1410 FRANRLAND AND MACGREGOR : NORMAL ETHEREAL CI1I.-The Normal Butylic, Heptylic, and Octylic Ethereal Xalts of Active Glyceric acid. By PERCY FRANKLAND, Ph.D., B.Sc., F.R.S., and JOHN MACGREGOR, MA., late Forster Research Scholar in Univereity College, Dundee. IN a previous communication (this vol., 511), we hare described the preparation and properties of the methylic, ethylic, propylic, iso- propylic, normal and secondary butylic, and isobutylic derivatives of both inactive and active glyceric acids. We there showed that the rotation was increased step by step in passing from the methylic to the isobutylic compounds ; but that the normal butylic glycerate exhibited a rotation not only inferior to that of the isobutylic, but eren lower than that of the propylic or isopropylic compounds.This anomalous behaviour of the normal butylic compound we attributed in part to the obvious impurity of the product (as proved by range of boiling point and unsatisfactory figures on analysis), but also, in part, to the rotation of the normal butylic glycerate being really excep- tional. The anomalous rotation exhibited by the product in question has led us not only to renew our endeavours to obtain the normal butylic glycerate in a state of purity, but also to prepare two moye extreme members of the normal series of ethereal salts, namely, the lieptylic and octylic glycerates, for the purpose of comparison, Our re-preparation of the normal butylic salt, which we have now ob-SALTS OF ACTIVE GLYCERIC ACID. 1411 tained in a state of purity, as well as the preparation, for the first time, of the heptylic and octylic salts of active glyceric acid, form the subject of the present paper.Active Normal Butylic Glycerate, OH*CH2* CH (OH) *C 0 0 *C H2*C El2* CH,.CH,. I n order t o insure the purity of this salt, we commenced by very carefully fractioning the normal butylic alcohol obtained from Kahl- baum, until a product was obtained of uniform boiling point, 116-5-117*8", and of density unalterable by further fractionation, namely, 0*8157-0.8159 at 15"/15". 33 grams of active glyceric acid were then digested with an excess of this purified normal butylic alcohol during eight hours a t about 160" in a sealed tube. The contents of the tube were fractionally distilled in a vacuum on an oil bath. 21.44 grams of crude salt were obtained, exhibiting an observed rotation (a)D = -13.40' at 17" (in 99.2 mm.tube). This product was dried with calcium chloride, and again fraction- ated in a vacuum; the principal fraction, weighing 1684 grams, had an observed rotation (a)= = -28.55" at 18" (in 198.4 mm. tube). This was refractioned in a vacuum, and the principal product now had an observed rotatiou ( a ) D = -14*50" at 17" (in 99.2 mm. tube). On again fractionating in a, vacuum, the principal product exhibited exactly the same rotation, indicating that the maximum degree of purity had been reached. The density at 15"/15" was found t o be 1.1084, from which follows the specific rotation -13.19 (at 17"). - - 14.5 = 0.992 x 1,1084 - The boiling point of this compound was 138.5" a t 16 mm.pressure. The purity of the substance was further demonstrated by hydro- lysis with alcoholic potash. Parts of KOH required to hydrolyse 100 parts of substance. (1) ........................ 34.88 Theoretical for butylic glycerate 34.57 (2) ........................ 34-741412 FRAXKLAXD AND RlACGREGOR : NORMAL ETHEREAL On combustion, 0.2789 gave 0.5272 CO, and 0.2186 H20. C = 51-56 ; H = 8.71. C7H1,04 requires C = 51.85 ; H = 8.54 per cent. Active Hepfylic GZycernte, OH*CH,*CH(OH)*COOC7H,,. Some difficulty was experienced in preparing this and the octylic salt, in consequence of the heptylic and octylic alcohols not dissolving glyceric acid, so that the alcohol and acid were only in very restricted contact with each other in the sealed tube, leading, of course, to a very much retarded interaction of the two substances.This difficulty was successfully overcome by placing the tube oven in which the sealed tubes were heated on a rocker, so that the contxnts of the tubes could be frequently shaken up during the digestion. The active glyceric acid was digested with an excess of heptylic alcohol (b. p. 175-177"), in sealed tubes, at 150-160" during six hours. The contents of the tubes were then submitted to fractional dis- tillation in a vacuum ; the principal fraction was dried with calcium chloride, and then repeatedly fractionated in a vacuum, until the principal fraction underwent no furt,her change, either in density or rotatory power, by further fractional distillation. I n this manner a product was ultimately obtained of density 1.0390 at 15"/15", and exhibiting a rotation (a)D = -11.65" at 1P (in the 99.2 mm. tube).from which follows the specific rotation -11.65 GdIc = o.992 1.0390 = -11.30 at 18". The boiling point was 173-175" under 14 mm. pressure ; tempera- The purity of this heptylic glycerate was ascertained by hydrolysis ture of oil bath, 210-220". - with alcoholic potash, and by combustion. Parts of EOH required to hydrolyse 100 parts of substance. (1) . . .. . . .. .. .. .. . . . . . . . . . . . - 27.93 Theoretical for heptylic glycerate 27.45 On combustion, 0.2242 gave 0.4810 CO, and 0.1982 H,O. C = 58.52 ; H = 9.82. 0.2120 ,, 0.4543 ,, ,, 0.1872 ,, C = 58.44; H = 9.81. 0.2925 ,, 0.6280 ,, ,, 0.2587 ,, C = 58.54; H = 9.83. C1,H2,0~ requires C = 58.82 ; H = 9-80 per cent.SALTS OF ACTIVE GLYCERLC ACID.141 3 111 order to determine whether the activity of the glyceric acid had remained unimpaired in the preparation of these salts, we performed the following experiment. 1.5671 gram of active heptylic glycerate was hydrolysed with an :Lqueous solution of barium hydrate ; the liquid was boiled to free it frjm the liberated heptylic alcohol, after which it was saturated with carbonic anhydride, and boiled to precipitate the excess of barium as carbonate. After filtration, the liquid containing the bariuin.glycerate was comentrated to definite bulk (18.4 c.c.), and the rotation ( X = -1.55) determined with the polarimeter ; the concentration was ascertained by means of a barium estimation in an aliquot part ( c = 7.045 per cent.), from which data follows the specific rotation of anhydrous barium glycerate The specific rotation of anhydrous barium glycerate has been pre- viously determined by one of us (this vol., p.299), as 10.01 for a solution of 10 per cent. concentration. The above result, therefore, clearly shows that the activity of the glyceric acid had undergone no change in the process of etherification. Active Octylic GZycerate, OH*CH,*CH(OH)*COOC8H1,. The octylic alcohol employed was obtained from Kahlbaum, but was fractionated before use, and only that portion boiling at 195-197" was employed. The active glyceric acid was digested with an excess of octylic alcohol in a sealed tube a t 145-155" during 10 hours. Owing to the immiscibility of the alcohol and acid, the rocking arrangement referred to in the preparation of the heptylic glycerztte was employed with advantage.The contents of the tube were subsequently re- peatedly fractionated in a vacuum, until the rotation underwent no change on further fractionation. The final product, which exhibited a faint fluorescence, distilled over at 181-183" (temp. of oil bath 285-229") under a pressure of 13 mm. of mercury. The density at 15"/15" was 1.0263, and the observed rotation in the 99.2 mm. tube, -10*40" at 19", corresponding to the specific rotation The rotation of this selt appears to be but very slightly aeected by1414 FRANKLAND AND MACGREGOR : NORMAL ETHEREAL differences of temperature ; thus the following rotations were observed at the temperatures indicated.[a],, = -10.40" at 17*2* = -10.40 ,7 19.5 - -10.4.5 ,, 20.5 - The octylic glycerate yielded the following results on hydrolFsis and combustion. Parts of KOK required to hpdroljse 100 parts of substance. (1) .. . . .. .. . . .. . . .. . . ,. .. .. . . Theoretical for o:tylic glycerate. . 26.17 25.69 I. 0.2000 gave 0.4414 CO,and 0.1803 H,O. C = 60.20 ; H = 10.01. IT. 0.1976 ,, 0.4371 ,, ,, 0.1795 ,, C = 60.32 ; H = 10.09. 111. 0.2238 ,, 0.4942 ,, ,? 0.2029 ,? C = 60.23; H = 10.08. C,,H,,O, requires C = 60.55 ; H = 10.09 per cent. On keeping a sealed tube containing some octylic glycerate in the cold, the contents were found to have become solid; the crystals, which were of a laminar form, melted at 22.5". The heptylic salt did not become solid under the same conditions.General Characters of the Ethel-eal Salts of Actice Gdyceric acid. In the following table (p. 1415), we have recorded the density and rotatory power of the several ethereal salts of active glyceric acid which we have up to the present prepared. The table exhibits the following relationships between the several members of this homologous series. 1. The densities continuously diminish and approach more arid more nearly to 1 as the molecular weight increases. The difference in density f o r each increment of CH, diminishes in ascending the series. These relations arc more conspicuously brought out in the following paper (see p. 1419) by means of curves indicating the densi- ties and molecular volumes of the several compounds. From the diagram there given, it will be seen that whilst the density curve tends to become asymptotic, the molecular volumes of the et'hereal salts with normal radicles lie on an almost straight line.The molecular t-olumes of the isopropylic and isobutylic salts lie on a straight line slightly above and parallel with the other. 2. The rotations exhibited by the members of this series are extremely interesting ; our reinvestigation of the rotatory power of the normal butylic, coupled with the additional information concern-Active Qlyceric acid. Ethereal salt. Methylic ................ Ethylic .................. Propylir (normal) ........ Isopropjlic .............. Butylic (normal). ......... Isobutplic ................ Butylic (secondary) ....... Heptylic ................ Octylic .................. Density, 15'1 15".--- 1 1 * 2798 l*lg21 1 *1448 i 1 -1303 1.1052 1.0263 - Difference. -- 0 *0877 0 *0473 0.0364 0 *0252 0 *0694 0,0127 Observed rotation in 198.4 mm. tube, (4 D. - 12 *2" -21 -7 -29.4 -26 *5 -29 '0 -31 '2 - 23 -2 -23 -3 -20 *8 , Sljecific rotation, Ca1D. -- - 4-80' - 9-18 - 12 *94 - 11 -82 - 13 '19 - 14 *23 - 10 '58 - 11 -30 - 10 *22 Molecular rotation, MCalD. 100 - 5-76 - 12 '30 - 19 '15 - 17 '4.9 - 21 '37 - 23 *05 - 17 -14 - 23 -05 - 22 '28 -27 -9 - 52 *8 -74 *9 - 67 -8 -77 *o - 82 *9 -61 -7 -68 *3 -62 '6 - ( P x 1 0 y u1 241 -8 * An explanation of [S]D and (? X lo6) is given later (scc next page).1416 FRANKLASD AXD BIACGREGOR : NORMAL ETHEREAL ing the optical activity of the heptylic and octylic salts, now place altogether beyond doubt the phenomenon which was already to be inferred from the data given in our former paper, namely, that the optical activity does not increase indefinitely in ascending such a n homologous series, but that a maximum rotation is rapidly reached, which is followed by a gradual decline as the molecular weight is further augmented.I n our series of homologous gljceric ethereal salts with normal radicles, the maximum is apparently attained in the specific rotation of the normal butylic compound, the activity of the lteptylic and octylic salts being distinctly lower, although the curve is very flat in this region (see diagram below). During the progress of this investigation, a paper by MM. Ph. Guye 2nd L. Chavanne (Crompt. rend., 116, 1454) has appeared, in which they endeavour to deduce from theoretical considerations this re- markable phenomenon of a maximum rotation in an homologous series of compounds.Thus, employing the figures contained in our preyious paper on the ethereal salts of glyceric acid, they make the following calculations for the series. Glycerates. Metliylic ............ Ethylic.. ............ Propylic (normal) . . , . Rutylic (normal). ..... I sobutplic ........... Propylic (secondary or Hutylic (secondary) ... iso-) Specific rotation, CalD- - 4-80 - 9.18 - 12 -94 - 11 '02 - 14 -2.3 - 11 -82 - 10 -5s Molecular rotation. - 5.7'6 - 12 '30 - 19 '15 - 17 -85 - 23 *05 - 17 -49 - 17 '14 [ a b -- -27.9 -52 *S -74 '9 -64.6 -85 -9 - 67 -8 -61 "i 2 39 345 358 3 47 358 (347) 347 It should be pointed out that in the above table [ S ] , stands f o r what Guje calls moZecuZar devia.tiou, and which is calculated from the formula in which a is the observed angle of rotation, L the length of the polarimeter tube, M the molecular weight, and d the density.The advantage claimed for this constant is that it compares the rotmations of lengths of liquid containing the same number of molecules ; P i n the above table, on the other hand, stands for the product of asymmetry, which may be calculated, if the four different masses (a, b, c, d ) attached t,o the asymmetric carbon atom are supposed to be concentrated a t the apices of the tetrahedron, by the formulaSALTS OF SCTIVE GLTCERIC ACID. 1417 In the figures calculated by Guye in the abwe table, the factor (I sin 0 1 ) ~ is neglected, whilst P is multiplied by lo6, in order to avoid fractions.Guye and Chavanne hare also prepared an honiologous series of ethereal salts, derired from an active valeric acid, and have made tho following similar calculations for them. Valerie acid. ................ Methylic valerate ............ Ethylic .............. Butylic (norm.) ,, ..... Isobutylic . . . . . . . B enzylic . . . . . . . Pi-opplic (norm.) ralerate ..... Cab -~ + 13 *64 + 16 -83 + 13 -44 + 11.68 + 10.60 + 10 -4s + 5-31 [Ax] D. - --- + 13 *91 + 19 -53 + 17 -47 + 16 -82 + 16 -75 t 16 -56 3- 10 -20 [SlD. --- +61 ,1 f 75 '5 + 61 *8 + 55 -2 + 51 -7 + 51 -0 1- 30.2 ( P % 106). I --- 218 1 374 364 351 I (351) I (221) i Thus, according to these calculated values of P, the predicted ma4ximum rotation falls on the normal propylic glycerate and on the etbylic Talerate for the two normal series of salts, whilst for the secondary series of salts, representatives of which have only been prepared in the case of glyceric acid, the predicted maximum falls on the isopropylic compound.The experimental maxima, again, as given in Ihe above tables, similarly fall on normal propylic glycerate for the normal series, and on isopropylic glycerate for the secondary series, whilst the experimental maximum falls on methylic valerate f o r the normal series. With the corrected rotation for normal butylic glycerate which we are now able to supply, however, the values for this salt become [a]D = -13.19; [&f]D = -21.37'; [6]D = -77.0, and are, therefore, the experimental maxima in the series.Thus, neither in the case of the glycerates nor in that of the valerates is the coincidence between the predicted and observed maximum rotation complete, for, in both series of compounds, the experimental maximum falls on the next higher term t o that on which falls the maximum deduced from the calculated value of P (product of asymmetry). Striking enough as are these very close coincidences between calculation and experiment, they require some criticism. Thus, in the first place, the series of valerates prepared by Gnye and Charanne cannot be regarded a s a very satisfactory one for a crucial experiment, inasmuch as the valeric acid employed was1418 NORMAL ETHEREAL SALTS OF ACTIVE QLYCERIC SCID. not a single substance, but must have been a mixture of an active and an inactive acid, as it was prepared by oxidatioii from an active amyl alcohol having a rotation [a],, = -4.40, whilst the purest active amyl alcohol prepared by Le Be1 (Bull.SOC. Chirn., [2], 21, 542) exhibited the rotation [aID = -5.6 to -5.7. But if the valeric acid employed consisted of a mixture of active and inactive molecules, it is surely open to very considerable doubt whether the active and inactive acids would etherify to the same extent or in the same pro- portions with the different alcohols employed. I n this connection, we need only refer to the phenomenon recorded by Le Bel, that active amyl alcohol and hydrogen chloride interact much less easily than inactive amyi alcohol and hydrogen chloride." Although, therefore, the values given above for the rotations of the valerates may be strictly proportional to the rotations of the unknown pure active valerates, the element of doubt which necessarily attaches to them obviously renders this series of compounds ill suited for testing the coincidence between theoretical calculation and experiment.The uncertainty attaching t o the results obtained for the series of valerates is em- phasised by the fact that Pierre and Puchct (Compt. rend., 116, 1456, foot-note) had previously obtained values for the rotatory power of these ethereal salts i n which the ethyl compound exhibited the maximum rotation. Our own series of glyceric ethereal salts is not open t o this sus- picion, inasmuch as all the members were prepared from pure active calcium glycerate, and we have proved, by actual experiment, t h a t the barium glycerate obtainable by the hydrolysis of the salt with barium hydrate has the same rotation as barium glycerate prepared directly from the active calcium glycerato with which we started. This homologous series of glyceric ethereal salts proves, beyond all doubt, that there is a maximum rotation, the highest rota- tion observed in the iiormal series being that of the butylic compound, whilst the rotations of the heptylic and octylic compounds are dis- tinctly inferior. As, however, we have not yet prepared the inter- vening normal pentylic and hexylic glycerates, it is impossible to sag whether the real maximum may not fall on one of these compounds, although, from the character of the curves given below, i t appears improbable that it should. The complete coincidence between calculation and experiment obtained in the case of the secondary ethereal salts of glyceric acid, based, as it is, on the examination of only two compounds (isopropylic and secondary bntylit; glycerate), caniiot either be regarded as more than a provisional confirmation of 31. Guye's suggestive theory. * A result recently confirmed by Rogers (this vol., p. 1130).