INDUSTRIAL AND ENGINEERING CHEMISTRY
2114
TABLE Iv.
Vol. 42, No. 10
2,2-BIS(4-HYDROXYPHENYL)PROPANEMODIFIED p-tert-BUTYLPHENOL (Moles CHaO per mole phenol, 1.6; basic catalyst, 0.6% NaOH)
RESINS
Tests on Varnishb
Resin
No.
Moles 2,2-bis(4Moles Hydroxyp-tertphenyl)Butylphenol propane
p H after Ha804 Treatment 8.0 6.4 5.5 5.0 4.6
19 7/s vt 1/1 '/t 20 '/8 '/* 21 22 VI VI 23 24 "/16 1/18 5.5 Commercial heat reactive oil-soluble phenolic resin BR-10282 (Bakelite Corp.) 4 C = compstible; I = incompatible. b T 'ang oil base; K = maleiniaedlinseed oil base.
:A
Tung oil
I
C C C
Compatibility4 Maleiniaed Y-bodied linseed linqeed oil 011
I C
C C
C C C C
C
C
Bodying time, %inch Gel string time a t 150' C., a t 135'C., min. hr.
...
I I I
C
C I C
.a/. .
12T 27K 25T 25T
1l/r 11/4 1' / t
27K
11/4 13/4
K
Viscosity 50% solids
1
... J E
... 42
... 30, swelied '
45 46 16
30, least swelling 35, least swelling 4 , swelled, soft
37 30
I
M
'/4
Cake Hardness Shore A Durometer Baked 5 hr. a t Baked 135' C. plus 48 6 hr. a t hr. a t l l O o C. in 135O C. transformer 011
E K
33, little swelling 15
-
TABLEV. A.S.T.M. D-115 TESTSON ELECTRICAL INSULATING V.4RNISHES Composition Example 21 (rnaleinieed linseed ,oil) Example 16 (Y-bodied linseed oil) plus medium oil length linseed alkyd
Drying Time a t l l O o C., Hr. 1
1.5
In conclusion it must be stated that the most interesting question is still unanswered-namely, what intermediate stages of the phenolic condensation and polymerisation process are produced aa a function of pH to influence this wide variation in properties. As yet quantitative analytical tools are either lacking or unproved. Thus far the author has been unable to corroborate methylol determinations by the method of Lilley and Osmond. It is hoped that either conventional analytical methods, such m those of Lilley and Osmond, or infrared spectra will ultimately provide the answer.
Heat Endurance a t 150° C., Hr. 18-20
Oil Proofneas Passes
84-96
Passes
Dielectric Strength Dry, Wet (24 hr.) volts/rnil 2430 1740
I
volts/mil 1281
950
LITERATURE CITED
(3) Hultssch, K.,J . prakt. Chem., 158,No.2,275(1941). (4) Lilley, H. S., Varnish Making, Oil & Colour Chem. Assn., 11321, Chem. Publishing Co.,Ino., New York (1940). ( 5 ) Lilley, s., and Osmo,,d, D. 'w. J,, J . sot, Chem. I d , (London), 66,425-7 (1947). (6) Lilley, H. S., and Osmond, D. 15'. J., Paint Technol., 13, 21724 (1948). (7) Lykken, Porter, Ruliffson, and Tuemmler, IND.ENG.CHEW, ANAL.ED., 16,219-34 (1944). (8)Singer, R. J. R., Kemisk, 23,49-61 (1942): (9) Turkington, V. H.,and Allen, I., presented before the Division of Paint, Varnish, and Plastics Chemistry at the lOlst Meeting, AMERICANCHEMICAL SOCIETY, St. Louis, Mo. (10) Turkington, V. H., Shuey, R. C., and Shechter, L., IND. ENQ. CHEM.,30,984 (1938).
w.,and Perrins, L., J . Oil &? CO~OUT Chemists' A8SOC., 30, No.324, 185 (1947). (2) Farmer, E.H., J . Chem. Soc., 1943,472.
RECEIVED October 8, 1949. Presented before the Division of Paint, Varnish, and Plastics Chemistry a t the 116th Meeting, AMXXWXN CHEMICAL SOCIETY, Atlsntic City, N. J.
(1) Charlton,
Catalytic Synthesis of Benzofurans J
J
CHROMIUM CATALYSTS FOR CY CLODEHYDROGENATION OF 0-ALKYLPHENOLS CORWIN HANSCH, CARLETON SCOTT1, AND HOWARD KELLER2 Pomona College, Cluremont, CaZg. This paper discusses the vapor phase catalytic dehydrocyclization of o-ethylphenol, o-isopropylphenol, 0allylphenol, and thymol to benzofuran, 3-methylbenzo-
furan, 2-methylbenzofuran, and 3,6-dimethylbenzofuran, respectively. The investigation of several chromium-oncharcoal catalysts for this reaction is reported.
I
conditions for a given reaction such as the conversion of heptane to toluene. With the exception of the interesting work being done by Orchin et al. ( 6 )little has been done with the more complicated molecules. The work reported here was directed primarily toward increasing the scope of the reaction, particularly with respect to heterocyclic molecules. The need for more suitable procedures for the industrial preparation of this increasingly more important class of compounds was one of the motivating factors in this research. This work was of a scouting nature; the expensive development of optimum conditions will be left to larger better equipped laboratories.
N LINE with increasing interest in the catalytic synthesis of aromatic molecules by dehydrocyclisation, a program has been undertaken in this laboratory to investigate the vapor-phase catalytic synthesis of some of the larger aromatic molecules, especially the heterocycles. Previous interest in the dehydrocyclization reaction has centered mostly on the conversion of relatively simple aliphatic hydrocarbons to simple aromatics. This work has been devoted to gaining an insight into the mechanism of the reaction ( 4 ) or toward improvement of processing 1 Present addrers, XIascnchusetts Institute of Technology, Cambridge. Masr. 2 Preqent address. Leffinqwell Company, Whittier. Calif.
INDUSTRIAL AND ENGINEERING CHEMISTRY
Odober lQS0
o-Alkylthiophenols (8), o-alkylphenols (a), and o-alkylanilines (1 ) have been dehydrocyclized to thianaphthenes, benzofurans, and indoles respectively, according to the following equation:
@7:+@ where X = S, 0, NH. I n the first work with alkylphenols ( 9 ) two compounds were investigated-ethylphenol and o-allylphenol. Cyclization of these to the corresponding benzofurans was effected a t temperatures of 550' to 625' C. using platinum and palladium catalysts. This paper discusses the use of chromium catalysts for this reaction and the extension of the reaction to two new phenolsthymol and o-ieopyopylphenol. The conversion of thymol to 3,&diethylbenzofuran according to the following equation was selected as the reaction for investigation of the catalysts. CHs
-+ C a
CHs
This alkylphenol waa chosen because it is easily obtained in pure form from commercial sources and because preliminary work indicated that the isopropyl group was more easily removed by pyrolysis than the ethyl group. Also it was of interest to see whether the methyl group attached to the benzene ring would be affected or not. CATALYST PREPARATION
The activated charcoal used in this work was Type BP-6x8, No. EZ-71-A, obtained from the Pittsburgh Coke and Chemical Company. 1. Chromiumon charcoal: In 150ml. of boilinqwater9.1 grams of chromic anhydride were dissolved, and to this solution were added, with stirring, 50 grams of activated charcoal. After thorough mixing, the excess was evaporated and the catalyst dried at loo'
grees above the temperature a t which the reaction was to be run. The liquid products were separated by means of a condenser and dissolved in about an equal volume of petroleum ether (boiling point 64' to 110' C.) The gaseous products were paeaed through a webtest meter to measure their volume. The phenolic producte were extracted three times with 10% sodium hydroxide, first with 100 ml., then with 70 ml., and finally with 50 ml. The combined sodium hydroxide solutions were washed with about 20 ml. of petroleum ether and the two ether portions combined. This solution waa dried over potassium carbonate or sodium sulfate and distilled. The sodium hydroxide extract was acidified with concentrated hydrochloric acid and extracted three times with petroleum ether. The combined extracts were washed once with water and then dried over sodium sulfate. These extracts were then distilled through a 20-plate spinning-band fractionating column. All boiling points given are corrected. ~ACETYL-&METHYLPHENOXYACETIC ACID. T h e 2-acetyl-5methylphenol used was prepared in 66% yield, b.p./4 mm. 100102', by the method of Rosenmund and Schnurr (7). The phenol (26.0 grams), 23.6 grams of chloroacetic acid, and 15.0 grams of sodium hydroxide were dissolved in the minimum amount of 50:50 alcohol and water. This mixture waa heated on a boiling water bath for 5 hours; a t the end of each hour 11.2 grams of chloroacetic acid and 10.0 grams of sodium hydroxide were added, with just enough alcohol and water to keep salts from separating from solution. Four additions were made. The mixture wag then cooled to room temperature, acidified to Congo-red, and extracted with ether. The ether was extracted with sodium bicarbonate solution, and the phenoxyacetic acid waa precipitated from the bicarbonate with hydrochloric acid. It waa recrystalliaed from boiling water. Yield: 17.5 grams, 48.5%, melting psint 140-141' C. Koelsch (6) reports the melting point as 142' to 144' C. 3,6-DIMETHYLBENZOFURAN. T W O grams Of anhydrous sodium acetate and 10 grams of the phenoxyacetic acid were placed in 9.8 grams of acetic anhydride. The mixture was slowly heated in an oil bath to the boiling point. A considerable amount of bubbling occurred owing to the evolution of carbon dioxide. The mixture was refluxed for 2 hours, then cooled and made alkaline with excess sodium hydroxide. This basic solution was steam distilled, and the distillate was extracted with ether. After drying over anhydrous sodium sulfate, i t waa distilled: boiling point (730 mm.) 216.5' C.; yield 6.3 grama, n# 1.5470. The physical constants of the 3,6-dimethylbenzofuran made catalytically and from the phenoxyacetic acid were identical: boiling point (730 mm.) 216.5' C.; n# 1.5470. The picrate melted at 75' to 76' C. Stoermer (9) reports the boiling point as 222' C.; &?1.5505; picrate melting point 76' C. One run was made with 61 grams of +allylphenol a t 550' C. (run waa started at 525' C. for the first 10 minutes) using catalyst 3 and space velocity 1750. The space velocity (S.V.) was calculated as milliliters of vapor at standard temperature and pressure per milliliter of catalyst per hour. The conversion to 2methylbenzofuran was 27.2% compared with 26.5% under the same conditions, using a palladium catalyst. In all the runs reported in this research no appreciable charring occurred in the catalyst tube. The condensate was usually lightcyellow in color. The catalyst gained 1 to 1.5 grams in weight; part of this waa liquid which remained in the catalyst tube. During the latter part of this research, equipment was available for gas analysis and such analyses were made with a Fisher gas analyzer of the Orsat type. The unsaturates were determined by absorption in fuming sulfuric acid, and the hydrogen was determined by converting it into water over copper oxide. The remainder of the gas sample was considered to be saturated hydrocarbons.
,o,rcHa + 2H1
c.
2. Chromium-copper on charcoal: In just sufficient water to effect solution, 11 a m of chromic nitrate and 1.2 grams of Cuand the mixture was heated to boil(NOa)2.3H20were ing. To this solution were added, with stirring, 25 grama of activated charcoal. Boiling was continued for 5 minutes, then the eolution waa filtered, and the catalyst was dried at 100' C. 3. Chromium-copper on charcoal: Activated charcoal (22.9 grama) waa treated as above with 36.1 grams of Cr(NO&.QHnO a d 0.88 of Cu(NOa)r.BH&. 4. Chromium-copper on charcoal: Activated charcoal (25 grama) waa treated with 2.0 grama of chromium nitrate and 0.5 of copper nitrate in the same fashion as described for cabE 2 . 5. Chromium-zirconium on charcoal: Charcoal (25 grama) was treated with 2.1 grama of chromium nitrate and 0.25 ram of ZrO(NOa)n.2Hn0in the same manner as described for catayyat 2.
%solved,
The silica used in this work was obtained from Eimer and Amend, Type %155,6-16 mesh. 6. Chromium-copper on silica-gel: In 50 ml. of boilin water 36.1 grams of chronuum nitrate and 0.88 gram of Cu(N0$.3H?O were dissolved, and to thia solution were added 22.9 g r a m of si]ica gel; the nuxture was boiled for 3 minutes. The catalyst waa removed by filtration and dried at 110' C. PROCESSING PROCEDURE
The apparatus used in this work is similar to that previously reported ( 4 ) . I n all runs, 5 ml. of the catalyst were placed in the catalyst tube, supported and covered by quartz chips. The aatalysta were reduced in a slow stream of hydrogen first for 45 minutes at 250' C., and then for 45 minutes at a few de-
2115
~
2116
Vol. 42, No. 10
INDUSTRIAL AND ENGINEERING CHEMISTRY
TABLE I. DEHYDROCYCLIZATION OF THYMOL TO 3,6-DIMETAYLBENZOFURAN hun No.
Thymol Uaed, Grams'
28 86 67 69 74 75 80
DMB,
%
Phenolsa,
%
Thymol Recoveredb, %
Gas Evolved, MI. 1st 10 Last min. 5 min. of run of run Total
Catalynt
S.V.
39.5 47.7 48.0 63.8 47.1 50.0 53.0
Pd
3"
885 1400 1380 735 1410 1500 1320
4.7 4.8 13.8 13.8 5.7 6.9 16.0
23.0 8.0 13.7 12.8 17.8 21.4 8.5
19.8 33.2 29.4 31.2 26.3 31.8 33.8
370 860 710 1020 340 380 500
290 230 390 370 280 290 280
5240 4000 5550 6480 3340 5250 5830
81
58.0
3
1350
17.5
13.1
37.6
470
310
6120
92 10
53i7 65.7
3 3
1340 1280
10.7 14.4
10.0 5.8
38.6 49.5
750 520
230 240
5600 5230
12
43.3
6
1220
.,
..
300
150
2480
1 2 2 5
*.
Gas Analyses Sample taken, min.
..
Unsatd., %
..
.. ..
HrC,
%
Satd., %
, .
..
..
..
..
.. .. ..
5 42 8 40 57
5 15 7 13 16
713 59.0 75.0 53.0 38.0
87:7 28.0 18.0 34.0 48.0
10 40
10 17
'O 19 42
36 l7 41
50:O 35.0 31.0 24.0 14.0
40:O 48.0 52.0 40.0 45.0
.
I
.. .. ..
..
..
Weight ? ofI o r i g i ~ material; l this fraction 197-215' C.) oonsiated of m-cresol, identified by its urethane, and 3,8-dimethylphenol, identified by ita benzoate aerivative. b This fraction (b.p. 225-230' C.) oonsisted of thymol and probably some 2-propenyl m-cresol C Pure hydrogen, after passing over catalyst 3 at 550' C., 8.V. 1000, contained about 3% saturaded hydrocarbons. a
TABLE 11. DEHYDROCYCLIZATION OF 0-ISOPROPYLPHENOL AND 0-ETHYLPHENOL Propylphenol Used Gra&
T:mz.,
86
90
22.8 48.8
4
46.6
Run No.
Phenolb, %
S.V.
MBO, %
640 575
1810 1540
12.2 18.6
9:9
600
1540
8.4
20.5
Gas Evolved, MI. Last min. 5 min. of run of run Total 1st 10
Gas Analyses Sample taken, min.
Unsatd., %
HI. %
Batd., %
36:O 66.0 28.0
8i:O 10.0 54.0
0 85.0 72.0 0 lo 16.0 10 23.0 1.5529, b. p. 191-193O C.; Steerrner (8)reports n*$ 1.5535 and b. p. 193-194O C.
35.0 13.0 84.0 62.0
35.0 31.4
510 870
490 370
2760 8380
43.7
800
360
6330
6 45 10
..
1 24 18
Ethylphenol1 Recovered,
Ethyi-
%3' Grams
I8opropylphenol0 Recovered, %
Beneofurand, %
%
87
38.5
800
2020
10.5
15.94
17.7
602
220
4040
2
44.9
625
1830
10.8
17.6
35.7
1110
290
8000
nv
8 40 10 45
MB = 8-Methylbenzofuran: refractive index was This fraction b.p. 170-19S0 C.) was phenol with some iaopropylphenol. This fraction tb.p. 195-214O C.) contained some unsaturated phenol also. d Benzofuran b.p. 170-1'11" C. I B. P. 198-201" C. e Phenol with some ethylphenol. b
e
DISCUSSION
THYMOL TO 3,6-DIMETHYLBENZOFURAN. The results on this reaction are summarized in Table I. Attempts to dehydrocyclize thymol with palladium-on-charcoal catalyst, which was effectivein converting o-allylphenol to 2-methylbenzofuran, were not successful. This might be expected since palladium is not as effective as chromium in the dehydrogenation of ethylbenzene. Thus one would not expect the isopropyl group on thymol to be readily dehyd-ogenated to a propenyl group which could then cyclize by addition of the hydroxyl group across the double bond. It seems likely, in accordance with previous theories of dehydrocyclization ( 4 ) , that the cyclization occurs by way of the intermediate olefin as follows:
0 -
-
CHI
CHs
\ -d---CHa
\ -(!X€-CHI
/ /--OH CHs
CHs
O - ~ C H ~ --t
/\ O \' CHa
OJ-CH~
/\ ' 0 CHs
Thus with palladium, conversions of only 3 to 5% were obtained, whereas with chromium-copper catalysts under the same conditions yields of 17% were obtained. All the runs were made at 550' C. except 92 and 10, which, were made at 575' and 525' C., respectively, since early runs indicated this to be themost suitable temperature. Higher temperatures increase the initial
rate of reaction but the activity of the catalyst decreases much more rapidly. At 525' C. the conversion was not quite as good, but a much better recovery of unreacted thymol was obtained. As shown in the gas analysis for run 80 the decomposition aa indicated by oiefins and saturates waa high at the start of the run and then dropped during the course of the run. In runs 81, 92, 10, and 12 the temperature for the first 10 minutes of the run was kept 50' below the temperature at which the run waa to be made. It was hoped that this would condition the catalyst in its cracking action. This seems to be true to a ceriain extent, In run 81 a much higher initial rate of hydrogen evolution was observed and a slightly better yield of 3,6-dimethylbenzofuran was obtained. The boiling point of this product waa 215' to 217' C. (730 mm.). In runs 81 and 10, where three analyses were made, the percentage of hydrocarbons formed by cracking increases with the age of the catalyst. The most effective catalyst used was 3, chromium promoted with a small amount of copper. Larger amounts of copper, aa used in catalyst 2, decreased the efficiency somewhat. Catalyst 5, containing zirconium, was much less effective than the copperpromoted chromium. Run 75 with a silica-supported chromiumcopper catalyst shows that silica is not as effective as charcoal. Earlier work (3) revealed that alumina-supported catalysts caused great decomposition of the phenols. o-ISOPROPYLPHENOL TO 3-METHYLBENZOFURAN. The resulttj obtained in the dehydrocyclization of o-isopropylphenol are summarized in Table 11. In runs 90 and 87 the temperature for the first 10 minutes of the run was kept 50' below the temperature a t which the run was to be made. All runs reported in this table were made with catalyst 3. The conversions at the various temperatures show that the effect of temperature is critical,
Ootober 1950
I N D U S T R I A L A N D E N G I N E E R I N G CHEMISTRY
and further study might reveal a more optimum temperature for the reaction. The isopropylphenol waa more stable to cracking action than was the thymol. The results for this reo-ETHYLPHENOL TO B~NZOFURAN. action are summarized in Table 11. A run, not recorded in the table, made at 550' C. and space velocity of 1970 yielded no benzofuran. I n comparing the ease of dehydrocyclization of the various phenols, the most easily cyclized is the unsaturated o-allylphenol. A run with the o-propenylphenol, With the double bond next to the benzene ring, gave essentially the same results. Since these compounds are unsaturated to start with, they cannot be compared directly with the saturated phenols. Of these, thymol was the most easily dehydrocyclired, then isopropylphenol, and finally ethylphenol. From this limited group, it might seem that electron-releasing groups promote this reaction. Other phenols are being investigated with respect to this point.
2117
ACKNOWLEDGMENT
This work was supported in part by a Frederick Gardner Cottrell grant-in-aid from the Research Corporation, 406 Lexington Ave., New York, N. Y.,and in part under Oflice of Naval Research contract NR45B149. LITERATURE CITED
(1) Gresham and Bruner, U. 8.Patent 2,409,676(Oat. 22, 1946). (2) Hansoh and Blondon, J . Am. C h .Sa.,70,1661 (lSaS), (3) Hansoh, Saltonetall,and Settle, Zbid., 71,943(1949). (4) Hoog, Verheus, and Zuiderweg, Trans. Farudau Soc., 35, 996 (1939). (6)Koelsoh,J . Am. C h .Soc., 67,672(1948). (6)Orohin, Reggel, and Friedel, Zbid., 71,2743 (1949). (7) Rosenmund and Schnurr, Ann., 460,66(1928). (8) Stoermer,Zbid., 312, 274 (1900). (9) Ibid., p. 290. RECEIVED Ootober la, 1949.
Low Temperature Calcination Rates of Limestone ARCHIE WAKEFIELD, JR.',
AND MACK TYNER
University of Florida, Gainasville, Fla. T h e calcination of partioles of natural limestone from 0.02 to 0.08 inch in diameter at temperatures between 1400' and 1700' F. has been investigated. The effects on cate of calcination of partiole size, temperature, and rarbon dioxide concentration in the furnace atmosphere were measured. An equation which represents the data was developed empirically. The data indicate that, in the range of conditions studied, rate of calcination is controlled by diffusion rate of carbon dioxide from particle.
T
L
H E calcination of limestone to form "quick" or unslaked lime is a very simple process consisting of heating the limestone to a temperature 8bove 1700' F. and holding i t a t that temperature until calcination is complete. Unfortunately, however, no general statement may be made about the time required to burn a given sample of limestone a t a given temperature. Few investigations have been made into this problem in this country, but much work has been done in other countries (usually on pure calcium carbonate); however, little agreement can be noted among the various investigators ( 1 , B , 7 , 9 , ID, 14, 16, 17). Most of these workers attempted to apply physical-chemical concepts to the reaction, and have determined reaction orders which vary from zero to about two. Also, some investigators (8,11) have noted that limestone, as is to be expected, is only qualitatively similar to pure calcium carbonate in the calcination reaction. It seems unlikely that the concept of reaction order has any bearing on the case, because the rate at which limestone particles calcine is mote likely to depend on the rate of some physical process, such as heat or material transfer, than on the velocity of the purely chemical portions of the process. Even fewer investigations have been made into this phase of the problem than into the chemical aspects ( 3 , 4 ) . Furnaa' (9)work indicates that heat transfer plays an unimportant role in the low temperature calcination of particles of limestone a few centimeters in 1
Preaant addreas. Wakefield Optical Company Charlotte, N. c
diameter. He heated pieces of limestone in a furnace for varying lengths of time and then cut them open to determine, if possible, the reaction mechanism, It was clear that the calcination proceeded along a concentric zone which advanced into the stone. Furnas was able to derive an equation for the rate of advance of this zone at any temperature, and he found this rate to be independent of particle size and external carbon dioxide concentration-Le., per cent carbon dioxide in the furnace atmosphere. Further, he found that at furnace temperatures below about 1700' F. the center of the body rose to the temperature of the furnace long before calcination was complete, while at furnace temperatures above 1700" F. the center temperature rose to a value of 1700' F. regardless of the furnace temperature and remained at that value until calcination was complete. In this second case, the center of the particle was in an un. stable condition but could not decompose until the zone of calcination penetrated to it. Because no heat went past the calcining zone to raise the center temperature, the calcining zone must have been utilizing heat as fast as it could be supplied; thus, the rate of penetration of this zone depends on the rate of heat transfer to this point. When furnace temperatures were below 1700' F., there was an excess of heat being supplied to the body above and beyond that which could be used for calcination because the internal temperature was the same as that of the furnace; consequently, heat transfer was not the deciding factor in these low temperature calcinations. I n the present investigation, since the particles used were much smaller than those used by Furnaa and a slightly lower temperature range was used, i t was thought that Furnas' data might not apply. However, it seemed entirely reasonable to assume that small particles would calcine in the same manner larger ones-i.e., along a concentric zone. If the rate of calcination depends on the rate of diffusion of the carbon dioxide released from the particle, the logical approach toward correlation of the experimental data would be to express this rate as a function of the partial pressure driving forceLe., the difference between the equilibrium disaociation pressure of carbon dioxide above limestone, and the external, or atmoepheric, partial prmure of carbon dioxide.
