Document text
THE UNIVERSITY
OF ILLINOIS
LIBRARY
VM<5
THE ARSENIC CONTENT OF SULPHATE OF
ALUMINIUM USED FOR WATER
PURIFICATION
BY
ARTHUR NORTON BENNETT
B. S. University of Illinois, 1907
THESIS
Submitted in Partial Fulfillment of the Requirements for the
Degree of
MASTER OF SCIENCE
IN CHEMISTRY
IN
THE GRADUATE SCHOOL
OF THE
UNIVERSITY OF ILLINOIS
1915
Digitized by the Internet Archive
in 2013
http://archive.org/details/arseniccontentofOObenn
UNIVERSITY OF ILLINOIS
THE GRADUATE SCHOOL
.June 5. 191 5
I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY SUPER-
VISION BY ^ETHJ.R..iID£T.0.N...3EUlIE.T.I _
ENTITLED fEHB-.. ARSMTQ fiOTTCWP OTP SIITiPTTATR OF ALUMTffTTIM
HSED....ED£....:MEH....PIJE.n,.XCA^.IQ]S..
BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE
degree of Mas..l.e.r......Q.f. ...jS.c.i.e;nc.e. _ _ _
In Charge of Thesis
Head of Department
Recommendation concurred in :*
Committee
on
Final Examination*
Required for doctor's degree but not for master's.
CONTENTS
Introduction 1
Historical 2
Method of Analysis 8
Analytical 11
Conclusion 16
\3l5
^3
INTRODUCTION.
Specifications requiring arsenic free sulfate of
*
aluminium for water treatment by several European purification
*Jour. f. Gasbel, 1913 (Sept)
plants suggested to us that it would be advisable to make
determinations of the arsenic content of the sulfate of aluminium
used in this country and particularly of that used in the state
of Illinois.
It is well known that products which are manufactured
with the aid of commercial sulfuric acid quite generally contain
more or less arsenic, depending upon the purity of the acid used.
The poisonous character of arsenic compounds, even when present
in small amounts, makes it of general interest and importance to
have definite knowledge of the presence or absence of arsenic in
any substance which enters directly or indirectly into foods or
drinks. Sulfuric acid is used in the manufacture of sulfate of
aluminium and it is thus quite essential, particularly to those
who are in public health work, to know whether arsenic in any
considerable amounts is being added to drinking water in the
process of purification with this chemical.
-2-
HISTORICAL.
We have been unable to find any definite published
data relating to the arsenic content of sulfate of aluminium.
G. A, Soper, several years ago made an investigation of this
problem and referred to his results while discussing a paper by
E. E. Wall* on "Water Purification at St. Louis, Missouri".
*Trans. Am. Soc. C. E., 60, 202-9.
Mr. Soper was discussing in particular the use of sulfate of iron
in water treatment and spoke in the following manner: "There is a
final point which the speaker hesitates to mention, but inasmuch
as, before this society, it will probably be taken in the
conservative spirit in which it is intended, and may lead to
useful inquiries, it may be referred to briefly. It concerns
the composition of the sulfate of iron used • What are the
impurities in this sulfate? How much arsenic is there in this
sulfate? Some years ago the speaker had occasion to examine
specimens of sulfate of aluminium from a good many filter plants
and found arsenic in nearly all of them. It is true, that usually
the arsenic was not present in large quantities, but it was
easily discoverable, and in some of the samples it was present in
sufficient amount to be of more than passing interest. The
arsenic, of course, came from the sulfuric acid used in making
the sulfate of aluminium, the sulfuric acid having been produced
-3-
from pyrites which contained arsenic."
From Mr. Soper's first remark it is evident he realized
that, due to the increasing public prejudice against the use of
any material containing arsenic or ether similar poison, a great
injustice might he worked upon both the manufacturer and the
plants that use sulfate of aluminium for water treatment by
giving widespread publicity to this matter, at least, before all
phases of the problem had been thoroughly investigated. We, too,
are of this same opinion and so have refrained from mentioning
the names of any manufacturers whose product we have examined and
have also omitted the names of all filter plants outside the
state of Illinois.
We have found but one other reference to the arsenic
content of water treatment materials. In further discussion of
the same paper and relative to sulfate of iron, Mr. E. E. Wall
in reply to Mr. Sopor's statements said: "The writer has a copy
of a report in which it is stated that no arsenic was
found in any of the samples tested and that the small quantity of
arsenic in the sulfuric acid used in cleaning steel is, without
doubt, removed in the cleaning tubs in the form of arsenureted
hydrogen, thus leaving the liquors from which copperas is made,
free from arsenic. Even if there should be a minute quantity of
arsenic in the sulfate of iron, it is scarcely possible that
this could remain in the water after treatment with the quantity
-4-
cf lime used at St. Louis." The explanation of the absence of
arsenic in sulfate of iron due to its loss as arsine is quite
feasible but such is not the case with sulfate of aluminium.
We have not been able to learn that anything has been
done in this country to regulate the amount of arsenic in sulfate
of aluminium. The purification plants (at least in the state of
Illinois) have made no effort to obtain an arsenic free article.
The manufacturers of sulfate of aluminium keep more or less
accurate records of the arsenic content of their product. We
have found only one producer who advertises "arsenic free alum".
Neither the government nor any of the states have promulgated
legislation regulating this product, although there is a regulatio
concerning arsenic in other substances entering into foods. The
government has set a limit for arsenic in coal tar dyes and in
baking powder of one part in 700,000. This very low limit,
particularly when it is considered that only relatively small
amounts of these substances are used in food preparation, shows
that considerable importance is attached to the presence of
arsenic and its compounds.
MANUFACTURE.
Sulfate of aluminium or filter alum as it is commonly
called, is manufactured by digesting finely ground bauxite, AlgO^,
HgO, with the required amount of sulfuric acid. After the reactio:
is complete the liquor is passed through filter presses or some
-5-
other filtration medium to remove silica and other insoluble
impurities. The clarified solution is then run into evaporating
pans and the excess water driven off. It is then allowed to cool
and solidify after which it is broken or ground to the desired
degree of fineness required by the trade. This product consists
largely of the normal sulfate, Alg(S04)3, 18 H^O, with a small
amount of a basic sulfate, so that there is an excess of Al^O^
over that amount required to combine with the S0_ present. All
3
specifiaations for this product call for at least 17$ AlgOg. The
following is a typical analysis. Total AlgO^, 17.5%; Free AI2O3,
1.5$; Total S03, 38.5$, Fe203, .50$; Water of crystallization,
40$.
There is a sulfate of aluminium on the market containing
22$ AlgOg. This is made by driving off sufficient of the water
of crystallization to make the required percentage of AlgOg. To
prepare this product approximately 60$ of the water of crystalliza-
tion would have to be driven off to raise the AlgOg content from
17$ to g2$. Its chief advantage consists in saving effected in
transportation charges.
It is readily seen from the process of manufacture that
practically all of the arsenic present in the sulfuric acid will
appear in the finished product. The arsenic content of the
sulfuric acid depends upon two factors; first, the amount of
arsenic present in the raw material from which it is manufactured,
and second, the method of manufacture.
-6-
The three principle materials from which sulfuric acid
is manufactured are pyrites, (J^Sg), zinc-blende (Zn3) and sulfur.
The arsenic in pyrites varies from very small amounts to quite
appreciable quantities so that sulfuric acid manufactured from
this material may vary widely in arsenic content. Zinc-blende
and sulphur are both comparatively free from arsenic and yield
an acid containing but little arsenic. Sulfuric acid manufactured
by the old lead chamber process contains a large percentage of
the arsenic present in the raw materials unless special means for
removing it have been taken. In the contact process arsenic
interferes with the catalytic action of the platinum, so that it
must be removed from the gases before they are passed over the
catalyser. Thus the contact process furnishes an acid practically
free from arsenic.
One water purification plant* manufactures its own
*Eng. Record, 71, 576.
sulfate of aluminium effecting a considerable saving in the cost
of the product. In this case it is not necessary to produce the
product in the solid form. After the action of the sulfuric
acid upon the bauxite is complete the mixture is ready for use.
In this way the filtration, evaporation and crushing of the product
are eliminated.
METHOD OF USING.
Sulfate of aluminium is used for the purpose of clarify-
ing turbid waters and for removing the soluble coloring materials.
It is added to the water at the rate of from 1 grain to 5 or 6
grains per gallon depending upon the character of the water treated
Aluminium hydroxide is formed by reaction with alkaline salts
present in the water according to the following equations:
A12(S04)3 + 3 CaHg(C03)2 = 2 Al(OH)s ♦ 3 CaS04 ♦ 6 COg
A12(S04)3 ♦ 3 MgH (CO ) = 2 A1(0H) ♦ 3 MgSO ♦ 6 C02
A12(S04)3 ♦ 6 KaHC03= 2 Alg(0H)3 ♦ 3 NOgS04 + 6 C02
On settling , this aluminium hydroxide floe mechanically carries
down the suspended matter including bacteria. The coloring matter
which is largely organic is precipitated with the probable
formation of an aluminium lake.
SAMPLES,
In order that our results might be of greatest value by
as
showing the condition of the sulfate of aluminium/it is actually
used, we first obtained as many samples as possible directly from
the water purification plants in Illinois. Twenty-six plants use
sulfate of aluminium in treating water. The purpose of our
investigation was explained to the managers of each plant. They
were asked to co-operate with us by furnishing a sample of the
product used, together with the name of the manufacturer or dealer
supplying the same. Twenty-two of the plants very promptly
complied with our request, and in nearly every case, expressed
decided interest in the subject with a wish to know the results of
-8-
our investigation.
Owing to expense of transportation practically all of
the sulfate of aluminium used in Illinois is supplied by three
manufacturers. In order to make our study more complete we have
extended the scope of our investigations and have secured samples
from practically all of the large manufacturers of sulfate of
aluminium in the country. In some cases the samples came directly
from the producer and, in others, from the water treatment plants.
The specimens were carefully sampled, ground and analyzed in
duplicate by the following methods.
METHODS OF AUALYSIS.
The method used in obtaining most of the data given is
a modified Gutzeit Method, developed by Claude R. Smith* in his
*U. S. Dept. Agr., Bur. of Chem. , Circular No. 102.
work on coal tar dyes and other food constituents. The results
obtained by this method were in several cases checked by the
Marsh -Berz el ius Method** and were found to agree. The Gutzeit
**U. S. Dept. Agr., Bur. of Chem., Circular No. 99.
Method has been investigated by Sanger and Black*** and others
***J. Soc. Chem. Ind., 26, 1115. (1907)
for quantitative work, and, when proper care is taken in the
-9-
manipulation, has been found to give satisfactory results. The
chief modification proposed by Smith is the use of paper sensitized
with mercuric bromide instead of mercuric chloride, which had
previously been generally used. The bromide gives more permanent
stains and the standards can be kept longer. The method depends
upon the formation of a dark orange stain when the generated
arsine is brought in contact with the sensitized paper. The
apparatus used is essentially as described by Smith. The generator
is a 50 cc. wide mouth Erlenmeyer flask. This is connected with
two upright tubes 8 cm. in length and 1 cm. in diameter, the
lower containing lead acetate paper and the upper filled with
cotton moistened with 5% lead acetate solution. Fitted into the
upper tube by means of a rubber stopper is a capillary tube 3 mm.
in internal diameter and 12 cm. in length. This capillary is
constricted at two points about 3 l/2 cm. from each end. By this
means the sensitized paper is held in the center of the tube thus
producing stains of equal length on both sides of the paper. Under
uniform conditions, the length of the stain varies with the amount
of arsenic present. A series of standard stains prepared from
known amounts of arsenic are used for comparison. A convenient
series is made from 2, 5, 7.5, 10 and 15 micro-milligrams. The
amount of arsenic in the weight of sulfate of aluminium taken is
determined by matching the stain it produces with the standards;
it is then a matter of simple calculation to determine the
percentage arsenic content or the parts per million of arsenic.
-10-
A one gram sample will contain as many parts per million of
arsenic as there are micro -milligrams of stain obtained. For
example, if one gram of sulfate of aluminium produces a stain
which matches the 5 micro -milligram standard stain, then that
sample contains 5 parts per million. One part per million is
equivalent to ,0001 of one per cent. A stain representing between
five and twenty-five micro-milligrams gives the most satisfactory
results. A stain between these limits can be obtained by varying
the weight of sulfate of aluminium used.
The sensitized paper is made from heavy, close-textured
drafting paper, cut into strips 2.5 mm. by 1£ cm. These strips
are soaked for an hour in a bfo alcoholic solution of mercuric
bromide. The excess solution is squeezed off and the strips
allowed to dry.
For the analysis of samples containing more than 30 parts
per million AS2O3 another method proposed by Smith,* was used.
*U. S. Dept. Agr., Bur. of Chem., Circular No. 102.
The generated arsine is passed into mercuric chloride solution.
(10 cc. of 5% EgClg diluted to 100 cc.) According to Smith
probably several different arsenides of mercury and some free
arsenic are formed. These are oxidized by the excess of mercuric
chloride slowly in the cold and rapidly on heating, forming
arsenous acid and mercurous chloride. The mercurous chloride can
be filtered off and weighed and the arsenous acid in the filtrate
-11-
determined by titration with iodine. In this way checks are
obtained in the one determination. The equation used for the
calculation of arsenic from the weight of mercurous chloride
obtained is :
2 AsH3 + 12 HgCl2 + 3 H20 = 12 HgCl + As203 ♦ 12 HC1.
Thus 1 AsgOg is equivalent to 12 HgCl.
As an alternative the conglomerate precipitate can be
titrated by means of iodine. Sufficient potassium iodide is
added to form the soluble double potassium mercuric iodide and
then an excess of standard iodine solution. YThen all the
precipitate has gone into solution the excess iodine is titrated
with standard thiosulfate. The iodine absorption represents the
oxidation of arsine to arsenic acid in which 1 As is equivalent
to 8 I.
As usual all reagents used were tested to prove their
freedom from arsenic •
DETERMINATION OF ARSENIC.
Five grams of finely ground sulfate of aluminium are
dissolved in the generating flask in 30 cc. sulfuric acid (1-4)
with the aid of heat. Four or five drops of a 40$ solution of
stannous chloride in concentrated hydrochloric acid are added and
the solution cooled. Four or five grams of arsenic free moss zinc
are now added and the lead acetate tubes and capillary containing
the sensitized strip are connected . The evolution of gas is
-12-
allowed to proceed for at least one hour. The stain, after
drying, is then compared with the standards. A steady, "brisk
hut not violent evolution of gas should he maintained. This can
he regulated hy varying the acidity, volume of solution, amount
of zinc and temperature. After a little experience very uniform
results can he obtained. In the determination of larger amounts
of arsenic it is necessary to allow the evolution of gas to
proceed for two or three hours.
The results from the samples obtained from the Illinois
purification plants are given in Table L, those from outside the
state in Table II. In all cases the arsenic is recorded as
arsenic trioxide, AsgO^. Twenty-four samples from Illinois plants
and seventeen from sources outside the state were analyzed.
The results obtained by analyzing sulfate of aluminium
used in Illinois clearly show that arsenic in exceedingly small
amounts is always present. We find a minimum of 0.8 parts per
million (.00008$) and a maximum of 4.0 parts per million (.00004$)
of arsenic as AsgOg in the sulfate of aluminium used by Illinois
water purification plants. If a water were treated with sulfate
of aluminium containing the maximum amount of arsenic found, at a
rate of 6 grains of alum per gallon, an amount which is very
seldom exceeded, and if all the arsenic were soluble and remained
in the filtered water, since arsenic is not a cumulative poison,
a person must drink 1285 gallons of the treated water at one time
to obtain a medicinal dose of 2 milligrams. From this it is
-13-
TABLE I.
ARSENIC AS AssOa IH SULFATE OF ALUMINIUM USED III ILLINOIS.
City Arsenic as Ase03
Pts
• per million
Percent
Gallons
Cairo
1.6
.00016
3213
Carlinville
1.8
. 00018
2856
Charleston
1,8
.00012
4283
Chicago and
Rogers Park
1.4
. 00014
3671
E. St. Louis and
Granite City-
.8
.00008
6425
Dec atur
1.4
.00014
3671
Elgin
1.6
.00016
3213
Evanst on
1.4
.00014
3671
Ft. Sheridan
1.2
. 00012
4283
Hamilton
1.4
. 00014
3671
Kankakee
.8
.00008
6425
Kenilworth
1.4
.00014
3671
Lawrenceville
3.0
.00030
1713
Macomb
1.6
.00016
3213
Moline
1.0
.00010
5140
Mt . Carmel
2.0
.00020
2570
Mt. Vernon
1.2
.00012
4283
Pana
1.2
.00012
4283
Quincy**
1.0
.00010
5140
Quincy**
4.0
.00040
1285
Rock Island
2.0
.00020
2570
Rock Island
Arsenal**
1.6
.00016
3213
Rock Island
Arsenal**
1.0
.00010
5140
Streator
3.4
.00034
1512
*Gallons of water containing a minimum medicinal dose of 2 mg. when
the water is treated with 6 grains of sulfate of aluminium per
gallon, provided that all the arsenic remains in solution.
**Two samples from different manufacturers.
-14-
readily seen that the arsenic content of sulfate of aluminium
used in Illinois is of no significance.
TABLE II.
ARSENIC IN SULFATE OF ALUMINIUM OBTAINED FROM SOURCES
OUTSIDE THE STATE OF ILLINOIS.
Sample No. Arsenic as Ase03
Pts. per mil.
Percent
Gallon*
1
0.5
.00005
10280
2
1.2
.00012
4283
3
1.2
.00012
4283
4
1.4
.00014
3671
5
2.6
.00026
1977
6
2.6
.00026
1977
7
4.0
.0004
1285
8
4.0
.0004
1285
9
5.0
.0005
1028
10
20.0
.0020
257
11
27.0
.0027
190
12
31.0
.0031
166
13
49.0
.0049
105
14
280.0
.0280
18
15
941.0
.0941
5.5
16**
1240.0
.124
4.0
17
1240.0
.124
4.0
*Gallons of water containing a minimum medicinal dose of 2 mg.
when the water is treated with 6 grains of alum per gallon,
provided that all the arsenic remains in solution.
**Nos. 16 and 17 were obtained from the same plant and are
probably from the same lot.
Somewhat to our surprise the samples obtained from
sources outside the state showed a wider range in arsenic content.
In one case there was 0.5 parts per million (.00005$), and in
-15-
nine cases there was more than in the highest Illinois sample,
the maximum "being 1240 parts per million of arsenic as Ass03
( .124$),
If a water were treated with alum containing 1240 parts
per million AS3O3 at the rate of 6 grains per gallon and
provided all the arsenic remained in solution, 0,13 parts per
million of arsenic as AS2O3 would he added and a medicinal dose
of 2 mg. would be contained in four gallons. This would he quite
an appreciable amount and is more than should be added in water
purification. However, owing to the insolubility of the arsenites
and arsenates of calcium, magnesium, aluminium and iron, a large
part of the arsenic would be removed with the precipitated
aluminium hydrate. To determine the extent of this removal, if
any, several experiments were carried out.
Two liters of water were treated at the rate of 6 grains
per gallon with sulfate of aluminium containing 941 parts per
million of AS2O3. By this treatment 188 micro milligrams of
AS2O3 were added. Forty micro milligrams of AS2O3 were recovered
from the filtered water and 144 from the sludge. Thus only 22$
of the arsenic remained in solution.
Some water was treated at the rate of 20 grains per
gallon with sulfate of aluminium containing 1240 parts per million
of AS2O3. Only 12$ remained in solution.
A sample of filtered water was obtained from the
filtration plant using this latter alum. It had been treated at
-16-
the rate of 200 Ids. of sulfate of aluminium per million gallons
or at the rate of 1.4 grains per gallon. Prom this water only
7% of the arsenic originally added was recovered from the
solution. The untreated water gave no test for arsenic. Thus a
removal of 9Z>fo of the arsenic was effected by the purification
process.
CONCLUSION
Sulfate of aluminium used for water treatment, as far
as our investigation shows, contains arsenic in amounts varying
from traces (0.5 parts per million) to 1240 parts per million.
Sulfate of aluminium used by water purification plants in
Illinois does not contain a significant amount of arsenic. Some
sulfate of aluminium used elsewhere contains a much larger
amount of arsenic, but since at least 75% of the arsenic added
in the treatment of water v/ith sulfate of aluminium is removed
with the precipitated aluminium hydrate, there is a strong
probability that in no case a sufficient quantity of arsenic
would be added to the filtered water to have therapeutic
significance. However, since sulfate of aluminium containing an
insignificant amount of Ass03 can be readily obtained, the
manufacturers should make an effort to keep the arsenic content
of their product at a minimum and water works officials should
demand an article practically free from arsenic.
-17-
A2JDEBBBM
Mr, M. 3. Litch, Franklin H. Zalbfleisch Company,
and the General Chemical Company very kindly consented to
check our analyses of several samples of sulfate of aluminium.
Five samples were submitted for analysis. They varied in
arsenic content from the minimum to the maximum amount found.
The following results were obtained.
Ass03 - Parts per million.
Our analysis Check analysis
1 0.5 0.6
2 1.4 2.5
3 1.4 1.8
4 4.0 3.0
5 1240.0 1550.0
With the possible exception of No. 2, these checks
are very satisfactory. When less than 2 parts per million of
Ass03 are present the limit of accuracy of the Sutzeit method
is 0.2 parts per million; with from 2 to 5 parts per million
the limit is 0.5; with 1000 or more parts per million the limit
of accuracy varies from 100 to 200 parts per million depending
upon the method used.
Any error due to sampling can be eliminated in this
case as only small samples were used and they were thoroughly
mixed •
-18-
We wish to express our appreciation of the assistance
rendered by the manufacturers and water works officials who
furnished us samples for examination.
We wish further to thank Professor Edward Bartow, who
suggested the problem, for his helpful interest and advice
throughout the progress of the investigation.
xm
WSMWm^m vma mm nR
§ ^1 -fK |w IP
m fe$$S$/i> $^J8g-'A |*%Sl8 P^S"
/ v.-^^.-j/ \%%&J \:'.s>^m