Specific Spot Detection of Chlorine in Inorganic and Nonvolatile Organic Compounds FRITZ FEIGL, ERWIN JUNGREIS, and VARDINA LIPETZ Laboratorio da Proclugio Mineral, Ministerio da Energia e Minas, Rio de Janeiro, Brasil, and Department o f Inorganic and Analytical Chemistry, Hebrew University, Jerusalem, Israel
b A new method was developed for the detection of chlorides in the presence of bromides cind iodides, based on the quantitative elimination of the latter by treatment of the test solution with permolybdic acid and sulfosalicylic acid, and subsequent precipitation of silver 'chloride. The upplication for the detection of organic bound chlorine in nonvolatile compounds is described.
I
T IS WELL K N O T ~ P ; that
if halides are heated with lead dioxide in the presence of acetic x i d , free bromine or iodine is liberated, whereas it is assumed that chlorides remain unaltered. This behavior should offer the basis for the detection of chlorides in miature \\ith o ,her halides after the elimination of the latter. ,4s analytical papers contain no references to this subject, we have examined the pertinent redox reartions to find the conditions for a Fossible analytical application. The following experiments have given an indication as to the behavior of halogen ions when treated with lead dioxide anc acetic acid. 1. To three test tubes, respectively, containing 3 ml. of O.lyo solutions of sodium chloride, sodium bromide, and sodium iodide, a few centigrams of lead dioxide and 2 t o 3 drops of concentrated acetic arid were added. The tubes were immersed in a bath of boiling water, arid each top was covered with a disk of filter paper moistened with an aqueous solution of p-amino dimethyl aniline chloride. A reddish-violet fleck (V7urster Red) indicates the formation of free halogens (4). This color reaction resulted immediately from broinide- or iodidecontaining mixtures, in contrast to chloride-containing mixture where no chlorine is formed even after heating for 10 minutes. 2. The bromide- and iodide-containing mixtures as used in 1 were heated for 10 minutes in boiling water, then filtered, and to the clear filtrate a fern drops of silver nitrate solution were added. Yo prec pitation of silver iodide occurred, wherws a strong precipitate of silver bromide was observed. This behaviclr indicates that the redox reaction wh2re.
PbOz
+ 21' + 4H4
-P
Pb+'
+ 2HzO + Iz
(1)
is quantitative, as the analogous redox reaction PbOz
+ 2Br' + 4"s Pb+,
+ 2H20 + Brz
(2)
leads to an equilibrium and therefore does not result in the quantitative removal of bromide. Thus the above reactions, followed by filtration and precipitation with silver nitrate, make possible the detection of chloride in mixture only with iodide, and not with bromide. The same results are obtained when manganese dioxide is substituted for lead dioxide in the above experiments. -4s shown in Equation 2, only the quantitative removal of bromine from the equilibrium may lead to the complete disappearance of bromide. For this purpose heating under the above conditions is not sufficient. Davidson and Pearlman (1) claim that prolonged heating (evidently longer than 10 minutes) eliminates the bromine. The reason may be that through the following partial reactions bromide is newly formed :
+ HBrO Pb+2(Mn+2)+ BrO' + H20 Pb02(Mn02)+ 2H+ + Br' Brz
+ H20
-+
HBr
(3) -P
(4)
It seems probable that a rapid elimination of free bromine from Equilibrium 2 may lead to a quantitative elimination of bromide ions. To achieve this aim, we have repeated the experiments described in 2 in the presence of sulfosalicylic acid which reacts quantitatively with free bromine according to:
+
C6H,(OH)(COOH)S03H BrZ+ C&Br (OH) (COOH)S03H HBr
+
(5)
Although hydrobromic acid is formed in Reaction 5, which has already been used for the elimination of free bromine, (Z), it may be expected that in the presence of excess manganese (lead) dioxide it would in turn be oxidized according to Equation 2. The complete elimination of bromide in this fashion is demonstrated in the following experiment. Three milliliters of 0.1N alkali bromide, a few centigrams of manganese
(lead) dioxide, and a few centigrams of sulfosalicylic acid were heated together in a test tube for some minutes. After filtration, a solution was obtained which remained clear upon addition of silver nitrate. Therefore bromides as well as iodides are quantitatively eliminated by heating with lead (manganese) dioxide, acetic acid, and sulfosalicylic acid. As chlorides remain unaltered by this treatment, their detection becomes possible, through precipitation u i t h silver nitrate. I t was to be expected that in the redox reaction which leads to the formation of free bromine or iodine the lead dioxide (manganese dioxide) could be replaced by other water-soluble osidants reacting faster than the solid oxidants. It was indeed found that this end could be achieved by hydrogen ~ peroxide activated by h l 0 0 ~ - ions (permolybdic acid) ( 3 ) . When a solution containing bromide or iodide is heated for 5 minutes with hydrogen peroxide-molj.bdate miature and sulfosalicylic acid, solutions are obtained which remain clear after addition of silver nitrate. It is noteworthy that heating of sodium iodide with permolybdic acid produces not only free iodine, but also iodic acid, which is precipitated with silver nitrate. However, when the treatment is carried out in the presence of sulfosalicylic acid, no iodate is formed. Hence it is evident that the primary formed iodine is fixed by sulfosalicylic acid, thus preventing its oxidation to iodate. The sulfosalicylic acid therefore acts under these conditions as an acceptor for free bromine as well as for free iodine; in the latter case probably an iodosulfosalicylic acid is formed. The specific detection of chlorides described here makes possible the detection of chlorine in nonvolatile organic compounds, provided a preliminary ignition with manganese dioxide is carried out. Thus the carbonhydrogen skeleton is destroyed liberating carbon dioxide, and water and the primary formed halogen hydride reacts with manganese dioxide according to: MnO,
+ 4HHal+
MnHalz
+ 2Hn0 5 Hal,
VOL. 36, NO. 4, APRIL 1964
e
885
hydrogen peroxide and concentrated acetic acid to which are added a few drops of 0.5M ammonium molybdate.) The tube is immersed in a boiling water bath for 10 minutes. After cooling, the solution is tested with 5% silver nitrate solution. A positive response is indicated by the appearance of silver chloride percipitate. Limit of identification: 10 pg. of sodium chloride in the presence of a few milligrams of sodium bromide and iodide. B. Test for chlorine in nonvolatile organic compounds. PROCEDURE. Minute quantities, or a drop of the solution of the test materia1 are placed in a micro test tube, mixed with a few The test for chlorides and for chlocentigrams of kIn02, and evaporated rine in nonvolatile organic compounds if necessary. The test tube is then can be carried out with macro amounts strongly heated over the flame of a as well as by the technique of spot micro-burner. After cooling, a few test analysis. The latter is described drops of dilute acetic acid are added, here. heated in a water bath, and then filtered A. Test for chloride in presence of bromide and iodide. PROCEDURE.or centrifuged. The remainder of the procedure is the same as described in A. In a micro test tube, a drop of perLimits of identification are 20 pg. of molybdic acid and some centigrams of p-chloro aniline; 20 fig. of chloro sulfosalicylic acid are added to a drop of acetic acid; 20 bg. of chlorobenzothe test solution. (Permolybdic acid is prepared from a 1: 1 solution of 30% phenone; 20 pg. of chloral hydrate;
I n this way, half of the halogen remains in the ignition residue as manganese halide side by side with unaltered manganese dioxide (or Mn20a formed from the MnOz). When this residue is heated with acetic acid and filtered (centrifuged), a solution of manganese halides is obtained, which can be examined for chloride after oxidative removal of bromide or iodide by heating with permolybdic acid and sulfosalicylic acid. Accordingly, the detection of chloride in organic compounds is possible even in the presence of organic bound bromide and iodide.
20 pg. of chlorobenzene; 20 pg. of o-chloro aniline; 20 pg. of 1-chloro naphthalene; 20 pg. of p-chlorotoluene; 5 pg. of 5;chloro-8-oxyquinoline; 10 pg: of 3,4-dichloroaniline; 16 pg, of 2,4-dinitrochlorobenzene. ACKNOWLEDGMENT
We are grateful to the Conselho National de Pesquisas (Rio de Janeiro) and the Department of Inorganic and Analytical Chemistry of the Hebrew University, Jerusalem, for material support LITERATURE CITED
(1) Davidson,
D., Pearlman, D., "A Guide to Qualitative Organic Analysis," 2nd ed., p. 15, New York, 1952. (2) Feigl, F., ANAL. CHEM. 27, 1318 (1955). (3) Feigl, F., "Chemistry of Specific,
Selective, and Sensitive Reactions," Chap. 5, Academic Press, New York, 1949.
(4)-T;emain, H. E., IND.ENG.CHEM., ANAL. ED. 3, 225 (1931). RECEIVEDfor review July 1, 1963. Accepted November 4, 1963.
Determination of Molybdenum and Sulfur in Solid Lubricants 1. L. KALNIN American-Standard Research Division, New Brunswick, N. 1.
,A combustion 'flow method has been developed in which wet oxygen is used to oxidize and separate molybdenum and sulfur (and halogen) contuined in solid lubricants from the underlying substrates in the presence of organic lubricant material. The results indicate that satisfactory analysis can be done on lubricant applied to certain Mo- and W-containing alloy (Hastelloy) substrates.
T
o EVALUATE the perrormance of solid lubricants based on MoS2, it is often desirable and sometimes necessary to determine their molybdenum and sulfur content before and after the experiment. Some of the MoSz may have reacted! ,particularly when subjected to oxidizing or corrosive ambients at elevated temperatures. The MoSz powder may begin to oxidize at 100' C. or less, depending on the particle size and humidity ( 1 4 ) . For sufficiently fine powders under humid conditions, a slow oxidation occurs even at room temperature (17'). According to differential thermal analysis experiments, MoSz powder used here began to oxidize above 300" C. in dry air. I n ambients containing chlorine, the MoS2 is readily 886
ANALYTICAL CHEMISTRY
converted to RloSCl2, and eventually to MoC16 at above 400' C. (8); but slow corrosion already begins at low temperatures (?: 250' C.) and,. probably, also depends on the particle size. After application, the MoS2 lubricant adheres strongly to the substrate and cannot be removed either mechanically or by conventional aqueous solvents without introducing a substantial portion of the substrate in the analytical sample. Even on inert substrates, the MoS2 is not readily removed by the strong acids or bases. The use of hot HXO3-H2SO4 mixture which does dissolve the MoSz precludes the determination of sulfur. A literature survey indicated that no standard method is available for the determination of M o S z in lubricants present on ferrous substrates. A method, therefore, has been developed which permits a rapid and quantitative separation and determination of molybdenum, sulfur, and also halogen from several alloy, ceramic, and quartz substrates. This method is essentially a modification of the combustion method used in the organic analysis for the determination of sulfur and halogen. It consists of oxidizing the MoSz together with the organic lubricant material, (if not
present in excessive amounts) and the substrate by a slow stream of met oxygen. The volatile oxidation products are SO2 and Moo3 which are removed from the furnace separately at different temperatures. Although dry oxygen or even air may also effect the desired separation, the met oxygen is superior for the following reasons. The formation of SOa, which dissolves in the absorbing solution to give the SO4+ instead of the desired SO3+, is suppressed (10); a faster oxidation of MoS2 a t a lower temperature occurs (1.4, 1 7 ) ; the sublimation rate of the Moo3 in the presence of water vapor above 700" C. is considerably increased because of the formation of a more volatile hydroxide, M O O ~ ( O H()7~) . EXPERIMENTAL
The combustion chamber was a quartz tube, 12-mm. 0.d. and 18 inches long, inserted in a Fisher microcombustion furnace. A t the tube inlet, a borosilicate glass saturator trap wrapped with heating tape and filled with distilled water provided the desired partial vapor pressure of H20. At the outlet end, the tube was joined to an empty U-tube intended to collect traces of JfOOs and then to a fritted disk absorber trap. Apparatus.
