April 1949
INDUSTRIAL AND ENGINEERING CHEMISTRY LITERATURE CITED
(1) Davis, “Chemistry of Powder and Explosives,” p. 392, New
York, John Wiley & Sons, 1943. Davis and Abrams, Prpc. Am. Acad. A& Sci., 61, 437 (1926). Davis and Rosenquist, J . Am. Chem. Soc., 59, 2114 (1937). Goldherg, B.S. thesis, Polytechnic Institute of Brooklyn, 1938. Interstate Commerce Conimission Legulations 4, “Transportation of Explosives and Other Dangerous Articles by Freight,” Sections 52 and 62(m), pp. 62, 69, 1941. (6) Pellizzari, Atti. accad. Lincei, 30, I, 171 (1921): Gam. chim. i t d , 51, 140 (1921).
(2) (3) (4) (5)
849
(7) Robertson, M .S. thesis, Polytechnic Institute of Brooklyn 1935. (8) Sahetta, M.S. thesis, Polytechnic Institute of Brooklyn, 1935. (9) Sabetta, Himmelfarb, and Smith, J . Am. Chem. Soc., 57, 2478 , (1935). (10) Schmookler, B.S. thesis, Polytechnic Institute of Brooklyn, 1932. (11) Thiele, Ann., 273, 133 (1893). (12) Weaver, B.S. thesis, Polytechnic Institute of Brooklyn, 1934. RECEIVED April 5, 1948. Published with the approval of the Chief of the Bureau of Ordnance, U. S. Navy. The opinions expressed are those of the authors and do not necessarily represent the official opinions of the U. S.
Navy.
Reactions of tert-Butyl Hypochlorite
with Vegetable Oils and Derivatives SOYBEAN OIL H. M. TEETER, R. C. BACHMANN, E. W. BELL, AND J. C. COWAN Northern Regional Research Laboratory, T h e compound tert-butyl hypochlorite has been investigated as a chlorinating agent for soybean oil and methyl esters of soybean fat acids. The reaction is cxothermic, and the products contain 18 to 24% of conjugation, principally of the diene and triene types. Thermal dehydrochlorination of the chlorinated oils at 150’ C. permits removal of about 50% of the halogen originally prcsent. The products contain increased amounts of triene and tetraene conjugation. Prolonged dehydrohalogcnation at higher temperatures removes as much as 75.5% of combined halogen and gives viscous products containing conjugation which is mainly diene. A satisfactory method for continuous dehydrochlorination is described.
I
N 1923 tert-butyl hypochlorite was fist described by Chattaway and Backeberg (4) who were continuing studies of alkyl hypochlorites initiated by Sandmeyer (16, 17) in 1885. Subsequently, Harford (8-11), Irwin and Hennion ( l a ) ,and Emling, Vogt, and Hennion ( 6 ) showed that halohydrins and ethers and esters of halohxdrins could be produced by chlorination of an ethylenic compound with tert-butyl hypochlorite in the presence of water, alcohols, and organic acids, respectively. Harford also reported (IO) the use of tert-butyl hypochlorite purely as a chlorinating reagent-for example, he claimed the formation of 1-chloro-2-methyl-1-propenefrom isobutene and of 2-methyl-3-chloro-2-butene from trimethylethylene. These results differ from those described by Kenner ( I S ) who obtained a 65% yield of 3-chloro-1-cyclohexene from the reaction of cyclohexene and tert-butyl hypochlorite under Ziegler’s conditions (18) for effecting halogenation with positive halogen compounds. Bergstrom ( 2 ) and Bolland and Koch (3) showed that, in the union of oxygen and methyl linoleate, conjugated hydroperoxides are formed. Bolland and Koch postulated that this effect resulted from rearrangement of an intermediate free radical obtained by removal of a proton from the central methylene group (C-11) of methyl linoleate. This explanation is in agreement with Farmer’s views (6) concerning the participation of free-radical mechanisms in thz oxidation of unsaturated fatty acids with gaseous oxygen. Another instance of movement of a double bond during substitution was obqerved by Ross, Gebhart, and Gerecht (16) in the reaction of maleic anhydride and methyl oleate. The facts outlined in the preceding paragraphs appeared to
U.S. Department
of Agriculture, Peoria, I l l .
indicate that subTtitution of an unsaturated fatty acid might, in general, be accompanied by shift of a double bond. Substitution of halogen into polyunsaturated fatty acidr by means of terl-butyl hypochlorite would therefore be expected to produce conjugation.
RCHZ-CH=CH-CH2-CI-I=CH-CH2-R’
+ RCHz-CH-CH=CH-CH=CH-CHz-R’ I
+ (CHa)8COCl
+
I
c1 Since halogenated fatty acids should not be subject t o deterioration such as that undergone by oxidized oils as a result of decav of hydroperoxide groups, the conjugation produced should be stable. The object of the present experiments was t o determine the characteristics of the reaction of soybean oil with tert-butyl hypochlorite, with particular reference to the amount and type of conjugation which might result, and to investigate the effect of dehydrochlorination upon the chlorinated soybean oil. Preliminary experiments were conducted with methyl esters of soybean fatty acids. One equivalent of tert-butyl hypochlorite reacted exothermically with one equivalent of these esters to yield products containing from 18 t o 24% of conjugation (Table I). The transfer of halogen from hypochlorite to oil was essentially quantitative.
TABLEI. CHLORINATION
OF &IETIIYL
FATTY ACIDS
Time, Min.
Temp.,
20
40 40
O
C.
Cia,
%
7 -
Diene
ESTERSOF SOYBEAN
Conjugation. ’%------Triene Tetraene
10 43 14.4 3.6 GO 10.70 14 8 4 3 20 7.28 11 7 GO 50 9.72 12 6 20 GO 10 56 11 7 5 8 GO GO 10.93 11.3 10 6 a Maximum vrtlue for monosubstitution is 10 7w0.
x
0 4
0 0 1 0 1
3 6 2 6 6
Total
18 19 20 21 18 23
4
4 1 3 1 5
-
An importaht feature of the reaction was the induction period of 20 to 40 minutes, observed when the first increment of t e i t butyl hypochlorite was added The behavior when subsequent portions of hypochlorite were added depended mainly on the temporature a t which the reaction mixture was held. -4t temperatures above 60” C. the induction period was negligible aftcr the reaction had once started, and subsequent reaction orcurred
850
Vol. 41, No. 4
INDUSTRIAL AND ENGINEERING CHEMISTRY
suggests thst the total conjugation may be independent of the SOYBE.4N FATTY distribution among the various types of conjugation. Tithin ACIDS WITH VARIOUS Awoums O F tert-BuTYL HYPOCHLORITE AT the probable experimental error, the percsntage of new conjugated 65" C. bonds was directly proportional to the percentage of chlorine. Ilypochlorite, Thib result is in harmony with the premise that one doublr bond Moles/Mole - - --Conjiigntion, 70----. of Esters Cia, yo Diene rrlene Tetracne Total should shift for each chlorine atom introduced. 0.1 . . . 9.0 1.1 0.05 10.15 Thermal dehydrochlorination of the chlorinated oils resulted in a 6.2 0.5 6.6 13.40 0.60 2.4 1.o 9.7 18.1 20.7 0.2 substantial inerearc in the number of conjugated double bonds, 12.2 1.5 16.2 4.2 20.8 0.4 m o d of thi5 increase being found in the triene and tetraene type, 2.4 14.4 2.o 16.7 19.3 0.2 20.8 7.0 5.0 0.0 0.0 7.0 of conjugation (Table IV). About 50% of the halogen originall\ 43.1 1.0 10.0 0.0 0.0 1,o present was removed. The addition of potential dehydrochlorinaa Maximum value for monosubJtitution is 10.7% tion catalysts-e g , twt-butyl hydroperoxide or decamethyleric glycol-appeared to have no significant effect The formation of TABLE 111. RESULTS O F RATCH CHLORIKidTIO?iO F SOYBEAN OIL coniiderable amounts of triene and tetraene conjugation i s of A T 60" c. interest since other methods of isomerization produce oil.; conNew taining conjugation predominately of the diene type. Prolonged Conjugated Sample Cla, -------Conjugation, 70-- -. KO. % Diene Triene Tetraene Total Bonds, % dehydrochlorination a t higher temperatures removed 75.594 01 1 15.2 0.5 0.1 15.8 32.3 7.42 less of the original halogen, but the products vere very viseour, 15.2 34.4 8.62 2 11.7 2.8 0.7 16.6 1.3 42.7 9.26 3 8.4 6.9 and the conjugation present was mosely diene. Data adequate 21.2 4 10.2 2.2 57.0 9.31 8.8 to permit detailed comparison of these resultq with those ob16.6 36.6 9.39 2.6 0.4 5 13.6 9.92 20.7 5.1 1.0 14.6 6 48.5 tained by others could not be found in the literature I t appearj, 19.9 48.3 6.3 12.5 10.00 1.1 7 17.2 3.2 10.13 0.3 38.2 13.7 8 however, that when soybean oil chlorinated with lert-butyl 16.4 3.7 10.33 2.6 9 10.1 48.1 hypochlorite is dehydrochlorinated, it behaves in much the same 19.9 11.9 10.52 1.2 6.8 10 49.0 20.5 50.2 12.5 10.87 6.8 11 1.2 manner as do other chlorinated oils ( 7 ) . Average 18.2 A O . 5 T o mitigate the effects of thp large amount of heat liberated hi a Maximum value for monosubstitution is 10.8%. the reaction between soybean oil and tertbutyl hypochlorite, an apparatus for continuous Chlorination was devised. In this apparatus terl-butyl hypochlorite vapor was forced bv a stream smoothly upon addition of further increments of hypochlorite. of carbon &oxide countercurrently through a descending stream Because of the induction period, caution must be exercised to of soybean oil in a packed column. The reaction none was make certain that each increment of hypochlorite added, espereadily maintained at 100' C by a combination of the external cially the first, has reacted before further addition of the reagent heating and the heat of reaction, arid the induction peiiod c a u 4 is attempted. If excessive amounts of hypochlorite be present no difficulty. when reaction occurs, violent ebullition and rapid temperature This procedure was noL entirelv successful because the result* were not reproducible. The particular type and amount of rise may cause the reactants to be ejected from the reaction vessel. I n addition t o the experiments where equivalent amounts of conjugation produced varied widely and apparently randomly. reactants were used, the reaction of methyl esters of soybean Whether this effect resulted from failure to control certain fatty &rids, both with more and with less than the equivalent variables adequately-for example, the rate of vaporization of amount of hypochlorite, was studied. The results (Table 11) the hypochlorite-or whether i t was due to differences in ieaction mechanism caused by the use of vaporized rather than liquid show that the maximum conjugation is produced witla 1to 2 moles of hypochlorite, and that increasing or decreasing the proportion hypochlorite is not known. of hypochlorite eventually decreases the amount of conjugation Methods of effecting continuous dehydrochlorination mere also emploved. The most successful device consisted simply of an produced. externally heated, evacuated, inclined tube down which thr Reaction of soybean oil and tert-butyl hypochlorite in equivachlorinated oil was passed. The advantage of this method wa< lent amounts gave products having the properties shown in Table 111. These experiments were conducted under constant condithp removal of about 65y0of the halogen originally present during contact times of about 30 seconds. -4considerable increase in tions in 5 0 far as this could be accomplished without special equipment. Chlorination varied from 7.42 to 10.8770'0,correthe number of new conjugated bonds formed was also noted sponding to a yield of chlorinated oil ranging from approximately Table V gives data for these experiments. 70 to lOOsl,. The amounts of diene, triene, and tetraene conjugaThe dehydrochlorinated oils dried rapidly, in the absence of tion in the individual samples varied widely. However, the low metallic driers, t o fairly hard, colorless films n-hich were frequently wrinkled or frosted. The set-to-touch time was iiquall, standard deviation of the avwage value for total conjugation
TABLE11. REACTIOKO F METHYLESTERS
OF
7-
.
.
I
TABLE IXr. Sample NO.
1 2 3 4
PIEOPERTIES O F ~ T i L O R I N . 4 T V V SOYBEAN OIL, BEFORE AND AFTER DEHYDROCRLORINATION
Treatment
Original No. 1 heated 3 hr. a t 120'' C. Original No. 3 heated 5.75 hr., 150-160° C. 5 No. 4 heated 8 hr., 175-186' C. 6 No. 5 heated 7.75 hr., 180-220° C . No. 6 heated 2 hr., 180-220° C. 7 Original 8 No. 8 heated 4 hr.a, 150-160' C . 9 No. 8 heated 4 hr,b, 150-160' C. 10 Original 11 12 No. 11 heated 4 hr., 150' C. 1% tevt-butyl hydroperoxide added. b 1% decamethylene glycol added.
C1,
%
9 a9 6 34 10 97 5 33
5 04 4 10
2 9 4 4 10 4
72 92 92 75 02 92
c1
Remoird,
x
ad
'4
.
51.4 54.0 62.6 75.2 .
50.5 52.2
...
50.9
%
38.9
,..
,
New Coniugated Bonds,
.
I
44,s
... ...
...
~
, . .
26.3 28.4
...
38.2
c---
Diene 13.6 8.0 16.6 11.0 12.8 13.2 11.3 14.6 12.1 7.7 15.7 8.3
-Conjugation, yo- Triene Tetraenr Total 16.6 2.6 0.4 27.5 4.9 13.3 19.4 0.3 2.5 31.1 15.7 4.4 31.6 15.9 2.9 20,7 0.6 6.9 14.3 0.2 2.8 20.7 5,l 1.0 28.2 13.8 2.3 27.0 3.6 l5,7 0 6 20.4 4.1 29.5 16.1 5. I
Color (Gardner) 10-11 10 10-11 13-14 13-14 14-15 16-17 13 13 13
.,. .
I
.
Viscosity (Gardner) E-F F
Set-tuTouch, Min.
..
H
J-K T-U 2% ZB
I
R N . 1
* *
,.
..
.. 26 20
..
30
April 1949
INDUSTRIAL AND ENGINEERING CHEMISTRY
851
cases, as with rubber or phenols, violent reaction occurred as soon TABLEv. RESULTS OF CONTINUOUS DEHYDROCHLORINATION as the reagents were mixed. In others, as with cyclohexene or OF CHLORINATED SOYBEAN OIL methyl oleate, mixtures with tertbutyl hypochlorite appeared to Property
Original
g;'r?moved, % New conjugated Conjugation, % bonds. % Diene Triene Tetraene Total Color (Gardner Viscosity (Gardner)
10.2
...
... 15 2 0 .o 0 3
-.
15.5
... ..
Oil Dehydroohlorinated at 300' C. 400' C. 3.66 3.51 64 65
59.3 10.7 16.1 .-5.3 32.1 17 P
51.7 12.9 13.9 2.8
29.6 18 S
of the order of 20 to 30 minutes. These are properties which would be expected for an oil containing substantial amounts of triene and tetrame conjugation. A few Preliminary studies of the Properties of these oils were made in ester-gum varnishes of 50gallon oil length. Both nonbreak soybean oil and soybean oil treated with 2% maleic anhydride were used for comparison. The principal disadvantage of the dehydrochlorinated oil was its tendencyto release additional hydrogen chloride a t the high temperah - e s required for Preparation of the varnishes. The Principal advantage was a considerable reduction in the time required to prepare the varnish and in the time needed for the film to become tack-free. The Engineering and Development Division of this laboratory has examined the experimental information, and reports that the Cost of this hypochlorite process should be low enough t o make it economically feasible. However, further research and development work is necessary before any final conclusions can be reached. tert-BUTY L HYPOCHLORITE
PREPARATION. The procedures reported by Chattaway and Backeberg (4)and by ~~~i~ and ~~~~i~~ (18) were modified to permit preparation of large amounts and to avoid use of low temperature baths. A solution of 14moles of sodium hydroxide in liters of water was prepared in a 12-liter three-necked, round-bottomed flask equipped with gas inlet and outlet tubes and a mechanical stirrer. When the solution had cooled to room temperature, 7 moles of ter& butyl alcohol were added together with sufficient watertoproduce homogeneous solution. The apparatus was then placed in a water bath a t 18'to 20°C.,andthecontentswerestirredvigorously while a current of chlorine was introduced well below the surface. Passage of chlorine was continued until formation of a second liqnid phase was complete, from 4 to 6 hours being required. The upper, oily layer was then separated, washed with sodium carbonate solution until neutral t o Congo red, washed four times with water,and dried Over calcium &loride, The yield was 547 to 750 grams (72 to 99%) of tertbutyl hypochlorite of 98% purity, as determined by iodometric titration, The product was used without further and stored in brown glass bottles at o, Beoause of the extensive work contemplated with STABILITY. this reagent, attempts .cvere made to find any conditions under which the use of the reagent might be haEardous. I t was found bath in all-glass that this reagentcould be distilled on a apparatus. Any residue appeared to be stable under these conditions. Exposure t o ordinary room illumination, daylight or artificial, appeared to have no effect. Ultraviolet light, however, caused exothermic decomposition which increased in violence as exposure to the ultraviolet light was prolonged, Decomposition stoppsd upon removal of the ultraviolet light, and there was no indication of spontaneous continuation of the decomposition. tert-Butyl hypochlorite reacted readily with a wide variety of substances containing active or replaceable hydrogen, such as acids, unsaturated hydrocarbons, phenoh rubber, and saturated and unsaturated fat acid esters. In some ~
be stable indefinitely unless they were irradiated with ultraviolct light. A third type of behavior was that with methyl undecylinate or esters of soybean fatty acids. Here violent reaction may not occur for periods up to an hour after the reagents are mixed. In all of these cases, the violent reaction which ensued could not be described as an explosion but rather as vigorous ebullition, caused by a sudden exothermic reaction. Nevertheless, ejection of the contents of the reaction vessel frequently resulted, and if the apparatus offered resistance, parts were forced loose and broken. Since insufficient data exist to permit prediction of the expected behavior, tart-butyl hypochlorite should never be mixed with another material except in quantities sufficiently small t o render any violent reaction harmless. All-glass apparatus should be employed, and no rubber connections should be utilized. Plastic tubing is satisfactory for Connections t o gas inlet and outlet tubes, if needed. Stirrers should be sealed either with mercury or by passage through long, close-fitting sleeves. As a further precaution, considerable time should be allowed before it is concluded that reaction has not occurred. Particular care should be exercised when hypochlorite is added in increments t o a reaction mixture because reaction frequently does not Occur until a sig. nificant concentration of hypochlorite is built up. Once the characteristics of a given reaction have been determined by cautious experimentation, it is usually possible t o arrange conditions which involve no hazard t o the operator. CHLORINATIONS
METHYLESTER& OF SOYBEAN FATTY A c ~ u s . A 1.35-gram sample of methyl esters of soybean fatty acids was placed in a small teRt tube, and 0.5 gram (1.0 equivalent') of tert-butyl hypochlorite Was added. The tube was then Placed in a bath maintained at the desire8 temperature for 20 to 60 minutes. Volatile material was removed from the sample by passing carbon dioxide through it for 1hour a t room temperature, under vacuum. was then detRrmined spectrophotometrically and chlorine analysis was obtained by combustion with sodium I gives the Of peroxide in a Parr bomb* shows the Of experiments Of this type. experiments in which more and less than one equivalent of hypochlorite was used' SOYBEAN OIL. Into a 500-ml. three-neck flask, equipped with having a a stirrer, condenser, thermometer, and dropping tube reaching to the bottom Of the flask, was Placed long 250 grams of alkali-refined soybean oil (iodine number 134.6). The apparatus was then placed in a constant temperature bath Of lert-butyl maintained at 60" C., and 92 grams (l.0 hypochlorite w?re added in increments of about 10 ml. A period of about 30 minutes elapsed after addition of the first increment before reaction occurred, as indicated by a rise in temperature of the reaction mixture to about 87"C. When the temperature had fallen t o 60", a second increment was added. This procedure was repeated until all of the hypochlorite had been added. The reaction mixture was allowed to stand for 1 hour at 60°, and then Volatile material was removed by distillation in vacuo a t 50" in a current of carbon dioxide. This experiment was repeated eleven times with the results indicated in Table 111. From 80 to 100% of the expected recovery of tert-butyl alcohol was obtained. The recovered alcohol usually contained a small amount of free acid due to slight dehydrochlorination during distillation.
c.
Calculated by assuming t h e average moleculei weight of methyl esters (or soybean oil) t o be t h a t of methyl oleate (or glyoeryl trioleate). This ssaumDtion introduces little error because t h e fat aoid radicals oresent differ only slightly i n molecular weight.
852
INDUSTRIAL AND ENGINEERING CHEMISTRY DEFIYDROCALORIYATIONS
CHLORIN~TF S DO Y R COIL. ~ ~ The sample n-as placed in a flask connecttd to an efficient trap. After evacuation of the system to 3 mm. or lesc, steam from ~tgcncrator operating at the same pressurr was paised through the oil while it was heated to the desired temperature for the allotted time. Table IV gives the results of several dehydrochlorinations. The percentage of new conjugated double bonds formed by dehydrochlorination was calculated as follows: The total number of conjugated double bonds per 100 fat acid radicals before dchy5 by taking the sum of twice the perdrochlorination ~ a found centage of diene conjugation, three times the percentage of triene conjugation, and four times the percentage of tetiaene conjugation. This sum was then deducted from the corresponding sum for the oil after dehydrochlorination. I n order to protect the mechanical pump used in these experiments a train of absorption tubes packed with calcium oxide and calcium chloride o as connected between the apparatus and the pump. After each use the pump Tyas drained and flushed. Hydrogen chloride seldom reached the pump; most of it wag retained by the steam caught in the trap and the remainder was renioved by the absorption tubes COKTINUOUS DICHYDROCHLORISAT~OX. The apparatus consisted of a Liebig condenser 50 cm. long, having an inner tube 10 mm. in diameter and a shell 17 mm. in diameter. The outer shell was wound with Nichrome ribbon. Spacing of the windnq \vas uniform except at the upper or feed end where spacing was gradually increased in order to provide a substantial temperature gradient. I n this way oil entering the column was brought bomewhat less suddenly up to temperature, and spattering $$asthus reduced. Temperature was measured with thcrmocouplcs inserted through the water inlets Thc conden+Pr wa- mounted a t an angle of about 30°, the upper end was equipped with a dropping funnel (Z), and the lower end was connected to vacuum through a conventional fraction cutter. The tube wa5 evacuated
Vo!. 41, No. 4
to a pressure of 3 mm. or less, and was heat,ed to the desired t,enipcrature with a variable voltage transformer. Oil was then passed down the tube at a rate which provided about 30-second coiit'act time in the hot tube. Table V gives results obtained by t)his method. ACKh-OWLEDGRIENT
The authors express their appreciation t o t,he .halyticiil arid Physical Chemical Division of this laboratory for supplying thc spectrophotometric analyses. The experiment'son continuous tlchydrochlorination were conducted by J. E. .Ta ckson. LITERA'KKE CITED
(1) Ashburn, G., and Frank. R. F., IND. Esa. CHEM., ANAL.ED., 16, 418 (1944). (2) Bergstrom, S., S a f u r e , 156, 717 (1945). (3) Bolland, J. I,., and Koch, I T . P., J . C h c m hoc., 1945, 445--7. (4)Chattaway, F.D., and Backeberg, 0 . G., I b i d . , 1923, 299!3r3603. ( 5 ) Emling, B. L., T'ogt. R . R.,and Henriiom G. F., J . Arn.-Chem. SOC.,63, 1634-5 (1941,. (6) Farmer. E. H . , Trans. Famriny Soc., 42, 228-36 (1946). (7) Gardner, €1. A., U. 6 . Patent 1,452,553 ( A p d 24, 1923'1. (8) Harford, C. G., I b i d . , 2,054,814 (Sept. 22, 1936). (9) I b i d . , 2,107,789 (Feb. 8,1938). (10) I b X , 2,179,787 (NOT.14, 1939). (11) I b i d . , 2,207,983 (Jul>- I O , 1940). (12) Irwin, C . F., and Henilion. G. I:., J . Am. ('hem. SOC..63, 858-60 (1941). (13) Kcniier, J . , Nature, 15F, 370 (1945). (14) Radlove, S.B., Teeter., H . A I . , Bond, IT. FI., CoTTvan, J. C.,and Kass, J. P.. IKD. EXG.CHEX.,38, 998 (1946). (15) Ross, J..Gebhart, A. I.. and Gerecht, J. F., J . Am. i'hem. S o c . , 68, 1373-6 (1946). (16) Sandmeyer, T., Her.. 18, 1767-89 (1835). (17) I b i d . , 19, 857-61 (1886,. (18) Ziegler, K., Spat,h, A%., Scliaaf, E., Schumaiin, W., a.nd Winkc!mann, E., Ann., 351, 80-119 (1942). RECEIVED October 31, 1947. Presented before t l ~ ePaint and I'rotcotive Coatings Group a t the 15th west Regional Meeiing, A x x ~ r c . 4C~I I I ; M ~ C A L S o r I m Y , Kansas City, 110.
Dielec
a SILICON DIQXIDE
ROBERT V. JELINEM, P-IENKY B. LISFQRD, E. IC. McMAHON, AND PHILIP W. SCHUTZ' Colunabia University, .Yew York 27, 3.. Y .
A
N EARLIER paper by Schutz and AlcMahon (4) was the first published work on the dielectric heating of brds of granular solids, and y a s concerned mainly with alumina, although some data on silica were also presented. The work discussed below is a continuation and expansion of t,he experiments with silica, and i t is hopcd that the data prcsented will help to clarify this problem. The earlier work has shown definitely that for both alumina and silica there exists a variation of heating rate with grain size. For alumina this variation was established as an apparently linear increase of heating rate with particle diameter over the range 0.103 to 2.04 mm., as shown in Figure 1 of this paper, a reproduction of the lower curve of Figure 6 of ( 4 ) . The earlier results for silica were less complete than those for alumina, and it was concluded merely that a general increase in heating rate with grain size exists for silica over the particle size range 0.664 to 3.53 mm. That this latter statement is not entirely true is 1
Deceased.
pointed out in this paper. However, since experimental t.rchniquc in t,he earlier work ivith silica ha,s now been establialictl as faulty, the new data are a clarification rather than a contradiction of thosc previously presented. THEORY
By means of conventional electrical relationships, tJlic usual equation for power input into a dielectric (1, 6) should hc:
where q,. = power input per unit volume (watts per ml.)
E/d
voltage gradient acrojs dielectric (volt3 prr em.) frequency (cvcles per sec.) K = dielectric constant tan 6 = dissipation factoi. K(tan 6) = loss factor =
f=
