US2771471A - Polycarboxy acid esters of oxypropylated hexides - Google Patents

Description

United States PatentO POLYCARBOXY ACID ESTERS F OXYPROPYLATED HEXIDES Melvin De Groote, St. Louis, Mo., assignor'to Petrolite Corporation, a corporation of Delaware No Drawing. Original application December 1, 1950,

Serial No. 198,756. Divided and this application December 2, 1952, Serial No. 323,732 i 8 Claims. (Cl. 260-3474) Practically, the hexides are derived from: hexitols, by an anhydridizing reaction, in consequence of which the hexides are named by reference to the hexitols from which they can be formed, by changing the suflix itol to ide. Thus mannitol yields mannide, sorbitol yields sorbide, dulcitol yields dulcide, etc. Each of the hexitols is capable of forming a number of isomeric hexides. In this application the terms mannide, sorbide, etc., are employed in the generic sense to cover all isomeric hexides formed from the respective hexitols. See U. S. Patent No. 2,387,842 to Soltzberg, dated October 30, 1945.

The hexides are appropriately the dihydroxylated ultimate anhydrides obtained from hexitols which may go through an intermediate stage involving the formation of compounds having 4 hydroxyls such as sorbitan, mannitan and dulcitan. Re-stating what has been said previously in more elaborate manner, attention is directed to U. S. Patent No. 2,322,821 to Brown, dated June 29, 1943. The following text is a substantially verbatim excerpt.

The formation of the ethers from the polyhydric alcohols involves dehydration which may be accomplished by heating and driving off water, The splitting off of water from 2 hydroxyls attached to polyhydric alcohols can proceed both internally and externally. In internal etherification where the two hydroxyls are both attached to the same molecule of polyhydric alcohol, formation. of the ether link leads to oxide or carbon-oxygen rings. Where the disposition of the hydroxyl groups allowsthe formation of rings having four to seven members, as is the case with the hexitols, then internal etherification toproduce cyclic internal ethers is, in general, the preferred reaction.

This internal ether formation can proceed in two stages I with the formation of one or two carbon-oxygen rings giving rise to monoor dianhydro products. In the case of hexitols these may be called generally hexitans and hexides; for example, sorbitan, mannitan, and dulcitan,

andlsorbide,'mannide, and dulcide. Examples of such possible ring structures are given herewithz" 'HQCHMOHOH .l r

' 0H, cHoHoHomoH "ice o Honk-onoH,

| +H1O H1 011- HOH In addition to internal'etherification, external etherification can take place. Since'this is a bimolecular reaction involving 2 molecules instead of the one molecule concerned in. internal etherification, this external etherification tends'to proceed to a lesser extent andbecomes of importance principally when the internal condensation becomes difiicult or impossible.

Specific reference is made to the two aforementioned patents as to appropriate procedures for producing sorbideymannide, or dulcide, all of which may be "considered as examples of suitable hexides.

Iffor convenience the hexide is indicated thus:

the product obtained by oxypropylation may be indicated thus:

H(OCsHs)1tOR'O(C3HeO)n'I-I with the proviso that n andn' represent whole numbers which added together equal a sum varying from 15 to 80,

and the products "of the present invention, i. e.,-the acidic esters'obtained by reaction of the polycarboxy acid, may

be indicated thus:

.of the polycarboxy acid doon R\ N and preferably free from any radicals'having more than v8, uninterrupted carbon atoms in a single group, and with the further proviso that the parent diol prior to esterification be'water-ins'oluble and kerosene-soluble. i

The products of this invention have particular value ,as demulsifying agents in a, process for resolving petroleum emulsions of ,the .waterdn-oil type that are commonly referred to as cut oil, roily oil, emulsified oil, etc., and which comprise finedroplets of naturally-occurring waters or brines dispersed in a more or less permanent state throughout the oil which constitutes the continuous phase of the emulsion. A process for resolving petroleum emulsions of. the water-in-oil type which utilizes the ,products described herein is described and claimed in my copending application Serial No. 198,756 filed December 1, 1950. I,

The products are also useful in a process for separating emulsions which have been prepared .under controlled conditions from mineral oil, such as crude oil and relatively soft waters or .weak brines. Controlled emulsificatipn and subsequent deinulsification under the conditions just mentioned are of significant value in removing impurities particularly inorganic salts from pipeline oil. They are alsouseful for other purposes, such as stabilizing emulsions, as Spreaders in the application of asphalt in road building and the like, as flotation reagents, as lubricants, etc.

Attention is'directed tothe C. M. Blair, I r. Patent No. 2,562,878, dated August 7, 1951, the application for which was copending with my copending application Serial No. l98,756'noted above and in which there is described, among otherthings, a process for breaking petroleum emulsions of the water-in-oil type characterized by subjecting the emulsion to [the action ofan esterification product of a dic'arboxylicacid and a polyalkylene glycol in which the ratio of equivalents of polybasic acid to equivalents, of'polyalkylene glycol is in the range of 0.5 to 2.0, in which the al kylene group has-from 2 to 3 carbon atoms, and in which the molecular weight of the product is between 1,500 and 4,000.

Similarly, there have been used esters of dicarboxy acids and polypropylene glycols in which 2 moles of the dicarboxy acid ester have been reacted with one mole of apolypropylene glycol having a.molecularweight, for ,e,xa mp,le, of 2,000 sq: asrto forrn an acidic fractional $50 1 Examination of what. is saidsubsequently herein as well as, the hereto. appended claims in comparison with ,theprevious example willshow the line of delineation between such somewhat comparable compounds. Of greater significance, owever, is what is said subsequently in, regard to the structure of the parent diol as compared to polypropylene glycols whose rnolecular weights may vary 1,000 to 2,000.

r.-P Yn en9 ha saidz er ina e l edivided into four parts:

Part 1 is concerned with the preparation of the oxypropylation derivatives of ,the dihydroxylated hexide;

Part 2 is concerned with the preparation of the esters from the oxypropylated derivative;

Part 3 is concerned with a consideration of the structure; of the diols, which is of significance in light of what is, saidsubsequently; and p Part 4 is concerned with, certain derivatives which can beobtainedfrom the oxypropylated diols. In some instances, such derivatives are obtained. by modest oxyethylation preceding the oxypropylation step, or oxypropylation followed by oxyethylation. This results in diols having somewhat different properties which can then be reacted with the same polycarboxy acids or anhydrides described in Part 2. For this reason a description of the apparatus makes casual mention of oxyethylation. For the same reason there is brief mention of the use of glycide.

PART 1 Fora number of well known reasons equipment, whether laboratory size, semi-pilot plant size, pilot plant particular alkylene oxide.

termined andset range, for instance, 95 to 120 C., and

size, or large scale size, is not as a rule designed for a invariably and inevitably, however, or particularly in the case of laboratory equipment and pilot plant size the design is such as to use any of the customarily avilable alkylene oxides, i. e., ethylene oxide, propylene oxide, butylene oxide, glycide, epicholorhydrin, styrene oxide, etc. In the subsequent description of the equipment it becomes obvious that it is adapted for oxyethylation as well as oxypropylation.

Oxypropylations are conducted under av wide variety of conditions, not only-in regard-to presence or absence of catalyst, and the kind of catalyst, but also in regard to the time of reaction, temperature of reaction, speed of reaction, pressure during reaction, etc. For instance, oxyalkylations can be conducted at temperatures up to approximately 200 C. with pressures in about the same range up to about 200 pounds per square inch. They can be conducted also at temperatures approximating the boilingpoint of water or slightly above, asfor example to'120" C. Under such circumstances the'pressure will beless than 30 pounds per square inch unless some special procedure is employed asis sometimes the case, to wit,

keeping an atmosphere of inert gas such as nitrogen in the vessel during "the reaction. Such low-temperaturel'ow reaction rate oxypropylations have been described very completely in US. Patent No. 2,448,664 to H. R. Fife,et al., dated September 7, 1948; Low temperature,

low pressure oxypropylations are particularly desirable where the compound being subjected to oxypropylation contains one, two or three points of reaction only, such as monohydric alcohols, glycols and triols.

Since low pressure-low temperature-low reaction speed oxypropylations require considerable time, for instance, 1 to 7 days of 24 hours each to complete the reaction they are conducted as a rule whether on a laboratory scale, pilot plant scale, or large scale, so as to operate automatically. The prior figure of seven days applies especially to large-scale operations. I have used conventional equipment with two added automatic features; (a) a solenoid controlled valve which'shuts oil? the propylene (b) another solenoid valve which shuts off the propylene oxide; (or for, that matter ethylene oxide if it is being 'used) if the pressure gets beyond a predetermined range,

such as 25 to 35 pounds. Otherwise, the equipment is substantially the same as is commonly employed for this purpose where the pressure of reaction is higher, speed of reaction is higher, and time of reaction is much shorter. In'such instances such automatic controls are not necessarily used.

Thus, in preparing the various low temperature oxypropylation examplcsI have found it particularly advantageous to use laboratory equipment or pilot plant which is designed to permit continuous oxyalkylation whether it be oxypropylation or oxyethylation. With certain obvious changes the equipment can be used also to permit oxyalkylation involving the use of glycide where no pressure is involved'except the vapor pressure of a solvent, if any, whichmay have been used as a diluent. 1

As previously pointed out the method of using propylene oxide is thesame as ethylene oxide. This point is emphasized only forthe reason that the apparatus is so designed and constructed as to'use either oxide;

' The oxypropylation procedure employed in the preparation of the oxyalkyl'ated derivatives has been uniformly the same, particularly in light of the fact that a continuapproximately 15 gallons and a working pressure of one "thousand-"pounds gauge pressure.

This pressure obviously is far beyond any requirement as far as propylene "oxidegoes unlessthere is a-reaction of explosive violence involved due to accident. The autoclave was equipped with the conventional devices and openings, such as the variable-speed stirrer operating ,at speeds from 50 R. P. M. to 500 R. P. M.; thermometer well andthermocouple for mechanical thermometer; emptying outlet; pressure gauge, manual vent line; charge hole for initial reactants; at least one connection for introducing the alkylene oxide, such as propylene oxide or ethylene oxide, to the bottomof the autoclave; along with suitable devices for both cooling and heating the :autoclave, such as a cooling jacket; and, preferably, coils in addition thereto, with the jacket so arranged that it is suitable for heating with steam or cooling with'water and further equipped with electrical heating devices. Such autoclaves are, of course, in essence small-scale replicas of the usual conventional autoclave used in oxyalkylation procedures. In some instances in exploratory preparations an autoclave having a smaller capacity, for instance, approximately 3% liters in one case and about 1% gallons in another case, was used.

Continuous operation, or substantially continuous operation, was achieved by the use of a separate container to hold the alkylene oxide being employed, particularly propylene oxide. In conjunction with the smaller autoclaves, the container consists essentially of a laboratory bomb having a capacity of about one-half gallon, or somewhat in excess thereof. Insome instances a larger bomb was used, to wit, one having a capacity of about one gallon. This bomb was equipped, also, with an inlet for charging, and an eductor tube going to the bottom of the container so as to permit discharging of alkylene oxide in the liquid phase to the autoclave. A bomb having a capacity of about 60 pounds was used in connection with the IS-gallon autoclave. Other conventional equipment consists, of course, of the rupture disc, pressure gauge, sight feed glass, thermometer connection for nitrogen for pressuring bomb, etc. The bomb was placed on a scale during use. The connections between the bomb and the autoclave were flexible stainless steel hose or tubing so that continuous weighings could be made without breaking or making any connections. This applies also to the nitrogen line, which was used to pressure the bomb reservoir. To theextent that it was required, any other usual conventional procedure or addition which provided greater safety was used, of course, such as safety glass protective screens, etc.

Attention is directed again to what has been said previously in regard to automatic controls which shut oif the propylene oxide in event temperature of reaction passes out of the predetermined range or if pressure in the autoclave passes out of predetermined range.

With this particular arrangement practically all oxypropylations become uniform in that the reaction temperature was held .within a few degrees of any selected point, for instance, if 105 C. wasselected as the operating temperature the maximum point would be at the most 110 C. or 112 C., and the lower point would be 95 or possibly 98 C. Similarly, the pressure was held at approximately 30 pounds within a 5-pound variation one way or the other, but might dropto practically zero, especially where no solvent such as xylene is employed. The speed of reaction was comparatively slow under such conditions as compared with oxyalkylations at 200 C. Numerous reactions were conducted in which the time varied from one day (24 hours) up to three days (72 hours), for completion of the final member of a series. In some instances the reaction may take place in considerably less time, i. e., 24 hours or less, as far as a partial oxypropylation is concerned. The minimum time recorded was about a 3-hour period in a single step. Reactions indicated as being complete in 7 hours may have been complete in a lesser period of time in light of the automatic equipment employed. In the addition of propylene oxide, in the autoclave equipment as far as possible the valves were set so all the propylene oxide if fed continuously would be added at a rate so that the predetermined amount would react within the first 15 hours of the 24-hour period or two-thirds of any shorter period. This meant that if the reaction was interrupted automatically for a period of time for pressure to drop or temperature to drop the predetermined amount of oxide would still be added in most instances well within the predetermined time period. Sometimes where the addition was a comparatively small amount in a 7-hour period there would be an unquestionable speeding up of the reaction, by simply repeating the example and using 4, 5 or 6 hours instead of 7 hours.

Whenoperating at a comparatively high temperature, for instance, between 150 to 200 C., an unreacted alkylene oxide such as propylene oxide, makes its presence felt in the increase in pressure or the consistency of a high pressure. However, at a low enough temperature it may happen that the propylene oxide goes in as a liquid. If so, and if it remains unreacted there is, of course, an inherent danger and appropriate steps must be taken to safeguard against this possibility; if need be a sample must be withdrawn and examined for unreacted propylene oxide. One obvious procedure, of course, is to oxypropylate at a modestly higher temperature, for instance, at 140 to 150 C. Unreacted oxide affects determination of the acetyl or hydroxyl value of the hydroxylated compound obtained.

The higher the molecular weight of the compound, i. e., towards the later stages of reaction, the longer the time required to add a given amount of oxide. One possible explanation is that the molecule, being larger, the opportunity for random reaction is decreased. Inversely, the lower the molecular weight the faster the reaction takes place. For this reason, sometimes at least, increasing the concentration of the catalyst does not appreciably speed up the reaction, particularly when the product subjected to oxyalkylation has a comparatively high molecular weight. However, as has been pointed out previously, operating at a low pressure and a low temperature even in large scale operations as much as a week or ten days time may elapse to obtain some of the higher molecular weight derivatives from monohydric or dihydric materials.

In a number of operations the counterbalance scale or dial scale holding the propylene oxide bomb was so set that when the predetermined amount of propylene oxide had passed intothe reaction the scale movement through a time operating device was set for either one to two hours so that reaction continued for 1 to 3 hours after the final addition of the last propylene oxide and thereafter the operation was shut down. This particular device is particularly suitable for use on larger equipment than laboratory size autoclaves, to wit, on semi-pilot plant or pilot plant size, as well as on large scale size. This final stirring period is intended to avoid the presence of unreacted oxide.

In this sort of operation, of course, the temperature range was controlled automatically by either use of cooling water, steam, or electrical heat, so as to raise or lower the temperature. The pressuring of the propylene oxide into the reaction vessel was also automatic insofar that the feed stream was set for a slow continuous run which was .shut off in case the pressure passed a predetermined point as previously set out. All the points of design, construction, etc., were conventional including the gauges, check valves and entire equipment. As far as I am aware at least two firms, and possibly three, specialize in autoclave equipment such as I have employed in the laboratory, and are prepared to furnish equipment of this same kind. Similarly pilot plant equipment is available. This point is simply made as a precaution in the direction of safety. Oxyalkylations, particularly involving ethylene oxide, glycide, propylene oxide, etc., should not be conducted except in equipment specifically designed for the purpose.

Example -1a- The particular autoclave used was one having a capacity of approximately 1 /2 gallons. The speed of the stirrer could be varied from 150 to 350 R. P. M. 500 grams of sorbide were mixed with 50 grams of caustic soda. The mixture was charged into the autoclave. The reaction pot was flushed out with nitrogen. The autoclave was sealed and the automatic devices set for injecting a total of 2700 grams of propylene oxide in four hours. The pressure regulator was set for a maximum of 35 pounds per square inch. This meant that the bulk of the reaction could take place, and probably did take place, at a lower pressure. This comparatively low pressure wasthe result of the fact that considerable catalyst was 7 Example 5a In this instance 1800 gramsof the reaction mass identified as Example 4a, preceding,were permitted to remain in the autoclave; No additional catalyst was added. 1035 grams of propylene oxide were added during this fifth stage. The conditions of oxypropylation were substantially the same as in Example 1a, preceding, particularly in regard to pressure and temperature. The time 1'e quired to add the propylene oxide in this stage was six hours. 1 added, in fact the percentage was very hlgh, as 1nd1cated. 10 The propylene oxide was added comparatively slowly at Example a rate of about 700 grams P hour more impoffant, in this instance 1800 grams of the reactionmass iden- Selected temperature l'ange was t0 -9 1. a tified as Example 5a, preceding, were permitted to remain sh y above thfi bolllng P of Water- The Initial in the autoclave. No additional catalyst wasadded. Ap- 1ntroduct1on of propylene oxlde was not started until the proximately 641 grams of propylene oxide were added heating devices had raised the temperature to approxiduring thi sixth stage. The conditions of oxypropylation m y the bolllng POlHt O Water. At the end of the were substantially thesame as in Example la, preceding, reaction a sample was taken and oxypropylation proceedparticularly in regard to pressure and temperature. The ed as 111 EXamPIe imm i ely Suc e g time required to add the propylene oxide in this final stage Exam le2a was Six hours What has been said herein is presented 1n tabular form 2,000 grams of the i f mass P P Y ldefltlfied in Tablel immediately following, with some added inas Example were permuted 0 main in the autoformation as to molecular weight and as to solubility of clave. No add1t1onal catalyst was added. To th1s there the reaction product in water, xylene and kerosene.

TABLE 1 Composition before Composition at end Y M. W. Max. Max. Ex. No. by hyd. temp., pres, Time,

. H. 0. Oxide Catalyst, Theo. H. 0. Oxide Catalyst, deter- 0. lbs. sq. hrs. amt., amt., grs. grs. m0l..wt. amt, amt., grs. min. in.

grs. grs. grs.

1 The hydroxylated compound is sorbide.

was'added 1089 grams of propylene oxide. The time required to add this smaller amount of propylene oxide in the second stage was three hours. In all other respects the operation was the same as in Example In, preceding. The temperature was the same and the pressure was the same. At the end of the reaction period part of the sample was withdrawn and oxypropylation proceeded as in Example 3a, following.

Example 3a Example 4a 1900 grams of the reaction mass identified as Example 3a, preceding, were permitted to remain in the autoclave, and subjected to further oxypropylation without the addition of any more catalyst. The conditions ofoxypropylation were the same as in Example la,,preceding, particularly in regard to temperature andfp're'sslure.v The amount of oxide added was 1262 grams and the addi- All the examples were insoluble in water, and soluble in xylene, and all but Example 1a, were soluble in kerosene.

The final product, i. e., at the end of the oxypropylation step, was a somewhat viscous very pale straw-colored fluid which was water-insoluble. This is characteristic of all various end products obtained in this series. These products were, of course, slightly alkaline due to the residual caustic sodaemployed. This would alsobe the case if sodium methylate were used as a catalyst.

Speakingof insolubility in water or solubility in kerosene such solubility test can be made simply by shaking small amounts of the materials in a test tube with water, forinstance, using 1% to 5% approximately based on the amount of water present.

Needless to say, there is no complete conversion of propylene oxide into the desired hydroxylated compounds. This is indicated by the fact that the theoretical molecular weight based on a statistical average is greater than'the molecular weight calculated by usual methods on basis of acetyl or hydroxyl value. Actually, there is no completely satisfactory method for determining molecular weights of these types of compounds with ahigh degree of accuracy when the molecular weights exceed 2,000. In some instances the acetyl value or hydroxyl value serves assatisfactor'ily as an index to the molecular weight as any other procedure, subject to the above limitations; and especially in the higher molecular weight range. Ifany difiicultyis encountered in themanufacture of the esters as described in Part 2 the stoichiometrical amount of acid or acid compound should be taken which corresponds to the indicated aeetyl or hydroxyl value. This matter has been discussed in the literature and is a matter of common knowledge and requires no further elaboration. In fact, it is illustrated by some of the examples appearing in the patent previously mentioned.

1 PART 2 As previously pointed out the present invention is concerned with acidic esters obtained from the oxypropylated derivatives described in Part 1, immediately preceding, and polycarboxy acids, particularly dicarboxy acids such as adipic acid, phthalic acid or anhydride, succinic acid, diglycollic acid, sebacic acid, azelaic acid, aconitic acid, maleic acid or anhydride, citraconic acid or anhydride, maleic acid or anhydride adducts as obtained by the Diels- Alder reaction from reactants such as maleic anhydride and cyclopentadiene. Such acids should be heat stable so they are not decomposed during esterification. They may contain as many as 36 carbon atoms as, for example, the acids obtained by dimerization of unsaturated fatty acids, unsaturated monocarboxy fatty acids, or unsaturated monoc'arboxy acids having 18 carbon atoms. Reference to the acid in the hereto appended claims obviously includes the anhydrides or any other obvious equivalents. My preference, however, is to use polycarboxy acids hav- 2 ing not over 8 carbon atoms.

The production of esters including acid esters (fractional esters) from polycarboxy acids and glycols or other hydroxylated compounds is well known. Needless to say, various compounds may be used such as the low molal ester, the anhydride, the acyl chloride, etc. However, for purpose of economy it is customary to use either the acid :=orthe-anhydride. A conventional procedure is employed. ,On a laboratory scale one can employ a resin pot. of the kind described in U. S. Patent No. 2,499,370, dated March 7, 19 50, to De Groote & Keiser, and particularly with one more opening to permit the use of a porous spreader if hydrochloric acid gas is to be used as a catalyst. Such device or absorption spreader consists of minute alundumthimbles which are connected to a glass tube. One can add a sulfonic acid such as para-toluene 'sulfonic acid as a catalyst. There is some objection to this because in some instances there is some evidence that this acid catalyst tends to decompose or rearrange the oxypropylated compounds, and particularly likely to do so if the esterification temperature is too high. In the case of polycarboxy acids such as diglycollic acid, which is strongly acidic there is no need to add any catalyst. The use of hydrochloric acid gas has one advantage over paratoluene sulfonic acid and that is that at the end of the reaction it can be removed by flushing out with nitrogen, whereas there is no reasonably convenient means available of removing the paratoluene sulfonic acid or other 'sulfonic acid employed. If hydrochloric acid is employed one need only pass the gas through at an exceedingly slow rate so as to keep the reaction mass acidic. Only a trace of acid need be present, I have employedhydrochloric acid gas or the aqueous acid itself to eliminatethe initial basic material. My preference, however, is. to use no catalyst whatsoever and to insure complete dryness of the diol as described in the final procedure just preceding Table 2.

.The products obtained in Part 1 preceding may contain a basic catalyst. As a general procedure I have added an amount of half-concentrated hydrochloric acid considerably in excess of What is required to neutralize the residual catalyst. The mixture is shaken thoroughly andv allowe d to stand overnight. It is then filtered and refluxed with the xylene present until the water can be separated in a phase-separating trap. As soon as the product is substantially free from water the distillation stops. This preliminary step can be carried out in the flask to be used for esterification. If there is any further deposition of sodium chloride during the reflux stage needless to say a second filtration may be required. In

any event the neutral or slightly acidic solution of the oxypropylated derivatives described in Part 1 is then diluted further with suflicient xylene decalin, petroleum solvent, or the like so that one has obtained approximately a solution. To this solution there is added a polycarboxylated reactant as previously described, such as phthalic anhydride, succinic acid or anhydride, diglycollic acid, etc. The mixture is refluxed until esterification is complete as indicated by elimination of water or drop in carboxyl value. Needless to say, if one produces a half ester from an anhydride such as phthalic anhydride, no water is eliminated. However, if it is obtained from diglycollic acid, for example, water is eliminated. All such procedures are conventional and have 'been so thoroughly described in the literature that further consideration will be limited to a few examples and a comprehensive table.

Other procedures for eliminating the basic residual catalyst, if any, can be employed. For example, the oxyalkylation can be conducted in absence of a solvent orthe solvent removed after oxypropyla-tion. Such oxypropylation end proudct can then be acidified with just enough concentrated hydrochloric acid to just neutralize the residual basic catalyst. To this product one can then add a small amount of anhydrous sodium sulfate (sufficient in quantity to take up any water that is present) and then subject the mass to centrifugal force so as to elimimate the hydrated sodium sulfate and prob-ably the sodium chloride formed. The clear somewhat viscous strawcolored amber liquid so obtained may contain a small amount of sodium sulfate or sodium chloride but, in any event, is perfectly acceptable for esterification in the manner described. t

It is to be pointed out that the products here described are not polyesters in the sense that there is a plurality of both diol radicals and acid radicals; the product is characterized by having only one diol radical.

In some instances and, in fact, in many instances I have found that in spite of the dehydration methods employed above that a mere trace of water still comes through and that this mere trace of water certainly interferes with the acetyl or hydroxyl value determination, at least when a number of conventional procedures are used and may retard esterification, particularly where there is no sulfonic acid or hydrochloric acid present as a catalyst. Therefore, I have preferred to use the following procedure: I

have employed about 200 grams of the diol as described in Part 1', preceding; I have added about 60 grams of henzene, and then refluxed this mixture in the glass resin pot using a phase-separating trap until the benzene carried out all the water present as water of solution or the equivalent. Ordinarily this refluxing temperature is apt'to be in the neighborhood of 130 to possibly 150 C. When all this water or moisture has been removed I also withdraw approximately 20 grams or a little less benzene and then add the required amount of the carboxy reactant and also about 150 grams of a high boiling aromatic petroleum solvent. These solvents are sold by various oil refineries and, as far as solvent effect act as if they were almost completely aromatic in character. Typical distillation data in the particular type I have employed and found very satisfactory is the following:

I. B. P., 142 C. ml., 242 C. 5 ml., 200 C. ml., 244 C. 10 ml., 209' C. ml., 248 C. 15 ml., 215 C. ml., 252 C. 20 ml., 216 C. ml., 252 C. 25ml, 220 C. ml., 260 C. 30 ml., 225 C. ml., 264 C. 35 ml., 230 C. ml., 270 C. 40'ml., 234 C. ml., 280 C. 45 ml., 237 C. ml., 307 C.

After this material is added, refluxing is continued and, of course, is at a high temperature, to wit, about to C. If the carboxy reactant is an anhydride ne'edless to sayno water o frea'ct'ion appears; if the carboxy reactant .is an acid water of reaction should appear and should be eliminated at the above reaction temperature. If his not eliminated I simply separate out another or cc. of benzene by means of the phase-separating trap and thus raise the temperature to 180 or 190" C., or even to 200 C., if need be. My preference is not to go above 200 C.

The use of such solvent is extremely satisfactory provided one does not attempt to remove the solvent subsequently except by vacuum distillation and provided there is no objection to a little residue. Actually, when these materials are used for a purpose such as demulsification the solvent might just as well be allowed to remain. If the solvent is to be removed by distillation, and particularly vacuum distillation, then the high boiling aromatic petroleum solvent might well be replacedyby some more expensive solvent, such as decalin or an alkylated decaiin which has a rather definite or close range boiling point. The removal of the solvent, of course, is purely a conventional procedure and requires no elaboration.

In the appended table Solvent #7-3, which appears in numerous instances, is a mixture of 7 volumes of the aromatic petroleum solvent previously described and 3 volumes of benzene. This was used, or a similar mixture, in the manner previously described. A large number of examples indicated were repeated employing decalin, using this mixture and particularly with the preliminary step of removing all the water. If one does not intend to remove the solvent my preference is to use the petroleum solvent-benzene mixture although obviously any of the other mixtures, such as decalin and xylene, can be employed.

The data included in the subsequent tables, i. e., Tables 2 and 3 are self-explanatory, and very complete and it is believed no further elaboration is necessary:

. 12 Recheck the hydroxyl or ac'etyl value of the oxypropylated product and use a stoich'iometrically equivalent amount of acid; (b) if the reaction does not proceed with reasonable speed either raise the temperatures indicated or else extend the period of time up to l2'or 16 hours if need be; (c) if necessary, use V2% of para- TABLE 3 Ex.No. Arut.s0l- Esterifi- Time of I of acid Solvent vent; cation esterifi- Water out ester (grs) temp, cation 1 cc 0. (hrs.)

7-3 225 148 2% None. 7-3 254 173 3% Do. 7-3 282 156 2 Do. 7-3 276 164 3% Do. 7-3 302. 183 4 5.8. 7-3 256 172 V 3% None. 7-3 223 163 2% Do. 7-3 241 181 4 Do. 7-3 237 152 3% Do. 7-3 254 176 2% 4.1. 7-3 282 152 5% None. 7-3 253 138 3% D0. 7-3 304 172 2% Do. 7-3 242 151 3 Do. 7-3 272 148 4% 1.8. 7-3 312 152 4% None. 7-3 270 183 3% D0. 7-3 255 158 1% Do. 7- 248 165 5 Do. 7-3 232 168 5 About 1 cc. 7-3 272 152 1% None. 7-3 311 165 3% D0. 7-3 303 150 4% Do. 7-3 296 143 2% D0. 7-3 251 156 6 Less than 1 cc.

toluene sulfonic acid or some other acid as a catalyst; (d) if the esterification does not produce a clear product a check should be made to see if an inorganic salt such as sodium chloride or sodium sulfate is not precipitating out. Such sa-lt should be eliminated, at least for explora- TABLE 2 Ex. No. Ex. No. of Theo. Theo Actual M01. wt. Amt. of Amt. of

of acid hydroxy M. W. hvdroxyl hydroxyl based on hyd. Polycarboxy reactant polyester empd. H. C. V. of value actual cmpd. carboxy H. C. H. V. (grs.) reactant 1, 447 77. 1 82. 8 1, 356 226 Phthalic anhydrido. 49 1, 447 77. 1 82. 8 1, 356 226 Maleic anhydride 3 1, 447 77 1 82. 8 1, 356 226 Citraconic anh ydridem. 37 1, 447 77. 1 82. 8 1, 356 226 Su'ocinic auhydride. 3 1, 447 77. 1 82. 8 1, 356 226 Diglyeollic acid 45 2, 296 48. 8 55. 7 2, 015 202 Phthalic anhydritle. 2, 296 48. 8 55. 7 2, 015 202 hdaleic anhydride 20 2, 206 43. 8 55. 7 2, 015 202 Citraconic anhydrida... 22 2, 296 48. 8 55. 7 2, 015 202 Succinlc anhydride 20 2, 296 43. 8 55. 7 2, 015 202 Diglycollic acid 27 3, S26 29. 3 .8 3, 234 216 Phthalic anhydridm 20 3, 826 29. 35 34. 8 3, 234 216 Maleic anhydride 13. 5 3, S26 29. 35 34. 8 3, 234 216 Oitraconic anhydridefln 15 3, S2 29. 35 34. 8 3, 234 216 Succinic anhydride. 13. 5 3, 826 29. 35 34. 8 3, 234 216 Diglycollic acid 18 6, 225 18. 05 26. 05 4, 168 209 Phthalio anhydride 15 6, 225 18.05 26. 95 4, 163 209 Malelc anhydride 10 6, 225 18. 05 26. 95 4, 168 209 Oitradonic anhydrlde 11 6, 225 18. 05 26. 95 4, 163 209 Succinic anhydride 10 6. 225 18. 05 26. 95 4, 168 209 Diglycollic acid 13. 5 8, 445 13. 3 22. 35 5, 035 202 12 8, 445 13. 32 22. 35 5, 035 202 Maleic anhydride 8 8, 445 13. 32 22.35 5, 035 202 Oitraeonic aullydride 9 8, 445 13. 32 22. 35 5, 035 202 Succinic anhydrlde 9 8, 445 13. 32 22. 35 5, 035 202 'Diglycollic acid 9 As has been pointed out previously an alkaline catalyst is generally employed in the oxypropylation of the hexide. 1 have found it convenient to remove the excess of catalyst using hydrochloric acid in the manner previously noted. If a trace of hydrochloric acid remains in the oxypropylated hexide esterification seems to proceed with considerable ease. This is particularly true in the case of the anhydrides and also in the case of diglycollic acid.

The procedure for manufacturing the esters has been illustrated by preceding examples. If for any reason reaction does not take place in a manner that is acceptable, attention should be directed to the following details: (a)

tion experimentation, and can be removed by filtering. Everything else being equal as the size of the molecule increases the reactive hydroxyl radical represents a smaller fraction of the entire molecule and thus more difficulty is involved in obtaining complete esterification.

Even under the most carefully controlled conditions of oxypropylation involving comparatively low temperatures and long time of reaction there are formed certain compounds Whose compositions are still obscure. Such side reaction products can contribute a substantial proportion of the final cogeneric reaction mixture. Various suggestions have been made as to the nature ofthesc compounds, such as being cyclic polymers of prop lene tion procedure can be repeated using an appropriately reduced ratio of carboxylic reactant. Even the determination of the hydroxyl value by convention-al procedure leaves much to be desired due either to the cogeneric materials'previously referred to, or for that matter, the presence of any inorganic salts or propy-lene oxide. Obviously this oxide should be eliminated. The solvent employed, if any, can be removed from the finished ester by distillation and particularly vacuum distillation. The final products or liquids are generally very pale straw color to amber incolor, and show moderate viscosity. They can be bleached with bleaching clays, filtering chars, and the like; However, for the purpose of demulsification or the like color is not a factor and decolorizat-ion is not justified. V

In the above instances I have permitted the solvents to remain present in the final reaction mass. In other instances I have followed the same procedure using decalin ora mixture of decalin and benzene in the same manner and ultimately removed all the solvents by vacuum distillation. Appearances .of the final products are much the same as the diols before esterification and in some instances were somewhat darker in color and had a reddish cast and perhaps somewhat more viscous.

PART 3 Previous reference has been made to the fact that diols such as polypropylene glycol of approximately 2,000 mo lecular weight, for example, have been esterified with dicarboxy acids and employed as demulsifying agents.

On first examination the difference between the herein described products and. such comparable products appears to be rather insignificant; In fact, the difference is such that it fails to explain the fact that compounds of the kind herein described may be, and frequently are,

or better on a quantitative basis than the simpler compound previously described, and demulsify faster and give cleaner oil in many instances. The method of making such comparative tests has been described in a booklet entitled Treating Oil Field Emulsions, used in the Vocational Training Courses, Petroleum Industry ondary alcohol group is unitedto a primary alcohol group, etherization being involved, of course, in each instance.

Usually no eifort is made to differentiate between oxypropylation taking place, for example, at the primary alcohol radical or the secondary alcohol radical. Actually, when such products are obtained, such as a high molal polypropylene glycol or the products obtained in the manner herein described one does not obtain a single derivative such as'HO(RO)nH in which n has one and only one value, forinstance, 14, 15 or 16, or the like.

I =Rathe'r', one obtains a cogeneric mixture of closely related or touching homologues. These materials invariably have high molecular weights and cannot be separated from one another by any known procedure without decomposition. The properties of such mixture represent the contribution of the various individual members of the mixture. On a statistical basis, of course, n can be appropriately specified.- For practical purposes one need only consider the oxypropylation of a monohydric alcohol because in essence this is substantially the mechanism involved. Even in such instances where one is concerned with a monohydric reactant one cannot draw a single formula and say that by following such procedure one can readily obtain or 60% or of such compound. However, in the case of at least monohydric initial reactants one can readily draw the formulas of a large number of compounds which appear in some of the probable mixtures or can be prepared as components and mixtures which are manufactured conventionally.

Simply by way of illustration reference is made to the De Groote, Wirtel and Pettingill Patent 2,549,434, dated April 17, 1951, the application for which was copending with my copending application Serial No. 198,756, noted above.

However, momentarily referring again to a monohydric initial reactant it is obvious that if one selects any such simple hydroxylated compound and subjects such compound to oxyalkylation, such as oxyethylation, or oxypropylation, it becomes obvious that one is really producing a polymerof the alkylene oxides except for the terminal group. This is particularly true where the amount of oxide added is comparatively large, for instance, 10, 20, 30, 40 or 50 units. If such compound is subjected to oxyethylation so as to introduce 30 units of ethylene oxide, it is well known that one does not obtain a single constituent which, for the sake of convenience, may be indicated as RO(C2H4O)aoH. Instead, one obtains a cogeneric mixture of closely related homologues, in which the formula may be shown as the following,

wherein n, as far as the statistical average goes, is 30, but the individual members present in significant amount may vary from instances where n has a value of 25, and perhaps less, to a point where 11 may represent 35 or more. Such mixture is, as stated, a cogeneric closelyrelated series of touching homologous compounds. Considerable investigation has been made inregard to-the distribution curves for linear polymers. Attention is di rected to the article entitled Fundamental Principles of Condensation Polymerization, by Flory, which appeared in Chemical Reviews, volume 39, No. 1, page 137.

Unfortunately, as has been pointed out by Flory and other investigators, there is no satisfactory method, based on either experimental or mathematical examination, of indicating the exact proportion of the various members of touching homologous series which appear in cogeneric condensation products of the kind described. This means that from the practical standpoint, i. e., the ability to describe how to make the product under consideration and how to repeat such production time after time without difliculty, it is necessary to resort to some other method of description, or else consider the value of n, in formulas such as those which have appeared previously and which appear in the claims, as representing both individual constituents in which n has a single definite value, and also with the understanding that n represents the average statistical value based on the assumption of completeness of reaction.

This may be illustrated as follows: Assume that in any particular example the molal ratio of the propylene oxide to the diol is 30 to '1. Actually, one obtains products in which n probably varies from 10 to 20, perhaps even further. The average value, however, is 15, assuming, as previously stated, that the reaction is complete. The

Furthermore, it becomes obvious that one now has an unsymm'etrialradical in the'majority of cases for the reason that in the .cogeneric formula which appeared previously, i. e.,

' u 1| nooownoto 03mma'owtntonoawoonw n and n are usually not equal. For instance, if one introduces 15 moles of propylene oxide, n and n could not-be equal insofar that the nearest approach to equality is where-the value of n is 7 and n is 8. However, even in the case of- 'an even number such as 20, 30, 40 or 50,

it is also'obvious that n and n will not be equal in light of what has been said previously. Both sides of the molecule are not going to grow with equal rapidity, i. e., to the same size. Thus the diol herein employed is differentiated from polypropylene diol 2000, for example, in that (a) it carries a hetero unit, i. e., a unit other than apropylene glycol or propylene oxide unit, (b) such unit is off center,--and (c) the effect of that unit, of course, must have som-e'eifect in the range with which the linear molecules can be drawn together by hydrogen binding or van der Waals' forces, or-whatever else may be involved.

What has been said previously can be emphasized in the following manner. It has been pointed out previously that in the last formula immediately preceding, n or n could be zero. Under the conditions of manufacture as described in Part 1 itis extremely unlikely that n is ever zero. However, such compounds can be prepared readily with comparatively little difficulty by resorting to a block- 'ing effect or reaction. 'For instance, if the oxypropylated dihydroxylate-d hexide is es-terified with a low molal acid such as acetic acid mole for mole and such product subjected to oxyalkylation using a catalyst, such as sodium methylate and guarding against the presence of any water, it becomes evident that all the propylene oxide introduced, for instance 15 to 80 molecules per polyhydric allcohol necessarily must enter at one side only. If such product is then saponified so 'as to decompose the acetic acid ester and then acidified so as to liberate the water-soluble acetic acid and the water-insoluble :diol -a separation can be made and such diol then subjected to esterification as described in Part 2, preceding. Such esters, of course, actually represent products where either It or n is zero. Also intermediate procedures can be employed, i. e., following the same esterification step after partial oxy propylation. For instance, one might oxypropylate with one-half the ultimate amount of propylene oxide to be used and then stop the reaction. One could then convert this partial oxypropylated intermediate into an ester by reaction of one mole of acetic acid with one mole of a'diol. This ester could then be oxypropylated with all the remaining propylene oxide. The final product so obtained could be saponified and acidified so as to elimimate the water-soluble acetic acid and free the obviously unsymmetrical diol which, incidentally, should also .be kerosene-soluble.

From a practical standpoint -I have found no advantagedmgoing to this extra stepbut it does emphasize the difference-in structure between the herein described diols employed asintermediates and high molal'polypropylene glycol, such as polypropylene glycol 2000.

PART 4 Previous reference has been made to oxyalkylating agents other than propylene oxide, such as ethylen-eoxide. Obviously variants can be prepared which do not depart from what is said herein but do produce modifications. The diol,-i. e., the dihydroxylated initial compound, to wit, the hexides, .can be reacted with ethylene oxide in modest amounts and'then subjected to oxypropylation provided that the resultant derivative is (a) water-insoluble, (b) kerosene-soluble, and (c) has present 15 to alkylene oxide radicals. Needless to 'say, in order 'to have water-insolubil-ity and kerosene-solubility the large majority must be propylene oxide. Other variants suggest themselves as, for example, replacing propylene oxide by butylene oxide.

More specifically then one molevof sorbide can be treated with 2, 4 or 6 moles of ethylene oxide and then treated with propylene oxide so as to produce a waterinsolu'ble, kerosene-soluble diol in which there are present 15 to 80 oxide radicals as previously specified. Similarly the propylene oxide can be added first and then the ethylene oxide, or random oxyalkylation can be employed using a mixture of the two oxides. The compounds so obtained are readily est-erified in the same manner as described in Part 2, preceding. Incidentally, the diols described in Part 1 or the modifications described therein can be treated with various reactants such as glycide, epichlorohydrin, dimethyl sulfate, sulfuric acid, maleic anhydride, ethylene'imine, etc. If treated with epichlorohydrin or monochloroacetic acid the resultant product can be further reacted with a tertiary amine such as pyridine, or the like, to give quaternary ammonium compounds. If treated with maleic anhydride to give a total ester the resultant can be treated with sodium bisulfite to yield a su-lfosuccinate. Sulfo groups can be introduced also by means of a sulfating agent as previously suggested, or by treating the chloroacetic acid resultant with sodium sulfate.

I have found that if such hydroxylated compound or compounds are reacted further so as to produce entirely new derivatives, such new derivatives have the properties of the original hydroxylated compound insofar that they are effective and valuable demulsifying agents for reso- 'lution of water-in-o'il emulsions as found in the petroleum industry, as break inducers in doctor treatment of sour crudes, etc.

This application is a division of my copending application Serial No. 198,756 filed December 1, 1950.

Having thus described my invention what I claim as new and desire to obtain by Letters Patent, is: v

l. Hydrophile synthetic products, said hydrophile synthetic products being characterized by the following formula inwhich R is a hexide radical; n and rtiare nu merals with the proviso that n and n equal a sum varying from 15 to 80, and R is the radical of a dicarboxy acid selected from the group consisting of acyclic and isocyclic dicarboxy acids having not more than 8 carbon atoms and composed of carbon, hydrogen and oxygen of the formula:

ooon R\ coon and with the further proviso that theparent dihydroxylated compound prior to esterification be water-insoluble.

2. Hydrophile-synthetic products; said hydrophile synthctic products being characterized by the following formula O (HOOC)R()(OC2Hu)nORO (UaHaOM RKIOOH) in which R is a sorbide radical; n and n are numerals with the proviso that n and n equal a sum varying from 15 to 80, and R is the radical of a dicarboxy acid selected from the group consisting of acyclic and isocyclic dicarboxy acids having not more than 8 carbon atoms and composed of carbon, hydrogen and oxygen of the formula:

and with the further proviso that the parent dihydroxylated compound prior to esterification be water-insoluble.

3. Hydrophile synthetic products; said hydrophile synthetic products being characterized by the following formula n i? (HOOC)RC(OGaHa).-ORO(OsHuOh OROIJOOH) in which R is a sorbide radical; n and n are numerals with the proviso that n and n equal a sum varying from 15 to 80, and R is the radical of a dicarboxy acid selected from the group consisting of acyclic and isocyclic dicarboxy acids having not more than 8 carbon atoms and composed of carbon, hydrogen and oxygen of the formula:

COOH coon is citraconic acid.

8. The products of claim 3 wherein the dicarboxy acid is diglycollic acid.

References Cited in the file of this patent UNITED STATES PATENTS Blair Aug. 7, 1951 De Groote Jan. 27, 1953 OTHER REFERENCES Schwartz-Perry: Surface Active Agents Interscience (1949), p. 209.

Claims (1)

1. HYDROPHILE SYNTHETIC PRODUCTS, SAID HYDROPHILE SYNTHETIC PRODUCTS BEING CHARACTERIZED BY THE FOLLOWING FORMULA