Countercurrent Distribution of High-Boiling Phenols from Low

Karr, P. A. Estep, and L. L. Hirst. Anal. Chem. , 1960, 32 (4), pp 463–475. DOI: 10.1021/ac60160a005. Publication Date: April 1960. ACS Legacy Archi...
0 downloads 0 Views 1MB Size
(2) Block, R. J., ANAL.CHEM. 22, 1327

(1950). (3) Bolling, D., Soper, H. A., Block, R. J., Federation Proc. 8, 185 (1949). (4) Closs, X., Haug, C. M.,Chem. & Znd. (London) 1953, 103. (5) Curzon, G., Giltrow, J., A-ature 172, 356 (1953). (6) Dagliesh, C. E., J . Chem. SOC.1952, 3940. ( 7 ) Dunn, M. S., Rockland, L. B., Advances in Protein Chem. 3,296 (1947). (8) Ebrey, P., Chemist Analyst 48, 36 (1959). (9) Fowden, L., A'uture 167, 1030 (1951). (10) Fujisawa, Y., J . Osaka City Med. Center 1, 7 (1951).

(11) Kotake, hl., Sakan, T., Nakamura, N., Senoh, S.,J . Am. Chem. SOC.73, 2973 (1951). (12) Lambooy, J. P., Zbid., 76, 133 (1954). (13) MdFarren, E. F., ANAL.CHEM.23, 168 (1951). (14) Mann, F., Mann Research Lab., 136 Liberty St., S e w York 6, N. Y., personal communication, April 8, 1959. (15) Rlehta, C. J., thesis, Univ. California at Los dngeles, 1957. (16) Meister, A., Levintow, L., Kingsley, R. B., Greenstein, J. P., J. Biol. Chem. 192,535 (1951). (17) Rockland. L. B.. Underwood.' J. C.. ~ ; X A L .CHEM:26, 1557 (1954).

(18) Toennies, G., Kolb, J. J., Zbid., 23,823 (1951). (19) Weichert, E., Acta Chem. Scand. 9, 547 (1955). (20) Wieland, T., Reiberg, O., Schon, W., unpublished results, 1955; quoted by Muller, E., et al., eds., "Methoden der Organischen Chemie (HoubenWeyl)," 4th ed., Vol. 11, Pt. 2, p. 319, G. Thieme, Stuttgart, 1958. (21) Zweig, G., ANAL. CHEW 31, 821 (1959). RECEIVED for review September 8, 1959. Accepted January 4, 1960. Work supported by grants from the American Cancer Society, U. S. Public Health Service, and University of California.

Countercurrent Distribution of High-Boiling Phenols from a Low-Temperature Coal Tar CLARENCE KARR, Jr., PATRICIA A. ESTEP, and LESTER 1. HIRST, Jr.' low-Temperature Tar laboratory, Bureau of Mines, U. S. Department of the Inferior, Morgantown, W. Vu.

b The countercurrent distribution of tar acids described here i s the first full demonstration of the capabilities of this method of fractionation for complex mixtures of high-boiling phenols. Tar acids boiling at 231 ' to 331 ' C. from a low-temperature bituminouscoal tar have been fractionated and analyzed by distillation and countercurrent distribution. The presence of 84 individual compounds was demonstrated; 24 were identified with respect to specific isomers and 36 with respect to structural type, the remainder being tar acids of unknown structure. Quantities of all components were determined or estimated.

A

s

of an extensive investigation of the identity of the constituents of a low-temperature bituminous coal tar, it was desired to fractionate and analyze the high-boiling phenols. The method of analysis chosen for this complex material 1%as countercurrent distribution, supplemented by ultraviolet and infrared spectrophotometry. Complel mixtures of phenols are frequently encountered in various fields of research, but the capabilities of countercurrent distribution for fractionation of these substances have never been fully demonstrated. The only previous use of countercurrent distribution for high-boiling phenols is that described by Golumbic et al. (6) for material obtained from coal-hydrogenation oils. This work was done more PART

Present address, California Institute of Technology, Pasadena, Calif.

than 10 years ago, before modern equipment was available with such advantages as all-glass construction, higher number of transfers, greater sample capacity, fraction collector, and completely automatic operation. I n addition, a much larger selection of authentic specimens and spectra has become available in recent years for qualitative confirmation. The net result of these new factors is an analytical method especially suited to extremely complex mixtures of high-boiling phenols, EXPERIMENTAL

Preliminary Fractionation by Distillation. A 200-gram portion of t a r acids, representing 3.54 weight yo of a low-temperature bituminous coal tar, was distilled a t 20 mm. of mercury through a column filled with glass helices. All material boiling u p t o 118' C . head temperature, equivalent t o 232' C. a t 760 m m . of mercury, was removed as a single fraction, leaving a residue of 50 grams of highboiling phenols, or 25 weight yo of the original tar-acid mixture. A charge of 41.68 grams of high-boiling tar acids was distilled at 2.9 mm. and a reflux ratio of 20 to 1 in a spinning band still, with the results shown in Table I. Infrared spectra were obtained on all fractions, which were then combined on the basis of qualitative similarity to give 10 samples. All samples except the lowest boiling one were subsequently fractionated by countercurrent distribution. Countercurrent Distribution. T h e instrument used was a 60-tube allglass model, x i t h 200 tubes in t h e fraction collector and a n automatic robot mechanism (IO). The tube capacity for each phase was 40 ml. The

Table 1. Fractionation of High-Boiling Tar Acids in Spinning Band Still

Fraction No. 1

2 3 4 5 6 7

8 9

10

11

12 13 14 15 76 -_

17 18 19 20 21

Boiling Range, C. Lvt., 2.9 mm. 760 mm. Grams 77.5- 78.0 231 0.71 7 8 0- 79 0 231-232 1.31 7916- 84:O 232-238 1.10 1.35 84.0- 8 8 . 0 238-244 1.14 88.0- 91.0 244-247 1.00 91.0- 94.0 247-251 1.40 94.0- 96.0 251-253 96.0- 98 0 253-255 0.86 0.86 98:0-102.0 255-258 0.85 102.0-103.0 258-260 0.51 103.0-106.0 260-264 0.96 106.0-110.0 264-270 0.87 110.0-116.0 270-278 1.15 116.0-117.0 278-280 1.15 117.0-124.0 280-290 124 1.17 - - - .0-729 - -. 0 . 290-297 129 .O-132.0 297-300 1.08 0.96 132.0-139.0 300-308 0.85 139.0-152.0 308-325 162.0-157.0 325-331 0.88 331 0.24 157.0

instrument was operated to give 100 to 105 transfers or plates. The average sample size was about 185 mg.; the upper phase consisted of Spectro grade cyclohexane, and the lower phase was a phosphate buffer made from different proportions of 0.5N Na3P04and 0.5-11 Na2HP04. A series of separations was made a t pH varied between 9.94 and 11.86. A particular value was chosen because of the suspected presence of certain phenols and their known partition coefficients. After the completion of each fractionation, 8 ml. of 1 to 1 hydrochloric acid was added to the tube to neutralize the buffer and mixed well to dissolve the phenols in the cyclohexane. Ultraviolet spectra were obtained on VOL. 32, NO. 4, APRIL 1960

463

each cyclohexane solution, and plots were prepared of total absorbance at two informative wave lengths us. tube number. The curves are shown in the upper halves of Figures 1 through 9, inclusive. On the basis of the curves and the qualitative similarity of the ultraviolet spectra, cyclohexane solutions were combined for infrared analysis. Almost all of the cyclohexane had to be evaporated off to obtain satisfactory spectra. To avoid the loss of easily sublimed phenols, the procedure \\-as as follons:

6

. _

About 150 ml. of Cyclohexane solution containing, on the average, 6 mg. of phenols, was placed in an ordinary distillation flask with a side arm and a cold finger maintained at 0' to 5' C. The cyclohexane was almost completely removed in about 45 minutes, with no detectable loss of phenols. by reducing the pressure to about 65 mm. and immersing the flask in a mater bath maintained at 40' to 50' C. The volume of the solution was reduced to 0.25 ml., just enough to fill a 0.5-mm. sodium chloride cell. A matched cell containing cyclohexane was used in the reference beam to obtain the infrared spectrum. Determination of Distribution Curves. On t h e basis of t h e combined qualitative results of t h e ultraviolet and infrared spectra, t h e constituents froin each countercurrent distribution fractionation were distinguished and assigned numbers (Table 11). T h e distribution curve of each constituent was readily visualized by following the appearance and disappearance of characteristic absorption bands in the spectra of consecutive countercurrent fractions. This was made easier by the fact that the distribu-

COLLECTOR^

-FRACTION

TUBE NUMBER

Figure 1.

Countercurrent distribution of phenols boiling at

238' to 251

C.

100 transfers; pH 11.58

Countercurrent Distribution of Phenols Observed Analytical Wave Lengths Spectrum Partition (UV in Mp, IR in p ) Reference Coefficient Boiling at 238-251" C., 100 Transfers, pH 11.58 0 285.2, 275.2, 252.5, 245.5 mp a 0 . 14b 281.2, 273.8 mp D 0. 16b 285.0,278.7 mp, 13.72, 12.52,12.32,11.85, 11.58, 10.57,9.95,9.80,8.97,8.66,8.42, 7.90, 7.73 !J 276.4,271 .O, 268.7 mp, 14.25, 13.07, 12.98, a 0 .31h 10.18, 10.00, 9.52, 8.72, 8.32, 7.85, 7.62 Table II.

No.

Constituent Identity

2 3

UnknovnI 3,5-Xylenol 3,4-Xylenol

4

4-Indanol

5

5-Indanol

6

3-Ethyl-5-methylphenol

7

3-Ethyl-4-methylphenol

8

4%-Propylphenol

1

P

464

ANALYTICAL CHEMISTRY

289.6,283,3,280.2m p , 14.45,13.49,12.52. 12.32, 11.91, 11.65, 10.68, 9.23, 8.82, 8.52, 7.92 p 280.9, 276.5,273 3,271 .O, 267.0,264 0 mp, 14.47, 12.02, 11 86, 11 67, 11.07, 10.40, 8.68, 8.58 12.40, 11.49, 10.81, 8.61, 8.31, 7.95, 7.70 u 285.4, 278.9, 276.0, 273.0, 270.0, 267.0, 264.0, 261.0 mp, 14.20, 13.08, 12.62, 12.20, 9.02, 8.56, 6.19 p

D

(7)

Peak Tube Calcd. Exptl: 0

12 13

12 14

23

21

0.3jh

26

26

0.59b

37

37

0.64 0 .87h

39 46

41 (Continued)

Table II. KO.

9

Constituent Identity 2,3,5-TrimethylphenoI

11

2,6-Dialkylphenol (2-ethyl6-n-propylphenol ) 2,3,5,6-Tetramethylphenol

12

4-Indenol

13

5-Indenol

4 5 14 15

PIndanol 5-Indanol Unknown I1 3,4,5-Trimet hylphenol

16

1-, 2-, or 3-methyl-5-indanol

10

( l-methyl-5-indanol?)

17

1-,2-, or 3-methyl-4-indanol

18

3,5-Dialkylphenol (3-methyl-5-n-propylphenol) 3,4-Dialkylp hen01 (4-ethyl3-n-propylphenol)

19

( 3-methyl-.i-indanol?)

20

Methyl-1-inclanone (4met hyl-1-indanone? )

21

Methyl-1-indanone (6methyl-1-indanone? ) 5-Indenol

13 5

22 16 23

5-Indanol 2,3- or 2,6-alkylalken-lylphenol (3-methyl-2propen-1-ylphenol ) 1-,2-, or 3-methyl-5-indanol ( l-methyl-5-indanol?) 3,4-Dialkylphenol (3-ethyl4-n-propylphenol )

24

2-Phenylphenol

20

25 26 27

Methyl-1-indanone (4methyl-1-indanone) 3,4-Dinuclearphenol Alkenylphenol I 1-, 2-, or 3-methyl-4-indenol

28 29 30

Unknown I11 Cycloalkenylphenol I? Methylindenol

31 32

Alkenylphenol I1 7-Methyl-5-Indanol

33

Methyl-5-inclanol ( 6methyl-5-indanol) 2,4- or 3,4-dialkylphenol?

34

Countercurrent Distribution of Phenols (Confinued)

Observed Analytical Kave Lengths Spectrum (UV in M p , I R in p ) Reference Boiling at 238-251" C., 100 Transfers, pH 11.58 a 282 4,277 9,273 4 mp. 12 06, 11 90, 10 30, 9 2 7 , s 77, 8 62, 8 17, 7 88, 7 62 p d 285 0, 277 5 mp, 13 83, 13 08 p 13.08, 12.95, 12.06, 9 . 1 7 , 8.46 p

(1

Boiling at 251-258' C., 100 Transfers, pH 11.20 299.5, 287.8, 252.5 mp, 14.45, 13.45, a 13.42, 12.20, 11.10, 10.98, 10.60, 10.00, 6.42 p 307.0,290,268,258 mp, 14.60,14.48,13.65, d 13.00, 10.52,9.40,9.00,8.25,6.42,6.29, 6.27 p 276.4, 271.0, 268.7 mp 289.6, 283.3, 280,2 mp a (KOdistinctive bands) 284. 7, 280.0, 276.0 mp, 14.33, 13.52, 12.04, a 8.82, 8.44, 8.06 p 288.8, 284.0, 280 mp, 12.54, 12.32, 11.95, d 10.68, 9.24, 8.55 p 276.5,271.5,269.0 mp, 14.27, 13.12,12.98, d 10.21, 8.73, 8.34 p 280.3, 276.6, 273.4 mp, 14.47, 12.04, 11.95, d 10.38, 8 . 7 2 p d 285, 278.3 mp, 14.34, 13.60. 12.84, 12.45, 12.35, 11.87, 10.53, 10.37, 8 . 7 0 ~

-,

(I

Boiling at 258-260' C., 100 Transfers, pH 10.49 314.0, 302.5, 296.0, 278.0, 257.4, 253.5, d 249.5, 245.6, 242.0 mp, 12.95, 8.65, 5.92 p 314.0, 278.0, 271.0, 257.5, 253.5, 249.7, d 242.0 mp, 5.93 p 307.0, 290.0, 280.0, 268.0, 258 mp, 14.60, d 13.00, 10.52, 10.38,8.25, 6.27 p 289.5,284.0, 280.0 mp, 14.45. 13.47, 12.55, 12.32, 10.68, 9.25, 8.82, 8.55 p d 296.4 mp, 14.10, 12.88, 10.20 p 288.5,282.9,279.4,274.5mp, 14.45, 12 .07, 12.30, 9.24, 8.55 p 280.5,273.5mp, 14.35, 13.61, 12.37, 11.88, 9.21,9.02,8.90,8.68p 14.25, 13.90, 13.70, 13.32, 13.02, 12.04 p

2.6b

Peak Tube Calcd. Exptl.' 25FC

23FCc

6FC

9.8 59b 0 . 13b

1FC

1FC

11

13 24

0.31 0.6Sb 0.79b 0.96 1.F

41 45

39 45 50 56

55

3.3

18FCc

3.8

16FC

6.0

lOFC 4FC

15

0.35

26

1.0

56

2.0

30FCc

2.96

24FC

20FC

20FC

3.0

d

13

5FC

d

13

5FC

(i

Boiling at 260-270' C., 101 Transfers, pH 11.86 314.0,302.0~257.0,253.0,249.0,246.0mp d 284.8, 280.7, 275.5 mp 292.0 mp 309.6, 299.7, 288.6, 265.0, 260.7, 254.5, 249.3 mp, 14.14, 13.15, 13.00, 12.89, 12.77, 12.42, 10.72, 10.51, 9.85, 9.15, 7.25, 6 . 2 8 , 6 . 1 7 , 6 . 1 4 p 303.0, 240.0 mp 253.0 mp 296.5, 261.0, 254.5, 249.3 mp, 10.42, 8.71, 8.45 p 308.0 mp 287.8, 283.0,279.5mp, 14.47, 13.10, 12.07, 10.53, 10.20,9.87,8.92,8.47, 8.07, 7.58, 7.52, 7.26, 6.22 p 287 8,283 0, 279 5mp, 14.45, 13.60, 12.60, 12 50, 12.10, 11.97, 10.52, 1 0 . 0 5 ~ 287 5, 281.5, 277.0, 272.5 mp

Partition Coefficient

d

d

e

d

34b

2FC

4FC

0.020

2

0.020 0.086 0.19

2 8 16

0.38 0.51 0.60

28 34 38

0.71 1.2

42 55

1.4

58

1.6

37FCc (Continued)

VOL. 32, NO. 4, APRIL 1960

465

Table II. KO.

35 24

Constituent Identity 2,4- or 3,4-dialkylphenol(P isopropyl-3-n-propylphenol) 2-Phenylphenol

36

2,4- or 3,4-dialkylphenol (2,4di-n-propylphenol)

37 38

Unknown IV Unknown V

39 40

Unknown VI 2-Naphthol

41 42 43 27

Unknown VI1 Ketone (alkyl indanone) Alkenylphenol I11 1-, 2-, or 3-methyl-4-indenol

28 30 44

Unknown I11 Methylindenol 5,6,7,8-Tetrahydro-2-naphthol UnknoFn VI11 5,6,7,8-Tetrahydro-l-naphthol 2-Phen ylphenol

Countercurrent Distri-

Observed Analytical Wave Lengths Spectrum Partition Reference Coefficient (UV in M p , IR in p ) Boiling a t 260-270" C., 101 Transfers, pH 11.86 d 287 5, 280.5, 273.4 mp 4.5

Peak Tube Calcd. Exptl.'

283.0,245.5 mp, 14.25, 13.90, 13.70, 13.32, 13.02, 12.04, 8.49 p 285.0, 279.0, 273.0 mp, 12.45 p

13FC

a

d

Boiling a t 270-280" C., 100 Transfers, pH 10.28 248.0 mp 307.5,301.0,296.0,286,0,265.4,256.5 mp. 14.57, 13.50, 13.00, 12.50, 10.84, 9.24 9.15, 9.07, 9.02, 8.78, 8.52, 8.32, 8.12, 6.34 p 243.5 mp, 14.69, 12.74 p 328.6, 324.0, 321.0, 314.0, 310.0, 307.0, 301.0, 285.4, 273.8, 263.5, 254.0 mp, 13.47, 12.52. 12.41, 11.97. 10.42. 8.95. 8.75,8.60, 6:60, 6.22, 6 . 1 2 ~ (KOdistinctive bands) 314, 263.8, 259.0, 255.0, 218.0 mp, 5.92 p 301.0 mp 309.6, 299.7, 288.6, 265.0, 254.5, 249.3 mp, 14.14, 13.15, 13.00, 12.89, 12.77, 12.4'2, 10.72, 10.51, 9.85, 9.15, 6.28, 6.17, 6.14

4.55

13FC

1FC

59

13 18

0.15 0.22

0.33 0 .5gb

13FC

37

25 37

0.79 1.0 1.2 2.2

44 50 56 27FCc

3.9 5.4 6.7b

9FC

l5FC llFC 9FC

9.8 25 ( 1 4 )

2FC

6FC 1FC

51b

1FC

1FC

48

24 35 48

Ir

45 46 24

38 47 40

48 49 50

51

52 53 54

Unknown V Cycloalkenylphenol II? 2-Sapht hol

Unknown I X I-, 2-, or 3-polyalkyl-Pindenol Methyl-2-naphthol (4methyl-2-naphthol? ) Dimethyl-1-naphthol (5,7dimethyl-1-naphthol?) 4-Methyl- 1-naphthol Methyl-2-naphthol (3methyl-Bnaphthol? ) 1-Methyl-2-naphthol

303.0,240.0 mp,14.55,10.87, 10.10, 8.54 p 296.5, 254.5, 249.0 mp 287.8,279.5 mp, 14.45,13.60,12.60, 12.50, 12.10, 11.97, 1 0 . 5 2 ~ 256.0 mp,10.40, 10.25, 8.92,8.70,8.50 p 279.0,273.0mpJ 14.10, 13.04, 1 2 . 0 7 ~ 283.5,245.5 mp, 14.25,13.90,13.70, 13.32, 13.02, 12.04, 8.49 p Boiling at 280-297" C., 101 Transfers, pH 9.94 307.5, 296.3, 265.0, 255.8, 243.5 mp 304.5,292.5,273,7,264.5,256.5mp, 13.30 p 328.6, 324.0, 321.0, 314.0, 310.0, 307.0, 301.0, 285.4, 273.8, 263.5, 254.0 mp, 13.47, 12.52. 12.42, 11.98, 10.42, 8.94, 8.80, '8.60, 6'.60, 6.24, 6.12 p (No distinctive bands) 312 (sh), 307.5, 299.5, 296.3, 265.0, 256.0 mp, 14.65,13.12,12.40,9.13,6.33,6.14p 331.5, 327.0, 324.2, 317.0, 290.5, 276.0, 267.0, 258.0 mp 327.4, 320.0, 312.0, 306.0, 303.0, 265 .O mp 326.0,319.0,312.0, 303.0,290.0 mp,13.15, 12.40, 6.30 p 331.5, 317.0, 290.0, 278.0 mp 335.0,320.0,304.0,265.0 mp, 13.48, 13.72, 12.45, 12.34, 10.55, 10.10, 9.83, 9.45, 8.82, 7 . 4 5 , 6 , 6 3 , 6 . 2 4 , 6 . 1 5 @ 328.2, 321 .O, 314.0, 300.0, 290.0 mp, 12.52

d

(1

1.2 1.8

55 32FC0

2.5

24FC

3.7

16FC

5.5b

llFC

lOFC

d

6.6

9FC

(11)

7.4

8FC

d

56

Alkyl-1-naphthol (2-ethyl1-naphthol?) 2-Cyclohexylphenol

279 mp, 13.30 p

58 47

Unknown V Cycloalkenylphenol II?

Boiling at 297-300" C., 100 Transfers, pH 10.10 307.5, 296.3, 265 .O,255.8, 243.5 mp 304.5, 292.5, 273.7,264.5, 256.5 mp, 13.30

80

2-Naphthol

65

0

0.31 0.53 0.91b

2FC

30

P

(4, 6)

2400 (14)

0-1FC

1FC 20 30

0.25 0.43

P

A66

ANALYTICAL CHEMISTRY

328.6, 324.0, 321.0, 314.0, 310.0, 307.0, 301.0, 255.4, 273.8, 263.5, 254.0 mp, 13.48, 12.52, 12.42, 11.98, 10.42, 8.60 p

0

0.72b

42

42

bution of Phenols (Confinued) KO.

49 57 58 50 59 60 54 61 62 55 56 40 49 50 60 63 64

Constituent Identity 1-, 2-, or 3-polyalkyl-4indenol Unknown X Unknown XI Methyl-2-naphthol (4methyl-%naphthol? ) Unknown XI1 CPhenylphenol 1-hlethyl-2-naphthol Dimethyl-1-naphthol (2,5or 2,7-dinic?thyl-l-naphthol?) Dimethyl-1-naphthol (2,6dimethyl- 1-naphthol ? ) Alkyl-1-naphthol (2-ethyl1-naphthol? ) 2-Cyclohexylp henol

2-Kaphthol 1-, 2-, or 3-polyalkyl-4-indenol Methyl-2-naphthol (4 methyl-2-naphthol?) 4-P henylp henol

70

Unknown XI11 Methyl-2-naphthol (7methyl-2-naphthol? ) Unknown XIV Unknown XV Unknown XVI: Unknown XVII Unknown XVIII Alkyl-1-naphthol (2-ethyl1-naphthol? 2-Alkylcycloalkylphenol I

40

2-Naphthol

71 60

Unknown XIX 4-Phenylphenol

65 66 67 68 69 55

72

Observed Analytical Wave Lengths Spectrum (UV in M p , I R in p ) Reference Boiling a t 297-300' C., 100 Transfers, pH 10.10 312 (sh), 307.5, 299.5, 296.3, 265 .O, 256.5 mpJ14.65,13.12,12.40,9.13,6.33,6.14p

(No distinctive bands) 309.0, 303.0, 297.0, 285.0, 276.0 mp 331.5, 327.0, 324.2, 317.0, 290.5, 276.0, 267.0, 258.0 mp, 13.40 p (No distinctive bands) 255.6 mp, 14.60, 14.05, 13.20, 12.05 p 335.0, 320.0, 304.0, 265.0 mp 327.0, 311.0, 303.0, 290.0, 280.0 mp

d

a

(11)

331.5 mp 328.2, 321 .O, 314.0, 300.0, 290.0 mp, 12.52 p 279 mp, 13.30 p

Peak Tube Calcd. Exptl.

1.4

58

1.6 1.9 2.5

38 31FCc 24FC

2.9 3.7b 5.5 7.9

16FC

d

21FC 16FC llFC 7FC 6FC

9.3

3FC

20 2100 ( 1 4 )

0-1FC

2FC

0. O4Ob

5

0.050

4 5

d

0.094

9

a

0.11b

(4, 6 )

Boiling a t 300-325" C., 105 Transfers, pH 11.85 a 328.6, 314.0, 285.0, 273.8 mp, 13.48 p 307.5, 299.5, 296.3, 256.5 mp 331 .O, 317.0,290.5,276.0, 267.0,258.0 mp, 13.40 p 255.6 mp, 14.60, 14.05, 13.20, 12.05, 11.02, 9.30, 8.97, 7.95 p 303.0, 296.0, 290.0 mp 331.0, 317.0, 275.0, 266.3 mp

Partition Coefficient

258.5 mp 305.5 mp 304.0, 294.0 mp 261.5, 255.0 mp 260 mp d 328.2,321 .O,314.0,300.0,290.0 mp, 14.53, 13.48, 12.52, 12.12, 10.07 p 287.0 mp, 13.30 p d Boiling at 325-331' C., 100 Transfers, pH 11.85 328.6,286,0,279,5,273.0mp, 14.00, 13.48, 12.52, 12.42, 11.98, 10.42 p 306.0, 299.0, 294.0, 275.2 mp a 256 mp, 14.60, 14.05, 13.20, 12.05, 9.30, 8.97, 7.95 p 331.0 m p 0

10

10

0.44 0.59

32 39

0.72 0.84

44

1.3

45FCc 31FC 20FC 4FC

48

1.9 3.0 15

4FC

15 0.042b

5

4

0.042 0.11b

10

4 10

Methyl-2-naphthol (60.14 methyl-Znaphthol?) a 73 3-P henylphenoi 281 .O, 249.5 mp, 14.40, 13.25, 8.67, 8.56 p 0. 16b d Methyl-2-, -3-, or -4-fluo74 311.0, 298.5, 282.5, 277.0, 258.0 mp 0.20 renol 75 Unknown X X 278.5 mp 0.27 76 Unknown X X I 280.0, 260.7, 254.8 mp 0.54 77 Unknown X X I I 302.0, 282.0, 268.0, 260.5, 255.0, 249.0 mp 0.89 78 8-Methyl-2-naphthol? 14.30, 13.76, 13.16 p d 2.3 Unknown X X I I I 79 (No distinctive bands) 3.5 80 Unknown XXIV (No distinctive bands) 6.6 d Alkyl-1-naphthol (2-ethyl55 328.2,321 .O, 314.0,300.0,290.0 mp, 14.53, 29.5 1-naphthol?) 13.48, 12.52, 12.12, 10.07 p 13.27 p d 81 2-Alkylcycloalkylphenol I1 59 0 This laboratory. Partition coefficient determined from authentic specimen, other values calculated from experimental peak tube. Tube in fraction collector. Covered in discussion. e Coal Tar Research Association, Leeds, England.

12 15

VOL. 32, NO. 4, APRIL 1960

14 16 21 35 47 26FC* 17FC 9FC 2FC 1FC

467

0

0

9

m

m

\

N

I

-g .0

e

0 rc,

0 N

0

I m

0

m

m

rc)

N

I

I

0 N

I 0

-

33NV88OSEV

468

ANALYTICAL CHEMISTRY

I

0

-

LD

o

I

m

I

P

0 m

N

SWV8EIIllIW

-

C

1

I

E

E

this n a v e length, 307 mp. The resulting value of milligrams per milliliter vias multiplied by 40 to obtain the milligrams of 2-naphthol in each tube. This technique could be used even for small amounts of unknoTr n structure. For example, having established the distribution curve for constituent 40, 2-naphthol (upper half of Figure 5 ) , the curve for constituent 42, a ketone. u as similarly obtained. The absorbances in the overlapping parts of these t1Y-o curves nere added, then subtracted from the total absorbance curve to give the shaded area in the figure. This gives the absorbance a t 307 mp due to constituent 41 (a phenol of unknonn structure) in each tube. Use of Partition Coefficients. Partition coefficients for the cyclohexanephosphate buffer system \yere obtained for a large number of authentic specimens of phenols a t a wide variety of pH values. These nere used t o calculate t h e peak tube number, using the folloning equations:

I

,

N

16-

l2-

8W

0

z a

m

4-

(r

0

2

0-

A-FC

2-

=

59

where .T = peak tube number in instrument .TFc = peak tube number in fraction collector n = n u m l m of transfers k = partition coefficient

32

27 24 t

1

0

10

20

30

40

50

41

31

21

I!

F F R A C CTO IL LO E CN TOR-

!

T U B E NUMBER Figure 4.

Countercurrent distribution of phenols boiling a t 260' to 270' C. 101 transfers; pH 11.86

tions are e=.sentially Gaussian as ne11 as symmetrical. Distribution peaks observed in this manner always coincided n i t h peaks in the plot of total absorbance vs. tube number. For example, constituent 6, 3-ethyl-5methylphenol, identified by its characteristic absorption, occurred in tubes 22 through 52. The calculated peak in tube 37 coincided ~ i l hthe total absorbance peak (Figure 1). Consequently, the distribution curves for those constituents producing small shoulders in the total absorbance curve could be delineated. Constituent 8, 4-n-propylphenol, is indicated in tubes 39 through 43 by its absorption a t 285.4 and 278.9 mh, The rewlting distribution peak in tube 41 corresponds with the loner shoulder in the total absorbance curve. The milligrams of each phenol in each

tube were determined from the absorptivities at characteristic wave lengths, as obtained from authentic specimens or literature data. Where such absorptivities a, ere not available, an arerage absorptivity was obtained from phenols with the most similar structure or, for constituents of unknown structure, from phenols with the most similar absorption bands and boiling points. Plots of milligrams us. tube number were prepared (Figures 1 through 9). One useful technique was to construct the distribution curve for the compound on tlie plot of absorbance us. tube number, as described, and as illustrated for constituent 40, 2-naphthol, in the upper half of Figure 5 . The absorbance value for each tube number n a s read from this distribution curre and divided by the specific absorptivity for 2-naphthol a t

Equatioii 1 is the one presented by Williamson and Craig ( I S ) . iThile Eyuation 2 is self-evident. The tubes are numbered from 0 to 59 in the direction of transfer in the instrument and from 0 up in the order of use in the fraction collector. Thus, if the last tube used in the fraction collector is number 40, it r-iill immediately follow tube 59 in the plot of milligrams tis. tube number. If the calculated peak tube agreed \vel1 with tlie experimental peak tube, this constituted additional proof of identity (Table 11). Khere the authentic 'specimen n-as not available, the partition coefficients of the many other available phenols vr-ere the basis of partition coefficient-structural correlations nhich were frequently useful and reliable in establishing the general structure of constituents. Quantitative Analyses. T h c weight' percentages of each constituent in t h e total high-boiling acids (0.89 weight % of the tar) \yere calcwlatcd from the distributioli cumes of milligrams u s . tube number and the aniouiits of t h e yarious distillat'e fractions (Table 111). For completeness, data ohtained b y other tccliniques are included. T h e t a r acids boiling from 231" to 238" C. could be analyzed as well by fractional distillation followed by infrared spectrophoVOL. 32, NO. 4, APRIL 1960

469

0 +

+ 0

du O b 9 1

I

0 N

I

I

-

0

N

-

I m

33NV'BtlOSEI

470

ANALYTICAL CHEMISTRY

I v

I

o

Q

I

E 0

were seen to be mostly long-chain saturated aliphatic carboxylic acids. There could have been a small amount of terminal unsaturation, but there was no internal unsaturation, as this gives a strong C=C stretching band which was not detected in any of the fractions. These acids have such low partition coefficients that they mere always restricted to the first few tubes in the instrument and did not interfere with the determination of the phenols. The weight per cent of total aliphatic carboxylic acids in each boiling range was determined by using the intensity of the C=O stretching band a t about 5.85 microns. A plot of absorptivity us. boiling point for various aliphatic and externally unsaturated carboxylic acids in the range 237" to 305" showed an exact linear relationship-that is, E = -0.305t 133, vhere E = absorptivity for a 0.1-mm. sodium chloride cell and t = boiling point in "C. The results of this quantitative procedure are summarized in Table 111.

+

55

:;14

DISCUSSION

2 5

'. 2

il

0

50

i

t h

0

IO

20

3C

40

50

40

30

20

I

IO

--.. 0

L F R A C T ~ O NCOLLECTOR

,

TUBE NUMBER

Figure 7.

Countercurrent distribution of phenols boiling a t

297"

to

300' C.

100 transfers; pH 10.1 0

tometry. Five phenols were identified that were not found in either the earlier gas-liquid chromatography work (8) or the countercurrent distribution fractionations discussed here : 3-ethyl-2methylphenol, 2-methyl-5-isopropylphenol, 3-n-propylphenol, 4-isopropylphenol, and 2,3,6-trimethylphenol. The evidence for the presence of the last two is good but not as complete as that for the other three. I n addition, 5ethyl-2-methylphenol, which was the structure proposed for an unknown phenol found in the gas-liquid chromatography work (8),has now been positively identified on the basis of recently published infrared spectral data (7'). These six compounds have been included in Table 111 for completeness.

The presence of 1-naphthol in the tar acids was readily determined by the ultraviolet and infrared spectra of the distillate fractions. However, this phenol is easily oxidized in the presence of air when dissolved in alkaline solutions, and none could be detected in the countercurrent fractions. 2-Kaphthol occurred in two distillate fractions in such large amounts that some of it crystallized out. The pure 2-naphthol isolated in this manner was not added to the charge to the countercurrent instrument. Thus, the amounts of l-naphtho1 and 2-naphthol shown in Table 111 were based on the distillate fractions. Aliphatic carboxylic acids were found in every countercurrent fractionation. From their infrared spectra these acids

K i t h respect to identification, the constituents fell into one of three groups. The first group comprised those compounds for which authentic specimens were available. For these compounds ultraviolet and infrared spectra and partition coefficients determined in this laboratory n ere available. The second group consisted of compounds for \\-hi& ultraviolet and/or infrared spectra were available from the literature and, in a few instances, partition coefficients. The third group 71-as composed of compounds for which only the literature boiling point was available. Compounds in the first two groups were nearly a l ~ a y sidentified with complete certainty. Compounds in the third group were almost never identified with complete certainty as to the specific isomer. However, by the cautious application of spectral-struptural correlations, partition coefficient-structural correlations, and boiling range-struct u r d correlations, the general structure of the compound could be established. I n some instances the name of a specific isomer is given in parentheses in the tables. These compounds were selected as most representative of the constituent; the selections must be considered only tentative. A somewhat detailed description of the identification of the compounds in the third group is presented as a general guide for other investigators. Alkylphenols. T h e spectra of constituent 10 indicated t h a t it mas a 2,6-dialkylphenol. I t s partition coefficient of about 10 a t pH 11.58 is only slightly Ion-er t h a n t h e 15 for 2,6-di-n-propylphenol, indicating t h a t it could be very similar in structure. Both 2,6-di-n-propylphenol and 2VOL. 32,

NO. 4, APRIL 1960

0

471

Table 111. Quantitative Analysis of Tar Acids Boiling at 231 to 331 C. Constituent Wt. %a Constituent w t . %a 3-Ethyl-2-methylphenol 0.05 Ketone (alkylindanone) 0.12 3-Ethyl-Pmethylphenol 0.05 3,4-Dinuclearphenol 0.22 5-Ethyl-2-methylphenol 0.05 Cycloalkenylphenol I? 0.02 3-n-Propylphenol 0.05 0.17 Cycloalkenylphenol II? 4-n-Propylphenol 0.51 2-C yclohexylphenol 0.81 4-Isopropylphenol 0.05 2-Alkylcycloalkylphenol I 0.06 2,3,6-Trimethylphenol 0.05 0.04 2-Blkylcycloalkylphenol I1 3,4,5-Trimethylphenol 0.53 hlethyl-2-, 3-, or 4-fluorenol 0.25 2,3,5,6-Tetramethylphenol 0.17 Total 2.49 2-Methyl-5-isopropylphenol 0.05 3,5-Dialkylphenol(3-methyl-52-Phenylphenol 0.31 n-propylphenol) 1.14 0.11 3-Phenylphenol 2,6-Dialkylphenol (Z-ethyl-64Phenylphenol 1.46 n-propylphenol) 1.03 1.88 Total 3,4-Dialkylphenol(3-ethyl-4-npropylphenol ) 0.47 1-Naphthol 0.80 3,4Dialkylphenol(4-ethyl-3-n6.17 2-Naphthol propylphenol) 0.95 5,6,7,8-Tetrahydro-l-naphthol 0.09 2,4- or 3,Pdialkylphenol (45,6,7,8-Tetrahydro-2-naphthol 0.09 isopropyl-3-n-propylphenol ) 0.54 0.16 PMethyl-1-naphthol 2,4- or 3,4-dialkylphenol (2,4Alkyl-1-naphthol (2-ethyl-ldi-n-propylphenol) 0.53 3.38 naphthol?) 2,4- or 3,4-dialkylphenol? 0.06 Dimethyl-1-naphthol (2,5- or 0.03 2,7-dimethyl- 1-naphthol? ) Total 6.28 Dimethyl-1-naphthol (2,6-diPIndanol 2.10 0.03 methyl-1-naphthol? ) 5-Indanol 1.97 Dimethyl-1-naphthol (5,7-di1-, 2-, or 3-methyl-4-indanol 0.18 methyl-1-naphthol?) ( 3-methyl-4-indanol? ) 0.36 0.06 1-Methvl-2-naahthol 1-, 2-, or 3-methyl-5-indanol Methyl:2-naphthol (3-methyl( l-methyl-5-indanol?) 1.45 0.03 2-naphthol?) Methyl-5-indanol (6-methyl-5Methyl-2-naphthol (4-methylindanol) 0.02 0.43 2-naphthol?) 7-Methyl-5-indanol 0.41 Methyl-2-naphthol (6-methyl0.04 2-naphthol?) Total 6.31 Methvl-2-naahthol (7-methyl0.18 4-Indenol 0.10 0.07 8-Met6yl-2-naphthol? 5-Indenol 1.09 1-, 2-, or 3-methyl-4-indenol 11.74 0.56 Total Methylindenol 0.33 illiphatic carboxylic acids 1-, 2-, or 3-polyalkyl-4-indenol 0.50 0.16 Boiling 238-251 Total 2.58 0.54 Boiling 251-258' 0.09 Boiling 258-260" 2,3- or 2,6-alkylalken-10.4i ylphenol (3-methyl-20.32 propen-1-ylphenol) 0.01 0.10 Alkenylphenol I 0.02 0.09 Alkenylphenol I1 0.004 0.20 Alkenylphenol 111 0.12 0.24 Boiling 325-331' Methyl-1-indanone (Pmethyl2.21 Total 1-indanone? ) 0.60 Methyl-1-indanone (6-methylUnknowns I through XXIV 1-indanone?) 3.48 0.05 Total On basis of 50-gram residue of high-boiling acids (231' to about 360") which constitutes 0.89 weight yoof tar. O

methyl-6-n-propylphenol were excluded as possibilities, as authentic specimens of these compounds were available for comparison. It was concluded, therefore, that constituent 10 could be 2-ethyl-6-n-propylphenol (boiling point 240 "). Constituent 18, a 3,5dialkylphenol, had a partition coefficient of 6.0 a t p H 11.20, calculated from its concentration peak in tube 10FC. The partition coefficient of 3-ethyl-5-methylphenol is 1.1, which would place its peak in tube 55. Therefore, the degree of n-alkylation must be greater than in 3-ethyl-5methylphenol. 3-5-Diethylphenol was excluded, as its infrared spectrum was available. It was concluded that constituent 18 could be 3-methyl-5-npropylphenol (boiling point 254"). Constituent 19, a 3,4-dialkylphenol,

472

ANALYTICAL CHEMISTRY

had a partition coefficient of 15 a t p H 11.20, only slightly loB-er than the 20 for 2,6-di-n-propylphenol. Therefore, the degree of n-alkylation must be only slightly less. Taking into consideration the boiling range of the distillate, constituent 19 could be 4-ethyl-3-n-propylphenol (boiling point 258'). Constituent 23, a 3,4-dialkylphenol, had a partition coefficient of 13 at pH 10.49, whereas the value for 3-methy1-4ethylphenol is 5.0. Because five out of eight major infrared absorption bands are essentially identical to major bands for constituent 19, this suggests that these compounds are closely isomeric, but not identical, as shown by the ultraviolet spectra. Therefore, constituent 23 could be 3-ethyl-4-n-propylphenol (boiling point 263 "). Constituent 35, a 2,4- or 3,4-dialkyl-

phenol. had a partition coefficient of 4.5 at pH 11.86, which indicates the alkyl groups must have structures and locations on the ring that would give only a moderately high partition coefficiente.g., there could not be two long chains close to the OH group. 2,4-Diisopropylphenol (boiling point 248") is too low boiling to be present; however, the presence of 4-isopropyl-3-n-propylphenol (boiling point 262"), is a distinct possibility. Constituent 36, also a 2,4- or 3,4-dialkylphenol, had a very high partition coefficient of 59 a t pH 11.86 requiring two n-alkyl groups with a t least one in the ortho position. 2,4-Di-npropylphenol (boiling point 263") is a likely possibility. Alkenylphenols. Constituent 22, a 2,3- or 2,6-dialkylphenol, had a n ultraviolet absorption band at 296.4 mp, indicative of an unsaturated side chain, probably a n o-propenyl group, conjugated with the benzene ring (2). 2-Propen-1-ylphenol has a partition coefficient of 3.5 a t p H 10.49, which is close to the 3.0 for constituent 22. 2-l\lethyl-6-propen-l-ylphenol (boiling point 243") is too low boiling to be present, but 3-niethyl-Z-propen-lylphenol (boiling point 256") could be present, Constituent 29 has the low partition coefficient of a n unsaturated phenol and an absorption band a t 253.0 mp, in common with 2-cyclopenten-lylphenol (boiling point 272"). Cycloalkylphenols. Constituent 70 has the 13.30-micron absorption band in t h e infrared, indicative of monoortho substitution. T h e very high partition coefficient of 15 at p H 11.85 is of t h e same order of magnitude as t h a t for 2-cyclohexylphenol, constituent 56. T h e higher boiling range for constituent 70 and the bathochromic shift for the ultraviolet absorption band indicate alkylation on the cycloalkyl ring. Constituent 81 is similar to 70 but not identical, as indicated by the slight shift in the mono-ortho band to 13.27 microns and the slight change in peak tube a t p H 11.85 from 4FC to 1FC. Indanones. Constituent 20 had no OH band in its infrared spectrum and had a strong band a t 5.92 microns for t h e C=O group. T h e ultraviolet spectrum was t h a t of a benzene ringsubstituted 1-indanone ( 5 ) . An infrared band a t 12.95 microns for three adjacent hydrogens on the benzene ring indicated that constituent 20 could be either 4-methyl- or 7-methyl-1-indanone. Constituent 21 was very similar to 20, except that there was no evidence for three adjacent hydrogens on the benzene ring, As the partition coefficient of constituent 21 is appreciably greater than that of constituent 20, the methyl substituent must be situated in the latter so as to decrease the extent of keto-enol tautomerism compared with the former.

whereas const'ituent 33 has a good match with 2,4,5trimethylphenol for six bands in the out-of-plane region, indicating 6-methyld-indanol (boiling point 263"). Ortho-alkylated isomers invariably h a r e the highest partition coefficients, a fact that is in agreement with the relative positions of these two constituents. Indenols. Constituent 13 is a meta-substituted phenol, with split bands at 6.27 and 6.29 microns indicative of a conjugated olefinic group. T h e absorption a t 290 and 370 mp indicates a dinuclear system such as in indenes, which absorb in the range 290 to 295 mp. There is a 16-nip increase in the longest wave length band betiTeen indan and 5-indanol. From this it might be expected that the longest wave length band of 5-indenol would be 293 16 = 309 mp, compared with the 307-mp band observed. The longest ITave length band of a 2,3-di-substituted-6-indenol has been given as 305 mp ( 2 2 ) . From all of this evidence it seems certain that const'ituent 13 is 5-indenol (boiling point about 259 "), or its nearly identical isomer, 6-indenol. The partition coefficient of 0.31 a t p H 11.20 is considerably lower than the 0.i9 for 5-indanol, as would be expected with the introduction of an olefinic group. The infrared spectrum of constituent 27 showed splits a t 6.14, 6.17. and 6.28 microns indicative of an olefinic: group conjugated with a benzene ring (9). A band a t 7.25 microns indicated an indene-type structure, and a band a t 14.14 microns indicated a 2:3,-subst'ituted phenol. Thus, this constituent must be a 4-indenol with no additional substitut>ionon the benzene ring; lio~vever, as 4-indenol boils around 250", constituent 27 must hare a methyl group on t'he five-membered ring. The methyl-4indenol structure n-as confirmed by comparing the ultraviolet bands a t 309.6, 299.7, 288.6, 265, 260.7, and 254.5 mp with the bands obtained mith a sample of 4-indenol synthesized from 4-indanol by R. E . Dean of the Coal Tar Research Ahsociation in Leeds, England. This sample consisted of more indanol than indenol, but characteristic bands a t 306, 300, 289.0. 265, 260, and 252.0 mp could be picked out. Constituent 12 was clearly 4-indenol; although the longest wave length band \vas obscured, t'here were three other ultraviolet bands and nine infrared bands t'hat checked well with the synthetic material. -4 synthetic sample of 5-indenol was also available, which likewise consisted of more indanol than indenol. The longest wave length band of this impure sample was 304.4 mp, rather than the 307.0 mp found for constituent 13. This discrepancy may be due to a difference in the ratio between 5-indenol arid 6-indenol, both of which could be produced from 5-indanol. However, two other

+

55

*

5-

4-

rn

I 4

Lz

?

3-

1 1

I

60 I

2-

0

IO

20

30

50

40

45

+35

25

FRACTION

TUBE

Figure 8.

Countercurrent distribution

15 5 C COLLECTOR^

NUMBER

of phenols boiling at 300" to 325" C.

105 transfers; p H 11.85

This suggests that constituent 20 is 4methyl-1-indanone, whereas constituent 21 is 6-methyl-1-indanone. Indanols. Constituent 16 had ultraviolet and infrared spectra t h a t clearly indicated a 5-indanol structure. There Irere infrared bands for out-of-plane deformation vibrations of two adjacent hydrogens, as ne11 as a single hydrogen. This indicated t h a t t h e indanol was substituted in the saturated ring, and probably t h e substituent would be a single methyl group to place this compound in the same distillate fraction in which 5-indanol (constituent 5 ) is the major component. Similarly. it n-as indicated that constituent 17 was a 4-indanol with a single methyl group in the saturated ring. Whereas the partition coefficients of 4- and 5-indanol a t p H 11.20 are

0.68 and 0.79, respectively, the values for the methyl-4- and -5-indanol are 3.8 and 3.3, respectively-that is, they are reversed with respect to the parent compounds. Seemingly, this mould require the methyl group in the l-position of the 5-indanol for the minimum increase of partition coefficient and the methyl group in the 3-position of the 4-indanol for the maximum increase. Constituents 32 and 33 are both 5indanols. I n the infrared bands for out-of-plane hydrogen deformation there are good matches between 4-indanol and 2,3-xylenol and between 5-indanol and 3,4-xylenol. Constituent 32, 7 methyl-5-indanol, identified by the ultraviolet and infrared spectra of an authentic specimen, has a good match with 3,4,5-trimethylphenol for Seven bands in the out-of-plane region

VOL. 32, NO. 4, APRIL 1960

473

ultraviolet bands and nine infrared bands checked well with the synthetic material. The relative positions of 4-indenol (constituent 12) and 5-indenol (constituent 13) are the same as those of 4-indanol and 5-indanol. The sample of 4-indenol had long wave length bands completely isolated from the rest of the spectrum, which allowed determination of the partition coefficient, even though the sample had more indanol than indenol. Using this value gave a good check between calculated and experimental peak tube results. Naphthols. The spectra of relatively few of the alkylated 1- and 2naphthols were available for comparison with those of samples isolated by countercurrent distribution. However, the ultraviolet spectra of a large number of alkylated naphthalenes were available, and the longest wave length absorption bands of the alkylated naphthols could be estimated closely from the spectra of the corresponding naphthalenes. For example, the value of the band for 3-methyl-2-naphthol can be estimated by adding to the value for 2-naphthol the value of the bathochromic increment between 2-methylnaphthalene and 2,3-dimethylnaphtlialene. Of course, these estimated values could not be used for positive proof of identity of individual isomers. The naphthols were all readily identified as either 1- or 2-naphthols from the highly characteristic shapes and relative intensities of the ultraviolet absorption bands, which were altered very little by the presence of alkyl groups. The degree of bathochromic shift of these bands from the wave length values for the parent compounds was an immediate indication of either monomethyl or dimethyl substitution. The hydrogen out-of-plane deformation vibration region around 11 to 15 microns was useful in determining possible arrangements of hydrogen atoms and, hence, of methyl groups. Constituent 55, a dimethyl (or ethyl)1-naphthol, had a strong infrared band a t 12.52 microns indicative of two adjacent hydrogens and suggesting 2-alkyl substitution as in 2-methyl-1-naphthol, 12.50 microns. This constituent almost certainly had ortho substitution to have a partition coefficient as high as 30 a t p H 9.94. It is not 2-methyl-1-naphthol itself, because its longest wave length band is only 324.4 mp, determined with an authentic specimen, compared with 328.2 mp for constituent 55. Strong infrared bands a t 14.53 and 13.48 microns indicate four adjacent hydrogens as in 2-methyl-1-naphthol, 14.55 microns, and 4-methyl-1-naphthol (constituent 52), 13.50 microns. Thus, it appears that constituent 55 must be a 2-alkyl-lnaphthol and could be 2-ethyl-I-naphthol.

474

ANALYTICAL CHEMISTRY

0 lD r J

N

3 0-

2 5-

2 0W 0

z a

m I5LL

0

m m a IO-

5-

0-

1 0

IO

20

30

50

40

40

30

1 20

IO

0

LFRAC COLLECTOR^ TION -UBE

Figure

NUMBER

9. Countercurrent distribution of phenols boiling a t 325" to 331" C. 100 transfers; pH 1 1.85

There were six distinct countercurrent distribution peaks that were readily determined as due to monomethyl-2naphthols. The 5-isomer was immediately excluded on the basis of the available infrared spectrum of this compound, leaving equal numbers of constituents and isomers. Constituent 54 was positively identified as l-methyl-2naphthol by the available infrared spectrum of this compound. Constituent 53, which immediately precedes 1methyl-2-naphthol in the countercurrent distribution, must also have ortho-substitution to have such a large partition coeficient. Thus, constituent

53 appears to be 3-methyl-2-naphthol (boiling point 302'). Constituent 50 had a strong band a t 13.40 microns indicating four adjacent hydrogens. The low partition coefficient of 2.5 a t p H 9.94 confirmed the exclusion of the 1- and 3-methyl isomers, already proposed for other constituents. Therefore, constituent 50 appeared to be 4-methyl-2-naphthoL Constituent 78 had strong bands a t 14.30 and 13.16 microns, indicative of three adjacent hydrogens. As the 5-isomer was excluded, it appeared that constituent 78 was 8-methyl-2-naphthol (boiling point 320 ') .

Fluorenols.

Constituent 74 has ultraviolet absorption bands very similar t o those for 8-methyl-2-fluorenol (9). However, the longest wave length band in the region around 31 1 mp is held in common b y other methyl-2fluorenols and methyl-3(and 4)-fluorenols. The longest wave length bands for methyl-1-fluorenols are a t wave lengths a t least 10 mp lower. Thus, constituent 74 is a methyl-2-, -3-. or -4fluorenol. ACKNOWLEDGMENT

Special thanks are due Joseph R. Comberiati for perforniing the spinning band distillation and Edward E. Childers for assistance in performing the

countercurrent distribution fractionations. REFERENCES

(1) Anchel, M., Blatt, A. H., J . Am. Chem.

SOC. 63, 1918-52 (1941). (2) Bader, A. R., Ibid., 78, 1709-13 (1956). (3) Elsner, B. B., Parker, K. J., J . Chem. SOC. 1957, 592-600. (4) Friedel, R. A,, J . Am. Chem. SOC.73, 2881-84 (1951). (5) .Friedel, R. A., Orchin, M., “Ultrs;; violet Spectra of Aromatic Compounds, Wiley, New York, 1951; spectra 47, 51, 78, 79. (6) Golumbic, C., Woolfolk, E. O., Friedel, R. A., Orchin, M., J. Am. Chem. SOC. 72,1939-42 (1950). (7) Irvine, L., Mitchell, T. J., J . A p p l . Chem. 8, 425-32 (1958). (8) Karr, C., Jr., Brown, P. M., Estep,

P. A., Humphrey, G. L., Fuel 37,227-35 -

I

(1958). (9) Morrison. A., Mulholland, T. P. C., J . Chem. SOC.1958,2702-5. (10) Post, H. O., Scientific Instrument Co., Inc., 6822 60th Road, Maspeth 78. N,Y.. ‘LPriceLists & Prints, Glass Countercurrent Distribution Instruments,” 1955, et. seq. ( 1 1 ) Sadtler, S. P. & Son, Inc., “The Sadtler Standard Spectra,” Sadtler Research Laboratories, Philadelphia, 1958. (12) Silverman, M., Bogert, M. T., J. Org. Chem. 11, 34-9 (1946). (13) Williamson, B., Craig, L. C., J . Bid. Chem. 168, 687-97 (1947). (14) .Woolfolk, E. O., Golumbic, C., Friedel, R. A,, Orchin, M., Storch H. H., U. S. Bur. Mines Bull. 487 (1950).

RECEIVEDfor review October 13, 1959. Accepted January 8, 1960.

Cocrystallization of Ultramicro Quantities of Alkaline Earth Elements with Potassium Rhodizonate Determination of Radiobarium in Sea Water HERBERT V. WEISS and MlNG G. LA1 U. S. Naval Radiological Defense laboratory, Sun Francisco, Calif. The feasibility of the cocrystallization of ultramicro quantities of elements from solution with organic reagents was studied with the alkaline earth-potassium rhodizonate system. Results indicated that crystallized rhodizonate was enriched with t h e alkaline earth microcomponent and that the distribution approximated the Doerner-Hoskinstype of equilibrium. These elements were quantitatively recovered with ease at concentrations of 1G - I 4 to 1 0-I6M. The cocrystallization process was applied to the determination of radiobarium in sea water. The cocrystallization step separated barium from sulfate ion, which interferes with its determination in radiochemical analysis by the carrier technique. The crystallized material was readily incorporated into a conventional radiochemical procedure, and reliable results were obtained.

T

DIRECT PRECIPITATION of a n element from solution is precluded when the quantity of substance t o be separated is less than the amount required to exceed the solubility product. Under such circumstances a nonisotopic carrier map be used; separation by this mechanism is known as coprecipitation. Details of the coprecipitation process HE

and the method of determining the Doerner-Hoskins logarithmic distribution coefficient (A) and the homogeneous distribution coefficient (D)are comprehensively reviewed by Hahn (4). Nearly all the coprecipitation reactions described involve the use of inorganic carriers (8). However, Kuznetsov (6) pointed t o the advantages of the coprecipitation of elements with organic reagents and reported a number of schemes in which minute quantities of a variety of elements were quantitatively recovered rrith such reagents. The method requires binding the trace elcment Rith ttn organic reagent and then precipitating the added agent M ith still another organic compound. If the compound formed upon the addition of the first agent is slightly soluble or only slightly ionized, then the element is usually carried efficiently. Kuznctsov emphasized the requirement for large molecular weight compounds to effect efficient coprecipitation and used combinations of organic reagents such as methyl violet and a n azo compound or tannins . The current study explores the possibility of isolating trace quantities of a n element by the direct crystallization of a n organic reagent from solution. Such a crystallization process simplifies the operating procedure. Furthermore, a

wide variety of reagents is available since the conditions required to crystallize organic reagents are usually easily achieved. If the trace element and the organic reagent combine to form a compound whose insolubility is greater t h a n that of the reagent, then according to Fajans’ rule ( I ) , the crj stallized reagent should be enriched with trace element. T o examine this premise as well as the feasibility of isolation of ultramicro quantities of an element by this process, the alkaline earth-potassium rhodizonate system was studied. This system was selected because the varying solubilities of the alkaline earth rhodizonates provide a basis for verification of the applicability of Fajans’ rule. The solubilities of the alkaline earth rhodizonates decrease with increasing molecular weight for calcium, strontium, and barium (2, 9). Although the radium rhodizonate salt has not been described, Iresumably it forms the most insoluble complex of this chemical group. The cocrystallization of elements of the alkaline earth group with rhodizonate upon crystallization of fractions of carrier with ammonium chloride or alcohol was studied. To measure ultramicro quantities of alkaline earth elements accurately, appropriate radioactive tracers were used. The informaVOL. 32, NO. 4, APRIL 1960

475