InUUSTRIAL AND ENGINEERING CHEMISTRY
1186
a small proportion of
the i r o n o x i d e and alumina remained in t h e citrate-insoluble residue. It is noteworthy that, although r e v e r s i o n had been noted in t h e case of sample 75-76d, the citrate-insoluble PzOs apparently was in the u same form as in the x XP,O, IN ROCK other runs, apparently FIGURE 3. EFFECTOF GRADEOF being present as fluorROCKON CONVERSION TO AVAILapatite. ABLE PHOSPHATE The mole ratio of the citrate-soluble Pz05 (total PzO5 minus water-soluble P,O5 and citrate-insoluble P&) to the citrate-soluble Fen03 plus A1203is given in the last column of Table IV. In all except the last three runs, this ratio is approximately unity, indicating that citrate-soluble compounds such as trialuminum phosphate and triferric phosphate are formed.
VOL. 28, NO. 10
In the last three runs of Table IV reversion was observed. I n these runs the mole ratios of citrate-soluble PzOF,to citratesoluble Fez03 plus AlpOa were 2.2, 3.0, and 3.1, respectively. This may indicate that the cause of reversion was the formation of di- or monoaluminum and ferric phosphates, thus causing a greater proportion of the acid to be combined with the iron and alumina, and causing a reversion of the monocalcium phosphate to fluorapatite. This phenomenon was especially marked and rapid when superphosphate samples were heated to about 300" F. (150" C.) for an hour or so, indicating that the formation of di- and monoaluminum and ferric phosphates undoubtedly proceeds, although a t a slower rate, a t room temperature.
Literature Cited (1) Copson, Newton, and Lindsay, IXD. ENO.CHEM., 28,923 (19363. (2) Hill and Beeson, J. Assoc. Oficial d g r . Chem., 18,244-60 (1935). (3) Hill and Hendricks, ISD. ESQ. CHBM.,28, 440 (1936). (4) Hill and Jacob, J . LIssoc. Oficial Agr. Chem., 17,487-505 (1934). (5) .Jacob, Hill, Marshall, and Reynolds, U. 8. Dept. Agr., Tech. Bull. 364 (June, 1933). (6j Marshall. Rader, and Jacob, IND.ESG. CHEU.,25, 1253 (1933). (7) Willard and Winter, IND. ESG.C H E K ,.%rial. Ed., 5, 7-10 (1933). RBCEIVED July 2, 1936
TERTIARY ALKYL ETHERS PREPARATION AND PROPERTIES
@I
N 1907 the Belgian chemist Reychler (8) heated trimethylethylene, methanol, and sulfuric acid together in a sealed tube and found that methyl tertiary amyl ether was produced: CH,OH
+ CsHlo +CH30CjHn (tertiary)
Regchler performed only this one experiment, and little attention was paid to the reaction, probably because the tertiary olefins requisite to the synthesis have not been readily accessible. Thus to secure trimethylethylene, it was necessary to isolate isoamyl alcohol from fused oil and then dehydrate the alcohol; the preparation of isobutylene necessitated the similar isolation of isobutanol as a first step. In recent years, however, the advances in the cracking and distilling arts in the petroleum ndustry have made these olefins readily available as major constituents of the hydrocarbon fractions of appropriate boiling point. These fractions consequently afford a conlenient source of the two lowest tertiary olefins, although in the form of solutions in other hydrocarbons. However, tertiary butanol and pentanol are likewise cheaply available a t present from petroleum. Hence they serve as ready sources of pure isobutylene and trimethylethylene when required. The availability of these materials led t o the present investigation of Reychler's reaction. The authors have endeavored to determine the scope of this reaction with regard to both alcohols and olefins, and to determine the most suitable conditions for producing these ethers in good yields
Starting Materials OLEFINS. Starting with methanol, the general reaction may be written: CHsOH CnH?, C- CHsOC,Ha,+i
+
T.W.EVANS AND K. R. EDLUND Shell Development Company, Emeryville, Calif
I
This reaction reaches an equilibrium which varies with the olefin used. By observing the change in yield, the suitability of the different olefks for this synthesis can be determined. In this reaction isobutylene reacts very rapidly and the equilibrium lies well on the ether side. Trimethylethylene is slightly less favorable than isobutylene, and the tertiary hexylenes derived by dehydrating tertiary hexyl alcohols are only moderately suitable. By the time diisobutylene is reached in the series, practically no reaction takes place. illso the introduction of chlorine into the isobutylene molecule inhibits the reaction, so that @-methylallyl chloride is not suitable. On attempting t o use secondary olefins such as 2butylene, much higher reaction temperatures and catalyst conccntrations are required and lower yield.; otitained than wit8hi-obutylene. ALCOHOLS. Similarly, to evaluate the alcohols used in this reaction, a fixed olefin, isobutylene, can be reacted with various alcohols and the change in yield noted. By this means it is found that, the more acid or labile the hydrogen in the hydroxyl group, the faster the reaction and the better the yield. Thus, as a class, the priinary alcohols are much inore suitable than the secondary alcohols, and the tertiary alcohols react slightly if a t all. [The existence of tertiary dibutyl ether is doubtful, although Pet'rov (8) claims t o have secured it.] Of these alcohols, methanol is the most reactive. However, the primary alcohols, a t least as high as isoamyl, may be employed. Of the secondary alcohols, isopropanol and secondary butanol give only moderate yields, while the secondary pentanols are only slightly reactive. Similar considerations apply to the polyhydric alcohols. If the acidity of the hydrogen atom is taken as the controlling factor, it is interesting to observe that, although methanol adds to 2-butylene a t only a slow rate to give low yields of methyl
OCTOBER, 1936
INDUSTRIAL AND ENGINEERING CHEMISTRY
secondary butyl ether, acetic acid under similar circumstances adds rapidly to give good yields of secondary butyl acetate. CATALYSTS.Sulfuric acid is the best of the catalysts used, although a number of substance$ are suitable. Thu5 the mineral acids in general, organic sulfonic acids, and salt4 such as aluminum chloride can be employed. From a practical standpoint, however, sulfuric acid possesses the advantages of cheapness and of yielding a good reaction rate, and a product easily purified.
Preparation of Ethyl Tertiary Amyl Ether As a specific illust'ration, the preparation of ethyl tertiary amyl ether may be considered. A satisfactory reaction mixture consists of 600 cc. of pure trimethylethylene (5.3 moles), 175 cc. absolute ethanol (3.0 moles), and 24 cc. sulfuric acid (0.45 mole). This mixture separates into two liquid phases which must he stirred to bring about the reaction. A pressure autoclave is used and a temperature of 60" C. is maintained. This temperature is chosen because it produces a satisfactory reaction velocity and favorable equilibrium. Higher temperatures lead to a leps favorable equilibrium, dehydration of the ethanol t o ethyl ether, and polymerization of the olefin. Equilibrium is reached in an hour; at this point the mixture is cooled and drained. There are still two liquid phases presenta lower acidic phase of 100 cc. and an upper hydrocarbon phase of 680 cc. Analysis shows that t,he upper layer contains only a trace of sulfuric acid, small amounts of polymers and tertiary pentanol, and roughly 40 per cent ethyl tertiary amyl ether, 1.5 r cent ethanol, and 57 per cent trimethylethylene. This p"eaves the lower layer with nearly all the original sulfuric acid, the bulk of the unreacted alcohol, and relatively small amounts of ether and olefin. Hence only the upper layer is worked for its ether content. First, the upper layer is neutralized with a little strong caustic to remove the sulfuric acid, and then distilled. Ethanol and trimethylethylene form an azeotrope containing about 5 per cent alcohol so that the first distillate consists of this azeotrope. The ethanol is complet'ely removed in this fashion before all the trimethylethylene has come off; by the time that occurs, the still bottoms consist of nearly pure ether. By continuing the distillation, t,hen, pure ether can be taken overhead. It is essential that all acid be excluded during this distillation. Otherwise, as the olefin and alcohol are removed, the ether will decompose in an effort to reiistablish equilibrium. After the upper layer is thus processed, the best. way to utilize t.he lower layer is to add fresh sulfuric acid, alcohol, and trimethylethylene equivalent to t'hat removed in the upper layer and again react the mixture at 60" C. for an hour. In this way the same two phases are secured again as at the end of the first reaction, and the process can be repeated indefinitely. The recovered pentylene-ethanol mixture may obviously be reprocessed from one run to the next so that overall yields of better than 90 per cent are readily obtained. The foregoing results take place with pure olefin. If the pent)anep e n t y l e n e fraction obtained by fractionating pressure distillate is used, the yield i.: reduced considerably, since bhe activity of the olefin is decreased by dilution with inert h y d r o c a r b o n s . Thus, starting with a hydrocarbon fraction containing 20 per cent tertiary pent'ylenes, the upper phase contained about 7 to 9 per cent, ethyl amyl ether in contrast to 40 per cent ether from pure pentylene. As an e x a m p l e o f t'he influence of the alcohol on t h i ? e q u i l i b r i u m , if methanol had been used instead of ethanol n-ith this same pentanepentylene fraction, the equilibrium upper layer rould have contained about 17 per cent m e t h y l a m y l ether as against the 7 to 9 per cent, obtained from ethanol.
1187
Inasmuch as the equilibrium depends on the activity of both the pentylene and the alcohol, i t might be expected that, by increasing the amount of ethanol, a higher concentration of ether would be present in the final upper layer. Theoretically this is true, but practically a limit is soon reached at which the excess alcohol dilutes the catalyst, thus slowing down the reaction so that equilibrium cannot be approached in a reasonable time. This method of synthesis possesses the advantage of giving a crude mixture in which the ratio of ether to alcohol is high, thus facilitating the separation of pure ether and leading to good alcohol efficiencies. This is in marked contrast to the resu!ts obtained by the conventional methods of synthesis from the two corresponding alcohols or by the Williamson synthesis. As an example of this huperiority, pure methyl tertiary butyl ether is readily obtained by this synthesis. Bennet and Philip (1) in 1928 desired this pure compound for solubility nieasurements and prepared it by the usual methods. Their product, however, boiled 1" C. low, had a density two units high in the second decimal place, and showed too great a solubility both for and in water. Apparently their sample was contaminated with alcohol. I n addition, this ynthesis readily yields isobutyl tertiary butyl ether, a compound which Reboul ( 7 ) was unable to prepare by the customary procedure. The example given of pure trimethylethylene and etliaiiol is particularly attractive because the azeotropic properties are such that both the unreacted olefin and alcohol are easily recovered in a form directly suitable for reuse. This is not always the case, however. Thus if normal butanol were suhstituted for ethanol, this method of recovery would not do because normal butanol and trimethylethylene do not form an azeotrope. I n cases of this kind i t is preferable first to extract the alcohol from the upper layer by washing either with water or dilute sulfuric acid, and then to distill to recover the ether. Although the foregoing examples have been of batch processes, the change to a continuous process is readily made. If, for example, i t is desired to prepare ethyl tertiary amyl ether continuously, a tower is filled about three-quarter. full with a n equilibrium lower layer such as t h a t described. Into the hase of the tower a solution of ethanol in trimethylethylene is injected. As the droplets of this solution rise through the acid phase, reactjon takes place and the top part of the tower
The direct addition of olefins to alcohols in the presence of catalysts to give ethers has been investigated as well as the properties of the resulting ethers. Those olefins which can be derived theoretically by the abstraction of water from a tertiary alcohol are most suitable. A large number of catalysts are available, but sulfuric acid is the most satisfactory. The synthesis is reversible, the reaction reaching a definite equilibrium. Consequently, the pure ethers are unstable in the presence of mineral acids, but are stable in alkaline or neutral solution. The methyl tertiary alkyl ethers form peroxides very slowly on storage in contrast to most ethers. The ethers generally have a pleasant, camphor-like odor, and those of the monohydric alcohols are relatively insoluble in water. The ethers of the polyhydric alcohols are partially soluble in water, according to the degree of etherification. Methyl tertiary butyl ether, the lowest boiling in the series, is like diethyl ether in its physical properties but possesses the advantage of lower vapor pressure.
INDUSTRIAL AKD ENGINEERlNG CHEIIISTRY
1188
VOL. 28, NO. 10
fills with equilibrium upper layer which is drawn off continuously. From time to time small additions of sulfuric acid are made to compensate for that removed in the upper layer.
tractives on a volumetric basis. Holyever, nleaaurements of the speed of diffusion of acid from the water phase to the ether phase show that equilibrium is reached more rapidly with the tertiary than with diethyl ether. On distilling the Properties of Ethers Formed resulting ext'racts, a further advantage of the tertiary ether Table I lists the compounds prepared by this synthesis (3) is noted in that, by virtue of its higher boiling point (55" during the present investigation; Table I1 gives additional against 35" (2.1, i t removes more water in its azeotrope than properties of the lower members of the series. does the diethyl ether; in addition, its azeotrope stratifies into two layers. __. The ethers of the monohydric alcohols are not solvents for TABLE I. ETHERSPREPARED SYNTHETICALLY the simple cellulose esters, although they show high dilution ratios with the accepted solvents. The lower ethers are solEther Point. Boilingr . dZz vents for the mixed acetate-propionate ester, however. The Methyl tert-butyl 55 0 7405 Ethyl tert-butyl i3 0 mono ethers of the polyhydric alcohols are cellulose nitrate Isopropyl tert-butyl 123-4 87-88 :,:!& solvents, as might be expected. n-Butyl tert-butyl Isobutyl tert-butyP aec-Butyl tert-butyln Ieoamyl tert-butyl" Methyl tert-amyl Ethyl tert-amyl Isopro yl tert amyla Methy? tert-hexyl" tert-Monobutyl ether of ethylene glycol tert-Dibutyl ether of ethylene glymla Mixed methyl tert-butyl ether of ethylene glyoola Mixed ethyl tert-butyl ether of ethylene glycol Mixed n-butyl tert-butyl ether of ethylene glycola tert-Monobutyl ether of 1,2-propylene glycol" tert-Monobutyl ether of glycerols tert-Dibutyl ether of glycerol tert-Monoamyl ether of ethylene glycol" tert-Monobutyl ether of diethylene glycolo Phenyl tert-butyl Benzyl tert-butyl 0
114 114-5 i3s-411 86-7 101-2 114-5 113 153
0 7516 0.7604 0.7662 0,7703 0.7667
..
,.
0.7815
83 (20 mm.)
0.8970 0,8266 01) . 88399 311 0 8317
151-3 9 3 4 (5 mm.) 8&2 (4 mm.) 50-5 (3 rnm., 72 (,2 mm.) 185-6 X2-83 , Hinin.
0.9947 0.8921 0,8993 0.9374 0.9214 0 921
171
131-2 148
0.8707
Not previously described.
__
~. - ~ . ~~~
~
One of the most interesting features of the methyl tertiary alkyl ethers is their resistance to autoxidation and peroxide formation on storage. This property has been observed for methyl tertiary butyl, amyl, and hexyl ethers over a period of several years. Others closely related to these ethers structurally, such as diisopropyl and ethyl tertiary butyl ethers, form peroxides fairly readily on storage (W,6),
___
.~~
TABLE111.
.&ZEOTROPES
Boiling ?& A Component R Point, C. by Weight Methanol 51.6 85 Rater 52.6 96 (1.7:4 Ethyl tert-hutyl Ethanol 66.6 79 see-Butanol S o azeotropism see-Pentanol No azeotropism Water 65.2 94 !4.0!' Methsi tert-anis1 Methanol 62.3 50 sec-Butanol 86.0 93 sec-Pentanol No azeotropism Water 73.8 91 (6.5)O Ethyl tert-amyl Ethanol 66.6 79 sec-Butanol 94.5 61 Water 81.2 87 j10.310 .4zeotrope splits into two phases on condensing; the figure in perenthesea i b the volumP per cent layer. Component A !Ether) Methyl tert-butyl
A characteristic of these ethers, and of et,hersin general, is that they form azeotropes with the alcohols boiling near their own boiling points. Some of these azeotropes have been determined and are listed in Table 111. Phenyl tertiary butyl ether was also prepared. _______-__This is an interesting compound because of the TERTIARY ETHERS T.4BLE 11. PROPERTIES O F LOWER difficulties which have attended attempts a t its Methyl Ethyl Llethyl Etbl preparation (4) ; the product generally resulting Ether tert-Butyl !ert-Butyl tert-.lmsl tart-Amyl is not the desired ether but para tertiary butylCH~OCIHO CrHsOCaHg CH.KOCHi1 C:HSOC~HII Formula 116 phenol. The reaction of isobutylene and phenol 88 102 102 Mol. weight in the presence of sulfuric acid leads largely to Density: d' 4 0.7456 0.7456 I1 i T 5 ( i 'I.7703 alkylated phenols, but some ether is formed and dzi 0.7405 0.7404 0.7703 11,7657 may be isolated by alkali extraction of the phenols. 0 7656 ll.7609 0 . 7 3 5 2 0 . 7 3 5 3 d2z The product obtained here was alkali-insoluble ,i 7561 d3z 0.7299 0.7300 0 7607 Refractive index, ng0 1.3689 1 3760 1 31888 1 :NE and, upon addition of aluminum chloride, re\'apor pressure a t 25' mm. Hg liy arranged to para tertiary butylphenol, A similar B. p. at 760 mm. H g ,C.,C. 5 245 5.2 7 130 2.8 8 675 .3 Heat vaporization, cal./gram 76.8 74.3 iR.0 1 reaction of diisobutylene and phenol is recorded Sp. heat (liquid, 25' C . ) , Gal./ [ I . 50 in the lit,erature. 0 52 0.51 0.51 gram/' C. Surface tension a t 24' C., dynes/ In general, the Reychler reacfion niakes easily 22.6 21.8 19.4 19.8 om. Change in b. p., a t 760 mm., accessible a large number of tert,iary mixed ethers 0.048 0.057 0.045 0.049 C./mm. Hg which are otherwise rather difficult to obtain. In Soly. of ether in water a t 20' C., grams/100 soln. 4.8 1.2 1.15 particular, it offers a cheap means of preparing Soly. of watergram8 in ether a t 20' C., grams/100 grams soln. 1.5 0.5 0.6 0 2 the lower members of the series on a commercial scale. The tertiary ethers are unstable in the presence of acids, since they are synthesized by a reversible reaction. Mineral acids in particular decompose them rapidly, organic acids a t a much lower rate. Because of the slowness with which their decomposition is catalyzed by acetic acid, they are suitable for its concentration from dilute solutions, either by extraction as in the Brewster process using diethyl ether, or by azeotropic removal of water. In tzhisconnection, the system methyltertiary butyl ether-water-acetic acid has almost the system for disame solubility relations &s the ethyl ether. Hence the two ethers are potentially equal es-
Literature Cited (1) Bennet a n d Philip, J. Chem. SOC., 1928, 1934. (2) Clover, J . Am. Ciiem,. soc., 44, 1107 (1922); 46,419 ( 1 9 ~ 9 ) . (3) E d l u n d and Evans, U.S. P a t e n t s 1,968,033 a n d 1,968,801 (1934).
it; J": (6) petro;-, R ~~~~~1
J.
~~;.",':~~3~;g31)~
~p h y~s . crhem. ~ ~. ~ s21,348 . , (18xuj
(7) R e b o u l , Bull. S O C . chim. [3]2 , 2 5 (1889). (8) Rewhler, Bull. am. c h h . B e b , 21, 71 (1907).
R E C E I V EJune D 23, 1936. Presented before t h e Division of Orgaiiic Chemistry a t the 90th Meeting of the American Chemical Society, San Frenoieoo, Calif., .kugust 19 to 23, 1935.