Patented Aug. 22, 1939 2,170,624

UNITED STATES PATENT OFFICE 2,170,624 PROCESS FOR THE PRE PARATION OF PENTAERYTHIRTOL Joseph A. Wyler, Allentown, Pa., assignor to Tro jan Powder Company, Allentown, Pa. No Drawing. Application January 11, 1938, Serial No. 184,457 5 Claims. (C. 260-635) This invention relates to the preparation of formation of pentaerythritol, and that by the pentaerythritol and particularly relates to a elimination of methyl alcohol from the reacting process for the preparation of pentaerythritol mixture improved yields of pentaerythritol, and from methanol-free formaldehyde and acetalde improvement in the quality of crude penta hyde. The principal object of my invention is to erythritol may be obtained, provide an improved process for the preparation When formaldehyde and acetaldehyde react, in of pentaerythritol. alkaline medium, to form pentaerythritol the re It has been known for more than forty years action taking place is usually expressed by the that formaldehyde and acetaldehyde react, in the following equation: 0 presence of an alkali to form pentaerythritol, and this knowledge has led to the development of numerous processes for the preparation of this where, M is a monovalent metal or one half of a substance. In these various processes, the effect divalent metal. of Such factors as temperature at Which the re This simple equation is, however, wholly in 5 action is carried out, concentration of reactants, adequate in that it does not convey any concep " duration of the period of reaction, treatment of tion of the types of the reactions which occur the mixture after completion of the condensa during the formation of the pentaerythritol, and tion, etc. are disclosed. consequently, the importance of the presence of My invention rests upon my discovery that catalytic agents and the absence of anti-catalytic methyl alcohol acts as a disturbing factor in the agents (which anti-catalytic agents are normally reaction of formaldehyde and acetaldehyde to present in the formaldehyde) has never been dis form pentaerythritol, and that in the absence of closed. methyl alcohol the reaction between formalde In the commercial preparation of pentaerythri hyde and acetaldehyde proceeds with greater tol, 4 to 5 molecules of CH2O are used for each 25 ease, and goes further toward completion with molecule of CH3CHO. These are caused to react the production of an increased yield of penta in an alkaline medium at 20-50° C. and in such a erythritol based upon the quantities of formalde Volume of Water as to provide a solution to con hyde and acetaldehyde used, than is the case tain less than 20% of aldehydes at the beginning when methyl alcohol is present in the reacting of the reaction. The alkalinity is usually ob 30 mixture. The significance of this observation will tained by the use of Ca(OH)2, Ba(OH)2 or NaOH. be apparent when it is understood that com After the condensation, the pentaerythritol mercial formaldehyde is always made by the must be freed from a metal formate and other Oxidation of methyl alcohol, and that all com products. Usually this is done by treating the mercial formaldehyde contains from 3% or 4% mixture after the condensation reaction, with 35 up to as much as 15% of methyl alcohol, as an Sulfuric or oxalic acid, or both, to pecipitate the impurity. Up to the time of the present work, Cal or Ba ions as Sulfate or oxalate. The in it was never known that this small amount of Soluble Ca or Ba salts are separated by filtration, methyl alcohol present as an impurity in form leaving the pentaerythritol, formic acid and vari aldehyde was in any way more harmful than Ous by-products in solution. This solution is sub 40 the amount of water that is also present in form jected to repeated evaporations and crystalliza 40 aldehyde, and which forms merely a harmless tions in order to separate the pentaerythritol in diluent. I have discovered, however, that the pure condition. former conception of the harmless character of Among the by-products formed are dipenta methyl alcohol is in error, and by removing erythritol, formose, aldol, possibly arabino-ketose, 45 methyl alcohol from formaldehyde, (down to a other compounds possessing primary and/or sec 45 limit of A of 1%, below which I have so far not Ondary alcohol groups, a substance of the prob been able to go) I have obtained yields of penta able composition of erythritol distinctly greater than those obtainable by the use of formaldehyde containing methyl (CHOH), ceHo CEC(CHOH), 50 alcohol, and very materially greater yields than C.O.C. can be explained by the mere removal of the acetals, etc. It is the formation of these which methyl alcohol when considered as an inactive causes the relatively low yields of pentaerythri diluent. In other words, I have discovered that tol-yields of 50-65% of theory-in the present methyl alcohol is actively harmful in the reaction commercial processes. 55 between formaldehyde and acetaldehyde in the Now, it is the purpose of my invention to pro SS 2 2,170,624 vide a process for the preparation of pentaeryth I do not limit myself to these specific details. ritol in which the yields are above 70% of I may, for instance, use a temperature as low as theory, and usually about 80%. s 10° C. and as high as 50° C.; a concentration of I accomplish this by using a formaldehyde aldehydes as low as 3% and as high as 20%; a which is either free from or contains less than proportion of CHO to CH3CHO of from 4:1 to 2 parts of CH3OH per 40 parts of CH2O. I have as high as 5:1; a rate of addition of lime cor found that the presence of CH3OH in the reac responding to as low a period of addition as one tion mixture producing pentaerythritol has a di hour and to as high a period of addition as rective effect upon the course of the reaction twenty hours; and I may use, in place of lime, between formaldehyde and acetaldehyde. If a any other suitable base-the essential charac O very large amount of CH3OH be present in the teristic of my process being that the reaction O reaction mixture above referred to, the yield of mixture is free from methanol or essentially so. pentaerythritol may be practically nil-a syrupy The other steps are well known in the art. product is all that forms. Whereas, if methyl It will therefore be clear that my invention alcohol be absent a crystalline product (mainly comprehends a process for making pentaerythri 5 pentaerythritol) of excellent color results. tol which constitutes a distinct improvement over 5 My invention does not refer to processes for present processes in that il, enables one to obtain making a methanol-free formaldehyde, but is re greater yields of pentaerythritol. I do not limit lated to the use of this type of formaldehyde, the Scope of my invention, except as indicated in regardless of how it is made, in the preparation the appended claims. of pentaerythritol. I claim: 20 As an example of how I may carry out my 1. The process for the preparation of penta process, I give the following: erythritol comprising the alkaline condensation 396 pa is of 30.34% formaldehyde (by weight), of acetaldehyde with formaldehyde essentially which is free from methanol, are mixed with 83 free from methanol, in aqueous solution. parts of 52.8% (by weight) of acetaldehyde and 2. The process for the preparation of penta 25 900 parts (by weight) of water. The mixture is erythritol comprising the alkaline condensation stirred in a suitable container and 35 parts of of formaldehyde and acetaldehyde in aqueous lime (CaO) in the form of a slurry, formed with Solution, said solution containing not more than 30 the aid of 166 parts of Water, are slowly added 2 parts by weight of CH3OH to 40 parts by weight during the course of four to five hours, maintain of CH2O. 30 ing the temperature of the reaction mixture at 3. The process for the preparation of penta about room temperature until tests indicate the erythritol comprising the condensation of form Substantial absence of CH2O and of CH3CHO. aldehyde and acetaldehyde in aqueous solution in A typical end point for the reaction is one indi the presence of calcium hydroxide; said solution cating 0.09% aldehydes present, as determined being essentially free from methanol. With iodine Solution. 4. The process for the preparation of penta The reaction being completed, it is necessary erythritol comprising the condensation of form to remove the lime in order to obtain the pure aldehyde and acetaldehyde in aqueous solution pentaerythritol. For this purpose, the mixture in the presence of calcium hydroxide; said solu 40 is treated with slightly less sulfuric acid than is tion containing not more than 2 parts by weight 40 theoretically equivalent to the lime used. This of CH3OH to 40 parts by weight of formaldehyde. precipitates the calcium as CaSO4. In order to 5. The process for the preparation of penta precipitate all of the lime Out of the solution, a erythritol comprising the condensation of form Small amount of Oxalic acid is added in the form . aldehyde and acetaldehyde in aqueous solution 45 of a Water Solution and the mixture stirred for in the presence of calcium hydroxide; said solu 45 one hour or more. The latter is then filtered, tion being essentially free from methanol, and the filtrate evaporated and the pentaerythritoi containing not more than 20% (by weight) of recovered by crystallization and washing. The combined aldehyde, not less than 4 molecules of total weight of dry crystals obtained is usually formaldehyde to 1 molecule of acetaldehyde and 50 about 95-105 parts. The remaining portion not more than 5 molecules of formaldehyde to 1 50 (after evaporation) is a Syrup. molecule of acetaldehyde. In the above example I ment?on certain specific operating conditions but I wish to emphasize that JOSEPH A WYER.