Determination of Dry Substance in Beet Sugar Juices. - Industrial

Determination of Dry Substance in Beet Sugar Juices. Robert J. Brown. Ind. Eng. Chem. , 1924, 16 (7), pp 746–748. DOI: 10.1021/ie50175a034. Publicat...
0 downloads 0 Views 458KB Size
746

INDUSTRIAL AND ENGINEERING CHEMISTRY

Vol. 16, No. 7

Determination of Dry Substance in Beet Sugar Juices' A Precision Method By Robert J. Brown GREATWESTERN SUQARCo., DENVBR,COLO.

APPARATUSAND PROHIS paper is a preResults checking within 0.01 per cent can be obtained on pressed cmum liminary report on juice, diffusion juice, and thin juice when using suficient care work done on the Two types of ovens were with the sand dry method. determination of a method available, the Spencer and For ordinary work the Spencer oven gives good results, and it is of analysis capable of givthe Freas vacuum oven. I n especially valuable when results are required in a short time. ing dry substance in thin the operation of the Spencer The largest single volatile impurity driven off from fresh pressed sugar juices a c c u r a t e l y o v e n , large volumes of juice during drying is carbon dioxide. I n no case did all other within 0.01 per cent on h e a t e d a i r a r e drawn volatile impurities amount to 0.007 per cent. original juice. through capsules containing If carrying completely to dryness would cause no more volatile This work is being done some material, such as asto be given off than is given off by the distillation method here dein connection with the debestos, in which the juice is scribed, the dry substance defermination can be made accurately termination of sugar in beet absorbed. The temperato 0.01 per cent. juices and sirups with the ture, volume of the air, and There is reason to believe that the dry substance determination object of ascertaining the the absorbing material can on ail juices as far as thick juice can be made accurately to 0.OI per purity of the solution accube varied. The writer cent. rately within 0.01 per cent. worked a t 105" C. and used The method of analysis will the largest volume of air he be strictly for research work, primarily in the calculation of could get, which with four capsules in the &en was about 212.4 elimination of impurities, since many of the refinements cu. dm. (7.5 cubic feet) per minute. Various absorbing matenecessary are entirely out of the question for control work. rials were used, and the best was unwashed, freshly ignited, The high degree of accuracy is made necessary by the fact extra long fibered asbestos. This gave the greatest absorbing that small errors in purity cause large variations in the capacity while still allowing a good flow of air. The writer was never able to get a constant weight on the calculation of elimination. An error of 0.1 per cent in purity, when calculating first carbonation elimination, will make an capsule, before or after addition of the sugar juice. While error of about 0.7 per cent in the percentage of impurities the variations were not great, they were greater than was removed, or about 2 per cent of the total. When calculating allowed, since an error of 0.0005 gram would cause an error battery elimination, a small error in purity may often cause of 0.01 per cent in dry substance, inasmuch as only 5 grams a negative elimination to be shown, when a positive elim- of a 12 per cent solution could be used. It is doubtful if the use of larger capsules would eliminate this difficulty, since ination may have been obtained. I n purity determinations there are three common methods the variation in weight using a larger quantity of absorbing of obtaining dry substance: (1) Brix, (2) refractometer, (3) material would probably be proportionally just as great. For ordinary work the results obtained would be considered actual drying. On account of the accuracy required it was impossible to use the first two methods. When working with practically perfect, since the true dry substance could be thin juices, the Brix dry substance is generally found to be checked, using pure sugar solutions, within 0.01 to 0.04 per sufficiently greater than the dry substance by actual drying cent. While the oven is not satisfactory for work demanding the to lower the purity 1 to 2 per cent, and the accuracy in reading the hydrometer is not sufficient to insure an accuracy highest degree of accuracy, it will give very good results in in purity greater than 0.35 per cent. The same objections control work when speed is more important than accuracy. hold true with the refractometer dry substance. The ac- If a large number of determinations art! to be run, the curacy in reading is not sufficiently great, and true dry sub- Spencer oven requires more man-hours than the Freas oven. Using the Freas vacuum oven it was possible to develop a stance is not shown, although it is generally closer than Brix. It was therefore decided to determine dry substance by method which gave satisfactory results. Results can now be actual drying. Solutions of about 12 per cent dry substance obtained, on 12 per cent pure sugar solutions, which check are used, and in order that the purity may be determined accu- within 0.01 per cent of the true value; and using pressed, rately within 0.1 per cent, the maximum error in the dry sub- diffusion, and thin juice, duplicate results check within this. stance determination must not be greater than 0.01 per cent. limit. The first decision was to run under vacuum at 70" C. The work to date has been limited almost entirely to pure sugar solutions and pressed juice from stored beets. A because of some results on decomposition of sugar given in small amount of work has been done on diffusion juice, thin Circular 44, Bureau of Standards. Upon heating sugar it juice, and thick juice. Eventually, all beet sugar juices and is decomposed with some change in weight, and caramel is formed. The Government figures show the time required sirups will be included. The first problem was to find the apparatus and procedure to form caramel to be equivalent to 0.01 per cent invert that would give results on dry substance checking within sugar when heating a t different temperatures. Their table 0.01 per cent. The second problem was to prove that, when follows : 50.0' C. 39.0' C. 79.6' C. 66.6' C. Temperature working on unknown sugar solutions, these results were 107.0 476.0 Time in hours 0.57 10.9 correct. If we extrapolate to determine the effect at 105" C. we see 1 Presented by R. J. Brown and H. W. Dahlberg before the Division that it is appreciable, while drying at 70" C. for 10 or 11 hours of Sugar Chemistry at the 66th Meeting of the American Chemical society, does not hurt the determination within the limits of error. Milwaukee, Wis., September 10 to 14, 1923.

T

July, 1924

INDUSTRIAL A N D ENGINEERING CHEMISTRY

The tests were made using sea sand and pumice, and glass and aluminium drying dishes were used. In the original tests check results were not obtained, nor even check weights on the empty dishes. I n these tests aluminium dishes filled with 25 grams of acidwashed and screened sea sand were used. The dishes were weighed several times after drying in the oven. Variations in weight as great as 1.8 mg. were noted. Tests on pure sugar solutions showed variations as great as 0.04 per cent, with the tendency to be high rather than low. I n the hope of obtaining greater constancy in results, glass dishes were used in place of the aluminium ones, but the results did not improve. Since the kind of dish used had no effect, the absorbing material was changed, and pumice was used as prescribed by the A. 0.A. C.2 The results were poor, the dry substance being high in all cases, and prolonged heating increased the weight. Sand was then tried again, and sand that had been previously acid-washed, ignited, and screened was treated again. It was given a thorough acid washing, dried and screened, that greater than 30 mesh and less than 60 mesh being discarded. A portion of this sand was ignited; the rest was not. The following method gave perfect results: Place 25 grams of sand in the aluminium dish containing a glass stirring rod, and heat for 15 hours a t 70" to 75" C. with 120 mm. absolute pressure. Cool in desiccator and weigh. Add about 5 cc. of the sugar solution from a weighing bottle, mix, and place in oven. Heat as described for sand, cool in desiccator for not more than 1 hour, and weigh rapidly. Keep fresh acid in the desiccator. In one test using this method, after the solution was dried, the dishes were allowed to stand in the desiccator for 48 hours before weighing, and the dry substance was 0.01 to 0.04 per cent high. They were then replaced in the oven for 15 hours, cooled in the desiccator for not over 1 hour, and weighed. All checked the actual within 0.01 per cent. Afterwards, dishes were never allowed to remain in the desiccator over 1 hour. These tests were repeated using sand ignited after acid washing, and excellent results were obtained. The work showed that correct results could be obtained on dry substance of pure sugar solutions (12" Brix) when proper care was taken, the following points being essential: (1) Absolutely clean sand (2) Properly screened sand 23) IXshes and sand dried for I 5 hours in oven a t .70" to 75 C., with 120 mm. absolute pressure (4)Sample of juice weighed into dish using weighing bottle i5) Fve-gram sample of juice dried for 15 hours in oven a t 70 to 75' and 120 mm. absolute pressure (6) Ilesiccators containing fresh acid (7) Samples weighed out of desiccator within 1 hour after removing from oven

It is not maintained that these conditions and none other should be used, but their use will give excellent results. This method not only gave correct results on known sugar solutions, hut also gave check results on pressed juice, diffusion juice, and thin juice. VERIFICATION OF RESULTS The work on the second problem, that of proving these results to be correct, has been limited almost entirely to pressed juice from stored beets. Before being able to determine the correct dry substance, we must define it. Ordinarily, dry substance is an arbitrary figure obtained by a certain procedure, and the results obtained on a given juice will vary with the procedure used. The primary consideration here is elimination of impurities,

* A,rsoc. Official Agr. Chem., Methods, 1920, p. 101.

747

and an impurity that volatilizes during evaporation is just as important as a nonvolatile impurity. This is especially true in the battery where viscosity plays an important part in diffusion. Therefore, it was decided to consider everything in the solution other than water as dry substance. It is probable that 0.01 per cent of free COa would not have a measurable effect on the viscosity, but until that has been proved, it must be considered as an impurity. Having decided that everything in the solution other than sugar and water is an impurity, the next step was to find a method for determining the volatile. Two methods have been used, both of which have their limitations. The first method was fixation of the acids present. It was believed that most of the volatile substance given off was the free acid present or decomposition products formed due to the presence of this free acid. In order to determine the true dry substance, the determinations were made by drying straight pressed juice, and pressed juice in the presence of an excess of ammonia and calcium carbonate, and calculating the result. It was found that when old pressed juice from poor beets was used, an appreciable amount of volatile acids was given off, calculating 0.015 per cent. With this same pressed juice the addition of a large excess of ammonia caused decomposition of some substances present in the juice and lowered the dry substance as much as 0.2 per cent. When working with fresh pressed juice no volatile acids were found to be present, and large excesses of ammonia did not cause lowering of dry substance. The limitations of this method are that if there is any decomposition of substances, such as proteins, .giving free ammonia, it would not be retained (in the drying dish), neither would any of the volatile neutral substances, such as alcohols, be retained. Our second method of determination of volatile substances was by distillation under vacuum a t approximately the temperature used in the dry substance determination. The vapors were passed successively through boiling sulfuric acid solution, boiling barium hydroxide solution, and the remainder was condensed. Vacuum was maintained by means of a filter pump, which was kept in operation less than 5 minutes per hour. Therefore, there should have been very little loss of volatile compounds. Distillation of a sample of old pressed juice showed 0.011 per cent volatile acids. By fixation of this same sample, 0.015 per cent had been found. All other tests showed less than 0.007 per cent total volatile compounds, including acids, bases, and neutral substances. Neutral substances were determined by oxidation to their respective acids by potassium dichromate and sulfuric acid. This method was not entirely satisfactory, since it was possible to obtain only 30 to 35 per cent oxidation of a known amount of ethyl alcohol by the same method. It has since been found that 90 to 95 per cent recovery of alcohol as carbon dioxide is possible by complete oxidation, and this method will be used in the future. However, the amount of volatile neutral substances is so small that it is not believed that the error in their determination seriously affects the results. The most serious fault with this method is that distillation is not carried completely to dryness, but since the same temperature is carried for a longer time than in the dry substance determination, no appreciable amount of volatile should remain undistilled. Tests to determine volatile have not yet been made on diffusion juice or thin juice, neither have they been made on pressed juice from normal beets. However, the amount of volatile should not be greater in any case than that obtained from the pressed juice used in these tests.

INDUSTRIAL A N D ENGINEERING CHEMISTRY

748

In no case has the carbon dioxide given off been considered, principally because no method of handling it in the dry substance determination is known. Only a trace is given off when boiling the juice a t atmospheric pressure, but when boiling under vacuum as much as 0.02 per cent carbon dioxide has been found, which was probably due to air leaks. At present it is believed that there is not a sufficient amount of volatile impurities given off from pressed juice, from normal beets, during the dry substance determination to affect the result to the extent of 0.01 per cent, provided carbon dioxide is not considered. Results on other juices must be determined later.

vol. 16, No._7

In view of some results obtained since this paper was written, the author wishes to state that perfect results could not be obtained during periods of high relative humidity. The original work was done under conditions which are normal for Denver, but during the past few months, a period of unusually high relative humidity, a decided relation was found between the results and the weather. Since an amount of moisture in the air that is considered high here would be normal in many localities, it is probable that the method herein described would only give perfect results when the work was done in a room supplied with "manufactured" air containing very little moisture.

The Gravimetric Determination of Lead in the Presence of Tin and Antimony' By Ernesto Stelling YALEUNIVERSITY, Nsw HAVBN, CONN.

I"ofAtinprevious article it has been shown that the oxides and antimony prepared in the wet way can be dissolved by means of concentrated sulfurous acid and hydrochloric acid, and that this method can be applied to the qualitative analysis of alloys containing these metals. In this article it will be shown that this fact can be applied to the quantitative separation of lead in alloys containing tin and antimony without preliminary removal of the latter metals. METHOD Treat 1 gram of the borings with 20 cc. of concentrated nitric acid and 10 cc. of water. Evaporate to dryness on the steam bath, add 50 cc. of water saturated at room temperature with sulfur dioxide and digest for 5 minutes a t approximately 60" C. Add 20 cc. of concentrated hydrochloric acid and boil. The oxides of tin and antimony dissolve completely. Add 10 cc. of concentrated sulfuric acid and evaporate until the sulfuric acid. fumes. Cool, add 20 cc. of an aqueous solution containing 10 per cent alcohol and 10 per cent sulfuric acid. Cool, filter on a Gooch crucible, wash with same solution, and ignite as usual. SAMPLES USED This method was tested by the analysis of two samples of lead alloys. One was a sample of lead base bearing metal (No. 53 of the Bureau of Standards) which had been analyzed by ten different chemists and whose analyses varied from 78.78 to 79.04 per cent of lead, averaging 78.89 per cent. It contained 10.91 per cent tin and 10.1 per cent antimony. The results obtained by the new method were: 78.80, 78.87, 78.87, and 78.90 per cent; averaging 78.86 per cent of lead. The second sample v a s a solder in which lead was first determined as sulfate by the standard method after removing tin and antimony. The results obtained by the standard method and by the new method were as follows: Standard Method Per cent 50.22 50.27 50.32 50.35 Average 1

2

50.29

New Method Per cent 50.26 50.29 50.30 50.36 .50.37 .. .

Average

Received May 12, 1924. THISJOURNAL, 16, 346 (1924).

50.32

CONDITIONS NECESSARY The accuracy and rapidity of the new method depend upon the conditions. The sulfur dioxide solution should be nearly saturated at room temperature, as the success of the operation depends primarily upon its concentration. If the solution is only half saturated, the oxides do not go completely into solution when hydrochloric acid is added. If a much larger volume of the solution is used, the hydrochloric acid is correspondingly diluted and the oxides will not readily dissolve unless a larger volume of hydrochloric acid is added, which will increase the time required in the subsequent evaporation. The digestion temperature may be varied somewhat, but experiments showed that if the temperature was below 40" or above 70" C., the oxides did not dissolve completely on subsequent treatment with hydrochloric acid. At lower temperatures the rate of adsorption of the sulfur dioxide seems to be too slow, and if the temperature is too high the sulfur dioxide is removed too rapidly by evaporation. The amount of hydrochloric acid should be more than 10 cc., 20 cc. being sufficient. Any larger amdunt will dissolve the oxide but is unnecessary. The amount of alcohol used for diluting and washing should be about 10 per cent of the total volume. If more is used the results come high in lead. With 50 pei cent alcohol the results were found to be from 0.1 to 0.3 per cent high. If pure water is used the results are low to the extent of 0.2 to 0.3 per cent. The writer wishes to emphasize the workability of the new method and also the accuracy of its results. ACKNOWLEDGMEBT The writer desires to express thanks to Professor H. W. Foote for his valuable suggestions and for his criticism of this paper. Glue in Coating Paper-The Bureau of Standards has conducted an investigation of the amount of glue required in coatings and the effect of glue of different grades. The results indicate that by properly selecting the grades of glue the amounts required may be materially decreased. Preliminary heating of the coating mixture was found to improve the quality of the coating by eliminating air bubbles. Some experiments were also made on the settling of clay in the coating mixture. I t is important that the clay be held in suspension in order t o obtain a uniform coating. I t ' w a s found that different glues vary considerably in clay-carrying power.

.