1409
NOTES
hIarch, 1950
ratio.’S2 These spectral differences have been used t o follow the changes occurring prior to gelation on the addition of an ammonium salt or acid and to detect differences between silicates of the same analytical composition but different properties. Figure 1 shows the absorption spectrum of 1.0 x M pinacyanol chloride in a mixture of 0.01 M NanSiOsand 0.01 M HzSO4 after the acidsilicate mixtures had aged a t twice this concentration for three minutes, one and five hours. The absorption spectrum changed gradually from that practically the same as in sodium metasilicate alone to the spectrum characteristic of the dye in a silicate with a silica to alkali ratio of 4 or 5 in mixtures aged for one hour.2 The a and /3 bands a t 600 and 548 mp, respectively, almost disappeared and the y band a t 498 mp became evident. During the subsequent fifteen hours the mixture changed so that the a and 0 bands of the dye’s spectrum again became well defined and the y band decreased in intensity and shifted to 514 mp. Aging mixtures of 0.1 M Naz0.3.3SiO~ 0.01 AI HzS04 produced a decrease in the intensity of the y band of the dye in these solutions with an accompanying increase in intensity of the a and B bands. These changes were reversed after about twenty hours of aging. Such results indicate that a maximum effect on the absorption spectrum of the dye is observed with particles of an intermediate molecular weight. The absorption spectrum of pinacyanol chloride was approximately the same in a 0.01 M Naz0.3.3SiO2-0.01 M (NH4)2S04 solution aged a t twice this concentration for twenty-one hours as in a freshly prepared mixture. The absorption spectra were therefore determined in 0.025 M solutions aged a t a concentration of 0.05 Af. In contrast to the above results with sulfuric acid, the addition of 0.05 M ammonium sulfate to 0.05 M sodium metasilicate produced within three minutes a profound change in the nature of these solutions as shown by changes in the absorption spectrum of the dye in them after dilution. The a and p bands had almost disappeared and a prominent y band appeared which increased somewhat in intensity on further aging. In mixtures of 0.05 M (NH4)~S04-0.05M Na~0.3.3Si02 only changes in the intensity of the y band of the dye were observed until the solution gelled. These differences in the behavior with pinacyano1 chloride of acid-silicate and ammonium sulfatesilicate mixtures during aging suggest differences in the mechanism of gelation. This is in agreement with a suggestion made previously on the basis of differences in the effect of concentration and of acid or ammonium salt to silicate ratio on the time of set.a Gelation of a silicate(1) R. C. Merrill, R. W. Spencer, and R. Getty, Tms JOURNAL, TO,
6.0
4.0
. 1
2 x 3.0 a
2.0
+
2460 (1948). (2) R. C. Merrill and R. W. Spencer, (bid., 70, 8688 (1948). ( 8 ) R. C. Merrill sad R. W. Spencer, J , Phyr. Colloid Chens., aubmittcd.
1.o
550
450
650
1,w.
Fig. 1.-Molar extinction coefficients of 1.0 X M pinacyanol‘chloride in 0.01 M Na2SiOs-0.01 M H&O, after acid-silicate mixture aged at twice this concentration for 3 min. - 0, 1 hr. - a, and 5 hr. - m.
ammonium salt mixture may involve formation of a readily coagulated ammonium salt of a silicic acid of intermediate molecular weight as well as the condensation reaction occurring on the addition of an acid. CHEMICAL DEPARTMENT PHILADELPHIA QUARTZ Co. PHILADELPHIA, PA. RECEIVED DECEMBER 16, 1949
The Preparation of 4-Ethyl-2-methoxy-2-phenylmorpholine BY ROBERT H. J O R D A S ~
ATD
ROBERT E. LUTZ
In our studies on ring-chain tautomerism of aethanolaminoketones2 i t was found that acidcatalyzed methanolysis of a-(ethylethano1amino)acetophenone (Ia) gave readily and in good yield the cyclic acetal (Ib),zb,dand there was observed no evidence of the formation of significant amounts of any other product. The formation in (1) Postdoctorate Fellow (1948-1949),supported by a grant-in u d from the National Institute of Health, recommended by the National Cancer Institute. (2) (a) Lutz, Freek and Murphey, Tar8 JOURNAL, 70,2015 (1948); (b) a papa presented at the Chicago meeting of the Arn8dc.n Chemical Oodety, Awil 20, 1948; (e) Lutr u d Yurphw, TPI JOURNAL Tl, 47E (1940); (d) Lutr and Jordam, W . ,71, 991 (1949).
1410
ES
VOl. 73
Preparation of Ib Hydrochloride.-Sample no. 1 (see Figs. 2 and 3), which proved to be unchanged Ib hydroI CJ%C=CHACH,CH? chloride, was obtained in 48% yield in accordance with the directions of Cromwell and TsouSbfor the preparation CH2 CH2 HCl of their “dimethoxy” compound (11) from Ib hydrochloCdIa I I OCHs OCHa ride; upon crystallizpion from methanol-ether mixture ‘G, /CH2 it melted a t 148-149 ; the mixture melting point with a RO/ o I1 sample of Ib hydrochloride prepared according to o;r original directions2d (of m. p. 148-149’) was 148-149 . I (a) R = H (Anal. Calcd. for ClaH1~N02.HC1:C, 60.57; H , 7.81; (b) R = CH, OCHs, 12.04. Found: C, 60.43; H, 8.17; OCHa, (c) R = CZH5 11.99.) X-Ray patterns from the sample prepared according to the Cromwell-Tsou directions for their “dithis reaction of a “dimethoxy” compound (11) methoxy” compound 11, are shown in Figs. 2 and 3, and such as was reported by Cromwell and Tsou3 are identical respectively with those of Figs. 1 and 4 obseemed to us y n l i k e l ~ . ~Our attempts to repeat tained from the original Ib hydrochloride.2d In a repetitheir preparation of “II”3bindeed gave only the ex- tion of this experiment the recovery of purified Ib hydrochloride was 70%. In another similar experiment where pected cyclic acetal (Ib hydrochloride), which was absolute ethanol was substituted for methanol, Ib hydroidentified by mixture melting point with an au- chloride was again recovered (80%). thentic sarnplelzd by comparison of X-ray difThe original preparation of Ib hydrochloride (B“) was fraction patterns obtained from the two samples repeated but with distillation of two-thirds of the methanol the reaction mixture (optimum yields had not pre(Figs. 1-4), and by a characterizing analysis for from viously been given). Using purified starting materials methoxyl. Furthermore it was shown that no the yield of crude (but nearly pure) hydrochloride (of significant transetherification occurred when I b m. p. 144.5-146’) was 95%; of liberated and distilled ~ 85% from hydrochloride was treated with ethanol under base (of b. p. 98-100” (1 mm.), n z s 1.5157) Ia; of pure hydrochloride (precipitated from 2 ml. of these conditions. 1-1 methanol-dry ether per g . by slow addition of 0.75 New preparations of Ib hydrochloride by meth- N ethereal hydrogen chloride until acid to congo, and addianolysis of Ia hydrochloride were carried out be- tion of excess ether) (of m. p. 147.5-148.5”), 79% from Ia. A preparation involving four days of refluxing 79% _ gave _ cause Cromwell and Tsou3had reported that “11” crude product. is a secondary product in the reaction. On treat- yield-of Dimomhism of Ib Hvdrochloride. Crvstal form-A was ment of Ia hydrochloride with boiling methanol for obtained‘ from methanol with solvent of crystallization two or five minutes, reaction was negligible; after which is very easily lost. Freshly prepared samples airone hour of refluxing considerable unchanged Ia dried a t room temperature for a half-hour lost approxi8% of weight i n vacuo overnight at room temperahydrochloride could still be recovered ; however mately ture. (Anal. Calcd. for C1aH19NOz.HC1: C, 60.57: N , when a small amount of free hydrogen chloride 7.81. Found: C, 60.52; H , 7.93.) Vacuum-dried was present, reaction was a t least half Completed samples exclusively were used to obtain X-ray diffraction after five minutes of refluxing, but unchanged ma- patterns. The compound crystallized from isopropyl alcohol in a different and solvent-free form (crystal form-B) terial was still present in very considerable which gave a different and characteristic X-ray diffraction amounts. pattern (Fig. 1). (Anal. Calcd. for C13Hl~N02.HCl: Ultraviolet absorption spectra of I b and IChy- C, 60.57; H, 7.81. Found: C, 60.95; H, 8.03.) Conin methanol showed slight but pro- version back into crystal form-A was effected by crystaldrochloride~~ = 0.22 and lization from methanol (the sample was identified by nounced maxima a t 256 mp ( E X melting point and X-ray diffraction pattern). Melting 0.24, respectively) , and the curves obtained were points of the two forms were identical and no depression very similar in fine structure to the curves re- was observed upon admixing; however, there was obboth ~ ~for Ib hydro- served in the case of crystal form-A a visible change in the ported by Cromwell and T S O U of the sample, which occurred a t 78-81’ under chloride and for their “dimethoxy” compound 11. texture heating from room temperature, and which was inIt would therefore appear that the work of slow terpreted in terms of transition to crystal form-B. This Cromwell and Tsou in respect to the formation of conclusion was tested by annealing a sample of form-A a t the compound supposed by them to be 113 is er- 100” for fifteen minutes; an X-ray diffraction pattern subsequently obtained from the sample was identical with roneous. that from a sample of crystal form-B obtained by crystalliExperimental6$’ zation from isopropyl alcohol. The hydroxymorpholine (Ia) hydrochloride (dry) crysThe reliability of the mixture melting point identification of the hydrochlorides in this series is supported by the tallized in one and the same form from both methanol and unmistakable 5 ’ depression given by the hydrochlorides from isopropyl alcohol; the identity of the two samples of the methoxy compound (Ib) and its ethoxy analog (IC). was shown by the identity of the X-ray diffraction patThe melting points, since they varied slightly when taken terns. (The pattern from the sample crystallized from under different conditions, were determined simultane- isopropyl alcohol is shown in Fig. 6.) I t is noteworthy ously, using similarly pulverized samples, with the bath that there was no evidence here of etherification of Ia temperature initially 10’ below that of the final melting hydrochloride (or transetherification of Ib or IC hydrochlorides) upon crystallizations from methanol or ethanol. and the rate of heating 1-2” per minute. Slow crystallization from methanol-dry ether mixture gave small well-formed crystals melting s:mewhat higher (3) Cromwell and Tsou, THIS JOURNAL, 71, (a) 993, ( b ) 995, (c) 994 than previously reported; m. p. 128-128.5 . (Fig. 4) (1949). Catalyst Requirement for Conversion of Ia to 1b.-In a (4) Cj.Bergmann and Weil, Bcr., 68, 1911 (1930); c j . also ref. 2d. typical experiment a solution of 1 g. of Ia in 5 ml. of meth(5) Determinations were made by Mr. Spencer M. King using a anol and 5 drops of saturated methanolic hydrogen chloBeckman type DU spectrophotometer. ride was refluxed for five minutes; the hydrochlorides were ( 6 ) Melting points are corrected. precipitated slowly portionwise by additions of dry ether. (7) Microanalyses w u e by Clark Microanalytical Laboratories, and by Dr. E.0. Cogbill and Mn, Anne Wilgur of thh Laborstor% The first crops comprising 35’% of the material were idenCZH5
P\
C2He
I
1
NOTES
Narch, 1950
Fig. I
1411
Representative X-ray dieraction patterns of crystalline hydrochlorides. Fig. 2 Fig. 3 Fig. 4 Fig. 5
-.
I b-1.01n - B Cryst. from isopropyl alcohol
..
Cryit. from
Sillll,,k S". 2. (blixt. of Ia and b),
CHIOH
Cryst. from CHiOH
Smlple s o 1.
tified by mixture melting wints and X-ray diffraction patterns as nearly pure Ib hydrochloride. The next Iraction (sample no. 2) was shown to be a mixture of Ib and l a hydrochlorides by X-ray diffraction patterns (Fig. 5 ) . In similar experiments without added hydrogen chloride Ia was recovered and was identified in the form of the base bv mixture meltinn ooints: and similar exmriments (withbut added hydr& chioride) were ckried out with similar results on la hydrochloride which had been freshly precipitated from an ether solution of la by hydrogen chloride. washed with dry ether and used dirrctlv without drvinn (these samdes had retained v e r ~ littleif any free hidrkhloric acid): The representative X-ray diffraction patterns given in Pigs. 1-6 are contact prints and were obtained by the Dowder wedxe techniaue in a cvlindrical camera with a i.16 cm. r a d k exposid to radiaiian from a copper target . X-ray tube wit h a nickelo~idefiltcrgivingwentiallyCu K radiation. COBS CHEMICALLAB. UNIVERSITY OF VIRGINIA CHARLOTIESVILLE. VIRGINIA RBcsrvgo A m 12. 1949
Fig. 6
ll,-v"r,,,h\
Iit
reported by the earlier investigators to warrant publication. Several methods of preparation were tried. Application of the classical esterification procedure, i. e., refluxing diethylene glycol with glacial acetic acid in the presence of concd. sulfuric acid, yielded no diacetate. The diester, however, was obtained by a Schotten-Baumann reaction using diethylene glycol in the presence of pyridine with either acetic anhydride or acetyl chloride. Diethylene glycol monoacetate was obtained in all cases, but it was not isolated in pure form. It has been found that acetic acid, formed during the preparation, must not be removed from the diacetate by alkali extraction for the diester was found to be extremely sensitive to bases. Thus, if a solution of the diester is shaken with aqueous potassium hydroxide (3 N)a t room temperatures for a short time, the diester is hydrolyzed to the monoester. It was not found possible to separate the monoThe Preparation and Properties of Diethylene and diacetates very well by fractional distillation Glycol Diacetate under vacuum as they formed a constant boiling BY SHOW CHULIANG,'R. W. WALKERAND N. R. TRENNBRmixture. The diacetate could be isolated, howI n connection with another investigation, a ever, by extracting with ether from an aqueous quantity of pure diethylene glycol diacetate was solution of the mixture of the two esters. The required, and consequently we followed the work monoester is more hydrophilic and remains in of previous a ~ t h o r s . ~An . ~ examination of the the aqueous phase. Diethylene glycol diacetate is hygroscopic. properties of the products so obtained revealed When pure and dry, i t melts at 18". On standing that they were of unsatisfactory purity, whereupou the more satisfactory methods described in air, the melting point gradually lowers as the below were developed. The physical properties diester absorbs moisture from the atmosphere of the pure diethylene glycol diacetate thus ob- and finally drops t o below 0". Quite unexpectedly, the diacetate is completely tained were found sufficiently differentfrom those miscible with water. This is very different from ( I ) Prercnf address: 31 I.smbtoo R o d . Oltsr.. Cnnmdn. the behaviors of structurally similar compounds (2) L. 11. Crelchcr i o d W. 11. Piltcocer. Turs JOURNAL. Clo165 such as ethyl acetate, diethyl ether and 1,4(1025). butauediol diacetate. (3) M.M*eleod. 1. C h m . Sw., 3082 (1928).
