Table 1.
Results on Synthetic ZirconiaYttria Mixtures
ZrOn Added, Found,
c/o
Yo
5.25 10.82 26.11 26.67 40.30 40 59 70 14 75 27 79.46 86.20 95.39
5.23 10 86 26.10 26.67 40.22 40 53 70 09 75 21 79.55 86.35 95.40
Mean error,
Rel. error,
Yo
7%
0.02 0.04 0.01 0.00
0.38 0.37 0.04 0.00 0.20 0 15 0 07 0 08 0.11 0.17 0.01
0.08 0 06 0 05 0 06
0.09 0.15 0.01
taminated with an average of 0.4 mg. of either coprecipitated yttria or alkali salts from the pyrosulfate fusion. The amount of contamination was independent of sample size or percentage composition of the mixtures. Yttrium was not detected by spectrographic analysis of the second cupferron precipitate. The method was tested by analysis of 11 synthetic mixture:, of zirconia and yttria ranging in zirconia content from 5 to 95oj,. Blanks were taken through each step of the prccedure. The results, given in Table I , show a mean error of 0.05%. The relative error ranges from 0.38 to 0 00% for samples between 5 and 95%, respectively. A
sample which had been found by a n ion exchange procedure in another laboratory to contain 52.11% zirconia, gave results of 52.03 and 52.09% by the double cupferron precipitation, showing good agreement between the two met hods. Analysis of a single sample may be completed in about 2 days elapsed time, requiring approximately 4 hours of the operator’s attention. Eight samples are a convenient number to be done in any one set. Twenty-four samples may be easily completed in 5 days. The double cupferron procedure would undoubtedly be also applicable to separation of titanium from yttrium and probably to the separation of other rare earths from titanium and zirconium. ACKNOWLEDGMENT
The authors thank Elizabeth K. Hubbard for performing the spectrographic analysis, and Martha S. Richmond and John R. Baldwin for furnishing the sample analyzed by ion exchange which was used as a control, and also for supplying the pure zirconia and yttria used to make u p the synthetic samples. LITERATURE CITED
(1) Alimarin, I. P., Tse, Y. H., Zh. Analit. Khim. 14, 574 (1959).
(2) Codell, M., “Analytical Chemistry
of Titanium Metals and Compounds,” p. 230, Iilterstience, Sew York, 1959. (3) Hettel, H. J., Fassel, V. A., ANAL. CHEM.27, 1311 (1955). (4) Hillebrand, W. F., “The Analysis of Silicate and Carbonate Rocks,” U. S. Geol. Survey Bulletin 700, p. 176 (1919). (5) Hillebrand, W. F., Lundell, G. E. F., Bright, H. A., HoHman, J. I., “Applied Inorganic Analysis,” 2nd ed., pp. 119, 572, 578, Wiiey, New York, 1953. (6) Kumins, C. A., ANAL.CHEM.19, 376 (1947). (7) Majumdar, A. K., Banerjee, S., Anal. Chim. Acta 14, 306 (1956). (8) Oesper, It. E., hlingenberg, J. J., ANAL.CHEM.21. 1509 (1949). (9) POPOV, A. I.,’ Wendlandti W. w., Ibid., 26, 883 (1954). (10) Sant, S. B., Sant, B. R., Talanta 3 , 95 (1959). (11) Schoeller, W. R., Analyst 69, 260 (1944). (12) Schoeller, W. R., Powell, A. R., “Analysis of Minerals and Ores of the Rarer Elements,” 3rd ed., p. 120, Harper, New York, 1955. (13) Scott, ,W. TV., “Standard Methods of Chemical Analysis,” 5th ed., p. 1100, Van Nostrand, New York. 1939. (14) Simpson, S. G., Schumb, W. C., J . Am. Chem. SOC.53, 921 (1931). (15) Gckery, It. C., “Analytical Chemistry of the Rare Earths,” p. 48, Pergamon Press, New York, 1961. (16) Wendlandt, W. W., ASAL. CHEM. 27, 1277 (1955). (17) Wilson, C. L., Wilson, D. W., “Comprehensive Analytical Chemistry,” 1-01. IC, . p. - 499, Elsevier, Amsterdam, 1962. (18) Wood, D. F., Turner, M., Analyst 84, 725 (1959).
RECEIVED for review February 11, 1964. Accepted March 19, 1964.
The Premixed, Fuel-Rich, Oxyacetylene Flame in Flame Emission Spectrometry A. P. D’SILVA,’ R. N. KNISELEY, and V. A. FASSEL Institute for Atomic Research and Department of Chemistry, Iowa State University, Ames, Iowa
b An improved premixed, oxyacetylene burner, which can b e safely operated under fuel-rich conditions, is described. Observafion of the atomic line spectra emitted in the interconal zone of this flame has made it possible to detect, for the first time, analytically useful lines for Ce, Hf, Ta, Th, U, and Zr in a simple flame. Sensitivities of detection for the strongest lines of 26 elements, whose lines are not emitted with significant intensity in stoichiometric hydrogen or l i yd roca rbon-fuel flames, are tabulated.
T
of a fuel-rich, oxyacetylene flame (1-3) has greatly increased the scope of application of Beckmantype burners in flaml. emission spectrometry. However, the high spectral background emitted by this flame has H E USE
limited the detection sensitivities attainable and difficulties have been encountered from encrustation of the oxygen orifice. .1 recent communication ( 4 ) described a simple modification of the standard burner which minimizes these difficulties. The first model provided a graphite premixing channel for the acetylene and oxygen after atomization had occurred. h more refined version of this burner is shown in Figure 1. The stainlmssteel insert in the Teflon barrel offers distinct advantages over the graphite channel. First, there is a reduced tendency for flooding from excess atomized solution since this excess readily drains from the tip into the reservoir. Second, contamination or so-called memory effects, occasionally exhibited by the graphite tube after prolonged use, are also minimized. If
contamination occurs, the tube may be readily cleaned and reused. However, the use of stainless steel requires that greater care be taken to protect the burner tip from excessive heat. Thus, continuous atomization of sample or solvent is required to provide adequate cooling of the burner tip, and the solvent level in the reservoir must be maintained. If stainless steel is attacked by the sample solutions, inserts of other structural materials possessing similar thermal conductivity and stability may be substituted. EXPERIMENTAL
The spectrometer and associated equipment have been previously described (1, s), as has the alignment Present Address: rlnalytical Division, A.E.E.T. Bombay, India. VOL. 36, NO. 7, JUNE 1964
1287
urn/_--
2so
200 IS0
COPPER lOmm
l2mm 80
0
Figure 2. Schematic diagram of premixed, fuel-rich, oxyacetylene flame
Figure 1
.
Improved premixed oxyacetylene burner
procedure for the premixing channel (4). The fuel-rich, osyacetylene flame was obtained when the burner was operated a t flow rates of approximately 3.5 and 4.3 liters per minute for oxygen and acetylene, respectively. The exact flowrates vary from burner to burner. Under these conditions the burner will freely aspirate approximately 1 ml. per minute of pure ethanol. Aqueous solutions have been used with this burner, but with limited success. For best results, it is recommended that the solution be composed of a t least 50% flammable solvent. The preparation of most of the solutions used in this study has been described (1, 3 ) . The solutions for hafnium,
J xxx)
tantalum, thorium, uranium, and zirconium were prepared as follows: Element Ta
Dissolution procedure Metal dissolved by heating for several hours with repeated additions of 1: 1 mixture of concentrated HiTOs-HF, evaporated to near dryness, followed by dissolution in absolute ethanol. Hf and Zr Oxychloride dissolved in water, evaporated to dryness, and dissolved in absolute ethanol. UsOs dissolved in 1: 1 "08, U evaporated to near dryness with 1: 1 " 2 1 0 4 , cooled, fol-
Th
lowed by dissolution in absoute ethanol without heating. Th( SOa)r.4H20 dissolved in 1 : l HCIO,, evaporated to near dryness, cooled, followed by dissolution in absolute ethanol without heating. RESULTS A N D DISCUSSION
The fuel-rich, oxyacetylene flame obtained from the premixed burner, with ethanol solutions aspirating, is shown in Figure 2. Three distinct zones are evident: a blue primary reaction zone in which molecular Ct and CH emission is very intense; an hterconal zone; and the outer secondary (combustion) zone. For the elements studied in the present investigation, the best sensitivities
L
cl
I 4000
I
I
5Ooo
WAVELENOTH (
I
6000
H
Figure 3. Comparison of background emission from standard Beckman burner (upper curve) and premixed burner (lower curve) when Loth are operated under fuel-rich conditions
1288
ANALYTICAL CHEMISTRY
4i HAFNIUM
moo Ppm
Table I. Estimated P1,actical Detection Limits Using the Prelmixed, Fuel-Rich Oxyacetylene Flame
DY Er
Eu Gd
Hf Ho La Lu Nd Nb Pr Re Sm
Sc
Ta Tb Th Ti Tm U V
Yb Y W Zr
length, A. 5697.0 5699.2 4186.8 4211.7 4008.0 4151.1 5826.8 4594.0 4346.5/ 4346.6 4401.9 4519.7 3682,2 4103.8 4187.3 5791.3 3312.1 4518.6 4883.8 4924.5 4058.9 4939.7 4951.4 3460.5 4783.1 4883.8/ 4884.0 5175.4 3907.5 3911.8 4020.4 4812.8 3901.4 4318.8 4326.5 5760.6 3653.5 3998.6 4094.2 4105.8 5915,4 4379.2 4408.2/ 4408.5 3988.0 4077.4 4102.4 4128.3 4008.8 3519.6 3601.2
Beckman Premixed burner bur ier
' 0.5 2
0.8305 25
d
0 3 40
10 10 0.3 0.1 0.3 1 1 0.0025 4
5 2 75 0.1
5 12 15 3 5
250 PW
URANIUM
iQOppm
8
%&
2.5
3I
0.
4-
WAVELENGTH [A)
Figure 4,
1 0.2 0.5 1 1 1 2
6
TUNGSTEN
200ppm
79
wave- Detection limit, p.p.mn Element Ce
TANTALUM
Recordings of most sensitive lines of Zr, Hf, Ta, W, and U
LANTHANUM 100 ppm
THORIUM
CERIUM 100ppm
p
IOOOpprn
2.
E
1
0.6 1.0 d
0.8 0.2 0.07 0.1 20 4
10 1 I
5 0.5 d
3 0.1.
3 90 d d
150 0.5 0.5 0.3 0.2 10 0.3 0.3 0.05 0.3 ' 0.3 0.5 4 50 75
a Concentration which yields line intensities equal to two times the standard deviation of the background fluctuations. * As observed in a fuel-rich, Beckman type burner ( 1). Atomic lines of t?is element not previously detected. Atomic lines previoiisly detected but were not analytically us4:ful.
were obtained by observing only the interconal zone and rejecting entirely the radiation emitted from the inner blue cone. Hence, only an 8-mm. vertical section of the intermediate zone
"5
WAVELENGTH(d)
Figure 5.
Recordings of most sensitive lines of La, Ce, and Th
was viewed by the entrance optics of the spectrometer. Figure 3 shows the difference in background emission between the standard Beckman and premixed burners when both are operated under fuel-rich conditions. The striking reduction in the background level for the premixed burner is reflected by the marked improvement in the sensitivities of detection as shown by the data summarized in Table I. In many cases the detection limits for previously reported lines are extended by an order of magnitude or more. The lower background level in the premixed flame also allows the use of other sensitive lines which are obscured by the background emission in the standard Beckman burner. Table I shows that analytically useful lines of Ce, Hf, Ta, Th, U, and Zr have been derived for the first time from a simple flame. Recordings of the most sensitive spectral lines of these elements are shown in Figures 4 and 5. The ability to excite analytically use-
ful spectral lines for almost all of the metallic elements in a simple flame should greatly extend the scope of application of the flame spectrometric technique. Moreover, since the enhancement in emission intensities has been attributed to an increase in the free atom population in the flame ( 2 , S), it should now be possible to detect these elements by atomic absorption procedures as well. LITERATURE CITED
(1 ) Fassel, V. A., Curry, R. H., Kniseley, R. S . , Speclrochim. Acta 18, 1127 (1962). (2) Fassel, 1'. A., Kniseley, R. N., Il'Silva, A . P., Curry, R. H., Myers, R. B., ANAL.C H E M . 36. 532 (1964). ( 3 ) Fassel, V. A. Geyers, R. B. Kniseley, R. N., Zcid., 19, 1187 (1063). (4) Kniseley, It. S . , D'Silva, .A. P., Fassel, I-.A.. Zbid., 35, 910 (1063).
RECEIVEDfor review Marrh 4, 1964. Accepted April 14, 1064. Work performed in the Ames Laboratory of the U. S.Atomic Energy Commission. VOL. 36, N O . 7, JUNE 1964
1289