Document 55wN5VrwZYD2GD86myvom58R

C- OiMFiLJ^'NTj i- 2369 AROCLOR 1242 I!!13ROVED XNTERDTVISIONAL TRAN ^wpV DIELECTRIC PROPERTIES TECHNOLOGICAL DEVELOPMENiTtSs ~ W. Gluesenkamp F. B. Zlenty D. T. Mowry To: Dr. J. H. Lum M. C. Throdahl Director, R&D J. S. Hayes Organic Chemicals Division AROCLOR 1242 WITH IMPROVED DIELECTRIC PROPERTIES DESCRIPTION OF PROJECT TR-23, R&E, 1959* October 15> 1959 2369 AROCLOR 1242 IMPROVED DIELECTRIC PROPERTIES A study of a market survey report by John S. Harris of Organic Division Development (Ref. l) and our own preliminary investigations led us to the conclusion that an attempt should be made to improve the following properties of Aroclor 1242: a) Dielectric constant (Dk) (5.825) Sb) Pour point (-19C) c) High temperature stability . We have had gratifying success in improving property (a) and our results constitute the subject of this report. - Result: A dielectric material, Aroclor 1242' with a Dk of 6.7-7.2, and a practical procedure for its preparation believed subject to patent coverage. HISTORICAL BACKGROUND "We have very few products of which we are the only manufacturer, and the Aroclor family is a prime example. It has had a long and profitable history and last year (1955) some 35 million pounds were sold for over $5 million - two-thirds of this for electrical use. Notwithstanding its low price, it gives us a very good return on investment. But with our protective patents gone, and a half dozen other chemical manufacturers looking at the Aroclors, we stand a good chance of losing our hold on these products at any time. With the electrical sales alone estimated in 195 from 30 to 55 million pounds, this would be quite a loss." This quote from Ref. 1 explains why we are carrying out a research program in this area. *Thls report was written by Harold Weingarten; Van R. Gaertner, Group Leader. ~' - DSW 621846 STLCOPCB4095877 2 The process described in this report was developed as a direct result of a three-phase fundamental study carried out during the last two years. Phase one, the study of the Gomberg reaction, permitted the labor atory preparation of high Dk chlorinated biphenyls leading to the discovery that an increase of one or two units in the Dk of chlorinated biphenyl mixtures did not adversely influence the power factor or resistivity. Phase two, the study of biphenyl chlorination, revealed 2,4'-dlchloroblphenyl as the precursor of most of the high Dk trlchloroblphenyls. A summary of this work is given in Table I. And Phase three, the study of the distri bution of isomers in Aroclors, revealed Aroclor 12J2 to be an excellent source of 2,4'-dichlorobiphenyl (see Table III). Figure 1 summarizes the isomer distribution study made on Aroclor 1242. The vapor phase chromatogram of Aroclor 1242' is included for comparison. Also included in Figure 1 are the estimated Dk's of the trlchloroblphenyls arising from the 2,4'-isomer. This project was begun in February of 1957 and the total expense through June, 1959, was $89,965. CHARACTERIZATION Process Method (a) 1. Aroplor 1252, made by direct chlorination of biphenyl (not by blending 1221 and 1242), is fractionally dis tilled* (see Figure 2) and the fraction rich in 2,4'dichlorobiphenyl (85 to 100$)** is used in the next step. 2. The fraction rich in 2,4*1-2d*ic4hlorobiphenyl is chlorinated to the trichloro level according to the standard Aroclor processing conditions. Purification is also carried out by the standard Aroclor method. 5* The low boiling fractions can be recycled through the chlorinators or used as Aroclor 1221. 4, The high boilers are chlorinated to higher levels for non-capacitor uses. A sample of Aroclor 1260' was pre pared in this way and sent to Organic Division for comparison with standard Aroclor 1260. Its physical properties and Dk were found to be essentially identical to the standard Aroclor 1260. Table II described the comparison (Ref. 4). Minor instability was apparent but can presumably be corrected by proper treatment. *For details see references 2 and 5. **Vapor phase chromatography is used to follow the distillation and determine the purity of the desired fractions. DSW 621847 % STLCOPCB4095878 3- - I TABLE I BIPHENYL CHLORINATION STUDY 24 26# 74 ,HoAC:CC14 'CI2 R.T. FeCl ^ CI2 400 43^ Benzene or CCI4 4 57# 22 78 <r No reaction in 40 days Cl 46 54 -i20# 24# 10# 60# 5# Must give all six isomers in about the same order of magnitude DSW 621848 STLCOPCB4095879 Figure 1. -4Aroclor Component Study STLCOPCB4095880 5 Hid (D ro ro STLCOPCB4095881 -6- TABLE II Property Aroclor 1260 Dayton Lot T-656 Aroclor 1260 Specifications Color, APHA Condition Sp. Gr. at 90/15-5C Acidity, mg ROH/gm. Moisture, ppm Viscosity at 210C, SUS Refractive Index at 25C Inorganic Chlorides Pour Point, C Distillation Range ASTM D-20 10$ Distilled by wt. 50$ Distilled by wt. 90$ Distilled by wt. 45 Clear 1.555 0.002 20 72.4 1.6459 NDA 28 592C 595 411 150, max. Clear 1.555-1.566 0.014, max. . 55, max. 72 - 78 1.6455 - 1.6470 No detectable amount 25 - 54 585-598C 390-404 400-420 Corrosion Test (6 hrs. at 210Cwith bright Change in wt. of Al. Color Condition Acidity Inorganic Chlorides Aluminum foil) None 45 Clear 0.005 0.1 ppm None 150, max. Clear 0.014, max. NDA - Dielectric Constant, 100C, lkc Power Factor, 100C, lkc Resistivity, 100C Dielectric Strength at 25C Monsanto Stability Test (l6 hrs. at 210C) 5.8 0.21$ q 825 x 10y 55 KV 0.8 ppm 5.6 - 3.8 -- 500, min. 30 KV, min. 0.7 ppm, max. Method (b) 1. Aroclor 1232-S is prepared by chlorinating biphenyl in the presence of FeCljj and sulfur as catalyst. Typical experimental conditions: Melt 4 kg of biphenyl in convenient size 4-neck flask fitted with stirrer, thermometer, gas inlet and outlet apparatus. Add 20 g FeCl} (anhyd. sublimed)* and 11 g of sulfur (flowers)*. *A ten-fold excess of catalyst was used here as a precaution against losses due to atmospheric moisture. DSW 621851 STLCOPCB4095882 -7 Pass in chlorine gas as rapidly as possible maintaining temperature at 100C. Continue chlorination until liquid density reaches 1.280 (25C) . Pass dry nitrogen through the crude maintaining temper ature at 80 to 90C until HC1 is removed. Add 30 g Ca(0H)2 (powdered) and distill through a 6 to 8" dis tilling head, taking all that canes over. The boiling range is about 130 to 215/10 mm. 2. The Aroclor 1232-S is then put through a fractional distillation exactly as described in method (a). Method (b) provides us with three advantages as a result of the difference in isomer distribution (Table III). TABLE III TYPICAL ISOMER DISTRIBUTION: AROCLOR 1232 STANDARD 2-Chloro 4-Chloro 2,2'- 2,4'- 4,4'- Trichloro 00 00 rH 14$ 6$ 52$ 15$ 9$ TYPICAL ISOMER DISTRIBUTION: AROCLOR 1232-S 1$ 0 12g 1$ 44$ 31$ 13$ The first advantage is flexibility. For example, we can remove essentially all monochlors without greatly increasing trichlors thus eliminating the recycle process. The second advantage is an increase in 2,4'-isomer con centration. And third, the 2,4-isomer is greatly reduced thus simplifying the fractionation since the 2,4-isomer is the most difficult contaminant to remove. ESTIMATED COST Based on a production volume for Aroclor 12421 of 15 million lbs. per year R. C. Binning has estimated the new distillation facilities would cost $600,000. He has further estimated that for an incre mental sales price of 4.0^ per lb. for Aroclor 1242' a 55$ return on this investment (after taxes) could be realized. J. 0. Bright of Organic Division Research Dept, has also prepared an economic evaluation of the Aroclor 1242r process (Ref. 5)* ` The estimate above does not cover the use of FeClj + S catalyst, which should be more favorable. DSW 621852 STLCOPCB4095883 8- - RELATIONSHIP TO EXISTING PROCESS One of the most favorable aspects of the process under discussion is the relatively small change it will cause in the existing process, requiring little more than the installation of the fractional distillation equipment (see Figure 2). Since the new process involves a fractionation the fate of the chlorinated biphenyl byproduct (low Dk dichlors) is of great importance. We feel certain this problem can be handled by the chlorination of the byproducts to higher Aroclors such as 1248, 1254, 1260 etc. These higher Aroclors are expected to have essentially the same physical properties as the corresponding standard Aroclors. This is borne out by the properties of Aroclor 1260 prepared from the low Dk dichloro byproduct (see Table II). Also important is the ratio of high Dk dichlors (to be converted to Aroclor 1242') versus the low Dk dichlor byproduct. Assuming optimistically that all of the Aroclor 1242 to be replaced by Aroclor 1242' and all of the low Dk dichloro byproduct to be converted to higher Aroclors, Table IV was constructed to show that a reasonable product balance is possible. Aroclor TABLE IV Yearly Production M Lbs. /based on Aug. 1958 to Apr. 1959 (Ref. 6)_7 Amount of . Dichloro _ Required M Lbs. 1242 1248 1254 1260 1262 1268 12.6 4.05^ 6.5 7.45 > 0.42 0.15 11.0 12.4 Product Balance Example (A) Aroclor 1242' High Dk Dichlor 12.6 M lbs. 11.0 M lbs. Low Dk Dichlor l6.4 M lbs. Example (B) 12.6 11.0 14.0 Example (C) 12.6 11.0 11.0 DSW 621853 STLCOPCB4095884 -9 The yearly production data (Ref. 6) is based on monthly averages from August 1958 to April 1959 for both the Anniston_and W.G.K. plants. 12.6 M lbs. of Aroclor 1242' requires 11.0 M lbs. of aichloro precursor while all of the remaining higher Aroclors require 12.4 M lbs. Three examples are given under product balance (Table IV) to describe the limits of high Dk dichlor versus low Dk dichlor distribution. Example (A) was actually calculated by R. C. Binning (see Refs. 1 and 2) -based on the composition of current Aroclor 1252 and shows a surplus of low Dk dichlor byproducts. Example (B) was estimated based on the composition of Aroclor 1252-S (see Table III) and the low_Dk dichlors are found to be much closer to the 'ideal' 12.4 M lbs. Example (C) is an estimated lower limit for the production of dichloro byproducts, assuming a lower degree of chlorination to give negligible trichlors. Both (A) and (B) can then be made to approach (C) more closely by a judicious selection of the degree of chlorination and the conditions of recycle of monochlors to recover more of the 2,4' values. PRODUCT Specifications Aroclor 1242' - Dk250 8.7 to 7*2 ASTM pour point -19C The other properties should be identical to those of standard Aroclor 1242 which are listed here for reference. '. '/ Density 25 1.580 Distillation range 525-380 Refractive index D-line 20 I.627-I.629 Outstanding Features Aroclor 1242' has a Dk between 6.7 and 7-2 compared to a Dk of 5.8 for standard Aroclor 1242. MARKETS (This section supplied by J. K. Craver, R&E Development) Based on estimates made by the Organic Development Department, there appears to be a market of from 5-15 million pounds per year for a high dielectric constant fluid to be used in low voltage power-factor correction capacitors. This is in addition to those markets now served by the regular Aroclors and would have to be developed over the next 5-10 years. In addition to satisfactory physical and electrical properties. Organic feel that any new product should be patentable and that the selling price should not be higher than 30^/lb. . DSW 621854 STLCOPCB4095885 10 The R&E Patent Department believe that we will have patent coverage on the processes for producing the high Dk Aroclor, A sales price in the range of 19-20^/lb. would yield 40$ pre tax return according to calculations made by Organic Research (Ref. 5 ) This estimate assumes the fractional sparation of a pure 2,41-2d*ic4h5l6o7rodiphenyl isomer which is then chlorinated to a trichlorodiphenyl of high Dk. The lower-boiling chlorinated diphenyls are recycled and the higher-boiling materials are used to produce the conventional Aroclors. Production rates of 5-15 million pounds per year are assumed. At this price level, we feel the improved Aroclors described in this report should be of considerable interest to the electrical industry, particularly in low voltage devices. We are aware that there are a number of other new dielectrics being considered by the trade; tolyl xylyl sulfone, 2,21-dichloro diphenyl ether and hexachiorobutadiene to name three that seem especially promising. However, the already established position of the Aroclors and their history of reliability, at low cost should make the job of introducing an improved grade much simpler than that of bringing out an entirely new product such as tolyl xylyl sulfone or hexachiorobutadiene. The picture on 2,2'dichlorodiphenyl ether is somewhat more enigmatic since much --: depends upon the relative costs of the raw materials. REFERENCES 1. J. S. Harris, "Liquid Dielectrics and Aroclor: A Market Survey", Development Department Report O.D. 1154, August 15, 1956. 2. R. C. Binning, Report to R. W. Schuler on Relative Volatility of Aroclor 1252 Components, Dayton, April 8, 1959* 5. R. C. Binning, Report to R. W. Schuler on Fractionation Requirements for Separation of 2,41-Dichloroblphenyl from Aroclor 1252, Dayton, June 5, 1959* 4. A. M. Ellenburg, memo to Harold Weingarten on Aroclor 1260 from Special Process, St. Louis, September 8, 1959* 5. J. 0. Bright, memo to F. B. Zienty, August 5, 1959, Economic Evaluation of High Dk Aroclors. 6. A. M. Ellenburg, memo to M. Kosmin on Aroclor Process, St. Louis, June 8, 1959* 7. H. Weingarten, memo to G. F. Deebel on Preparation of Aroclor 1242' and 1260', Dayton, July 28, 1959* DSW 621855 STLCOPCB4095886 11 8 . H. Weingarten, memo to A. M. Ellenburg, on Compositions of Aroclor 1232 by Catalytic Variations, August 13# 1959 9 H. Weingarten, Progress Report on Dielectrics, Issue 23, June, 1959- RECOMMENDATIONS FOR FUTURE WORK At Organic Division We recommend that a study be carried out to determine the optimum level of 1 dichlorination1 and recycle chlorination. The optimum level will, of course, be related to the desired product balance between high Dk dichlors and low Dk dichloro byproduct. We further recommend that the process details be firmed up and moved into the pilot plant stage as soon as possible since there is likely to be a two-year lag between the time we present our customers with samples and the time they complete their evaluations for acceptance. _ At R&E Division The pour point of Aroclor 1242 is another property we are anxious to improve while retaining a high Dk. A study of the viscosity of a series of chlorinated biphenyl isomers and the three lower . Aroclors (Ref. 9) convinced us that the viscosity and pour point are related to the degree of chlorination and not related to isomeric structure. We are, therefore, in the process of testing . high Dk Aroclors intermediate between 1232 and 1242. Estimated completion time - 3 months. We also plan to finish our chlorination and component studies and examine the possibilities of improving the Aroclors (pour point and Dk) with ether type additives. Estimated completion time - 12 months. PATENT SITUATION ' We have filed a patent application (C2167) directed to the dielectric composition of matter, the process of preparing same described above as methods (a) and (b), and electrical capacitors employing the new dielectric composition. To date we have not received the first Office action from the Patent Office. Additionally we have filed a patent application (02088) directed to high Dk Gomberg chlorinated biphenyls as dielectric composi tions of matter, the method of preparing same, and electrical capacitors utilizing these Gomberg mixtures. The first Office action has been received from the Patent Office and five claims directed to the electrical capacitors have been held to be allowable. DSW 621856 ; 1 j ! j j ; j j j j j I ! j i j 1i I STLCOPCB4095887 12 ACKNOWLEDGMENTS The acknowledged items listed below are to be given by the Organic Chemicals Division via memoranda between General Managers to be attached as addenda to each copy of this report. A. Acceptance, reason and data. B. Requirements for and availability of manpower to initiate and carry on the project. C. Program and time schedule for. further development of the project. ^^* / Associate Director bw DSW 621857 STLCOPCB4095888