Quantitative analysis of amines via amine boranes - American

Quantitative Analysis of Amines via Amine Boranes. Robert E. Lyle and Everett W. Southwick. Department of Chemistry, University of New Hampshire, Durh...
0 downloads 0 Views 390KB Size
Quantitative Analysis of Amines V i d Amine Boranes Robert E. Lyle and Everett W. Southwick Department of Chemistry, University of New Hampshire, Durham, N. H. 03824

THE BORON-HYDROGEN bond of borane or diborane has been shown to be hydrolyzed quantitatively with many protic solvents ( I ) . The conversion of the borane to a complex with an amine. has been shown to stabilize the boron-hydrogen bond to hydrolytic or oxidative reaction conditions (2-6). These observations suggested a feasible analytical procedure for estimating the molar quantities of amine based on formation of amine boranes. EXPERIMENTAL

Procedure. A weighed sample of the amine was placed in a flask which had been flushed with dry nitrogen. To this was added 20 ml of dry solvent (tetrahydrofuran in most instances) and the flask was attached to a gas buret. A similar apparatus was prepared in which no amine was added. To each flask was added a measured volume of solution of borane in tetrahydrofuran. After being stirred for 0.5 hour, the solution was treated with 2 ml of methanol, The volume of gas evolved from each reaction flask was measured and the volume from the reaction flask containing no amine was used to calculate the molarity of the boranetetrahydrofuran solution. The evolution of hydrogen was usually complete within 5 to 10 minutes and any reaction after this time probably results from decomposition of the amine borane. Such complications were avoided by changing the solvent to one of higher dielectric constant or by lowering the temperature of the reaction during the analysis. On the basis of the volume of hydrogen evolved from the flask containing amine, an estimation of the amount of excess borane present was made. The difference between this value and the moles of borane added gave the amount of (1) F. E. Martin and R. R. Jay, ANAL.CHEM., 34,1007 (1962).

(2) “Boron Nitrogen Compounds”, Advances in Chemistry Series, Vol. 42, American Chemical Society, Washington, D. C., 1964. (3) H. C. Kelly, F. R. Marchelli, and M. B. Giusto, Znorg. Chem., 3, 431 (1964). (4) H. C. Kelly, ibid., 5, 2173 (1966). (5) M. A. H. Landorf, Ph.D. Thesis, Purdue University, W. Lafayette, Ind., June, 1966. (6) J. T. Murray, Ph.D. Thesis, Purdue University, W. Lafayette, Ind., August, 1963.

borane protected as the amine borane and, therefore, indicated the number of equivalents of amine present. Typical results of this analytical procedure are given in Table I. RESULTS

A procedure for the estimation of amines was developed which involved the treatment of the amine with an excess, but measured volume, of a standardized solution of diborane in tetrahydrofuran. The excess diborane was decomposed by the addition of water or alcohol, and the volume of hydrogen which was liberated was measured. A similar volume of the diborane so1ut:on was hydrolyzed to determine the titre of the solution. The difference in the volume o phydrogen evolved in the two experiments indicated the amount of amine borane which was formed and thus the molar equivalents of borane that were stabilized to hydrolytic conditions. Because the nitrogen-hydrogen bond of amines is not broken by reaction with diborane ( I ) this analytical techn:que may be effective for primary, secondary, or tertiary amines. Such an analytical procedure could be used to determine the quantity of amines in a solution, the molecular weight (equivalent weight) of an amine, or the composition of a mixture of two amines of different molecular weights. The analysis depends on the amine borane being kinetically stable to the hydrolytic conditions used to destroy the excess borane. With the strongly basic, unhindered amines, such as the cyclic amines, there was no difficulty in achieving selective decomposition; however, to increase the rate of hydrogen evolution a solvent of low dielectric constant, such as dioxane, was used. With less stable amine boranes formed with amines of large steric requirements or of weak basicity the selective hydrolysis was more difficult. The amineborane of the large amine, diisopropylethylamine, underwent partial hydrolysis in dioxane; however, a satisfactory analysis was accomplished by using tetrahydrofuran as solvent and a temperature of 0 “C for the methanolysis of the excess borane. With the weakly basic diethylaniline the amine borane was too unstable to be compatible with methanol in either solvent. These results are in agreement with hydrolysis rates reported for amine boranes (3-6). Varying the analytical procedure to fit the require-

Table I. Typical Analytical Results

Amine 1-Methylpiperidine

Amine, mmole Borane, mmole 0.70 1.15 0.76 1.15 or-Methylbenzylamine 0.75 1.07 N,N-Diisopropylethylamine 0.46 1.18 Piperidine 0.95 1.14 a

Hydrogen evolved mla mmoleb 30.8 1.27 27.7 1.15 22.7 0.94 53.8 2.22 14.9 0.62

Borane excess mmolec 0.42 0.38 0.31 0.74 0.21

Determined Experimental amine mmoled error, 0.73 4.3 0.77 1.3 0.76 1.4 0.44 2.3 0.93 2.1

Volume uncorrected for temperature and pressure. Corrected. mmoles Hz mmoles BH8 = 3 . mmoles amine = mmoles BH3 Added - mmoles BH3 excess,

VOL. 40, NO. 14, DECEMBER 1968

e

2201

ments of the amine under analysis permits the estimation of the molar quantity of aliphatic amines with a deviation of about 3 which compares favorably with other analytical procedures for amines (see Table I>. Amines which form borane complexes of low stability, such as aniline derivatives,

may be too easily alcoholized to permit analysis by this method(3-6). RECEIVED for review July 12, 1968. Accepted September 6, 1968.

Chemical Microscopy of 10-Methylacridiniumctaloride Reactions with Platinum Metals and Gold Harold F. Schaeffer Depurtment of Chemistry, Wesfmiizster College, Fulton, Mo. 65251 BY MEANS OF chemical microscopy it has previously been shown ( I ) that an acid solution of acridiniumchloride, without benefit of any other complexing agent, could serve as a desirable reagent for the detection or identification of certain platinum metals and gold when present in acid solutions of their chlorides. The purpose of this paper is to report on the relative efficiency of the derivative 1O-methylacridiniumchloride as a micro reagent for these same metals. EXPERIMENTAL

The reagent solution was a 0.2M solution of 10-methylacridiniumchloride in 1N HCl; this appeared to be near the limit of solubility of the compound at room temperature. Just as in the previous experiments with acridiniumchlo:Ide, test solutions of the respective precious metal ions were prepared by appropriate dilution of stock solutions of gold(III), platinum(IV), palladium@), iridium(III), rhodium(III), and ruthenium(II1); the stock solution of osmium, however, was based on potassium osmate. All stock solutions were in 1N HC1, and all dilutions thereof were made with the 1N acid. Individual tests under the microscope were carried out by allowing a 20-pl reagent drop to flow into a similar droplet of sample solution on a slide. Most observations were made through an 8-nim objective and a 1OX ocular. Tests were considered satisfactory if appropriate crystals separated within 2 minutes. RESULTS AND DISCUSSION

With an acidified osmate solution containing the equivalent of one part osmium in one or two thousand, the reagent caused a prompt separation of an abundance of bright yellow crystals (Figure 1). Many were in the form of plumulose, branched X's, and other aggregates. In plumules the barbs joined the shaft at an angle of approximately 45". By polarized light the crystals appeared faintly dichroic, ranging from a very pale yellow to a slightly deeper shade. As samples were made more dilute there was a decreasing tendency to form the more complex branched aggregates, which were displaced by small yellow prisms. Two principal forms of the latter were observed; one type appeared as parallelepipeds having an extinction angle of approximately 30", while another form resembled slender rods, pointed at both ends, and exhibiting parallel extinction. With an osmium content of only one part in 40,000, none of the plumulose aggregates separated. The reaction permitted the identification of osmium in samples containing as little as one part metal in 150,000, or 7 pprn. (1)

H.F. Schaeffer, Microchem. J.,

2202

a

ANALYTICAL CHEMISTRY

9, 492 (1965).

Figure 1. Crystalline derivative obtained by reaction of 10-methylacridiniumchloride with solution containing osmate Using a 20-pl sample, this was equivalent to the identification of approximately 0.13 pg. For the purpose of merely detecting osmium, assuming the absence of interfering ions, the sensitivity of the reagent could be considered as 0.07 pg, because an adequate crystalline precipitate separated from solutions containing as little as one part osmium in 300,000. Because of the small dimensions of the crystals formed at high dilutions, a 4-mm objective should be employed. At high dilutions, the observations are also facilitated by the use of crossed polars. Solutions of gold(II1) chloride in HCl promptly reacted to form a yellow precipitate of very fine crystals (Figure 2). Even with a gold concentration of only one part in 50,000 a rather abundant precipitate appeared. In general, the crystals were so small that the use of a 4-mm objective was indicated. As a rule, crystals which appeared as elongated rectangles exhibited an extinction angle of 45", while those with pointed ends showed parallel extinction. Identification was possible with a gold concentration below one part in 200,000, or the equivalent of 5 pprn. Another metal for which the reagent showed a high sensitivity was platinum. The yellow precipitate contained small prisms occurring singly, or in the form of crosses, daggers, and some more complex aggregates (Figure 3); there were also occasional parallelepipeds. Many of the individual crystals exhibited parallel extinction, but angles of approximately 10" and 25" were also observed. Positive identifica-