Application of the Dimethyl Sulfate Test for Determining Small

Application of the Dimethyl Sulfate Test for Determining Small Amounts of Petroleum or Asphalt Products in Tars. Charles S. Reeve, and Richard H. Lewi...
0 downloads 0 Views 372KB Size
velopment of heat on moistening dry coal support this theory. The heat development by the moistening of dry coal may contribute appreciably to the spontaneous heating of some coals in storage. APPLICATION OF THE DIMETHYL SULFATE TEST FOR DETERMINING SMALL AMOUNTS OF PETROLEUM OR ASPHALT PRODUCTS IN TARS’ B y CHARLESS. REEVEA S D R I C H ~ RH D.

LEWIS

INTRODUCTORY

The recent production of mixtures of t a r with various asphalt and petroleum products for use in the road and paving industry has necessitated a satisfactory method for determining the proportions of each type of material present in such mixtures. The authors have based their research on the well-known dimethyl sulfate test proposed b y Valenta? and later investigated and reported upon b y Graefes and Harrison and Perkin.4 A. Sommers modified the test in order t o render i t applicable for the estimation of asphalt in tar-asphalt mixtures, and Chapin6 now uses it in a modified form for the detection of adulterants in coal t a r creosote sheep dips. -4 brief review of the literature upon the test t o date will, therefore, prove of value, before submitting our data. THE DIMETHYL SULFATE TEST

Valenta’s purpose in originating the test was t o determine the amount of t a r oils present as a n adulterant in paraffin and rosin oils, and he claimed as a result of his investigations t h a t while aromatic hydrocarbons are freely miscible with dimethyl sulfate, the open-chain hydrocarbons are entirely insoluble in t h a t reagent. I t was, therefore, only necessary to thoroughly agitate a measured volume of the sample under examination with a measured volume of dimethyl sulfate, and the increased volume of the latter would indicate the amount of aromatic hydrocarbons which had gone into solution. This is in general, but not absolutely, correct, and the real facts regarding the test as brought out b y later investigations are briefly summed up b y Chapin as follows:: .‘I, Aromatic hydrocarbons and derivatives thereof are completely soluble in dimethyl sulfate. being miscible in all proportions. ”2 . Open-chain hydrocarbons show varying degrees of solubility in dimethyl sulfate, ranging from zero t o a considerable percentage. light oils as a rule being the more soluble. “ 3 . Treatment of a mixture of aromatic and openchain hydrocarbons with dimethyl sulfate results in the formation of two layers; the upper contains most of the open-chain hydrocarbons plus a varying b u t usually small amount of aromatic hydrocarbons retained in solution; the lower layer contains practically all of the dimethyl sulfate, holding in solution most of the aromatic hydrocarbons plus a varying 1 Paper presented a t t h e Eighth International Congress of .Ipplied Chemistry, New York, September, 1912 2 Cheni. Z t g . , 30, 266-7 (19061. 3 Chemische Rezwe ueber die Fetl- und H a m - l i i d u s t r i r , 14, 112-5 (1907). Amdust, 33, 2-11 (1908). THISJOURNAL, 2. No. 5 :1910’1 ’’ L-.S . Dept. Agric.. Circular 167 7 I b i d , 167. 2.



but usually small amount of the open-chain hydrocarbons. ‘- of the retort. -1 I O O cc. 1

Loi- cit.

T H E J O C R S A L OF I-VD 1-STRI.4 L A S D E S G I S E E R I A Y G C H E L V I S T R Y

2 94

sample was taken and Table I shows the results obtained. I t will be understood t h a t volume percentages are referred t o throughout this paper unless otherwise noted. TABLEI-DIMETHYL SULFATETESTS o x DISTILLATES 270-315°C.

-

315-350°C.

350-375'

C.

u

m, u. b

p"

E

i

2 3 4

5

6

ON

DISTILLATES FROM 107, MIXTURE 315-350'

To. of sample

Mixture 9Oyr coke oven tar lOYc oil asphalt S o . 2 2 90Yr coke oven tar l 0 y r fluxed Bermudez asphalt 3 ?OVc coke oven tar IO'% residual petroleum 4(a) Refined tar preparation Free carbon = 19.36Yc

Pertentage distillate

C. fraction

----.

350-375' C. fraction

Percentage insoluble Perin centage (CH3bSO4 distillate

Percentage insoluble in (CH&5O4

1

a B

1

T A B L E111-TESTS

Vol. 5 , NO. 4

Oil asphalt Penetration 3 . 6 mm. Oil asphalt Penetration 17.0 mm. Oil asphalt Penetration 15.2 mm. California oil asphalt Penetration 7 . 3 mm. Fluxed Bermudez asphalt Penetration 12.2 mm. Residual petroleum Specific viscosity Engler 50' C.-65.1

2.2

notest

11.8

80.0

28.8

90.0

3.7

notest

15.5

85.0

28.0

90.0

3.2

notest

16.8

67 5

1i.2

90 0

13.0

85.0

32 5

82.5

7.0

i5.0

4 0

notest

12.0

85.0

17.1

&5,0

0.6

notest

6.8

90.0

25.2

90.0

I t will be noted t h a t in evey case, except Xo, 4 , the dimethyl sulfate insoluble tends t o be greater in the higher fractions. The z70-31j' C. fractions are too small for making the dimethyl tests with one exception, which in a measure indicates t h a t i t would be impossible t o detect their presence in such fractions from mixtures containing small amounts of petroleum or asphaltic compounds. These materials were then mixed with a crude coke-oven t a r whose characteristics were as follows : TABLE11-TESTS

ON

CRUDE COKE OVEK T A R

General character.. . . . . . . . . . . . . . . . . . . . .

Thin fluid

. . . . . . . . . . . . . . . . . . . 1.173 95 23Yc Soluble in C C 2 . . . . . . . . . . . . . . . . . . . . . . Insoluble in CE2 (free carbon). . . . . . . . . . . . 4 . 7 1 C; Ash . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 0.06L7, DISTILLATION

Fractions Water up t o 1100 110-1700

Percentage by volume

c.

c.

170-270° C . 270-315°C. 315-350' C. 350-375' C. Pitch (over 375' C.)

0.6 1.6 0.8 17.5 9.4 11 .o 9.7

Percentage by Character of weight distillates 0.5 Cloudy 1.3 Clear 0.6 Clear 15 . 5 1/2 solid 8.2 solid 9.7 Solid 8.4 Solid

49.4

55.8

__

__

100 0

100.0

Hard, brittle, glossy

Insoluble in (CHa 1 2 5 0 4 Sone None None Sone Sone None

11.5

5.0

17.5

25 .O

12.0

5.0

13.2

12.5

11 .6

10.0

13.0

22.5

14.0

15 .O

12.2

18.0

!a) Said t o contain 10Yc oil asphalt.

I n these I O per,cent. mixtures a n excellent indication of the presence of oil asphalt is shown by the dimethyl sulfate test even in the 31j-3joO C. fraction, b u t there was no trace of open-chain compounds in t h e 270-315' C. fraction. No evidence of a n oil product could be found in the total distillate obtained b y distilling t o coke by Sommer's method. On account of the ease with which I O per cent. could be detected by testing separate fractions, the test was not carried out with all the materials. TABLEI\--TEsTs

ON

DlSTILLhTES

315-350'

FROM

C. fraction

570 MIXTURES

350-375'

C. fraction

7 * _ _ _ 7 -

Pers o . of centage sample Mixture distillate 1 95'% coke oven t a r 5mc oil asphalt S o 1 9 2 2 Y S % coke oven tar 5'7, oil asphalt S o 2 15 5 3 9 5 % coke oven tar 5Cc oil asphalt N o 3 11 0 4 9 5 7 r coke oven t a r 5 % Calif oil asphalt 9 6 5 95?? coke oven tar 5 7 c fluxed Bermudez asphalt 11 0 6 95VC coke oven tar 5 % residual petroleum 8 5

Percentage insoluble Perin centage (CH3)PSOI distillate

7

Percentage insoluble in (CH3:2S04

iione

9 4

7.5

Sone

9 7

12 5

2 5

10 2

12 5

rone

10 8

10.0

None

10.8

8.0

7.5

12.5

12.5

I n these j per cent. mixtures, the asphaltic base can still be detected in fairly large percentage in t h e last fraction (350-37jo C.), but in only two cases (KO. 3 and No. 6) out of six was there present a dimethyl sulfate insoluble in a lower fraction (315350° c.). T A B L E\--TESTS

Oh'

DISTILLATES FROM 3Yp MIXTURES 350-375' C. fraction

The mixtures were made, unless otherwise stated, Percentage by adding the given percentage by n-eight of petro- s o . r1f insoluble in Percentage distillate Mixture (CH3hS04 leum or asphaltic product t o the weighed amount of sample 1 977< coke oyen t a r tar. The materials when necessary were rendered 5 .O 3c; oil asphalt No. 1 9 7 fluid by heating before mixing and in no case was 2 Y7YC coke m-en t a r 5 .O 8.9 3Yc oil asphalt N o . 2 there any separation of oil apparent. Distillations 3 977< coke oven t a r were then made in the manner above referred to, and 4.0 8 7 3 q oil asphalt No. 3 fractions were taken a t 270' C.. 3 1 j O C.. 350' C., 4 ?7% coke oven t a r 2.5 37, Calif. oil asphalt 7.8 and 3 j j O C. Four cc. of the distillate from each frac5 97% coke oven tar tion were treated with 6 cc. of dimethyl sulfate in a 1.25 8.1 3% fluxed Bermudez asphalt I O cc. cylinder graduated t o 0 . 2 cc. The resulting 6 977< coke oven tar 10.0 10.4 3% residual petroleum supernatant layer of insoluble oil was read and calculated t o its percentage of the volume of sample taken. I n the 350-375' C. fraction of the distillations of The results are shown in the following tables. the 3 per cent. mixture, the insoluble residue ranges

Apr.. 1913

T H E JOL-Ra\-.4L OF I.\-DYSTR19L

from I O per cent. t o 1 . 2 5 per cent. so t h a t the detection of even three per cent. of asphalt is made possible b y the method of fractionation which has been employed in this work. The 3 1 j O C. t o 350’ C. fraction did not give the slightest trace of insoluble products. CONCLUSION

While the above results do not give any absc.lute idea of either the amount or exact nature of the material which has been fluxed with the t a r , the authors feel t h a t the improvement which they have made upon former methods may still prove of value, until a more exact method shall be forthcoming. Specifications for tar-oil or tar-asphalt mixtures will be useless without some method of checking up the material submitted under them. We now think it possible, however, by specifying t h e grade and character of both the t a r and petroleum or asphalt product t o be used, and then obtaining d a t a on similar laboratory mixtures. t h a t we can determine with reasonable certainty whether our specifications are being fulfilled. I n order t o more closely determine the character of the material which has been fluxed with a tar. a volatilization test in conjunction with the dimethyl sulfate test will no doubt prove of great importance, and we feel t h a t some accumulated data on these two tests will place us in a much safer position for specifying and identifying bituminous mixtures which present such difficulty t o the chemist. 1-

s

PUBLIC ROADS WASHINGTON. D. C.

OFFICE O F

T H E DETERMINATION O F OXYGEN I N METALLIC TUNGSTEN POWDER AND SOME NOTES ON T H E DETERMINATION O F OXYGEN I N STEEL B y CHARLESMORRIS

A*YD E.VGINEERIA’G C H E - V I S T R Y

295

gen in steel; i. e . , the ignition of the substance in a stream of hydrogen, which method is credited t o Ledebur. The electrically heated furnace introduced b y the author‘ in 1908 for the direct determination of carbon in iron, steel and alloys is utilized in the process which is described in detail in this paper. Walker and Patrick,’ in a paper read at the Eighth International Congress of Applied Chemistry, attack the accuracy of the Ledebur method on the ground. t h a t any oxides of manganese or silicon present in the steel would not be reduced. The author regards the Ledebur method as more practical t h a n the proposed new one2 above noted; even if the former process does not reveal the total oxygen present, i t certainly shows enough of i t t o furnish a basis for judgment of t h e quality of the steel. If the steel is sufficiently dirty and poorly melted in actual openhearth Bessemer or crucible practice t o contain oxides of manganese and silicon, then it would surely contain enough oxide of iron t o condemn i t . .4 P P A R A T U S

The arrangement of apparatus is indicated in t h e drawing and the accompanying notes. The towers (or jars) are the author’s design as are also J , I and C, and were first used as part of a combustion train.3 I n this laboratory four furnaces are placed side by side. By the use of a P tube at t h e outlet of jar F, one train from F t o E: can be made t o serve two furnaces. Of course a separate set of A, B and C is necessary for each furnace. If, after making a large number of determinations, the blank begins t o show a gradual increase, the contents of the various jars must be renewed.

JOHNSON

Received January 22, 1913

11E T H 0 D

I t has been found a distinct advantage both in the manufacture and use of tungsten powders t o know their oxygen content. In one of t h e laboratories

Blank.-Before introducing anything into the electric furnace, close all points marked “screw pinch cock.” At B make a connection with a straight

APPARATUS A-Ignited asbestos and dry stick KOH. B-Glass wool plugs and PzO:, (anhydrous powder) C-Solution of K O H . D-A clamp t o prevent the quartz tube from moving since it was found t h a t results \-aried if the tube moved back and forth. E-Quartz tube. F-Ignited asbestos and P205 in alternate layers. G-Ignited asbestos and dry stick KOH in alternate layers.

under the author’s direction. this determination is a matter of daily routine. The method involves the‘same principle used in the determination of oxy-

H-Asbestos and small lump (size of pea) CaC12 in alternate layers. 1-50 cc. alkaline solution of pyrogallic acid, 5-50 cc. KOH solution. K-Kipp generator ( 2 quart) SOLUT~OSS Potassium hydroxide: 1 gram KOH t o 1 cc. water. Pyrogallic acid: 3 grams pyrogallic acid and 2 cc. water to 3 cc. * KOH solution. Kipp Generator: Use stick zinc and dilute hydrochloric acid ( 1 : 1 ) .

glass tribe instead of the 1

u

tube shown.

Insert

J . A m . Chem. SOC., 30, i73. Proc. 8th Intern. Cong.A p g l . Chrm.,21, 139; alsoTHIs JOURNAL, 4, 799. J . A m . Chem. SOC.,28, 8 6 2 .