Potash from Kelp. IIIâThe Preliminary Examination of Kelp Distillates

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

786

LABORATORY 3AND PLANT POTASH FROM KELP. 111-THE PRELIMINARY EXAMINATION OF KELP DISTILLATES By G. C. Spencer U. S. KELPPOTASHPLANT,SUMMERLAND, CALIFORNIA Received March 9, 1920

The process whereby kelp is treated a t the experimental plant for the extraction of potash and other values as operated by the U. S. Department of Agriculture a t Summerland, Cal., prescribes its reduction t o a charcoal from which potassium chloride and iodide are easily leached. This reduction is essentially a destructive distillation and is best carried out in closed retorts. Under such conditions large quantities of distillation products are obtained, and i t was the purpose of the research here described t o determine their nature, from t h e point of view of their possible commercial value.2 T h e work was begun in the Chemical Section of the Forest Products Laboratory, a t Madison, Wis., in February 1918,t h a t laboratory being chosen on account of its superior equipment for handling tarry distillates. I n addition t o t a r distillations, the writer made sixteen experimental distillations of dried kelp a t low temperatures in the wood distilling retort of the l a b ~ r a t o r y . ~ This low-temperature kelp distillation showed t h a t the nature of kelp t a r varies greatly with the temperature as in the case of coal distillations. Since dried kelp contains about the same percentage of nitrogen as bituminous coal (about 2 per cent) i t was hoped t h a t the presence of this nitrogen would promote the formation of aromatic compounds in the distillate a t higher temperatures corresponding t o those used in destructive coal distillation. Subsequent distillations of kelp a t high temperatures in Summerland, Cal., did not, however, warrant this speculation. Kelp t a r then stands in a class by itself as regards its properties and constituents. Three series of investigations were conducted t o determine the nature of the condensates from high-, intermediate-, and lowtemperature distillations of kelp, respectively. EXAMINATION O F LOW-TEMPERATURE

TAR

The so-called low-temperature kelp t a r made in the wood retort a t t h e Forest Products Laboratory was produced a t temperatures ranging from 2 4 3 ’ t o 305’ C. Each charge of kelp varied from 2 5 t o 3 0 lbs. T h e amount of crude t a r reported in connection with t h e various distillation runs averaged I. I per cent, which was, roughly speaking, double t h e actual amount of t a r present owing t o suspended water, 1 The second paper of this series entitled “Potash from Kelp. II-Experimental Distillation of Kelp a t Low Temperatures,” by the present 12 (1920), 682. writer, appeared in THISJOURNAL, 2 E. C C Stanford, “The Distillation of Seaweed,” Chem. News, 34, N o . 888, 237. D. M. Balch, “On the Chemistry of Cprtain Algae of J. W. Turrentine, “Note the Pacific Coast,” THISJOURNAL, 1 (19091, 7 7 7 on the Distillation of Kelp,” Proc. 8th Intern. Cong. A p p l . C h e m , 15 (1912), 313. J. S. Burd, “The Economic Value of Pacific Coast Kelps,” California Agr Exp. S t a , Bull. 248 (1915), 183.

* LOC.

cit.

Vol.

12,

No. 8

I

which could be removed only by subsequent treatment. The amount of dried t a r varied from 45 t o 65 per cent of the crude tarry substance taken. This low recovery of t a r from kelp is probably due t o the slow rate of heating and the low maximum temperature (320’ C.), which did not entirely expel the last of the volatile organic matter from the kelp material in any case. D R Y I N G O N L A B O R A T O R Y SCALE-The systematic examination of kelp t a r showed t h a t the entrained water must be removed before the t a r proper could be distilled, since its presence caused the latter t o froth and overflow into the condenser with resultant inconvenience and delay. This difficulty was overcome by heating the t a r t o a temperature of 1 1 0 ’ t o 140’ C. with an oil or sand bath, while a current of air was drawn through the t a r and condenser into a receiving apparatus. Later on it was found necessary only for the air t o pass through the flask t o sweep out the water and oil vapors as they formed, and the practice of bubbling air through the liquid was discontinued. This method of drying the t a r was troublesome and time-consuming and out of the question f o r larger quantities of material. A better method which was followed a t Summerland will be described later. The t a r thus prepared was nearly free from moisture and could be distilled with a free flame. It was. a black viscous liquid having a specific gravity ranging from 1.093t o I . 1 0 2 , and rather more pasty in i t s consistency t h a n t h a t made in Summerland. FRACTIONATION-When 406 g. O f this t a r were distilled with a free flame, 1 8 2 . 4 g. of an oily distillate with a characteristic odor came over between 230’ and 270’ C. On allowing this oil t o stand in a separatory funnel for several hours it separated into two. layers, the upper of which was dark in color, had a specific gravity of 1 . 0 2 2 , and weighed 139.I g. T h e light red lower layer proved t o be an aqueous solution of ammonium salts and of a water-soluble organic body similar to, if not identical with, t h a t which remains when the aqueous kelp distillate is evaporated. A fuller discussion of this water-soluble compouncl: will be given below. The upper layer of oil was shaken with successive portions of 30 and I O cc. of 4 N sulfuric acid t o remove all the bases. As is generally t h e case with such kelp oils, a considerable amount dissolved in the dilute acid, the total amount of bases and other compounds thus removed being 2 5 . o per cent. To determine whether the oils in the sulfuric acid were in a state of emulsion i t was shaken with a n equal volume of saturated sodium chloride solution. Only a slight amount of oil separated on standing. The sulfuric acid solution was next washed with ether, then the acid was carefully neutralized with sodium carbonate solution t o liberate any weaker bases without setting ammonia free. The brown oil which separated was taken up with ether and the ether solution shaken with dilute acetic acid t o form acetates of the free

Aug., 1920


bases if any were present. On evaporating off the ether the residue contained no salts. The oil which had been shaken with sulfuric acid was next shaken with successive portions of 4 N sodium hydroxide solution (30 cc. and I O cc. each) t o take up acids and phenols. The crude t a r acids, which were liberated from the caustic soda solution by sulfuric acid, differed distinctly from those of the higher temperature tar in odor, and possessed a specific gravity of only 0 . 9 7 9 . After standing for some time crystals were formed, apparently voluminous, but weighing very little when separated and purified. They were recovered from the oil by filtering through a dry filter paper and washing with low-boiling gasoline. After removing as much as possible of the excess oil by pressing between sheets of filter paper, the crystals were dissolved in glacial acetic acid and the solution evaporated slowly, a t first on the steam bath, later a t room temperature. The crystals thus obtained melted a t 6 3 O C. This, however, is not a trustworthy value, but the small yield did not permit another recrystallization. Alcohol and ether were only partial solvents. The main bulk of liquid acids gave no test for phenols when the oily mixture was shaken with sodium carbonate and the remaining oil treated with bromine. It is probable t h a t none could be formed a t the low -distillation temperature. The residual kelp oil made in this way is in general similar t o t h a t made from other kelp tars, although it has a different odor. The low-temperature kelp tar has a composition -which may be outlined as follows; Kelp -art

Pitch

: 5 5 . 0 per cent

Oil-+

Aqueous solution

/ 10.6 per cent

Ammonia- and acid-soluble oil

I

\

Water-soluble liquid 9 . 3 per cent

Ammonia 0.10 per cent

Acids 10.8 per cent

Residual oil 14.9 per cent

Low-temperature kelp tar has fewer desirable properties, from the industrial point of view, than t h a t made a t higher temperatures and SO far as known a t present -its chemical constituents would be of little more than academic interest. PITcH-The pitch from the low-temperature tar was black and brittle and had a specific gravity of I . I I ; its melting point stood a t 103 ’, and in general the sample did not seem t o have unusual or valuable properties. OF

TAR

PRODUCED

AT

INTERMEDIATE

TEMPERATURES

The tarry product from intermediate temperature -distillation was a mixture of pasty black tar with varying amounts of water (averaging 40 per cent), and about 20 per cent volatile oil. The odor of the t a r suggested the simultaneous presence of various sulfur and nitrogen compounds. The anhydrous t a r prepared from this crude material had a specific gravity of I .09z. I t s nitrogen

787

content was 3 17 per cent and i t contained no free carbon. TABLE I-SUMMARY

OW

First Runnings: Water.. Oil Fractions : 200°-210

RESULTSOn DISTILLATIONOF DRIBDT A R Per cent

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

210°-240

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

‘............................

240’-275 ............................ Pitch (by difference).

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

2.4 1.4 6.0 16.8 12.0 61.4

The entire oily distillate consisted of phenolic compounds (tar acids), neutral compounds (residual oil), ammonium compounds, and some water-soluble liquids. The fraction t h a t came over between 240’ and 275’ C. darkened very rapidly on exposure t o the air. The above fractions are not natural distillation cuts. The temperature rose so uniformly while the distillate was coming over t h a t only arbitrary temperature ranges can be assumed. The objectionable odor of the oils during distillation was considerably lessened by receiving the distillates in approximately 4 N sulfuric acid. The sulfuric acid not only removed ammonia but also dissolved a constituent of the oil. When a n attempt was made t o recover the bases by liberating them with alkali and shaking out with ether this acid-soluble oil separated with the ether and remained with the ammonium salts. It was probably identical with t h e acid-soluble oil mentioned under the heading of “LowTemperature Tar.” Attempts were first made t o fractionate the distillate directly from the dry tar, but the later practice was to collect the entire oily distillate a t once, a n d t o extract this as follows: The oil from the distilling flask was received in 30 cc. sulfuric acid (approximately 4 N ) and was shaken out once more with I O cc. of the acid; then finally washed with a saturated common salt solut,ion t o remove the excess of free mineral acid. Common salt solution was used because the oil, having nearly the same density as distilled water, showed a strong tendency t o emulsify if shaken even moderately with water alone. The separated oil was shaken twice in convenient portions with an excess of 4 N sodium hydroxide solution t o separate the total tar acids and phenols as water-soluble compounds, and the remaining oil was named for convenience “residual kelp oil.” RESIDUAL K E L P oIL-This crude residual oil soon turned dark in the air; its specific gravity was 0.96. A fractionation of 103.7 g. gave results which are typical of the various samples t h a t were distilled. The residue was, after cooling, a thick viscous liquid. Fractions

c.

EXAMINATION

,

90-95 220-230 235-240 250-260 270-300

Weight of Oil Grams 1.9 10.8 17.8 28.0 29.4

Per cent of Oil 1.8 10.4 17.1 27.0 28.3

The amount of residual oil usually obtained averaged about 2 0 per cent of the weight of dry tar. Its sulfonation value was 7 t o 8 per cent (with 3 7 N sulfuric acid), and the residue was water-white and not affected by strong caustic soda solution. The refractive index of this water-white residue was found t o be I . 4660 a t room temperature.

T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY Samples of t h e residual oil and of t h e crude kelp t a r oil were shown b y test t o be satisfactory flotation oils for ore concentration. T A R ACIDS-These were separated from the caustic soda solution by acidifying with dilute sulfuric acid and washing out the excess of the mineral acid with a saturated common salt solution. The t a r acids thus obtained had a density of 1.024 t o 1.041. The yield was found t o average I I . I per cent of the weight of dried tar. They were nearly black in color and cresolic in odor. Di- and triatomic phenols may be present with the cresols and their homologs, as their presence would account for the rapid darkening in color of these acids when they are brought into contact with air. T h a t these so-called t a r acids are almost entirely made up of phenolic bodies is shown by the fact t h a t only slight amounts of oily material could be separated by shaking with sodium carbonate solution. A rectification of this crude phenol ( 2 7 . I g.) gave the following results: Fractions O

c.

209-2 15 215-220 220-22s 2 2 5-2 5 0

Weight of Fractions Grams 6.5 3.3 2.8

6.4

Per cent of Oil 24.0 12.1 10.3

23.6

T h e first fraction gave a decided blue color with ferric chloride, indicating the presence of cresols, especially o-cresol. No guaiacol was found on careful examination. There was no sulfonation residue. Under the direction of Mr. E. P. Devlin, of t h e Forest Products Laboratory, a disinfectant solution was made with the crude t a r acids as follows: Mix 45 cc. of residual kelp oil with 1 5 cc. of ,tar acids. Dissolve 5 g. of caustic soda in water and add 2 5 g. of rosin. Warm until dissolved; then evaporate the water from the soap. Add the above oil mixture t o the soap, stir until dissolved, and make up t o I O O cc. with water. Samples of crude t a r acids, crude residual kelp oil, and the disinfectant made from them according t o t h e above formula, were submitted t o the Bureau of Chemistry for a test of their toxic properties. T h e report from the Bureau of Chemistry follows: The sample of crude tar acids and crude kelp oil were nonmiscible in water and non-susceptible to Rideal-Walker and Hygienic Laboratory tests for phenol coefficients. 0.I cc. of the crude tar acid sample sterilized IO cc. of culture of B. typhosus in 15 min. 0.I cc. of the same sample failed to sterilize 30 cc. of culture in I hr. These mixtures would be 1-100and 1-300 proportions, respectively. Phenol under like conditions kills 8.typhosus in 15 min. in 1-100dilution and fails to kill in I hr. in 1-200 dilution. I cc. of the crude kelp oil sample killed 4 cc. of culture of B. typhosus in I hr. It failed to kill in 3 0 rnin. I cc. failed to kill 9 cc. of culture in I hr. The disinfectant from kelp tar was found to havean R. -W. coefficient of between I . 5 and 2 .o. A sample of the disinfectant from kelp t a r was also examined by the Pathology Section of the Forest Products Laboratory for its toxic power against Fomes annosus Fr., a wood-destroying fungus. The relative toxicities of kelp-tar disinfectant compared with pure coal-tar creosote and zinc chloride may best be summed up in a quotation from an informal report on the work in the files of the Forest Products Laboratory:

Vol.

12,

No. 8

Kelp-tar disinfectant is one and one-fourth times as toxic as. pure coal-tar creosote, and twice as toxic as the zinc chloride and gum arabic emulsions of coal-tar creosote. The killing points are 0.25, 0 . 3 5 , 0.5, and 0 . 5 5 percent, respectively. TAR BASES-with the exception of ammonia, no bases have been distilled from kelp tar. All tests for pyridine and amino compounds have been negative, whether the supposed bases were removed by steam distillation or by taking up with ether after liberation from the acid solution with caustic alkali. I n each case a n oily substance of pleasant odor was recovered which appeared t o be basic in its properties but, when neutralized with acids, yielded nothing more t h a n a mixture of ammonium salt and an oil. Attempts t o m a k e a double compound with platinic chloride were fruitless. It is extremely improbable t h a t nitrogen can be combined t o any large extent as ammonium s a l t s in kelp tissue. It is more probably combined in a n unstable compound which breaks up readily on heating into ammonia and hydrocarbon derivatives. A n attempt was made t o isolate a base by treating t h e acid solution of kelp bases with sodium carbonate solution, thereby avoiding the evolution of ammonia, but no better results followed. The same oil separated, but i t did not form stable compounds with acids. PITCH-The residual pitch from this t a r was a black, brittle substance with a nitrogen content of 3. 5 3 per cent; its specific gravity was I . I 7 , and its melting point was 130' C. When fused with caustic alkali a n d iron filings a small amount of Prussian blue was recovered. Free carbon was absent. E X A M I N A T I O N O F HIGH-TEMPERATURE TAR

By high-temperature kelp t a r is meant the t a r r y distillate t h a t is formed a t temperatures ranging from 800' t o 950' C., as produced in the retorts a t the Summerland plant. D I S T I L L A T I O N APPARATUS-This t a r gave S O much trouble by frothing t h a t a special distilling apparatus had t o be devised t o guard against losses b y overflowing and t o condense the greatest possible amount of oil vapors from the distillate. A 2 gt. iron retort was heated by a sand bath and the vapors from the hot t a r were carried into the flask A, by a current of hea$ed air which was aspirated through the whole system. Flask A was surrounded by hot water, thus permitting t h e lower boiling vapors .to pass on into Flask B, while the heavier oils collected in A. B was a flask surrounded by ice water from which any still uncondensed vapors led into a reflux spiral condenser, thence into a final wash bottle, C. I n this way the greater part of the oils were condensed. While a succession of condensers seems be necessary for t h e recovery of most of the distillate, the oils collected in different condensers showed nearly the same properties so there was no real separation of the various constituents of the distillate. Approximately I O per cent of alcohol intimately mixed with each charge of t a r was found t o diminish t h e frothing by reducing the surface tension of t h e t a r bubbles and a t the same time carrying off t h e water. The melted t a r was easily miscible withi alcohol. tal

Aug.,

1920


A retort was finally devised a t t h e Summerland kelp plant which contained a perforated disk at the base of t h e dome. A jet of air under pressure was directed against this disk from above which broke the foam as i t came through t h e perforations. I n this way several gallons of tar were successfully distilled at a time. P R O P E R T I E S O F '(THICK" A N D "THIN" TARS-The two samples investigated have been referred t o as t h e "thick" and "thin" tars for convenience. The thick t a r was collected under such conditions t h a t a large part of its water and some of the volatile oil had already been distilled out b y t h e heat of t h e furnace. The thin t a r was taken directly from the condenser a n d probably represented t h e true tarry distillate from kelp better than the other. It is interesting t o note t h a t the percentages of volatile oils in each are about t h e same and, so far as could be learned, the constituents of these oils are nearly identical. The thick t a r had a nitrogen content of 4.24 per cent, but the thin t a r had only 2.64 per cent of nitrogen. A description of the first distillation of thick t a r i s given as typical of t h e others. The distillation proceeded slowly but smoothly, owing, n o doubt, t o t h e presence of alcohol. Alcohol commenced t o come over a t 85'. The air condenser was not connected until t h e temperature had risen t o 100'. After 105' had been passed, t h e flame was carefully increased from time t o time. Above 115' the distillate became a yellow oil. At 1 2 0 ' some t a r froth seems t o have been carried over, t h e flame was deCreased, and the distillation continued. At 130' t o I 53 ' %he distillate reacted strongly alkaline, and a t 240' i t was a thick dark-brown semi-solid, while above 2 5 5 ' i t became limpid and lighter colored. TABLETI-DISTILLATION DATA ON THICKKELP TAR Weight Fract. Fract. Fract. of Alco- 100;- 150'- 200'Sample hol 150 200" 280' Total Total Pitch Pitch G. Cc. G. G. G. G. P e r c e n t G. P e r c e n t 314 42 21 1 44 66 21.0 194 61.7 690 60 92 31 29 152 22.0 401 58.1 673 70 62 14 131 19.4 433 64.3 668 70 114 . . . . 114 17.0 462 6 9 . 1 70 138 14 152 23.7 403 62.8 641

.. ..

Distillates A and B showed specific gravities of 20' C., and their composition must be similar also, because on shaking successively with acid and alkali A yielded 69. 5 per cent of residual oils and 2 5.7 per cent of phenols, while B yielded 72.3 per cent of residual oil and 2 2.8 per cent of phenols. The corresponding oils from A and B were accordingly mixed for further treatment. 0 . 9 8 9 and 0 . 9 8 6 , respectively, at

TABLE 111-DISTILLATIONDATAON THIN KELPTAR Weight of Sample Y D i s t . A-Dist. B-Pitch-G. G. Percent G. Per cent G. Per cent 1002 1096 1172 123%

174.0 114.5 158.5 108.5

17.3 14.5 13.5 8.8

29.5 102.5 87.5 50.0

2.9 9.3 7.4 4.0

367.0 352.0 404.5 395.3

36.6 32.1 34.5 32.0

As has been stated, i t was hoped t h a t t h e distillation of such a nitrogenous body as dried kelp a t t h e temperatures employed in coal distillations might possibly yield a tar t h a t would be similar in many ways t o coal t a r and furnish some of t h e highly desirable hydrocarbons t h a t are obtainable from coal tar. So far as known, kelp has never before been distilled on

789

a large scale, i. e . , in ton lots, so the possibilities have heretofore been a matter of conjecture only. With the exception of phenolic substances, however, none of the well-known compounds of t h e coal-tar distillation have yet been recognized. PITCH-The residual pitch from the t a r distillation apparently has little value as pitch, but its nitrogen content runs from 3 . 5 t o 5. I 5 per cent and this nitrogen is in such a form a s t o permit the preparation of Prussian blue by fusion of the pitch with iron turnings and caustic alkali in an iron crucible. The yield of pigment is very small, but i t must be remembered t h a t only a small amount of nitrogen is ever recovered as cyanide by this process. If the pitch should ever be prepared in large quantities the production of Prussian blue might assume great importance as a kelp byproduct. Kelp pitch is a brittle substance with a speciThere was little, fic gravity of I . 17,and melts a t 130'. if any, free carbon in the samples t h a t were tested. E X T R A C T I O N O F DISTILLATES WITH ACID A N D ALKALI

-The manipulation of this oil with acid and alkali is by no means simple, as the oil has a great tendency t o emulsify and t h e resulting foam is broken up only with considerable difficulty. Slow and careful mixing is the only way t o prevent this trouble. The best strength of acid and alkali for this work was found t o be approximately 4 N . Sulfuric acid and caustic soda were always used b u t an experiment showed t h a t hydrochloric acid of equivalent strength served equally well and had the same solvent action on t h e acid-soluble oil which has already been mentioned under low-temperature tars as separating with the ammonia. The procedure in handling t h e oily distillates was t o effect t h e separation of neutral oil and phenolic substances by shaking out with acid and caustic soda, then t o rectify each of t h e products by distillation. The first extractions with acid and alkali were never complete and i t was necessary to distil the resulting oils and once more t o shake out these distillates with acid and alkali. The first distillates from tar were far from being uniform mixtures, hence t h e percentages of phenols and resiaual oils varied after the extractions of these distillates with acid and alkali. For example, four of these extractions of a n oil t h a t had been collected in Flask A varied among themselves as indicated in t h e following table. Oils collected in Flask B also varied t o some extent. Flask

Residual Oil Per cent 76.9

A , . . ..................

73.1

B .....................

Phenols Per cent 15.3 15.0 13.5 10.2 20.3 22.8

Such figures as the above are by no means accurate for t h e reason t h a t some of t h e aqueous solution is taken up by t h e oil in a n emulsion t h a t is impossible t o break up by any means except distillation. The oil also contains water-soluble constituents. I n one case 5 4 . 0 g. of oil were shaken successively with two 20-cc. portions of distilled water. The weight of residue after evaporating t h e water was 2 g,

T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

7 90 DISTILLATION

OF

RESIDUAL

OIL

FROM

FLASK

A-

When a sample of 217 g. of the residual oil in Flask A was heated with a free flame t o 120°, a mixture of oil and water distilled over. The oil was separated from the water, returned t o the distilling flask, and the di&illation resumed. The residue was rejected. Each fraction turned dark red on exposure t o air. Fractions Grams 12Oo-19O0.. . . . . . . . . . . . . . 6.0 190°-250: . . . . . . . . . . . . . . . 87.0 250"-296 68.0

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

TOTAL ................

161.0

Per cent 2.7 40.0 31.3

74.0

T R E A T M E N T O F FRACTIONS-Distillate 120°-1900: 6 g. shaken out with sulfuric acid and alkali. The greater part of the oil was soluble. Distillate 190~-25oO: 87 g., treated as above. Grams Neutral oil. . . . . . . . . . . . . . . 63.0 Phenols . . . . . . . . . . . . . . . . . . 21.8

Distillate 250'-296':

Per cent 72.4 25.0

68 g., treated as above.

Grams Neutral o i l . . . . . . . . . . . . . . . 5 5 . 5 Phenols.. . . . . . . . . . . . . . . . . 6.2

Per cent 81.6 9:1

A dark oil separated when the sulfuric acid extract was neutralized but there was no odor of ammonia. The total neutral oil recovered amounts t o 54.6 per cent of the crude oil distilled from tar and t o 7.6 per cent of the original t a r itself. Likewise the phenols are only 1 5 . 7 per cent of the crude oil and 2 . z per cent of the tar. D I S T I L L A T I O N O F C R U D E PHENOLS-The crude phenolic acids were purified by distilling and afterwards shaking out the fractions with caustic soda solution.

Vol.

12,

No. 8

The higher kelp distillation temperatures seem t o produce more phenols of cresolic nature t h a n t h e lower temperatures. Likewise a comparison of the residual oils of t h e t a r produced a t 480' t o 650' C. with the high-temperature t a r can best be made by observing the distillation fractions of corresponding temperature ranges as set forth in t h e following table: Tar Distillation 48Oo-65O0 Per cent Fractions 220°-2500 27.5 25O0-3OO0 55.3

I

Per cent 47.4 21.4

T a r 8OO0-95O0 Per cent Per cent 40.0 58.6 31.3 20.7

Per cent 68.I 16.5

The results indicate t h a t the high-temperature distillation of kelp produces a larger amount of lowboiling .oils than the lower temperature distillation does. FRACTIONATIOh'

O F THE R E S I D U A L O I L F R O M K E L P T A R

I t is unnecessary t o take u p the repeated rectifications of the residual kelp oil t h a t were necessary t o secure working material for analysis and further investigation. The final results of the fractionation a t atmospheric pressure, together with the densities of the distillates, are set forth in the following table: Fractions O

c.

175-200 200-2 10 210-220 220-230 230-240 240-250 250-260 260-270 270-280

Weights Grams 58.9 45.3 58.6 48.0 55.6 28.2 100.6 54.1 28.4

Densities 200 c. 0.9.30h 0.9454 0.9517 over sodium 0.9522 over sodium 0.9958 over sodium

....

'

0.9620 0.9700 0.9750

It was noticed t h a t all reacted with sodium as long as t h e metal lasted. The higher boiling fractions Fractions Grams Per cent seemed t o react more vigorously than the lower boil15O0-180: ................ 4.4 2.06 15.7 180"-215 ................ 33.5 ing fractions. 215°-2600................ 79.6 37.3 Fractions Z I O ~ - Z Z O ~ and 220°-z300 had so nearly Fraction 1 j 0 ~ - 1 8 0 (4.4 ~ 9.) was shaken with 11 cc. the same specific gravities t h a t they were combined of 4 N NaOH. and finally rectified by distillation in a vacuum. Grams The oil showed great tendency t o froth when distilled Insoluble oil.. ................. 0.7 Phenols recovered.. ............ 2.7 in vacuum, and this property seems t o be common to Soluble residue.. ............... 1.0 all kelp distillates. It is evident from the above fracFraction 1 8 0 ~ - - 2 1 5(~3 3 . 5 9.) was shaken with 7 7 . 5 tionation table t h a t this substance is a mixture possicc. NaOH. bly of many oils of different boiling points. The steady Grams rise in temperature during the entire process of rectiInsoluble oil.. . . . . . . . . . . . . . . . 2.4 Phenols recovered.. . . . . . . . . . . 29.0 fying the crude residual oil is a rather discouraging Soluble residue.. ............. 2.11 indication t h a t individual compounds can be isolated Fraction 215~--260~(79.6 g.) was shaken with 184 only with much difficulty and in small yields. cc. of the alkali. The product of the first vacuum distillation came Grams over between 7 4 O and 80' C. a t 5 mm. pressure. It Insoluble oil.. . . . . . . . . . . . . . . . . 5.2 Phenols recovered., ........... 74.8 was pale yellow in color when fresh but turned darker Soluble residue. ............... None on standing. Density 0.9524 a t 2 0 ' ; refractive index Grams Per cent Total' phenols finally separated.. . 106.5 50.0 1.4972 a t room temperature. The oil showed no Insoluble oils.. ................ 8.3 optical rotation. ' Soluble residues., .............. 3.1 The distillate thus prepared (weighing 7 7 . 7 g.) TOTAL ...................... 117.9 was further purified by shaking in a separatory funnel T h e separated phenols were finally all mixed and with 40 cc. of 5 per cent sodium hydroxide solution. 61. 7 g. were redistilled. After washing with water the oil was next shaken with Weight of Fractions Fractions Grams Per cent of Oil 1 5 cc. of saturated sodium thiosulfate solution and 150'-2 15 23.6 38.2 (24.0) once more washed with water. It should be remarked 2 15 0-220; 8.1 13.1 (12.1) 220'-240 15.0 24.3 (33.9) here t h a t the oil still showed a tendency t o emulsify The numbers enclosed in parentheses are correspond- if shaken too violently, also it was noticeable t h a t the ing percentages of distillate from the t a r acids made wash water was slightly colored after being shaken at temperatures ranging from 480° t o 650° (p. 788). with t h e oil. Under these circumstances, too, t h e

T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING CHEMISTRY

Aug., I 9 2 0

oil became sticky and adhered t o the glass walls of the funnel, which made a good separation from water more difficult. The oil was dried over anhydrous sodium sulfate and transferred t o a distilling flask from which i t was finally distilled over metallic sodium i n a vacuum of 3 mm. During the distillation the oil required much care t o keep down spurting. The main distillate came over between 73' and 80' C. and was immediately analyzed by combustion. Any formula t h a t might be derived from this analysis has no significance because subsequent treatment with bromine showed the oil t o be still a mixture. Weight taken, 0.1560 g. Weight of water, 0.1299 g.

Weight of carbon dioxide, 0.4730 g

c .......................... H ...........................

Found Per cent 82.7 9.2

Another fraction, 250'-260', was also rectified in The first drop came over a t 9 1 ' a t 1 8 mm. pressure and t h e following fractions were collected: zlacuo.

Fractions

Grams 2.7

9 1 '-1 12' 1 120-12 10 121 O-132' 132'-136'

Fraction

II2' -I

2I

Weight taken, 0.1548 g. .of water, 0.1300 g.

27.0 26.3 8.4

'

was analyzed by combustion.

Weight of carbon dioxide, 0.4678 g.

c . . . . . ..................... H... ........................

Weight


This analysis shows a close agreement with t h a t of t h e distillate just described. Bromination in ether solution, however, showed t h a t both fractions are mixtures of a t least two different compounds. Five grams of the 73' t o 80' distillate were dissolved in 2 0 cc. ether and 5 . 5 g. of bromine added from a buret while keeping t h e ether solution cooled by ice water. A dark, heavy oil immediately settled out and the ether solution was colored slightly red. The ether was decanted from the heavy oil and the solvent driven out by a blast of air, leaving a residual oil which was next washed with sodium carbonate solution t o remove free hydrobromit acid, etc. Weight *of ether-insoluble oil remaining, I . 8 g. ; of ether-soluble oil, 6 . 5 g. The ether-insoluble oil is very smeary and gives n o promise of being a satisfactory working material. The ether-soluble oil, however, is more promising. When a small quantity was heated in a test tube there were indications of decomposition, so the oil was first subjected t o a steam distillation. Five grams of oil were weighed into a flask and I O cc. alcohol added. During the steam distillation a pale yellow oil with a sharp odor came over. When distilled at atmospheric pressure a nearly colorless oil was obThere tained which boiled between 2 1 0 ' and 220'. was some indication of decomposition, showing t h a t only vacuum distillations should be employed. A Q U E O U S DISTILLATE F R O M K E L P

This distillate corresponds t o the so-called "ammonia liquor" from coal-gas retorts. It is a n amber-colored solution of ammonium salts and organic matter. The

79 1

liquid usually reacts alkaline t o litmus, b u t not in every case. I t s specific gravity ranges from 1 . 0 1 in t h e first runnings t o 1 . 0 6 5 in the later distillates. These aqueous products, as made in the oil-jacketed retort in the Forest Products Laboratory, were always acid t o litmus paper, but those made a t higher temperatures were always alkaline. The acidity was never established because, as the end-point was approached in titration, the solution became so turbid t h a t i t was impossible t o make out when neutrality was reached. When kelp was heated in a wood retort t o t h e temperature t o which wood is subjected in distillation, there were many marked differences in the behavior of the two substances. With wood a n exothermic reaction sets in a t about 230' and the heat of this reaction often raises t h e temperature t o 400') while with kelp there was no appreciable exothermic action. The greater part of the alcohol and acetic acid are formed, in the case of wood, during this spontaneous decomposition, and as there appears t o be no corresponding breaking up of cellular tissue in kelp distillation, one can hardly suppose t h a t any appreciable quantity of methyl alcohol and acetic acid is produced under such conditions.l The residue, on evaporation of I O cc. of the aqueous liquor, varied according t o the temperature employed in the distillation. The residues from the low-temperature distillations a t the Forest Products Laboratory averaged 9 . 3 3 g. per IOO cc., while those obtained from the liquors produced a t Summerland ranged from 3 . 6 6 g. t o 7 . 9 3 g. per I O O cc. The condensation of the distillate in t h e wood-distilling retort was very complete, but in the large retorts of the government plant there was much greater opportunity for the escape of steam and of substances volatile with steam. The ammonia content of t h e low-temperature distillate averaged 0 . 3 2 per cent of NH3, which corresponds t o a 1 . 7 oz. ammonia liquor, i. e., one gallon of the liquor would furnish I . 7 oz. of ammonium sulfate. The kelp liquor made in Summerland a t high temperature, however, always smelled of ammonia. One sample was found t o contain 0 . 7 6 per cent NHI, which corresponds t o a 4 oz. gas liquor, Such a liquor is well worth working u p for ammonia recovery, T h e residue, after evaporation of this aqueous distillate, is a pasty, amber-colored substance with a faint odor, and is a mixture of compounds including ammonium salts. When 1 0 9 . 7 g. of this substance were distilled on a sand bath a small amount of water mixed with a few drops of a very evil-smelling oil and some ammonia first came over a t 104'. After this the temand t h e principal perature rose rapidly t o over 200' distillate followed. The principal fraction (275'-285') was a liquid of reddish color with a strong green fluorescence, and 1 D. R. Hoagland, "Study of the Organic Constituents of Pacific Coast Kelps," J . Agr. Res., 1, 39.

792


somewhat less viscous than glycerol. It was soluble in water and alcohol and had a slightly astringent burning taste. I t s water solution was mildly alkaline t o litmus paper and decidedly so t o methyl orange, but a qualitative test for nitrogen was negative. The specific gravity at 23’ was 1.3431 and the refractive index was 1.5013 a t 28’. There was no optical rotation. Products Grams 4.6 9.0 59.5 19.5

Temperature

c.

104-186 186-275 275-285 285-295

The fluorescence disappeared on dilution with water; hence it must have been due t o slight amounts of foreign substances. When chilled in a freezing mixture of ice and salt the liquid showed no tendency t o solidify. Several unsuccessful attempts were made t o form a n acetyl derivative with acetic anhydride. When distilled in a vacuum there was a t first a great tendency t o frothing, so t h a t a 500 cc. distilling flask was needed t o prevent the liquid from being carried over with t h e vapors. The final rectification of these distillates was made in a 4 mm. vacuum. The weight taken was 80 g. Distillation commenced a t 75’ but most of the distillate came over a t 1 2 jo. The amount collected was 6 j . 9 g., or 82.3 per cent. The distillate finally obtained was a straw-colored liquid which distilled between 109’ and 113’ a t 2 mm. An analysis by combustion gave the following results: Weight taken, 0,2107 g. of water, 0.1318 g.

Vol.

12,

No. 8

from the residue on evaporation might be used as a glycerol substitute for mechanical purposes. ACKNOWLEDGMENT

Full acknowledgment is here made t o the officials of the Forest Products Laboratory for the use of space and apparatus, and for courtesies extended; likewise t o Prof. E. B. Fred, of the University of Wisconsin, for use of laboratory space and apparatus. THE TESTING OF SACCHARIMETERS BY MEANS OF THE TELESCOPIC CONTROL TUBE1 By C. A. Browne NI3W YORK SUGAR

TRADE LABORATORY, h C . , 80 NEWYORK,N. Y.

SOUTH

ST.,

A graduated telescopic polarization tube in saccharimetric analysis was first used by Jellet2 in 1864. Although i t possesses applications which greatly simplify certain problems of analysis, the possibilities of this ingenious instrument have been generally overlooked. I n 1884, just twenty years after the publication of Jellet’s paper, Schmidt and Haensch3 described a telescopic observation tube for controlling the accuracy of their saccharimeters, t o which they gave the name of “Control Observation Tube.” . . . .

Weight of carbon dioxide, 0.3668 g. Weight

c ................... H ...................


Calculated for CloHlaOe Per cent 51.7 6.9

Considerable work is necessary t o identify this material fully, but i t is undoubtedly a hydroxyl compound pf the glycol Or glycerol series* This is supported by the fact that it liberates hydrogen gas when treated with sodium. Its close resemblance to glycerol in physical properties indicates that it could doubtless be recovered from the aqueous distillate in the ordinary distilling apparatus used for the recovery of glycerol from liquors, and it might be of value as a glycerol substitute for mechanical purposes. SUMMARY

I n the foregoing account it has been shown t h a t kelp t a r can be separated into groups of compounds which may be designated t a r acids, neutral oils, ammonia, and pitch. These groups have been studied and their general characteristics determined, although the chemical identity of the constituent compounds has not been established. The crude kelp creosote has a high toxicity and the residual kelp oil has value as a flotation oil. These products would find a ready market. The pitch residue from kelp-tar distillation is a source of nitrogen for Prussian blue and cyanogen compounds. The aqueous kelp liquor contains ammonia in workable quantities and the water-soluble distillate

..

.

.

,.

FIG. 1-TELESCOPICCONTROL TUBE

The control tube, as manufactured by Schmidt and Haensch and as described by Landolt4 in his classic work upon L(Optical Rotation,,, consists of a telescopic metal tu& (Fig. I) which can be adjusted by means of a Screw to give a column of solution for any length between 2 2 5 mm. and 410 mm. By of a vernier t h e exact length of solution a t any point within this range can be read upon the scale to o,I mm. Before using the control tube i t is important t h a t its scale be accurately verified. The tube employed in the present work showed by the Landolt gage, a t the time of purchase, a n error of about 0.2 mm. a t all points of the scale. A more accurate calibration by the Bureau of Standards showed errors ranging from 0.15 mm. a t the 410 point of the tube scale t o 0.24 mm. a t the 2 3 0 point. The control tube has been subjected constantly t o a number of uses in the work of this laboratory during the past ten years. It has been employed: (I) for 1 Read before the Suaar Section a t the 59th Meetine - of the American Chemical Society, St. Louk, Mo., April 12 t o 16, 1920. 2 Proc. Roy. Irish A c a d . , 8 (1864), 279. 3 Z . Instrumentenk., 4 (1884), 169. 4 “Das Optische Drehungsvermijgen,” 2nd Ed., 1898, p. 398.

Potash from Kelp. IIIâThe Preliminary Examination of Kelp Distillates

Recommend Documents

Potash from Kelp. IIIâThe Preliminary Examination of Kelp Distillates