y-f r-'
»'. , ; V-'WiJv/
\m
^^if
n-aii
^'
. ^^'
•i^^ft^im
/ ••-/
^:^v/^
a\-^^ A'
m\y,
\^i yj
I, ON RELATIONS HKrW^EN THE COLOR AND T}{E COMPOSITION
AND CONSTITUTION OF THE ALKALI SALTS OP NITROPHENOLS
II. COMPARISON OF THE METASULPHAiaNEBENZOIC ACIDS MADE
BY DIFFERENT KETJIODS.
A Dissertation presented to the Board of
University Studies of the Johns Hopkins Uni-
versity for the Degree of Doctor of Philoso-
phy,
by
Joseph Cliristie Whitney Frazer,
1901.
Johns Hopkins University,
Baltimore, Maryland.
^rioi
Contents.
Page.
Ackno wl eagement
PART I. -
Relations between the color and the Composition and
Constitution of the Alkali Salts of Nitrophenols , 1
Introduction 1
Preparation of Material 2
Orthonitrophenolates 5
Metanitrophenolat es 14
Paranitrophenolates n 1C>
(1:2:3) NitrocresQlates 18
Conductivity of Ortho , Meta and Paranitrophenol -- 21
Conclusion 24
PART II.
Metasulphaminebenzoic Acids made by different
Methods 27
Historical Introduction 27
Preparation of Metasulphaminebenzoic Acid
Preparation from Metasulphobenzoic Diamide 30
Preparation from Metasulyjharninebenzonitrile 32
" Paratoluicinemetasulphonic
Acid 32
" Parabrorrmetatoluenesulphonic
Acid 33
" Metasulphonchlorbrnzoic Acid 34
Effect of Heat on the Melting-point of I'etasulpha-
minebenzoic Acid 35
Conclusions 37
Biographical 38
Acknowledg ment.
This investif^ation was suggested and carried out
under the supervision of Professor Remsen, to whom the au-
thor desires to express his gratitude both for helpful sug-
gestions during the course of the work and for encourage-
ment and instruction received .
I also desire to express my obligation to Professor
Morse, Professor Ames and Dr. Jones,
I, ON RELATIONS BETWEEN THE COLOR AND THE COMPOSITION
AND CONSTITUTION OP THE ALKALI SALTS OF NITROPHENOLS,
II, COMPARISON OF THE ^1ETASULPHAl^aNEBENZ0IC ACIDS MADE
BY DIFFERENT METHODS,
Introduction.
The alkali salts of the nitrophenols , so far as they
have been made, show quite a wide variation in color and
this difference of color has been stiidied by Carnelly and
(1)
Alexander in the case of the sodium and potassium salts of
ortho- and paranitrophenol. Since this work of Carnelly
and Alexander the subject has not been further investigated
and in some cases, notably in the case of metanitrophenol ,
the number of salts is extremely small. In none of them
has either the rubidium or caesium salt been made.
During the course of organic preparations ortho-
and paranitrophenol. were imde and some of each one convert-
ed into the sodium and potassium salts. Because of this
difference in color among these ortho and paranitropheno-
lates and, too, of the difference in color between these
salts of orthonitrophenol itself it was suggested to me to
prepare the sodium, potassium, rubidium and caesiiim salts
of the three nitrophenols , the object being, to see if any
connection between their color and composition could be
worked out. The salts of these metals were chosen as fur-
nishing the best series for such a study. In the case of
none of the nitrophenols has either the caesium or rubidium
salt been made and of metanitro phenol only the potassium
salt is recorded.
The lithium salts too would have been included, but
no lithium salts v/ere available for the work. As the work
progressed (1:2:3) nitrocresol v/as included in the list of
substances whose salts were made and analyzed.
PREPARATION OF MATERIAL.
Ortho- and paranitrophenol were made in the usual
way by treatment of phenol with nitric acid and after sep-
arating th.i oily layer from the acid the orthonitrophenol
is distilled off with a current of steam. As the orthoni-
trophenol comes over in this operation it is siiff iciently
pure for use. The paranitrophenol is extracted from the
tarry mass by hot, strong hydrochloric acid and this solu-
tion boiled with animal charcoal. It is obtained perfectly
pure by several recrystallizations in this manner.
Metanitrophenol was obtained from metanitraniline
by first diazotizing and boiling the resulting diazo com-
pound with dilute sulphuric acid, and then this solution is
extracted with ether. Durin^^ this operation the directions
(1)
of Bantlin were followed. By recrystallizing from strong
hydrochloric acid, as in the case of paranitrophenol the
substance is obtained pure.
In making the (1:2:3 ) nitrocresol orthotoluiiine is
the starting-point. Prom this ortho-cresol was made . By
treatment of the orthocresol with nitric acid in acetic ac-
(2)
id solution action takes place at once and after standing
a short while the whole is poTired into water. There sepa-
rates out a brownish-red substance and from this, by dis-
tillation with steam the (1:2:5 ) nitrocresol passes over
while the (1:2:5 ) nitrocresol which is formed at the same
time is left behind. In this condition the substance is
not quite pure enough for use, but can be obtained perfectly
pure by recrystallizing from dilute alcohol, and melted
sharply at 69.5 .
The preparation of the nitrophenolates is a simple
matter as they can be obtained from either the carbonates
or the hydroxides. In making the nitrocresolates it is
(1) 'B XI., 2099. (2) B XIV., 567.
best to use the hydroxide oi' the metal as the substance is
such a 'veak acid tliat it requires heating for some time to
drive out carbondioxide completely from carbonates, and
this is objectionable on account of the volatility of the
nitroci'esol with water vapor.
Some of these salts are obtained in pure state only
with some difficulty. This was found to be true of sodium
and caesium orthonitrophenolate and of some of the nitro-
ci'esolates. This is probably due to the fact that the salt
in solution undergoes slight decomposition while evaporat-
ing and some of the nitrophenol or nitrocresol separates
with the salt. That slight decompositions do go on in so-
lution of these salts is indicated by the fact that the
odor of orthonitrophenol and nitrocresol can be observed
when solutions of these salts have stood for any length of
time. The dry salts too undergo very slow decomposition,
the orthonitrophenol and nitrocresol escaping and the sub-
stance turning white. This may be due to the tendency of
the carbondioxide of the atmosphere to convert the substan-
ces into carbonates.
In order to purify those salts which separated out
slightly impure the following method was adopted. The salt
was treated in the cold with strong alcohol which would
free it from any excess of the nitrophenol or nitrocresol.
It was then dissolved in hot alcohol which would leave be-
hind any excess of carbonate. The salt thus extracted was
recrystalllzed from water. The analysis was made on the
sample as soon as possible, because of the fact that some
of the salts undergo slight decomposition on long standing
even when tightly corked as indicated by the odor.
The y;ater of crystallization was not determined as
it was probable in some cases some decomposition would take
place. So that the metal was determined as sulphate and
the formula deduced from these results.
Orthonitropheno lates.
SODIUM ORTH ONIT ROPIIENOT.A TE . - With this salt con-
cordant results could not be obtained till the method of puri-
fication outlined above was applied, analyses made on each
sample were concordant but varied with the sample used.
This is shown by the following analyses:
Amount of sub- Na ^ 80^ found X Na found,
stance taken.
( 0.5284. 0.2178 13.35
{
( 0.7593 0.3116 13.26
(0.5189 0.2250 14.04
(
(0.5224 0.2246 13.94
(0.5768 0.2348 13.22
(
(0.3283 0.1388 12.90
0.4749 0.1976 13.47
(1.0314 0.4429 13.90
(
(1.4349 0.6115 13.80
(:.8945 0.3194 14.0
(
(0.9554 0.3389 13.94
There were six different samples made and of the
analyses above the first and second were made on one of
these samples, the third and fourth on another, the fifth
and sixth on another, the seventh on another, the eighth
and ninth on another, the tenth and eleventh on another and
in all of them the percentage of sodium is too small for
the anhydrous salt. By purifying the salt by the method
above the following results were obtained.
Amt. of substance Na^SO^found X Na found Theory
0.3805 0.1679 14.29 14.28
0.3511 0.1537 14.18
This agrees with the formula C^ H^ ^ ^ Ua '^'^^ ^^-^^
crystallizes as is v;ell known in large, splendid, scarlet-
red needles, often as much as two centimetres in length.
P OTASSIU?, ! ORTHONITROPHENOLATF - It is ret as difficult
to obtain this salt pure as it is to obtain the sodium salt
above. It crystallizes from water solution in large orange-
red plates, often as much as a centimetre in length, or in
orange-red needles which have been obtained as much as three
centimeters long. This salt has been made and analyzed by
(1)
Fritzsche, who, from his analysis, assigned to it the for-
mula C^ H^ :^ I/2 K^O and later was made by Post and L'ehr-
(2) ^^ -0.^
tens who gave it the formula C H ^ 'H^O. The analyses
\ OK
made of this salt by myself agree with the formula first
given it by Fritzsche, as seen from the following results.
Amount of salt taken K^SO^ found % Y^ found Theory
1.1583 0.5387 20.86 20.96
1.1251 0.5230 20.83
RUBIDIW I RTHON I TRO PHENOL AT E - This salt has not been de-
scribed. It crystallizes from water in large beautiful
plates resembling very much the co-rresponding potassium
salt above, except that it is just slightly lighter in col-
ID A, 110 - 153.
8
or than this salt. It crystallizes well in orange-yellow
crystals which have a composition analogous to the corres-
ponding potassiiAm salt above, it having the composition
represented by the formula 0, II , 1/2 H as seen from
*" ^ ^ORb
the agreement of the following analyses with the theory.
Amount of substance taken Rs^SOy found % Rb found Theory
0.4010 0.2297 36.67 36.74
0,4563 0.2611 36.62
As v/ill be seen later this is the usual fonr. in
which this salt crystallizes, )xit not, however, the only
one. There is another form of this salt and the passage of
this last into the more common form is an interesting one.
Of this I will speak more fully later on.
CAESIUM ORTHONITROPHEi:OLATE . - This salt has not been de-
scribed before. It is very soluble in water but crystal-
lizes from this strong solution in large crystals, having
in this respect, the tendency shown by the orthonitrophen-
alates above of crystallizing well. It is one of the salts
with which difficulty was encountered in obtaining it pure.
The follov/ing two analyses made on different samples show
this.
t
Amt. of salt taken CSiSO^foand X Cs found
0.4341 0.2782 47.06
0,2992 0.1931 47.43
On purification by the method adopted above the fol-
lovang results were obtained.
Amt. of salt taken C^^So^ found X Cs found Theory
0.2990 0.1994 48,93 49,07
0.2829 0,1890 49.09
This salt differs in crystal form from the other ortho-
nitrophenolates above; it crystallizes neither in needles
nor in plates but in thick, massive crystals of a dark
scarlet-red color, resembling sodium orthonitrophenolate in
color. By reflected light the crystal faces show a slight-
ly greenish color.
In this series of four orthonitrophenolates two
crystallize without water of crystallization and two with
one-half molecule of water each. The first two, sodixom
and cresium orthonitroplienolates have, so far as I have ever
noticed no tendency to crystallize with water of crystal-
lization. On the other hand, it has been found possible to
crystallize potassivira and rubidiiim orthonitrophenolate with-
out water of crystallization. Fritzsche had early remarked
10
on the difference in color between sodium and potassium or-
thonitrophenolate and had actually tried to obtain the po-
tassium salt in the dark-red form corresponding to the so-
dium salt. He tried to do this by crystallizing the salt
from alcohol, but found that even from absolute alcohol the
salt crystallized in the lighter form and still contained
one-half molecule of water of crystallization.
On crystallizing the rubidium salt from water it was
observed that at times it would crystallize out in dark-red
needles resembling the sodium salt very much, though it was
not quite so dark as the latter. On setting the beaker
down on the desk the solution was probably disturbed, and in
contact Y/ith the mother liquor began to change into the
lighter, ordinary form of the salt. On dissolving the salt
by warming the solution it came do^vn again in the dark v«Aai,
But it was found that the salt in this form had such a ten-
dency to pass over into the lighter vi^-I^*^ as to make it im-
possible to filter it off and obtain it in dry condition.
When sligjitly disturbed by touching with a stirring rod the
change begins at the point of disturbance and passes over
the whole mass of crystals. Further, the change can be
brought about by sprinkling a small amount of the lighter
crystals on the red crystals when they have separated out
11
from the solution. By considering the colors of the ortho-
nitrophenolates described above we might expect this darker
yft/iuct^ of the nibidium salt to be anhydrous as is the case
with the sodium and caesium salts which it resembles in
color, the change of color being brought about by the salt
passing to the hydrated condition which is the usual (U*.^twrr,
in which it separates out from water solution.
In order to obtain the salt in pure condition then,
if this were true, it would be necessary to exclude water
from the solvent used. For this purpose absolute alcohol
was chosen as the best solvent. Crystallization was effec-
ted from absolute alcohol, but as the time for evaporation
at ordinary temperature is so great that the substance be-
gins to change because of absorbing atnospheric moisture,
the evaporation was hastened by keeping the solution warm.
Under these conditions the salt crystallizes out in red
needles, lighter in color than sodium orthonitrophenalate
but is much darker than the common foinn of rubidium ortho-
nitrophenolate. The salt has the formula C(,Ha^
as seen from the following analysis.
Amt. of salt taken Rb^ SO^ found % 'Rh found Theory
0,2331 0.1397 38.39 38.23
That the increased temperature can have nothing to
12
do with the change of color is shown by the fact that the
same result can be gotten without the aid of heat, by simp-
ly causing the evaporation to take place in an atmosphere
which is kept dry,
I have obtained sodivmi, rubidium and caesium orthoni-
trophenolate in the dark red Tt>iLi£., The potassium salt,
alone, of those investigated here, remains and under ordi-
nary corcvimstances shows no tendency to crystallize in the
(1)
darkei-TT'u'v-tC,. It was this salt which Pritsche had tried to
crystallize in the darker form and for this purpose he tri-
ed to obtain the desired result by crystallizing from abso-
lute alcohol and presiimably in an open vessel as no mention
is made of a different procedure. As a result he found
that the substance crystallized constantly in the lighter
b-»>iia^ and always contained its half molecule of water.
In using the same solvent as Pritsche had used it
was soon found that alcohol simply distilled over quick-
lime still contained enough water to transfoxTn the salt at
least to some extent into the hydrate. So that alcohol
which had been distilled over sodium was used. Some potas-
sium orthonitrophenolate was dehydrated by heating in an
air-bath to 120 and dissolved in hot, absolute alcohol and
(1) A, 110 - 103.
13
quickly filtered into an Erlennieyer flask, while still hot.
The flask and its contents were kept almost at its ]:)oiling
point and air, dried by passing through two wash-bottles
containing sulphuric acid, was drawn through the flask.
The crystals which separated at first were red and very
much like the darker rubidium salt above. As evaporation
went on the crystals which had separated began to turn
lighter and before evaporation was complete only the yellow
\}vu4jL remained. The red form of this salt was not gotten
pure until evaporation of the solution was effected in a
current of air dried over phosphorus pentoxide. Under
these circumstances the salt crystallizes from absolute al-
cohol in small red needles similar in appearance to the ru-
bidium salt eiitzU..*^ in a similar manner. The remarkable
fact brought out by this experiment is that this salt has
the pov/-er to take up water from air dried over sulphuric
acid. It v/as found by analysis to be the anhydrous salt as
seen below.
Amt of salt taken. IQ SO,, found X K found Theory.
0.1915 0,0935 21.88 22.00
14
Metanitrophenolates.
SODim.^ MET AIIIT ROPHENOLA TE - This salt crystallizes in
rathei* small needles of an orange-red color, resembling
somewhat in color potassium orthonitrophenolate. It was
found on analysis to have the composition represented by
/NO,
the formula C, H^ H as seen below,
Amt of salt taken Na^SOc). found X Na found Theory
0.2190 0.0771 11. 4f 11. 67
POTASSIUM LIETAKITROPHENOLATE - This salt like sodiiim meta-
nitrophenolate crystallizes in small needle-shaped crystals
but are much lighter in color than the latter. They are
orange-yellow. On analysis it was found to have the compo-
/NO^
sition represented by the foriruila C, }I ^ HO.
\ OK.
Post
and Mehrtens giving it the formula
NO^
Amt of substance taken K^SO^ found X ^' found
Theory
0.3206 0.1206 16.89
16.88
0.2793 0.1059 17-0:
15
RUBIDIUM HETA NIT ROPHEMOLATE - Two salts of rubidium with
metanitrophenol have been isolated, one the neutral salt
the other an acid salt. The difference in color between
these two substances is very marked and is suggestive of
the difference in color observed in the two v»AU*i«< of rubid-
ivun orthonitrophenolate above. But the difference here is
due to another cause.
The lighter of these tvifo is the neutral s-^lt which
vas shown by analysis to have the formula
^ORb
It crystallizes in small yellow needle-shaped crys-
tals. Its analysis gave the following results.
Amt of substance taken Ri»^SO„ found ^ Rh found Theory
0.3459 0.2065 38,21 38. 2Z
The acid salt mentioned above crystallizes in clus-
ters of well formed needle-shaped crystals radiating from
a common point. They are of a bro^vnish-red color by trans-
mitted light and brown by reflected light, and have the
composition represented by the formula
^ NO^ T'O-.
^ ORb OH
and do not contain water of crystallization. Its anal-
16
ysis save the following results.
Arnt. of salt taken Rb, SO,, found X Rh found Th-^ory
0.4031 0.1485 23.^8 23.5 6
CAESim.^ METAll^yrROPKENOLATE - This salt crystallizes in beau-
tiful, almost blood-red crystals which resemble caesium
orthonitrophenolate in color but is a clearer and more
brilliant red than this latter salt. It is, like one of the
rubidium salts above, an acid salt having the composition
as seen from the
Amt. of salt taken Cs^SO^ found /o Cs found Theory
0.2854 0.1267 32. -^fc 32.44
Paranitrophenolates
represented by the formula
NO^
,N0,
^ OCs
OH
following analysis.
SODIim PARAWITROPIIENOLATE - This salt crystallizes from
water in large lemon-yellow prisms, which on standing ex-
posed for even a short time lose their transparency and be-
come opaque witho^^t any noticeable alteration of color.
This change has been studied and found as would be expected
that the change is due to loss of water of crystallization.
17
As the salt fii'st crystallizes it contains four mole-
cules of water of crystallization and on standing exposed
two of these are lost.
POTASSim/I PARANITROPKENOLATE - This salt differs from the
others by its lack of crystallizing power. On evaporating
slov/ly it was found to come down as a lemon-yellov/ crystal-
line powder. On analysis it was found to have the composi-
tion represented by the formula
NOv
Ct Hy
,^0K
,2-H^O
Amount of salt taken
kv SO^ found
% K f oiind
Theory
0.2592
0.1172
20.26
20.0
0.2323
0.1052
20/29
This salt has been made and analyzed by Post and
(1)
riehrtens who found it to have the composition
.NO^
C(,H^ ( 2H^0
^OR
RUBIDim^ PARAM IT ROPHEIIOLATE - This salt crystallizes from
water in large, bright yellow scales having the composition
/NO^
C^H^ H^O from its analysis below.
^ORb
Amt. of salt taken Rb^So^ found % Rb, Theory.
0.2111 0.1161 35.20 35.37.
0.2277 0.1250 35. If
(1) B, VIII. , 1552.
18
Rubidium paraniti"ophenol«te crystallizes from water also
in beautiful v/ell formed yellow needles which when filtered
off from the mother liquor begin to break up at once into
small parts, a change which may probably be due to change
in the amount of water of crystallization in a manner some-
what analogous to the change observed in the case of sodium
paranitrophenolate. The rapidity of the change prevented
obtaining the substance in a state of sufficient purity for
analysis.
CAESIUM PARANITROFKENOLATE - This salt crystallizes in
large lemon-yellov plates having the composition
G ^ H^ ^ 3H^0
^ OCs
Amt, of salt taken C^j^SO^ found % C6 found Theory
0,2234 0.1240 40.78 40.92
0.3371 0.1872 40.81
(1:2:3 ) Nitrocresolates.
Nitrocresol is a weaker acid than the nitrophenols and be-
ing a volatile substance its salts are more easily decom-
posed and it is harder to obtain them pure. This seemed
specially true of the caesiiam salt, solutions of which when
left to evaporate underwent continuous decomposition and
19
nitrocresol separating. This may have been caused by the
slow and continuous action of carbon dioxide of the atmos-
phere; the odor of nitrocresol being always noticeable from
solutions of its salts. On account of the weak acid prop-
erties of nitrocresol its salts are made from the hydrox-
ides of the metals and the purification of the salts is the
same as in the nitrophenolates.
SODIim NITROCRESOLATE - This salt crystallizes from water
solution in small, light-red needles having the composition
given by the formula
C(,H^ (NO, ) CH, . ONu. 2H^0 as seen from
the agreement of the following analyses with the percentage
required by this formula,
Amt. of salt taken Na^SO^ fo;md % No, found Theory
0.4165 0.1398 10.85 10.90
0.3672 0.1243 10.90
POTASSITO NITROCRESOLATE - This salt crystallizes in light-
red scales resembling in color the sodivun salt above, it
being of a slightly lighter color and has a composition
given by the formula
C, Hg (CH, ) (NO.^ ) OK 1/2 H,^
20
Its analysis is as follows:
Amt. of salt taken K^SO^ found % K foimd Theory
0.1752 0,0760 19.45 19.56.
0.3233 0.1418 19.66
RUBID i m.I NITROCR ESOLATE - This salt crystallizes in beau-
tiful cherry-red scales having a very high luster. By slow
evaporation the salt is deposited in well-formed monoclinic
diamond-shaped crystals of dark cherry- red color. It has
the composition
Cf^Hg.f CH3). ( MO J. ORb. H^O
Amt. of salt taken Rb^SO^ found % Kb found Theory
0.2127 0.1122 33.59 33.43
Caesium nitrocresolate was made but as has been sta-
ted above, at each recrystallization partial separation of
occurred
the nitrocresol, so that the salt was not obtained pure enough
for analysis. For some reason this salt was more unstable
than the others. It is however of a red color resembling
sodium nitrocresolate in this respect .
The conductivity of ortho- and paranitrophenol has
been determined by Hantzsch who gives the following tables:
21
Paranitrophenol (at 25)
V
64
128
256
512
^/
IT
0.
,89
1.
,28
1,
,79
2.
,53
Orthonitrophenol (at 25 )
V
128 1.13
256 1.52
512 2.17
1024 3.14
Those niambers differ considerably from those of Ba-
der who found numbers considerably greater than these.
The determinations of my own below differ somewhat
from those of Kantzsch and of SchiUmann, but all of the
three sets of determinations ap;ree fairly well for a sub-
stance of such low conductivity and they all show that or-
thonitrophenol is a slightly v/eaker acid than paranitrophe-
nol, while metanitorphenol is a somewhat stronger acid than
paranitrophenol.
22
Voluir;e
32
64
128
256
Paranitrophenol (temperature 25 )
, U^r
0.924
1.201
1.567
1.752
Metanitrophenol (temperature 25 )
Volume ,^,^
32 1.18
64 1 . 20
128 1.52
256 2.01
Volume
128
256
512
1024
Orthonitro phenol
1.02
1.85
2.39
3.77
From my own determinations meta- and paranitrophenol
are dissociated to about the same extent while both are
dissociated somev^hat more than orthonitrophenol.
23
The absorption spectra of some of the salts of ortho
and paranitrophenol were examined but nothing of any value
was found. These substances when in concentrated solution
were found to absorb all the violet end of the spectrum
even down, through the yellow in the case of orthonitrophen-
olates. On diluting, more and more of the violet end of the
spectrum was transmitted till at great dilutions only a
diminution of the intensity of the violet end was noticed
without at any time a tendency to show a bonded spectrum.
The only difference in the case of solutions of paranitro-
phenolates is that the light is not absorbed so far down
towards the red end of the spectrum as is the case with the
same dilution of an orthonitrophenolate.
The work of Carnally and Alexander mentioned above
was not at hand till the work had progressed considerably
and some of their conclusions have been strengthened by the
study of the additional salts made.
The authors mentioned above found: -
1) That without exception the color passes toward the
red end of the spectrum as the temperature rises.
2) That the color of the ortho compound is nearer the
red end than the corresponding para compound. Para-com-
pounds have a higher melting-point than ortho- compounds ,
24
the latter he concludes from this to be at a relatively
higher temperature and according to this should have a col-
or nearer the red end of the chromatic scale than the high-
er melting para-compound which accords with the first con-
clusion above.
3) By a comparison of the nitrophenolates of the same
subgroup the color passes towards the red end as the atomic
weight increases.
4) The color passes towards the red end of the chromatic
scale as the amount of water decreases.
Objections were raised when the paper was read, it
being thought that the conclusions were broader than war-
rented by the experiments.
Conclusion.
1) Speaking in a general way concerning the color of the
salts described above we may say of the three classes of
nitrophenolates that the orthonitrophenolates possess a co-
lor which is nearest the red end of the chromatic scale
while the metanitrophenolates come next and the paranitro-
phenolates are furthest removed from the red end.
2) There is a gradation of color in the case of each
series of nitrophenolates above, and in the direction point-
ed out in the third conclusion of Carnelly and Alexander
25
above, that is the color growing lighter in each series as
the atomic weight of the metal incl'^^ases. This, however, is
not very noticeable in the paranitrophenolates as only
little difference of color is observed in passing from one
salt to the other in this series.
3) In the same way water of crystallization causes lit-
tle change of color in the paranitrophenolates while the
change of color due to water of crystallization is greatest
with the orthonitrophenolates and next in order in this re-
spect come the metanitrophenolates.
4) Paranitrophenol and metonitrophenol have the power
of forming acid salts while orthonitrophenol has not been
found to have this power.
5) The orthonitrophenolates above when dehydrated are
all red as is the case with the metanitrophenolates thoiigh
these are lighter in color than the orthonitrophenolates.
6) Paranitrophenolates when dehydrated and cold are all
yellow except sodium paranitrophenolate which has a reddish
color; this last instance is in accordance with the fourth
generalization of Carnelly and Alexander, above.
7) The nitrocresolates are all red though they are not
so dark as the orthonitrophenolates. With the nitrocreso-
lates the variation in color from one salt to another is
26
less than with ortho-and metanitrophenolates. Water does
not produce so much difference in color in the case of ni-
trocresolates which is ti'T.ie also of the paranitropheno-
lates.
27
PART II.
METASULPHAIvIINEBENZOIC ACID,
Metasulphamine benzoic acid has been made by several
workers and by the use of several methods, and the lack of
agreement which is found in these descriptions of the prop-
erties of the substance woiold justify a more careful study
of the substance when made by several methods. The point
of greatest difference is regarding the melting-point which
different investigators have found for the substance. Even
when using the same method different workers have found
melting-points which differ quite markedly from each other.
In fact, from these results it might fairly be doubted
whether the same substance had always been under investiga-
tion, though the methods employed would leave little room
for such doubt. Fox- this reason it was suggested to me to
prepare this substance by several methods and decide if the
same product is always obtained and if so to settle the
question concerning the melting-point of the substance.
In view of the interesting influence which contin-
ued heating has been found to exercise on parasulphamine-
benzoic acid and of similar work which has been done on
28
metasulphaminebenzoic acid itself the knowledge of a defi-
nite stateinent concex'ning the melting-point of metasulpha-
minebenzoic acid acquires more than ordinary significance.
Metasulphaiainebenzoic acid 'vas first obtained b^' Lim-
(1)
pricht and Uslar from inetasulphobenzoic diamide by heating
with potassium hydroxide. Under these conditions the car-
bamine group is saponified and the sulphamine acid precip-
itated on acidifying.
CONH, ^COOX
CcH, ^ + KOH = C,H^ ( NH^
The authors describe the substance as separating in
crystals resembling potassium chlorate in appearance. They
did not determine its melting-point but found the substance
()
to melt somewhere above 20n ,
(2)
Remsen and Palmer obtained the acid from metatolu-
enesulphon a.nide by oxidation with potassium permanganate.
They found the substance they obtained to melt at 235 (un-
corrected) at the same time suffering slight decomposition.
(3)
Moyes and V/alker have made the acid from metatolu-
enesulphon amide by oxidation with potassium f erricyanide.
(1) A, 106, 36, 37, 42-46.
(2) Am. Chem. Jour. , 4 - 144.
(3) Am. Chem. Jour. , 8 - 188.
29
The potassium salt of the acid they obtained >/as washed
with strong alcohol to free it from any unoxidized diamide.
The acid liberated from this salt they found to have a melt-
ing-point of 246-247 (uncorrected).
(1)
W. Jones has made the acid by three different meth-
ods; first, by a slight modification of the method of Lim-
pricht and Uslar, when he obtained an acid melting at 248 ,
By the second he obtained the acid by the saponification of
metasulphaminebenzoic nitrile with dilute caustic potash.
The nitrile itself he obtained from the diamide by treat-
ment with phosphorus pentachloride and, after distilling
off the phosphorus oxychloride, crystallizing the product
from warm water. The acid obtained in this way was the
same as that from the first method above. Of this second
method he says, "No method is better adapted to the prepa-
ration of the acid in pure condition than this one".
He obtained metasulphaminebenzoic acid by treatment
of metasulphobenzoic chloride with aqueous ammonia. The
acid obtained by this method he found to melt at 248 -
248. 5.
(2)
Griffin obtained metasulphaminebenzoic acid by ox-
idation of metataluenesulphonamide with chromic acid. On
purifying he obtained a substance melting at 233 (uncorrec-
(1) Private Communication. (2) Am.Ghem. Jour., 19-180.
30
ted) .
(1)
Nakaseko, working in this laboratory, next made the
substance. He obtained his acid by the method recommended
by Jones as being the best, the saponification of metasul-
phamine benzonitrile. After recrystallizing the acid sev-
eral times he fotmd it to melt at 237°- 238° (correct ed) .
He then converted the acid into the anmonium salt and re-
crystallized this. The purified ammonium salt was convert-
ed into the lead salt and this in turn recrystallized sev-
eral times. The acid was liberated from the lead salt by
treatment with sulphuric acid and then recrystallized sev-
eral times. During this treatment he found the melting-
point had remained unchanged at 237 - 238 .
It v/as because of this variation that the following
work was undertaken.
Metasulphobenzoic diamide was made from acid potas-
sium, metasulphobenzoate by treatment with phosphorous pen-
tachloride, Metasulphobenzoic dichloride thus obtained is
treated with aqueous ammonia when the diamide is obtained.
This diamide is saponified with caustic potash by the meth-
od used by Jones and the product acidified with hydrochlo-
ric acid. When the metasulphaminebenzoic acid crystallizes
(1) Dissertation, 1S98.
31
out the precaution mentioned by Jones is observed, that is
not to allow the solution to become perfectly cold, other-
wise the unchanged diaraide will separate alon^ with the
sulphamine acid. The substance thus obtained was recrystal-
lized four times observing the same precaution. The sub-
o
stance then h-id a melting-point of 238 , v/hen recrys -
tallized a fifth time in the same way^ after boiling with
o
animal charcoal, a melting-point of 237,5 was obtained.
Some of this product was then converted into the barium salt
by treatment with barium carbonate and the bari-um salt re-
o
crystallized from water and repeatedly washed with 95 al-
o
cohol, recrystallized, warned with 95 alcohol and after
cooling, washed with alcohol and again recrystallized from
water. From the salt thus purified the acid was liberated
by treatment with hydrochloric acid and the acid recrystal-
lized from hot water. The metasulphamine benzoic acid
crystallized in beautiful and perfectly white scales which
o o
melted at 257.5 - 238 .
The metasulphobenzoic diamide above is then treated
with phosphorus pentaciiloride which converts it into meta-
/Cl
sulphaminebenzonitrile hydrochloride C.U^ -^ ' "^ NH which
SO^NH,
on dissolving in warm water forms meta sulphaminebenzoni-
32
CN
trile C| H^ This substance is boiled five hours
with dilute caustic potash. At the end of this tine there
o
separates out an acidifying with hydrochloric acid, a volu-
minous precipitate which does not crystallize anything like
the rnetasulpharnine-benzoic acid. It can be separated by
fractional crystallization into a crystalline portion from
which the sulphamine acid can be obtained in pure condition.
Not a very good yield of the sulphamine acid was obtained
and the large amount of precipitate formed may have been
the acid potassium salt which came do'jvn on acidifying,
though this was not determined. After six recrystalliza-
tions the acid had every appearance of the metasulphamine-
benzoic acid made by the method above and was found to have
o o
a melting-point of 237 - 238 .
Crude parataludine metasulphonic acid was purified
by boiling the solution acidified with hydrochloric acid
and adding animal charcoal. Mitrogen trioxide was conduc-
ted into a thin paste made by mixing this with alcohol ac-
(1)
cording to the method given by Griffin. The diazo- compound
thus obtained was decomposed in absolute methyl alcohol in
presence of anhydrous sodium carbonate according to the di-
rections of the same author. After the decomposition of
(1) Am. Chem. Jour., 19 - 164.
33
the diazo- compound is complete the alcohol is evaporated
off and the dried residue converted into metatoluene sul-
phon chloride by treatment with phosphor0uspentachloride
and this on treatment with aqueous ammonia is converted into
metatoluene sulphonamide. This amide was oxidized with
chromic acid according to the directions given by Griffin.
To free the acid formed from the unoxidized taluene sul-
phonamide it is converted into the potassiijun salt by treat-
ment with a solution of potassium carbonate and filtering.
On liberating the free metasulphSmine benzoic acid and re-
crystallizing several times it was found to have all the
appearance of that obtained by the methods above and was
found to have a melting point of 237.5 - 238,5 .
The metasulphamine benzoic acid was next made from
parabroiaemetataluenesulphonic acid. The reduction of this
acid has been accomplished by Miiller by treating its aque-
ous solution with sodium amalgam. For the complete removal
of the bromine the action is continued for about eight days
At the end of that time it is found that after precipitat-
ing the .bromine with silver nitrate no bromine is left un-
reduced. The solution is then acidified with sulphuric
acid and evaporated when by allowing it to cool a great
deal of the potassium sulphate can be eliminated. The
(/) A. Z^?- SI
34
solution which is this obtained from the sulphate which has
separated, is then evaporated to dryness and treated with
phosphorus pentachloride as done by MiJliex-, the oily chlo-
ride after washing with water is converted into the amide
by treatment with aqueous ammonia. The metatoluene sulphoii-
amide thus obtained is oxidized with chromic acid as above
and the metasulphaminebenzoic acid obtained is purified in
the same way. The substance has every appearance of that
already obtained and was f oiind to melt at 237,5° - 238° , -The
acid was next obtained from the metasulphonchlorbenzoic
acid by treatment with aqueous aiimonia. It may not be out
of place to state that this chloride used was obtained from
benzoyl chloride by treatment with sulphurtrioxide. The
sulphonic acid group enters the position meta to *he car-
bonyl group and by an action not yet understood the above
substance is obtained. This as stated gives the metasul-
phaminebenzoic acid on treatment with aqueous antnonia and
was found to melt at 2<37,5° ,
The thennometer used in all this work was carefully
compared with two standard thermometers and the corrections
always applied.
There was one other point concerning the melting-
point of this acid which was tested.
35
It is stated by Jones that his metasulphaminebenzoic
o
acid could not be made to melt below 248 . Me kept a small
quantity of the acid at a temperature of 240°- 242 for an
hour; the temperature was then gradually raised and while
o o
doing so was kept at temperatures from 242 - 246 for three
hours, but the acid would not melt below 248^ On the other
hand, Nakeseko found quite different results when proceed-
ing in a similar way. While he found the melting point of
the metasulphaminebenzoic acid to be 23y - 238 he says it
can be melted qiiite lower than this.
After drying the acid carefully in an air-bath at
o
140 for two hours it was placed in specimen tubes and
o o
heated for four hours at 210 - 215 . At the end of this
time it did not melt, but by raising the temperature slowly
o
till it reached 220 it melted completely.
I placed about one gram of the dried metasulphamine-
benzoic acid in a specimen tube and heated it for one hour
° o ^ . .
at 215 , once the temperature rose to 220 for a short time
and at the end of this time the substance had melted. This
result is similat to that obtained by Nakeseko above.
This being the case, in determining the melting
points of the metasulphaminebenzoic acid made by the dif-
36
ferent methods above, the substance was heated as rapidly
as possible till near the melting-point so as to avoid any
change of the melting-point , such as was found above, to be
caused by continued heating.
37
Conclusion.
I. Metasulphaminebenzoic acid made from the following
substances by the methods outlined above is the same sub-
o o
stance meltxng at 237 - 238 ,
1)
CONH,.
From Metasulphobenzoic Diamide C^Uq (m)
SO^NH^
_. CN
2) From Metasulphaminebenzonitrile C(^H,/ v^ (m)
,,COOH
3) From Metasulphonchlorbenzoic Acid 0,U^ (m)
^SO^Cl
4) From Parabromt^metatoluenesulphonic Acid
H^
<^ . S Oj H
c
Br
5) From Paramidometato luenesul phonic Acid
(M M,
II. The melting-point of metasulpliamine benzoic acid can
be considerably lowered by continued heating at a temper-
ature near 215 .
38
Biographical,
The author was bom near Lexington, Kentucky, Octo-
ber 30, 1S75. He received his early education in the pub-
lic school near Lexington and then entered the State Col -
lege at Lexington, Kentucky, from which he graduated in 1S97
receiving the degree of Bachelor of Science, and Master of
Science the following year; he then entered the Johns Hop-
kins University where he has since pursued studies in Chem-
istry^ with Physics and Physical Chemistry as subordinate
studies. He obtained an Honorary Hopkins Scholarship in
1900 and was appointed fellow the following year.
V' /"^' /^y
->'>*:/•
^°«'.i«
^'^^f^Z::.m^
i'..
«!M',^^'' '■
:^6M>ri^
Ix Mfi
H>' ,.
f ^ '^8-, /:
,,, J'/^-^
t
'jjy
.-7 ■
Si '
■''■'■'Vq:''''- ! ••■ ■
