T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y ,
38
sidered negligible. T h e final volume was arbitrarily selected, as it was t h e greatest volume a t which 0.1 cc. of 0.2 per cent dextrose solution would make a distinct change in color. A test t u b e of t h e size given prevents much exposure t o t h e air, and its liability t o boil over when heated too rapidly, prevents excessive evaporation. Numerous trials were at first made with ferrous sulfocyanidel a n d iodide starch mixture,2 b u t it was found t h a t both have a common defect; namely, t h e y become colored on exposure t o air without addition of Fehling solution. Attempts t o obviate this destroy their sensibility t o a large extent. The writer had used sodium monosulfide for some time for colorimetric copper estimations and was impressed b y its sensitiveness. Furthermore, there ought t o be no tendency for t h e reoxidation of t h e copper on mixing with t h e indicator, as t h e medium is a reducing one. After a few minutes a yellow color will appear, owing probably t o solution of some of t h e copper sulfide. The absence of color, however, is permanent for sufficient time for a n observation t o be made, while when copper is unreduced, t h e appearance of t h e color is immediate. After t h e operator has become familiar with t h e end point, it can be judged very closely by t h e appearance immediately on addition of t h e drop. As long as unreduced copper is present, a yellowish color is perceptible without agitation, while when all t h e copper is reduced, t h e only shade is t h e reddish brown of t h e cuprous oxide. Proteins, a n d metals which give colored sulfides interfere with t h e end point. For t h e former, t h e writer boils a measured volume a n d adds alumina cream while still hot, then cools and makes up t o a convenient volume, adding sufficient water t o correct for t h e volume occupied b y t h e alumina. The solution can t h e n be allowed t o settle and t h e clear supernatant liquid used, or if time is a n object, i t may be filtered. This method works very well on most urines a n d colored solutions. For removing t h e metals, methods will doubtless suggest themselves. RESULTS OBTAINED
Various dilutions of a sugar solution were made u p and given t o a member of t h e staff (without information as t o their sugar content), who obtained t h e following figures. No. of cc. used in titration 23.30 15.65 11.70 9.35
M g . total dextrose present 47.60 47.75 47.60 47.75
It must be understood t h a t this was not pure dextrose, as t h e purpose was only t o see whether results were constant and independent of t h e concentration of t h e titrating solution over comparatively wide limits. The above results were each obtained with one titration besides t h e preliminary one. Impure maltose solutions were titrated in a similar manner. Ling Rendle and Jones, Allen’s “Commercial Organic Analysis,” 4th edition. 2 F. F. Harrison, Sutton’s “Volumetric Analysis.” 10th edition. 1
No. of cc. used in titration 23.2 15.5 11.6 9.25
Vol. 7 , NO.
I
Mg. total maltose present 92.8 93.0 92.4 92.5
The following results show t h e remarkable constancy of results over long periods. The titrations are on j cc. of mixed Fehling solution, which was the original manner of using t h e method, b u t i t was changed t o increase t h e accuracy of t h e end point. 6/11/1914 5 cc. Fehling reduced by 44.43 mg. crude maltose 8/26/1914 5 cc. Fehling reduced by 44.47 mg. crude maltose
For each value four concentrations were titrated. This constancy is not confined t o maltose, b u t is also t r u e of dextrose. The dextrose factor for j cc. in M a y was 23.6 mg. and in August 23.8 mg. The writer advises t h e standardization of t h e Fehling solution b y each user, a n d hence his factor should not be accepted as final. A mean of four determinations on pure glucose from two different sources in which t h e maximum variation from t h e average was * O . I S mg., gave 4 7 . j mg. dextrose as t h e value of I O cc. of a Fehling solution whose copper content had been accurately adjusted b y a sodium thiosulfate solution standardized with pure copper. As can be seen, t h e factors, once obtained, need be determined only at rare intervals. Pure dextrose, prepared by t h e Bureau of Standards, is now available for this purpose. SUMMARY
The writer describes a modification of Fehling’s volumetric method designed t o do away with some of its sources of error. The use of a new indicator is described which, so f a r as is known, is here first used for t h a t purpose. DISTILLERS LABORATORY 503 KENTUCKY TITLEBUILDING LOUISVILLE, KY.
A COMPARISON OF THE GUNNING-COPPER METHOD WITH THE KJELDAHL-GUNNING-ARNOLD METHOD FOR THE DETERMINATION OF NITROGEN By OVP F. JBNSEN Received June 3, 1914
I n 1908 t h e Association of Official Agricultural Chemists officially adopted’ a modification of t h e Gunning method for t h e determination of nitrogen which prescribes t h e use of 0.1 t o 0.3 g. of crystallized copper sulfate in addition t o t h e potassium sulfate. Penny,* in his report on the determination of nitrogen in 1907, saps of this modification: “ T h e claims made for t h e catalytic action of copper sulfate seem t o be justified b y experience, a n d there is reason t o suppose t h a t this reagent will do much t o shorten t h e time of digestion.” Although possessing many advantages over t h e other methods in use a t t h a t time, i t was not generally adopted. More recently another modification known as t h e Kjeldahl-Gunning-Arnold3 method has been proposed. This method differs from t h e GunningCopper method, only in the use of metallic mercury instead of copper sulfate. I n a recent number of THIS JOURNAL,^ Trescot shows, very conclusively, 1 2 8 4
U. S. Dept. Agr., Bureau of Chemistry, Bull. 122, p. 183. I b i d . . 116, p. 43. I b i d . . 108, p. 15. Vol. 6 (L913). 914.
T H E J O U R N A L OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
J a n . , 1915
t h a t the Kjeldahl-Gunning-Arnold method reduces t h e time of digestion over t h e original Kjeldahl or Gunning methods from four hours t o one and one-half hours. This is a very important item in laboratories where a large number of nitrogen determinations are being made. The Gunning-Copper method has been in use in this laboratory for t h e past nine years, and t h e general practice has been t o digest from two t o three hours after the solution becomes clear. Since this method possesses several advantages in manipulation over t h e Kjeldahl-Gunning-Arnold method, a comparison of t h e two methods with differe n t periods of digestion was undertaken, using bone meal, dried blood, cyanamide, and linseed meal as representative substances. I n Table I , t h e period of
39
this point, determinations were made on a number of substances such as are encountered in a fertilizer or food laboratory. Three determinations were made on each substance by each method. The data are given in Table 11. The results obtained for these substances by the two methods show a very close agreement, and in point of, accuracy there is no choice between them. The Gunning-Copper method possesses several distinct advantages over t h e other which may be enumerated as follows: I-The use of potassium sulfide is unnecessary, and t h u s one detail of manipulation (addition of K2S) is eliminated, slightly shortening t h e time required for t h e determination. Also t h e presence of hydro-
TABLE I-EFFECT OF DIFFERENTPERIODS OF DIGESTION Period of digestion after solutions became clear
ON PERCENTAGES OF NITROGENFOUND 11/2 hours 2 hours '
1 hour No. of analyses Max. Min.
METHOD BONE M E A L : Gunning-Copper ............................ 6 Kieldahl-Gunninn-Arnold.. . . . . . . . . . . . . . . . . . . 6
.......... ,. 6 . . . . . . . . .... . . 6 . . . . . . . . . .. . . 6 ... 6 LINSEED MEAL: Gunning-Copper ............................ 6 Kjeldahl-Gunning-Arnold., . . . . . . . . . . . . . . . . . . 6 ,
2.18 2.25
2.16 2.19
Av. 2.16 2.20
No. of analyses Max. Min. 4 4
2.20 2.20
2.18 2.18
Av. 2.19 2.19
No. of analyses Max. Min. 6
6
2.20 2.23
2.18 2.19
Av. 2.19 2.20
3 hours
No. of analyses Max. Min.
6 6
2.23 2.22
2.20 2.20
Av. 2.20 2.20
14.11 14.03 14.07 14.32 14.11 14.24
6 14.24 14.07 14.13 6 14.32 14.19 14.29
6 14.32 14.15 14.22 6 1 4 . 3 0 14.19 14.25
6 14.40 14.23 14.29 6 14.34 14.15 14.26
15.75 15.37 15.52 15.62 15.46 15.55
6 15.62 15.50 15.53 6 15.54 15.41 15.49
6 15.67 15.46 15.55 5 15.62 15.50 15.55
6 15.58 15.50 15.53 6 15.62 15.50 15.55
5 4
6
~
5.53 5.62
5.45 5.53
5.49 5.56
digestion in all cases refers t o t h e length of time after the solutions became clear. I n t h e length of time required t o become clear, no difference was noted in t h e two methods. Using a moderate flame, t h e time required for t h e sulfuric acid solution t o become clear was as follows: Cyanamide.. . . . . . . . . . 20 min. Bone meal.. . . . . . . . . . . 25 min. Dried blood.. . . . . . . . . . 25 min. Linseed m e a l . . . . . . . . . 35 min. The results seem t o indicate t h a t a digestion of one and one-half hours is sufficient t o produce a quantitative yield of ammonia in either method, except in t h e dried blood, where from two t o three hours are necessary in t h e Gunning-Copper method. Hibbardl found three hours t o be necessary, using this method. On t h e whole, t h e results are slightly in favor of t h e Kjeldahl-Gunning-Arnold method, as a quantitative yield of ammonia is produced after a digestion of from one t o one and one-half hours. However, i t is doubtful if there are many substances for which t h e Gunning-
5.62 5.62
5.50 5.59
5.57 5.61
6
5.62 5.59
5.50 5.53
5.55 5.58
6 6
5.64 5.62
5.56 5.56
5.62 5.58
gen sulfide in t h e laboratory from this source is avoided. 11-In adding the sodium hydroxide before distilling, t h e copper sulfate acts as a n indicator, so t h a t a large excess of alkali is easily avoided, and the bumping during distillation is much lessened. 111-The difference between t h e price of copper sulfate and metallic mercury, and the elimination of potassium sulfide, makes t h e Gunning-copper method somewhat cheaper. SUMMARY
A quantitative yield of ammonia in dried blood is produced sooner in t h e Kjeldahl-Gunning-Arnold method t h a n in t h e Gunning-Copper method. I n the case of all t h e other substances studied, a digestion of one and one-half hours proved equally efficacious for either method. The Gunning-Copper method possesses advantages in manipulation which make i t preferable t o the Kjeldahl-Gunning-Arnold method, especially where a TABLE11-DETERMINATIONS OF NITROGEN ON VARIOUS SUBSTANCES large number of determinations are t o be made. METHOD Gunning-Copper Kjeldahl-Gunning-Arnold Period of digestion
SUBSTANCE
Max. Gelatin. . . . . . . . . . . . . . . 15.14 E g g albumin (dried). . . . 12.86 Peptone. . . . . . . . . . . . . . . 14.49 Casein. . . . . . . . . . . . . . . . 13.87 Feathers. . . . . . . . . . . . . . 14.49 Leather. . . . . . . . . . . . . . . 4 . 8 3 Fish s c r a p . . . . . . . . . . . . . 6 . 4 8 Animal t a n k a g e . , . . . . . . 6.32 Garbage tankage.. . . . . . 2.99 Beef scraps.. . . . . . . . . . . 9 . 1 4 Castor bean pomace. ... 4.66 Cocoa. . . . . . . . . . . . . . . . . 4.01 Cottonseed m e a l . . . . . . . 7.33 Cottonseed meal.. . . . . . 6 . 4 3 Silage (dried). . . . . . . . . . 1 . 3 0 Flour. . . . . . . . . . . . . . . . . 2 . 4 3 Bran, . . . . . . . . . . . . . . . . . 2.60 Bone meal.. ........... 3 . 0 6 Peat. . . . . . . . . . . . . . . . . . 2.41
11/2 hours Min. Av. 15.12 15.13 12.75 12.80 14.38 14.45 13.81 1 3 . 8 3 14.43 14.47 4.72 4.76 6.35 6.43 6.31 6.26 2.96 2.98 8.97 9.04 4.55 4.59 3.94 3.97 7.30 7.31 6.40 6.42 1.26 1.28 2.41 2.42 2.55 2.58 3.02 3.04 2.37 2.38
Max.
11/z hours Min. Av.
Copper method will not give a n equally high yield with t h e same length of digestion. I n order t o test 1
THISJOURNAL,a , 464.
CHEMICALLABORATORY STATION MICHIGANEXPERIMENT EASTLANSING
COMPARISON OF A FEW METHODS FOR TOTAL PHOSPHORIC ACID IN SUPERPHOSPHATE' B y C. A. PETERS Received October 5, 1914
A 'comparison of several methods for the determination of total phosphoric acid in superphosphate has been made, emphasizing t h e interference of silica in t h e gravimetric method, and showing t h a t shorter methods give as accurate results as t h e customary double precipitation as phosphomolybdate and ammonium 1 T a k e n from a thesis b y Arthur G. Weigel, optionally presented for t h e degree of B.S. at the Massachusetts Agricultural College a t Amherst.
