ium
-
S FOR THE
lu
O
s of C
PA.
THESIS
N,
mb
OF THE GRADUATE SCHOOL
QUIREMENT
DEGREE OF DOCTOR OF PHILOSOPHY
BY
aSe
PHILADELPHIA
， ， ， .-.-
· ·, e , * *： ：
…！-
！º
GERMANTOW
THE JOHN C. WINSTON CO.
li
FULFILMENT OF THE RE
OF THE UNIVERSITY OF PENNSYLVANIA IN PARTIAL
T 3.
' > *
ſ=
-]
$2
Ci-
CCla
±
|-
-· O
(l) {| ſ==
~~）; Q
[×]
ſ-
Z.
[r]
-(/)
ſrl
[×
n-
|-$§§
！ *
§@₪§§§-|------· - ----·-
§§*,\！---- --·-·--
|-··ë:ºsºſ，-·---·- --*.*）
§§$$
∞§§
¿¿-??¿ $§§§-- ： * …---- **. ::--~~~ ~ ~ ~;~~§§|×-ſae~~~~·-*
-->sº： ，，№·|---.•:---- ~- - ..|----- - --§§§§§§- - - - --×--！！！！ -- -|- *… * *.- --§§

"I'm
-: * I .
* * *
* * *
f : , ,
■**
* . . .
# º','!'."
'', , , , ,
, ,
ſaeſ
§§§
¿№ ž.）;'）;
、、、。
șºğrgº; łOE ș «
¿
（5））
3}}§.
§§§§
�ț¢º º
ſaeſ）
§§§§
ºº！،
§¶√∞'，-·
-！ 3-，§§ſae
¿），··
·§§§:
ſae
§: ***
§§*（）;、、、、、、、、、、、）
§¶√≠√∞ſe ºff，¿șº: º№saeº.
§§§§§§§
¿
«№

Chemical Libay
G. W.
\% \
.Nº
\\? $*\
Halide Bases of Columbium
THESIS
PRESENTED TO THE FACULTY OF THE GRADUATE SCHOOL
OF THE UNIVERSITY OF PENNSYLVANIA IN PARTIAL
FULFILMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY
BY
HERBERT S. HARNED
GERMANTOWN, PA.
1913
PHILADELPHIA
THE JOHN C. WINSTON CO.
1913
ACKNOWLEDGMENT.
The author wishes to express his sincere thanks to
DR. EDGAR F. SMITH, under whose suggestion and direction
this work was carried out.
`
*S
**
:
INTRODUCTION.
In 1907 Chabriël reported a lower hydrated chloride
of tantalum to which he gave the formula TaCl2.2H2O.
He obtained this substance by heating 7.2 grs. Of tantalum
pentachloride with 15.2 grs. Of 3% sodium amalgam in a
tube of hard glass which was sealed at One end and evac-
uated by means of a suction pump. The reaction mass
was allowed to cool completely in a vacuum and then
treated with boiling water containing a little hydrochloric
acid. A deep emerald green solution was obtained and
from this, on evaporation, a green powder separated. This
was dried and analyzed. From the figures obtained by the
analysis, Chabrië felt justified in assigning the formula
TaC12.2H2O.
In 1908, Van Haagen” was working with the halides
of tantalum and the products of reduction obtained by
passing tantalum pentabromide through a red hot tube in
a current of hydrogen. In this latter reaction, he invariably
noticed, beyond the metallic deposit, a greenish black sub-
stance which was soluble in water to form an intense emerald
green Solution. Every attempt to produce by this method
a sufficient quantity of this substance for investigation
proved of no avail.
During the winter of 1908–1909, Chapin8 continued
this work. He decided to use the Chabrië method of reduc-
tion on tantalum pentabromide with the hope of being able
to obtain an amount of the green compound sufficient to
be investigated. For this purpose, he prepared 200 grs.
Of tantalum pentabromide. This was mixed, 10 grs. at a
time, with four parts by weight of 3% sodium amalgam,
and heated in an evacuated tube until reaction set in. The
temperature of the tube was then raised to a bright red
* Compte rend. (I907), I44, 804.
* Thesis—University of Pennsylvania, 1908.
*Journ. Amer. Chem. Soc., 32 (1910), 323.
(3)
4.
heat and the tube was allowed to cool to room temperature
with the vacuum pump still in use. The reaction mass
was boiled with water containing a little hydrobromic acid.
An intense emerald green solution similar to that described
by Van Haagen was obtained and, on evaporation, black
crystals separated. These were filtered off and dried.
From 200 grs. of tantalum pentabromide, only 20 grs. of
these crystals were obtained. When ground to a powder,
this substance had a green color. It was very soluble in
water, the Solution possessing a remarkable Coloring power
(1 gr. in 100 cc. of water giving such an intense coloration
that a layer half an inch thick was opaque). Ammonia
precipitated a brown hydroxide from this solution.
There was a doubt at first, on account of the small
yield, whether this was a compound of tantalum at all.
Four grs. were converted to the double fluoride. Analysis
of this compound proved without a shadow of doubt that
the metal contained in the green compound was tantalum.
Chapin carried out a number of analyses which indicated
that this body possessed the complex structure represented
by Tag Brix] Brº.7H2O. Molecular weight determinations.
by both the freezing and boiling point methods also indicated
the above constitution. Only one-seventh of the bromine
acted as ion. The tantalum seemed to be present in a com-
plex cation.
From this compound Chapin was able to prepare the
compounds indicated by the following scheme:
[Tag Brial Cl2.7H2O
i à
--- (TaBril(OH), IoH,0.
! E:
[Tagbriz]I2.7H2O
(TasRTlal Brº.7H2O NaOH equivalent to
two bromine atoms.
He also prepared Chabriè's chloride from tantalum penta-
chloride and proved that this was [TagCl2]Cl2.7H2O.
5
The object of the present investigation was to find out
whether these compounds were peculiar only to tantalum
and whether it was possible to obtain similar compounds
with any of the other members of the fifth group. Molyb-
denum in group six furnishes compounds of similar but not
analogous constitution e. g. [Mo;C14]Cl2.6H2O, [Moabral-
Br3.6H2O. Columbium on account of its great similarity
to tantalum and its position in the periodic system, prom-
ised best. In this investigation, the Chabriè reaction has
been used in an attempt to prepare a similar lower chloride
of columbium.
PRELIMINARY WORK.
The steps in the preliminary work were: (1) Purifi-
cation of the double fluoride of columbium and potassium.
(2) Conversion to oxide. (3) Preparation of columbium
pentachloride. It will not be necessary to go into details.
Reference to the papers of Hall' and Balke" will be sufficient.
The pentachloride was obtained by passing vapors of Sul-
phur monochloride and chlorine gas over the moderately
heated oxide; 200 grs. of the three times distilled penta-
chloride were obtained. Analysis of a 2 gr. Sample gave
a very satisfactory result.
REDUCTION PRODUCTs of PENTAVALENT COLUMBIUM.
It has been known for a long time that columbium
pentachloride dissolves in concentrated hydrochloric acid
and that, on addition of metallic zinc to this solution, reduc-
tion occurs, giving first an intense blue and finally, on con-
tinued reduction, a brownish black solution. It was thought
that strong hydrochloric acid might precipitate from these
Solutions a compound similar to that prepared by Chapin.
It was found that a blue and a black powder could be pre-
cipitated by concentrated hydrochloric acid but neither of
these compounds answered the description of Chapin's
* Hall and Smith: Proc. A. Phil. Soc., 44, 177 (1905).
* Balke and Smith: Journ. Amer. Chem. Soc. (1905).
6
compound since they readily oxidized, and since they gave
the corresponding blue and dark brown oxides on dilution
with water. Incidentally, it was found that when the
hydrochloric acid solution of columbium pentachloride was
boiled, a white powder separated which dissolved on dilution.
This substance was evidently related to the blue and black
powders precipitated by concentrated hydrochloric acid
from the reduced solutions, for both of these, if kept in
strong hydrochloric acid, oxidize to the white powder. The
following table will show to relations of these compounds
to one another:
Valence. Concentrated HC1. d.º.º. Diluted and Boiled.
V. . . . . . . . . . . . . White powder... Colorless solu- | White oxide.
tion.
III. . . . . . . . . . . . Blue powder. . . . . Blue solution . . . . Blue oxide.
II–III. . . . . . . . . Brown black Brown black Brown black
powder. solution. oxide.
These substances were all very easily decomposed and
the conditions of reduction seemed to exclude the possibility
of preparing a stable hydrated chloride of columbium by
this method. It was thought that if this substance was
to be prepared at all, higher temperatures would be neces-
sary. This attempt was abandoned and it was decided to
try the Chabriè reaction.
REDUCTION OF COLUMBIUM PENTACHLORIDE BY MEANS OF
SODIUM AMALGAM AT HIGH TEMPERATURES.
The details of the method and the form of apparatus
used in this reduction are given accurately in Chapin's paper,
so it will not be necessary to describe them here. It seemed
desirable to investigate a little more closely the temperature
conditions. To this end, 10 grs. of columbium pentachloride
were mixed with 70 grs. of 3% sodium amalgam and allowed
to react at a temperature just high enough to cause the
7
reaction to propagate itself entirely through the mixture. On
treating the reaction mass with water a brownish black oxide
was obtained and a brown solution which quickly decomposed
giving more of the brown oxide. On repeating this experi-
ment and on heating the first product of the reaction to a
red heat for a period of one hour, then allowing to cool to
room temperature without breaking the vacuum, and finally
boiling the mass with water, an intense olive green solution
was obtained and, at the same time, a precipitate of the
brown oxide. Two reactions seemed to take place here:
firstly, at lower temperatures, the columbium pentachloride
was reduced to a chloride which was easily decomposed to
the dark brown oxide on the addition of water, secondly,
this chloride, at higher temperatures was partially converted
into the substance which gave the green solution. The
olive green solution was stable and promised to be very
similar to the emerald green solution obtained by Chapin;
150 grs. of columbium pentachloride were reduced by this
method, the most satisfactory temperature used being just
sufficient to collapse the Jena tube. The substance which
gave the olive green solution was only soluble in boiling
water and a large quantity was needed for its complete
solution. The hot solution was filtered from the brown
oxide and evaporated after the addition of a small quantity
of hydrochloric acid. Small black shiny crystals sepa-
rated; these were filtered off and recrystallized. The
crystals obtained from this second crystallization seemed
perfectly homogeneous and, since they were so difficultly
soluble in water, it was thought that a high degree of purity
must have been reached. They were filtered on to a Munroe
crucible and dried in an air bath at 80° C. to constant
weight. From 150 grs. of columbium pentachloride only
6 grs. of this stable lower chloride were obtained.
PROPERTIES OF THIS COMPOUND.
This substance could be heated to 100° C. without
decomposition. The crystals were so small that the crystal-
8.
line form could not be observed. When powdered the sub-
stance was olive green in color. It was insoluble in cold
water, but soluble in boiling water, giving an intense olive
green colored solution. On addition of ammonia to this
solution, little or no decomposition took place, and, even
after long-continued boiling with ammonia, it was not com-
pletely decomposed. Chapin's bromide was completely
decomposed by boiling ammonia. On the addition of con-
centrated nitric acid to a boiling solution of this compound,
decomposition took place, giving first a brown coloration,
and finally the Solution oxidized to the pentavalent form
and columbic hydrate separated. The compound was com-
pletely soluble in concentrated alkalies giving a dark brown-
ish green solution.
ANALYSIS OF THE OLIVE GREEN CHLORIDE.
Since the chloride was not easily decomposed by boil-
ing ammonia, the following method of analysis was adopted:
The chloride was dissolved in boiling water, concentrated
nitric acid was added to the boiling solution until complete
decomposition took place. This required about one minute.
An excess of ammonia was then added and the solution
was again boiled for a few minutes. The columbic hydrate,
thus completely precipitated, was filtered off, washed with
boiling water, ignited and weighed. Ammonium carbonate
and nitrate solutions can not be used in washing columbic
hydrate precipitated in this manner, because they dissolve
quite a considerable quantity of it. On washing the pre-
cipitate with boiling water, a small quantity of the hydrate
invariably dissolved forming a colloidal solution. This was
boiled with a little more nitric acid and again precipitated
with ammonia, filtered off, ignited and weighed. The
filtrate from the columbic hydrate was acidified with nitric
acid and the chlorine was precipitated in the usual way
with silver nitrate, filtered on to a Gooch and weighed.
The following results were obtained and all through the
9
following discussion on the constitution of this chloride
will be compared with the results calculated from the
formula [Ch;C112]Cl2.7H2O, analogous to [TagBria]Br2.7H2O:
WT. SAMPLE=0.2Ooo
Obs. Calc. Obs. Calc.
Wt. CbzOs . . . . . . . . . . O. I349 O. I353 %Cb 47. 25 47.38
Wt. AgCl. . . . . . . . . . . O. 33.95 O. 3394 %Cl 4I .99 4I .98
It was thought that it would be well to determine the
columbium in this compound by two methods entirely dif-
ferent from the above: (1) To ignited a weighed sample
with nitric acid and obtain columbic oxide directly. (2)
To decompose the substance with concentrated sulphuric
acid, heat to fumes, dilute with water and boil, filter off the
precipitated columbic hydrate, wash with boiling water
until free from sulphuric acid, and finally ignite to the
oxide and weigh. The following results were obtained:
WT. SAMPLE=0.2000
Obs. Calc. Obs. Calc.
Method I. . . . . . . Wt. CbzOs o. 1342 o. I353 % Cb 47. Oo 47.38
Method 2 . . . . . . . Wt. CbzOs o. 1349 . . . . . . % Cb 47.25 . . . . .
DETERMINATION OF THE IONIZED CHLORINE.
This determination seemed at a glance to afford con-
siderable difficulty. The determination of the ionized
chlorine by the gravimetric method would necessitate the
precipitation of silver chloride from a cool solution not
containing nitric acid, for, if nitric acid be present, decom-
position of the complex cation would be likely to occur.
If a hot solution be used, a high chlorine result was to be
expected, due to the Secondary dissociation of the complex
radical. Chapin found in the case of [Tag|Brig) Brº.7H2O
that dilution affected the result considerably. Since 0.2 gr.
of the chloride of columbium required at least 100 cc. of
water for its solution, high results were to be expected It
was thought best to titrate the ionized chlorine with a
10
0.1 N silver nitrate solution using a silver concentration
cell." The E.M.F. of the cell will remain nearly constant
as long as chlorine ions are present in the solution. After
all the chlorine ions have been removed, the E.M.F. will
rise rapidly due to the rapid increase of silver ions in the
solution. The apparatus used was identical with that
described by Hildebrand and Harned" in their paper on the
determination of magnesium by means of the hydrogen
electrode with the single exception that a silver coated
platinum electrode was used in place of the hydrogen
electrode.
The determination of the ionized chlorine in the chloride
was carried out as follows: 0.2 gr. sample was dissolved in
150–200 cc. water and the solution was allowed to cool to
room temperature. The silver electrode and calomel elec-
trode were then placed in the solution. Leakage of the
solution of potassium chloride in the calomel electrode into
the columbium chloride solution was found to occur. To
prevent this, a solution of potassium nitrate, free from
chlorides, was drawn up into the side arm of the calomel
electrode. This proved entirely satisfactory. 0.1 N silver
nitrate was then added, 0.2 cc. at a time and a reading of
the E.M.F. of the cell taken after each addition. The
curve (see page 11) was obtained by plotting E.M.F. of the
cell against ce. of 0.1 N silver nitrate. The inflection in the
curve shows that 3.9 cc. of 0.1 N silver nitrate were used to
titrate the ionized chlorine.
WT. SAMPLE=0.2OOO
Obs. Calc. (one-seventh total chlorine)
Grs, ionized Cl. . . . . . . . . . . O. O.I.38 O. OI2O
This seemed to indicate that one-seventh of the total chlorine
in this compound was ionized in Solution. This was exactly
analogous to the behavior of Chapin's bromide.
* Behrend: Zeitschr. Phys. Chem., II, 426 (1893).
7 Hildebrand and Harned: Communications, Vol. I, Eighth
International Congress of Applied Chemistry.
11
The contents of the cell were then washed into another
beaker, the silver and calomel electrodes were washed and
the columbium and total chlorine were determined by the
EMF
jj
030 H.
jitſ#Rºnſ, -0.053, Cſ.
070
(). Iſ)
| | | |
000 1.00 200 j00 400 500
ſ: H9N0,
method outlined above, namely, decomposition by means
of concentrated nitric acid and ammonia. The following
results were obtained:

12
WT. SAMPLE=0.2Ooo
Obs, Calc. Obs. Calc.
Wt. CbzOs . . . . . . . . . O. I358 O. I353 % Cb 47. 55 47.38
Wt. AgCl, . . . . . . . . . O .3396 O. 3394 % C1 42. OO 4I .98
WATER DETERMINATION.
The method used for the determination of water in
this compound was the same as that used by Chapin to
determine water in his bromide. The chloride was in-
timately mixed with about 10 times its volume of ignited
1ead oxide and heated to 300° C. in a tube in a current of
dry air. The water was carried over into a calcium chloride
tube and weighed. For further details, see Chapin's paper.
The following results were obtained:
WT. SAMPLE=0.2OOO
Obs. Calc. Obs. Calc.
I. Wt. H2O . . . . . . . . . O. O.224 O. O2 I3 %H2O II .. 2 IO. 64
2. Wt. H2O . . . . . . . . . O. O2 I& o.O2.I.3 %H2O IO. 9 IO. 64
The results obtained by this time show conclusively that
this compound is analogous to [TagBria]Br2.7H2O and that
its formula must be [CB6Cl12]Cl2.7H2O.
NOMENCLATURE.
Chapin gave to his compound the name “bromo-
tantalum bromide” and to the complex cation the name
“bromo-tantalum.” The other compounds he called by
the corresponding names, e. g., bromo-tantalum hydroxide,
bromo-tantalum chloride. The same nomenclature will be
adopted here and throughout the remainder of this paper,
the chloride will be referred to as chloro-columbium chloride.
PREPARATION OF CHLORO-ColumBIUM HYDROxIDE.
One gr. of chloro-columbium chloride was dissolved in
as little water as possible and the Solution was allowed
to cool to room temperature. Just enough 0.1 N sodium
13
hydroxide was added to replace the two ionized chlorine
atoms with hydroxyl groups. On standing a black micro-
crystalline body separated. This was filtered on to a Mun-
roe crucible, washed with cold water and then with ether.
It was dried in a desiccator over calcium chloride. One gr.
of chloro-columbium chloride gave 0.6 gr. of this supposed
chloro-columbium hydroxide.
PROPERTIES OF THE HYDROxIDE.
This compound was insoluble in water but soluble in
both acids and alkalies. The acid Solutions were olive green
in color and probably contained the corresponding salts
(e.g., chloro-columbium nitrate, chloro-Columbium sulphate).
The alkaline solution was dark brownish green in color and
probably contained [ChéC112](ONa)2. aq. but attempts to
isolate anything from this solution stable enough to analyze,
proved fruitless,
ANALYSIS OF THE HYDROxIDE.
The method used was essentially the same as that
used in the analysis of the chloride. The weighed substance
was washed into a beaker and about 100 cc. of water were
added. The solution was brought to boiling, concentrated
nitric acid was added and the boiling continued until the
nitrate, which was first formed, was completely decomposed
and oxidized. The Solution was made ammoniacal and the
analysis was continued in precisely the same manner as
described under chloro-columbium chloride. The results
seemed to indicate the formula [Ch;C11ol(OH)2.8H2O as
shown by the following:
WT. SAMPLE=0.2OOO
Obs. Calc. Obs. Calc.
Wt. Cb2O5 . . . . . . . . . O. I379 O. I376 % Cb 48.28 48. I5
Wt. AgCl. . . . . . . . . . O. 2998 O. 2956 % C1 37. O8 36. 56
This chlorine result was not very satisfactory. Therefore,
another 1 gr. sample of chloro-columbium chloride was
14
converted to the hydroxide by the method outlined above.
Analysis gave the following:
WT. SAMPLE=0.2Ooo
Obs. Calc. Obs. Calc.
Wt. CbzOs . . . . . . . . . O. I380 O. I367 % Cb 48.32 48. I5
Wt. AgCl. . . . . . . . . . o. 2968 O. 2956 % Cl 36.70 36.56
Water was determined as with chloro-columbium chloride,
only in this case, it was necessary to raise the temperature
of the tube higher in order to drive off the last traces.
WT. SAMPLE=0.2Ooo º,
Obs. Calc. Obs. Calc.
Wt. H2O. . . . . . . . . . O. O275 o.o.278 % H2O I3.75 I3.90
This analysis corresponds very closely to that calculated
from the formula [CbéCl12] (OH)2.8H2O.
Per cent. Calc. Obs.
Cbs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48. I5 48.32
Cliº . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36.56 36.70
9 H2O. . . . . . . . . . . . . . . . . . . . . . . . . . . . I3.90 I3.75
O (by difference) . . . . . . . . . . . . . . . . . . I. 39 I. 23
| PREPARATION OF CHLORO-COLUMBIUM BROMIDE.
Two-tenths of a gram of chloro-columbium hydroxide was
weighed out in a platinum crucible and was then dissolved in
dilute hydrobromic acid. The solution, was evaporated to
dryness on a water bath and the crucible and its contents
reweighed. It was to be expected that the bromide would
have an analogous constitution to the chloride and there-
fore be [Ch;C112 Brº.7H2O. The theoretical increase in
weight calculated from the ratio
[CbgCl2] (OH)2.8H2O to [CbsCliº) Bra.7H2O
equals 0.0185 grs. The observed increase in weight was
0.0186 grs. This is confirmation of both the formula of
15
the hydroxide and bromide. The bromide was similar in
appearance to the chloride but differed by the fact that it
dissolved in a small quantity of cold water. Ammonia
precipitated from the solution of the bromide small shiny
crystals, probably the hydroxide, but no decomposition took
place. The analysis of the bromide was carried out by the
same method as that used for the analysis of the chloride
and hydroxide. The following results were obtained and
agree closely with those calculated from the formula
[CbgCl2]Br2.7H2O : -
WT. SAMPLE=O.2 I86
Obs. Calc. Obs. Calc.
Wt. CbzOs . . . . . . . . . . . . . O. I377 o. I375 % Cb 44. II 44. O7
Wt. AgCl-HAgBr . . . . . . . o. 362 I o. 36O2 % C1-HBr 46.25 46.03
A water determination of the bromide was not made.
THE INVESTIGATION OF THE ALKALINE SOLUTION OF
CHLORO-COLUMBIUM CHLORIDE
I. Koppel” in a paper on “The Constitution of the
Halogen Compounds of Bivalent Molybdenum” mentions
the series of compounds prepared by Chapin and their
similarity to the above-mentioned molybdenum com-
pounds. Since (Moscl)Cl·6H2O is soluble in concen-
trated alkalies and since double salts of the type
(MoAC1)C13.2KC1.2H2O are precipitated from these sol-
utions by concentrated acids, the question is naturally
raised as to the constitution of the alkaline solution of
chloro-columbium chloride and whether double salts can be
formed by the addition of strong acids to this solution.
Chapin had mentioned that bromo-tantalum hydroxide was
soluble in concentrated alkalis and that the solution probably
contained Tagbriel(ONa)2, aq. He went no further than
this point with the investigation of the alkaline solution.
It has been mentioned that both chloro-columbium
* I. Koppel: Zeits. anorg. Chem.: Band 77, Heft 3 u. 4.
16
chloride and hydroxide were soluble in concentrated alkalies
giving a dark brownish green Solution. On evaporating
this solution in a vacuum, no Substance of a homogeneous
nature separated. On the addition of a large quantity of
concentrated hydrochloric acid, a brown powder was pre-
cipitated.
INVESTIGATION OF THE BROWN POWDER.
One gr. of chloro-columbium chloride was dissolved in
a 20% caustic potash Solution and to this, a large excess of
concentrated hydrochloric acid was added. The brown
powder, thus precipitated, was filtered on to a Munroe
crucible, washed with cold water, then ether and dried in a
vacuum desiccator over calcium chloride. Yield 1 gr.
PROPERTIES OF THE BROWN COMPOUND.
This substance appeared perfectly homogeneous.
When ground on a mortar, the color of the powder was
brown and was distinctly different from the olive green
color of chloro-columbium chloride. It was soluble with
greater difficulty in boiling water than chloro-columbium
chloride and its solution at first appeared a yellowish green
but changed on boiling to an olive green. From this olive
green solution, chloro-columbium chloride could be obtained
by evaporation after the addition of a little hydrochloric
acid.
ANALYSIS OF THE BROWN COMPOUND.
The same method of analysis was adopted as with
chloro-columbium chloride and hydroxide, There was a
possibility that this substance contained potassium and
was a double salt. To determine whether this was true, the
filtrate from the silver chloride precipitate, which contained
the excess of silver nitrate, was boiled with hydrochloric
acid and the precipitated silver chloride was filtered off.
The filtrate was evaporated to dryness, heated until the
17
ammonium salts were driven off, and the weight of the
residue was obtained. This amounted to only 0.3 mg.,
showing that the brown compound was not a double salt.
The following analyses were obtained and agree very closely
to the values calculated from CbgC114.9H2O.
- WT. SAMPLE=0.2Ooo
No. 1 Obs. Calc. Obs. Calc.
Wt. CbzO5 . . . . . . . . . O. I285 O. I3 I3 % Cb 45. OO 45.98
Wt. AgCl. . . . . . . . . . O. 332 I O. 3293 % Cl 4I. O.8 4O. 74
WT. SAMPLE=0.2OOO
No. 2 Obs. Calc. Obs. Calc.
Wt. CbzO5 . . . . . . . . . O. I3 I7 O. I3 I3 % Cb 46. IO 45.98
Wt. AgCl. . . . . . . . . . O. 33 I4. O. 3293 % Cl 4O. 99 4O. 74
WT. SAMPLE=0.2OOO
No. 3 Obs. Calc. Obs. Calc.
Wt. CbzO5 . . . . . . . . . O. I3 I4. O. I3 I3 % Cb 46. OO 45.98
Wt. AgCl. . . . . . . . . . O. 3322 O. 3293 % C1 4I. O8 4O. 74
... A water determination was made by the method previously
used and the following result was obtained:
WT. SAMPLE=0.2OOO
Obs. Calc. - Obs. Calc.
Wt. H2O. . . . . . . . . . . O. O.28O O. O.266 % H2O I4. O I3.28
Since on going into Solution the brown chloride was con-
verted to the green chloride, an ionized chlorine determina-
tion was not made.
These figures seemed to indicate that this brown com-
pound had the formula CbºCl4.9H2O. If the difference
between the brown chloride and the green chloride was due
only to two molecules of water of crystallization, simple
heating of the brown chloride to 100° C. should give the
green chloride and the loss in weight should indicate the
completeness of the reaction. Therefore, 0.2 gr. of the
brown chloride was heated to 100° C. but, at this tempera-
18
ture, it decomposed, giving off a white powder which coated
the inside of the crucible. At the same time, no change
from the brown to the green chloride was noticed. On the
other hand, the action of boiling water or the action of water
at room temperature for a long period of time (a week)
gave an olive green solution, from which the green compound
could be precipitated by evaporation with dilute hydro-
chloric acid. These experiments would indicate that there
was a constitutional difference between the brown and the
olive green chlorides.
Since the action of water converts the brown to the
green compound, it was thought that, if the solution of the
green compound be precipitated by concentrated hydro-
chloric acid, the brown compound could be obtained. A
powder was found to be precipitated by concentrated hydro-
chloric acid but, when this was dried and ground in a mor-
tar, it gave the same olive green color as [Cb5C112]Cl2.7H2O
and on analysis proved to be this compound as shown by
the following:
WT. SAMPLE=0.2Ooo
Obs. Calc. - Obs. Calc.
Wt. CbzO5 . . . . . . . . . O. I34O O. I353 % Cb 46.92 47.38
Wt. AgCl. . . . . . . . . . O. 3394 O. 3394 % Cl 4I. 98 4I. 98
The fact that the brown compound could be prepared from
the alkaline solution but not from the solution of [Ch;Clel-
Cl2.7H2O in water afforded additional evidence that there
was a constitutional difference between the brown and the
green chlorides.
The present investigation had to end at this point
because the material had been exhausted and lack of time
prevented the preparation of more. As a final proof that
the element contained in these compounds was columbium,
a portion of the oxide obtained from the analyses was con-
verted to the double oxyfluoride of columbium and potas-
sium and this was analyzed. For this purpose, 2 grs. of the
oxide were fused with potassium bisulphate. The melt
19
was boiled with water and the precipitated columbic hydrate
was washed on a filter with an ammonium carbonate solu-
tion until the filtrate gave no test for sulphuric acid. The
oxide was then dissolved in hydrofluoric acid and a calcu-
lated amount of pure potassium carbonate was added.
Plates of the oxyfluoride separated. These were recrystal-
lized with centrifugal drainage and finally dried in a desic-
cator over calcium chloride. A sample was weighed out
into a platinum crucible, carefully decomposed by concen-
trated sulphuric acid and finally heated to fumes in order.
to be certain that all the hydrofluoric acid had been driven
off. The contents of the crucible were then boiled with
water and the precipitated columbic hydrate was filtered
off, washed with boiling water until free from sulphuric
acid, ignited and weighed. The filtrate was evaporated
to dryness and the residual potassium Sulphate was ignited
and weighed. The following results were obtained and are
compared with the values calculated from K2CbQF5.H2O:
WT. SAMPLE=o.3 Io2
Obs. Calc. Obs. Calc.
Wt. Cbz O5 . . . . . . . O. I384 o. 1378 % CbzOs 44. 6 I 44.42
Wt. K2SO4. . . . . . . o. 1792 - O. I796 % K2SO4 57.77 57. 90
SUMMARY.
1. Columbium furnishes the compound represented by
the formula [Ch;C112]Cl2.7H2O exactly analogous to the
bromide of tantalum prepared by Chapin.
2. The table on the following page sums up the most
important characteristics and relations of the members of
this group of compounds: -
HC1
NaOH
[ CbóCl il Cl2. 7H2O e *
> [CB&Cliº) (OH)2. 8H2O
equivalent to 2 Cl
Olive Green (Sol. in boiling H2O) Black (Insol. in H2O)
O
O º
# .# ſº.
E. 3| c.
§ #| 3.
O 3. #| #
** 1. Cú {I}
;I. alkalies , [CbCliº) (ONa).aq . concentrated #
(Brownish green solution) •
e O
(Formula given is hypothetical) 3.
CbsC114.9H2O concentrated [Ch. ClijBrº. 7H2O
*
Olive green
(Easily sol. in H2O
Brown
Š

Xà，
¿?
§§
ș，
§§
§§§§§§
§§§§§
ſºſae;（&#§§-ſae
§§
¿
íč，
·§§§§-¿№
$（/^（<（}（ſ）\，\！;¿ș§§§§§-¿
§§§§§§§§§§§§§§§§§§ºwł.:-ſºg§§§§§！
§§§
§§§ğă
º:
} § x
#
$º
&
g
*
ſº
§:
*:）*?< */****¿§，:
|-。
§ 3：3，§§§§§
Ř■**¿
-¿
（§§§º,
§§
*、、。
§§;：? .·、&&
！ 3·&ș，ſae}§§***ſ.ſ， ſºº§§§§§§§§§§§§§ſaeae
* &çº, º;：. →- ~ -…-… （...*_--，。：
º?
§
§§
§:
#:
&
§<
�
* * ?.
|-Ř
3,3Ģ*|-
∞
∞∞∞
N
§§§
§§§
|
~，ģ¿?}，
≡ ≈≠≤ , **）:
§§§§§
¿
O
CO
* ¿
-、、。
Ķ
§
gº
~º.ſae;:§§§&}。º,
：k!：§§§§§§£
- ：：*， ** * *·、。、、、。žºſ،¿:، ×~§§： -§§！ ！！ !！ ！！ ！!§§！ № №r， ſº
º
§¶√≠√∞
§§
:··§@₪¿º：ſaeſaeſë- · %§， №.）€:ſae№ae，----
·≡§§§§§§§§§§§§§§§§§§§§;&#;&#ffffffae&#ffff;&ēšķī£§§§§§§§§§§§§：¿
§§§§§§§§、、~--~
§§§§§§§
–==========
_–=====*
== CO
=—
…–=====*
=====
===）
§
º
¿№g
、：→ ·
§§§§§§§
ſae
§:ſſa
§
;
Ll–
O
ëſ
ſ CD
și CC
iſ;ĢĒ：…º…ſaeſae… ·***¿.*; *（）??± m.§§§§§
&3&$$$$$$
ſ.
¿
|
90.1505
E
§
3
º
º
§§§
§§§§§§§§§§
UNIV
|
~！}§§§
Ģe -， º§§§§"， și -Ķī£§！§§§§§¿?
- º¿S$§§§§§
* ſae*}}： -*$،§§§， - ：-） … ►}*** ¿ $¢ £ €§§∞ √∞:： ·：x.·§§· →·-->…--~~، ！- ،§§§：-（ ... -：- ：●ſ-~-----， ！№ºr№
ا§§、、、。、、。：·~~）; №. ！！！.§§§§§§§§）. №§§§§§§§§§§§§§§§§§§§§§§§§§§§●
ſae！§§ſae----±：§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§§ºº_y（:，-
§§§§§§§§§§§§§§§§::：* 3 ：
§§
§§
nº … •ſërae ·
ſ:）;ſae:ģ§§
：• …，---， -， -， -：-º.}§§šķ§§§§
¿?·¿·ſaew：،¿?ſae§:ſae<！--（ :-（ ：-）¿¿.
¿$¿№£3，5±（√≠√∞aeſae
：E, v， 3 ， 33;№r= №sſae
~·¿¿.*¿¿.*§§•
3;&§ 2ºº5º?§§§§¿§§§
§§§$3 $,！º：· -
0
。
șğģķğģ
ſaeae§§。ğ§§
ſae
§§§§§§§
。
| 3¿§§§§§
#№ſ:...…. ~，。
-§§§§§§
§§§§§
¿?
-§§§
·ſae:¿}ī£8
ſ'，§§šð|-
¿
#
**,
º Y.
§:
~--~--~× ×
• ！= ∞
tº:
{
§
#
Saeeae
rºſsº！！！！！！！！！！！！！！！！！！！！！！！!！?！！?！… ►

* * * § a
ºf º
3.
.
à
%
.."
%
:
2.
%
.*.*
*/
‘e
3
%
3.-
:
.
º
§º f
sºº
ſº
º