Sulfuric Acid Absorption and Iodine Values of Various Petroleum

hydrocarbons, or of straight or branched chain or cyclic olefins, or diolefins, acetylenes, or cracked gasolines. 11 Bur Mines, Tech. Pager 181 (1917)...
1 downloads 0 Views 447KB Size
December, 1926

INDUSTRIAL A N D ENGINEERING CHEMISTRY

I259

Sulfuric Acid Absorption and Iodine Values of Various Petroleum Products and Cracked Distillates Obtained Theref rom‘ By Jacque C. Morrell and Gustav Egloff UNIVERSAL OIL PRODUCTS CO., CHICAGO, ILL.

LTHOUGH Brooks and Humphrey2and later Faragher, Gruse, and Garner3 pointed out the errors involved in the determination of unsaturated compounds in petroleum products and cracked distillates by sulfuric acid absorption, the petroleum industry still continues to use this method as a general test for unsaturates. Brooks and Humphrey have shown that not only are the sulfuric acid derivatives soluble in the acid layer formed, but, in addition, polymerization of the olefins takes place to a great extent, the polymerization products being relatively stable toward the sulfuric acid and remaining in the oil layer. The formation of these polymers as pointed out by Brooks may be definitely shown by comparing the boiling point ranges of the oil before and after treatment with sulfuric acid. Although not a close quantitative method for the determination of unsaturated hydrocarbons, sulfuric acid absorption has its uses for qualitative and comparative tests. The determination of iodine numbers or per cent iodine absorbed as a measure of unsaturation and the variables which affect this determination have been frequently discussed. The displacement of the Hub14 solution (iodine and mercuric chloride dissolved in alcohol) by the more rapidly acting and more stable solution of Wijs,l consisting of iodine monochloride and glacial acetic acid, to the more general adoption of the Hanuse method complete one phase in the history of iodine number determination. The method of McIlhiney7 in extension of Allen’s8 making use of bromine absorption was a recognition of the probability of substitution as well as addition by the halogens. Johansene has taken advantage of the principles involved in the LIcIlhiney method, and applied them to the determination of iodine addition and substitution. To accomplish this a neutral organic solvent, carbon tetrachloride, replaces the glacial acetic acid used by Hanus and Wijs. I n the matter of iodine absorption, the present work concerns itself with the application of the Hanus method in the determination of iodine numbers, and the general conclusions which can be drawn from such a study. This method as applied to petroleum products and cracked distillates obtained therefrom is admittedly not a quantitative measure of unsaturation and, as the results will show, must take into consideration several. factors before even consistent results can be obtained. As old as the field is, the writers believe there is still much to be done in the working out of a consistent and reliable method which will give the iodine addition values in a mixture such as is represented by cracked distillates.

A

Presented before the Division of Petroleum Chemistry at the 68th Meeting of the American Chemical Society, Ithaca, N. Y.,September 8 t o 13, 1924. Revised manuscript received June 1. 1925. * J . A m . Chcm. Soc., 40, 822 (1918). a ~ m JOURNAL, s is, io44 (1921). 4 Dinglms potytcch. J . , ass, 281 (1884). 6 Ber., 81, 750 (1898). 0 2. Nahr. Gcnussm., 4 . 913 (1901). 7 J . A m . Chem. SOC.,81, 1084 (1899). 8 “Commercial Organic Analysis,” 2nd ed. * THISJOURNAL,14, 288 (1922). 1

Sulfuric Acid Absorption

The percentages of various oils dissolved or absorbed by 1.84 specific gravity sulfuric acid were determined by shaking 5 cc. of the oil with 4 cc. of the acid in a IO-cc. graduate and centrifuging to separate the two layers. The results of these tests are shown in Table I. Table I-Sulfuric KINDOF OIL Mexican Gas Oil: (a)

Acid Absorption A. P. I. Per cent gravity gasoline

26.3 46.4

39.5 65.7

35 17

27.6 36.9

35.3 44.3

22 21

21.0 53.0

41.7 75.8

84 20

13.4 30.5

23.3 75.5

100 19

(a)

12.6 49.7

23.6 76.8

100 16

8

13.4 43.3

6!.5

29.7

98 18

26.9 54.0

36.7 71.7

28 12

(b)

California Gas Oil: (a) (b)

Mexican Crude Straight-Run Distillate:

$1 Venezuela Crude Oil:

(a) (b) Mexican Crude (Panuco) :

(b) California Crude (El Doro): ’

Per cent absorbed

Louisiana Fuel Oil (paraffin base):

(4

(b) ( a ) Yields based on charging oil. ( b ) Yields based on cracked distillate.

All these cracked distillates were produced by commercial operation in a Dubbs unit. The experiments were made with untreated distillate. Table I shows that not only in all the cases was a greater percentage of the charging oil absorbed by sulfuric acid than that of the cracked distillate, but in several cases the absorption of the charging oil was complete. If the sulfuric acid absorption method is to be taken as a complete index of the unsaturate content, these latter oils would be completely unsaturated and, in general, instead of unsaturated compounds being formed in cracking the reverse would be true. That these oils may not be entirely unsaturated is indicated by the fact that Mexican gas oil from Panuco crude shows only 35 per cent sulfuric acid absorption. The mechanism of the phenomenon of complete absorption may be explained as follows: Some of the unsaturates, including asphaltic compounds, are acted on by sulfuric acid, the products of which in turn are miscible with the unactedon portion of the oil, giving a one-layer system. Determination of Iodine Numbers

The method of Hanus was used in these determinations. The oils were prepared for the reaction mixture by dissolving 5 grams of oil in a quantity of chloroform and making up to 500 cc., 1 cc. being equal to 0.01 gram of oil. As recommended by Smith and Tuttle,’O a 30-minute absorption period in the dark was allowed. I n every case, 25 cc. of the Hanus solution was used, insuring an excess. Special Erlenmeyer flasks with gutter tops were used for the reaction mixtures. Blank tests were made for every three determinations. 10

Bur. Slandards, Tech. Paper 81 (1914).

INDUSTRIAL AND ENGINEERING CHEMISTRY

1260

The results are shown in Table 11. Eight average determinations, using varying quantities of oil, were made on each sample, but only the extreme values are shown in the table. The curves, however, show the intermediate values. It is brought out clearly here that the iodine values decrease with increasing proportions of oil, keeping the Hanus solution constant. of A m o u n t of Oil on Mexican Crude Mexican Calif. Straight-Run Gas Oil Gas Oil Distillate (a) (b) (0) (b) (a) (b) 84.2 63.4 87.7 54.1 101.8 88.3 33.2 15.7 37.1 15.6 39.3 23.5 Charging oil. ( 6 ) Cracked distillate.

Table 11-Effect Oil Gram

0.06 1.00 ((I)

Iodine Numbers Venezuela Crude Louisiana Distillate Fuel Oil (4 (b) (a) (b) 84.5 68.8 92.1 51.8 35.8 16.5 38.9 10.3

Vol. 17, No. 12

pointed out as a determining factor, but the crux of the whole matter in simplifying the analytical procedure and obtaining consistent results is not the amount of oil nor the excess of iodine considered as such, but the relative proportions of iodine and oil, as can be definitely seen from Table 1V. I n order to determine the concentration effects, two oil solutions-using cracked distillate from Mexican gas oilof different concentrations were made up, a “dilute” solution containing 0.01 gram per cubic centimeter of solution, and a “concentrated” solution containing 0.04 gram per cubic

‘K .b

I n order to determine the effect of excess Hanus solution, varying quantities of the latter were used, keeping the oil constant-that is, 25 cc. of chloroform solution containing 0.25 gram of the cracked distillate from the Mexican gas oil. Table I11 shows the results obtained. Here again only the extreme values of seven average. experiments are shown. The intermediate values are shown on the curves, however.

$2 B

I

4

Table 111-Effect

of Varying A m o u n t s of Hanus Solution on Iodine Number Hanu8 s o h . Iodine cc. Titern Absorptionb “Excess”c number 25 39.21 13.10 3.00 60.2 90 165.81 22.50 7.40 103.3 Unabsorbed Hanus solution in cubic centimeters Na&Os. b Difference in cubic centimeters Na&Ot solution between blank for corresponding quantity of Hanus solution and titer. oRatio between titer and absorption.

centimeter. The iodine numbers varied with the quantity of oil, as would be expected, but for each pair of dilute and concentrated oil solutions they were practically constant. Table V shows the extreme values of six pairs of results.

It is seen, as has been pointed out by Smith and Tuttle and others on various types of oils, that the iodine number increases as the excess of Hanus solution is increased. On the othqr hand, by keeping the relative proportions of oil to Hanus solution constant, using the.same type of oil cited in Table 111, the iodine number is almost a constant, as is seen in Table IV.

Table IV-Effect of Constant Proportion of Oil to H a n u s S o l u t i o n Oil Hanus solution Iodine Gram c c. number 45 77.8 ,(a) 0.25 0.15 27 78.0 ~~

(b)

0.10 0.15 0.25

20 30 50

82.6

(c)

0.125 0.15 0.25

30 36

88.6 87.7 88.7

0.08 0.16 0.24

30 60

111.4 111.8 111.3

(d)

eo

90

82.4 83.1

The question of excess of iodine has been discussed from time to time. The amounts of oil used have also been

Oil Gram

0.06 1.00

Table V-Effect of Concentration NUMBER-IODINE Dilute Concd. 84.2 84.5 33.2 33.2

Bromine Addition and Substitution Values

As is pointed out by Dean and Hill,“ the results obtained by the McIlhiney method are not concordant. For the cracked distillate used here i t was difficult to obtain duplicate determinations by this method, although all conditions were kept strictly constant. The Hanus method also gives consistent results with oleic acid. Regarding the time factor, Smith and TuttlelO point out that after 15 minutes little or no effect will be observed on the iodine value. The effect of temperature is greater than that of time, and becomes even greater with increased excess. Miskin,l* however, showed with cracked distillate that the maximum value is approached as the time is increased; for example, when allowed to stand 24 hours the excess effect is not very noticeable. Miskin used the Wijs solution which Faragher, Gruse, and Garner3 state gives results similar to those given with the Hanus solution. The ratios of the iodine number to the sulfuric acid absorption shown by Dean and Hill” are much more nearly constant than the results obtained in the present investigation, which would be especially expected when the charging oils are considered. Dean and Hill, however, point out that the ratio of 6.5 given by them should never be employed for oils containing appreciable quantities of pitchy materials. Close consideration of the two methods and the mechanism of the reaction would show that there is no particular reason’ for expecting constant ratios. Faragher, Gruse, and Garner3 have concluded that the Hanus solution does not cause any appreciable substitution in the molecules of simple paraffin, cycloparaffin, or aromatic hydrocarbons, or of straight or branched chain or cyclic olefins, or diolefins, acetylenes, or cracked gasolines. 11 12

Bur Mines, Tech. Pager 181 (1917). J . I n s f . Pelroleurn Tech., 10, 299 (1924).

INDUSTRIAL A N D ENGINEERING CHEMISTRY

December, 1925

This conclusion is partly based on the statement (p. 1048) that the values as determined did not increase with long time periods. Miskin,’* on the other hand, found that increasing the period for cracked gasolines has the same result as a n excess of reagent-namely, a n increase in the iodine value. It is significant, however, that actual tests on the pure hydrocarbons were made by Faragher and eo-workers without any noticeable reaction; further, that the hydrocarbons used were of low molecular weight not predominantly present in cracked oils. With cracked products vapors similar to those formed by the halogen acids in a moist atmosphere may be noted upon addition of the Hanus solution. If this is the result of substitution the extent to which it takes place is an important question. Johansenu has undertaken to determine the extent to which substitution has taken place. I n carrying his purpose into effect he replaced the glacial acetic acid as a solvent in the Hanus solution by carbon tetrachloride, using strictly equivalent amounts of iodine and bromine. Some of Johansen’s findings were: (a) The new solution absorbed a greater total percentage of halogen than the Hanus solution. ( b ) Addition values do not change to the same extent as do the substitution values, and that, in general, increase in the absorption values is due mainly to increase in substitution. (c) Certain types of oils give negative addition values.

1261

tion if the same relative proportions of iodine-bromine solution are used. This does away with the unwieldy consideration of “excess” iodine-bromine solution and change in the weight of oil. Concentration effects with the latter do not vary the results within certain limits. 3-It is pointed out, in the light of various inconsistencies which have arisen in comparing the results of the analytical methods studied here, that much further investigation will be required in a study of the correct mechanism of the various reactions involved and proper interpretation of the results obtained. It is suggested that the substitution reactions of the unsaturated hydrocarbons of higher molecular weight particularly be studied.

Analysis of Materials Containing a Mixture of Metallic Iron and Iron Oxides’ By Henry C. M. Ingeberg TECHNICAL UNIVBRSITY

OF

NORWAY, TRONDHJEM, NORWAY

T

HE various existing methods for analyzing materials containing a mixture of metallic iron and iron oxides are all based upon the use of some solvent which is All these findings have been checked by the present writers within certain limits (Table VI). There is a considerable supposed to remove the metal and leave the oxides intact. change in the addition value with change in relative propor- After filtration the content of iron in filtrate (ferrous iron) tions of the reactants, although this change is not nearly so and residue (ferrous and ferric iron) is determined by one large as the variation in substitution values. It is difficult of the usual methods. It has also been proposed to determine to explain here the negative addition value, as well as the the metallic iron indirectly, by measuring hydrogen evolved change from negative to positive addition values with the or copper precipitated during its solution. Such materials, however, differ very widely as to structure same type of oil simply by a change in the relative proportions of the reactants. The great change in substitution values and reactivity of both the metal and the oxides they contain. with the change in relative proportions of reactants would Therefore, even if a solvent, when used on a certain type of also require investigation even from the limited viewpoint material, dissolves the metal completely without appreciably attacking the oxides, it may give entirely false results when of obtaining constant results. used on other materials, either because i t dissolves part of Table VI-Variation of Addition and Substitution Values the oxides also or because it leaves part of the metal undis(0.25 gram oil in 25 cc. solution) solved. For instancQ a solution of copper sulfate, which I-Br AbsorpSubstitution Substisoln. Blank Titer tion S 25 tution Addition has been repeatedly recommended as a solvent for this cc. cc. cc. Cc Cc. Cc. values values separation,* is known to react with ferrous oxide.3 If the Cracked Disfillate (Mexican Gas) ferrous oxide is present in the form of natural magnetite this 25 47,2 b:y5 8 . 30 44 79 ..42 22 99 .. 26 reaction will no doubt be very slow, and the method may 45 84,9 {g;:;: 16.35 6.1 12.2 72.3 24.6 give accurate results for mixtures of metallic iron and mag16.30 6.0 12.0 71.1 25.6 Original Charging Oil (Mexican Gas) netite. On the other hand, if the material to be analyzed is a 9.3 55.1 -1.2 25 4 7 , 3 5 { :3::5 8.9 4.65 ferrum reductum preparation or a partly reduced iron ore, 8.95 4.5 9.0 53.3 -0.3 the ferrous oxide will be present in a much more active form, 10.6 62.7 2.7 45 85.26 { ;7:Z5 11.06 5.3 11.11 5 4 10.8 64.0 1.8 and the error due to its reaction with the copper sulfate may 65.8 1.3 5.55 11.10 6o 113,52 { i;?Z:; 11.32 then be of considerable magnitude. 11.22 5.50 11.0 65.2 1.3 Obviously, under these circumstances, checking of any one It is evident from a study of the results obtained by the method by means of known mixtures of metallic iron and iron three methods that much further investigation will be re- oxides will be futile, except when the method is used to analyze quired in a study of the correct mechanism of the various that special type of mixture-and known mixtures of the same reactions involved and proper interpretation of the results type as, say, ferrum reductum, cannot be prepared; conobtained before definite generalizations can be made upon sequently, when the method is used to analyze such products, this subject. the reliability of the results will always be questionable. An attempt was therefore made to solve the problem along Summary and Conclusions somewhat different lines, by the method here described. 1-The sulfuric acid absorption method may be applied Method only as a rapid method for determining the relative unsaturation of different oils of a particular class, particularly light SOLUTIONS-sOhtiOn of KCuC&. Dissolve 560 grams of distillates. The heavier oils, even though they be distillates, CuC12.2H20 and 245 grams of KC1 in 1 liter of distilled water. containing pitchy or asphaltic matter, give results which Received May 12. 1925. * Peck, Chem. Ztq , 23, 642 (1899); Frerichs, Arch. d . Pharmacie, require further study for a proper interpretation. _ .Z . anpew. Chem.., 221.. 1224 (19091: . 2-Maintaining all other conditions, such as time, tem- 246, 190 (1908): Coblentz and Mav. Williams and Anderson, THISJOURNAL, 14, 1057 (1922); Sims and Larsen, perature, etc., constant check results may be obtained in the Ibid.. 17. 87 (1925). determination of the iodine value by the use of Hanus solub Foerste; andHerold, Z. Elekfrochem., 16, 461 (1910).

:::;

1

I ,