Titrimetric Determination of Ionic and Coordinated Azide Ion by Oxidation with Excess Nitrite Ion in Acid Solution Wilton R. Biggs and Richard W. Gaver' Department of Chemistry, California State University, San Jose, San Jose, Gal$ 95114 DURINGA RECENT INVESTIGATION involving certain azidocobalt(II1) complexes, it became necessary to have a method that would permit the total assay of coordinated as well as ionic azide ion at the 1-mmole level. Azideionin simple salts is quite easily determined by oxidation with excess cerate ion and back titration with ferrous sulfate ( I ) . This and other useful methods have been surveyed by Clem and Huffman (2). However, except where oxidation by nitrite ion in acid solution was employed, these methods were unsatisfactory for accurately determining small amounts of coordinated azide ion. The suitability of the nitrite reaction for the determination of coordinated azide ion was suggested by the success of Haim and others in applying this reaction to the quantitative removal of coordinated azide ion from cobalt(II1) complexes to be used in kinetic studies (3-5). Several analytical applications of this reaction have been reported. Azide ion in acid solution may be titrated directly with nitrite ion (at 0 "C for best results) using dicyanobis(1,lOphenanthroline)iron(II) as a redox indicator, provided oxygen is excluded from the reaction mixture (6). Millimolar amounts of azide have been estimated by measuring the hydrogen ion consumed in the oxidation of azide ion by nitrite ion at pH 3 (2). The first method is quite inconvenient for small amounts of coordinated azide and may be subject to side reactions between the indicator and the metal complex containing the azide ion. The second method is the more accurate but the required pH titration is subject to error when weak acid aquo metal complexes or hydrolyzable metal ions are present. Finally, both methods require a second method and reagents to differentiate between ionic and coordinated azide ion. In the present work, a method is described that permits the total azide assay of samples containing both ionic and coordinated azide ion. Oxidation of all azide species in acid solution is accomplished by the addition of a measured excess of nitrite ion. The excess nitrite ion is then determined by a standard cerate method (7). The method is fast, accurate, and subject to few interferences. Coordinated azide then may be estimated by subtracting from the total azide the amount of ionic azide determined with a separate sample by cerate oxidation of the azide ion to molecular nitrogen ( I ) . No new reagents would be required.
To whom correspondence should be addressed. (1) J. W. Arnold, IND.ENG.CHEM.,ANAL.ED., 17,215 (1945). (2) R. G. Clem and E. H. Huffman, ANAL.CHEM.,37, 366 (1965). (3) A. Haim and H. Taube, Znorg. Chern., 2,1199 (1963). (4) D. Loeliger and H. Taube, ibid.,4, 1032 (1965). ( 5 ) R. Barca, J. Ellis, M. Tsao, and W. K. Wilmarth, ibid., 6 , 243 (1966). (6) A. A. Schilt and J. W. Sutherland, ANAL.CHEM.,36, 1805 (1964). (7) G. F. Smith, "Cerate Oxidimetry," G . Frederick Smith Chemical Co., Columbus, Ohio, 1942, p 32. 1870
Table I. Comparison of Standard Cerate and Indirect Nitrite Titration of Sodium Azide Solutions Millimoles of azide No. of Av Mean Method samples Added found error Std dev 0.005 0.011 Cerate 10 1.302 1.297 Cerate 0,593 0.007 0.011 10 0.600 Nitrite 10 1.302 1.295 0.007 0.010 Nitrite 10 0.600 0.589 0.011 0.010
EXPERIMENTAL
The synthesis of cis- and tr~ns-[Co(NH~)~(N~)~]N3 has been described (8). The trans isomer is most easily isolated as the double salt [ C O ( N H ~ ) ~ ( N & N ~ NaN3. The cis isomer is isolated as the single salt. Both products were purified by recrystallization from NaN3 solution. Analysis for N and H was performed by the University of California Chemical Analytical Service, Berkeley. Analyses: Calcd for tr~ns-[Co(NH~)~(N3)2]N~ .NaN, 1/4 H20: N, 69.46; H, 3.91%. Found: N, 69.78; H, 3.97x. H, 4.78 Calcd for c ~ s - [ C O ( N H ~ ) ~ ( N ~ )N, ~ ] N71.94; ~: Found: N, 71.62; H, 4.66%. Solutions. A 0.0500M cerate solution was prepared from reagent grade ceric sulfate tetrahydrate and 0.5M sulfuric acid. Using a slight modification of Smith's procedure, it was standardized against primary standard sodium oxalate (Baker and Adamson) in 0.5M perchloric acid with Ferroin as the indicator (9). A 0.0500M ferrous iron solution was prepared from ferrous sulfate pentahydrate and 0.5M sulfuric acid. Its titer was established each day with the standard cerate solution, using Ferroin as the indicator. A 0.05OOM sodium nitrite solution was prepared from reagent grade material and standardized by cerate oxidation to nitrate using the procedure subsequently employed in the azide analysis scheme (7). Solutions of the azide samples were freshly prepared from the purified complex salts or from reagent grade sodium azide, as appropriate. Procedure. A solution of the sample, containing between 0.5 and 3.5 mmoles of azide, is mixed with a known excess ( 5 5 ml) of standard nitrite solution. The pH is lowered to 5 2 by the dropwise addition, with stirring, of 6 M HC10,. After a reaction time of at least 30 sec, the sides of the flask are washed down and the solution is diluted to about 100 ml. A large (>lo ml) measured excess of standard cerate solution is introduced. The sample is momentarily heated to 50 "C to complete the oxidation of excess nitrite ion. The flask is cooled and the excess cerate is titrated with a standard ferrous solution, using Ferroin as the indicator. Calculations. The amount of azide in the sample is equal to the mmoles of nitrite ion it consumed according to the Equation (3):
-
-
x.
HN3
+
"02
=
N2
+ N10 + HzO
(1)
(8) G. Linhard and H. Flygare, 2. Anorg. Allgem. Chem., 262, 328 (1950). (9) G. F. Smith, ANAL.CHEM.,27, 1144 (1955).
ANALYTICAL CHEMISTRY, VOL. 44, NO. 11, SEPTEMBER 1972
Complex cis-[Co(NH,),(N,)z]Na rruns-[Co(NH,),(N,)z]N3 NaN3.'1, Ht0
Table 11. Total Azide Analysis by the Indirect Nitrite Method No. of Millimoles of azide samples Added Av found Mean error 10 1.434 1.427 0.007 10 40
1.209 1.302
In terms of the reagents used, the total azide is given by: mmoles N3-
=
[ml N a N 0 2 X 0.0500M
(ml Ce(S04)2 X 0.0500M
-
- ml FeS04 X O.O5OOM)] (2)
RESULTS AND DISCUSSION
Samples of sodium azide were analyzed by both the nitrite method and a standard cerate titration ( I ) . At the 1.3-mmole level, the difference between the two methods was less than 2 ppt (Table I). The total azide content of cis- and transdiazidotetraamminecobalt(II1) azide was determined by the indirect nitrite method with excellent results (Table 11). The reagent concentrations employed in this procedure give optimum results for samples containing from 0.5 to 3.5 mmoles of azide. By increasing the concentration of the nitrite solution to OSM, the upper limit easily may be extended to about 25 mmoles of azide. A possibly serious side reaction in the nitrite method is the iron(I1) reduction of the residual complex during the final back titration. In some cases, a preliminary separation may be necessary. However, in the case of cobalt(II1) complexes such as those used in this experiment, the initial product is the diaquo complex, [ C O ( N H ~ ) ~ ( H ~ O )whose ~ ] ~ + , subsequent reduction by aqueous ferrous ion is negligibly slow under the conditions used. Although other interferences were not extensively investigated, anions such as chloride, nitrate, and sulfate did not interfere in amounts up to 250 ppm. High concentrations of bridging ligands such as chloride ion should be avoided since they would facilitate the undesirable reduction of the residual cobalt(II1) complexes by iron(I1) in the final step. The reaction of thiocyanate ion and nitrous acid to form unstable nitrosylthiocyanate should not interfere under the conditions used ( 2 ) . High results can be expected if uncoordinated ammonium ion is present; it reacts slowly with nitrite ion to give elemental nitrogen and water. Perchloric acid was used to adjust the pH for the initial reaction not only because it is a strong acid but also because its effect in increasing the oxidation potential of ceric ion is expected to help in the subsequent oxidation of excess nitrite
1.194 1.294
Std dev 0.009
0.015 0.006
0.013 0.006
ion. Acidification of the solution should be done carefully to prevent localized areas of high acid concentration since nitrous acid is unstable in strongly acidic solution. The possible loss of HNOzand/or H N I during the acidification step was of some concern but was not noted when using the recommended procedure. The possible loss of HNO, during the subsequent reaction with ceric ion was considered. However, no appreciable increase in precision could be gained by slowing the reaction through cooling the sample solutions, as long as large excesses of nitrous acid were avoided. Although this procedure was developed for azide ion in cobalt(I11) complexes, it should be quite generally applicable to ionic and coordinated azide ion in other compounds. For example, we recently used it to determine the azide content of azidopentaamminechromium(II1) bromide. Direct Titration of Ionic Azide Ion. The amount of uncoordinated azide in ~ ~ ~ ~ s - [ C O ( N H ~ ) ~NaN, ( N , ) ~was ]N~. determined in initially neutral solution by direct potentiometric titration with the standard cerate solution using a platinum indicator electrode GS. SCE. For a sample containing an estimated 0.571 mmole of ionic azide, the amount found was 0.580 mmole. Although the result is within the standard deviation, a slightly high value might be expected if partial hydrolysis of one of the coordinated azide groups occurs near the end point because of the accumulation of acid from the cerate solution (IO). The direct titration method gave low results for c ~ s - [ C O ( N H ~ ) ~ ( N ~ )The ~ ] Nsupposedly ~. uncoordinated azide did not react with cerate under the conditions used for the trans isomer. The unusually strong association of this azide group has been verified by conductivity measurements and is currently under investigation. ACKNOWLEDGMENT Many helpful suggestions from John Neptune and Donald West are gratefully acknowledged. RECEIVED for review June 30, 1971. Resubmitted March 27, 1972. Accepted May 1, 1972. (10) A. Haim, J. Amer. Chem. Soc., 85, 1016 (1963), note 7.
ANALYTICAL CHEMISTRY, VOL. 44, NO. 11, SEPTEMBER 1972
0
1871