direct determination of nanogram amounts of iodine by atomic-absorption spectroscopy using a...

Tuhra. Vol ?I. pp 573-579 Pergamon Press. 1974 Prmted m Great Bnram

DIRECT DETERMINATION OF NANOGRAM AMOUNTS OF

IODINE BY ATOMIC-ABSORPTION SPECTROSCOPY USING A GRAPHITE-TUBE ATOMIZER

M. J. ADAMS, G. F. KIRKBRIGHT and T. S. WEST

Chemrstry Department, Imperial College. London SW7. England

(Receiued 7 January 1974. Accepted 17 January 1974)

Summary-The direct determmatron of iodine by AAS at its 183.0 and 178.2 nm resonance lines by usmg a small graphite-tube atomizer, electrodeless discharge-lamp source and vacuum monochromator is described. Opttmum conditions for the determination of iodine have been established; similar sensitivity 1s obtamed when iodide or iodate samples are examined. With 10 ~1 aqueous samples sensitivities (for 17; absorption) of 4 x lo-” g and 2 x lo-” g of I were obtained at 183.0 and 17X.2 nm respecttvely; a detection limit of 2 x 10e9 g was observed at both lmes. Non-specific molecular absorptron from common inorganic salts causes interference with the determmation; the rodme non-resonance line at 184.4 nm may be employed to correct for this interference when moderate amounts of common salts are present.

Recent work in our laboratory has been concerned with the development of direct methods for the determination of non-metallic elements by the techniques of atomic absorption and emission spectrometry. As a number of these elements, notably sulphur, phosphorus and iodine, exhibit their useful ground-state atomic lines at wavelengths shorter than 200 nm it is necessary that the optical path through the atom cell and the instrumental assembly should be transparent at these wavelengths. We have described the direct determination of sulphur,’ phosphorus2 and iodine3 by atomic-absorption spectrometry utilizing the fuel-rich nitrogen-separated nitrous oxide-acetylene flame and by atomic-emission spectrometry with a high-frequency induction-coupled plasma source.4*5 In both techniques sample solutions are nebulized into the flame or plasma cell, a relatively large volume being required, and the sensitivity is limited by the difficulty of achieving high sample concentrations in the cell owing to dilution by the support gases. Both limitations of flame cells can be overcome by use of non-flame atomizers such as the graphite tube or graphite filament. The nature of flames and plasmas also imposes limitations at short wavelengths where atmospheric absorption is appreciable, as it is difficult to avoid degradation of the observed signal: noise ratio caused by fluctuating absorption at the in- terface between the flame or plasma and the atmosphere. For these reasons, the direct atomic-absorption spectrometry of elements such as iodine, sulphur and phosphorus by using non-flame atomization should give high sensitivity. L’vov and Khartsyzov6 have reported sensitivities for the determination of sulphur, phosphorus and iodine by atomic-absorption spectrometry (AAS) using pulse vaporization in a graphite cuvette. This paper describes the direct determination of iodine by AAS at its ground-state atomic lines at 178.3 (6?P3 ?-5p5 ‘P, 2) and 183.0 nm (6s4P 5 2-5p5 ‘P3 z), by using a vacuum monochromator,

573 T-\L \,?I 21 \<I h z

574 M. J. ADAM$ G. F. K~KBKIGHT and T. S. WEST

a nitrogen-purged optical system, an iodine microwave-excited electrodeless discharge- lamp source and a small heated graphite-tube atomization cell.

Fig. 1. Schematic diagram of ~nst~mentat~on employed.

EXPERIMENTAL

The instrumental a~ang~~nt is shown ~hemati~ll~ in Fig. 1. An iodine electrodeless djs~harge-lamp [EDL) was used as the source. This was made from silica tubing (i.d. 8 mm, I mm wall thickness) to form a bulb 200 mm in length, and containing a few mg of iodine. The EDL was supported in a 3/4-wave resonant cavity (Model 2 IOL, Electromedical Supplies Ltd., Wantage, U.K.) by means of a metal holder which allowed reproducible position- ing of the lamp. The source was powered by a 2450 MIIz Microtron Mk. III microwave generator (EMS Ltd., Wantage), using a reIIected-power meter (EDT Ltd., London S.W.6) connected in series between the generator and cavity to allow the lamp/cavity assembly to be tuned to minimum reflected power. Radiation from the source w%+s focused into the graphite-tube atomixer by a biconvex calcium Ruoride lens (25 mm diameter, 50 mm focal Length). The radiation emerging from the atomizer was then brought to a Focus on the entrance slit of a f -metre vacuum grating monochromatot (Rank Precision Instruments- Margate, U.K.) by a similar calcium fluoride lens. The optical path was formed from glass tubing (25 mm id.) placed between the source and graphite atomizer and between atomizer and monochromator entrance slit; this path was maintained under a slightly positive pressure of oxygen-free nitrogen. The I-metre vacuum monochromator had a reciprocal linear dispersion of 1.6 nm/mm andwas operated in conjunction witha rotary roughin~b~cking pump and an oil diffusion pump. While pressures ofless than ltY3 mm Hg are possible with this arrangement the operating pressure employed m this work was cct. @ 1 mm Hg; this is suf%cient to provide good transmission of radiation at the wavelen~bs employed. An EMI 62563 end-window photomuitiplier tube with a 50 mm diameter silica window was attached to the monochromator at the exit siit by a rubber Q-ring seal so that a vacuum-tight seal was obtained at the slit. The photomultiplier tube was operated with a Brandenburg model 47SR EHT unit at the voltage which gave the best signal: noise ratio with the light levels available (1620V). The photomultiplier output was led to a low-noise amplifier (Brookdeal type 450, Brookdeal Electronics Ltd., Berks.. England) and phase-sensitive detector (Brookdeal type 411) and a model DM64 oscilloscope (Telequipment Ltd., U.K.).

Moulton of the radiation From the EDL source was achieved by means of a rotatrng sector placed between tk s/Q-wave cavity and the lens. The rotating sector chopped the light beam at 285 Hz; the sector also provided a referena vaitage for operation oE the phase-sensitive detector. The rotating sector was also purged with nitrogen.

The graphite-tube atom cell employed is shown in Fig. 2 and is similar to that described by Dagnall, Johnson and West.’ A graphite tube (high-purity grade) of 6 mm o.d., and 4 mm i.d. and 20 mm in length was machined from graphite rod (Ringsdorf Company Ltd.). A sample introduction port was provided by a 2 mm hole drilled through the wall of&e tube. The graphite tube was supported by a graphite split-ring at each end and the assembly was then retained by two stainless-steel L-shaped holders_ One holder was secured to a stainless-steel support column and the second rested on a second column and was earthed t&a a length of copper braidmg attached to the base of the unit. One holder was thus allowed to “float” to simplify the alignment procedure and prevent fractures in the graphite tube owing to its expansion on heatmg. Roth support columns were water-cooled.

Nanogram amounts of iodine 515

m u Water cooling

Fig. 2. (a) Graphite-tube furnace assembly. (h) Detail of construction of graphite tube and holders.

The graphite tube was heated electrically; power was supplied via a Variac transformer (20-A, Claude Lyons Ltd., Liverpool. U.K.) and a large power transformer (rating 10 kW, Foster Transformers Ltd., London S.W.19) to provide 10 V and up to 450 A across the furnace. The entire cell unit was maintained under a constant flow of oxygen free nitrogen by means of a glass housing fitted with a removable cover to allow the sample to be transferred to the tube, and calcium fluoride windows to permit transmission of the source radiation. The atom- cell was nitrogen-purged by a supply separate from that used for the optical system, so that the flow-rate could be adjusted without affectmg the transmrssion of the optical path. Sample solutions (10 ~1) were transferred to the graphite tube by Eppendorf micropipette.

RESULTS AND DISCUSSION

Operation of iodine EDL source

The relationship between emission intensity and applied microwave power for the iodine electrodeless discharge-lamp source was studied at both the 183-O and 178.2 nm iodine resonance lines. The variation in line emission intensity at 183-O nm with microwave power is shown in Fig. 3; the curve obtained at 178.2 nm was similar. The decrease in intensity with increasing applied power reflects the increasing temperature attained in the resonant cavity with increasing applied power and subsequent condensation of the iodine onto the cooler wall of the EDL, not contained in the cavity. The discharge was difficult to sustain at less than 10 W. At between 10 and 15 W the emission intensity remained relatively constant. The absence, of a well-defined region at low power in which the discharge can be maintained and where the emission intensity changes rapidly with variation of applied microwave power (i.e.. the power curve has a steep positive slope) results in inabi- lity to modulate the radiation electronically. As shown in Fig. 3, if modulation is applied

576 M. J. ADAMS, G. F. KIRKBRIGHT and T. S. WEST

0.c. component of the emitted radiation

3 IO IS 20 2s 30 35 40

Applied microwave power, W

Fig. 3. Variation of emission intensity at 183.0 nm from iodine EDL source with applied microwave power.

at the 10-15 W applied-power level the resulting radiation exhibits only a small a.c. component. Mechanical modulation of the source intensity at 285 Hz with a rotating sector as described above was therefore empioyed. For the determination of iodine by AAS the power to the EDL source was maintained at 10-12 W; this power was found to give the best signal: noise intensity ratio at both resonance lines.

Operation of graphite-tube furnace for the determination of iodine

The optimum operating conditions for the determination of iodine with the graphite- tube furnace and instrumental assembly described, were established oia atomization and measurement of the peak absorbance obtained at 183.0 nm with 10 ~1 aliquots of aqueous potassium ammonium iodide solution. Figure 4a shows the variation in absorbance at 183.0 nm for 2 ppm of iodine with variation in voltage applied across the graphite-tube furnace. All measurements were made with an applied voltage of 8.5 V. Figure 4b shows the variation in absorbance at 183.0 nm with flow-rate of nitrogen in the furnace chamber. A nitrogen flow-rate as high as 15 l./min may be used before any decrease in absorbance is detected in the atomization of 10 ~1 of a 2-ppm iodine solution (potassium iodide). Above this flow-rate, cooling and dilution effects are observed and the peak absorbance de- creases and becomes less reproducible. When a nitrogen flow-rate of less than 0.5 l./min is used the purging time required between samples and between the drying and atomization steps is lengthened considerably. A flow rate of 1 l./min was maintained in all further work.

The exit and entrance slits of the monochromator were set at 30 pm (spectral band pass 0.05 nm) and a photomultiplier EHT voltage of 1620 V was employed. These settings were found to give optimum signal:noise and signal: background intensity ratios. The photomultiplier output was led to the amplifier, which was adjusted to produce a signal for the phase-sensitive detector such that its d.c. output voltage was 1OV at 100% transmission. The variable time-constant of the phase-sensitive detector was optimized to produce the best signal:noise ratio without distortion of the analytical signal by overdamping. The value of the time-constant used was 100 msec (bandwidth 3 Hz).

Nanogram amounts of iodine 51-l

0.3 -

e oz-/-

Y) 9 0.1 -

7.0 I I I I

7.5 8.0 65 9.0

Voltage across tube. V

0.3

t

f g 0.2

: 0.I ‘-1 I

(a)

(b)

I I I I I I 0 0.5 I.0 I.5 2.0 2.5

Flow rate through chombrr (N,, l./min)

Fig. 4. (a) Variation of absorbance at 1830 nm for 10 ~1 of 2 ppm iodide solution with voltage applied across graphite tube.

(b) Variation of absorbance at 183.0 nm for 10 ~1 of 2 ppm iodide solution with flow-rate of nitrogen in atomizer chamber.

AAS characteristics of iodine in the graphite-tube furnace

With the optimum operating conditions established calibration graphs for iodine atomic absorption at 183-O nm were constructed the samples being introduced as aqueous solutions of potassium iodide, ammonium iodide and potassium iodate. Calibration graphs of identical slope and linear range were obtained in each case; 10 ~1 samples produced calibration graphs linear up to 6 ppm and the sensitivity (for 1% absorption) was 0.04 pprn, i.e., 4 x lo- lo g. For the 178.2 nm iodine resonance line the sensitivity (for 1% absorption) was found to be better by a factor of two, i.e., 002 ppm in a 10 ~1 sample. This corres- ponds to an absolute detection limit of 2 x lo-” g. Although higher sensitivity is obtained at 178.2 nm the recorded emission intensity from the source is lower at this line than at 183.0 nm, so the precision of measurement is lower and a similar practical detection limit of 2 x lo- 9 g is observed. The absorption signals recorded were wholly atomic in nature; no non-specific absorption from aqueous samples solutions was observed at the 179.9, 184.4 and 187.6~nm iodine non-resonance lines from the source when 10 ~1 aliquots of 20 ppm iodide solutions were atomized.

Efict offoreign ions

In many sample types in which the determination of traces of iodide is of interest the predominating inorganic species present from the matrix are salts of the alkali metal and alkaline earth ions. Thus, for example. in biological and botanical samples, foodstuffs and water, large excesses of sodium and potassium (and frequently calcium and magnesium) are present. After sample pretreatment to destroy organic matter by wet oxidation, these

578 M. .i. ADAMS, G. F. KIRKBRIGHT and T. S. WEST

Table 1. Effect of common inorganic salts on iodine absorption

Salt Concentration studied ratio,

MX MX: iodide

183.0 tztn

Change in Increase in Absorbance absorbance, % absorbance 184.4 nm

NaCl 1:i:

0 0 0 +25 o-05 0.05

KC1

Na,SO.+

NaNOs

loo: 1 1OOO:l

1:l 10: 1

100: 1 1000: 1

1:l IO:1

100: 1 1cQo: 1

1:l 10: 1

loo:1 1000: 1

+100 0.22 022 100% abs6rption of radiation at each line 0 0

8 0 0

-640 O-10 0.10 loo”/, absorption of radiation at each line

8 : 0 0 +40 0.10 0.10

100% absorption of radiation at each line 0 0 0 0 0

+15 0.03 i.03 +200 040 0.40

Na,HP04 1:l 0 0 0 10: 1 +25 0.05 0.05

100: I 1000: 1

+75 0.15 0.15 100% absorption of radiation at each line

ions may be present in conjunction with one or more common anions such as chloride, sulphate, nitrate or phosphate. The effect on the iodine determination of different weights of several common inorganic salts likely to be present after sample pretrea~en~ has been investigated. The effect of these salts on the absorbance recorded for 10 4 aliquots of 2- ppm potassium iodide solution (20 ng of iodide) at 183.0 nm is shown in Table 1. The inter- ferences recorded were all positive. This effect may be attributed wholly to non-specific absorption at the 183-O nm line by molecular species such as NaCl. This was demonstrated by measurement of the absorption at the nearby iodine 184.4 nm non-resonance line for similar sample solutions. In each case the same absorbance was obtained at this line as that at the 183.0 nm resonance line for blank solutions containing only the salt investigated, and this was the same as the increase in absorbance recorded at this wavelength for iodide solutions when the foreign salt was added. It is therefore possible to correct for the non-specific absorption interference observed in the presence of moderate concentrations of the salts investigated, by subtraction of the absorbance obtained at 184.4 nm from that at 183-O nm. In the presence of lo-fold w/w quantities of most of the salts studied, however, virtually complete absorption of incident radiation at both wavelengths is obtained and such a correction cannot be made. In the presence of a lOOO-fold amount of sodium nitrate, however, it was possible to record absorption signals at both wavelengths, and correct for non-specific background absorption. This relative transparency obtained with nitrate is possibly due to more complete atom~ation of this salt under the conditions employed. Conversion of alkali metal salts into the nitrates before atomization would thus be preferable in methods developed for examination of biological and botanical samples. As nitrate may be employed in many sample pretreatment procedures for the

Nanogram amounts of iodine 579

oxidative degradation of organic material, this conversion might be accomplished without complication.

Conclusion

The graphite-tube atomizer reported here can provide an efficient atom-cell for the direct determination of nanogram amounts of iodine by atomic-absorption spectrosyopy at 183.0 nm. Though non-specific absorption interference is encountered from moderate amounts of simple inorganic salts, the iodine 184.4 nm non-resonance line may be employed for its correction.

Acknowledgements-We are grateful to the Paul Instrument Fund for a grant for the purchase of equipment and to Britrsh Steel Corp. and the Science Research Council for the CAPS studentship awarded to one of us (M.J.A.).

REFERENCES

1. G. F. Kirkbright and M. Marshall, Anal. Chem., 197244, 1288 2. Idem ibid., 1973.45, 1610. 3. G. F. Kirkbright, T. S. West and P. J. Wilson, Atomic Absorption Newsletter, 1972, 11, 53. 4. G. F. Kirkbright. A. F. Ward and T. S. West, Anal. Chzm. Acta, 1975 62,241. 5. Ident. ibid., 1973. 64, 353. 6. B. V. L’vov and A. D. Khartsyzov, Zh. Prikl. Spektrosk., 1969, 11,413. 7 R. Dagnall. D. Johnson and T. S. West. Atzal. Chrm. Acta, 1973.67, 79.

Zusammenfassung-Die direkte Bestimmung von Jod durch AAS an seinen Resonanzlinien bei 183.0 und 178,2 nm wird beschrieben. Es wird eine Atomquelle aus einem kleinen Graphitrohr, eine elektrodenlose Entladungslichtquelle und ein Vakuum-Monochromator verwendet. Die besten Bedingungen zur Bestimmung von Jod wurden ermittelt; eine ahnliche Empfindlichkeit wird erhalten. wenn man Jodtd- oder Jodatproben untersucht. Mit 10 ~1 waBriger Liisung wurden Empfindlichkeiten (fur 1 y0 Absorption) von 4 x lo-” und 2 x lo-” g J bei 183,0 bzw. 178.2 nm erzielt; bei beiden Linien betrug die Nachweisgrenze 2 x lo-’ g. Unspezifische Molektilabsorption haufig vorkommender anorganischer Ionen stiirt bei der Bestimmung; sind mlRige Mengen gewohnlicher Salze anwesend, dann kann man mit der Jodhnie bei 184,4 nm ohne Resonanzcharakter dafur korrigieren.

Rhumiz-On decrit le dosage direct de I’iode par spectromttrie d’absorption atomique avec ses raies de resonance 183.0 et 178.2 nm en utilisant un petit atomiseur a tube de graphite, une lampe a dicharge sans electrode comme source et un monochromateur sous vide. On a Ctabli les conditions optimales pour le dosage de l’iode; une sensibilite similaire est obtenue quand on examme des Cchantillons d’rodure ou d’iodate. Avec des tchantillons de 10 ~1, on a obtenu des sensibilites (pour 19, d’absorption) de 4 x 10-r’ g et 2 x lo-” g de I a 183,0 et 178,2 nm respectivement; on a observe une hmite de detectron de 2 x lo-’ g pour les deux raies. Labsorption moleculaire non specifique des sels mineraux communs provoque une interference dans le dosage: la raie de non-resonance a 184,4 nm peut etre employee pour tenir compte de cette interference quand des quantites mod&es de sels communs sont presentes.

direct determination of nanogram amounts of iodine by atomic-absorption spectroscopy using a...

Documents