OF ORGANIC SUBSTANCES IN SOLUTION. 145 XVL--Action OJ Sugars on Arnmouaiaca I Silver Nityute. By JAMES HEN DERSOX, B. Sc., 1851 Exhibition Scholar, University College, Dundee. WHILE recently pursuing a research in which it was essential that very small quantities of sugars produced in the course of experiment should be estimated quickly and accurately, and as the requirements were such as to preclude the adoption of any of the various copper- reducing methods, the author was led to undertake this research, with the view of discovering a method whereby such an estimation could be satisfactorily accomplished. The principle of the method adopted in this case, like Fehling's, is based on the fact that alkaline solutions of certain metals undergo reduction when heated with certain oxidisible compounds, such, for example, as sugars.Tollens VOL. LXIX. M146 HENDERSON :- ACTION OF SUQARS (Be),., 1883, 16, 921) has shown that when dextrose is heated with smmoniacal silver nibrate, the number of atoms of silver reduced for each molecule of dextrose varies according as a great>er or smaller excess of the silver solution is employed, a3 shown in the following results which he obtained. Ratio of 1 mol. of dextrose to atome Silver taken, Silver reduced. of silver. 104.3 102.44 12.29 115.6 108.00 12.96 176.8" 147.39 1'7.69 The quantity of silver nitrate originally taken was thus only slightly in excess of that used up by the dextrose. Experiments conducted under the above conditions cannot be expected to yield very constant. results, as the amount of silver nitrate employed in each case is evidently insufficient to ensure the complete oxidation of the sugar to some definite final stage ; it appeared to me, however.that this could be readily effected by heating the dextrose for a suffi- cient length of time with a fairly large excess of silver nitrate. Preliminary experiments were accordingly made with a view to determine what influence, if any, time, and the presence of tt larger or smaller quantity of ammonia exerted on the reduction of the silver nitrate by the sugar. .A Holutioii of aninionia of sp. gr. = 0.88, diluted with three times its own volume of water, was added to a given volume of decinormal solution of silver nitrate until the preci- pitated oxide was just redissolved ; a knoivn volume of a standard solution of dextrose was then added, and the mixed solutions intro- cluced into a boiling tube, and heated by means of steam issuing freely from a small copper boiler, into the neck of which the boiling tube was slipped, the greater part of the tube being thus subjected to the action of the vapoui-. After heating the solution at 100' for a definite time, a ltnowii volume was taken out by means of a pipette, cooled, diluted with water, and the reduced silver collected on a filter, and thoroughly washed with distilled water; the mixed filtrate and washings were then acidified with strong nitric acid, and titrated with K/50 ammonium thiocyanate, iron alum being used as an indicator.This strength of ammonium thiocyanate solution was used throughout, except in estimating the reducing powers of cane sugar, dextrin, and starch, where a decinormal solution was employed.This mode of estimating the unreduced silver nitrate was invariably nsed throughout the whole of the research. The influence of time m i the reducing action of the sugar was first of all ascertained in the following manner. A known volume of a standard solution of Jt In presence of caustic soda.ON AMIIOSIBCAL SILVER KITRATE. 147 dextrose was added to 40 C.C. of decinoi-ma1 silver nitrate, containing ammonia just sufficient t o keep the precipitated oxide in solution. and the mixed solutions heated a t 100' in the manner alreadj- described. Five C.C. of the solution were taken oat a t various inter- vals, and the amount of unreduced silver nitrate estimated.Time of heating Dextrose. at looQ c. reduced. 20 milligrams. 0.33 min. 1.64 C.C. I 5.0 mins. 1-66 ,, Amount of AgNO, N,'10 This, then, would seem to indicate that with the above ratio of con- centration o€ dextrose to silver, the reducing action of the sugar ceases in about half a minute from the start. The next experiments were made with the view of ascertaining the influence of ammonia on the reducing power of dextrose. A known volume of a standard solution of dextrose was mixed with 20 C.C. of clecinormal silver nitrate containing 5 C.C. of ammonia solution, and the mixture heated a t 100'. The same volume of ammonia solution was added in every case, although the strength of the solution varied throughout the series from 1 : 3 to 1 : 7.Five C.C. were taken out at various intervals, and the excess of silver nitrate estimated as before. The final ratio of silver nitrate reduced to that remaining in the solution was the same in each case. namely, 1 : 2. Dextrose. Strength of tmmonia solution 1 : ' i 1 : 6 1 : 5 1 : 4 1 : 3 1 : 3 Time of heating. 0.33 mine. 7'25 ,) 12'41 ,, 0'41 ,. 4-00 ,, 13-05 *, 0.33 ., 0.70 ,, 5 ' 5 8 :. 5-41 ,. 14-58 ,, i 27 -08 ,, BgXO, Nil5 reduced. 1-92 C.C. 1.64 ), 1-46 ,, 1-68 ,, 1.46 ,, 1-64 ,, 1.53 ,) 1-53 ), 1.55 ,, 1-50 ,. 1-49 ,. l . G i ), 1'52 ,, 1.45 ), 1-44 ,. The resnlts obtained above are sufficient to denionsirate the ratliev interesting fact that, provided the concentration of silver nitrate to tlextrose remains t h e same throughout the series, the strength of the ammonia solution employed does not appreciably influence the amount M 2148 HENDERSON : ACTION OF SUQARS i Time of heating.of reduction of the silver nitrate ; and further, when the above con- centration is adopted, it would appear, from the numbers obtained, that prolonged heating of the solution affects the reducing power of the sugar to a very limited degree. The next series of experiments was undertaken with the object of ascertaining the effect produced by varying the amounts of dextrose, the quantity of silver nitrate taken being the same in each experiment. A known volume of dextrose solution was added to 20 C.C. of decinormal silver nitrate containing 5 C.C. of ammonia solution, made by diluting ammonia, sp. gr. 0.88, with thrice its volume of water.The solution was then heated at looo, 10 C.C. taken out at intervals, and an estimation made of the unreduced silver nitrate in that volume o l solution. Ratio of dextrose taken to A:!zcz!10 silver nitrate reduced, expressed in milligrams. Dextrose. 10 -0 milligrams 10.0 ,, 7.0 , Y 10’0 ,, 4 - 0 :, 17% ,, ----__.----- 30 C.C. 6 *65 C.C. 1 : 3’51 30 Y Y 6-64 ,, 1 : 3.51 30 9 , 4.60 ,, 1: 5.52 30 Y, 6 *65 ,) 1 : 3-51 30 ,? 2-60 ,, 1 : 10 ‘53 40 9 , 11’73 ., 1 : 2’42 Prom these results we dednce the fact that, provided the amount of silvernitrate takenis the same in every case, the arnoiint of reduc- tion of the silver nitrate present in the solution is proportional to the amount of dextrose present. From the quantity of silver nitrate reduced by a given volume of a standard solution of dextrose, we am able to calculate the “ factor ” for ammoniacal silver nitrate, that is, the number of molecules of silver nitrate which are equivalent to 1 mol.of dextrose. The following series of experiments was undertaken, with the object of ascertaining the true value of this factor. I I I Ratio of silver AgNo, N/lO. AgNO, (N/10) Dextrose‘ 1 nitmte reduced to I reduced. 1 that remaining in solution. Factor. -- 11 -97 11 -95 11-97 11 -80 11 5% 12.06ON AMMONIACAL SILVER NITRATE. 149 AgN03 (NilO). Five C.C. of ammonia solutionof the same strengthas before (1 : 3) were mixed with 30 C.C. of decinormal silver nitrate, and to this was added a definite volume of a standard dextrose solution. After heating a t 100" for eight minutes, the whole solution was cooled, diluted, and the reduced silver filtered off, the excess of silver nitrate being esti- nisted in the usual manner.The dextrose used in the above experiments was a pure specimen obtained by recry stallisation from water. With ordinary commercial glucose, the following results were obtained. The following results were obtained. Ratio of silver AgKO, (Silo) nitrate reduced to reduced. that remaining in Factor* solution. Glucose. --- 17 *5 milligrams 10'0 )) Taking the mean of the results of the first series, we obtain the factor value 11-9. Soxhlet (J. pr. Chem., [2], 1880, 21,227) has clearly established the fact that in the case of the reduction of Fehling's solution by dextrose the dexlrose equivalent has by no means a constant value, but, on the contrary, is entirely dependent on the particular circumstances under which the reduction is effected.The causes chiefly affecting the reduction of the copper solution in the case of dextrose are dilu- tion? concentration of solutions, and the time of heating. As a consequence of this, therefore, when estimating dextrose in different, solutions, the results are comparable only when the experiments are conducted under precisely the same conditions of dilution, &c. As has already been shown in the case of the reduction of silver nitrate by dextrose, the influence of time ceases after 5-10 minutes ; more- over, it would appear that, provided the solution be heated for a sufficient, length of time, namely, 8-10 minutes, and prorided also that the ratio of concentration of the reduced silver to the silver left in the solution does not fall below 1 : 2, the amount of reduction of the silver nitrate is dependent solely on the quantity of reducing sugar present in solution. This at once suggests a method for the quantitative estimation of dextrose in dilute solutions.A series of estimations was next pereormed for the purpose of t,esting the accuracy of the value obtained for the dextrose factor. 2.5 C.C. ammonia solution of the usual strength (1 : 3) were added to 10 C.C. of decinormal silver nitrate, a known volume of standard dextrose solution was then added, and the solution heated at 100' for150 HENDERSON : ACTION OF SUGARS Milligram taken. 5.0 5 . 0 10 n o eight minutes. The reduced silvei- was next filtered off, and the excess of silver nitrate determined by titration. The fsctor used was that obtained in the previous experiments, namely 11.9.Milligram found. 5 *o 4.9 9 . 4 Dextrose. Cane sugar. AgNO (N/10) reduced. dgNOs @/lo). Time of heating. Ratio of silver nitrate used to that remaining in solution. 1 : 3.07 1 : 3.2 1 : 1.59 AgN03 (N,/10) reduced. 3.32 C.C. 3-25 ,, 6-23 ), Evidently, therefore, with a ratio rarying from 1 : 2 to 1 : 3 fairly accurate results can be obtained. CYnrLe Sugar. A pure specimen of cane sugar was prepared by recrystallisation from ethylic alcohol. A known volume of a standard solution of the pure sugar was heated at 100' for a certain length of time 1vit.h 20 C.C. of decinorma1 silver nitrate and 5 C.C. of ammonia solution (1 : 3) ; at the conclusion, the solution remained perfectly clear, exhibiting no signs of reduction of t,he silver nitrate, subsequent titration of the whole solutlion proving that no action had taken place.I I 160 milligrams 20 9 , 10 mins. 1 25 9 ) A dilute solution of cane sugar is therefore incapable of effecting the reduction of ammoniacal silver nitrate when the above concentratioii is used, but if stronger solutions of the sugar aye used, and more especially if the mixed solutions are heated for a longer period, the cane sugar begins to be slowly oxidised at the expense of the silvey nitrate ; this in all probability is due to the slow hydi-olysis of the disaccharide molecules by the alkali, sugars being formed which of course are capable of being oxidised by silver nitrate.This secondnr.y action is even more marked in the case of Fehling's solution ; at first no signs of reduction are observable, but, on continued boiling, they very soon begin to manifest themselves. This inability on the part of the silver solution to effect the oxidation of thc cane sugar may be ntilised as the basis of a niehhod for the estimatioii of dextrose in presence of cane sugar.ON AMMOSIACAL SILVER NITRATE. 151 Milligrams klien. 6.04 grams of dextrose and 4.063 grams of cane sugar were dis- solved in water, and the solution made up to a litre. A given volume of this solution was then heated a t loo3 with a known volume of decinoimal silver nitrate and 2.5 C.C. of ammonia solution (1 : 3) for a, certain time, the reduced silver filtered off, and the excess of silver nitrate determined in the usual wa,j-.Milligmms fourd. Dextrose. A gNO, (NilO) AgN03( N/10) reduced. 4- *85 C.C. 3-26 ,) 3-30 ,, 3.15 ), 3.w ,, Time. 8 inins. l2 J? 18 ) > 23 J ) J J Ratio of silver nitrate reduced to that left in solution. 1 : 1-06 1 : 2.1 1 : 2'1 1 : 2 '03 1 : 4'5 The presence of cane sugar in the solution appears to retard the oxidising action of the silver nitrate to a slight extent. A quantity of silver nitrate sufficient to oxidise a known amount of dextrose to a certain stage in a given time is incapable of doing so in presence of cane sugar in the same time ; this retarding action may be remedied either by increasing the time of heating, or perhaps better by increw- ing the amount of silver nitrate.Dextrin urzd Stas.ch. A specimen of ordinary dextrin was purified by dissolving it i n water, filtering, and precipitating with 90 per cent, alcohol ; the pre- cipitate was then washed with strong alcohol and dried in a vacuum desiccator. The reducing power of each substance was then estimatect by heating a definite volume of the standard carbohydrate solution for eight minutes at loo3 with 10 C.C. of decinormal silver nitrate and 2.5 C.C. of ammonia solution (1 : 3). AgNO, (N/10) reduced. AgNO, (NjlO). Dcxtrin : 16 -2 milligrams , . . . . . . . Starch: 4.05 ,) ~ ~~ Hence dilute solutioiis of dextrin and starch, like cane sugar, do Reducing sugars can be esti- not reduce ammoniacal silver nitrate. mated, therefore, in presence of either or both of these compounds.152 Lsvulose.,-- 10 milligrams 10 ,. 10 7, 10 11 HENDERSON : ACTION Ol!' SUGARS AgN03(N/10). *gNo3 (N/lO) Time Of reduced. heating. ---- ---- -_- 30 C.C. 6 -38 C.C. 8mins. 30 ,) 5-38 l l 15 ), 40 9 1 5 '57 1, 8 ,, 60 11 5-80 J l 8 9 1 Scheriiig's Lev 14 lose. The method used in determining the reducing power of this sugar was identically the same as that adopted in the case of dextrose. ~ I &]actow. AgNo~(Njlo) Time Of reduced' heating- I I Ratio of silver 1 nitrate reduced to 1 Factor. I that left in solution. Ratio of silver nitrate reduced to that left in solution. --- 1 : 4-57 1 : 4 '57 1 : 6.17 1 : 9.34 10 d milligrams 10'0 ,, 7.0 11 - Factor. -.- ----'-- 6 -73 C.C. 1 8 mins. 1 :3*4 12.11 I 1 : 3.4 12.07 4-70 l . 1 : 5-37 12 '08 9 -7 9.7 10 '0 10 -4 As with dextrose, the reducing power of levulose is but slightly influenced by the time of heatinK after eight minutes, Galactose.A pure specimen was obtained by recrystdlisation from ethylic alcohol. A standard solution of the pure sugas was then prepared, and a given volume heated at 100' with 30 C.C. of decinormal silver nitrate and 5 C.C. of ammonia solution (1 :3). After a certain interval, the reduced silver was filtered off and the excess of silver in the solution estimated as usual. The resalts of these experiments were as follows. - ~~ Factor = 12.08 (mean). As in the case of dextrose and levuloee, time exerts little or no influence on the reducing power of galactose after eight minutes' beating at 100'. It, will be obseived that the mean of the values obtained for the factor is almost equal to that of dextrose (11.9).Lactose. Tbe sugar used in these experiments was obt.ained by recrystallisa- tion from aqueous alcohol. A standard solution of this pure sugar was prepared and its reducing power estimated as in the previous cases.ON AMRIONIACAL SILVER NITRATE. 153 40 C.C. 40 Y , 4Q 7 9 40 9 9 40 9 , ~~ ~ Lactose. 2 T O C.C. 4.19 ,, 4.60 ,, 5 - 3 0 ,, 6-75 ,, AgNO, (N110). -- 8mins. 1 : 13.80 16 ,, 1 : 8-54 20 ,, 1 : 7.69 25 ,, 1 : 6.54 30 ,, 1 : 4.92 AgNO, (N 110) reduced. Time. Ratio of silver nitrate reduced to that left in solution. Factor calcula. ted for Qm 9 -72 14 -76 16 5 6 19-08 24 '30 ~~ ~~ ~ ~ Maltose. The specimen of maltose used in these experiments was prepared in the following way.A saturated solution of the sugar in alcohol was allowed to remain for more than a fortnight in a corked flask ; the solution was then filtered from the maltose which had separated, and the alcohol allowed to evaporate slowly in the air. A standard solution of the sugar thus obtained was then prepared, and the reducing power estimated in the usnal manner. AgNO, (N, 10) required. 0 -95 C.C. 1.27 ,, 7-25 y y 1-07 ,) Rat,io of silver nitrate reduced to that left in solution. 1 : 20 '05 1 : 17.6 1 : 14.69 1 : 4 . 5 Factor calc ula- tea for c12. 3 *40 3 -85 4'57 26 *lo Hence when either lactose or maltose is heated with ammoniacal silver nitrate under conditions sufficient to oxidise the same amount of any of the monosaccharides, dextrose? galactose, or levulose, to a definite final stage, only a very small proportion of the disaccharide becomes oxidised. The results of the time experiments in the case of lactose and maltose would lead 11s to suspect a gradual hydrolysis of the sugar by the ammonia as the beating is continued. The results obtained may be thus sumniarised. (1) When dextrose, levulose, and galactose are heated with am. moniacal silver nitrate under the given conditions, a definite fzctor can be obtained in each case. (2) Cane sugar, starch, and dextrin, when heated under the same conditions, exert no reducing action on ammoniacnl silver nitrate.154 TILDEN AND BARNETT: THE NOLECULAR WEIGHT (3) In tbhe case of lactose and inaltose a definite factor cannot be got, owing to the gradual hydrolysis of the disaccharide molecnles by the ammonia. Uizhersity C d e g e , Dzwdee, Deceinbes, 1895.