A New Rapid and Accurate Volumetric Method for the Determination of

The returned water from a plant is soft and free from mineral matter. If oil is present it should be removed and the water used as many times as possi...
0 downloads 0 Views 398KB Size
May, 191 I

T H E J O U R A ' A L OF I.'\'DL-STRIAL

organic matter and containing no chloride of calcium or magnesium. This subject requires alarge amount of scientific study. SUMLIARY.

Water containing not more than 3 grains of sulphates and carbonates per gallon can usually be treated by heat alone. Waters of medium hardness may be treated within or without the boiler, preferably without. Very hard water requires chemical treatment in a wvll designed softening plant. Some waters cannot be treated economically and are useless for boiler purposes. The returned water from a plant is soEt and free from mineral matter. If oil is present it should be removed and the water used as many times as possible. If i t corrodes the boiler one or more grains of alkali per gallon may prevent this action. Condensed water from steam turbines and surface condensers is oil-free, This is an argument in favor of their use. Lime is ideal as a water-purifying agent as it practically removes the carbonates of lime and magnesium-it should be free from magnesium oxide. Any softening agent that leaves sulphate of soda in solution favors foaming. The ideal purification removes the sulphate entirely. Monthly analyses should be made and kept on file of the water from the intake and from the boilers. Blowing down boilers will often save much labor and expense in the treatment and removal of tubes. I wish to conclude this paper with a plea for more systematic investigation of the causes and prevention of scaling, corrosion, priming and foaming-conditions that present themselves in thousands of power houses annually. [CONTRIBUTIONS

FROAI T H E

HAVEMEYER LABORATORIES, COLUYBIA U S I N o . 191.1

VERSITY.

A NEW RAPID AND ACCURATE VOLUMETRIC METHOD FOR THE DETERMINATION OF MANGANESE AND ITS APPLICATION TO THE ANALYSIS OF IRON AND STEEL. B Y F .I. METZGER A N D L. E. MARRS. Received April 10, 1911.

It has been observed for some time' that the presence of manganese is a disturbing factor in the determination of ferrous iron in rock by the sulphuric-hydrofluoric acid method, and that the more hydrofluoric acid present, the greater the tendency to produce high results. This fact led t o the supposition that in the presence of sufficient hydrofluoric acid, manganese and permanganic acid would react quantitatively, thus giving a method for the determination of the former. Preliminary experiments proved that in the presence of sufficient hydrofluoric acid, the manganese, after the reaction, is all in the trivalent form; thus: Mn0,8 H f - e - >In+++ 4H,O 4(~-) and 4Mn++ + 4(-) -+ 4Mn+++, or 8H+ +4Mn++ --+ jMn++' 4H,O. Mn0,From this i t is apparent that the value of the per-

+

+

+

+

+

1 See Hillebrand, C . S . Geol. Survey, Bi.rll. 422; ''i'he Analysis of Silicate and Carbonate Rocks." pp 162, 163.

AKD E*YGISEERISG CHEMISTRY.

333

manganate solution in terms of iron multiplied b y 0 . 7 8 6 8 2 ( 1 9 1 1 atomic weights) gives the value in terms of manganese. E X P E R I h l E K T A L. Solittions Csed.-( I ) Solution of potassium permanganate, approximately :V/30 ; standardized against specially prepared hlohr's salt ; I cc. = 0.0017 2 I gram iron or 0.001354gram manganese. ( 2 ) A solution of manganous sulphate accurately standardized by precipitation as manganese ammonium phosphate and weighing as manganese pyrophosphate after ignition; I cc. = 0 . 0 0 2 7 2 4 gram manganese. .Vethod.-Titrations were made in wax beakers, obtained by cutting off the tops of the ordinary white ceresine hydrofluoric acid bottles. A total volume of I o o - I j o cc. was used, containing I O cc. sulphuric acid ( I : 2 ) and 2 5 cc. hydrofluoric acid. Measured volumes of the manganous sulphate solution were titrated in the presence of water, sulphuric and hydrofluoric acids with varying amounts of ammonium fluoride as follows: KlInO& sol used. cc

\Veight of ammon. fluoride present Grams.

Weight of l f n taken Gram.

Weight

1 . .,, , . . . .

20.15

5

0 02724

2 . .,

20.90

5

3 . . . . , . , . . 20.15 4 . . . . . . . .20.90 .

10

0,02833 0.02724 0,02833

0,02728 0.02830 0.02728 0,02830

NO.

, ,

,

, ,

,

10

of ?dn found. Gram.

Error. Gram. +0.00004 -0.00003 t0.00004 -0.00003

The endpoint is a distinct pink which lasts several minutes ; the reaction proceeds somewhat slowly toward the end and the titration must be completed drop by drop. Sometimes there is a slight brown color a t the end of the titration but the appearance of the pink end point is distinctly and easily seen. With amounts of, manganese above about 40 mg., some experience is necessary for the recognition of the endpoint which is slightly masked by the brown color due t o manganic salts. Ammonium fluoride increases the speed of the reaction but the effect of I O grams is not sufficiently greater than that of 5 grams to justify the use of so much of the reagent. Next, titrations were made similar t o the above with varying amounts of ammonium fluoride and in the presence of 0.5 gram of iron (added in t h e form of ferric ammonium alum) in each titration. The results were as follows:

so 5...... 6.. .. 7 . . . 8 . . . .

9. . . .. 10. , . . . . 11.. . . . 12 . . . . . . 13 . . . . .

ElInO, sol. used. cc. 2 0 2

202 20.18 20.27 20.35 20 5 5 20.31 2255 20 75

Weight Weighr XYeight of JIn of NHIF Fe"' taken. present. present. Gram. Grams. Gram. 1 0.5 0 02732 2 0 5 0.02724 0,02724 3 0 5 4 0.5 0 02729 5 0 5 0.02754 6 0 5 0 02ii8 i 0 5 0 02749 8

0 5

10

0 5

-

Sum of 13. 267 .46

__

0,02806

W t . of ?.In found. Gram. 0 02735 0 02735 0 02732 0 02745 0 02755 0.02772 0.02750 0,03053 0,02810

__

___

0.36164

0,36203

t0.00039

0.03054

___

Error. Gram. +0.00003

t0.00011 tO.00008 f0.00016

L O 00001 -0.00006

tO.00001 00001 ' - 0 00004

-0

I t is seen that the presence of ferric iron does not affect the accuracy of the method. One gram of ammonium fluoride completely decolorizes 0.5 gram of

T H E ,J O U R N A L OF I N D C S T R I A L A N D E N G I N E E R I h l G C H E M I S T R Y .

334

ferric iron; the same effect can be obtained by using 2 grams of potassium or sodium fluoride, but as these dissolve with difficulty, the readily soluble ammonium fluoride is to be preferred. From the sum of titrations 1-13, 267.46 cc. of the permanganate solution is equivalent to 0.36164gram manganese or I cc. KMnO, = 0.001352 gram Mn, which is in very good agreement with the value obtained by calculation from the iron standard of the permanganate. The great accuracy of the above method is apparent when we consider the low equivalent of the permanganate solution in terms of manganese, together with the fact that the end-point is very easily obtained within one drop of the permanganate solution. The Analysis o j Steel.-Weigh out about I gram of sample; dissolve in I O cc. of nitric acid ( I : I ) in a covered casserole. Cool slightly; add I gram of ammonium persulphate and let stand until effervescence ceases; boil briskly for a few seconds, remove cover and evaporate t o dryness (do not bake) ; take up with 20 cc. of sulphuric acid ( I : 2 ) and 30 cc. of water and boil until the solution is clear. Cool, transfer to a wax beaker, add j grams of ammonium fluoride, 2 5 cc. hydrofluoric acid, dilute with water t o a volume of 100-1j o cc. and titrate. The solution is colorless and the pink endpoint is very easily seen. As much as 2 cc. of concentrated nitric acid has no effect on the titration; hence, if the residue has become baked add I or 2 cc. of dilute nitric acid to make complete solution take place readily. The above method was applied to twelve samples , of steel purchased from the Bureau of Standards, Washington, D. C. The results are given here in tabular form.

No. 14 15 16 17 18 19 20 21 22 23 24 25

Bessemer 0.1 C . . . . . . . . . . . . . . . . Acid open-hearth 0.1 C ,. . . . . . . . Basic open-hearth 0.1 C . . . . . . . . . Bessemer 0.2 C . . . . . . . . . . . . . . . . . Acid open-hearth 0.2 C. . , , . . , , . Basic open-hearth 0.2 C . . . . . . . . . Bessemer 0.4 C . . . . . . . . . . . . . . . . .Acid open-hearth 0 . 4 C.. . . . . . . . Basic open-hearth 0 . 4 C . . . . . . . . . Basic open-hearth 0.6 C . , . , , . , . . Basic open-hearth 0.8 C . . . . . . . . . Basic open-hearth 1.0 C . , , . , . , , , Average.

Found, Mn, p e r Cent.

0,485 0.388 0.506 0.866 0.726 0.436 0.835 0.460 0.376 0.524 0 626 0 320

0.545 0.413 0.519 0.946 0.795 0.513 0.934 0.533 0.442 0.588 0.688 0.465

0.513 0.412 0.528 0.890 0.760 0.464 0.872 0.486 0.406 0.568 0.654 0.405

Average.

_--c--

0.547 0.438 0.538 0.961 0.806 0.515 0.940 0.540 0.443 0.599 0.689 0.472

.

Difference between authors’ and B. of S’s avges.

0.546 0.425 0.529 0.953 0.800 0.514 0.937 0.537 0.442 0 593 0.688 0 468

__

.....................

The results obtained average 0,037 per cent. higher than the Bureau of Standards’ averages and 0 . 0 0 2 per cent. lower than their highest determinations. Nine separate determinations are given for each steel in the Bureau of Standards’ certificates, with an average variation of 0.076 per cent. and a maximum variation of 0.14 per cent. The authors’ greatest variation in duplicates is 0 . 0 2 j per cent. The authors ’ results are uniformly slightly higher than the az’erage results published by the Bureau of Standards, while they are almost identical (-0.002 per cent.) with the highest results, and on this account we wish t o call attention to the fact that in the method described there are neither precipitations nor filtrations

1911

and therefore no possible chances for loss of manganese, whereas, in other methods there are both precipitations and filtrations which might result in a small loss of manganese. I t is, of course, necessary that no carbon compounds exist in solution when titration is made, and the above treatment with ammonium persulphate in nitric acid solution insures their complete removal. Several other methods were employed to remove carbon compounds; namely, nitric acid alone, nitric and sulphuric acids, and ammonium persulphate in sulphuric acid, but none of these gave satisfactory results with all of the twelve samples given in the above table. Duplicate analyses were made of each of the twelve steels by simply dissolving in nitric acid, evaporating to dryness, taking up in dilute sulphuric acid, then adding ammonium fluoride and hydrofluoric acid, and titrating as usual. Although the duplicates in all cases agreed very well, they were all slightly high, the average of the thirteen samples being 0 . 1 1 1 per cent. higher than the average of the results reported by the Bureau of Standards. Our results showed further that this procedure will give accurate results if, instead of calculating’the value of the permanganate from an iron standard, the permanganate is standardized against a steel made by the same process and of known manganese content. Analysis O J Pig Iron.-Proceed as in steels: the combined carbon is all oxidized ; the graphitic carbon is entirely without effect on the permanganate and can be left in the solution as it obscures the endpoint only very slightly. If the graphite is large in amount and its removal is desirable, the solution is filtered just before adding the ammonium fluoride and hydro-

Bureau of Standards, Mn, per cent. _____----Low. High. Average. 0.552 0.440 0.560 0.911 0,800 0.520 0,924 0.520 0.443 0.620 0.703 0.460

May,

+0.033 +0.013 +0.001

+0.063 f0.040 +0.050

+0.065 +0.051 +0.036 fO.025

t0.034 +0.053 +0.037

fluoric acid. Results obtained with two Bureau of Standards ’ irons are given below. B . of S . , M n , p e r cent. __-__-->

Sample.

Low.

Found, Mn,

High.

Ave.

Per cent.

C . . . . . . . . 0 62

0.68

0.66

0.727l 0.673 0.683 0.678

D . . . . . . . . 1.38

1.43

1.41

Average,

Average, 1

Graphite not removed by filtration

Variation from B. of S., p e r cent.

0.690

f0.030

1.49’ 1.49l 1.52 1.48 1 ,495

+0.085

May, 19I I

T H E .JO L-R-YAL OF IiVD ( J S T R I A L A N D EiYGINEERIhrG C H E M I S T R Y .

*

The method possesses all the advantages of simplicity, ease of manipulation, rapidity and accuracy. Fourteen of the analyses included in the above tables were made, complete, in one afternoon. There are no precipitations or filtrations (unless it is desired to filter off the graphite from pig iron) and therefore no possible chances for loss of manganese during the process of analysis. The method is now being applied to the analysis of spiegels, rock, manganese ores, etc., and vie hope to report on these in this journal in the near future. QUAWTITATWE LABORATORIES, HAVEXIEYER HALL, COLUMBIAVSIVERSITY.

[ C O S T R l B ~ ' T I O N FRO51 T H E 'I'EXAS

AGRICULTURAL

EXPERIUENT STATIOS.]

EFFECT OF IGNITION ON SOLUBILITY OF SOIL PHOEPHATES. By C:.

S . FRAYS.

Received February, 24. 19 11.

I t is known that ignition of the soil increases the quantity of phosphoric acid dissolved therefrom by acid. Stewart' uses this as a method for estimating the organic phosphoric acid of the soil. He extracts the original soil and the ignited soil with I z per cent. hydrochloric acid, and assumes that the increased quantity of phosphoric acid secured originates from organic phosphates. The object of the work here reported was to study the effect of ignition upon mineral phosphates such as may occur in the soil. M e t h o d of TT'oiiz.-The quantity of phosphate containing 0.1 gram phosphoric acid was weighed into a platinum dish and ignited for ten minutes, a t a low red heat. I t was transferred, dish and all, to a bottle, 2 0 0 cc. of 1 2 per cent. hydrochloric acid added, allowed to stand 24 hours, filtered, washed with hot water, and made up to , j o o cc. Phosphoric acid was then determined in z o o cc. Another series of experiments was made, in which the ignited phosphate was digested j hours a t 40' with joo cc. of AT/s nitric acid. RESULTS.

The results of the experiments are reported in the table. Ignition increased the solubility of the phosphoric acid of these phosphates decidedly. With fifth-normal nitric acid, about ten times as much phosphoric acid is dissolved from the ignited phosphates, as from those not ignited. With the strong acid, ignition rendered practically completely soluble all the phosphates which were not already completely soluble in this solvent. It is evident that no analytical method for organic phosphoric acid can be based on ignition. If the strong acid dissolves all inorganic phosphates, then any method which determines the remaining phosphoric acid will determine the organic phosphoric acid. If the strong acid does not dissolve all the inorganic phosphates (and we have evidence that it does not) then the remaining inorganic phosphates will be ren1

Bull. 145, Illinois Experiment Station.

335

dered partly soluble in the acid by the ignition, so that the ignition method would not be accurate. PHOSPHORIC A C I D DISSOLVEDFROM I G N I T E D AND O R I G I N A L MINERALS. (In Percentage of the Quantity of Phosphoric Acid Used.) With 1 2 per cent. hydrochloric acid.

___--__-

Laboratory number 240 Wavellite., , , , . , 7 2 1 IT-avellite.. , , . . 726 Wavellite.. , , . . . 714 Dufrenite.. . . . . . 718 Dufrenite.. . , , . . 716 Variscite. . , . . . . 724 Triplite.. . . . . . .

Original mineral.

.. 19 18 100 86 26 100

With

*V/5 nitric acid.

_ _ A Y_ _ _

Ignited . mineral.

Original mineral.

100 100 100 100 96 100

4.8 4 .5

Ignited mineral.

.

2 0 4.8 8.0 12 . o

80.7 97.5 58 . o 35.5 75 . 0 100.0

IO0 .. ... Phosphoric acid is not the only constitutent of the soil whose solubility in acid is increased by ignition. We have examined twenty-five soils, and ignition increases the oxide of iron and alumina dissolved by the acid from every one of them. The increase is great in some instances. For example, in soil 3368, ignition increased the acid-soluble oxide of iron and aluminafrom 2.63 to 5.48 per cent., and in soil 1361, from 1.25 to 6.50 per cent. Full details of this work will be published in Bull. I35 of the Texas Experiment Station.

S U M M A R Y AKiU C O N C L U S I O S S .

Ignition increases about ten times the solubility of the phosphoric acid of wavellite, dufrenite and variscite in fifth-normal nitric acid. 2 . Ignition renders variscite, dufrenite and wavellite almost completely soluble in 1 2 per cent. hydrochloric acid. 3. Ignition of the soil will probably render inorganic phosphates soluble in acid, and therefore is not a method for estimating organic phosphoric acid. 4. Ignition of the soil renders considerable quantities of iron and aluminium oxides soluble in acid. I.

COLLEGESTATION, TEXAS. February 15, 1911.

DEVELOPMENT OF THE SUGAR INDUSTRY.' By Aucusr

VOX

WACHTEL

Received Apnl 2 6 , 1911

The sketch submitted to your kind attention treats an extensive subject and necessarily will give only a general impression and not one adequate t o the importance and the technical development of the manufacture. The sugar industry was the pioneer in vacuum boiling a rational extraction of raw products by cheap evaporation and filtration, and solved first the problem of handling a difficult crystallization on the largest scale and in an entirely modern way. Still this industry has the reputation of not being strictly a chemical manufacture-a wrong impression, as it must be either classified with logwood or tannin extractions where the methods of sugar manufacture are simply copied, or i t can be compared with the manufacture of cream of tartar or alkaloids, which are all extracted from some natural raw material and a nearly chemically pure product turned out. 1

Read before the Ne% York Section of

dustry, March 1 7 , 1911

the Society of Chemical In-