Spot Test for Diketones and Quinones Based on Catalytic Effect

File failed to load: https://cdn.mathjax.org/mathjax/contrib/a11y/accessibility-menu.js. ADVERTISEMENT · Log In Register · Cart · ACS · ACS Publicatio...
3 downloads 7 Views 348KB Size
V O L U M E 28, NO. 3, M A R C H 1 9 5 6

397

tions followed a straight line which was parallel to the line eypected if no iron were present. Hence, the amount of fluoride complexed by iron remained constant and independent of the amount of rare earth ion present. Kickel also interfered, but not in a predictable fashion. The data on nickel are given to show that it interferes, but they are not necessarily a guide for estimating its effect. The data for the titration deviate from a straight line and give a poor end point. Therefore, nickel should be eliminated prior to the determination of fluoride. Cliromium appears t o interfere in a predictable fashion. Data in Table VI1 show that about half of the fluoride is removed as a nontitratable species. The data show that for each mole of chromium present, there are about 2 moles of fluoride removed as a complex ion; hence, the interfering species can be depicted as difluochromium(II1) ion, which Wilson and Taube have described (8). According to their data, this ion should be formed in high yield under the conditions of the experiment described here. As with iron, the titration lines were parallel to the line expected if no chromium were present. Improvement of Procedure. I t should be possible to improve the accuracy of the titrations by measuring the excess titrant much closer to the end point-e.g., within 10 or 20y0 excess of t it rant,

As lanthanum fluoride is more insoluble than samarium fluoride ( I ) , perhaps more dilute fluoride could be titrated with lanthanum ion. I n order to ensure complete carriage of the europium tracer, the precipitation and flocculation should be rapid. Radioactive lanthanum-140 could be used as the tracer, but ita half life of 40 hours is too short to be practical for routine work. ACKNOW LEDGM EST

The authors wish to thank Sue Krainock and Louis Geoffrion for technical assistance. and J. F. Suttle for the tracer. LITERATURE CITED

(1) Kury, J. W., Univ. of Calif., thesis, July 1953; U. S. Atomic

Energy Commission Document UCRL-2271.

(2) Langer, Alois, ANAL.CHEM.22, 1288 (1950). (3) Langer, Alois, J . Phys. Chem. 45, 639 (1941). (4) Lockett, E. E., Thomas, K. H., .Vucleonics 1 1 , S o . 3, 14 (1353) ( 5 ) Popov, A . I., Knudson, G. E., AKAL.CHEM.26, 292 (1954). (6) Reynolds. D. S., Hill, W. L., I N D . ENG.CHEM.,A N A L . E D . 1 1 , 21 (1939). (7) Willard, H. H., Winter, 0. B., Ibid., 5, 7 (1933). (8) Wilson, -4. S., Taube, Henry, J . Am. Chem. SOC.74, 3509 (1952).

RECEIVED for review August 4, 1955. -4ccepted December 9, 1955. Work U. S. Atomic Energy Cornmission.

was done under auspices of

Spot Test for Diketones and Quinones Based on Catalytic Effect FRITZ FEIGL and CLAUD10 COSTA N E T 0 Laboratorio d a Produfio Mineral, Ministerio d a Agricultura, Rio d e Janeiro, Brazil

Translated by R A L P H E. OESPER University o f Cincinnati, Cincinnati, O h i o

The slow reaction between formaldehyde and 1,Zdinitrobenzene, which yields a violet alkali salt of the aci- form of o-nitrosonitrobenzene, is hastened by the addition of 1,Z-diketones and quinones. It is assumed that an intermediate catalysis is involved. Microgram quantities of the catalytically active compounds can be detected by drop reactions if certain simple conditions are maintained. New microtests for anthracene, phenanthrene, and inositol are made possible by the ready conversion of these compounds into anthraquinone, phenanthraquinone, and cyclic polyketones, respectively.

F

ORMALDEHYDE functions as a hydrogen donor in alka-

line solution and thus reduces 1,a-dinitrobenzene to the n c i - form of 1,2-nitrosonitrobenzene ( I ) , giving a violet alkali salt. However, the reaction

be due to the fact that formaldehyde in alkaline solution reduces them to 1,2-hydroxyketones as shown in Equation 2, and these products in turn reduce the 1,2-dinitrobenzene to the violet o-quinoidal alkali salt as shown in Equation 3. The diketone is thus regenerated and can react again according to Equation 2. Reactions 2 and 3, which occur again and again, proceed faster than Reaction 1. Addition of the partial Reactions 2 and 3 gives a net reaction xhich is identical with Reaction 1, in nhich the diketone does not appear, even though it is continuously consumed and regenerated. 2

2

-co -Lo

+ 2CHzO + 2 0 H -

-CHOH 1

-co

+

-CHOH --t

2

u=so*-

+ 2 H C 0 0 - + 3Hz0

(1)

proceeds slowly even with relatively large quantities of formaldehyde in sodium carbonate solution, and hence, has no analytical value. The redox reartion has been found to be rapid in the presence of 1,a-diketones and quinones, which act as catalysts. This effect is the basis of a sensitive and specific test, xhich may be used as a spot reaction, for these compounds. The catalytic hastening of Equation 1 by 1,a-diketones may

+ 2HC00-

(2)

+ 3HzO

(2)

+ 20H-=KO

-co

-

=SO>-

fi-so-

1

-CO

t 2

1

-co

There is no doubt as to the existence of Reagtion 3, because microgram quantities of acyloins and benzoins (compounds containing the group -CHOH-CO-) have been found to yield the violet color characteristic of organic hydrogen donors when they are warmed with an alkaline-alcohol solution of 1,2-dinitrobenzene ( 2 ) . No reports could be found in the literature about the reduction of diketones by formaldehyde as in Reaction 2. Attempts to convert benzil quantitatively into benzoin by warming an alcohol solution with a sodium carbonate solution of formaldehyde failed. However, the evaporation residue of such a reaction mixture contained benzoin, as revealed by the color reaction with l,2-dinitrobenzene and also by the production of hydrogen

ANALYTICAL CHEMISTRY

398 sulfide on heating with sulfur to 140" C. Production of hydrogen sulfide on fusing with sulfur is characteristic of benzoin and for compounds with secondary alcohol groups ( 3 ) . A comparison test with the evaporation residue of a sodium carbonate-formaldehyde solution gave no color reaction with 1,2-dinitrobenzene and only a weak hydrogen sulfide reaction after heating with free sulfur. The latter response was due to the formation of some hydrogen sulfide when hydrous sodium carbonate plus sulfur is heated to 140' C., by the action of the superheated water vapor released a t this temperature. These findings are additional supporting evidence for the mechanism proposed above. Despite the fact that Reaction 2 gives but slight yields of 1,2-hydroxyketone, its participation in the catalysis is entirely feasible. The decisive factor is not extensive reaction, but only whether such amounts of 1,2-hydroxyketones as are produced can participate rapidly in Reaction 3.

treated with 1 drop each of 25% sodium carbonate solution, 4y0 formaldehyde, and a 5% solution of 1,2-dinitrobenzene in benzene. The mixture is shaken and placed in a boiling water bath. The shaking is repeated intermittently. A more or less intenee violet color appears in 1 to 4 minutes, depending on the quantity of catalytically active material present. 9 blank comparison test is recommended. These amounts were detected: Compound Y Diacetyl 0.05 Bend

2

Furil

0.2

Phenanthraquinone

0.002

3-Sitrophenanthraquinone

0.002

Ninhydrin

0.5

,Jr- co-co J,,I

DETECTION OF 1.2-DIKETONES AND QUINONES

If the catalytic action of 1,a-diketones is to be employed in analysis, it is necessary to maintain conditions a t which the uncatalyzed redox reaction, Equation 1, proceeds with the lowest possible velocity, because the test is based on the establishment of the differences in the reaction rates. After many trials it was found that the greatest reliability and sensitivity are attained by conducting the reaction in strong carbonate solution a t the temperature of a boiling water bath, and by using a benzene solution of l,2-dinitrobenzene. This may be obtained in surfacerich form by evaporating off the benzene and dissolving the residue in hot water to form an approximately 0.m solution. Under these conditions, Reaction 1 becomes apparent only after 4 to 5 minutes, whereas even minimum amounts of catalytically active diketones produce the color reaction much sooner. Solutions of the test material may be used if necessary. It was found that quinones exhibit a similar catalytic action. This is probably ascribable to the fact that these quinones lead to hydroxy compounds via redox reactions which are analogous to Reaction 2. O = a = O

-

0

30

SH 2-Methyl-l,4-naphthaquinone (vitamin K:)

0.01

+ 2CH20 + 2 0 H -

-

=OH

+ 2HC00+ 2HC00-

OH

03':" 11

0

Sodium 1,2-naphthaquinone- 0 . 5 4-sulfonate

\O These products function as hydrogen donors to 1,2-dinitrobenzene and so bring about the color reaction. Consequently, it is advisable to make a preliminary trial with the sample in the absence of formaldehyde, using the prescribed procedure. If no color appears, or a pale violet a t most, the test should be repeated with the inclusion of formaldehyde. A violet color or a more intense result of the color reaction is then proof of the presence of catalytically active 1,a-diketones or quinones. A simple procedure for detecting 1,a-diketones in the presence of numerous reducing agents, which give the color reaction without the addition of formaldehyde, consists of adding alkali hypobromite to oxidize the sample and then proceeding with the catalysis reaclion. (Compare with the detection of inositol in the presence of reducing sugars, described later.) However, the interference due to quinones and polyphenols, which are oxidized to quinones, is not avoided by this modification. If 1,Pdinitrobenzene is used, a red color develops. The limits of identification were not determined for this reagent; they seem to depend on the nature of the test material. I t is probable that 3,Pdinitrobenzoic acid may also be used as the reagent, because it has been employed (6) for the chromatographic detection of reducing sugars

Procedure. The test is conducted in a micro test tube. One drop of the aqueous or benzene solution of the test material is

0 !I

+ 2CHz0 + 2 0 H - HO(T>OH

0

Isatin

Anthraquinone

0 'I

03"

0.05

$0 0

Sodium anthraquinone2-sulfonate

0.5

0

8 Sodium rhodizonate

0.5

0

V O L U M E 28, NO. 3, M A R C H 1 9 5 6 Dehydroascorbic acid, p-benzoquinone, anthraquinone disulfonic acid, and chloranil gave strong responses. The fact that vitamin Ka catalyzes the reaction makes it probable that vitamins KI and Kz, which also contain a naphthaquinone nucleus, would function in this manner. Sodium dehydrotartrate showed no catalytic effect despite the fact that it is a 1,2-diketone. The sensitive detectability of 1,a-diketones and quinones, through their catalytic activity in the formaldehyde-1,2dinitrobenzene system, makes possible new tests for anthracene, phenanthrene, and inositol, because these compounds are easily converted into corresponding catalytically active compounds. DETECTION OF ANTHRACEIVE AND PHENANTHRENE

Anthracene and phenanthrene are converted into anthraquinone and phenanthraquinone, respectively, by evaporation with concentrated nitric acid. These products react satisfactorily with the reagent mixture, if the reagent is precipitated in a surface-rich form from its benzene solution by evaporation of the solvent. Procedure. One drop of the benzene solutionpf the sample ip evaporated to dryness in a micro test tube. A drop of concentrated nitric acid is added and the evaporation is repeated. A drop or two of benzene is added to the evaporation residue, and the procedure for 1,a-diketones and quinones is then followed. The depth of the color indicates the quantity of anthracene or phenanthrene involved. Limit of identification is 2 y of anthracene; or 3 y of phenanthrene. Because nitric acid merely nitrates naphthalene, this test provides a means of detecting anthracene in naphthalene if a comparison is run with pure naphthalene.

399 which are based on the production of colored alkali earth salts of rhodizonic acid (6). Procedure. A drop of the aqueous solution to be tested for inositol is evaporated to dryness in a micro test tube. One drop of concentrated nitric acid is then added and the evaporation is repeated to remove the unused nitric acid. The residue is then treated according to the procedure for diketones and quinones. Limit of identification is 5 y of inositol. The foregoing test may not be employed directly if reducing sugars are present, even though the sugars are partially converted to oxalic acid by evaporation with concentrated nitric acid. Consequently, a test for reducing sugars must be made prior to the evaporation. This is done by adding a drop of the benzene solution of l,2-dinitrobenzene to the sodium carbonate test solution and warming. If reducing sugars are present, a violet color develops. If inositol is to be detected in the presence of reducing sugars or ascorbic acid, these can be quantitatively oxidized by means of alkali hypohalogenite which does not affect inositol. Procedure. The test is conducted in a micro test tube. A drop of the aqueous test solution is treated with 1 drop of strong bromine water and 1 drop of 0.5N sodium hydroxide. The mixture is warmed for 1 to 2 minutes in a water bath. The excess hypobromite is then decomposed by adding a drop of 10% hydrogen peroxide solution and evaporating to dryness. The evaporation residue is then treated as described for the detection of inositol. This procedure satisfactorily revealed 10 y of inositol in the presence of 1000 y of glucose. ACKNOWLEDGMENT

Ernesto Silva collaborated in working out the test for inositol. The support of the Conselho Nacional de Pequisas is also gratefully acknowledged.

DETECTION OF INOSITOL LITERATURE CITED

As shown above, the formaldehyde-1,2-dinitrobenzene reaction is catalyzed by sodium rhodizonate or rhodizonic acid. Although sodium rhodizonate is readily oxidized, by itself it has no action on 1,2-dinitrobenzene because it is not a hydrogen donor. Inositol is converted by concentrated nitric acid into rhodizonic acid ( 4 ) along with other aliphatic cyclic polyketones. A new test for inositol has been developed from these facts. This test is far more sensitive than the tests previously used,

(1) Bose, P.K., 2. anal. Chem. 37, 110 (1932). (2) Feigl, F.. Stark, C., ANAL.CHEM.27, 1838 (1955). (3) Feigl, F..VokaE, L., Mikrochim. Acta 1955, 101. (4) Fleury, P., Balatre, P.. "Les Inositols," Chap. 11, Masson & Cie.,

Paris, 1947. (5) Ibid., Chap. 11'. ( 6 ) Weygand, F., Hofmann, H., Ber. 83, 405 (19.50).

RECEIVED for review August 1, 1955. Accepted December 20, 1955.

Photometric Determination of Boron in Titanium and Its Alloys R. C. C A L K I N S a n d V. A. STENGER Main Laboratory, The Dow Chemical Co., Midland, Mich. A method was desired which would incorporate the precision of a photometric method into the determination of small amounts of boron in titanium. Application of the carminic acid method for boron requires complete elimination of the titanium. Conditions have been found under which this may be done by cation exchange removal of the peroxy-titanium(1V) complex. Boron is determined on the residue from a neutralized and evaporated aliquot of the column effluent. Vanadium is the only interfering element that has been encountered. This interference is negligible if the alloy contains less than 0.3% vanadium with 0.005 to 0.10% boron, or if not more than 10 times as much vanadium as boron is present in the higher ranges.

I

N A method for the determination of boron in titanium and its alloys, recently published by Norwitz and Codell (6), the sample, dissolved in hydrochloric acid, is passed through a cation exchange column to remove most of the titanium. The remainder is precipitated by boiling with calcium carbonate, after which boric acid is determined by titration with a standard base in the presence of mannitol. It occurred to the authors that the method might be made more precise for small quantities of boron if a photometric determination, such as that with carminic acid (2, 7 ) , could be substituted for the titration. Preliminary experiments revealed that titanium in quadrivalent form interferes strongly with the carminic acid test for boron. Absorption spectra for the complexes, presented in Figure 1, illustrate the extent of this interference. -4s attempts to keep