Use of ammonium fluoride in determination of zirconium and other

Jul 1, 1970 - View: PDF | PDF w/ Links. Citing Articles; Related ... Constance C. Butler and Richard N. Kniseley. Analytical ... Klaus Dittrich , Alan...
0 downloads 0 Views 449KB Size
avoided if an appropriate volume of iron(II1) solution is used as an indicator. A practically constant titer was obtained with less than 0.2 ml of 5 % iron(II1) chloride (or 10% iron alum) solution per 10 ml of solution. In the titration of 20-ml portions of 0.2N pyridine with 0.2N HCl in the presence of 0.2 ml of 5 % FeCl3.6Hz0 and from 0.1 to 1.0 gram of KzS04,all results were accurate to within 0.5%. Similar results were obtained when 0.2 ml of 10% iron alum replaced both FeC13 and KzS04 was added to 20-ml portions of 0.2Npyridine and titrated with 0.2NHC1. Interferences. In both the method using sulfuric acid titrant and the method using hydrochloric acid titrant, neutral salts which react with Fe3f to form precipitates like Pod3-

disturb the determination. Small amounts of NaCl, NaN03, MgS04, K2Cr207,Hz02,and Lipon-F (a commercial neutral detergent) do not interfere. The method has been successfully applied to a- and ypicoline and also to pyridine (Py) isolated from RuOl 2Py (7) and ZnClz.2Py. e

RECEIVED for review October 30, 1969. Accepted March 19, 1970. (7) Y . Koda, Znorg. Chern., 2, 1306 (1963).

Use of Ammonium Fluoride in Determination of Zirconium and Other Elements by Atomic Absorption Spectrometry in the Nitrous Oxide-Acetylene Flame A. M. Bond Department of Inorganic Chemistry, University of Melbourne, Parkville, Victoria, 3052, Australia DURINGTHEIR DEVELOPMENT of the high temperature nitrous oxide-acetylene flame as a means for improving the sensitivity of the atomic absorption analytical method for certain metals forming refractory oxides, Amos and Willis ( I ) observed that fluoride enhanced the absorbance of zirconium in this flame. This observation was later confirmed by Bond and O’Donnell (2) and by Sastri and coworkers ( 3 , 4 ) . Fluoride has also been used to improve the sensitivity of the spectrographic determination of zirconium in the carbon arc (5). Enhancement effects in atomic absorption spectrometry can be either physical or chemical in origin. However, for fluoride interaction with the zirconium system, chemical intrepretations have been favored (2-4). Bond and O’Donnell ( 2 ) observed that the enhancement of zirconium absorbance by the ammonium ion had very similar characteristics to that caused by the fluoride ion. More recently, Bond and Willis (6) reported that a considerable number of other nitrogen-containing compounds also caused an enhancement of zirconium absorbance in the nitrous oxideacetylene flame and that the effect appeared to involve the formation of a zirconium-nitrogen bond. The effect of the ammonium ion was, furthermore, found to be additive to the enhancement provided by fluoride (2,6). Even in the nitrous oxide-acetylene flame, the sensitivity and limit of detection of zirconium are not particularly good. It was therefore felt that use of a suitable concentration of ammonium fluoride would improve the atomic absorption (1) M. D. Amos and J. B, Willis, Spectrochim. Acta, 22, 1325, 2128 (1966). (2) A . M. Bond and T. A. O’Donnell, ANAL. CHEM.,40, 560 (1968). (3) V. S. Sastri, C. L. Chakrabarti, and D. E. Willis, Can. J. Chem., 47, 587 (1969). (4) V. S . Sastri, C. L. Chakrabarti, and D. E. Willis, Talanta, 16, 1093 (1969). ( 5 ) A. A . Frishberg, Russ. J. Appl. Spectrosc., 5 , 8 (1966). (6) A. M. Bond and J. B. Willis, ANAL.CHEM., 40,2087 (1968).

932

0

ANALYTICAL CHEMISTRY, VOL. 42, NO. 8, JULY 1970

method for zirconium determination by utilizing the enhancement effects of both the ammonium and fluoride ions. A fairly high concentration of ammonium fluoride should be tolerable, without any matrix interference due to formation of solid material in the flame, because ammonium fluoride should readily decompose to give the highly volatile products, NH3 and HF. Interelemental and chemical interferences in the nitrous oxide-acetylene flame are quite widespread (1, 7-9). In an unpredictable fashion, certain elements or compounds can cause either suppression or enhancement of the absorbance of the element being measured. As described later in this paper, many of the interferences in the determination of zirconium are completely eliminated or greatly suppressed by use of ammonium fluoride. Thus the possible use of ammonium fluoride to improve the atomic absorption method for determination of zirconium in the nitrous oxide-acetylene flame has been investigated in detail. Its use in improving the determination of some other elements forming refractory oxides has also been examined. Results, optimum concentrations, and further considerations of the mechanism of the ammonium fluoride enhancement are also reported. EXPERIMENTAL

Reagents. All chemicals used were of reagent grade purity. Zirconium solutions were prepared from zirconium oxychloride octahydrate (ZrOClz.8Hz0). All solutions contained 0.006M potassium chloride to suppress ionization, which can otherwise be appreciable for many elements in the nitrous oxide-acetylene flame (8). Sufficient hydrochloric acid was added to each solution to maintain all zirconium species in solution. (7) J. B. Willis, Appl. Opt., 7 , 1295 (1968). (8) D. C. Manning, At. Absorption Newslett., 5 , 127 (1966). (9) W. Slavin, A. Venghiattis, and D. C . Manning, ibid., p 84.

C

I

0

0.1

0.04

[ADDED

0.0 8

SPECIES

0.12

J pj

Figure 2. Absorbance of zirconium as a function of concentration of added species. [Zr] = O.01M

0

0.01

Czrl M

0.0 2

0.03

Figure 1. Influence of various species on absorbance of zirconium a. (X) Zirconium b. (0) Zirconium with 0.1M NHICl c. (0) Zirconium with 0.1M NaF d . (u) Zirconium with 0.1M NHIF Calcium, magnesium, titanium, palladium, tin, manganese, cobalt, copper, and nickel solutions were prepared from their respective chlorides. Platinum and rhodium solutions were prepared from the sodium salts of their chlorides; hafnium and uranium from their nitrates; molybdenum from molybdic acid or molybdenum trioxide; tungsten from sodium tungstate ; vanadium from vanadium pentoxide ; and tantalum from tantalum potassium fluoride. Apparatus. Atomic absorption measurements were made with a n AA-100 atomic absorption spectrophotometer, (Varian Techtron Pty. Ltd., North Springvale, Victoria 3170, Australia), fitted with a 50- X 0.5-mm slot burner for use with nitrous oxide-acetylene mixtures. Absorption was usually measured in a fuel-rich, slightly luminous flame having an interconal zone about 30 mm high. Hollow cathode lamps made by Atomic Spectral Lamps Pty. Ltd., Australia, were used as light sources for all elements. RESULTS AND DISCUSSION

Enhancement of Zirconium Absorbance of Ammonium Fluoride. All atomic absorption measurements for zirconium were made a t the 3601.2 A line. Figure 1 shows absorbance/concentration curves for various zirconium solutions. This figure shows the considerable sensitivity improvement obtained by the use of 0.1M ammonium fluoride. The absorbance for the same concentration of zirconium is increased eight to ten times by the presence of 0.1M NH4F. The presence of 0.1M N a F or 0.1M NH4C1 also produces a n increase in zirconium absorbance but not as marked as with 0.1M NH4F. The ammonium and fluoride enhancements are approximately equivalent and additive. Figure 2 shows that n o appreciable improvement in absorbance of a 0.01M zirconium solution is gained by using ammonium fluoride concentrations in excess of 0.1M . Figure 2 also shows the effect on the absorbance of a 0.01M zirconium solution of ammonium chloride and sodium fluoride. The basic shape of all graphs is the same. At concentrations less than that of zirconium, the plots are approximately linear. At concentrations roughly equal to that of zirconium, curvature is observed. When approximately four-

a.

(W)

NHiCl

b. (0) NaF C. (X) NHiF

to eightfold excess is reached, the absorbance becomes almost constant. Influence of Ammonium Fluoride on Absorbance of Other Elements. Table I shows the effect of ammonium fluoride on the absorbance of a number of elements along with the stability constants of the fluoride formed in solution, the melting points of the likely oxides formed in the flame, and the melting points of some fluoride compounds. Very few data are available on ammonia complexes or compounds. Several relationships can be derived from examination of this Table. All elements which are enhanced by fluorideuiz., zirconium, hafnium, tantalum, titanium, and uraniumexist in noncomplexing media as oxy or hydroxy complexes, which are often polymeric and aggregated in nature (10). These species would dry out in the flame as refractory oxides with high melting points and thus atomization would be inefficient. These same elements all form very strong fluoride complexes in solution, and in the presence of fluoride they could therefore dry out in the flame as volatile fluorides which would lead to greater efficiency of atomization than in the absence of fluoride, and give rise to the observed enhancement of absorbance. This mechanism is essentially the one put forward in detail for zirconium previously (2), but it would appear to hold for the other elements as well. Some support for this mechanism can be found in Frishberg’s work (5). In the spectrographic determination of zirconium in a carbon arc, zirconium is normally converted to nonvolatile zirconium oxide (Zr02), and consequently the sensitivity of the method is poor. Addition of ammonium fluoride increases the sensitivity of’ the spectrographic analyses in a similar fashion to the atomic absorption method. Frishberg showed that the improved sensitivity in the carbon arc resulted from the formation of volatile zirconium tetrafluoride. Comparison of spectrographic and atomic absorption conditions should be valid, so this lends some support to the mechanism put forward previously to explain enhancement effects of fluoride in the atomic absorption method for zirconium and now for some other elements. Examination of Table I shows that four conditions appear to be necessary to observe enhancement of a particular element by fluoride. (10) F. A. Cotton and G. Wilkinson, “Advanced Inorganic Chernistry,” 2nd ed., Interscience, New York, 1966. ANALYTICAL CHEMISTRY, VOL. 42, NO. 8, JULY 1970

0

933

Table I.

Correlation of Physical Properties of the Oxides and Fluorides, and the Influence of NH4F on Absorbance, for a Number of Elements Influence Oxidation state of of NHdF Strength of M.P. of Element solutions examined Line A on absorbance fluoride complexes” M.P. of oxide,b,c“C fluoride,csd “C Zr (IV) 3601.2 Enhances Strong 2700 (ZrO2) 912 (ZrF4) 3072.9 Enhances Strong 2812 (HfOz) 956 (HfFa) Hf (IV) Ti (111) and (IV) 3642.7 Enhances Strong 1825 (TiOt) 400 (TiFisitbl) 2130 (Tho,) 1200 (TiF3) Ta 2714.0 Enhances Strong 1800 (TalO:) 97 (TaFs) Enhances6 3514.6 U Strong 2500 (UOz) :9: 17 (MoFe.) 795 (Moos) None 3132.6 Mo Weakf W 2551.4 None Weakf 1473 (wos) 2 (WFd 2246.1 None Strong 1127 @no2) Sn 705 (SnFmbl) 4226.7 Weak Ca 2580 (CaO) 1360 (CaFz) None 2852.1 None Weak 1266 (MgF2) 2800 (MgO) Mg 2794.8 Weak Mn 1705 (Mn304) 920 (MnFz) None 2483.3 Weak [(II) state] 1020 (FeF2) None Fe 1420 ( FeO) 1565 (FeZO3) Strong [(111) state] 1000 (FeF3dec) 2474.5 870 (PdO) Weakf Pd None ... 2659.5 ... Weakf 450 (PtOt) Pt None 1125 (RhzO,) 3434.9 Rh Weakf 600 (RhFssubl) None 1970 (V201) 3184.0 Enhanceso Weakf V 19 (VFs) cu 3247.5 None Weak 785 (CUF~) 1200 (COF~) 2407.3 None Weak 1990 (NiO) 1450 (NiFz) 2320.0 None Weak a Compiled from Reference I I . * Most stable oxide tabulated. c Values from Reference 14. d Values from References 12 and 13. e Low sensitivity and high noise level made magnitude of enhancement difficult to assess. f No absolute values available but evidence is that weak or no complex formation exists. 0 Enhancement due to ammonia only.

.)

K;1

Table 11. Some Interferences Encountered in Determination of Zirconium by Atomic Absorption. Effect on absorbance Effect on Effect on Effect on absorbance of 10-2M Zr/O.lM NaF absorbance of 10-2M absorbance of 10-2M Interfering species of 10-2M Zr solution solution Zr/O.l M ”8.3 solution Zr/O.lM NH4F solution Ammonia as NHaCl or Marked enhancement Marked enhancement Enhancementb None up to 0.05M NH4SCN Enhancementc Marked enhancement None up to 0.05M Marked enhancement Fluoride as NaF Enhancement Enhancement None up to 0.05M Depression Nickel as NiBr2 Enhancement None up to 0.05M Enhancement Cobalt as C O ( N H ~ ) ~ C ~Enhancement ~ Depression Depression None up to 0.01M Depression Sulfate Depression Depression None up to 0.01M Depression Nitrate 5 “Enhancement” indicates absorbance increases by less than 0.05, “Marked Enhancement” indicates absorbance increases by more than this. b Additional 0.05.44 NH4C1to the 0.1 M already present causes very slight enhancement. c Additional 0.05M NaF to the 0.1M already present causes very slight enhancement.

(i) The metal should exist as any oxy-type complex in aqueous solution. (ii) The metal should be capable of forming a very strong fluoride complex in aqueous fluoride solutions. (iii) The most stable oxide in the solid state should be refractory. (iv) The metal should form a volatile fluoride complex in (1 1) Chemical Society (London), “Stability Constants of Metal Ion Complexes,” Special Publication No. 17, 1964. (12) R. Colton and J. H. Canterford, “Halides of the First Row

Transition Metals,” Wiley, Interscience, London/New York/ Sydney, 1969. (13) J. H. Canterford and R. Colton, “Halides of the Second and Third Row Transition Metals,” Wiley, Interscience, London/ New York/Sydney, 1968. (14) “Handbook of Chemistry and Physics,” 48th ed., Chemical Rubber Co., Cleveland, Ohio, 1967-1968, pp B-149 to B-242. 934

*

ANALYTICAL CHEMISTRY, VOL. 42, NO. 8, JULY 1970

the solid state with a melting-point considerably below that of the refractory oxide. Unless all four conditions apply, no enhancement is observed. These observations are consistent with the postulated mechanism for fluoride enhancement. Very little positive discussion can be made about the enhancement due to ammonia because of lack of thermodynamic data. However, in cases where ammonia does cause enhancement, the available evidence suggests a chemical mechanism similar to that operating with fluoride. The previous work (6) on zirconium- and nitrogen-containing compounds was certainly consistent with formation of zirconiumnitrogen bonds. However, whether compound formation occurs in solution or in the drying out process in the flame is debatable (6). Suppression of Interferences of Zirconium Absorbance by Use of Ammonium Fluoride. A number of elements inter-

fere with the absorbance of zirconium. Table I1 summarizes the interferences caused by various species on the absorbance of 0.01M zirconium in the presence of (a) no added material, (b) O.lMNaF, (c) 0.1MNH4C1,(d) 0.1MNH4F. The zirconium solutions containing 0.1M NH4F are considerably less susceptible to interference than are the others. Presumably the enhancements caused by ammonia and fluoride are such that extra enhancement by other species does not occur unless very high concentrations of the potentially interfering species are present. Furthermore the very high stability in solution of the zirconium-fluoride complexes presents interference from both the weakly complexing nitrate and strongly complexing sulfate ions (11) which otherwise form complexes which decompose in the flame to give ZrOz. Analytical Use of Ammonium Fluoride. For determination of zirconium by the atomic absorption method, the presence

of 0.1M NH4F in all standard and unknown solutions is recommended. Such a procedure gives an improvement in sensitivity of about eight- to tenfold and a corresponding improvement in the limit of detection. Furthermore, many interferences are either removed or greatly suppressed in this manner. For the same reasons the addition of ammonium fluoride is also useful in the determination of hafnium, titanium, and tantalum. ACKNOWLEDGMENT

The author thanks J. B. Willis for his interest and assistance in the planning of this work and for many helpful discussions. RECEIVED for review February 11, 1970. 1970.

Accepted April 15,

Preparation of Thermally Stable Gas-Chromatographic Packing Materials A. H. Al-Taiar, J. R. Lindsay Smith, and D. J. Waddington Department of Chemistry, University of York, Heslington, York, YO1 5DD, England

GAS-LIQUIDCHROMATOGRAPHY, when used in conjunction with mass spectrometry, is generally only practicable up to about 250 "C. Above this temperature, the vapor pressure of the liquid phase is too high ( I ) . Although a trap may be inserted to absorb the vapor of the liquid phase (2), the lifetime of the column is reduced. Gas-solid chromatography does not have these drawbacks, but on solids such as silica gel and alumina, polar solutes are irreversibly adsorbed on the surface active groups. The attempts to modify silica gels by chemical deactivation with, for example, trimethylchlorosilane (3), with dimethyldichlorosilane (4), with a cyanoalkylchlorosilane (3, and with alcohols (6, 7), have not been particularly successful for the monolayers of organic material are not dense enough to prevent specific interactions between the solutes and bulk solids (3, 5). Stationary phases which are chemically-bonded to the solid support will be more thermally stable than conventional liquid materials. Moreover, in theory, they will give more reproducible results than the latter, for GLC separation of solutes occurs in solvent pools in the capillaries of the solid support, and separation is dependent, in part, on its structure (1) R. Teranishi, R. G. Buttery, W. H. McFadden, R. T. Mon, and J. Wasserman, ANAL.CHEM.,36, 1509 (1964). (2) R. L. Levy, H. Gesser, T. S. Herman, and F. W. Hougen, ibid., 41, 1480 (1969). (3) A. V. Kiselev, Yu.S.Nikitin, V. K. Chuikina, and K. D. Shcherbakova, Russ. J . Phys. Chem., 40,71 (1966). (4) N. V. Akshinskaya, A. V. Kiselev, Yu. S.Nikitin, R. S.Petrova, V. K. Chuikina, and K. D. Shcherbakova, ibid., 36, 597 (1962). (5) A. V. Kiselev, Discuss. Faraday SOC.,40, 205 (1965). (6) C. Rossi, S . Munari, L. Cengarle, and G. F. Tealdo, Chim. Ind. (Milan),42, 724 (1960). (7) I. V. Borisenko, N. I. Bryzgalova, T. B. Gavrilova, and A. V. Kiselev, Neftekhimiya, 5, 129 (1966).

which varies from batch to batch (8, 9). Celite has been modified by treatment with hexadecyltrichlorosilane (10) and with octadecyltrichlorosilane (11) followed by hydrolysis, yielding materials which are chromatographically equivalent to Celite coated with a silicone elastomer. The surface hydroxyl groups on porous glass particles have been esterified with 3-hydroxypropionitrile to give a material which is highly efficient (12) and significantly more thermally stable than chromatographic materials using this nitrile as a liquid phase. In this paper, we report some of our exploratory studies on the preparation of materials suitable for high temperature gas chromatography. EXPERIMENTAL

Operating Conditions of the Gas Chromatograph. All results were obtained isothermally on a Pye 104 gas chromatograph equipped with a flame ionization detector coupled to an RE 511 (Goertz Servoscribe) recorder. Glass columns (1.5 m X 4 mm i.d.) and stainless steel columns (5.5 m X 1.2 mm i.d.) were used. In some experiments, the chromatograph was connected to an A.E.I. MS 12 mass spectrometer, at 70 eV electron energy, 100 pA trap current, 8 kV accelerating potential, with a source temperature of 165 "C. The carrier gas, nitrogen (B.O.C. "oxygen-free"), was deoxygenated and dried by passing it through successively a solution of chromium(I1) chloride in hydrochloric acid over zinc amalgam, concentrated sulfuric acid, potassium hydroxide pellets, silica gel, and molecular sieve 5A. (8) N. C. Sana and J. C. Giddings. ANAL.CHEM., 37,822 (1965). (9) M. Krejfi, Collect. Czech. Chem. Commun., 32, 1152 (1967). (10) E. W. Abel, F. H. Pollard, P. C. Uden, and G. Nickless, J . Chromatogr., 22, 23 (1966). (11) W. A. Aue and C. R. Hastings, ibid, 42, 319 (1969). (12) I. Haldsz and I. Sebastian, Angew. Chem., 81, 464 (1969). ANALYTICAL CHEMISTRY, VOL. 42, NO. 8, JULY 1970

935