1 i7S TABLE I CRVSTAIMNE FORMS OF XYIJTOL Form Metastable Stablc hl p , "C 61-61 3 M-94 5 ,, 1 5l?J 1 549 Y 1 MF. 1 .566 3 1 521 1 551 Crystal system Monoclinic Rhombic 21' = 38 * 5" dxialangleand f 211 = 32 * 6" dispersion ( r > Y) weak ( r < Y) weak Optical character (+:I 1f i WESTERN REC~IONAL RESE.4RCH LABORATORY BUREAU OF AGRICULTURAL AND INDUSTRIAL CIIb.,MISTK\ AGRICULTURAL RESEARCH ADMINISTRATION IT. DEPARTMENT OF A4GRIOULTURE RECLIVOIJ ~ I A 11, Y 1943 ALBANY, CALIFORNIA
s.
Suggested Method for Replaceable Hydrogen BY HERMAX A . LIEBHAFSKY
Dilute sodium amalgams react rapidly neither with acetonitrile nor with water dissolved therein at low concentrations. If silicon tetrachlxide is added, however, the amalgam will, upon violent shaking, rapidly liberate hydrogen derived from the water initially present. Unfortunately, the reaction is not quantitative even under conditions chosen after considerable experimentation : the hydrogen is liberated in amounts ranging from one to two gram atoms for each mole of water.
TABLE KhPLacE ABLF U"
HYDROGP \T I Y SOlUtP
Watei Water Watcr Watcr Water Water Water Water Water Kthyl alcohol Ethyl alcohol Methyl alcohol
1
ACETONITRILE SOLUrIONS % Solute Cr, Solute added found
0,008
0 . 004
,015 030
010 019
045 ,060 ,075 ,I3 .13
03 1 ,045 ,055 .08 . (I8 .12 .17
. 1 :i , 13
. 13
. 13
,17 ,1'1
The method suggested here is subject to conplications. Silicon tetrachloride and sodium amalgam are both consumed in side reactions. 'I'able 1 shows that the stoichiometry is not simple; water gives lower results than the alcohols, but this is not surprisiiig because water (or an --OH grouf.') is the more likely to be retained in an inactive condition by the reaction products. The amalgam becomes covered almost instantly by a film, presumably siliceous, that bars i t lrom reacting when its surface is a t rest; this simplifies the manipulation. Sodium amalgam can be slowly oxidized by the air in the system; the occurrence of this reaction, the rate of which is usually negligible, c;m be turned to account as a signal that the hydrogen chloride has been consumed with amalgam still present. In spite of these complications, the proposed method may find its place where other methods are inapplicable or too slow. Ko doubt other substances that react like silicon tetrachloride could be used. One experiment showed that dioxane could replace acetonitrile. The method grew out of experiments being carried on here by Dr. C. E. Reed. 1 am grateiul also to Dr. K. 0 . Sauer for supplying the silicon tetrachloride, and to Alr. L.H . Rronk for helping with the de t erin ina ti om.
Samples of known water (or alcohol) content were prepared with acetonitrile that had been distilled from phosphorus pentoxide; the solvent gave a negligible blank. 50 cc. of such a sample was measured into a 100-cc. acetylation flask attached t o a mercury-filled gas buret. 'Ten drops of freshly distilled silicon tetrachloride was added with a dropper and dissolved by shaking the closed flask; 10 cc. of sodium amalgain (about 0.0570 by weight) was now poured in, and the closed flask shaken violently for three ten-second periods; a gas buret reading was iaken after each period had elapsed, the final reading being taken when temperature equilibrium had been re-estahlished. At water concentrations below 0.05% by weight, more hydrogen could not be liberated by this procedure when amalgam or silicon tetrachloride was subsequently added. A t the higher concentrations, these substances had to be added alternatcly in the amounts given above KESEAKCH LABORATORY until 110 more hydrogen was evolved. Some of the results GENEKAL ELECTRIC COMPANY obtaiiied are given in Table I ; here all percentages are by SCHEXFCTADY, ?;EW UORK K t i C s I V E D J U N h 7, 1948 weight, and water is assumed to have two replaceable . __. -. .. hydrogens. Two samples of acetonitrile in which U.013 atid 0.0770 The Preparation of p-Primeverose Heptaacetate water had been obtained with the Karl Irischer reagent by and 0-Vicianose Heptaacetate Mr. Stanley Mills, Niacet Chemical Company, gave only 0.003 aud 0.0Z70 by the present method when the results 13v CHESTER &I.~ ~ C C ~ . O S KA NEDYGEORGB ~ H. COLEMAN \vere calculated as in the last columii of the table. (An Recently in this Laboratory we have had occaempirical correction would, of course, improve the agreesion to prepare primeverose and vicianose in coninent.) \.\.'hether the Karl Fischer reageut reacted with >uhstanresother than water in these samples is riot knowii ( 1 ) Research I'ellow of t h e r o r n I'roductz Refining V u
