016
FRANK E. YOUNG
A
B
C
Fi’g. 4.--Potmtinl curvw for condensed gases (curves A, :ind C arc for successive stngcs of compression).
n,
in@;is also complete at about 70°K. There is thus littrle difference in cooperation between the four l a y m and the bulk solid. As the coverage is lo\rwed, the temperaturo at which the line narrowing is complete falls until a t the monolayer the temperature is about 21°K. As the coverage is further lowered, this temperature starts to increase slightly and continues to increase slightly until about 0.3 monolayer. Experimental difficulties make it hard to get data below this coverage. This increase of temperature below the monolayer indicates an increase in co6peration which can only be due to an increase in density due to compression. illthough the line shape changes with covrrngc and the range of temperature for broatfen-
Vol. 61
ing of the line to the maximum also varies, the general conclusion from the above result i R not changed. This is a striking confirmation of the theoretical conclusions reached in the foregoing discussion. The fact that titanium dioxide has other sites than those of the flat surfacee does not affect the argument since more than half of the sites in the monolayer are those of a flat surface. The methane in the high energy sites could only change the line shape slightly and would not affect greatly the de,gree of cooperation. Previously Mastrangelo and one of us have shown18 that there is a sharp inflection in the 2.41 OK.isotherm of helium on titanium dioxide a t the end of the second layer. This was attributed to a decrease in surface pressure occurring rather suddenly in the formation of the sccond layer. It was pointed out that such an cffect implied that the density of the film was relatively constant throughout a layer. It was thus concluded that “the film consists largely of regions of the surface completely .filled.” Acknowledgments.-The cooperation of Dr. William A. Bteele in the many discussions of the material presented in this paper is gratlefully acknowledged as is also the helpful criticism of Dr. Julian Eisenstein of the Department of Physics. (13) 8. V. (1981).
R. Mastrangelo and J. Q. Aston. J . Chern. l’huo., 19, 1370
D-GLUCOSE-WATER PHASE DIAGRAM BY FRANK E. YOUNG’ Contribution from Western Ulilizalion Research Branch, Agricultural Research Service, U . S. Department of Agriculture, Albany 10, California Received Novembrr $6, 1068
The n-glucose-water phase diagram ha6 been investigated between -30 and +02”. Solubility curves have been eetablished for all of the known solid phases: ice, a-D-glucose, p-n-gliirove and (Y-D-lucose monohydrate. The ice-a-n-glucose monohydrate and ice$-n-glucose eutectics have been observed 4yerimentalfy. Mutarotatory equilibrium between CYnnd p-isomers in D-glucose solubions i A found to be noarly independent of temperature but may be affected by concentration.
The discovery of new crystalline hydrates in recent studics of phase equilibria in the sucrosewater2 and n. fructose-watera systems, the practical importance of the crystallization of a-nglucose monohydrate i n fruit pre~ervation,~ and inconsistencies in available data on the D-glucosewater system prompted this new investigation of phase equilibria in aqueous D-glucose solutions. The temperaturcs at which ice and n-glucose solutions are in equilibrium have been reported by Abegg6 and by Poxner and Amerkhanov.6 The results of these two studies are in good agreement for solutions more dilutc than 23% hy weight of anhydrous sugar, but a t higher concentrationfi significant differences occur which increase witfh sugar content. Abegg’s ice curve, based on meas( 1 ) Callfornia State Polytechnic Institute, San Luis Oblspo, Calif. (2) F. E. Young and F. T.Jones. T H ~JOURNAL, S 63, 1334 (1949). (a) F. E. Young, F. T. Jon08 and H. .l. LewiB, ibad., 66, 1093 f1052). (4) R. W. Olmn and E, L. Moore, Food Tech., 8 , 175 (1954); G. b l . Cole, {bid., B, 38 (10551. (6) It. A. Ahcan. A. phyaik. Chem., 16, 200 (1894). (6) E. Pozner and A, X, Amerklinnov, J . I’hycl. Chem,, U,R.R.R., 16, 1137 (1041).
urctments on solutions containing up to 42y0 gluOORC, is essentially the same as that found for fructose solutions.a On the other hand the Russian investigators found higher ice points at high sugar contents, the result a t 71% glucose being alwut 12” higher than the ice point of a fructose solution of the same concentration and only 1.5” lower than 7101, sucrose solution. Soluliility data for a-D-glucose and its monohydrat)e have been reported by Jackson and Silsbee.’ Their two measurements a t 28” for the an. hydrous sugar arc in poor agreement with the value ohtained by cxtrapolatiori of their other data obt,ained a t higher temperatures. No solubility data have been reported for 8-D-glucose. New determinations of the solubility curves of ice, a-D-glucose, @-r,-glucoseand a-D-glucose monohydrate are reportcd here for the range from -30 tjo + 6 2 O . These curves have established new values for the eutectics of this system. No crystal(71 R. F. Jackaon and C. G. Silshoe, Nall. Bur. Slandards Sei. Papers. 17, 715 (1922)(No.437).
*
D-GLUCOSE-WATER PHABE DIAQRAM
May, 1957
611
usually aft,er 20 to 30 minutes, thP Audden evolution of heat caused a momentary incrrage in roncentration before crystallization rendered the solutiou too pasty to filter. Each result obtainod I)y this technique was chcckod by repenting the procedure with a solution whose initial concentration was the equilibrium value indicated by earlier mmsurements. It was not feasible to check these results with initially oversaturated solutions, because crystallization of the rnetastable solid phase proceeds so slowly that equilibrium is not Experimental attained before crystallimtion of a stnl)le solid phase begins. Materials .-Measurements were made on solutions which Frequent spontaneous crystallization of glucose monohad hecn allowed to stand overnight at room temperature hydrate and extremely slow mutarotation made it imprerto reach mutarotatory equilibrium. These solutions were tical to hold these solutions long enough to ensure mutaprepared from distilled water and reagent rade a-D-ghlcose rotatory equilibrium a t the temperature of measurcment which gave [u]% 52.79" in water (c 4, f4). At several Separate experiments were performed to determine how points, the results were checlred on solutions of National much the solubilities of anhydrous D-glucose measured as Hureau of Standards dextrose Standard Sample No. 41. just described were in error due to lack of attainment of p-u-Glricose was prepared from the above a-D-ghCOSC by equilibrium between the a- and p-isomers in solution. For tlia metshod of Hudson and Dale,s after which i t was dried this purpose ammonia was used to catalyze mutarotdon. in vacuo over anhydrous crtlcium chloride to remove alcohol Measurements first were made to doinonstrate that amand water. Since the solubilities of both a-and fl-D-glUCOSe monia is effective in speeding mutarotation a t low temperarcinninetl unchanged when the ratios of solid phase to solu- tures. A D-glucoAe solution to which 2 drops of concention were increased, t,he amounts of impurities remaining trated ammonium hydroxide had been added was allowed to in these materials were too small to affect the results. e uilibrate at -10.65O in the presence of solid a-D-glucose. Analysis.-Concentrations were determined by refrac- T%o solution was then diluted 0.40% and again allowed to tometer readings a t room temperature. These were con- equilibrate in the presence of the solid phase. The concenverted to n-glucoee percentages by the table of Young and tration quickly rose to 50.25%, its value before dilution. Jones.@ All concentrations in this paper are expressed in Repetition of t,he dilution and equilibration in the presence of solid a-D-glucoRe produced the same concentration. Had grams of anhydrous D-glucose per 1 0 g. of solution. Isothermal Measurements.-The solubility curves for mutarotation not occurred rapidly in the presence of ama-D-glucose, its monohydrate and ice in the regions in which monia, each successive dilution would have reduced the thcy are the stable solid pha.ses were determined b equilib- final value by an amount corresponding to the reduction of rium mcaflurements in constant temperature batts as de- the P-D- lucose concentration. mibed previously.* Equilibrium was not considered atSolubhties determined by this procedure were higher for tained until the concentrations of solutions which had n-D-glUCOSe by 0.35% at -10.65" and by 0.10% at 0', and npproached equilibrium from both above and below satura- lower for 8-D-glucose by 0.25%. a t - 16.80", than those tion differed by less than 0.10% and remained constant for measured without adding ammonia. Tliese figures indicate several days. Agreement of measurements for initially the equilibrium between a- and &isomers is not shifted much under- and oversaturat,ed solutions also ensures attainment by rather large chan es in temperature, a conclusion in acof mutarotatory equilibrium when the solid phase is not cord with data of Istell and Pigman!* at 0 and 20" for 8 ice. The final result for each temperature represents the 4% D-@UOOSe solution. The magnitude and direction of the average of a t least 4 measurements. Although several shifts indicated by our solubility data are also in agreement succesaful measurements were made by this method on the with their rotation data. The small corrections indicated metastable portion of the ice curve, down to -go, the fre- hy these figures have been applied to solubility data obtained quent spontaneous crystallization of a-o-glucose monohy- for a- and p-D- lucose below 20" in the absence of ammonia. drate made it necessary to investigate the remainder of the P.t 40" or higaer, mutarotatory e uilibrium was attained phase diagram by other methods. before measurements were complete%. Warming Curves.-Points on the ice curve were alRo deThe solubility measurements on anhydrous e- and 8-Dberminetl hy the warming curve method* over the entire glucose a t or below 40" wcre made on about 100 g. solution range reported hew. Except for t,he point at -29.55' each ill a 25 X 200 mm. test-tube in a constant temperature bath. result is the avcragr of two or more determinations. That The solutions were agitated with a rotary stirrer and temice poin'~above -9" ohhinod from warming curves are not peratures were read on a ralibrated thermometer raduated i n error from lack of time for mutarotation is shown by their to 0.1". Above 40". solution and excess solid Rase were agreement with those det8ermined by isothermal measurtr- rapidly tumbled end-ovcr-end in partly filled viafs in a conments. That ire points ut lower temperatures are also not stunt temperature bath. affected I)y mutarotation wtts indirat
