John Almon.

The American journal of science and arts online

. (page 76 of 102)
Online LibraryJohn AlmonThe American journal of science and arts → online text (page 76 of 102)
Font size
QR-code for this ebook

an excess of free acid not combined with the oxyd of the metal
to be determined, this must be first separated by evaporation to
dryness in the usual manner. The presence of alkalies and
alkaline earths is of course without influence on the result, but
on the other hand even very small quantities of iron, alumina
and various other bases make it almost impossible to determine
the point of saturation with precision, these oxyds in solu-
tion giving with cochineal ana logwood specific reactions not
easily distinguished from those produced by the alkalies in ex-
cess. For uiis' reason the method does not apply when oxyds
of this class are present, and this case is precisely that which

Am. Joub. Sci.— SBOOin) Sbribs, Vol. XUV, No. 181. - fiBPT., 1867.

Digitized by


210 W. CHhbs-^Coniribuihns to Chemistry.

moit frequenfly oocais in practice. The method will be aopb'
cable to all these cases if nereafter a ooloring matter shonla be
discovered sensitive only to acids or bases in excess, but not pro-
dacittg specific coloration with salts which aiB neutral in consd-
tation. it is possible that cyanin, the remarkable properdes of
which hare been described by Schdnbein, or the rosocyanin
of Schlumberger,* may fulfil the condition indicated, but I have
had no opportunity of experimenting with either.

The method of precipitation above described may be used
with advantage in preparing a pure normal acid for titritioo.
Pare crystalhzed sulphate of copper is to be powdered and
heated in a porcelain crucible placed within a Hessian cracihie
for about an hour, the texs^rature being gradually raised and
not allowed to exceed a low red heat. The anhydrous sulphate
is then while still hot to be transferred to a perfectly dry glass
tube which can be closed with a good cork covered with tin fofl^
After coolin^^ the tube is weighed, the contents poured into a
flasky the salt dissolved in water and the copper precipitated
at a boiling heat as above. The filtrate and washings are then
to be made up to a known volume. From the weight of the
anhydrous sulphate employed the qimntity of sulphiffic acid

E resent in the solution is known. In experiments made in tbi»
iboratory by Mr. & Ghauvenet this method has been £>and
very accurate and expeditious.

%2. On the predpUaHon of copper by hypophoephanjm aeUL

In a memoir on the hypophosphiteSy A. Wttrtzf has shown
that when solutions of copper are heated to 70^ G. with hypo-
phosphorous acid a hydruret of copper is precipitated, which on
boiling is reduced to metallic copper with evolution of hydrogen^
On repeating this experiment 1 found that the precipitation of
copper is complete, and as the alkaline Inrpophoaphites are now
to DC had in commerce, it ap(>eared probable that the process
might be applied to quantitative estimatiour Experiments to
determine this point have been made by Mr. R Ghauvenet with
the following results r

The copper should be in solution as sulphate, the liquid con-
laining a little free acid. The precipitation from the nitraie is
always incomplete. When chlorbydric acid or chlorids are
present the method firils entirely, the copper being redaoed to
subchlorid and remaining in solution. The solution must not
be too dilute ; the precipitation is complete if the saturated sola-
tion of sulphate be diluted with not more than ten times its balk
of water, before the addition of the hypophosphites. In order
to avoid the sudden evolution of hydrogen gas, and also to ob-

* BolL de la Bo&M Chimique, Man. 1866, p. 194.

t Ana. de Ghimle et de Phjiiqae, 3d serieSi vol. tI, p. 199.

Digitized by


W. Gibbi—CoiUributions U Chemistry. 311

tain the preoipitate in a spongy coherent form, it is best not to
allow the liG[uid to boil. The solution of hypophosphite having
been added in the cold and in excess, the temperature is to be
gradually raised uutil, after standing for some minutes between
80'' C. and 90° C, the hydmret of copper has entirely separated
in coherent masses. It is easy to determine when the preoipitar
tion is complete by taking out a drop of the clear liquid with
a rod and testing upon a porcelain plate with a drop of sulphy-
dric acid solution. No niter need be used if the precipitatioa
be effected in an assay flask; the copper is easily washed by de-
cantation, and may then be transferred to a porcelain crucible by
the well-known method of inversion, dried and gently ignited in
a current of hydrogen. The following analyses wUl serv; to
illustrate the accuracy of this method. In all of them hypo-
phosphite of magnesium was employed as the precipitant

1. 1*1650 gr. pure sulphate of copper gare 0*2966 gr. coppers25-46 p. c. (CSiavreoei)

2. lJJ690gr. '• «« «* 0-S970gr. " ss26*46 " "

5. 1*4266 gr. " •. « 0^M26gr. •* ss26*48 « •«

4. 1-8060 gr. " u u 0-8827 gr. " =26*42 <* (KEGarmaii)

6. 0*8208 gr. " « « 0*2087 gr. « =b26*42 « (E. F. Gale.)

In (4) and (5) a large excess of salphate of nickel was present
The formula €u&0- +5aq gives 25*42 per cent of copper. In

the third analysis sulphates of iron, manganese, nickel and zine

in veiy large excess were added to the solution of copper.
I. In a very pure subsulphid of copper from Arizona, Mn

Chauvenet found in four analyses 74*24, 7487, 74*86 and 74*41

per cent copper,
n. In an alloy of copper and nickel

6. 0*4245 gr. gave 0*8606 gr. of copper =84*92 per eeot (Chauvenet)

7. 0*8616 gr. " 0*8070 gr. " =84*92 « •«

8. 0-1880 gr. « 01170 gr. « =84*85 « «

9. 0-1980 gr. •• 0-1680 gr. " =84-84 «* «

ni. In brass wire

10. 1*6800 gr. gave 1.0706 gr. of copper =65-67 percent (Chauvenet)

11. 1.8655 gr. ** 1-2240 gr. " =65*61 *♦ "

12. 1-6770 gr. " 1-1010 gr. « =65*65 « "

In the last seven analyses the alloj^ was dissolved in sulphurie
acid, nitric acid being added from time to time to assist in solu-
tion. The solution was then evaporated until the last traces of
nitric acid were expelled. The presence of iron in the form of
sulphate does not in any way interfere with the completepre-
cipitation of copper by hypophosphite of magnesium. Wnen
sesquichlorid of iron is present, however, the copper is always
reduced to subchlorid ana is not precipitated as metal or hydru-
ret A solution of a hypophosphite reauces sesquichlorid of iron
to protochlorid : the reduction is particularly rapid and corn-

Digitized by


212 W. OMi—CantrOnawtu to Chemisiry.

plete when a salt of oopper is also present and the liquid oon-
tains free chlorhydric acid. I have endeavored to base nnoQ
this redaction a method for determining iron yolometrically, W
all the experiments failed in consequence of the difficulty of de-
termining the exact point at which the reduction of the inm is
complete. Sulphocyanid of potassium, proposed for this puqxjse
bj WinUer* in his process with subchlorid of oopper as a ledu*
cmg agent, was not found to give sharp indications. When eop*
per and iron are present together as chlorids, the addition of
bypophosphile of magnesium simply reduces the copper to sub-
chlorid as above stated. If in this case we add an alkaliQechlo-
rid to keep the subchlorid of copper dissolved, the copper may
be easily precipitated as subsulphid by sulphydric add gas.
Wi)en arsenic or antimony are present with copper these must
first be separated before precipitating the copper as hydruret, as
careful experiments by Mr. G. JAllj have shown that both arsenic
and antimony are precipitated with the copper. Mr. Lilly ob-
tained the following analytical results when arsenous acid was

1-2690 gr. BulpbaU of oopper g»Te 0-32'79 gr, m«tallic copper s:±25*8S pr. ct copper.
1-6127 gr. " •• 0-8906 gr. " =26-77 ••

0-9688 gr. *• " 0-2609 gr. ** *=26-08 ••

The formula gives 25*42 of metallic copper. In presence of
antimonious acid —

0*7100 gr. sulphate of copper gave 0*2454 gr. copper =z 34-56 per cent

After addition of Sb^O, and Bochelle salt

0*0875 gr. Bulphate of copper gave 0*2426 gr. oopper = 24*56 per cent

Bepeated analyses by Mr. Lilly also showed that copper could
not be determined accurately in Schweinfurt green by hypophcw-
phite of magnesium, and that the presence of Bochelle salt did
not completely prevent the precipitation of arsenic with the oop-
per when arsenous or arsenic acid were mixed with sulphate of

in assaying copper ores it is usually desirable to bring the
metal at once into tne form of sulphate. Numerous experiments
made in this laboratory fully justify me in recommending the
following method. The finely pulverized ore (sulphide of oopper
and iron) is to be mixed in a porcelain crucible with three or
four times its weight of a mixture of one molecule of bisal*

Ehate and one of nitrate of potassium. The mixture is then to
e slowly heated to low redness, which is b^t accomplished in
a muffle. The metallic snlphidis are completely oxydized with-
out the least frothing of the heated mixture. Enough strong
sulphuric acid to convert all the sulphate of potassium into bisol*
* Zeitechriil Sir Aoaljtische Ghemle, Bd. iv, p. 428.

Digitized by


W. Gibbs — Contributions to Chemistry. 218

Ehate is then added, and the crucible is to be again carefully
eated until the contents run to a clear fused mass. On cooling
the mass usually separates readily from the crucible which is
not attacked, and on solution the iron and copper are found com-
pletely converted into sulphates. This process has been tried
suocessfullv with a great variety of ores. The whole operation
requires about an hour. In the case of ores containing much
bisulphid of iron it is best to heat the powdered ore first as long
as sulphur is given off, and afterward to add the oxydizin^ mix-
ture and heat as above. The sulphids of lead, zinc and anti-
mony are completely oxydized by the same process.

§ 3. On the precipitation of copper and nickel by aUcaUne car'


The precipitation of copper by zinc or by the electrolytic
method requires that the metal should be present in the form
of sulphate or chlorid and does not succeed with the nitrate.
As stated above the employment of the hypophosphites is lim-
ited to the case in which the metal exists as sulphate. The old
mode of precipitating copper as oxyd by caustic potash has dis*
advantages which are familiar to all chemists, but on the other
hand is independent of the nature of the solution of copper em-
ployed so lon^ at least as no organic matter is present Accord-
ing to Bose* the alkaline carbonates precipitate copper less com-
pletely than caustic alkalies. This statement, however, is not
accurate for all the conditions under which the experiment may
be performed ; and I have found that copper may oe completely
precipitated from the sulphate, nitrate or chlorid when the solu-
tions are boiled together for a sufficient time and are sufficientlv
dilute. Mr. E. B. Taylor, who has made a careful study of this
method of determining copper, has arrived at the following as
the best method of conducting the process. The solution of
copper is to be diluted with water until the liq^uid contains not
more than about one gram of the metal in one litre. A solution
of carbonate of potash or soda is then to be added in small ex-
cess, and the whole boiled for about half an hour. The boiling
proceeds quietly and without succussions; the blue green carbo-
nate soon oecomes dark brown, and has a fine granular character
which renders it extremely easy to wash. After washing it is to
be ignited in an atmosphere of hydrogen, and the copper weighed
as metal ; it will be found to be free from alkali. In this manner
Mr. Taylor obtained in five analyses the following results:

l'S884 gr. pare sulphate of copper gaye 0*4688 p, metallic copper r= 25*44 pr. ct
1 *7 144 gr. " metallic cop. diraoWeS in aqua regta gaTe 1 *7 1 6 1 gr. cop. =100*09 p. e,
1-8860 gr. " •* " •* « " 1.8868 gr. - = 9»-98 "

1-4667 gr. " " " •* nitric add •• 1^670 gr. " =100^ **

1-4686 gr. « « « « « « 1-4684 gr. - = 99-65 •

* Handboch der Analytiachen Chemie, ii, 175. Sedwte Auflage.

Digitized by VjOOQ iC

S14 W. Gibbs — Contributions to Chemistry.

The filtrate is perfectly free from copper if tke process has beeo
well conducted.

The ignited oxyd is in a state of great subdivision, and the
ignition must therefore be conducted with much care to avoid
loss. A small portion of the oxyd or basic carbonate usaaUy
adheres to the sides of the vessel in which the boiling takes place.
This is to be re-dissolved, and again precipitated, but great care
must be taken not to add a large excess of the alkaline carbo-
nate, which gives a solution from which the copper is not pre-
cipitated by boiling.

Nickel may be completely precipitated from its solatioDS by
precisely the same process. The green basic carbonate may M
washed much more readily than the oxyd precipitated by caus-
tic alkali; it is to be ignited and weighea as oxyd. in two
analyses Mr. Taylor obtained the following results.

1-9808 gr. anhydrous salpbate of nickel gave 0-9551 gr. Ni0=37*79 p. c.
1-4601 gr. " " " " 0-7008 gr. Nie=87-64 **

The formula siSO* requires 87-69 (iFi=68). Dr. F. A. Gtenth
informs me that he has also used the alkaline carbonates in pre-
cipitatiog nickel, and with most satisfactory results.

The precipitation of cobalt by an alkaline carbonate can only
with much difficulty and by long boiling be made complete;
As a means of determining cobalt it is not to be recommended.
On the other hand Mr. FTW. Clarke has found that cobalt is
completely and easily precipitated by the process of oxydatiou
first given by Popp,^ which consists in neutralizing the solution
with carbonate of sodium, adding acetate of sodium and then
boiling with an excess of an alkaline hypochlorite, taking care
to keep the solution alkaline. The hydrated aesquioxyd (7) of
cobalt thrown down may be readily washed. After reduction
in hydrogen the metal is found to be free from alkali Ifickel
may, as ropp has also shown, be precipitated in the same manner,
but the process given above seems to me preferable.

In this connection I may be permitted to state that the method
of separating cobalt from nickel by means of peroxyd of lead
attributed to myself in the new edition of Bose'sf BD&ndbuch der
Analytischen Ghemie and also ascribed to me by Ghiuhe^ was
never even propc^ed by me. *

Cobalt and nickel may be precipitated from neutral solutions
of their sulphates, nitrates and chlorids by adding first an exoess
of oxalic acid to the concentrated solution and then a laige ex-
cess of strong alcohol. After standing a few hours the filtrate
is perfectly free from metal. The oxalates are very easily
wanhed. This method is, however, rarely available for analyV

* ZeitMlirift fUr Analytische Chemie. f Secbst* Aofljige, Bd ii, p. I43w

t Zeittehrift flir Analytiacht Chemie.

Digitized by


W. Oibbs — Contributions to Chemistry. 315

ical purposes, since it fSsdls entirely when salts of ammonium or
of tne alkaline metals are present. The oxalates are also in
such a state of subdivision that it is almost impossible to ignite
them without loss. The oxyds of copper, cadmium, zinc, manga-
nese and magnesium, are also completely precipitated from their
sulphates by oxalic acid and alcohol, bat not in presence of alka*
line salts. The same is true of both mercurous and mercuric
nitrates. In the few cases in which this mode of precipitation will
find application in practice it will probably be best to determine
the oxalic acid in tne oxalate by hypermanganate of potash.

In a former paper I have stated that the sulphids of cobalt
and nickel thrown down from boiling solutions by a boiling so-
lution of sulphid of sodium may be washed without oxydation
upon the filter. The difficulty of preparing pure sulphid of
sodium has, however, been an objection to this method.* This
difficulty may easily be removed by dissolving the crystallized
tetrahedral sulphid, Na^S-fOaq, in alcohol of 90 per cent, fil-
tering and allowing the solution to crystallize. After two or
three crystallizations the pure sulphid may be dried over sulphu-
ric acid in vacuo and the white effloresced mass preserved in a
well stoppered bottle. The sulphid is chemically pure.

§ 4. O/i the employment of sand and glass Jitters in qttantitoHve


Sufficient attention has not been paid to the advantages of
filters of sand and glass over those of paper, when precipitates
are to be dried upon the filter at a definite temperature. By
choking the throat of a funnel with coarse fragments of glass
and then placing upon these successive layers of powdered glass
or sand, tne upper layer being of the finest powder, it is easy to
make a filter upon which almost any precipitate may be filtered
off and washed out completely without the slightest loss. The
funnel with its contents may then be dried at any temperature
below that at which the glass softens or at which the precipitate
undergoes chemical or physical change. Mr. E. R Taylor, who
has carried out this su^estion with the greatest care and thor-
oughness, obtained the following results in three analyses of tar-

I. 0*0917 gr. tartar-emetic gave 0*0463 gr. Sb^S, =: 36*00 per cent Sb.
XL 0*6286 gr. « •• 0-8149 gr. " =s: 86*08 " "

TIL 0*1766 gr. " •* 0*0891 gr. " = 36*07 ** "

The formula requires 85-92 per cent (Sb=120).

In the first and second analyses the precipitated sulphid was
dried at 875° 0. In the third the funnel had the shape of a tube

* This Journal, yol zzzTii, p. 860, alio Zaitsdirift ft^ Analytiaofae Obmi*, Bd.
ill, p. 892.

Digitized by


816 W. Gibbs — Contributions to ChemUtry.

tapering at one end and dilated iu the middle to a sort of biilb^
and the precipitated sulphid of antimony after drying was ig*
nited in the filter tube itself in a current of carbonic acid gas.
This form of funnel, which is due to Mr. Taylor, will be found
very advantageous. A small common funnel may be inserted
into the top, and after drying, the tube funnel closed with a oork
in weighing.

§5. On the e9tim(Ui(m of manganese (M pi/raphofphaie.

The existence of an ortho-phosphate of manganese and am-
monium corresponding to the well known salt of magnesium,
was long since ascertained by Otto.* The subject has more
recently Deen studied by Debrayf who has described a series
of analogous phosphates, all of which are remarkable for thdr
insolubility. Otto^s salt, P208Mn2(NH4)2+2H20, from its highly
crystalline structure, the facility with which it is formed, and
its insolubility, appeared well adapted to the quantitative esti-
mation of manganese, and the following analyses show that this
metal like magnesium may be advantageously precipitated as
ammonia-phosphate and weighed as pyrophosphate.

To the solution of manganese; whicn may contain salts of am-
monium or of the alkaline metals, disodio orthophosphate is to
be added in large excess above the quantity required to pre-
cipitate the manganese as orthophosphate. The white precip-
itate is then to be redissolved in excess of sulphuric or chlorhy-
dric acid, heated to the boiling point and ammonia added in ex-
cess. A white or semi-gelatinous precipitate is produced which,
on boiling or standing for some time even in the cold, gradually
becomes crystalline and finally is completely converted into
beautiful talcose scales which nave a pearly luster and a pale
rose color. It is best to precipitate each time in a platinum ves-
sel in which the ammonia-phosphate may be boiled for ten or
fifteen minutes, and to allow the salt to remain at a temperature
near the boiling point of the liquid for an hour after it has be-
come crystalline. The ammonia-phosphate may then be filtered
off and washed with hot water. The washing takes place with
extraordinary facility on account of the crystalline character of
the salt. The orthophosphate after drying and ignition yields
pyrophosphate of manganese as a nearly white powder. In
this manner

1. 0-9565 gr. MnSe4 gave 8985 gr. PaOyMoj, = 4668 p.c MnO

2. 1-1400 gr. " " 1-0717 gr. " =46-67 ** •

3. 0-8145 gr. " « 0-7646 gr. " =r 46-63 " **


* Bull de la Soci6t6 Ohimiqae. Noavelle S6tm ii, p. 11.
t Ana der Chemie und Fharmacw, yiii, 178.

Digitized by


W. Gihbs — Contributions to Chemistry. 217

4. 0-9464 gr. MnSO^ gave 0-8886 gr. P^e^MDa = 46-66 p.c. ffnO
6. 1-8181 gr. " * 1-2890 gr. " =46-68 " "

6. 1-0666 gr. '« " 0-9960 gr. " =46-76 " "


The formula requires 46*67 per cent (Mii=54). The sulphate
employed was pure and perfectly anhydrous. In two analyseB
of crystallized chlorid of ma nga nese not quite free from mechan-
ically mixed water, Mr. F. W. Clarke obtained 2?-08 and 27-07
per cent of manganese^ In the same salt the percentage of
chlorine was found to be 85*68 which corresponds to 2714 per
cent of manganese.

The advantf^e of this method orer that commonly employed
for the estimation of manganese, is that the process permits us
to weigh the metal in the form of a perfectly definite compound
and not as an oxyd which cannot be safely assumed to consist
wholly of Mn^Oy When manganese is associated with the
alkaline earths it is of course first to be separated as sulphid, or
by Schiel's method as a hydrate of the sesquioxyd. The ammonia-
pnosphate is almost absolutely insoluble in boilin^water, in
ammonia and in solutions of salts of ammonium. The salt is
nearly white but sometimes becomes a little more red upon the
filter. If it assumes a, rather deep dull red color the whole of the
phosphate of manganese has not been converted into ammonia-
phosphate. The precipitate is then to be redissolved in dilute
chlorhydric acid, more phosphate of sodium added and then am-
monia in excess, after which the boiling is to be repeated. This
repetition is very rarely necessary, a little practice enabling the
analyst to judge when the conversion from the fiocky-gelatinous
to the ciystalline 'condition is complete. The filtrate from the
crystalline salt is perfectly free from manganese. Phosphoric
acid cannot be determined by precipitation as ammonia-phos-
phate of manganese, because the crystalline character of the salt
upon which the success of the process depends is only produced
by digestion with an excess of phosphate. Bette* has aescribed
an ammonia-phosphate of zinc which, like the corresponding
manganese salt, is almost absolutely insoluble in water. Debravf
has analyzed similar salts of nickel and cobalt, and Ottoj: has also
describea the analogous ammonia-phosphate of iron. I have
myself prepared an ammonia-phosphate of cadmium which, like
the other salts of this group, is extremely insoluble in water.
All of these salts however, are more or less readily soluble in
ammonia and in salts of ammonium, and after repeated trials I
have not succeeded in rendering any of them available for an-
alytical purposes.

Cambridge, June 20th, 186*7.

* Ann. der Chemie nnd Pharmacie, xt, 129. f ^^* ^^t.

X Ann. der Chemie and Pharm. xri, 199.

Am. Joub. Sci.— Sbookd Sibibs, Vol. XLIV, No. 181.— Skpt., 1867.

Digitized by


316 On the action ofperoxyd of Manganese an ttric add.

Abt. XXI. — Note on the action of peroxtd of manganese upon une
acid; by C. Gilbbbt Wheeleel

The oxydizing action of the peroxyds upon organic sab-
stances rarying to some extent according to the peroxyd em-
ployed, I have investigated the action of peroxyd of manganeso
upon uric acid.

If uric acid and peroxyd of manganese are heated together
with a like quantity of water and sulphuric acid is added in
small portions at a time until no farther action is to be observed,
the black pasty mass then filtered, and the filtrate evaporated
to about one-fourth of its original volume, there is obtained
after considerable time, a quantity of large hexagonal crystals,
which by analysis and characteristic reactions was found to be
parabanic acid.

If uric acid is heated with a large quantity of water only,
until the latter is brought to the boiling point and then per-
oxyd of manganese added as long as evolution of carbonic acid
occurs and the mass filtered, there remains on the filter pe^
oxyd of manganese and oxalate of manganese, while the filtrate
on being somewhat concentrated yields crystals, which if again
dissolved and treated with animal charcoal may be obtained color*
less and quite pure. They were tasteless, rather diffieoltly sol-
uble in cold but readily soluble in warm water ; the solution

Online LibraryJohn AlmonThe American journal of science and arts → online text (page 76 of 102)