotation decreasing 2.4° in 2.25 years.
Conversion of the
acid to the ethyl ester gave a product having a specific rotation,
= -35.5*.
This value for the supposedly optically pure ester is
probably too low because some racemization took place during the esterification.
Different specific rotations of the ester prepared in several
experiments established this fact.
£
Fischer and Warburg also prepared 1-o-bromopropionic acid by the
action of nitrosylbromide on d-alenine.
1-alanine.
The d-acid was obtained from
Yields of about 65 percent were obtained but the product
contained about 15 percent inactive acid.
Abderhalden and Wybert®
modified Fischer and Warburg’s procedure for converting d-alanine to
1-a-brcmopropionic acid.
Fischer and Warburg
Better yields were obtained.
a
prepared 1-o-bromopropionyl chloride by re­
acting thionyl chloride with the acid.
They had previously tried
using phosphorus oxychloride but the necessary excess of the latter
was too difficultly removed from the product.
The only physical
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
property of the chloride which they reported was its boiling point,
B.P.,_
= 27°.
12 ram.
was made.
Apparently no determination of specific rotation
The physical properties determined on inactive chloride
are as follows:
B.P. = 131*-133°, D11 = 1.6979 .
The preparation of an active form of a-bramopropionamide has not
yet been described.
Bischoff^® prepared the dl-amide by the intro­
duction of dry ammonia to the benzene solution of dl-a-bromopropionyl
branide.
The amide was purified by recrystallization from benzene.
Jacobs and Heidelberger11 prepared a-iodopropionamide by the dropwise
introduction of a-iodopropionyl chloride to a cold, vigorously turbined
aqueous solution of ammonia.
The amide separated from the mixture and
uras recrystallized from benzene, toluene, or water.
cedure Hamilton and Simpson
12
By the same pro-
prepared 3-branopropionamide.
Moureu and Brown13 prepared dl-a-bramopropionitrile in good yield
by dehydration of the amide with phosphorus pentoxide at 250*.
The
nitrile, after removal from the reaction mixture by distillation under
reduced pressure, was redistilled from phosphorus pentoxide.
= 59*;
D
0
20
20
“ 1.5808; D
= 1.5505; n
= 1.4585;
4
4
D
fractivity = 23.61 (calculated = 23.69).
B.P-
24 mm
molecular re-
In this research optically active a-brcmopropionitrile was pre­
pared several times.
A typical preparation and the modifications of
the accepted procedure which were tried are given in the following
page s.
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
1-g-Bromopropionlc Acid
A.
Preparation from d-Alanine by the Method of Abderhalden and
Wybert®.
40 g. (0.45 mole) of d-alanine.
H s i e i “ 3‘01‘ <3 -15 «•
in 50 cc. of aqueous solution) was dissolved in 365 cc. of 10$ hydrobromic acid solution and cooled in an ice-salt bath.
96 g. (0.6 mole)
of bromine was added and a vigorous stream of nitric oxide was passed
through the solution for 3.5 hours.
was added in anall increments.
During this time 32 g. more bromine
Air was blown through the solution to
remove the major portion of the excess bromine.
effected with sulfurous acid.
Complete removal was
The 1-a-bromopropionic acid was removed
by extraction with ether, and the ether solution dried over calcium
chloride.
The ether was evaporated and the residual acid twice frac­
tionated in vacuo.
the theoretical),
B.
B.P.= 80°-82°.
o mm.
L J5461
Yield, 29 g. (42 percent of
= -8.52°.
g
Resolution of dl-q-Bromopropionic Acid with Cinchonine .
200 g. (1.3 mole) of dl-a-bromopropionic acid was dissolved in 5 1. of
water at 45° C. and, with vigorous stirring, 200 g. (0.68 mole) of
powdered cinchonine was added in small increments.
After complete
dissolution of the cinchonine the solution was placed in a water bath
maintained at 35°-45* and evaporated under reduced pressure to half
its original volume.
This evaporation was carried out as rapidly as
possible since prolonged treatment of a-bromopropionic acid with wann
water results in partial conversion to lactic acid
14
.
Shortly after
the start of the evaporation crystals of the cinchonine salt began to
appear and at the end approximately 200 g. of salt was removed from
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
the chilled solution*.
This was dried, powdered and dissolved in ap­
proximately 3.25 liters of water at 40°.
This solution was evaporated
in vacuo to about 2 l/4 liters, whereupon 155 g. of salt separated.
In this manner, the fractionation was repeated three more times leaving
82 g. of the cinchonine salt of predominantly 1-ct-bromopropionic acid.
To isolate the free acid frcn the salt, the latter was dissolved in
about 100 cc. of concentrated hydrochloric acid, diluted to 400 cc.,
and the resulting solution extracted with ether.
The combined ex­
tracts were dried with sodium sulfate and the ether removed by dis­
tillation.
The residual acid was fractionated in vacuo, 38 g. of acid
a|D
= -13.79°.
Complete
7
resolution was not affected, nor was it necessary, since Raxnberg had
completely resolved the acid and found for the optically and chemically
pure enantiomorph
This value enables the percent
optical purity of any subsequent preparation to be calculated; on this
basis the present sample was found to be a mixture of 74.2$ 1 and
25.8$ d.
*
The first two attempts to resolve a-bromopropionic acid resulted
in the separation of the salt as an oil which defied attempts
to obtain it in crystalline form by evaporation of solutions in
other solvents and by adding first an excess of cinchonine and
then of the acid. Success in obtaining seed crystals was at­
tained only after allowing a portion of the oil to stand several
weeks, during which time patches of crystals appeared.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
d-g-Bramopropionic Acid
A fraction of the acid having a substantial dextro rotation was
obtained by treatment of the mother liquors.
The first mother liquor
of 2 1/2 liters volume was evaporated to 1 1.
No salt separated even
though it was seeded and left to stand several days.
The acid was
isolated from this salt solution by the procedure previously described.
Yield, 48.9 g.
B.P..
= 72°.
4 ram.
a OR = 6.2° (1 dm. tube).
<so
The re-
maining mother liquors from the fractional crystallization were com­
bined, evaporated to 500 cc., and the acid isolated.
27.8 g. was obtained.
B.P..vs nun
__ • = 72“-73*.
An additional
This was combined with
the 48.9 g. portion to give 76.7 g. of acid having a rotation
6.40° (1 dm. tube).
This was dissolved in 2 1. of water at 40° and
77 g. of cinchonine added.
Upon evaporation of this solution to 1 1.,
82 g. of salt separated and was filtered off.
mother liquod 29.5 g. of acid was obtained.
^
=
Prom the remaining
B.P.^ 5 ^
= 78°-78.5°.
= +10.17°, corresponding to a mixture of 67.8$ dextro and 32.2$
levo forms.
l- q - B r o m o p r o p io n y l C h l o r id e
38 g. (0.25 mole) of a-bromopropionic acid having a specific ro­
tation,
= -13.79° vras heated under reflux for 4 hours with 100 g.
(0.84 mole) of thionyl chloride.
Excess thionyl chloride was removed
by distillation under diminished pressure and the acid chloride frac­
tionated in vacuo.
chloride).
Yield, 40 g. (containing considerable thionyl
B.P.„_
45 _
ram. = 54®-55°.
No effort was made to remove com-
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
pletely the thionyl chloride since this would necessitate subjecting
the active chloride to repeated heating, possibly resulting in seme
loss of activity as well as product.
Furthermore, the thionyl chloride
was found not to interfere in the subsequent use of the acid chloride.
The necessity of using a large excess of thionyl chloride and pro­
longed warming was established by the discovery that using just slightly
more than the theoretical amount of thionyl chloride resulted predomi­
nantly in dehydration of the acid to form the anhydride.
In one in­
stance, 46.5 g. of bromopropionic acid, having a specific rotation of
•
*1 OR
[aJjj
= +10.38°, was treated with 45 g. of thionyl chloride at 55°-60®
for 1.5 hours and the resulting mixture distilled in vacuo.
About 35
g. of clear, colorless liquid remained undistilled at 100° under 6 ram.
pressure, a temperature well above the boiling points of the acid and
acid chloride.
The odor of the anhydride resembled that of the free
acid and a drop dissolved in water with the evolution of considerable
heat.
In a 1 dm. tube it had a rotation of
agg = +26.18°.
After
treatment with 100 g. of thionyl chloride for 4.5 hours at 55°-60° and
removal of excess thionyl chloride by distillation under reduced pres­
sure, 41.2 g. of a-bromopropionyl chloride was obtained (79 percent of
the theoretical yield).
3.P.25 ^
= 42°-43°.
OgQ = +14.96° (1 dm.
tube).
1-q-Bromopr opi onamide
40 g. (0.23 mole) of 1-a-bromopropionyl chloride was dissolved in
150 cc. of sodium-dried benzene, cooled in an ice-salt mixture, and
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
with accompanying vigorous stirring dry ammonia was passed into the mix­
ture until its odor could he detected.
About 9 g. was required.
The
mass turned mushy and assumed an orange-yellow color as a result of the
thionyl chloride present in the acid chloride.
The solid was filtered
off, washed with cold, dry benzene, and repeatedly treated with hot
*
benzene to extract the amide from the ammonium salts.
The amide
crystallized frcsn the benzene in the f o m of colorless, feathery needles.
Yield, 29 g. (82 percent of the theoretical).
M.P. 124®-125*.
J a J ^ = -12.9* (c = .0972 g. in 25 cc. of benzene).
An alternative procedure for the conversion of the acid chloride
to the amide was that of Jacobs and Heidelberger^.
The acid chloride
is added dropuvise, with stirring to a dilute, aqueous ammonia solution
maintained at -10* in an ice-salt mixture.
The amide separated from
the mixture almost immediately after its formation.
washed with cold water and dried at once.
It was filtered,
After complete desiccation
of the solid product the amide was extracted with hot benzene and
crystallized from the latter.
It was quite necessary that the benzene
used for recrystallizing the amide be dried over sodium.
In one
preparation of amide the recrystallization from benzene was carried
out before the crude amide had thoroughly dried.
resulted.
Partial racemization
It was also found that recrystallization of the amide from
water resulted in more or less complete racemization.
In view of the ease of racemization of active a-bromopropionamide
in water solution it was thought that perhaps the amide could be pro­
duced in an optically active form by dissolving dl-amide in an aqueous,
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
highly asymmetric environment and allowing it to crystallize out.
Accordingly, 2 g. of dl-a-bromopropionamide and 5 g. of a-d-glucose
were dissolved in 5 cc. of water.
This solution was boiled gently
for several minutes, treated with activated carbon, filtered and cooled.
About 0.8 g. of amide crystallized out.
It was inactive.
The amide
was redissolved in the solution, 5 cc. more water added, the solution
boiled gently for several minutes and then set aside for several days
to evaporate spontaneously.
The amide which crystallized out was in­
active.
1-g-Bromopropionitrlle
1-a-Bromopropionamide was dehydrated to l-a-brcanopropionitrile by
a modification of the method of Moureu and Brown
13
.
These authors
heated the mixture of amide and phosphorus pentoxide to 250° before
reducing the pressure to distill off the nitrile.
For the case at
hand, in which it was desired to minimize racemization, heating to
250® was considered too drastic, so the mildest possible conditions
under which the dehydration could be carried out were determined.
The procedure which was adopted is described in the following prepara­
tion.
Twenty-seven and one-half grams (0.18 mole) of 1-a-bromopropionamide, poxvdered in an agate mortar, was mixed intimately with 35 g.
(0.25 mole) of phosphorus pentoxide in a 250 cc. flask connected by
an all glass delivery tube to a receiver-trap immersed in an icehydrochloric acid freezing mixture.
The flask was heated in an oil
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
bath to 65°-70® for about 45 minutes, during which time the mass began
to appear moist.
The pressure was reduced by means of an oil pump to
2 mm. and the liquid distilled into the receiver.
The temperature of
the oil bath was very slowly raised, over a period of 2-3 hours, to
120° while the pressure was kept at 2 mm. or lower.
Yield, 15 g. (62
percent of the theoretical) of crude nitrile, the greater part of
which had distilled at the lower temperature.
In attempts to dehydrate a-bromopropionamide to the nitrile under
conditions as mild as possible, the dehydration with phosphorus pent­
oxide was attempted in dry benzene.
The amide liras dissolved in the
benzene and phosphorus pentoxide suspended in this solution.
After
three days the benzene was removed by distillation on a water bath.
The residue was dry indicating that no nitrile had been formed.
The dehydration was also tried using thionyl chloride.
Eight
and one-half grams (.056 mole) of the amide was dissolved in 30 g.
(.25 mole) of thionyl chloride and the solution warmed at 65°-75°
under reflux for about 3 hours or until gas was no longer evolved.
Dehydration took place and optical activity was retained, but the
nitrile was difficultly purified.
To remove the excess thionyl
chloride the nitrile had to be repeatedly fractionated in vacuo, each
distillation leaving behind a tarry residue with resultant loss of
nitrile.
A portion of the crude material was washed with water to
remove thionyl chloride and dried over calcium chloride, but this
treatment resulted in about 50 percent racemization.
Viere there no
necessity of retaining maximum rotatory power this procedure would
be satisfactory.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
The crude nitrile was mixed with phosphorus pentoxide and frac­
tionated in vacuo using a 19 in. silvered, vacuum-jacketed column
packed with Fenske glass helices and equipped with a total condensationpartial take-off type of still head.
a 20-junction copper-Advance theimel.
The temperature was measured with
The receiver was immersed in ice
and protected with a calcium chloride drying tube.
The system included
a 20 liter bottle and was evacuated continuously with a water pump.
After discarding the first few drops of distillate, the main fraction,
13.2 g. was collected between 44.5° at 15 mm. and 43.2° at 13.7 mm.
Specific rotation,
Jjaj^ — -7.35°.
Analysi s (KJ eldahl):
Calculated for CHgGHBrCN:
Tound:
10.46 percent N
10.44, 10.50, 10.49 percent N
This material, between periods of use, was sealed in ampoules.
Occasionally, after exposure of the nitrile to the atmosphere during
measurements, it was redistilled from
before being sealed up.
Samples on which these precautions were not taken were observed to
produce small crystals on the walls of the vessel after several days
with resultant diminishing of the rotatory power of the nitrile.
The
crystals were more soluble in water than a-bromopropionamide so it is
supposed that they were ammonium a-bromopropionate resulting from the
hydrolysis of the nitrile by moisture absorbed from the atmosphere.
After obtaining the first fraction of 15 g. of crude nitrile the
dehydration mixture was again heated in vacuo, this time very slowly
to 135® where it was kept for about an hour.
Approximately 5.5 g.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
more crude nitrile was obtained.
Upon fractionation from phosphorus
pentoxide in the 19 in. column, 4 g. of pure nitrile,
was obtained.
mm
“
The specific rotation of this fraction was
indicating that some racemization took place during the dehydration of
the amide to form the nitrile.
This fraction was mixed with phosphorus
pentoxide and stored in an ampoule for later use.
The total yield of
pure nitrile in this preparation was 17.2 g. (71 percent of the theoret­
ical).
Three other preparations in which larger proportions of phos­
phorus pentoxide were used gave yields of 62.1, 62.5 and 63.3 percent
in the dehydration.
The specific rotations of these other prepara­
tions of the nitrile are summarized in a later section.
THE PROPERTIES OP cl-BROMOPROPI QMITRILE
A.
The Melting Point of q-Bromopropionitrile.
The Purity of the
Present Sample.
The purity of organic liquids is usually estimated from a compari­
son of some physical constant or constants, such as density, boiling
point, viscosity, refractive index, or freezing point, with the ac­
cepted values in the literature.
This presupposes that the latter
values were accurately determined on pure materials, a supposition
seldom warranted.
Hence these comparative methods are unsatisfactory.
Another test of purity applied to liquids is the constancy of the
values for these physical constants during a fractional distillation.
The frequent occurrence of azeotropic mixtures, however, renders this
R e p ro d u c e d with perm ission of th e copyright owner. F urther reproduction prohibited without permission.
method somewhat untrustworthy.
Skau
15
pointed out that the most easily obtainable and most posi­
tive single evidence of purity is based on the shape of the heat con­
tent temperature curve when it is determined by a static or integral
method in the region where the solid changes to liquid.
With experi-
ence the curve may even be used to estimate the amount of impurities
present.
The heating curve of the nitrile was determined by a procedure
and with an apparatus (Figure 7) similar to that described by Skau
3_5
The chief difference was that the shield and main thermels, used for
determining the temperatures of the shield and sample respectively,
consisted of two copper-Advance junctions instead of one.
It was not
necessary to calibrate the former since it was used to measure only
the rate of temperature rise of the shield surrounding the sample.
(The desired rate of rise in the temperature of the shield was ob­
tained by applying across the 30-ohm heating element the appropriate
voltage from the secondary of a "Vari-Tran" transformer.)
The main
thermel was calibrated at the melting points of methylcyclohexane
(-123.3®), chloroform (-63.4®) and carbon tetrachloride (-22.9°).
Samples of these substances were obtained from the Bureau International
des Etalons Fhysico-Chimiques, Brussels, Belgium.
checked at the melting point of mercury (-38.87®).
The calibration was
The results are
given in Tables 2 and 3 and are represented graphically in Figures 8
and 9.
The temperatures are all given in millivolts as read on a
"Queen” potentiometer; the values in parentheses are the temperatures
of the shield.
The curve for carbon tetrachloride (Table 3 and
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without perm ission.
Dewar flasks
(M9
Heater
Copper shield
Shield thermel
Sample
Main thermel
—
V3
Sample tub©
before filling
Vacuum Line
Figure 7.
Heating Curve Apparatus
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
Table 2.
Heating Curves
Time
2
min.
2.5
3
4
5
5.5
6
7
7.5
8
8.5
9
10
11
11.5
12
12.5
13
14
14.5
15
16
17
17.5
18
19
20
20.5
21
22
22,5
23
23.5
24
25
26
27
27.5
28
28.5
29
Methylcyclohexane
8.15 mv.
(8.08)
8.10
8.06
(7.96)
8.01
Chloroform
Mercury
4.55 mv
(4.43)
4.52
4.50
4.47
4.44
4.42
(4.27)
4.40
2.84 mv.
2.80
(2 .66 )
7.97
4.39
4.38
4.37
2.77
2.74
2.73
7.95
(7.83)
7.94
7.93
4.36
(4.11)
4.36
4.36
2.73
(2.54)
2.73
2.73
7.93
7.93
7.93
(7.70)
7.93
7.93
7.93
4.36
4.36
4.36
(3.94)
4.36
4.36
4.36
2.73
2.73
2.73
(2.40)
2.73
2.73
2.73
7.93
7.93
(7,58)
7.92'
4.36
4.36
(3,78)
4.36
2.73
2.73
7.92
7.91
7.91
7.90
(7.46)
7.89
4.36
4.35
4.35
4.35
2.73
2.73
2.73
2.73
(2 .1 1 )
2.73
7.88
4.34
(3.61)
4.33
2.73
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
Table 2.
Time
Met hylc yc1ohe xane
30
min.
31
32
32.5
33
34
35
36
37
37.5
38
39
40
41
42
42.5
7.87 mv.
7.85
7.82
(7.34)
7.78
7.74
7.69
7.62
7.53
(7.23)
7.46
7.40
7.35
7.30
7.26
(7.12)
(Continued)
Chloroform
4.33 mv.
4.32
4.32
(3.50)
4.30
4.28
4.25
4.21
4.16
(3.38)
4.12
4.06
4.00
Mercury
2.73 mv
2.73
(1.98)
2.73
2.72
2.72
2.71
2.70
(1.86)
2.69
2.64
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
57*
Time, minutes
c\2
to
00
>> O
O
>»
r*“
n
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
Table 3.
Heating Curve
Time
2
min.
2.5
3
4
5
6
7
7.5
8
9
10
11
12
12.5
13
14
15
16
17
17.5
18
19
20
21
22
22.5
23
24
25
26
27
27.5
28
29
30
31
32
32.5
33
34
35
36
37
Carbon Tetrachloride
3.57 mv
(3.38)
3.52
3.46
3.40
3.35
3.32
(3.10)
3.31
3.31
3.31
3.31
3.31
(2.86)
3.31
3.30
3.29
3.28
3.26
(2.64)
3.23
3.19
3.12
2.96
2.82
(2.44)
2.72
2.62
2.54
2.48
2.43
(2.26)
2.38
2.33
2.29
2.24
2.20
(2.09)
2.17
2.14
2.11
2.07
2.04
Time
37.5 min.
38
39
40
41
42
42.5
44
46
47.5
48
50
51
52
52.5
53
54
55
56
57
57.5
58
59
60
61
62
62.5
63
64
65
66
67
67.5
68
69
70
71
72
72.5
73
74
75
76
78
Carbon Tetrachloride
(1.94) mv
2.01
1.98
1.95
1.92
1.89
(1.80)
1.84
1.78
(1.68)
1.73
1.68
1.66
1.65
(1.56)
1.64
1.63
1.63
1.63
1.63
(1.45)
1.63
1.63
1.63
1.62
1.61
(1.36)
1.59
1.57
1.55
1.52
1.49
(1.28)
1.45
1.40
1.36
1.32
1.29
(1.21)
1.26
1.24
1.22
1.19
(1.14)
R e p ro d u c e d with perm ission of the copyright owner. F urther reproduction prohibited without permission.
59
o>
o
o
l.Zo
~Time., mi nu tes
U)
Nj
^
*ff9Ai]llLU
/
?- S ’-/a cj£~£J?J
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
Figure 9) includes, in addition to the melting point, a temperature of
transition between two solid forms.
The melting point of a sample can be found directly from the heat­
ing curves which have been run to test its purity.
The exact point
taken as the melting temperature has to be more or less empirically
chosen; on the basis of considerable work, Skau decided that the tem­
perature of the sample two minutes before the beginning of the change
in direction of the curve at the end of the ’’flat" could be taken as
the melting point.
This rule was used to determine the e.m.f. of the
thermel at the melting point of each of the standard samples and the
values are summarized in Table 4.
Using these values to evaluate the
Table 4.
Standard
Accepted M . P .
E.M.F. at M.P.
Methyleyelohexane
-123.3°
-.00793 volts
Chloroform
- 63.4°
-.00436
Carbon Tetrachloride
- 22.9°
-.00163
constants, the equation relating the e.m.f. of the thermel to the tem­
perature was found to be
t®
=
13.9 x 103 E (volts) - 8.33
X
104 E 2 + 1.62 x 107 E 3
(17)
With this thermel the melting point of mercury was found to be -38.9*
which agrees perfectly with the accepted value, -38.87°.
The transition temperature of carbon tetrachloride was found to
be -47.3®.
Wahl^® reported it to be -47°.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
The results on the heating curve for purified l-ct-bromopropionitrile are given in Table 5 and are graphically represented in Figure
10.
It is to be noted that the first signs of melting appear only
very shortly (approximately 3 minutes) before the time when the shield
reaches the temperature of the "flat” .
With extremely pure compounds,
such as the standardizing substances studied, this initial evidence
of melting appears at or shortly after the time when the shield tem­
perature is at the temperature of the flat.
However, Skau has shown
that the time of the beginning of inflection is extremely sensitive
to impurities; a sample of chlorobenzene containing 0.5 mole percent
of meta-dichlorobenzene first showed evidence of melting about 10
minutes before the time when the shield temperature reached that of
the "flat", while with pure chlorobenzene this time of initial
evidence of melting appeared about 2.5 minutes previous to the time
of the shield temperature reaching that of the "flat".
In view of
this sensitivity and the fact that the approach to the melting point
of the nitrile was rather abrupt it can be concluded that the method
of purification employed gave" samples of a-bromopropionitrile which
were quite pure.
The e.m.f. at the melting point of a-bromopropionitrile was
0.00436 volts which corresponds to a temperature of -63.4°.
The sample of chemically pure nitrile which was used for the
determination of the heating curve of this substance consisted of a
mixture of unequal proportions of the two optical antipodes.
There­
fore, in view of the simple nature of the heating curve, it can be
concluded that there is no racemic compound formation.
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
Table 5.
Time
12
min.
12.5
14
16
17.5
18
20
22
22.5
24
25
26
27
27.5
28
29
30
31
32
32.5
33
34
35
36
37
37.5
38
39
40
41
42
42.5
l-g-Bromopropionitrile
4.61 mv.
(4.57)
4.57
4.53
(4.49)
4.51
4.48
4.46
(4.42)
4.45
4.45
4.44
4.43
(4.35)
4.43
4.42
4.42
4.42
4.42
(4.28)
4.41
4.41
4.41
4.40
4.40
(4.21)
4.39
4.39
4.39
4.38
4.38
(4.15)
Time
43
min.
44
45
46
47
47.5
48
49
50
51
52
52.5
53
54
55
56
57
57.5
58
59
60
61
62
62.5
67
67.5
68
69
70
71
72
72.5
73
l-g-Bromopropionitrile
4.37 rav
4.37
4.37
4.37
4.36
(4.08)
4.36
4.36
4.36
4.35
4.35
(4.01)
4.34
4.34
4.34
4.33
4.32
(3.93)
4.31
4.30
4.28
4.27
4.26
(3.83)
4.12
(3.71)
4.08
4.06
4.01
3.97
3.93
(3.60)
3.89
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
f>3.
DO
o
m
”
7ime,
minutes
o
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
B.
The Density of g-Bromopropionitrile.
The density of a-bromopropionitrile had to be accurately known in
order to calculate molecular refractions and specific rotations.
Since
the latter were to be determined at several temperatures, the density
as a function of temperature was measured.
for their preparation
D° = 1.5808
and
Moureu and Brown
13
found
= 1.5505, values found
to be too high compared with those of the present work.
The density of the nitrile was measured at 0*, 25°, 50* and 75°
by a dilatometric method.
The dilatcaneter had a capacity of about
7 cc.; the capillary tube was of pyrex glass specially chosen for
uniformity of bore and was graduated in millimeters.
In calibrating
the capillary, it was found that 3.2136 g. of mercury formed a column
43.50 cm. long at 22°.
This corresponds to a volume of 0.00546 cc.
per centimeter.
Water was used for the calibrating fluid.
It was boiled to re­
move dissolved gases, cooled, and introduced into the evacuated
dilatometer through the capillary.
Calibrations were made at the
temperatures at which the density of the nitrile was to be determined.
Weighings were corrected for the buoyancy of air.
The thermostat was a 4 liter beaker of water heated with an
immersion heater.
The temperature was manually controlled to
* 0.03°
and was measured with a calibrated thermometer graduated in tenths
of a degree.
The coefficient of expansion of the nitrile was so great that In
raising the temperature 25* between determinations the liquid level
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
passed completely out of the graduated capillary.
A portion of the
sample was therefore removed each time so that the level was approxi­
mately at the calibration point for that temperature.
Correction was
made for any difference by use of the volume-per-unit-length calibra­
tion of the capillary.
The latter was assumed to be constant.
The results on the densit3r of the nitrile are given in Table 6
and are graphically represented in Figure 11.
Table 6.
Density of q-Brcmoproplonitrile
t_
Density, g. per cc.
0“
1.5777
25°
1.5383
50°
1.4982
75°
1.4577
These data can be represented over this temperature range by the
following equation:
=
where
1.5777 - 1.557 x 10-3
t - 0.80 x 10“6 t2 + 3.2 x 10“9 t3
t = degrees centigrade.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
65*
PQ
>
-r-f
to
to
to
Vo
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
C.
The Refractive Dispersion of a-Bromopropionitrile; Molecular
Refractivity.
In view of the importance of group polarizabilities in connection
with some of the theories of optical activity it seemed desirable to
obtain as complete information as possible on the refractive dis­
persion of a-bromopropionitrile.
Data were obtained in the visible
spectrum at temperatures of 0.5®, 25® and 50*.
A Pulfrich refractometer (of Zeiss manufacture) completely
equipped with jackets for circulating water at constant temperature
about the sample and prism was used.
The sample cup was originally
cemented to the prism with a wax which was found to be soluble in
the nitrile.
The cup was therefore removed, cleaned and refastened
with glycerine-litharge cement.
After the latter had been allowed
to set overnight it was washed with water, alcohol and chloroform
prior to use.
A mercury arc lamp and a helium-filled Geissler discharge tube
were employed as multichromatic light sources.
A rapid stream of thermostating water was circulated through the
refractometer from a well-stirred, 70-liter thermostat by means of a
small gear pump.
The temperature was read on a calibrated thermometer
contained in the heater thimble which was immersed In the sample.
The temperature was maintained constant to
t. 0.02°.
At the lowest
temperature dry air was blown on the face of the prism and the cup
in order to prevent the condensation of moisture from the atmosphere.
For the same purpose a thick rubber gasket was placed around the
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
heater thimble so that when it was in position in the .sample cup the
latter was effectively stoppered.
The measurements v/ere made in the customary manner; the precaution
being taken to approach the reading always from the same direction.
Three or more readings were taken at each wave length and averaged.
After the low temperature measurements the sample was replaced by a
fresh portion.
In order to calculate the refractive index of the sample from the
measured angle of refraction correction must be made for refraction
due to the prism.
This is done by means of the equation
n
where
(18)
t
n^ = refractive index of the sample at temperature t, N^. = re­
fractive index of the prism at temperature t, and i is the observed
angle of refraction (corrected, of course, for the zero of the in­
strument).
In tables furnished with the prism used, these corrections
have been made only for the wave length of the sodium D line.
There­
fore, to get refractive indices at the other wave lengths employed,
the values of the refractive index of the prism at these wave lengths
had to be known.
The latter were obtained by using the dispersion
data furnished with the prism (Table 7) to calculate an equation of
the form'*'7
(1 9 )
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
with \vhich the refractive index of the prism at the desired wave length
2
2
In this equation a^.,
k^, m^. and 1^
are constants which
was calculated.
were found to have the values given in the following equation expressing
the refractive index of the prism, N as a function of wave length,
2
N 20
=
2.55703 - 0.007562
\ 2
A
0.023072
+
A
(20)
- 0.029846
Points 1, 2, 3 and 5 in Table 7 were used in evaluating the constants of
Table 7.
X » microns
N
0.4340
1.64351
2.
.4861
1.63309
3.
.5461
1.62504
4.
,5893
1.62085
5.
.6563
1.61596
No.
1.
equation (20).
For point 4 the equation gives a value of N = 1.62083.
The interpolated values of
o
at the wave lengths of the mercury and
helium lines anployed are given in Table 8.
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
Table 8.
Source
He
microns
N
0.7065
2.60164
2.
.6678
2.60910
3.
.5876
2.62758
.5791
2.63003
.5461
2.64076
.5016
2.65919
7.
.4922
2.66382
8.
.4713
2.67534
9.
.4471
2.69116
.4358
2.69970
.4047
2.72810
1.
4.
Hg
5.
6.
10 .
He
Hg
11 .
The dispersion data for the glass of the prism was obtained at
20°.
Therefore if the refractive index of a sample is measured at
some other temperature correction must be made for the difference in
refractive index of the prism between 20* and the temperature of the
measurement.
A table of temperature corrections for the prism used
was supplied by the makers.
The correction, which is applied to the
observed index, is a function both of the magnitude of the latter
and the wave length.
For indices of the order of magnitude 1.44 to
1.48 encountered in this work, the variation of temperature correction
with wave length is given in Figure 12.
The ordinate is given in
units of the fifth decimal place per degree difference of the tem-
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
71.
QOS)'
*3
•a
oooy
jf
iO:
oqSS'
ooof
•
In
K
l
^ 3"fd I ' 0 » d H-fS"/•
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
perature from 20*.
rection.
An illustration will make clear the method of cor­
At 0.5* the observed index of refraction of a-brcmopropioni-
trile at wave length 0 . 5 8 7 6 was 1.46936.
The value of the cor­
rection at this wave length is found to be 0.31 units of the fifth
decimal place per degree.
The total value of the correction is 0.31
(0.5°-20°) = -6.0, which upon addition to
nuncorr
gives
n corr
=
1.46930.
The observed angles and calculated refractive indices are listed
in Table 9.
These are graphically shown in Figure 13.
The values at
25° were determined on samples from two different preparations of the
nitrile, very closely agreeing results being obtained.
These data
were fitted to equation (19) to give the following equations:
n2
°*
P
ngQ
-
-
2.120706 - 0.006421
\2
+
2.085080 - 0.002319
\2
A
+
\p
2.046961 + 0.002028 A
■- ■.
A 01291.1--X
- 0.025440
0.0131368
X2-
(21a)
(21b)
0.020936
o 013394
+
X
- 0.016829
The refractive indices at the wave length of the sodium D line
were obtained by use of the above equations and the values were em­
ployed in the calculation of molecular refractivities by means of the
Lorenz and Lorentz equation,
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
(21c>
•H |
•
03
CO
o
o
>
1
—i
i
n
*— i
in
m
in
D~
D-
t— 1
.
I— 1
•
»— i
•
rH
IN
rH
rH
•
«— 1
CO
rH
to
CO
o
o
O
in
a
u
o
o
a
in
•
CO
Nil
in
CO
•
rH
•
rH
^H
•
o
I-i
to
in
rH
•
i— I
In
e
0
•
lO
rH
rH
£
o
4
O
o
4
4
0
•*H
rH
CO
rH
cn
rH
CO
rH
05
CO
CO
tO
Tj<
oa
rH
to
CV2
t
o
05
rH
•
tD
O-
•
cr>
o-
EN
a»
CO
<— i
•
cco
CO
■<*
in
<x>
CD
Cin
co
o
CO
in
o
o
o
•
rH
•
i— i
.
i— i
•
rH
•
»— 1
rH
r— i
rH
•
rH
D-•
cr*
o•
r
H
rH
o•
o-•
cn•
05
co
•
O
o•
CO
03
rH
CO
rH
CO
0
<*«
o
o
o
O
rH
rH
o
ID
9
rH
to
rH
to
DlO
rH
•
«— 1
in
•
in
rH
in
CO
in
C\3
CO
■<*
o
•
o
03
05
rH
in
o
to
•
i
—I
rH
•
rH
in
CO
«o
u
u
o
o
a
•
rH
CO
■"4*
tO
CO
in
to
rH
O
IQ
rH
O
to
cn
to
rH
•
in
<H
to
DO*
to
rH
in
CO
1— 1
o
•
•
rH
rH
*
rH
1—1
•
03
to
o*
rH
•
CO
to
o
•
in
to
a>
05
o
rH
■<*
O
o-
to
o
to
1—i
•
in
o
rH
rH
rH
in
o-
to
£>
CO
o-
•
rH
•
rH
•
rH
05
In•
D~
o
3.
3•
•— i
rH
03
47® 58.5’
9
32.0'
03
1.48762
in
to
i
f—ni
I
o
C\3
o
03
o
o
O
CO
CO
CO
rH
in
to
o
GO
to
to•
to
oco
in
o
oin
rH
rH
•
•
CN
•
o
©
O
fn
O
o
CO
to
In•
in
CV3
o
o
o
O
o
3
3
3
3
tO
rH
tn
to
rH
O
in
CO
to
i
—i
rH
cr>
xF
•
•
•
a>
W
•
o
03
rH
•
o
3
<D
in
rH
Cn
O
i
CO
£50
CO
03
CO
in
UD
CO
R e p ro d u c e d with perm ission of the copyright owner. Fu rther reproduction prohibited without permission.
45® 52.8’
CO
4047
03
CO
to•
51.7'
in.
o
11.8’
o
R e fr a c t iv e
Index
•
•
H
03
in
i— i
in
«
of
a -B r o m o p r o p lo n ltr lle
to
•
o
oo
in
50® 12.8’
in
o
3•
1.47525
pH
f-l
o
o
a
1,46225
73.
74
to
o
<V-i
in
o
i
c\ni
on
i
Ml
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
75.
k
=
n2 - 1
M
. _
n2 + 2
The values found were:
Rn «. =
0.5
23.67
R25
23.75
=
R 50
23.81
The molecular refractivity obtained by adding atomic and group refractivities
18
was found to be
R
=
23.56.
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
, ,
(2 2 )
D.
The Specific Rotation and Rotatory Dispersion of* g-Brcgnopro-
pionltrile.
a-Branopropionitrile was prepared three times in an optically ac­
tive condition.
The specific rotations of these preparations are
summarized in Table 10.
Table 10
[a]|5 for Optically Pure
Nitrile
Composition
1
5.30*
2
+ 4.56°
31.8
ft
" ; 68.2
7.42*
74.2
If
It
3.
67.1 percent levo; 32.9 percent dextro
25.8
15.5®
+ 12.5®
ft
15.3°
tr
The rotations were measured with a Schmidt and Haensch double­
field polarimeter the analyzer scale of which could be read to a
hundredth of a degree.
The instrument was covered on three sides by
a black, light-tight box and at the analyzer end by a black curtain
to keep out stray light.
The light source, a shielded sodiurn-vapor
lamp, was placed outside the box and the light admitted to the polar­
izer through a metal tube.
Measurements were made on the pure liquid
contained in a 7 cc. one decimeter, water-jacketed tube.
Readings
could be duplicated to -0.03° and the average of at least five read­
ings was used to calculate specific rotations.
Temperature control
to ±0.02° was obtained by circulating a rapid stream of water from a
thermostat through the water jacket.
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
The composition given in the second column of Table 10 are those
of the various samples of a-bromopropionic acid from which the nitrile
was prepared.
(Rambergts^ value of
H i 5 - -28.5° for the completely
resolved acid was used as the basis for calculating the degree of opti­
cal purity of the acid.)
The optical composition of the nitrile does
not exactly correspond, however, to that of the acid since it was evi­
dent that partial racemization took place in the dehydration of the
amide to the nitrile, and it is highly probable that other steps in
the conversion xvere not carried out without some racemization.
There­
fore it must be emphasized that the values of the specific rotation
of a-bromopropionitrile given in the third column of Table 10, and
in particular the highest value,
= -15.5*, represent minimum
values of the specific rotation of this compound.
The true value,
somewhat higher than 15.5°, could not be determined because of the
ease of racemization under the conditions of the preparation, not
only of the nitrile but of the intermediate compounds.
Optically active a-bromopropionitrile undergoes slow autoracemization.
A sample which had a rotation of a = -8.15° in a
one decimeter tube at 25.8° at the time of its preparation was
sealed in an ampoule and kept in a dark cupboard.
Fourteen months
later it had a rotation of a = -1.90° (1 dm. tube; t = 25°) and
after 18.5 months its rotation had dropped to a = -1.76 (l dm. tube;
t = 25*).
This sample was dry.
Another sample of the nitrile,
which upon redistillation over phosphorus pentoxide was collected in
trap cooled in a dry-ice-alcohol bath, had a rotation of a = +7.01°
(1 dm. tube; t = 25°).
In spite of protection of the receiver with
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
a calcium chloride tube moisture condensed in it, as evidenced by the
fact that the nitrile became cloudy at low temper atlires.
Upon frac­
tionation of this material at temperatures under 45® and without fur­
ther drying, the rotation dropped to a = +4.93® (1 dm. tube; t = 25°).
This represents a loss of 33.7 percent in rotatory power while only
22 percent of the material was removed in the fractionation.
Thirty-
three days later the rotation was found to have dropped 0.26® further.
It was immediately placed over phosphorus pentoxide and upon redis­
tillation from the latter the rotation was found to be unchanged.
These observations seem to indicate that the autoracemization is
catalyzed by moisture.
In order to check on the optical purity of the nitrile to find,
if possible, the extent of racemization in its preparation from the
active a-bromopropionic acid, attempts were made to convert it back
to either the acid or its ester, compoundc which are known in an
optically pure state.
This conversion, to be of any value, had to
be carried out under conditions which would be expected to produce
as little racemization as possible.
The most promising method ap­
peared to be the alcoholysis of the nitrile to the ester by absolute
alcohol in the presence of hydrogen chloride.
Twelve grams (0.09 mole) of the nitrile (
=
-7.19®) and
4.1 g. (0.09 mole) of absolute alcohol were dissolved in 20 cc. of
absolute ether, and the solution was saturated at 0° with dry hydro­
gen chloride.
Crystals of a-bromoprionimino-ethyl-ether hydrochloride
separated from the reaction mixture after a day and a half.
The hy­
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
drogen chloride-saturated ether was distilled from the crystals in
vacuo.
The latter were dissolved in ice water, and the solution im­
mediately and continuously shaken with ether to remove the ethyl
a-bromopropionate as fast as it was formed by the hydrolysis of the
iraino ether.
The ether solution of the ester was dried over sodium
sulfate, the ether removed by distillation, and the residual ester
fractionated in vacuo.
Yield 7.5 g. (46 percent of the theoretical).
B -P -S1 mm. = 60--61*.
[a] ! 3 = -5.09*.
On the basis of Ramberg’s^ value for the specific rotation of
the nearly optically pure ester, this sample obtained from the nitrile
contained only a 14.3 percent preponderance of the levo form.
The
acid from which the nitrile was prepared contained a 48.4 percent
preponderance of the levo form; thus, assuming no racemization in
the alcoholysis of the nitrile, the extent of racemization in the
preparation of the latter was 34.1 percent.
The maximum value of
the specific rotation of optically pure a-bromopropionitrile would
then be about 23.1°.
However, the assumption that no racemization
took place in the alcoholysis of the nitrile is of questionable
validity.
It can only be concluded that the specific rotation,
JaJ^, of optically pure a-bromopropionitrile lies between 15.5°
and 23.1*.
On first sight, this relatively large value for the specific
rotation of a-bromopropionitrile indicates that greater importance
should be attached to the higher approximations of Kirkwood's
polarizability theory (see Chapter I).
Kirkwood 19 has suggested,
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
however, that it is not unquestionable evidence, since it is possible
that lateral optical anisotropy of sufficient magnitude to account for
the observed rotation could be induced in the
the
C-Br
dipole.
CN
group, probably by
He has not as yet carried out any further calcula­
tions to test the validity of this view.
The rotatory dispersion of a-bromopropionitrile in the visible
portion of the spectrum was measured by the method of Perkin 20 , who
resolved the lines from multichromatic sources by mounting a directvision spectroscope in front of the eyepiece of the polarimeter.
The
polarimeter, polarimeter tube, and temperature control have been
described above.
The light sources employed were the sodium-vapor
lamp previously mentioned, a high intensity, Hanovia quartz mercuryvapor lanp operating at 5 amperes, and a large (10 x 0.4 cm.), aircooled, hydrogen discharge tube capable of operating for short in­
tervals at 200 milliamperes.
line was employed.
Prom the hydrogen spectrum, only the C
All of the visible lines of the mercury discharge
except that at 0.4046
were employed.
The latter was almost com­
pletely absorbed by the nitrile.
The error in the polarimeter readings varied inversely with the
intensity of the line employed.
The maximum error, encountered in
the measurements on the blue mercury line,
, which has
very low intensity, was certainly less than 0.10°, or approximately
0.5 percent.
Measurements were made on the pure nitrile; the average
of three or more readings on each line being used to calculate
specific rotations.
The latter were multiplied by a factor, F, to
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
obtain the specific rotation of the optically pure nitrile, the assump­
tion being made that no racemization took place in the formation of
the nitrile.
The results are given in Table 11 and are graphically
represented in Figure 14.
Those on d-a.-brcmopropionitrile (prepara­
tion 2, Table 10) are to be considered as merely preliminary data
since they were obtained with three different instruments* and with­
out temperature control of the sample.
The l-ot-bromopropionitrile
was preparation number 3 of Table 10 and the measurements on it were
made with the apparatus and procedure described above.
The tempera­
ture coefficient of the rotatory dispersion was determined in order
to provide as complete optical data on this compound as could be ob­
tained with the facilities available.
The values of
given in Table 11 were plotted as a function
of wave length (Figure 14a). The linear relationship indicates that
the rotatory dispersion in the region studied can be adequately ex­
pressed by a two-constant Drude equation:
Accordingly the data were appropriately weighted, and fitted to such
an equation by the method of least squares.
The following equations
were obtained:
*
The author is indebted to Professor Everett S. Wallis and Prince­
ton University for the privilege of obtaining some of these
data with their instrument.
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
82.
OlpM
o k r
rH|'--1
•
e
P=4 tC
to 1-- 1©
• d ,'d
Os »■ <
tO
•
to
©
Td
w
<D
<D
U
<D
OS
CD
•
CQ
O
c\2
CO o>
CO
• • • • •
CQ CO to o co
H
rH 1— I rH 02
rMJii
1__ 1
d
OJIh
Olf— (
ora;
rHj*1 | J
* •
fH tjfl
e 1-- 1©
O <t * . cd
in l ii
0
02
■—(
.
H4
II
p
o
•H
P=4
CO
0?
0>
Ed
a>
..--- >
o C - CN cn C"» o
tO tO CD to CO tO
m • •
• * •
in c^ o
i-H
CO
0*
d
to
to
•
o
•
02
to
II
IV,
a
P
a>
P
co
®
CD
i
—i
P
p
Id
p
o
P3
•H
Pr
O
P
Pr
0
1
in
p
02
o
m
i
CQ
02
•
tO
i-H
1
02 O CO
tO r—1 CO
« •
.
to CO in
rH rH 02
1
1
i
M O
02 t
in
in 02 to o
CO
CO c o in cj! 02
Ph
to
a>
CS.'P
(2.0
O
co
02 02 in as co
CO rH CO in 02
.
.
.
n
no
co
in
CP CO (O H IO
CD t- ^ o> n
h
«o cd
to
_ in ~ip
■* in tJ4■«*
?
Pc
0
1
t n CO CO
i— I
i— Ii— I i— I 02 CO
II I II
to
in
O-
TO
CD
rH o
H4O
a» o to
CO to 02
......
CD
P
o
02 cd tsos
o
to
©
CD —
• 0*
o
in «s
02 rH
H 4
•
o
p rH
i— 1
•d 1
ID
■— 1 tl
O
o 0
o
<D
a t
3
02
CO
as
H4
t>>
p
•H
CQ
P
(D
CO
CD
to
in
>s
-p
m
a
CD
I I I I I I
m
i
to
o i— I ©
■dc .d.d
O
>> e*
p
rH 02 CO CD cO
rH I— I (— I rH 02
CO 02 CD 02 02
(M H CO tO CD cO
P
......
03
CO W
^ O
CO
(O (O i n C# 02
h
O
I—l
11 8
c f IQ
I I I I II
<D
■
•H
olp
O r—
old
Td
«
CD
■—
.
H
I
«
10
CO CO tO
ii i ii
.
•H
P
CQ CJ
P *rH
0-
tO
Td
N jc CD
8
1—i• cd
CO
§ b r
N
p
CQ
O
O
P
O
o
o
•h
to CO
o
• •
o CO
H 4 02
1 1
CO to CO
o as to
• • «
CD H 4 in
i— 1 02 to
1 1 i
in in
• «
i— { CQ
to in
I i
.~
o
in «
02 o
.
o
p r—1
r—t
Td 1
a>
rH II
O
O sO a
to 02 o
O
p
o
o
<£>
in in
. •
H4
CO <o
1 «
a>
CO
m •
i— 1 in
r— 1 rH
1 1
CQ
•
in
02
i
Olfi.
p
•rH
P
cd
"tt4 CT> in CQ CO
. • • • •
CO o i— \ CQ
1
— i i— i i— 1 r H
1 1 i 1 1
i
P
-rH
p
EH
<D
CO to 02 02 in as
CD CO
in CV2 D~
.
•
•
CO rH rH
rH rH
1 1 t
CD
m
c- rH O
eCO
• •
H 1 rH CO
tO tO in
1 1 i
00 tO
c- cn
. •
cn o
CQ 02
i
tO
H
4 to
CO
'f
as ■
H 4 02 o- i— 1 Cv
. • • • . •
rH in to CO in in
1— I i— i 1— 1 I— 1 02 £Q
1 i 1 1 i 1
CO
•
CQ
CD
1
to
©
Td
w ss ta
c» a> o
H4
......
to
rH a s c o in
as
rH m m co
h4
i—I r—i i—I i—I02 CO
I I I I I I
P
o
CD
t*. p
p ©
•H p
CO
«
CD ----
02 ^ CD O
I— I i— i rH 02
I I I I
CO
tO
in
to
tO O
OS CD
co oin in
rH tO
to rH
H 4 os
in tt4
CO
in
co
h 4
©
o
p
rH
" a
CQ P
P *H
CQ
O
P
o
in ©
rH• -8
Cd
0 - CO OS H 4 rH D CO i n rH O IN CO
I
P
C
—d
t>---T
02 CD Ci)
w a w = = =
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
83.
<»-i
•p
CQ
in
k/*/< lenyth, fx.
o
to
CQ
cwf
«*•
V
'
permission
prohibited without
" R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited
without permission.
83a.
C\2
micrans
O
O
O
PO
P O
o
to
W>
«0
S i 1—1
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
84.
4.192
H0,4
X2 - 0.0702
4.266
25
X2 - 0.0689
H
4.286
H
X2 - 0.0699
50
r 169.5
_
4.500
X2- 0.0694
It is seen that the constant
X 2
is essentially independent of tem­
perature while the constant A varies in the manner shown in Figure 15.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
85.
o
VP
to
•H
h
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
IS.
The Absorption by g-Bromopropionitrile in the Hear Ultra-Violet,
The absorption of a-bromopropionitrile at 20° in the near ultra­
violet portion of the spectrum was measured* with a Bausch and Lomb
sector photometer employing a tungsten spark source of radiation.
The
solvent was 95 percent ethyl alcohol which had been redistilled from
potassium hydroxide.
The spectra were recorded on Eastmann iH0 plates.
The absorption characteristics of the nitrile between 0.420y/t-*-' and
0.210^/44' , the limits of the instrument, are shown in Figure 16.
logarithm of the absorption coefficient,
€
The
, is plotted as ordinate.
is given by the expression
£
=
7 d
los
i
where c = concentration of the nitrile in moles per liter of solution,
d = 2 is the length of the sample in centimeters, and
is the ratio
of light intensities passing through the pure solvent and through the
solution.
Log i2. was obtained from the setting of the rotating
sector.
*
The author is greatly indebted to Miss Margaret McLean of this
Laboratory for her help in obtaining this absorption spectrum.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
87.
•
o
0
1)
•H
F®<
3
U
-P
-«H
a
o
*H
S’
o
P
Ph
o
§
p
PQ
1
d
o
a
o
•H
•P
&
O
+>
©
rH
O
£
a)
P
+3
■H
&
a)
S3
a>
-p
cl
■H
CQ
Wa/e
Lenath, microns
s
O
O
I
to
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
BIBLIOGRAPHY FOR CHAPTER II.
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
1.
Brockway,
J. Am. Chem. Soc., 58, 2516 (1936).
2.
Weissberger and S&ngewald,
J. Chem. Soc., 157. 855 (1935).
3.
Bretscher,
Helv. Phys. Acta, 2., 257 (1929).
4.
Le Fevre and Vine,
J. Chem. Soc., 140. 1878 (1938).
5.
Ramberg,
Ber., 33 B . 3354 (1900).
6.
Fischer and Warburg,
Ann., 340. 168 (1905).
7.
Ramberg,
Ibid., 570. 234 (1909).
8.
Abderhalden and Wybert,
Ber., 49 B . 2456 (1916).
9.
Collet,
Bull. Soc. Chim.
[3], .15, 717 (1896).
10.
Bischoff,
Ber., 50 B . 2312 (1897).
11.
Jacobs and Heidelberger,
J. Biol. Chem., 21, 146 (1915).
12.
Hamilton and Simpson,
J. Am. Chem. Soc., 51. 3158 (1929).
13.
Moureu and Brown,
Bull. Soc. Chim.
[4], 27, 907 (1920).
14.
Fittig and Thomson,
Ann., 200. 79 (1880).
15.
Skau,
Proc. Am. Acad. Arts and Sci., .67., 551 (1933).
16.
Wahl,
Z. phys. Chem., 8 8 . 141 (1914).
17.
Tilton,
Bur. Standards J. Research, 17, 646 (1936).
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
18.
Gilman,
Organic Chemistry, An Advanced Treatise, p. 1739 (1938).
19.
Kirkwood,
Private Communication.
20
.
Perkin,
J. Chem. Soc., 89, 616 (1906).
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
C H A PT ER I I I .
THE PREPARATION, PROPERTIES, AND ATTEMPTED
RESOLUTION OF BRCMOCHLOROE1IJOROMETHANE
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
THE PREPARATION, PROPERTIES, AND ATTEMPTED
RESOLUTION OF BROMOCEELOROFLOGRGMETHANE
The ideal compound on which to test various theories of optical
rotatory power and more particularly to test unambiguously the im­
portance of the higher terms in Kirkwood’s theory would presumably be
a pentatomic, optically active molecule.
Such a molecule would be of
further theoretical interest in view of the fact that no molecule con­
sisting of five atoms has ever been prepared in an optically active
state.
In fact it was once suggested'*' that optical activity does not
necessarily result from the presence of an asymmetric carbon atom in
an organic molecule unless other carbon atoms are attached to the
asymmetric center. However, this idea was disproved when Pope and
p
Read succeeded in resolving chloroiodomethanesulfonic acid, CHCIISO3H.
This result proved that any compound the molecule of which consists
of one carbon atom united to four inorganic atoms or groups is capable
of existence in an optically active condition.
There are five possible pentatomic, potentially asymmetric, or­
ganic molecules:
1.
Bromochlorofluoroiodomethane, CBrClFI
2.
Bromofluoroiodomethane, CHBrFI
3.
Chlorofluoroiodomethane, CHC1FI
4.
Bromochloroiodomethane, CKBrClI
5.
Bromochlorofluoromethane, CHBrCIF
Of these compounds only the last two have been prepared and of these,
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
attempts at resolution have been made only on the latter.
Bromochloroiodomethane was prepared by Garino and Teofili
rz
who ob­
tained it by the treatment of bromoehloroiodopyruvic acid with alkali,
They reported it to be a liquid, m.p. about 5°, unstable in air, light,
and in solutions of most of the common organic solvents.
Undoubtedly
because of this great instability no attempts have been made to separ­
ate it into its enantiomorphs.
Bromochlorofluoromethane was prepared by Swarts
4
by the reaction
of dibronochioromethane with equimolecular quantities of antimony tri­
fluoride and bromine.
He also obtained it from the treatment of either
brcmochlorofluoroacetaldehyde or bromochlorofluoroacetic acid with
5
alkali .
Swarts described it as a highly mobile, colorless, volatile
liquid (B.P. 38*) having an agreeable chloroform-like odor.
It turned
light yellow in sunlight, attacked glass only at red heat, was stable
to nitric acid, and was destroyed completely by alkali.
It was
definitely identified by elementary analysis and vapor density de­
terminations.
He recognized the fact that it was the simplest
asymmetric molecule and attempted its resolution with the limited
amount he had available.
The only attempt which he described con­
sisted of a fractional crystallization of the solid complex which he
stated the compound formed with saiicyiide.
The complex contained
two molecules of the substituted methane, and since it could not be
obtained in two different crystalline forms he concluded that one
molecule of each enantiomorph of bromochlorofluromethane was present
in each molecule of the complex.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
Swarts
5
prepared ‘the strychnine as well as the cinchonine salt of
bromochlorofluoroacetic acid and in each case succeeded in getting by
fractional crystallization two isomeric salts of different specific
rotatory power.
After separation of the alkaloid from the salts, solu­
tions exhibiting optical activity vrere obtained but the optical activity
was of a fugitive character and could not be exactly measured.
This
almost instantaneous racemization of bromochlorofluoroacetic acid in
solution precluded the possibility of obtaining active bratiochlorofluoramethane by decarboxylation of the active acid in alkaline solu­
tion.
THE PREPARATION OF BRCMOCHLOROFLUOROMETHANE
The reported difficulty of preparation and instability of bromochloroiodomethane together with similar information, gained in this
work, concerning substituted methanes containing iodine plus other
halogens, indicated that any of the iodine-containing compounds listed
above would not only be difficult to obtain but unsuited to attempts
at resolution.
Therefore, attention was centered on Swarts* compound,
bromochlorofluoromethane.
The method of preparation of bromochlorofluoromethane was
basically that used by Swarts.
were adopted.
Several modifications of details
As in his procedure, the dibromochloromethane from
which the bromochlorofluoromethane was prepared, was derived from
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
ethyl chloroacetal by the procedure of Jacobsen and Neumeister
as
follows:
2Br«
CBrgClCHO
CHgClCH(OCgH5 );
+
2CgH5Br
+
HgO
NaCK
1
CHBrgCl
+
HCOQNa
Some ethyl chloroacetal was prepared by the chlorination of ethyl al7 8
9
cohol * and of paraldehyde followed by treatment with ethyl alcohol ,
but a better method was that of Eilachione"^ who obtained it by the
addition of chlorine to vinyl acetate in the presence of ethyl alcohol:
3C 3H 5 CB
CHo-CH- OCOCHn
CHg—CH-OC OCHrj + Clg
I 2 I
Cl
|H 2-G H (0 ° ^ 5 J8
3
Cl
Cl
+ CgHgOCOCHg + HC1 + HgO
Ethyl Chloroacetal
9
A.
Preparation from Vinyl Acetate .
About 350 g. of vinyl
acetate was purified by distillation, the fraction boiling at 70°-72°
being collected for use.
Three hundred and forty-four grams (4 moles)
of the purified acetate and
1200
cc.
(20
moles) of absolute ethyl al­
cohol were placed in a three-necked flask immersed in a dry ice-alcohol
cooling mixture.
a towel.
The flask, where exposed to light, was covered with
Accompanied by vigorous stirring, a stream of dry chlorine
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
was passed into the solution until the increase in weight was 284 g.
(4 moles).
The mixture was allowed to warm to room temperature and to
stand overnight.
It was then poured into an equal volume of ice water,
and the ethyl chloroacetal, which separated as an oil, was taken up in
ether.
The ether solution was washed with cold water, then with sodium
bicarbonate solution until it was neutral to litmus, then washed once
more with cold water, and finally dried with calcium chloride.
The
ether was removed by distillation from a water-bath, and the residual
ethyl chloroacetal was fractionated in vacuo.
The yield was 410 g.
(69 percent of the theoretical) of ethyl chloroacetal boiling at 67®68° (32 mm. ).
B.
7 ft
Preparation from Ethyl Alcohol ? .
While being cooled with
running water 1.5 liters (24.9 moles) of 95 percent ethyl alcohol was
treated with chlorine until the specific gravity at 25° was 1.02.
750 cc. of alcohol was added and the mixture heated several hours at
60°-70° until hydrogen chloride no longer came off in large quantities.
The mixture was then cooled and allowed to stand over marble until
carbon dioxide was no longer evolved.
After filtering, the product
was precipitated by addition of water, the water-insoluble layer
separated, dried over calcium chloride and fractionated.
Yield,
270 g. (21.3 percent of the theoretical) of ethyl chloroacetal boil­
ing at 150°-160°.
£.
Q
Preparation from Paraldehyde ♦
Chlorine was passed into
264 g. (2 moles) of paraldehyde at 20®-25® with vigorous stirring
R e p ro d u c e d with perm ission of the copyright owner. F urther reproduction prohibited without permission.
until the latter had taken up 260 g.
One-half liter of alcohol was
added and the mixture, after standing overnight, was poured into an
eqiial volume of water.
The water insoluble layer was separated, dried
over calcium chloride and fractionated.
Yield, 120 g. (13.2 percent
of the theoretical) of ethyl chloroacetal boiling at 50°-55° (16 mm.).
g
Dlbromochloroacetaldehyde
Four hundred and ten grams of ethyl chloroacetal (2.67 moles) was
stirred under reflux, and 850 g. (5.3 moles) of bromine was added dropwise.
Approximately three-fourths of the bromine reacted rapidly, the
heat of reaction being sufficient to keep the mixture at 50°-60*.
When
this was no longer the case, the mixture was maintained at the elevated
temperature by a water bath until several hours after completion of the
bromine addition.
The mixture was allowed to stand overnight at the
ordinary temperature.
The condenser was set downward for distillation
and the mixture heated gradually to 100° to distill off excess bromine
and ethyl bromide.
The residual crude dlbromochloroacetaldehyde was
used in the next preparation without further purification.
Dibrconochloromethane
The crude dibromochloroacetaldehyde obtained in the above prepara­
tion was added dropivise to a vigorously stirred solution of 140 g. of
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
sodium hydroxide in 500 cc. of water cooled by an ice bath.
Upon com­
pletion of the addition, the excess alkali in the aqueous layer was
neutralized with hydrochloric acid and the dibramochloromethane steam
distilled from the mixture.
The product was separated from the aqueous
layer, dried with calcium chloride and fractionated under a nitrogen
atmosphere.
Yield, 402 g. (72 percent of the theoretical, based on
the amount of ethyl chloroacetal used) of dibramochloromethane boiling
at 119*-121° (765 mm.).
Bromochlorofluoromethane
Antimony trifluoride was prepared by dissolving antimony trioxide
in a slight excess of cold, 48 percent hydrofluoric acid contained in
a platinum dish which was surrounded by ice.
The solution was evapor­
ated to dryness on a sand bath, moistened with hydrofluoric acid and
again dried.
The resulting antimony trifluoride was powdered and
stored in a waxed bottle.
Four hundred and two grams (1.92 moles) of dibromochloromethane
was mixed with 116 g. (0.55 mole) of antimony trifluoride and 104 g.
(0.55 mole) of bromine in a copper vessel.
The latter was fitted by
means of a rubber stopper to a water-cooled copper condenser set down­
ward for distillation.
alcohol cooling mixture.
The glass receiver was immersed in a dry iceThe reaction vessel was heated in an oil
bath to 60°-70* for 12-16 hours during which time most of the dibrcsnochloromethane was converted to more volatile products.
Some gas
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
having a phosgene-like odor passed uncondensed through the receiver.
Forty-five grams of unchanged reactant was recovered by steam distilla­
tion of the residue.
The receiver containing the distillate was con­
nected to an ampoule cooled in a dry ice-alcohol mixture and the former
allowed to warm to room temperature.
About one-fourth of the distill­
ate distilled into the ampoule which was sealed and set aside.
This
material was probably largely chlorodifluoromethane (B.P. -39.8*) con­
taining brcmodifluoromethane (B.P. -14.5°) as well as same free bromine.
Free bromine was removed frcsn the higher boiling portion of the products
by shaking with mercury, 190 g. of crude bromochlorofluoromethane being
obtained by distillation.
The product had a sharp phosgene-like odor
which was not removed by refluxing over magnesium oxide for more than
two hours.
at
It was therefore necessary to resort to washing the material
0° with dilute alkali followed by water. Following this treatment,
168 g. remained.
It was dried with calcium chloride and fractionated
over phosphorous pentoxide using the 19 in. packed
in Chapter II.
column described
The main fraction, 120 g., was collected at 36.3°-
36.8° and stored in a sealed ampoule wrapped in black paper.
This
corresponds to a yield in the last step of 47.5 percent of the theoreti­
cal or an overall yield of approximately 23.6 percent.
Analysis:
0.2097 g. of sample gave 0.4693 g. AgCl + AgBr
Calculated for 0.2097 g. CHBrCIF;
0.4711 g. AgCl + AgBr
The sample was decomposed in hot, alcoholic potassium hydroxide in a
sealed ampoule.
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
In efforts to obtain better yields of bromochlorofluoromethane, dibromochloroacetaldehyde and aibromochloroacetic acid obtained by the
oxidation of dlbromochloroacetaldehyde with concentrated nitric acid
were fluorinated by Swarts’ reaction and the resulting bromochlorofluoroacetaldehyde and bromochlorofluoroacetic acid converted to bramochlorofluoromethane by the action of alkali .
The yields were much poorer
than those obtained by the method described in detail above.
Dibromochloromethane was also fluorinated by the method of Henne^.
The principle of this method consists in fluorinating by means of
nascent mercuric fluoride.
In practice the method consists in stirring
red mercuric oxide in any compound containing one or more halogen atoms
to be replaced by fluorine and passing a stream of hydrogen fluoride
through the mixture.
The mercuric oxide combines instantly with the
hydrogen fluoride, and the nascent mercuric fluoride fluorinates the
organic material.
Twenty-one grams (0.1 mole) of dibramochloromethane
and 11 g. (0.05 mole) of mercuric oxide were placed in a copper flask
fitted with a copper reflux condenser and a copper stirrer.
The mix­
ture was cooled in an ice bath and during vigorous stirring, hydrogen
fluoride was passed in through a copper tube.
The products, chloro-
difluoromethane and perhaps some bramodifluoromethane were gases and
passed out through the condenser.
There remained a little dibromo-
chloromethane in which no bromochlorofluoromethane could be found.
The antimony analog of Henne’s method with nascent mercuric
fluoride was tried, using antimony trioxide instead of mercuric
oxide.
At room temperature no reaction took place.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
THE PROPERTIES OP BROMOCHLOROELUORCMETHANE
A.
The Melting Point and Bolling Point of Bramochlorofluoroanethane.
The Purity of the Sample 'Used in the Determination of Physical Properties.
A 140 g. sample of bromochlorofluoromethane boiling at 36.3“-36.8°
was refractionated with the 19 in. packed column in order to get a
middle cut for the determination of the properties of this substance.
The first refractionation yielded a 65 g. middle fraction boiling at
36.26®-36.37®.
(The pressure dropped from 762 mm. to 749 mm. during
the 8 hours in which this fraction was collected.)
The 65 g. fraction
was again fractionated (reflux ratio, approximately 12:1) and a 36.7 g.
middle fraction was taken from it.
This material boiled at 36.11®-
36.18° (756.0-756.2 mm.) the temperature being measured with a cali­
brated, 20 junction copper-Advance thermal the reference junctions of
which were maintained at 0°.
Sealed in an ampoule and stored in the
dark, it remained colorless.
This highly purified portion of brono-
chlorofluoromethane, in addition to being used for the determination
of the boiling point, was used for the determination of the heating
curve at the melting point, the melting point, the density, the re­
fractive dispersion, and the structure bjr the method of electron
diffraction*.
*
Por this purpose a sample of the branochlorofluoromethane was
sent to Professor L. 0. Brockway at the University of LCichigan.
Up to the time of this writing, the results had not yet been
publi shed.
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
The heating curve and melting point were determined with the appa­
ratus and by the technique described in Chapter II.
The data are given
in Table 12 and are graphically represented in Figure 17.
As before,
the values in parentheses are the temperatures of the shield.
It is
seen that the first signs of melting appeared at approximately the
time the shield temperature was at the melting point, substantiating
the evidence gained from the boiling point that the sample was quite
pure.
The e.m.f. of the thermel at the melting point of the brcmochlorofluoromethane was -0.00744 volts which corresponds to a temperature of
-115°.
B..
The Density of Bromochlorofluoromethane.
The density of
bromochlorofluoromethane was determined at 0°, 10®, 20° and 25® with
the same dilatometer and by the same procedure as was used for the
determination of the density of a-bromopropionitrile (Chapter II).
Swarts4 reported the density at 16® to be 1.9058 g./cc.
The vslues
of the density of this substance at 0.5®, 10® and 20® were necessary
for the calculation of the molecular refraction at these temperatures.
The results are given in Table 13.
Table 13.
The Density of Broraochlorofluorcmethane
t_
0°
10®
20®
25®
Density, g. per cc.
1.9771
1.9503
1.9214
1.9068
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
101.
Table 12.
Heating; Curve of Bromochlorofluoromethane
Time, minutes
Time, minutes
E.K.F. . itw.
20
7.465
(7.62)
21
7.46
3
7.70
22
7.45
4
7.67
22.5
(7.07)
5
7.65
23
6
7.62
24
7.44
7
7.60
25
7.44
(7.46)
26
7.44
8
7.575
27
7.44
9
7.555
27.5
10
7.535
28
7.43
11
7.525
29
7.42
12
7.515
30
7.41
(7.32)
31
7.40
13
7.505
32
7.38
14
7.49=
32.5
15
7.49
33
7.36
16
7.485
34
7.31
17
7.475
35
7.24
(7.20)
36
7.16
2.5
7.5
12.5
17.5
18
7.47
19
7.465
37.5
in
7.73
•
c*-
2
E.U.F.. mv.
(6.95)
(6,83)
(6.72)
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
xi
I—I
Time,
m in o tr e i
O
*y
Q
IS.
■*rC
////ct-f '
'S
S-
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
o
03
These data can be represented by the following equation:
t
=
where
t
£.
1.9771
- 2.51 x 10“3 t
- 2.03 x 10”5 t2 + -3.3 x 10“7 t3
is the Centigrade temperature.
The Refractive Dispersion of Bromochlorofluoromethane.
The
refractive dispersion of bromochlorofluoromethane in the visible spec­
trum was measured at 0.5°, 10° and 20° in the same manner and with the
same monochromatic light sources as were used in similar measurements
on a-bromopropionitrile (Chapter II).
The observed angles of re­
fraction, i, and calculated refractive indices, n, are listed in Table
14.
The refractive dispersion isotherms are graphically represented
in Figure 18.
These can be expressed by the following equations:
n~
0.5
=
2.008301
- 0.004322
n
=
2.004050
- 0.018275
n
10
20
0.011751
X2 - 0. 018322
0.0087264
A 2 - o.039543
=
1.977795
- 0.006096
X
0.010821
X2 -o. 022556
The molecular refractivities of brcmochlorofluoromethane at the
wave length of the sodium D line were calculated by means of the
Lorenz and Lorentz equation.
0.5
R
10
20
The values found were:
19.24
19.26
19.306
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
Reproduced with permission of the copyright owner. Further reproduction prohibited without permission.
Table 14.
The R e f r a c t i v e I n d e x o f B r o m o c h lo r o flu o r o m e th a n e
10c
0 .5 *
Source
1.
He
i
jiiic r o n s
0 .7 0 6 5
49*
i
nc o r r .
20°
i
nc o r r .
nco rr.
6 .0 ’
1 .4 2 4 8 5
50*
4 .5 '
1 .4 1 8 9 5
51*
2 .4 '
1 .4 1 3 1 5
2.
.6 6 7 8
49° 1 8 .5 '
1 .4 2 6 2 0
50* 1 6 . 8 ’
1 .4 2 0 3 5
51" 1 5 . 5 ’
1 .4 1 4 4 8
3.
.5 8 7 6
49° 5 2 .4 ’
1 .4 2 9 2 5
50* 5 0 . 6 '
1 .4 2 3 4 5
51* 5 1 . 0 ’
1 .4 1 7 4 4
.5 7 9 1
49* 5 7 . 4 ’
1 .4 2 9 6 0
50* 5 4 . 8 '
1 .4 2 3 8 9
51* 5 5 . 3 '
1 .4 1 7 8 8
.5 4 6 1
50* 1 7 . 0 '
1 .4 3 1 3 9
51* 1 5 . 0 ’
1 .4 2 5 6 3
52* 1 6 . 0 ’
1 .4 1 9 6 1
.5 0 1 6
50* 5 1 . 2 '
1 .4 3 4 4 0
51* 5 1 . 5 '
1 .4 2 8 4 6
52* 5 2 . 6 '
1 .4 2 2 4 8
7.
.4 9 2 2
50° 5 9 .1 ’
1 .4 3 5 2 2
52"
0 .9 '
1 .4 2 9 1 5
53*
1 .9 ’
1 .4 2 3 1 9
8.
.4 7 1 3
51* 2 1 . 5 ’
1 .4 3 7 0 1
52* 2 3 . 2 '
1 .4 3 0 9 8
53* 2 4 . 9 '
1 .4 2 4 9 8
9.
.4 4 7 1
51* 5 1 . 9 ’
1 .4 3 9 5 1
52* 5 4 . 0 ’
1 .4 3 3 4 9
53* 5 7 . 0 '
1 .4 2 7 4 1
.4 3 5 8
52“
8 .7 '
1 .4 4 0 8 2
53* 1 0 . 1 '
1 .4 3 4 8 8
54* 1 3 . 7 ’
1 .4 2 8 7 8
.4 0 4 7
53*
5 .4 ’
1 .4 4 5 1 3
54*
1 .4 3 9 1 8
55* 1 3 . 7 '
1 .4 3 2 9 5
4.
Hg
5.
6.
10.
11.
He
Hg
7.9’
104
105.
<v-i
oo
<4-1
to
o
o
oa
o
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
If from the latter value of the molecular refractivity of bromochlorofluoromethane the atomic refractivities of bromine, chlorine, carbon
and hydrogen
12
are subtracted, the atomic refractivity of fluorine at
and 20° is found to be 0.956.
Since this
is not in complete agreement with such values previously reported it
seems to be further evidence that the atomic refractivity of this
element in organic molecules is dependent to a larger than usual ex­
tent upon the other atoms present, particularly halogens, as well
the aromatic or aliphatic character of the molecule.
Locke, Brode
and Henne13 found values ranging from 0.68 to 1.223 for fluorine in
ten different ethanes and ethylenes the hydrogens of which were com­
pletely substituted by chlorine and/or bromine in addition to the
fluorine.
From measurements on 30 liquid substances with aromatically
bound fluorine Schiemann^ found the atomic constant for fluorine to
be
Fp
=
0.997.
This is in good agreement with Swarts
aliphatic compounds.
15
data on
From data on three monofluorinated aliphatic
hydrocarbons Swarts found:
Fa
=■ 0.989;
F^
=
0.973
and
Fy
=
1.005.
D.
The Attempted Resolution of Bromochlorofluoromethane.
The
resolution of bromochlorofluoromethane presents a difficult problem
because of the fact that it contains no functional group allowing the
formation of diastereoisomeric compounds which could be separated by
the usual methods.
The possibility of its production in an optically
active condition by the decarboxylation of active bromochlorofluoroacetic acid was virtually eliminated by Swarts'
observation that the
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
latter undergoes almost instantaneous racemization upon being liberated
from its salts with stryehinine and cinchonine.
Of the various extraordinary methods which have been used for the
separation of enantiomorphous substances those of possible value for
the resolution of bromochlorofluoromethane appear to be
1. fractional
crystallization or fractional precipitation of complex compounds
formed with optically active substances;
2. selective solubility or
selective association of one of the active components with an opti­
cally active substance;
and
3. selective adsorption on an active
solid substance.
Several resolutions have been effected by the crystallization of
complex addition compounds.
ft
Windaus, Klanhardt and Weinhold
tained almost complete resolutions of the
dl
1g
ob­
forms of a-terpineol and
ac-tetrahydro-j3-naphthol by fractional precipitation of their digitonides.
Further application of this method has not been made, pre­
sumably because no situation has arisen which would justify use of
the expensive digitonin rather than simpler compounds commonly used
for the resolution of phenols, alcohols and thiophenols.
Goldberg
17
Sobotka and
found that the tendency towards choleic acid fomation is
greater in one of the two enantiomers of a racemic mixture, thus per­
mitting resolutions to be accomplished by choleic acid formation.
They obtained choleic acids with predominating levorotatory
enantiomers
from racemic camphor, phenyl ethyl ethanol, dipentene, and methyl ethyl
acetic acid.
They pointed out that this phenomenon offers the possi­
bility of resolving racemic hydrocarbons and other substances devoid
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
of functional groups that would allow resolution by the customary
Pasteurian methods.
Weiss and Abeles
18
reported still a third type
of addition compound formation in which one of the enantiomers of a
racemic mixture showed greater tendency toward compound formation
with, an optically active substance.
They added one mole of d-fJ-
naphthylcamphylamine to tv/o moles of dl-secondary-butylpicramide in
alcohol to form the 1:1 addition complex, which crystallized out.
The mole of secondary butylpicramide remaining in the mother liquor
was predominantly 1evorotat ory.
They reported no further attempts
using this method but suggested its possible value for the splitting
of racemates of "an indifferent chemical nature such as aromatic
nitro compounds, hydrocarbons, and halogenated hydrocarbons (bromochlorofluoromethane among others)".
van't Hoff
19
surmised that in an active solvent, the solubility
of the dextro-form of another active substance might differ frcsn
that of the levo-form; and a priori considerations are, apparently,
so much in favor of this being the case that in spite of many fail­
ures repeated attempts have been made to carry out resolutions on
this basis.
Tolloczo
20
dissolved r-xnandelic acid by shaking with
ether and a concentrated aqueous solution of fructose; while racemic
acid was similarly distributed between water and 1-amyl alcohol.
In
both cases the acid recovered from each pair of solvents was optically
inactive.
Goldschmidt and Cooper
21
the same solubility in d-limonene.
found that d- and 1-carvoxime had
Cooper‘d
showed that the solu­
bility curves for sodium ammonium 1- and d-tartrates in dextrose
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without perm ission.
solution are identical, with the limits of experimental error,
Jones 2 3
got negative results on dissolving d- and 1-camphor and d- and 1-cam24
phoroxime in d-pinene and 1-amyl bromide, Ebert and Kortum
got neg­
ative results for potassium acid 1- and d-tsrtretes in water and in
aqueous solutions of mannitol, and of d- and 1-camphor sulfonic acids
in benzene and in benzene solutions of 1- and d-csmphor.
There were
also no relative changes in surface tension which could be detected.
tf
25
Schroer
reported a successful experiment. He dissolved r-mandelic
acid in d-carvone and fractionally extracted with water.
The acid
extracted first was slightly levo; subsequent extracts had diminishing
levo rotation, and finally, further extractions removed dextro-acid.
The opposite sequence was observed when 1-carvone was used as a sol­
vent,
In addition to this observed difference of solubilities of
enantiomers in an asymmetric solvent, there has been reported by
Patterson and Buchanan
26
the observation that enantiomers have small
but appreciably different molecular solution volumes when dissolved
in an active solvent.
The difference for isobutyl 1-tartrate and
isobutyl d-tartrate dissolved in 1-menthyl acetate (approximately
5 percent solutions) was five times the observed difference (equal
to the probable experimental error) when dissolved in benzyl benzoate.
For ethyl diacetyl 1-tartrate and ethyl diacetyl d-tartrate in
1-menthyl acetate the difference was about thirty times that ob­
served in benzyl benzoate solutions.
The difference between the
molecular solution volumes of the ethyl diacetyltartrates in the
asymmetric solvent is not only much larger than that for the iso­
butyl esters but is also in an opposite sense, which makes the re-
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
suits considerably more convincing.
Thus the predicted asymmetry of*
solvent action seems to be fairly definitely established experimentally
and it is to be expected that refined measurements may, in some cases,
disclose differences in other physical properties such as surface
tension, viscosity, and, more important, vapor pressure.
The method of resolution by selective adsorption of one of the
antipodes on an asymmetric adsorbent has been deemed possible for
some time but is just now being developed to what is hoped will be a
practicable stage.
ft
In 1904 Wilstatter
2 7 attempted unsuccessfully to
demonstrate a selective adsorption of one of the active components of
a racemic alkaloid on wool or silk.
Porter and Ihrig28 in 1923 claimed
to have effected an almost complete resolution of dl-m-fi-naphtholazomandelic acid by preferential adsorption of the d- form on wool, but
this resolution could not be confirmed by Brode and Adams 2 9 .
Very
small partial resolutions of dyes by adsorption on wool have been
r*/*\
described in one or two isolated cases by Ingersoll and Adams
*z-i
by Morgan and Skinner^ .
and
rzp
Fischgold and Ammon
obtained evidence of
optically selective adsorption of mandelic acid in the presence of
alkaloids on optically inactive adsorbents.
Experiments with racemic inorganic and metallo-organic complexes,
e.g., dl-chlorobisdimethylglyoximodiamminocobalt, were made by
Tsuchida, Kobayashi and Nakamura 33 , in which a warm saturated solu­
tion of the salt was allowed to cool over powdered d- or 1-quartz.
The residual solutions after decantation were generally, although
not invariably, faintly active, the siga of activity in any given
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
case varying with that of the quartz employed.
Following this work
Karagunes and Coumoulos^, adopting the sarae technique of chromatographic
adsorption analysis as that used by Henderson and Rule 35 , obtained evi­
dence of the differential adsorption of the d- and the 1- component of
racemic{triethylenediamine chromichloride upon powdered active quartz.
The first resolution by selective adsorption which resulted in
sufficient separation so that the polarimetric observations were out­
side the range O.Ol'-O.lB® obtained in the experiments mentioned above,
was carried out by Henderson and Rule 36 .
By successive selective ad­
sorption on solid lactose they completely resolved dl-phenylenebisiminocamphor.
1_.
Attempts to Resolve Bromochlorofluoromethane by
Addition Compound Formation
Experiments were carried out to see whether bromochlorofluoromethane would form a solid addition compound with any of the optically
active solids available in this laboratory.
The substances tried in­
cluded most of the common sugars, some amino acids, camphor, tartaric
acid, ascorbic acid, cinchonine, salicin, aesoxycholic acid and digitonin.
The general procedure consisted of preparing about 2 cc. of
a near-saturated aqueous or dilute alcoholic solution of the solid,
adding several drops of bromochlorofluoromethane and cooling the mix­
ture in an ice bath for several days during which time it was shaken
frequently.
Attempts with desoxycholic acid were also made in non-
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
aqueous media.
In view of the reported addition compound-forming prop-
erties of desoxycholic acid 37 and digitonin 38 greatest hope was enter­
tained for these to enter into compound formation with the halogenated
methane.
The only success was achieved with digitonin.
Approximately 100 mg. of digitonin was dissolved in about 4 cc.
of warm water.
After cooling, several drops of bromochlorofluororaethane
was added, the mixture shaken vigorously and then cooled in ice.
A
cloudiness appeared after a few minutes and the precipitate slowly in­
creased in amount until, upon standing overnight, the entire mixture
was a semi-solid mass.
To see whether digitonin itself would pre­
cipitate in such a manner under these conditions the addition compound
was broken down by gentle boiling of the mixture until the odor of
bromochlorofluoromethane could no longer be detected in the vapor.
This required but a few minutes.
The solution was clear at this
stage and remained so even though seeded with digitonin and cooled
overnight in ice.
With chloroform a digitonide was obtained in the same manner as
above.
Weinhold
This is contrary to the reported observation of Windaus and
38
that chloroform as well as other halogenated paraffins do
not form digitonides.
ooauici
CMUOj.1
The chlorofoxm-digitonide was found to be un-
^uauoxuj
Ui
washed thoroughly with water.
j. u
vv
x J_ _L
U
XJ/CSil
Uile
tiOlUUl O H
CLilCl
It smelled strongly of chloroform and
gave a copious precipitate of silver chloride when qualitatively
tested for chloride.
Upon drying, the chloride test became decreas-
ihgly positive and was completely negative when the solid had dried.
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
The bromochlorofluoromethane digitonide was similarly unstable*
In a preliminary attempt to resolve bromochlorofluoramethane by
means of digitonide formation,
g. of digitonin was dissolved in
1.0
100 cc. of water by warming to 80®.
The solution was cooled in ice
and 16 g. of the substituted methane added and shaken frequently.
Overnight a copious precipitate appeared, the entire mass being mushy.
It was allowed to warm to room temperature and upon evacuation about
8
g. of brcmochlorofluoramethane was drawn off and condensed in a trap
surrounded with a dry ice-alcohol cooling mixture.
this material was,
ttgg = -0.05°.
The rotation of
A second fraction was obtained from
the mixture after the removal of most of the excess bromochlorofluoromethane as described above.
The mixture, containing the insoluble
addition compound, was placed in a distilling flask arranged for dis­
tillation and slowly warmed.
When the temperature reached 60*-70° the
compound started to break down and bubbles of brcmochlorofluoromethane
passed up through the mixture.
Almost an hour was necessary for the
complete disappearance of the insoluble addition compound.
of distillate was obtained.
Its rotation was,
About 4 g.
(I25 *= +0.03*.
A sam­
ple of the residual aqueous solution of digitonin gave an appreciable
precipitation with strongly acidified silver nitrate solution indi<io4,4 r>
W « —. V/ 4 . 4 ^ ^
in the formation
ft
W
e
*
ft-P
AA j r A J . W .A . j r A J X . t->
SrntjL.
of halogen acids.
f
tftVi 1 ft*r»ftP1 ^>ft-nft»*»ft
4
-V
if
tv
tf
t <
f
t
f
t
1T 4
<
<
vy
X _1_ UXV«<X
UAACUAO)
X *2 O U X
m
The hydrolytic action of such a
solution at elevated temperatures on digitonin probably accounts for
the fact that only 0.80 g. of the latter could be recovered from the
solution as the amyl alcohol digitonide.
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
The polarImetric measurements were made on the Schmidt and Haensch
instrument described in Chapter II using the sodium vapor lamp.
A one
decimeter polarimeter tube, specially constructed to have a small vol­
ume (about 0.7 cc.) was employed.
It had a side tubulature for filling
and the end plates were cemented on with glycerine- 1 itharge cement
which was thoroughly washed with organic solvents after setting.
Since the diameter of the tube (about 3 mm.) was less than that of the
light beam it was found to be advantageous to grind the inner walls to
cut down light reflection from them.
assured.
A straight bore was also thus
About 10 readings were usually taken in making one measure­
ment, the maximum deviation from the mean being about 0.02°.
Care
was exercised to get the polarimeter tube containing the sample in
approximately the same position as the empty tube had been for the
zero readings.
In subsequent attempts to effect a partial resolution employing
the digitonide the solvent was 25 percent alcohol.
This permitted
solution of both the digitonin and the bronochlorofluoromethane but
the digitonide was insoluble.
The latter was separated from the
mother liquor by centrifuging before being broken down to obtain the
bromochiorofluoromethane.
To a solution of 14.8 g. of digitonin in 400 cc. of 25 percent
alcohol, 10 g. of bromochiorof luoromethane was add-id.
The bottle
containing the solution was tightly stoppered and placed in ice.
7/ithin 15 minutes there wa 3 a copious precipitate but the mixture
was not centrifuged until after 5 hours had elapsed.
The supernatant
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
liquid was drawn off the centrifugate and the latter warmed to 55° in a
distillation apparatus.
A slow stream of air was passed through the
mixture to sweep the evolved methane into the dry ice-cooled receiver.
About 0.9 cc. of distillate having a rotation,
tained.
a
= +0.15°, was ob­
It contained some alcohol and water but was sealed in this
state in an ampoule.
to be inactive.
Upon reexamination two days later it was found
This apparent ready racemization of bromochloro-
fluoromethane accords with the conception of extreme mobility of
groups attached to isolated atoms.
The ready racemization of asym­
metric assemblages associated with isolated atoms of cobalt, chromium,
tin and in quinquevalent nitrogen compounds of simple molecular con­
stitution has been observed.
It has frequently been suggested that
this is an explanation of the difficulties experienced in preparing
optically active substances containing fewer than three carbon atoms,
and is in agreement with the fugitive optical activity in aqueous solu5
tions of bromochlorofluoroacetic acid observed by Swarts , chlorosulfoacetic acid observed by Backer and Burgers
acid
40
McMath.
and bromochloromethanesulfonic acid
39
41
, and bromochloroacetic
demonstrated by Read and
On the other hand, there is the optically stable chloroiodo-
methane sulfonic acid discovered by Pope and Read
Read and McMath^0 .
42
and confirmed by
Its stability has been postulated as being due to
some specific effect of the large volume of the iodine atom.
Two days after the digitonide had been separated from the mother
liquor in the experiment just described the latter was heated to 85°
and maintained there until about
8
g. of distillate was obtained.
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
This was mostly alcohol and was optically inactive.
The digitonin and
mother liquor were mixed and the experiment repeated twice in succession
with negative results.
In each of these attempts the insoluble dig­
itonide was much slower in forming than it was the first time.
The
solution, in the absence of bramochlorofluoromethane, remained turbid
and contained considerable halide ion.
Only 9 g. of the digitonin
could be recovered as the amyl alcohol digitonide.
The amyl alcohol digitonide was heated in an Abderhalden dryer at
80° for
8
hours but since this did not completely remove the amyl al­
cohol it was suspended in water and the latter boiled until the odor
of amyl alcohol had disappeared from the vapor.
Repetition of the ex­
periment with 9 g. of digitonin and 10 g. of bromochlorofluoromethane
produced 2.9 g. of distillate having a rotation,
a 25 ~ +Ch02®.
Again
negative results were obtained when the experiment was twice repeated
using this same digitonin solution.
This time only 5 g. of digitonin
could be recovered, first by precipitation as the ether digitonide
(after removing most of the alcohol from the solution), followed by
amyl alcohol digitonide precipitation, and finally the solution was
treated with £-naphthol.
before the last stage.
The digitonin was apparently all removed
Upon evaporating same of the remaining liquid
in vacuo a sticky solution remained which still retained some foaming
properties characteristic of digitonin solutions.
A final experiment using 5 g. of digitonin and
6
g. of bromo-
chlorofluoramethane was carried out in the same manner as before.
From the digitonide approximately 1 cc. of distillate having a rota­
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
tion,
a = +0.03° was obtained.
A day later 0.7 cc. was distilled from
the mother liquor; it was inactive.
Again another attempt using the
same digitonin solution was futile.
The second formation of digitonide
required several days whereas the first required about 15 minutes.
It is interesting to note that successful results could apparently
be obtained only xvith fresh digitonin solutions and that the treatment
resulted in considerable destruction of digitonin as well as bromo­
chiorof lu or omethane.
The rotations, with one exception, are so small
as to be only a little greater than the error in making readings.
However, their uniformitj1- of direction along with the one relative^
large value are rather convincing evidence that a partial resolution
has been achieved.
2.
Attempts to Resolve Bromochlorofluoromethane Based
on the Asymmetry of Solvent Action
It seems fairly definitely established
35 36
’
that enantiomorphous
substances exhibit more or less pronounced differences in physical
properties under asymmetric conditions not involving so-called primary
valence forces.
It therefore appeared worthwhile to investigate the
possibility of effecting a partial resolution of bromochlorofluoro­
methane by utilizing differences in vapor pressures of its antipodes
over its mixture with an optically active solute.
In the absence of
any basis for an estimate of the magnitude of such vapor pressure
differences it xvas decided to try a separation using optically active
solutes which seemed likely to associate to a considerable extent
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
118.
■with brcmochlorofluoromethane.
Because of the asymmetry of such a sol­
ute, it should have different degrees of association, and hence differ­
ent solubilities, in the enantiomere of the solvent; this should re­
sult in a difference of vapor pressures which might possibly, in some
cases, be of sufficient magnitude to allow separation.
Haloforms in general have been shown to form complexes by hydro­
gen bonding with substances containing electron donating atoms such
as oxygen, nitrogen and sulfur.
from infra-red absorption
AA
4-5
*
Evidence for this has been obtained
and Raman
/fi
spectra studies, and from
An
measurements of deviations from Raoult*s law
ing
a Q
’
, of heats of mix-
49 50
51
52
* , of viscosities , of dielectric constants , and of freezing
points
5*5
.
The order of increasing effect of the halogen atoms in
causing complex formation of haloforms with donor substances by hy­
drogen bonding is
I < Br ^ Cl <C,F, hence brcmochlorofluoromethane
should be expected to associate to a large extent with ethers, esters,
ketones, amines and amides.
In particular, d-camphor and diethyl d-tartrate, prepared by the
method of Fischer and Speier
54
were tried.
In preliminary experiments,
concentrated solutions of these substances in bromochlorofluoromethane
were each sealed, after partial evacuation, in a flat bulb on one end
of a U-tube of large bore.
The other end was placed in a dry ice-al-
cohol cooling mixture for about
liquid had distilled.
2
weeks or until a small quantity of
The distillates were washed with concentrated
sulfuric acid and redistilled in order to remove any of the active
solute which may have been distilled over.
The resulting bromochloro-
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
f luoromethane was inactive in each case.
In view of the inefficiency
of fractionation in such a distillation the vapor pressure differences
vrould have to be quite appreciable to effect any separation.
The nega­
tive results merely emphasized the fact that the magnitude of any such
difference must be very small.
It was therefore decided to try the
separation in a reasonably efficient counter-current tower.
As a system less precious than one involving brcmochlorofluoro­
methane for trial experiments, the combination dl-secondary butyl
alcohol-desoxycholic acid was used.
This pair should constitute a
particularly favorable case since butyl alcohols form stable choleic
acids.
Accordingly this alcohol was distilled in a column having a
125 cm. section packed with single-turn glass helices.
The column
was vacuum jacketed, the outside of the jacket being maintained at
the same temperature as the top of the column.
The column was pro­
vided with a double still head so designed that the major portion of
the vapor from the column was condensed into an extraction thimble
containing the desoxycholic acid before being returned as reflux to
the column.
The distillation was run at about 40 mm. pressure since
a low temperature favors the stability of the complexes.
column had run at total reflux (reflux rate
20
After the
drops per minute) for
about 1.5 hours, sufficient distillate was collected at a reflux
ratio of
10
to
1
to fill a 20 cm. polarimeter tube.
the original alcohol was distilled.
One-fifth of
The distillate had no rotation
within the sensitivity of the polarimeter.
As would be expected
similar experiments with desoxycholic acid at atmospheric pressure,
R e p ro d u c e d with perm ission of th e copyright owner. F urther reproduction prohibited without permission.
and with d-tartaric acid, also gave negative results.
In these cases
the vapor pressure difference appears to be too snail to be of prac­
ticable value.
It remains to be seen whether any can be found large
enough to permit accomplishing a significant resolution.
Time limita­
tions have prevented any further efforts in this direction.
_3.
Attempted Resolution of Bromochlorofluoramethane
by Optically Selective Adsorption
A search was made for a solid, optically active substance which
would adsorb bromochiorofluoromethane frcm. its solution in a suitable
solvent.
Because of the expected small rotatory power of this sub­
stance as well as the uncertainty regarding the magnitude of any differ­
ence in adsorptive powers of the enantiomers on an asymmetric substance
it was concluded that a change in refractive index due to adsorption
from-solutions would be a more sensitive measure of degree of ad­
sorption than any production of rotatory power that might occur.
Therefore the choice of solvent was based upon its having an index
of refraction differing materially from that of bromochiorofluoro­
methane; the solvent should also be non-polar to minimize attraction
for the methane and solvent action on the adsorbent, it should be
completely miscible in all proportions with the methane derivative,
it should be available in quantity and easily purified, it should be
rather high boiling to permit easy recovery of the low boiling bromo­
chiorof luoromethane, and it should be liquid at
0°
since it was de­
cided to work at or near this temperature in order to favor maximum
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
adsorption and minimize loss by volatilization of bromochiorofluoro­
methane.
Toluene appeared to be a solvent with properties closely ap­
proaching those desired.
Toluene was dried by distilling off approximately 10 percent.
The remainder was used without further treatment.
For use in deter­
mining composition from refractive index the refractive indices (for
the wave length of the yellow helium line, 0 . 5 8 7 6 )
of bromochloro-
fluoromethane toluene mixtures at 0.5° C. were determined and are
shown in Figure 19.
Plotted as ordinate is the angle of refraction
as observed with the Pulfrich refractometer (Chapter II).
The substances investigated as possible adsorbents were powdered
in an agate mortar and dried 2 days over potassium hydroxide in a
vacuum desiccator.
Two grams of the dry, powdered material was placed
in 4 cc. of an approximately 25 percent solution of broanochlorofluoro­
methane in toluene and the glass stoppered bottle containing the mix­
ture immersed in ice.
Approximately 1.5 cc. of the supernatant liquid
was withdrawn and its refractive index at 0.5° quickly measured and
compared with that of the original solution.
The substances investi­
gated and the changes they produced in the angle of refraction are
listed in Table 15.
The positive value for
A i
lactose is probably due to soluble impurities.
in the case of
The more hopeful ad­
sorbents appear to be arabinose, maltose, sucrose, and mannitol.
These experiments had to be discontinued because of lack of time for
further work.
It is hoped that they may be completed at a future
date.
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
Weight of BromocUlorefluarorneih&nt.
by
Percent
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
Table 15.
Changes In Refractive Index Produced by Adsorption
of Bromochlorofluoromethane from. Its Solution in Toluene
t = 0.5°
Adsorbent
Micrometer Reading
A i
Control Solution
3° 5 4 . 6 '
Arabinose
3° 4 6 . 3 *
- 8.3’
Glucose
3° 5 3 . 8 ’
-
Lactose
3° 5 5 . 7 *
+ 1.1'
Levulose
3° 5 4 . 1 '
-
Maltose
3° 4 7 . 7 *
- 6.9*
Mannitol
3° 4 9 . 4 *
-
5.2'
Sucrose
3° 4 8 . 0 '
-
6.6*
Control Solution
2° 4 1 . 6 *
Alanine
2° 3 6 . 5 *
- 5.1*
Albumen
2° 3 9 . 4 *
- o pt
Casein
2° 3 7 . 3 *
- 4.3’
Cystine
2° 3 8 . 4 *
-
3.2’
Glutamic Acid
Hydrochlori de
2° 4 0 . 8 *
-
0.8’
Tartaric Acid
2° 4 1 . 1 *
- 0.5*
Tyrosine
2° 3 9 . 8 *
_
(i = 3 8 ° 3 0 . 8 ’ )
0.8*
0.5*
(i = 3 9 ° 4 . 8 * )
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
1.8’
In summarizing these experiments xvhich have aimed at the resolution
of dl-brcmochlorofluoromethane, it seems that a definite resolution was
obtained by means of the digitonide.
Systematic application of this
method to obtain complete resolution would be extremely difficult, even
if at all possible, because of the apparent ease of racemization of
the bromochlorofluoromethane under aqueous conditions.
Furthermore,
large amounts of the expensive digitonin would be required.
The experiments based on the asymmetry of solvent action and on
selective adsorption have not been carried beyond an exploratory stage.
The latter method would appear to be the more promising, and it is
hoped that further experiments can be carried out.
Certainly the
importance of the problem from the theoretical point of view merits
a thorough trial of all possibilities, in spite of the great diffi­
culties encountered, not the least of which is the preparation of
the relatively large amounts of bromochlorofluoromethane which would
be necessary.
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
BIBLIOGRAPHY FOR CHAPTER III.
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
1.
van’t Hoff-Eiloart,
Arrangement of Atoms in Space, p. 25 (1898).
2.
Pope and Read,
J. Chem. Soc., 105. 811 (1914).
3.
Garino and Teofili,
Gazz. chim. ita*L., j56, 847 (1926).
4.
Swarts,
Bull. Acad. roy. Belg.
[3], _26, 102 (1893).
5.
Swarts,
Ibid., [3], 31, 28 (1896).
Ibid., Memoirs Couronnes, 54, 54 (1896).
6.
Jacobsen and Uemneister,
3er., 15, 601 (1882).
7.
Lieben,
Ann., 104, 114 (1857).
8.
Fritsch,
Ibid., 279, 300 (1894).
9.
Freundler,
Bull. soc. chim.
(4], <1, 70 (1907).
10.
Filachione,
J. Am. Chem. Soc., 61, 1705 (1939).
11.
Henne,
Ibid., 60, 1569 (1938).
12.
Gilman,
Organic Chemistry, An Advanced Treatise, p. 1739 (1938).
13.
L-j ike, Brode and Kenne,
J. Am. Chem. Soc., 56, 1726 (1934).
14.
Schiemann,
Naturwissenschaften, _19, 706 (1931).
15.
Swarts,
J. chim. phys., 20, 30 (1923).
16.
r.7indaus, ICLanhardt and Weinhold,
Z. physiol. Chem., 126, 308 (1923).
17.
Sobotka and Goldberg,
Biochem. J., _26, 905 (1932).
R e p ro d u c e d with perm ission of th e copyright owner. Further reproduction prohibited without permission.
18.
Weiss and Abeles,
Monatsh., J59, 202 (1932).
19.
van’t Hoff,
Lagerung der Atome in Raume, p. 8 (1908).
20.
Tolloczo,
Z. physik. Chan., 20, 412 (1896).
21.
Goldschmidt and Cooper,
Ibid., 26, 714 (1898).
22.
Cooper,
J. Am. Chem. Soc., J23, 255 (1900).
23.
Jones,
Proc. Cambr. Phil. Soc., 14, 27 (1907).
24.
Ebert and Kortum,
Ber., 64, 348 (1931).
25.
Schroer,
Ibid., 65, 966 (1932).
26.
Patterson and Buchanan,
J. Chem. Soc., 142. 290 (1940).
27.
Wilstatter,
Ber., 37, 3758 (1904).
28.
Porter and Ihrig,
J. Am. Chan. Soc., 45, 1990 (1923).
29.
Brode and Adams,
Ibid., 48, 2193, 2202 (1926).
30.
Ingersoll and Adams,
Ibid., 44, 2930 (1922).
31 •
iuOrgsJi tine!.
32.
Pischgold and Ammon,
Biochem. Z., 234, 39 (1931).
53.
Tsuchida, Kbbeyashi and Nakamura,
J. Chan. Soc. Japan, 56, 1339 (1935).
34.
Karagunes and Coumoulos,
Praktika, 15, 414 (1938).
Nature, 142, 162 (1938).
Skinner ,
J. Chem. Soc., 127. 1731 (1925).
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
35.
Henderson and Rule,
Ibid., 141. 917 (1938).
36.
Henderson and Rule,
J. Chem. Soc., 141, 1568 (1939).
37.
Sobotka,
Chemistry of the Sterids, p. 109 (1938).
38.
Windaus and Weinhold,
Z. physiol. Chem., 126. 299 (1923).
39.
Backer and Burgers,
J. Chem. Soc., 127. 233 (1925).
40.
Read and McMath,
Ibid., 128, 2183 (1926).
41.
Read and McMath,
Ibid., 127, 1572 (1925).
42.
Pope and Read,
Ibid., 105. 811 (1914).
43.
Read and McMath,
Ibid., 154. 2723 (1932).
44.
Gordy,
J. Am. Chem. Soc., 60. 605 (1938).
45.
Buswell, Rodebush and Roy,
Ibid., 60, 2528 (1938).
46.
Dadieu and Kohlrausch,
Physik. Z . , 31, 514 (1930).
47.
Zellhoefer, Copley and Marvel,
J. Am. Chem. Soc., 60, 1337 (1938).
48.
Zellhoefer, Copley and Marvel,
Ibid., 60, 2666 (1938).
49.
McLeod and Wilson,
Trans. Faraday Soc., 31, 596 (1935).
50.
Zellhoefer and Copley,
J. Am. Chem. Soc., _60, 1343 (1938).
51.
McLeod,
Trans. Faraday Soc., 30, 482 (1934).
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
52.
Glssstone,
Ibid., 33, 200 (1937).
53.
Wyatt,
Ibid., 25, 43 (1929).
54.
Fischer and Speier,
Ber., 28, 3255 (1895).
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
SUMMARY
R e p ro d u c e d with perm ission of the copyright owner. Further reproduction prohibited without permission.
SUMMARY
a-Bromopropionitrile, a substance of simple molecular
structure, has been prepared in an optically active state.
In addition to its more common physical properties, its
rotatory and refractive dispersions in the visible spectrum,
at several, different temperatures, have been measured.
A procedure for the preparation of pure bromochloro­
fluoromethane in yields of 20-25 percent of the theoretical
has been developed.
The refractive dispersion of this sub­
stance in the visible spectrum, at several different tem­
peratures, in addition to its more common physical proper­
ties, have been determined.
Attempts were made to resolve bromochlorofluoromethane.
These vrere based upon addition compound formation, vapor
pressure difference of enantiomers due to asymmetry of sol­
vent action, and optically selective adsorption on an active
adsorbent.
Partial success was achieved only by the first
of these three types of procedure.
Bromochlorofluoromethane
was found to form a digitonide with which a partial resolution
of the former was effected.
R e p ro d u c e d with permission of the copyright owner. Further reproduction prohibited without perm ission.
Документ
Категория
Без категории
Просмотров
0
Размер файла
5 088 Кб
Теги
sdewsdweddes
1/--страниц
Пожаловаться на содержимое документа