The relative calibration points found in this way agree reasonably well with the absolute values. The mean of 18 estimates of the effective sample volume calculated from chromatograms was 0.095 & 0.005 pl. which agreed well with the value found by weighing, 0.100 =!= 0.005 pl., and thus confirmed the accuracy of the absolute calibrations. It is therefore considered practicable t o calibrate accurately the response of sensitive gas chromatographic systems by a combination of the absolute method described here and the relative method
already widely applied and used as a check in this work.
Thompson, R. J., Proc. of Bnd Sqinp. on Gas Chromatography, Amsterdam (1958) p. 165, Butt,erworths, London, 195s.
LITERATURE CITED
(1) Beare, W. G., McVicar, G. A., Ferguson, J. B., J . Phys. Chem. 34, 1310 (1930). (2) Butler, J. A. V., Thompson, D. W., Maclennan. W. H., J . Chem. SOC.1933, 674. (3) Desty, D. H., Geach, C. J., Goldup, A., Proc. of 3rd Symp. on Gas Chromatography, Edinburgh (1960), p. 46. Butterworths, London, 1960. (4) Loyelock,. J. E., Zbid., p. 16. (5) Primavesl, G. R., Oldham, G. F.,
( 6 ) Purnell, J. H., Ilept. of Physical
Chemistry, Cambridge, England, private communication, 1962. ( 7 ) Scott, R. R., Stannard, B. W., Chem. Zntl. (London) 1960, 1259. (8) . . Swoboda. P. A. T.. Zbid., P . 1262. and Proc. of 3rd Sump: on G& tography, Edinburgh (1960) p. 354,
Butterworths, London, 1960. M. G. BURNETT P. A. T. SWOBODA Low Temperature Research Station Cambridge, England
Titration of Aromatic N-Oxides in Acetic Anhydride with Perchloric Acid SIR: Wimer has reported that both amides (9) and sulfoxides (10) when dissolved in acetic anhydride can be titrated with perchloric acid in acetic acid. Following the completion of the present work i t was learned that Wimer had also titrated certain amine oxides (11) dissolved in acetic anhydride with perchloric acid in dioxane. Mackenzie and Winter (6-8) and Burton and Praill (1, 8) have presented considerable evidence that. a solution made from perchloric acid, acetic anhydride, and aceti: acid has CH~COZH‘, (CH3CO)zOH-’, and CH3CO+as the acidic species in equilibrium and that their acidities increase in the order listed. In this work eight aromatic N-oxides dissolved in a mixture of acetic anhydride and acetic acid have been titrated with perchloric acid in acetic acid. The results of the titration studies are summarized in Table I.
N-acetoxy tertiary amine perchlorates were isolated from many of the titrated solutions. These have been identified by elemental analysis and by conversion to the parent N-oxide when treated with sodium hydroxide. For example, N - acetoxy - 2 - methylquinolinium perchlorate, m.p. 152-3’ C., (Anal. Calcd. for C12H12N02: C, 47.77; H, 4.01; N, 4.64. Found: C, 47.76; H, 4.23; N, 4.52) yields 2-methylquinoline N-oxide dihydrate when reacted with sodium hydroxide. Based on the isolation of the Nacetoxy tertiary amine perchlorates the success of the titrations could be due to the reaction of either CH&O+ or (CH3C0)20H+ or both with an Noxide. It is postulated that the Noxides in the present study reacted predominately with protonated acetic anhydride as shown by the following equation.
Table 1. Titration of Aromatic N-oxides in Acetic Anhydride with Perchloric Acid
A’-oxide of Pyridine 2-Methvlpyridine 3-Methylpyridine 4-Methylpyridine 2,BDimethylDvridine &&Loline Iaoquboline 2-Methylquinoline dihydrate 2-Carboxypyridine 4-Nitro-2methylpyridine
Meall Calcd. Observed Neut. Neut. VariEquiv. Equiv. ance 95.1 95.5 0.1 109.1
109.6
0.1
109.1
109.5
0.2
109.1
109.2
0.2
123.1 145.2 145.2
123.3
146.3
145.3
0.3 1.3 0.2
195.2
195.7
0.5
139.1 Unsucc. 154.1 Unsucc.
d r-coca, Jaffe (5) has found that pyridine Noxides substituted with electron-withdrawing groups are weaker bases than are pyridine N-oxide and its homologs. Wimer (11) reports that N-oxides of 4 nitro- and 4-cyano-pyridine can be titrated in a system in which CHaCO+ is likely to be present, whereas in the present work N-oxides of 2-methyl-4 nitropyridine and 2-carboxypyridine could not be titrated. If CHIGO+ is a
,oo/ 700
200
y.
I
0
2
4
6
0
10 12
14
16 18
2 0 22 24 26
ML. O F O.SN ncio4
Figure 1 . Tiiration of 3-picoline-Noxide ( 1 ) and isoquinoline-N-oxide (2)
stronger acid than (CH3COz)zOH+, then i t is reasonable that the difference between Wimer’s results and our results is due to a predominance of CH3CO+ as the acidic species in his titrations, whereas (CH8C0)20H+predominated as the acidic species in our titrations. Also, it is possible for some aromatic Noxides to react with xetic anhydride without the presercc I acid (4). Additional subs4 uted pyridine N oxides and somc aliphatic N-oxides will be studied L~ an effort to learn more about ’ le generality of the titration method. EXPERIMENTAL
The N-oxides listed in Table I were distilled or recrystallized prior to dissolving them in 10 to 20 ml. of acetic acid. The samples of N-oxide usually weighed between 0.2 and 1.5 grams. Generally 10 to 25 ml. of acetic anhydride was added as well as 2 to 3 drops of methyl violet solution, and the resulting solution was titrated immediately with 0.1 to 0.5M perchloric acid in acetic acid (3). A Beckman Model H2 glass electrode pH meter with a glass electrode and a calomel electrode was used. For convenience the titrant VOL. 34, NO. 9, AUGUST 1962
1163
was added rapidly until the indicator was green. After the needle on the millivolt scale assumed a constant value, additional titrant was added in increments of 0.1 ml. The end point was determined by noting which 0.1-ml. increment gave the largest change in millivolts. At the end point the indicator gradually changed from green to yellow. In the few cases in which i t was tried, back titration with sodium acetate in acetic acid was satisfactory. Figure 1 shows the potentiometric infections obtained for the titration of two compounds. ACKNOWLEDGMENT
F. E. Cislak, Reilly Tar and Chemical Corp., Indianapolis, Ind., gave us
generous samples of many of the Noxides studied. For this help we are very appreciative.
(8) Zbid., pp. 243-53. (9) Wimer, D. C., ANAL.CHEW30, 77-80 (1958). (10) Ibid., p. 2060. ( 1 1 ) Zbid., 34, 873 (1962).
CHESTER W. MUTH ROBERT S. DARLAK H. ENQLISH WILLIAM ALLENT. HAMNER
LITERATURE CITED
(1) Burton, H., Praill, P. F. G., J. Chem. SOC.1950, 1203-6. (2) Ibid., pp. 2034-8. (3) Fritz, J. S., “Acid-By; Titrations in Nonaqueous Solvents, p. 13, a.
Frederick Smith Chemical Co., Columbus, Ohio, 1952. (4) Furukawa, S., Yakugaku Zasshi 79, 492-9 (1959); C . A . 53, 180296 (1959). (5) Jaffe, H. H., Doak, G. O., J. Am. Chem. SOC.77, 4441-4 (1955). (6) Mackensie, H. A. E., Winter, E. R. S.,
Trans. Faraday SOC.44, 159-70 (1948). (7) Zbid., pp. 171-81.
Dept. of Chemistry West Virginia University Morgantown, W. Va. I n part from Robert S. Darlak’s M.S. Thesis at West Virginia University (1961).
Robert S. Darlak held an N.D.E.A. Fellowship, and William H. English and Allen T. Hamner held N.S.F. Summer Undergraduate Fellowships at West Virginia University during 1960 and 1961, respectively.
Spectrophotometric Ultramicrodetermination of Inorganic Phosphorus and Lipide Phosphorus in Serum SIR: Adaptations of the Fisk-SubbaRow microprocedure (6) have been reported (4, IO, 11), but improvement in its sensitivity is desired. Heat has been applied to intensify the color of the mixture (2), and the method has recently been revived for a sensitive phosphorus assay in lipide column chromatography (1). The procedure described in this paper’ involves: (1) increasing the acidity of the sodium trichbroacetate (TCA) filtrate with relatively strong sulfuric acid and adape ing the heating modification of Bartlett ( I ) for a sensitive inorganic phosphorus assay; (2) providing simultaneous lipide extraction after the TCA aliquot is removed, since the plasma phospholipide can be quantitatively precipitated by TCA (13) and extracted by alkalinealcoholic solution (14). In the method described, the amount of serum used is approximately l/m of that required by the Fisk-SubbaRow procedure. PROCEDURE
Reagents and Materials. Hydrogen peroxide (30%, Baker’s, phosphorus-free) and ammonium molybdate (5% in deionized water) are used. The Fisk-SubbaRow reagent is prepared by adding 250 mg. of purified l-amino-2-naphthol-4-sulfonic acid (Eastman Organic Chemicals) with mechanical stirring to 100 ml. of freshly prepared 15% sodium bisulfite (anhydrous), followed by 500 mg. of anhydrous sodium sulfite. The solution is filtered and stored in a dark bottle. It is freshly prepared weekly. All the tubes are soaked in 5Oy0 nitric acid overnight, rinsed with distilled deionized water five times, and dried. 1164
ANALYTICAL CHEMISTRY
Inorganic Phosphorus. From 5 to 10 pl. of serum is added to 100 pl. of water in a 10-ml. centrifuge tube. After mixing, 40 pl. of 40% trichloroacetic acid is added, mixed with a mixer (Vortex Jr.) for 1 minute, and allowed to stand for 5 minutes in an ice water bath. The mixture is centrifuged for 15 minutes a t 2500 r.p.m. in a refrigerated centrifuge and 100 pi. of the supernatant fluid is transferred into another centrifuge tube. Then 300 pl. of 10N sulfuric acid, 250 pl. of water, and 200 p l . of 5’% ammonium molybdate are added. After mixing, 50 pl, of FiskSubbaRow reagent is added a t the bottom of the tube and mixed well. The solution is placed in a boiling water bath for 7 minutes and then cooled in an ice water bath. The color is read a t 830 mr with a Beckman spectrophotometer with a red-sensitive phototube. A microcuvette, 4 X 10 x 3 mm. (Pyrocell Manufacturing Co.), is used. A standard solution containing from 0.1 to 1.0 pg. of phosphorus is made for the preparation of the standard curve. In an alternative procedure for the Coleman Jr. spectrophotometer. 20 pl. of serum in 100 pl. of water is precipitated with 30 fil. of 50% TCA. The final color mixture is read a t 700 mp, using a microspace adapter (Coleman 6-319). Lipide Phosphorus. After the supernatant fluid is decanted, the precipitate is washed twice with 0.5 ml. of 5y0 TCA. Then 0.2 ml. of 0 . 2 N potassium acetate in absolute ethanol is added to the precipitate and mixed well with the mixer for 1 minute. The tube is allowed to stand 10 minutes a t room temperature, and the contents are mixed and centrifuged. The supernatant fluid is carefully transferred to a borosilicate glass centrifuge tube. using a capillary pipet with constricted end. The precipitate
is re-extracted twice with 0.3 ml. of absolute ethanol and the extract is added to a corresponding centrifuge tube. The collected lipide extracts are evaporated to dryness, 0.5 ml. of 10N sulfuric acid is added, and the extracts are placed in an oven at 230” C. for 10 minutes. After the tube is cooled, 1 drop of H202is added and the tube is placed back in the oven for 5 minutes. Excess peroxide is removed bv adding 1 drop of 5% urea solution and reheating for 5 minutes. When the tubes are cool, 1.3 ml. of water and 0.4 ml. of 5% ammonium molybdate are added to the digest. After mixing, 100 pl. of FiskSubbaRow reagent is added, and the procedure for inorganic phosphorus determination is followed. A standard curve must be prepared. RESULTS AND DISCUSSION
A known control serum (Versatol) was diluted to various concentrations and used to compare the ultramicromethod with the Fisk-SubbaRow method. Each filtrate was individually assayed according to the procedure outlined in Table I. The color produced by the two methods was read both a t 830 rnb with the Beckman DU and a t 700 mp with the Coleman Jr. spectrophotometer. Beer’s law was followed (Figure 1). The spectra of the color produced in these two methods were compared (Figure 2). By these figures, it is clearly demonstrated that the color produced by the present method is about 20 times (at 830 mp) or 5 times (at 700 mp) as intense as that obtained by the Fisk-SubbaRow procedure. The pattern of the spectrum obtained from the inorganic phosphorus on TCA filtrate is very closely correlated with that of lipide phosphorus previously
