Magnesium Determination in Bovine Blood Serum MITSUO OKAMOTO and J. W. THOMAS Bureau of Dairy Industry,
U. S. Department o f Agriculture,
A
STUDY was made t o compare the three most widely used methods for the determination of magnesium in serum. Magnesium values in calf serum were determined following the Simonsen et a2. ( l a ) ,and procedures outlined by Kunkel et al. (8), Hoffman (7). The first method (8)employs a dye, Titan yellow, which forms a colored lake with magnesium hydroxide. The intensity of this lake can be measured by a photoelectric colorimeter. The second method (12)employs precipitation of magnesium as magnesium ammonium phosphate from the calcium-free filtrate, and the colorimetric measurement of the phosphorus in the precipitate. The third method ( 7 ) employs the precipitation of magnesium with 8-quinolinol and the quinoline concentration is estimated colorimetrically as ferric quinoline. Because some modifications have been made and also because it is the principal method employed in the authors’ laboratory, the first method is briefly discussed here. However, for more deand Orange tails the publications of Heagy ( 5 ) ,Kunkel et al. (8), and Rhein (11)should be consulted. To 4 ml. of a tungstic acid filtrate, add 1 drop of alcoholic methyl red indicator, sufficient 0.1N sodium hydroxide to change the indicator color from red to yellow, 1 ml. of 0.1% polyvinyl alcohol, 1 ml. of 0.05% Titan yellow solution, 2 ml. of 1.5h’sodium hydroxide, and sufficient deionized water to make the total volume to 10 ml. The resulting colored solution is read against a blank at 540 mp, To test the accuracy of the methods, recovery experiments were performed for the Titan yellow and 8-quinolinol methods. The Titan yellow method gave an average recovery value of 100.3y0 with a standard deviation of h 4 . 2 (see Table I), and the 8-quinolinol method gave an average recovery value of 92.2% (three values, range 82.3 t o 98.8%).
Table 1.
Calf S o . x 569 X 569 h-2780 X 569
Recovery from Known Amounts of Magnesium Added to 1 M1. of Serum M g in 11g RfS Mg
Seruni,
Added,
7
Y
20 2 12.1 12 9 12.5 28.9 11.5 18.3 15.4 19.4 16 4
10 10 10
Calcd.. Y 30 2 22 1 22.9 22.5 38.9 24.5 28.3 25.4 29.4 20.4
Found,
10 10
21.2 21.7 20.6
29 7 19 n 23 5 22.7 38.5 24.4 29.3 25.7 29 6 21 1 21.5 21 6 22.7 21.5
Hoffman ( 7 ) 10 X 589 14 8 10 X 582 2 9 10 580 15 0 0 Standard deviation, 4.2
24 8 12 9 25 0
20 4 12 7 24 7
S614 S 572 S 173
N 572 N 173 N 3188
Mixture Mixture Mixture Mixture
x
11.6 11 2
11.7 10.6
10
10 10 10 10 10
4 10
in
21.6
Recovery, To
Y
98 3 86.0 102.6
100 0 99.0 99.6 103.5 101 2 100 7 102.9 99.5 101 9 101.6 104.4 A,..
Washington,
In comparing the Titan yellow and the ammonium phosphate methods, 94 samples were analyzed. In 70 samples (Group 1) the magnesium ammonium phosphates were left for precipitation for 2 hours as suggested by Simonsen et al. (fd); in 24 samples (Group 2 ) the solutions were left overnight. Samples in Group 1 gave an average value of 90.1 & 14.8%, and the samples in Group 2 gave an average value of 101.6 i 6.0% as compared to the Titan yellow values based at 100%. Thc average values for these comparisons are shown in Table 11. The Titan yellow method was chosen for routine estimation of serum magnesium for its rase, rapidity, and reproducibility.
Table 11. Average hlagnesium Serum Values from Three IIethods, Titan yellow,
A further test of the reproducibility of the two methods was performed by determining the serum magnesium values in 23 samples by both methods. Basing the Titan yellow values as loo%, the 8-quinolinol values, in comparison, were 84.9 & 12.33%, the range being 24.8 to 140.8% (see Table 11). The 15% discrepancy in some part can be accounted for by the loss of magnesium hydroxyquinoline in the washing process. The 8-quinolinol method appeared somewhat nonreproducible and unreliable and the breakage of the special tubes used in the method was EO excessive that this method was discontinued.
Method .\lgNHaP01,
Hydroxyquinolate, No. of Detns. mg. % mg. Yo mg. R 22 2.02 1.61 70 1.82 1’630 24 1.82 1,855 Magnesium ammonium phosphate preciiiitated for 2 hours. b Magnesium ammonium phosphate preciiiitated overnight.
Q
~ _ ~ ~ _ _ ~_ _ ~ ~ There has been some controversy as t o whether calcium interfered in the determination of magnesium by the Titan yellow method (d, 3, 6, 9). Therefore, magnesium content in serum was determined in two ways: directly in serum and in a calcium-free filtrate. In 15 samples the average values were 1.54 mg. % for ’ for the direct serum. I n the calcium-free filtrate and 1.47 mg. % the calculations of these values two standard curves were employed. The calcium-free filtrate values were read against a standard curve containing no calcium, while the direct serum values were read against a standard curve containing 1007 of calcium per 10 ml. The 0.07 mg. % difference in the two different methods of estimating the magnesium concentration is negligible for routine analysis; however, for more accurate results in the direct serum determination, an appropriate amount of calcium must be added to the blank and standard solutions. When the transmittance readings, for the direct serum, were read against the standard curve containing no calcium, thr values averaged 1.31 mg. % (see Table 111).
Table 111. Serum Magnesium Concentrations Using Calcium-Free and Calcium-Containing Deproteinized Filtrate
1no.v 82 3 98 4 98 8
D. C.
Calciiim-Free Filtrate,
Calf No. N 574 N 575 N 576 N 580 N 825 N 2783 N 3158 N 574 N 578 X 580 N 582 S 750 h- 3159 K 3165
Ms. 53
Direct Serums, M g . 7c 2 11
Direct Seriimb,
Ms. 70
2.30 2.28 1.28 1 47 1.43 2 06 2.22 2.08 1.20 1.37 1.36 1.38 1.53 1.53 0.48 0.37 0.39 0.96 1.20 1.55 1.37 1.51 1.75 1.27 1.40 1.32 0 94 1.20 1.41 1.52 l,65 1.68 0.85 1.00 1,08 1.46 1.62 1.76 1.28 1.42 1.70 AV. 1.54 1.31 1 47 a Value determined rising blank and standard containing no calcium which was also used for determining values in column 1. b Value determined using blank and standard solutions containing added calcium similar in amount to that found in serum.
1072
1073
V O L U M E 2 6 , NO. 6, J U N E 1 9 5 4 T o prove that the excess ammonium and oxalate ions, introduced in precipitating calcium, had no effect, 1 ml. of 1% ammonium oxalate [(?JHl)zC204HzO] was added t o the 10 ml. of blank and all known solutions. S o significant difference in the ammonium oxalate added etandards and the other standards were found. In working with a standard curve containing a definite amount of calcium, it is likely that an error would be introduced, as all samples will not contain the same amount of calcium. Probably this accounts for the 0.07 mg. % difference in the aforementioned values. This finding, that the amount of calcium found in bovine serum docs interfere with the estimation of magnesium by the Titan yellow method, is in contrast t o that found by Gilliam ( 3 ) )Garner (%), and Ludwig and Johnson (9). m7hole blood values of magnesium in cattle were reported by two investigators to be approximately the same as that of serum ( I I 7 ) , while Green and Macaskill (4)reported that two thirds of the magnesium was contained in the corpuscles while one third was located in the plasma. In this experiment a measured amount of whole blood from young calves was laked by the addition of water and magnesium u-as estimated in the manner described by Kunkel et al. (8). Orange and Rhein (11) stated that iron in the solution of ashed whole blood cells interfered in the estimation of magnesium; however, since in this method samples are not ashed, the small amount of iron is probably removed during deproteinizatiori n ith tungstic acid (IO). I n making the comparison 27 samples were divided into three groups. 1.
Those with values of serum magnesium over 2.00 mg. %.
2. Those with values of serum magnesium between 2.00 and
1
.oo?&.
3. Those with values of serum magnesium under 1.00 mg. %. Group
Av. Serum, Av. Whole Mg. % Blood M g , M g . %
No. of Analyses
Ratio
The tabulated results indicate that the drop in magnesium content of the red corpuscles takes place more slowly than in serum. This is in agreement with the work done by Tufts and Greenberg (IS). The calves in Group 1, which would be considered normal, had a ratio of 1: 1 or the same amount of magnesium in their serum as in whole blood. This is in agreement with the published work of Kunkel et al. (8)and Eveleth ( 1 ) . LITERATURE CITED
(1) Eveleth, D. F.,J . B i d . Chcni., 119, 289-92 (1937). (2) Garner, R. J., Biochem. J . , 40, 828-31 (1946). (3) Gilliam, s. W., IND.ENG.CHEM...\NAL. ED.,13,449-501 (1941). (4) Green, H. H., and XIac*sakill, E. H., J . -4gr. Sci., 18, 384-90 (1928). ( 5 ) Heagy, F. C., Can. J. R e s f a r c h , 26, 295-8 (1948). (6) Hirschfelder, A. D., and Serles, E. It., J . B i d . Chem., 104, 63545 (1934). (7) Hoffman, W.A , , I h d . , 118, 37-45 (1937). (8) Kunkel, H. O., Pearson, P. B., and Schweigert, B. S.,J. Lab. Clz'n. M e d . . 32. 1027-33 (1947). (9) Ludwig, E. E., and Johnson, C. K.,IND.ENG.CHEM.,ANAL. ED.,14, 895-7 (1942). (10) Mikkelsen, D. S., Toth, 5. J , and Prince, A. L., Soil S e i , 66, 385-92 (1948). (11) Orange, Af., and Rhein, H. C., J. Biol. Chem., 189, 379-86 (1951). (12) Simonsen, D. G., Westover, I,. XI., and Wertman, XI., Ibid., 169, 39-47 (1947). (13) Tufts, E. V., and Greenberg, D. II.,Ibid., 122, 715-26 (1938). ~I
RECEIVED for review July 21, 1953. Accepted October 1, 1953.
Determination of Ring Content of Aromatic Petroleum Fractions R. NICHOLS HAZELWOOD Line M a t e r i a l Co. ( A M c G r a w Electric Co. Division), South Milwaukee,Wis.
A
ItOhIATIC fractions from petroleum have been analyzed by dens'ity-temperature coefficient and molecular weight ( 5 ) , refractive index-temperature coefficient and molecular weight (5),and numerous other correlations of physical properties and structure. Recently Martin and Sankin ( 7 ) described a method for determining aromatics in petroleum concentrates using characterization functions derived from molecular weight, density, and specific dispersion. These authors pointed out that it should be possible to eliminate the specific dispersion factor and to substitute refractive index in its place. This is highly desirable because measurement of refractive dispersion is difficult, whereas refractive index can be determined with ease. DERIVATION OF EQUATIONS
Trial attempts to find a group of functions which would give satisfactory calculation of aromatic content indicated that the refractive index did not correlate directly with the aromatic content. Using the "characteristic factors" of Smittenberg (O), it was found that the correlation curves of aromatic ring content, RA, with &fAn were logarithmic in form rather than linear and that there was a n appreciable spread of the values. This spread was later found to be due to total ring content. It was found by trial that the factor M A r , , which is derived from the refractivity intercept, T ( , could be correlated with total ring content, RT, in combination with the factor M a d . In these equations,
ill = molecular weight
- 0.8510 nk0 - 1.4i50
Ad = d:"
An = Ari =
(nk"
-
+ );
Ard = An
-
-
1.0495
Ad 2
RT = total number of rings R A = number of aromatic rings By plotting X A r , versus NAd for the data given in Table I for a series of aromatic petroleum concentrates and alkyl aromatics, a chart is obtained which has isobaric lines for total ring content. This chart is shown in Figure 1. Determination of the slope of the RT lines in Figure 1 permits calculation of relative values of the two fartors. The reciprocal of the slope is incorporated into the expression:
RT = aM(bAr,
+ Ad) + C
ahich is of the form derived by Van Xes and Van Westen (IO) for relating MAn and MAd to RF. l/slope = b in this equation. It was found that b = -0.67. Therefore,
RT
= aiM(Ad
- 0.67 Ar,)
+c
