THE ANALYST. 167 THE VOLUMETRIC DETERMINATION OF REDUCING SUGARS. Part 111.-The Determination of Sucrose and Invert Sugar in Mixtures. BY ARTHUR R. LING, F.T.C.,ANL) THEODORE RENDLE. (Read at the Meeting, Jfarch 4, 1908.) IT is well known that sucrose, although by constitution a non-reducing sugar, does reduce Fehling’s solution to some extent when boiled therewith, and that in mixtures of sucrose and invert sugar, such as raw sugars, cane molasses, and cane syrups, it is necessary to apply a correction when estimating the invert sugar in these by any method involving cupric reduction. Tables giving corrections for the influence of sucrose in mixtures of the two sugars referred to have been drawn up by Meissl, Zulkowski, Herzfeld, and by others. These corrections, however, refer specially to the particular modifications of the cupric reduction method used in each case.Since, however, the determination in question is purely empirical, each modification of the method requires its own table of corrections. In the present paper we place on record the data necessary for correcting the results obtained by the method previously described by us (ANALYST, 1905, 30, 182). The invert sugar used in the xperiments was prepared by hydrolysing sucrose with hydrochloric acid in the manner previously described (Zoc. cit.). The sucrose employed was of high purity. The ash content was practically nil, and the reducing power, per se, equal to less than 0.04 per cent. of invert sugar.168 THE ANALYST. I n the following tabulated results : Column A gives the amounts in grams of sucrose present in 100 C.C.of the sugar Column B gives the percentages of sucrose present expressed on the total Column C gives the percentages of invert sugar present expressed on the total ColumnjD gives the number of c.c.'s of sugar solution required to reduce 10 C.C. Column E give8 the percentages of invert sugar on the total sugars found by Column F gives the differences between the values given in columns C and E. The values given under column D are in each case the mean of at least two solutions. sugars. sugars. of Fehling's solution. direct experiment. determinations differing by not more than 0.1 C.C. made by different observers. TABLE I. Each solution contained, in addition to the sucrose shown under column A, 0.15 gram of invert sugar per 100 C.C.No. of Exlieriment. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12, A. Nil 0.005 0.025 0.05 0.10 0.25 0.50 1 *oo 2-50 5.00 10.00 15-00 B. Nil 3.2 14.3 25.0 40.0 62.5 "(3.9 86.9 94-3 97 4 98.5 99.0 C. 100.0 96.8 85.7 75.0 60 -0 37.5 23.1 13-1 5.7 3.0 1-5 1.0 D. 34.70 34.70 34.70 34.60 34.45 34.20 33.75 33-05 32.85 31.90 31.25 29.75 E. 100.0 96% 85.7 75.1 60.3 37.9 23.7 13.7 5.92 3.26 1.67 1.17 F. Nil Nil Nil 0.1 0.3 0.4 0.6 0.6 0.22 0.26 0.17 0.1'7 The conwntration of the invert sugar solutions in Table 11. is that at which we prefer to work-namely, 0.2 gram per 100 c.c.--and the experiments recorded in this table are in consequence more numerous than those in the others. A critical examination of the results given in Table 11. shows that the influence of sucrose is practically negligible until the proportion on the total sugar amounts to 30 per cent.(see column B), at which point the invert sugar is over-estimated by 0.2 per cent. This influence of sucrose increases progressively in the same direction until the proportion, expressed on the total sugars, 99.3 per cent., is reached, beyond which it has not been determined (see tho Iast result of the series). At this point the invert sugar is over-estimated to the extent of about 15 per cent. I t mustTHE ANALYST. 169 be remembered, however, that the magnitudes representing the percentages of invert sugar decrease as those representing the percentages of cane-sugar increase, and it will be seen that the correction to be applied is actually the greatest when the percentage of mcrose on the total sugars is between 50 and 80.In the case of a mixture of equal parts of sucrose and invert sugar, the latter would be returned as 50.4 per cent., instead of 50 per cent.-that is to say, it would be over-estimated by 0.8 per cent. ; whilst in the case of a mixture of sucrose 99 parts and invert sugar 1 part, the latter would be returned as 1.14 per cent., instead of as 1 per cent.; or, in other words, it would be over-estimated by 14 per cent. TABLE 11. Each solution contained, in addition to the sucrose shown under column A, 0.2 gram of invert sugar per 100 C.C. No. of Experiment. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26, 37. 28. 29. A. Nil Nil Nil Nil 0.01 0.03 0.05 0.10 0.20 0.30 0.40 0.50 0.60 0.70 090 1-25 1 -50 1-75 1-75 2.00 2 *oo 2.50 3.00 5.00 7 .OO 10.00 20-00 25.00 30.00 B.Nil Nil Nil Nil 4.8 13.0 20.0 33.3 50.0 60.0 66.6 71.4 75.0 77.7 80.0 86.2 88.2 89 -7 89.7 90.9 90.9 92.5' 93.8 96.1 97.2 98.0 99.0 99.2 99.3 C. 100.0 100*0 100.0 100.0 95.2 87.0 80.0 66.7 50.0 40.0 33.4 38.6 25.0 22-3 20.0 13.8 11.8 10.3 10.3 9.1 9.1 7.5 6.2 3.9 2.8 2.0 1.0 0.8 0.7 D. 25.65 25.65 25.60 25-70 25.60 25.60 25.60 25.55 25.45 25.40 25.55 25.30 25.20 25.15 25.10 25-05 24.95 24-85 24.80 24.70 34.80 24.80 24.70 24.20 23.60 92.95 83-40 22.25 22-25 E. - - - - 9530 87.10 80.10 66.90 50.40 40.40 33.80 29-00 25.40 22-70 20-40 14.10 12.10 10.60 10.60 9.45 9.41 7.76 6.44 4.05 3.04 2.23 1-14 0.92 0.80 F. - - - - 0.10 0.10 0.10 0.20 0.40 0-40 0.40 0.40 0.40 0.40 0.40 0.30 0.30 0.30 0.30 0.35 0.31 0.26 0.24 0.15 0.24 0.23 0.18 0.12 0.10170 THE ANALYST. TABLE 111.Each solution contained, in addition to the sucrose shown under column A, 0.25 gram of invert sugar per 100 C.C. No. of Experinien t. 1. 2. 3. 4. 5 . 6. 7. 8. 9. 10. 11. A. Nil 0.01 0.05 0.50 1.00 5.00 10.00 15.00 20 00 25.00 0-10 B. Nil 3-8 16.6 28.3 66.6 80.0 95-2 97.5 98.3 98.7 99.0 C. 100.0 96.2 83.4 71.7 33.4 20.0 4.8 2-5 1.7 1.3 1.0 D. 20.20 20.25 20.20 20.10 19.85 19.70 19.30 18.70 18.10 17.45 17.10 E. - - - 72-10 34.00 20.50 5.02 2.70 1-92 1-53 1.20 F. - - -_ 0.40 0.60 0.50 0.22 0.20 0.22 0.23 0.20 We believe that for practical purposes it will be found best simply to add the values shown in the tables under column F (which represent the influence of definite percentages of sucrose, expressed on the total sugars present) to the percentages of invert sugar, determined by direct titration.For this purpose, how- ever, it is necessary to know the percentage of sucrose, not calculated on the sample, but on the sugars (invert sugar and sucrose) in the sample. This can be determined either by the Clerget method or by the method of double titration before and after hydrolysis with hydrochloric acid. Since only the approximate percentage of sucrose is required in order to deduce the correction, the value calculated by the usual formula when the titration method is used is sufficiently accurate. This formula is as follows : 95(I’ - I) 100 ’ s= ~ in which S is the approximate percentage of sucrose, I is the apparent percentage of invert sugar-i.e., the direct titration value-and I’ is the percentage of invert sugar obtained by titrating the sample after complete hydrolysis.The appropriate value given in column F is subtracted from the value of I and added to the value of S, the respective results giving the corrected per- centages of invert sugar and of sucrose. In order to express these on the sample, S + I each of the values is multiplied by ----. To be exact, the value added to the 100 approximate percentage of sucrose, S, should be diminished by 5 per cent. ; but the accuracy of the method does not warrant this refinement, seeing that the corrections to be applied are values of comparatively small magnitude. The titrations recorded in this paper were practically all carried out by Messrs.George McLaren and R. C. Denington, to whom our best thanks are due.THE ANALYST. 171 DISCUSSION. Mr. CHAPMAN said that a good many years ago he was in the habit of using the volumetric process entirely, as being the best then in existence. In regard to the end-point, the accuracy with which a trained normal eye could determine the disappearance of colour without the aid of a chemical indicator was extraordinary ; but that, of couree, only applied to solutions which were colourless, or nearly so. During the past two years he had had considerable experience with the use of the ferrous thiocyanate indicator, and he should like to say how admirably it had worked in his hands, and how much it had contributed to making the volumetric process a more accurate and a more generally useful one.He entirely agreed with the authors as to the necessity of adding the sugar solution to the Fehling’s solution in small quantities at a time. Under ordinary circumstances it was not well to run in quantities of more than 4 or 5 C.C. at the commencement, and, of course, much smaller quantities towards the end of the titration. That the authors had been able to so increase the accuracy of the volumetric process as to be able to get results correct to within 1 part in 250 was a notable achievement, and one on which they were to be congratulated. For a good many years he (Mr. Chapman) had used the gravimetric process, and he had found no dificulty in obtaining duplicate weighings of oxide of copper agreeing within 1 milligram, but, of course, the difference might occasionally be a little greater.Some years ago he had made very numerous experiments as to the copper-reducing powers of invert sugar, dextrose, and maltose, for the conditions under which he was in the habit of working in his own laboratory, and he was accustomed to use his own numbers in place of those of Brown, Morris, and Millar, although there could, of course, be not the slightest doubt as to the high degree of accuracy which those chemists had reached. I n regard to the effect of concentration, his results had shown that the copper-reducing power of maltose was very much less affected than that of dextrose or of invert sugar, and that, in consequence, a factor could be used for the conversion of oxide of copper into maltose without introducing any serious inaccuracy. He had also made some experiments with lmulose, but at that time it was exceedingly difficult to obtain that substance in a state of sufficient purity, and so the experiments were not carried very far.In the estimation of invert sugar, the error due to the influence of cane-sugar was not usually very serious, although in the case of molasses and similar products it might be quite appreciable. It would, however, be very useful to have the accurate figures which the authors had given, and to know what the error really amounted to. He should like to hear how the lamdose used by the authors had been prepared and purified. Mr. W. T. BURGESS thought that the necessity was not sufficiently realised of seeing that small pipettes were always quite accurate and used in a, definite manner.He thought it very unlikely that if the same 10 C.C. pipette were used by three observers the respective quantities delivered would agree within 1 part in 300, or .03 C.C. Care would therefore be necessary in measuring the small quantity of Fehling’s solu- tion (10 c.c.), otherwise an error of 1 in 300 might be expected, apart from other experimental errors in the authors’ process.172 THE ANALYST, Mr. JULIAN L. BAKER said he was convinced of the great accuracy of which the volumetric process was now capable, and he felt able to rely entirely upon it, and to eliminate the gravimetric process. Years ago, when potassium ferrocyanide was the only indicator used, the volumetric process was capable of a certain degree of accuracy, provided the amount of reducing sugar measured was small; but the use of this indicator had the great disadvantage that the solution containing an excess of cupric salt had to be filtered through paper, and it had been pointed out by Dr.Divers that the results were apt to be erroneous, owing to absorption of the copper salt by the paper. With pure sugars, etc., giving clear solutions, he thought that, as Mr. Chapman had mentioned, an indicator was hardly necessary, but it was essential in dealing with dark sugar products like caramel, malt extracts, black beers, etc. H e asked the authors if they had investigated the influence of the time of boiling upon the results, because different people’s ideas of bringing to a boil varied somewhat.He was looking forward to studying the authors’ maltose curve, because in some recent work he had obtained very pure maltose-pure to the extent of crystallising from water-and had found its reducing power to be from 101 to 1014. Mr. J. H. Millar had obtained the figure 101, working with pure maltose. Mr. H. HULTON agreed with what had been said as to the necessity for running in the sugar solution in small quantities. He did not think that an indicator was necessary when titrating starch degradation products, because a very good indication of the end of the reaction was given by the characteristic frothing which took place, He thought that the dieculty with Iwulose which had been mentioned was due to the great susceptibilitg of ltwulose to decomposition by strong alkali, and that while part of the lamdose was reducing the copper, some of it was being decomposed by the excess of soda, and he should therefore expect that no amount of repetition would give concordant results.Mr. E. H. JEFFERS asked if the authors had made any comparison between the limits of error when ferrocyanide and ferrous thiocyanate respectively were used. He thought that really the only objection to the use of ferrocyanide was that most operators filtered rather a large quantity in order to get the reaction on a plate, but by the method recently suggested by Messrs. Watts and Tempany, of simply with- drawing a drop on a rod and placing it on a couple of sheets of filter-paper, it was possible to obtain the reaction in a similar way to that adopted by the present authors.Mr. LING, in reply, said that the sugars they had used in their experiments had all been highly purified, recrystallisation being continued until the specific rotatory powers (based on readings in 8 400 mm. tube) were practically constant. The publication of the ltievulose values was justified, not only because they were able by averaging these values with those obtained with dextrose to check the values obtained from the product of the inversion ” of sucrose, as they carried it out, but also because it was now usual to return the percentages of dextrose and lz?vulose separately in the analysis of commercial invert sugars, these latter sugars being sometimes adulterated with starch sugar (dextrose).Seeing that the ltlevulose used was of such a high degree of purity, he did not think much advantage would accrue from a repetition of the values obtained with that sugar. He agreed with Mr.HultonTHE ANALYST. 173 that, in all probability, the lesser accuracy of the values they had obtained with levulose as compared with the other sugars was to be ascribed to its ready de- composition under the influence of alkalies. With regard to the question of time, this was quite as constant with each individual worker in the volumetric process as it was in the gravimetric. The limit of accuracy of the two processes was discussed at length in the paper by Mr. Jones and himself. He had not made any comparison between the ferrocyanide and ferrous thiocyanate indicators recently, but whereas with the latter duplicate burette readings agreed within 0-1 c.c., when ferrocyanide was used it was difficult to get them closer than 0.5 C.C. Over twenty years ago he had used the method of applying the ferrocyanide indicator on a pad of filter-paper, but he did not think that it had any advantages over filtration. Mr. JONES agreed with Mr. Burgess that a pipette which delivered 10 C.C. of water would not deliver that quantity of Fehling’s solution, but a pipette could be chosen which did deliver 10 C.C. of Fehling’s solution ; or, more simply, any ordinary pipette might be marked and set aside for use in this work. Their own practice was to use a graduated pipette, sold to deliver 10 C.C. of water between the marks 0 and 10, and it was a matter of indifference to them whether it delivered 9.9 C.C. or any less quantity, as the difference between any two deliveries was exceedingly small, and the pipette used in standardising the solution was always used in analytical work with that solution.