PREPARATION OF SOME NEW HYDROXY A(’n>S BY THE ACTION OF CHLOR HYDRINS ON SODIUM M ADONIC ESTER BY HERBERT ORION CALVERY TMESIS FOR THE DEGREE OF BACHELOR OF ARTS I N CHEMISTRY COLLEGE OF LIBERAL ARTS AND SCIENCES UNIVERSITY OF ILLINOIS 1921 Afl Vb*\ C\2> UNIVERSITY OF ILLINOIS CNJ uu My_25,__ I 92 l__ THIS IS TO CERTIFY THAT THE THESIS PREPARED UNDER MY SUPERVISION BY __K2HBEHT ORION _CALVERY _ ENTITLED ?.HEPAMTI ON _ QF_ J3 0ME_ J$W_ H YD ROXY _ AO I _D_ S_ BY _THE_ _AG1 I OIL _ . OP _ JO H L OR H YD R_IN S _ OH. _ SOB I U M JAAL 0 II I _C _ E S TER IS APPROVED BY ME AS FULFILLING THIS PART OF THE REQUIREMENTS FOR THE DEGREE OF ^_^_he_lor__o_f _ Ar ts__in _Chanis_t ry_ C* J. Instructor in Charge Approved : HEAD OF DEPARTMENT OF 4A3508 . - 2 - ACKNOY.T1EDGMENT . The writer wishes to express his appreciation to Dr. Marvel for his constant and helpful directions and suggestions during these preparations. His advice and helpfulness has been greatly appreciated. Digitized by the Internet Archive in 2016 https://archive.org/details/preparationofsomOOcalv -3- TABLE OF CONTENTS. I . Acknowledgment . 2 . II. Introduction. 4. III. Historical. 4. 1. Ethylene chlorhydrin. 4. 2. Trimethylene chlorhydrin. 5. 3. Trimethylene iodhydrin. 8. 4. Trimethylene chloracetate . 8. 5. Trimethylene iodacetate. 9. 6. Phenoxy propyl bromide. 9. 7. Phenoxy propyl diethyl malonate . 9. 8. Normal nhenyl octyl ether. 10. 9. Normal octyl bromide. 10. 10. Normal octyl alcohol. 10. IV. Theoretical. V . Experimental . 1 . Prenaration of 2. Prenaration of 3. Prenaration of 4. Prenaration of 5. Preparation of 6. Preparation of 7. Preparation of 8. Prenaration of 9. Preparation of 10. Prenaration of 11. Prenaration of VI. Summary. VII. Bibliography. 11 . p-hydroxy ethyl malonic ester. 14. trimethylene chlorhydrin. 14. trimethylene iodhydrin. 16. trimethylene chloracetate. 17. trimethylene iodacetate. 17. phenoxy propyl bromide. 17. phenoxy propyl diethyl malonate. 18. phenoxy valeric acid. 19. phenoxy propyl butyl diethyl malonate. 19. normal phenyl octyl ether. 20. normal octyl bromide. 20. 20 . 21 . ' , x . . . -4- Introduction. Within recent years a great amount of work has been done on the preparation of chlorhydrins . The result of this work has been some cheap methods of their preparation and making them easily a- vailable. Malonic ester, from which sodium malonic ester is pre- pared, is also quite accessable . Since theoretically a condensa- tion of these two classes of compounds was possible it was thought that some new hydroxy acids might be synthesized which would be of interest . The expensive and inefficient methods by which these acids have been made have made their use quite limited. If the method of preparation which the writer tried and will discuss in this paper had been successful a new series of compounds of the form 0=G 0 CII* -CHi-6-CHj.-CHa 6 6=0 would have been prepared. These are dilactones. They would have been made by the following reaction: H00C 0=9 0 IIO-CH* -CH 2 .-C-CHs.-CH 2 . -OH — > CH* -CH^-C-CIIi-iHa. C00H 6 C=0 The acids which were possible by these condensations were the p y^and c9-hydroxy-dicarboxylic and also mono -carboxylic acid. HISTORICAL. 1. Ethylene chlorhydrin. - Ethylene chlorhydrin was first pre- pared by Wurtz 1 in 1859. He saturated some ethylene glycol with hydrochloric acid gas . He sealed the saturated solution in tubes and kept it at a temperature of 100 degrees for several hours . On removing the product from the tubes he found a product which had a constant boiling point of 128 degrees. This on analysis proved to be ethylene chlorhydrin. - I • v . -5- Carius 2 prepared ethylene chlorhydrin by an entirely differ- ent method in 1863. His method of preparation was to treat ethy- lene with 2-3 io hypochlorous acid. He fitted a large flask with a tightly fitting stopper and filled it with ethylene gas . He then poured on the gas 2-3 $ hypochlorous acid. The hypochlorous acid was in water solution. After a short time he distilled the result- ing product and obtained a product which gave all tests for ethyl- ene chlorhydrin. He gave its boiling point as 130 degrees. Schorlemmer 3 prepared ethylene chlorhydrin by saturating ethylene glycol with hydrochloric acid gas and allowing it to stand for several hours. Ladenburg^ modified the methods which had previously been used and placed the glycol in a distilling bulb and passed dry hydrochloric acid gas through it at a temperature of 148 degrees. After the reaction started he raised the temperature to about 160 degrees and collected the distillate which contained water, chlor- hydrin, hydrochloric acid gas, etc. He extracted the distillate with ether which removed the ethylene chlorhydrin and dried it over potassium carbonate. He evaporated off the ether when the product was dry and distilled the residue. He redistilled it a second time in order to purify it, obtaining a yield of 60 $. Ethylene chlorhydrin has a boiling point of 128 degrees (Wurtz). At 8 degrees its specific gravity is 1.24. It is mis- cible with water in all proportions. Unon reduction and hydrolysis it gives hydrochloric acid and alcohol. 2. Trimethylene chlorhydrin*- Trimethylene glycol from which trimethylene chlorhydrin is made was first prepared in 1871 from trimethylene bromide by Germont n< . The bromide treated with gla- . • ‘ . t . J. • •] ■ ;:•••• ' ■. «?:■ ■; , • ; • . ! , r • ' / -G- cial acetic acid and silver acetate yielded the diacetate which boiled at 203 degrees. This compound was treated with barium hy- droxide and yielded a compound boiling from 208-218 degrees. Analy- sis and reactions proved it to be trimethylene glycol. Reboul^ also prepared it in 1874 by the same method giving the boiling point as 216 degrees. By accident Freund"^, while trying to prepare normal butyl alcohol by fermentation of glycerol, found that some other com- pound was being formed in considerable quantity. This compound when distilled with steam and then redistilled proved to be pure trimethylene glycol. The most important method is that discovered by A. A. Noyes and W. H. Watkins 7 in 1895. A soart concern near Boston was having trouble getting glycerol of the proper specific gravity. The cause was attributed to some impurity. They sent some of the light mater ial to Noyes and Watkins who subjected it to fractional distillation They found that it contained 38 $ trimethylene glycol. Another method is that of Niederist^. His method was to treat trimethylene bromide with moist silver oxide. The method that is most extensively used, in fact it is almost entirely used, is the fermentation method. It is the most inexpensive and the best yields are obtained by it. The only use of trimethylene glycol in this problem is its use as the main product in the preparation of trimethylene chlorhydrin. Trimethvlene chlorhydrin was prepared first by Reboul 9 in 1774 by heating trimethylene glycol saturated with hydrochloric acid gas in sealed glass tubes at a temperature of 100 degrees for several hours . When the mixture was removed from the tubes and * « • - ' ' • >' . •> . /t rtvl ' . • . ! ■ 1 • ■ • r ; r • submitted to fractional distillation it was found that both tri- methylene chloride and trimethylene chlorhydrin were formed. They were separated by fractional distillation. The chlorhydrin method of Reboull°> as it is called, is the method that has been used previously. It consists in passing dry hydrochloric acid gas through trimethylene glycol in the cold. His method was to pass the gas through for 16 hours and then fraction- ally distill the resulting product. By this method 50 $ yields were claimed but the writer followed it out carefully and was never able to obtain more than a 40 ^ yield. This method of preparation was used by Derrick and Bissell 11 but was modified slightly, in the time especially in which they allowed the gas to pass through the glycol. Into 130 grains of the glycol they passed hydrochloric acid gas for three hours and then fractionally distilled the product. They obtained a 25 gram yield. Trimethylene chlorhydrin boils at 160 degrees and has a spec- ific gravity of 1.123 at 17 degrees and is only slightly soluble in water • The best method for the preparation of trimethylene chlor- hydrin is a method which is similar to the method used by Reboul^ 0 for the preparation of ethylene chlorhydrin. It is the method used here in the University of Illinois laboratories and has been im- proved by the writer. It will be described in full in the theor- etical and experimental part of this paper. It eliminates the seal- ed tube method which is a great advantage over most of the other methods used. The corresponding bromhydrin has also been prepared by Fruhlingl^ i n 1882. A mixture of 100 parts of trimethylene glycol and 100 parts of hydrobromic acid (48$) was saturated with hydro- ' -- * - fc . - 8 - broraic acid gas and heated from 4-5 hours on a water bath. Both trimethylene bromide and trimethylene bromhydrin were formed. The bromhydrin is slightly soluble in water while the bromide is not and a complete separation may be effected by repeated washing with water. The wash water was neutralized with sodium hydroxide and shaken with ether. The ether was distilled off on the water bath and the residue distilled at a higher temperature under diminished temperature. The product distilled at from 98-112 degrees under a pressure of about 180 mm. .and has a specific gravity of 1.5304 at 20 degrees. It is soluble in six parts of cold water. 3. Trimethylene iodhydrin. Trimethylene iodhydrin was pre- pared by Henry* 3 in 1879 from the chlorhydrin. A solution of the chlorhydrin in methyl alcohol was treated with sodium iodide and the mixture allowed to stand. Afterwards the product was distilled and the iodhydrin was found to boil at 225 degrees without decom- position and had a specific gravity of 2.349 at 13 degrees. It had a sweet smell and a sharp taste. It is soluble in water al- cohol and ether. The writer has prepared the iodhydrin by the same method except that acetone was used as a solvent instead of methyl alcohol because of the fact that sodium chloride is insol- uble in acetone while the sodium iodide is very soluble. 4. Trimethylene chloracetate . Derrick and Bissell 14 prepared this compound to use in the preparation of trimethylene oxide. Their method was to take 300 grams of the chlorhydrin in a flask and allow 250 grams of acetyl chloride to drip slowly on the chlor- hydrin. The flask was then fitted with a condenser for downward distillation and the mixture distilled. 375 grams of trimethylene chloracetate were obtained. It is a clear colorless liquid with a sharp o dor and a boiling point of 164 degrees . ♦ a-tn - . - . < •- -9- 5. Trimethylene iodacetate. Trimethylene iodacetate was pre- pared hy Henry* 5 in the same manner as the iodhydrin. He dissolved the corresponding chlor compound in methyl alcohol and added sod- ium iodide. The iodine replaces the chlorine just as in the other cases . The iodacetate is also prepared hy the same method used in the preparation of the corresponding chloro compound. It was used hy Ladenburg 15 in 1883. It has a boiling point of 207-210 degrees. 6. Phenoxy-propyl-bromide . Phenoxy-propyl-hromide was pre- pared hy Lohmann* 7 in 1891. His method was to take one mole of phenol, one mole of trimethylene bromide, and about two moles of sodium ethylate and reflux on a steam hath for a few hours. After distilling off the alcohol he distilled the residue under reduced pressure and obtained the desired product in good yields. The same method has been used here at the University of Ill- inois and better results have been obtained by varying conditions. The method of Wohl and Berthold 18 for preparing phenoxy-ethyl- bromide was used by the writer and excellent results were obtained. This method is to take a large flask fitted with a reflux condenser and place in it 100 grams of phenol, 125 grams of trimethylene bromide, and 1000 grams of water. Then dissolve a slight excess of sodium hydroxide in 250 cc . of water and add this to the solution in 50 cc. portions every half hour and then reflux the mixture two hours longer before removing the flame. There are two layers and the heavy one contains the compound desired. Separate the two layers in a separatory funnel and distill under reduced pressure. The excess trimethylene bromide can all be recovered. 7. Phenoxy-propyl-diethyl-malonate . This compound has only L . r\ ‘ ' • ' 't ■ ... . - t ■ * ■ • ; ' - 10 - 1 Q "been prepared by Gabriel . He took 34 grams of raalonic ester, 34 grams phenoxy-propyl-chloride , 8 grams of sodium and dissolved the whole mixture in absolute alcohol dissolving the sodium first. He refluxed this mixture on the water bath for a few hours and then distilled off the alcohol, added water to dissolve the sodium chloride in which his product was an insoluble oil. He separated the two by means of a separatory funnel and obtained a yield of 45 grams of the crude ester without distillation. Gabriel then sapon- ified the crude ester and obtained upon neutralization the dicar- boxylic acid. He obtained a 20 gram yield of the dicarboxylic acid He then heated this acid and by driving off carbon dioxide he ob- tained the mono-carboxylic acid. It crystallizes from ligroie in nice white crystals which have a melting point of 64 degrees. From this compound he obtained the delta hydroxy acid by splitting off the phenyl group with hydrochloric acid in sealed tubes and then treating the chlor compound with silver oxide and water. 8. Normal-phenyl-octyl -ether . Perkins 20 prepared normal- phenyl-octyl-ether in 1896 by the following method. He took one mole of octyl iodide, which had been prepared from octyl alcohol, with three moles of phenol and one mole of sodium hydroxide with a little water and placed the substances in a Wurtz flask and heated until he thought the reaction complete. He then removed the con- tents and distilled. He obtained normal-phenyl-octyl-ether. It is a solid at 8 degrees and boils at 285 degrees under atmospheric pressure . 9. Normal octyl bromide. Theodore Zincke21 prepared octyl bromide by treating octyl alcohol with hydrobromic acid gas. The gas is passed through the alcohol and the reaction takes place. . ' 1 1 'a ■ • , ■ ■ - 11 - The bromide has a boiling point of 198 degrees. 10. Normal octyl alcohol. The first and only method of ob- taining octyl alcohol was from nature. Branchimont and Zincke 22 found it in combination with acetic acid and butyl alcohol in the plant Heracleum Spondylium and separated them by fractional dis- tillation. It has a boiling point of 195.5 degrees. THEORETICAL. The first compound to be prepared in the synthesis of hydroxy acids is the condensation product of one of the chlorhydrins , for example ethylene chlorhydrin, with sodium raalonic ester. The sod- ium malonic ester is prepared by the treatment of malonic ester with sodium dissolved in absolute alcohol. The condensation should take place according to the following reaction COOC^H* COOCxHiT HO-CH a -CH x -Cl-f Na-C-II > HO-CH^-CH^-C-H + NaCl COOCaH* C00C a H^ It was observed that the theoretical amount of sodium chloride was obtained but when the condensation product was distilled under re- duced pressure that only about 15 $ of the theoretical yield of the hydroxy ester was obtained. This yield is about the same as that Fittig and Roder 2 ^ when they made the same product by treating sodium malonic ester with ethylene oxide. The general method for running malonic ester syntheses is to reflux the mixture until neutral. In making these runs however the time was varied with apparently no change whatever in the results. A list of the substances used in these attempted preparations is as follows:- ethylene chlorhydrin, trimethylene chlorhydrin, chlor ethyl acetate, ethylene bromide, trimethylene bromide, tri- methylene iodhydrin, trimethylene chloracetate , trimethylene iod- - • I > '• •' ' - 12 - acetate, and phenoxy propyl bromide. Of these only the first and last gave any yields at all, the first giving a yield of never more than 15 and the last giving as high as a 63 ^ yield. It was quite interesting to note however what happened when trimethyl- ene chloracetate and trimethylene iodacetate were used. In the process of distillation under reduced pressure every time, when the temperature had reached about 140 degrees under 20 mm., the contents of the flask became a deep red gel which would expand and fill the flask and not distill at all. It was not identified. Since the phenoxy propyl bromide would react giving good yields the work of Gabriel was repeated as far as the following re- actions indicate. COOC* H s 0000*115 (1) O'OCHj.-CIU-CHi-Br Na-C-H >0OCH Z -CH Z -CIV-C-H + NaBr 0000*11$- C00C*IV 0000*115* COOH (2) 0"QCH* -CH^-CH^-C-H 0OCII Z -CH Z -CH -C-H COO C* Hi" COOH (3) COOH 0OCH Z -CH^ -CH Z -C-H 6 oo h heat 0 OCH* -cii^-ch^-c^-cooh M.P. 64 degrees. Because it was desired to observe some more reactions with the same compound the writer did not follow Gabriel’s work any farther but attempted the preparation of normal oc£yl alcohol by the following series of reactions. COOCJV (1) 0OCH Z -CH Z -CH, -Br Na-C-CH* -CH*-CH Z -CH, > COOC* H r COOC H ^0CH z -CH* -CH* -C-CH^-CH* -CH, -OH 3 -f NaBr COOC H This compound upon saponification gives the dicarboxylic acid and we have, V • t: - -13- COOH fuse (2) OOCH^-CH.-CH.-C-CH.-CH^CH.-CH^ soda _ llme QOCH^ ( CH^ ) t CH , (3) QOCH (CH ) CH HBr (48$)— > Br-CIL^CI^ ) 6 CH^ ~h CJI^OH (4) Br-CH^ -CI^-CH ,-CH^-CH* -CH i -CII x -CH 3 -A& 9 . ,_j> Hj 0 ( IIO -CH^ -CH 2 -CH * -CH^ -CH ^ -CII ^ -CH a -CH 3 For lack of time the last step was not run at all and the pre- paration of the bromide was tried only once . The first reaction was carried out just exactly the same as with malonic ester using butyl malonic ester instead. A yield of 63 $ of the calculated amount was obtained. In the fusion with soda-lime to drive off carbon dioxide it was interesting to note that the phenyl group was removed also and some ocfcyl alcohol was obtained at this point. EXPERIMENTAL. 1. ^-hydroxy ethyl diethyl malonate. The method used in the preparation of this compound is the general method given by W. A. Noyes for carrying out a synthesis with malonic ester. To ion cc. of absolute alcohol in a round bottom flask fitted with a reflux condenser add 5.8 grams of metallic sodium and allow the mixture to cool, then add 40 grams of malonic ester and shake well. After shaking add through the condenser 20 grams of dry ethylene chlorhydrin, place on a steam bath and reflux until neu- tral. The time required for this is about three hours. Now change the condenser and arrange for downward distillation. Dis- till off the alcohol. This requires about three hours. Pour into the flask enough water to dissolve the sodium chloride and separate the two resulting layers in a separatory funnel. The brown oily layer contains the product desired. Disregard the water portion. - . , -14- Distill the oil under reduced pressure. There are two main frac- tions one of which is malonic ester which distills from 120-130 de- grees under 25 mm. of pressure, the other is the ester which dis- tills from 170-180 degrees under this pressure. The yields vary but the best obtained was 15 $ of the calculated amount . Refluxing for a longer time did not change the yield. Neither an excess of malonic ester nor of ethylene chlorhydrin changes the per cent yield. The condensation with trimethylene chlorhydrin, trimethy- lene iodhydrin, trimethylene iodacetate, trimethylene chloracet- ate, chlorethyl acetate, trimethylene bromide and ethylene bromide all gave practically negative results. There was always a malonic ester fraction and a varied amount of tar. 2. Trimethylene chlorhydrin. A hydrochloric acid generator is made in the following manner. A five liter round bottom flask is fitted with a two hole rubber stopper through which are inserted a dropping funnel and an outlet tube. In the dropping funnel is placed commercial sulfuric acid. In the flask are placed 4 pounds of salt and about two liters of 37 $ hydrochloric acid. A 100 cc. wide mouth round bottom flask is fitted with a four hole rubber stopper through which are inserted a dropping funnel, a thermometer reaching nearly to the bottom of the flask, one glass tube reaching to the bottom of the flask, and another glass tube extending just through the stopper is connected to a water condenser. This small flask is clamped to a ring stand and 25 cc. (never more than 40 cc.) of trimethylene glycol are run into it. About one third of the bulb is immersed in an oil bath and the bath heated until the thermometer reading is 160-180 degrees . The tube extending to the bottom of the small flask is con- nected to the hydrochloric acid generator and a very rapid stream . -15- of hydrochloric acid gas is passed through the glycol. Trimethy- lene chlorhydrin, water, hydrochloric acid, and some side re- action products begin to distill over and are collected; excess hydrochloric acid gas being absorbed by the water. As rapidly as the glycol is used it is replenished from the dropping funnel, al- ways keeping the amount in the flask constant. The process is con- tinuous and as much material may be run through as is desired with- out changing the apparatus, except to replenish the hydrochloric acid and salt in the generator. It has been found as the following table shows that one kilo of glycol may be run through this apparatus in four hours, at which rate best results are obtained. Place the distillate on a steam bath and allow all the hydro- chloric aeid gas to be distilled off that can be. About one hour is required for this. Afterward place it in a flask with a frac- tionating column and carefully distill, using pieces of clay to prevent bumping. The fraction distilling up to 158 degrees is collected in a separate container. The part distilling from 158- 164 degrees is good chlorhydrin and is not treated any further. The re xt fraction boiling from 164-180 degrees is collected and re- distilled once, collecting the fraction from 158-164 degrees as above. The fraction distilling above 164 degrees is combined with the original high boiling fraction, and the whole is distilled collecting what comes over below 220 degrees. This is mainly tri- methylene glycol with some chlorhydrin in it. This fraction may be run again and good yields obtained from it. Above 220 degrees there is a high boiling bar which will not distill (possibly high boiling ethers). The first fraction contains water and about 50 trimethylene chlorhydrin. This distillate is neutralized with ’ fc , - -16- commercial sodium carbonate and the chlorhydrin separates out. It is then poured off and distilled, a small amount of low foiling product "being obtained. By running the fraction from 164-220 degrees a second time and treating the distillate as described above a 70 $ yield was obtain- ed. Glycol grams H SO lbs . Salt lbs. HC1 lbs Chlor- hydrin grams . Recov. Glycol grams Low- boiling f rat . Tar grams ~ ■ Time hrs . Times fract- ionated. 190 4 5 2 120 11 80 31 2 5 440 291 25 50 50 5 3 220 148 15 30 45 3 3 200 138 24 20 50 2 3 210 4 5 2 158 10 50 2 2 1000 12 10 6 876 110 55 210 7 2 500 7 5 5 487 30 95 2 2 500 8 5 5 378 105 10 85 2 ' 2 The above table shows a record of the runs made. The seventh and eighth runs were the best giving the highest yields. This is due to the fact that a very rapid stream of hydrochloric acid gas is passed through the glycol at a temperature of 160-180 degrees. 3. Trimethylene iodhydrin. This is prepared by the method of Henry with the modification that acetone is used for a solvent in- stead of methyl alcohol. Put 400 cc. of acetone in a liter round bottom flask. Dis- solved one mole of sodium iodide in the acetone. To this solution . -17- is added one mole of trimethylene chlorhydrin and the whole is al- lowed to stand over night. Sodium chloride is filtered off, the acetone distilled off and the final product distilled under reduc- ed pressure. A slightly brown product was obtained. The yield was 45 of the theoretical amount. 4. Trimethylene chloracetate . In a round bottom flask is placed one mole of trimethylene chlorhydrin and to it is added in course of half an hour one mole of acetic anhydride. Add in small portions and shake well. After all the acetic anhydride has been added fit the flask with a condenser for downward distillation and distill. The chloracetate comes over at 164 degrees in good yields As high as 95 $ of the calculated amount is obtained by this pro- ceedure . 5. Trimethylene iodacetate. 100 grams of trimethylene chlor- acetate are added to a solution of sodium iodide in acetone and the mixture allowed to stand over night. The sodium chloride is then filtered off and the acetone distilled off on the steam bath. The residue is distilled under reduced pressure. A product is obtained which boils at 208-211 degrees under atmospheric pressure. This is trimethylene iodacetate. The yield is 53 ^ of the theoretical amount . 6. Phenoxy propyl bromide. The method of V/ald and Berthold is used in the preparation of phenoxy propyl bromide. It is the same method used for the preparation of phenoxy ethyl bromide. A five liter flask is fitted with an upright condenser and a dropp- ing funnel. In the flask is placed 1500 cc . of water 700 grams of trimethylene bromide and 310 grams of phenol 120 grams of solid sodium hydroxide is dissolved in 500 cc of water and this placed in the dropping funnel. The contents of the flask are raised to boil- - I . * . -18- ing and at about 40 minute intervals 25-50 cc . of* the sodium hy- droxide solution is admitted to the flask. After all the sodium hydroxide solution is added the contents are refluxed for two hours longer. There are two layers in the flask and when this is allowed to cool they are separated in a separatory funnel and the heavy layer which contains the phenoxy propyl bromide is distilled under reduced pressure. The water layer containing the sodium bromide is neglected. The distillation is repeated once and the phenoxy pro- pyl bromide comes over at 180 degrees under 25 mm. of pressure. 250 grams of trimethylene bromide may be recovered in some cases. Calculating on the trimethylene bromide which enters into the re- action 87