Flame Spectrometry in Analysis of Refractory Oxide Single Crystals

This chapter is devoted to the use of flame atomic-emission and atomic-absorption spectrometry techniques to determine the stoichiometric composition, as well as studying the regularities of distribution of impurities and dopants in single crystals based on refractory oxides used as active media for solid-state lasers (leucosapphire, ruby, yttriumaluminium garnet), laser light modulators (strontium titanate), scintillation detectors (cadmium tungstate), solid electrolytes (┚-alumina), medical implants (leucosapphire). Growth of high quality single crystals requires not only adherence to specifications and the growing parameters but reliable methods of analytical control of their composition. Last years along with so-called F-centre forming elements, the increasing attention is given to alkali and alkali-earth impurities which also may cause worsening of quality of single crystals (Dobrovinskaya et al.,2007; Nagornaya et al., 2005; Tupitsyna et al., 2009). Digestion of refractory oxide single crystals is one of the most complicated and important stage of the procedure of analysis. Condensed phosphoric acid is recommended as an effective reagent for sample preparation of single crystals of oxide compounds. For the rapid flame spectrometry determination of alkali metals and calcium impurities, procedures of the sample preparation of water extracts of αand γ-forms of aluminum oxide, yttriumaluminum garnet and magnesium-aluminates used as raw materials for single crystals growing are proposed. It is also shown the effectiveness of ultrasonic sample preparation technique followed by the direct determination of Co and Ni in suspensions of α-Al2O3.


Introduction
This chapter is devoted to the use of flame atomic-emission and atomic-absorption spectrometry techniques to determine the stoichiometric composition, as well as studying the regularities of distribution of impurities and dopants in single crystals based on refractory oxides used as active media for solid-state lasers (leucosapphire, ruby, yttriumaluminium garnet), laser light modulators (strontium titanate), scintillation detectors (cadmium tungstate), solid electrolytes ( -alumina), medical implants (leucosapphire).Growth of high quality single crystals requires not only adherence to specifications and the growing parameters but reliable methods of analytical control of their composition.Last years along with so-called F-centre forming elements, the increasing attention is given to alkali and alkali-earth impurities which also may cause worsening of quality of single crystals (Dobrovinskaya et al.,2007;Nagornaya et al., 2005;Tupitsyna et al., 2009).Digestion of refractory oxide single crystals is one of the most complicated and important stage of the procedure of analysis.Condensed phosphoric acid is recommended as an effective reagent for sample preparation of single crystals of oxide compounds.For the rapid flame spectrometry determination of alkali metals and calcium impurities, procedures of the sample preparation of water extracts of αand γ-forms of aluminum oxide, yttriumaluminum garnet and magnesium-aluminates used as raw materials for single crystals growing are proposed.It is also shown the effectiveness of ultrasonic sample preparation technique followed by the direct determination of Co and Ni in suspensions of α-Al 2 O 3 .

Phosphoric acid digestion of refractory oxide single crystals followed by flame spectrometry analysis
The most commonly used methods of digestion of sparingly soluble substances have a number of shortcomings.Alkali fusion techniques result in high blank values for alkali-earth elements and totally unsuitable for alkali elements determination.Heating of single crystals samples with different acid mixtures is time consuming and not always effective procedure even when autoclave digestion method is used (Foner,1984; Haines, 1988; Homeier et al.,  1988; Krasil ' shchik et al., 1986; Morikava, 1987; Otruba, 1990).It is shown, that for the acid digestion of minerals and oxides condensed phosphoric acid can be successfully applied (Bock, 1979;Hannaker, 1984;Mizoguchi, 1978).It can be obtained by dissolving of P 2 O 5 in www.intechopen.comAtomic Absorption Spectroscopy 106 75 % orthophosphoric acid.Also, upon heating orthophosphoric acid, condensation of the phosphoric groups can be induced by driving off the water formed from condensation (Mizoguchi, 1978;Zolotovitskaya et al., 1984).Sometimes, dehydration is performed at the reduced pressure and temperature 260…300 ºC (Corbridge, 1990).Condensation of orthophosphoric acid leads to forming the mix of polyphosphoric acids which are strong complexing agents having high dissolving ability (Zolotovitskaya et al., 1984(Zolotovitskaya et al., , 1997;;Trachevskii et al. 1996).In this section the main results on application of condensed phosphoric acid for the digestion of wide range of compounds are presented.Among the investigated materials were aluminium oxide single crystals and fusion mixtures; magnesium aluminium spinel doped with Fe, Ni, Co, V, Ti; yttrium aluminium garnet; gallium scandium gadolinium garnet doped with Nd; cadmium tungstate; strontium titanate and other functional materials.

Thermally activated acid-base transformations in the system "condensed phosphoric acid -oxide material"
The composition of the polyacids obtained after heating of an orthophosphoric acid, the mechanism of solvent action of polyphosphoric acids as well as the composition of complex compounds formed during digestion of analysed materials in condensed phosphoric acid has been studied by heteronuclear NMR spectroscopy (Trachevskii et al., 1996).The NMR spectra ( 31 P, 17 O, 27 Al, 113 Cd) of liquid and solid samples were recorded on Bruker CXP-200 spectrometer using single and multipulse sequences as well as magic angle spinning techniques.Heating-up to 400 °C of 75 % orthophosphoric acid leads to driving off the water and to forming linear polyphosphoric acids.Identification of the obtained substances was performed using 31 P NMR spectroscopy of water solutions of corresponding salts as well as water solutions derived from melts partly neutralized by ammonia.In the investigated temperature range (20…400 ºC) a set of processes of orthophosphoric acid transformations can be expressed by the following equation: where n = 1…12.From the point of view of prediction of reactivity of dissolvent the effect of upfield shift of signals in 31 NMR spectra, corresponding to consecutive thermal generation of polyphosphatic homologues (with п=2, 3, 4) (Trachevskii et al., 1996) is important.The analogous trend of change of signals was observed at shift of acid-base equilibriums towards augmentation of protonation of anions: With driving off the water, the equilibrium 3 4 + 2  3 + + 2 4 -is shifting towards formation of molecular form 3 4 .At these conditions the strongest base In Fig. 2 dependences of dissolution rate of different fusion mixtures and single crystals on temperature are presented.The mass of samples was 0.05…2.00g and volume of H 3 PO 3 was 10 ml for Al 2 O 3 or TiO 2 and 5 ml for WO 3 .Dramatic increase in dissolution rate is observed at 210…270 °C for charge and at 250…300 °C for single crystals.Optimum values of temperature for dissolving of fusion mixtures and single crystals were 270 and 300 °C, respectively.It was unreasonable to dissolve such materials at higher temperatures due to formation of sparingly soluble vitreous products.The obtained data were used for optimization of sample preparation procedures of some oxide materials (Table 1).NMR-spectroscopy data (Trachevskii et al., 1996) has allowed us to explain high hydrolytic stability of phosphate decomposition products.Analysis of 31 NMR spectra for the system of K 2 WO 4 -3 4 has shown that after dilution there are significant amounts of diphosphate anions and their complexes with tungsten in solutions.Presence of specified diphosphate complexes as well as solvable orthophosphate complexes is the main reason of stability of diluted solutions.Obviously, similar processes cause the same behavior of other oxide materials studied in this work.

Material
Sample mass, g where n can vary from 1 to 3, m and k -from 1 to 5, x -from 0 to 0,2.Flame spectrometry measurements were performed on a spectrophotometer "Saturn" (Ukraine) with use of acetylene-air, propane-butane-nitrous oxide and acetylene-nitrous oxide gas mixtures.The processes taking place during flame atomization of materials and excitation are well studied (Havezov et al., 1983;Haswell, 1991;Hill, 2005;L'vov et al., 1975;Magyar, 1987;Welz , 1999).However, application of flame spectrometry in analysis of solutions of the complex composition especially containing metal phosphate complexes, demands to study of influence of various factors on absorption and emission signals of analytes.Most essential of them are the main composition of solution and conditions of atomization of material in a flame.The last depends on the chosen gas mixture, a fuel/oxidant ratio, zone of photometric observations.The numerous data available in the literature on this problem are ambiguous.

Effects of redox characteristics of the flame and the height of the burner on analytical signals
The redox characteristics of a flame defined by the fuel/oxidant ratio were estimated visually, basing on the height of the inner cone (l, mm) which was changed over a range from 1 to 5 for acetylene-air, propane-butane-nitrous oxide flames and from 1 to 40 for acetylene-nitrous oxide flame.It was considered that acetylene-air and propane-butanenitrous oxide flames were oxidative, stoichiometrical or reducing when l was equal to 1…2, 3…4 and 5…6, respectively.The corresponding values for the acetylene-nitrous oxide flame were 1…2, 2…35 and 40…45.
In order to investigate effect of redox characteristics of the flame on analytical signals they were recorded at different distances above the burner orifice while l value was constant.It has been found that analytical signals of all elements of interest depend on redox characteristics of a flame.Most distinctly these dependences are appeared for the metals forming hardly dissociated compounds (alkali-earth and rare-earth elements, Al) even when a high-temperature flame is used (Maitra, 1987;Pupyshev. et al., 1990).
As an example, in Fig. 3  It can be seen that at determination of these elements, especially by atomic-emission method, it is necessary to select a fuel/oxidant ratio very carefully.For alkali-earth metals, Mg and metals of subgroup Fe composition of gas mixture is less critical, however also is important enough.So, sensitivity of the determination of alkali-earth metals in the presence of 3 4 in acetylene-air oxidizing flame is ~2 times lower, than in stoichiometrical or reducing flame.In Table 2 optimum operating conditions for the determination a number of elements in phosphate solutions are resulted.It is found that the optimal fuel/oxidant ratio does not depend on the sample matrix.Fig. 4 demonstrates dependences of analytical signals of Mg, Ca, Sr, Ba, Al and Y from the height of the burner for the solutions of different composition.It can be seen that there is a correlation between the maximum of analytical signal and the height of the inner cone.In the presence of a "heavy" matrix the specified maximums shift to a hotter flame zone which is located directly over an inner cone or close to it.In some cases (see curves a, e) maximums are diffuse.Buffer additions of chlorides of alkali metals have no effect on the position of maximum.Also, it is found that analytical s i g n a l s o f a l k a l i m e t a l s d o n o t d e p e n d considerably on the height of the burner.

Element Method
Wavelength, nm Flame1 Gas flow rate, L h

Solvent effects on the absorption and emission spectra of analysed elements
Effect of phosphate ions on analytical signals of alkali, alkali-earth, transition and other elements was studied in detail in many papers (Magyar, 1987;Smets, 1980).However, these results look rather discordant.This fact can be explained by variety of spectrometers and hardware parameters used as well as by quite different concentrations of phosphate ions and analysed elements.Therefore, effect of phosphate ions on emission signals of Al, alkali, alkali-earth elements and absorption signals of Mg, Al, Y, Ti, V, Fe, Co, Ni in different flames at the constant concentration of the phosphoric acid (12.4 mas.%) has been investigated.The results presented in Fig. 5 show that effect of a phosphoric acid on analytical signals is various.
In an acetylene-air flame 3 4 shows strong depressing effect on all alkali elements.This effect increases in the order Li <Na, K <Rb <Cs, that is in a good agreement with the literary data (Henrion et al., 1979).In a propane-butane-nitrous oxide flame effect of 3 4 on Li, K signals practically misses.For Rb and Cs the depressing effect is smaller, and for Na is greater, than in an acetylene-air flame.
It is noted that effect of phosphate ions on analytical signals of analysed elements depends essentially on a composition of gas mixture.For example, in an oxidative acetylene-air flames an emission signal of K in the presence of  Probably, it can be explained by decrease of partial pressure of elemental oxygen and the considerable facilitation of atomisation of oxides of analysed elements in the presence of aluminium (Grossmann, 1987).
In an acetylene-air flame Al demonstrates depressing effect on analytical signals of nickel and cobalt which is more substantial for Ni.On the contrary, absorption signals of Co and Ni in a propane-butane-nitrous oxide flame are increased.
It is known from the literary (Henrion et al., 1979) that W has a depressing effect on emission signals of alkali metals which is incremented in the order Li < Na < K < Rb < Cs.
As have shown results of our investigations, in the presence of cadmium and phosphate ions, tungsten acts quite differently.As can be seen from Thus, character and power of matrix effect in phosphoric acid solutions of examined materials depend on the nature of the element to be analysed, measurement conditions and a composition of a gas mixture.Optimisation of these parameters allows analyst to minimise matrix effect.

Interference effects of analysed elements in phosphate solutions
Avoiding of interference effects is of significant importance in the high-precision determination of a stoichiometric composition of functional materials.Study of interferences of Mg, Al and Y in yttrium-aluminium garnet and magnesium aluminium spinel was performed using solutions prepared by digestion in a phosphoric acid of oxides and mixtures of oxides of these elements.The content of each oxide in a mixture matched to stoichiometric relationships 3Y 2 O 3 • 5Al 2 O 3 and MgO•Al 2 O 3 or by 2-5 times differed from them.The content of dissolved oxides in solutions was 0.5-3 g L -1 .It was found that in an acetylene-air flame a strong depressing effect of Al on absorption signals of Mg is observed.In a propane-butane-nitrous oxide flame this effect is weaker, whereas in an acetylene-nitrous oxide flame it is insignificant (Fig. 6).Authors (Havezov, 1983;Lueske, 1992)  It was also found that interferences of Na, K, Ca, Mg during their determination in CdWO 4 , Al 2 O 3 , yttrium-aluminium garnet and magnesium aluminium spinel can be eliminated by using of a cesium chloride buffer.In this case, it is possible to determine all these elements from the same solution.

Some practical applications of phosphoric acid digestion of refractory oxide single crystals followed by flame spectrometry analysis
Our investigations have shown efficiency of use of a condensed phosphoric acid for dissolution of different refractory oxide single crystals.The data presented in the previous sections were used for the development of new procedures for determination of main components, dopants and microimpurities in single crystals and raw materials.For the determination of a stoichiometric composition of yttrium-aluminium garnet and magnesium aluminium spinel, powder reference materials 3Y 2 O 3 •5Al 2 O 3 and MgO•Al 2 O 3 were used.The content of metals in these compounds was determined precisely by chelatometry.Calibration standards were prepared by dissolving of different masses of a powder reference material in condensed phosphoric acid.All solution contained CsCl as a spectroscopic buffer.An acetylene-nitrous oxide flame was used in the most cases but the determination of Mg is also possible with use of a propane-butane-nitrous oxide flame.Tables 4 and 5 contain data on the methods accuracy check and comparison with the results of chelatometry (in terms of oxides).As can be seen from the Table 4, high concentrations of Al can be determined by both FAES and FAAS methods but the last one is more preferable.It was found that K and Na can be determined in sparingly soluble tungstates KGd(WO 4 ) 2 and BiNa(WO 4 ) 2 after their dissolving in a condensed phosphoric acid with a confidence interval not exceeding 0.5…0,7 mas.%.Moreover, the absence of matrix effect allowed carrying out the analysis with use of calibration standards containing solvent and analysed elements only (CsCl was used a spectroscopic buffer for the determination of Na).Accuracy of developed procedures for the determination of dopants in aluminium oxide based materials was checked by comparison with results obtained by an independent method as well as using a different method of sample preparation (Table 6).The proposed procedures are more precise than arc atomic emission analysis and do not require large amounts of samples as is required in X-ray fluorescence analysis.

Material
Table 7. Metrological characteristics of procedures of flame spectrometry analysis of single crystals with phosphoric acid digestion at the stage of sample preparation

Analysis of aqueous extracts of powders based on aluminium oxide
The traditional methods of sample preparation of aluminium oxide such as fusion with fluxes and selective acid dissolution in autoclaves are time consuming and unsafe.Digestion of oxide materials in condensed phosphoric acid is express and reliable method.However, a strong depressing effect of phosphoric acid on analytical signals in atomic spectrometry does not allow to provide low detection limits for many elements.Recently, procedures of the direct analysis of oxide materials from aqueous suspension are widely used, but this method (see section 4) requires use of standards identical to analysed samples.This section contains results of the study of aqueous extraction of K, Na, Ca, V, Ni and Co from -, -Al 2 O 3 , mAl 2 O 3 • nY 2 O 3 and mAl 2 O 3 • nMgO used as raw materials for single crystals growing.
Leaching procedure was carried out in quartz beakers (50 ml) at ambient temperature by mixing of powders with ultrapure water to obtain stable suspension followed by spontaneous sedimentation of a solid phase.Heating of the sample has appeared unacceptable owing to oxide adherence to beaker walls.It was also found that the spontaneous sedimentation of Al 2 O 3 solid phase and Al(O ) 3 formed during leaching is an optimum technique of sample preparation.Centrifugation of suspension is found to be unsuitable because of formation of very stable suspension.Especially, this is typical for fine powders of γ-А1 2 3 and Y 3 Al 5 O 12 .Effect of α-A1 2 O 3 particle size on the recovery of Na, K and Ca is shown in Fig. 7.One can see that the full extraction of elements is observed up to particle size < 60 microns and for K It is possible to make the following assumptions concerning effects of transferring of impurities of alkali metals and Ca in an aqueous phase during the extraction procedure irrespective of Al 2 O 3 modifications:  along with increase of specific surface of materials during grinding, a great number of microinhomogeneities and defects are appeared on a surface of particles (Dobrovinskaya et al., 2007;Hanamirova, 1983;Foner, 1984);  γ-А1 2 3 and Y 3 Al 5 O 12 fine powders (<< 1 µm) are partly hydrated and some amounts of Al(OH) 3 are always present;  after calcination of alum at temperature of 1200 ºC there is a part of unevaporated sulphates of alkali metals in the obtained aluminium oxide.The full decomposition of suphates takes place at higher temperatures (Chebotkevich, 1975;Hanamirova,1983).Sulphates of alkali metals have a good solubility in water;  it was experimentally confirmed by other authors (Ermolenko et al., 1971;Hanamirova, 1983;Uksche, 1977 that at temperatures of 600-1200 ºC aluminate of alkali and alkaliearth element of spinel type are generated.Owing to a "quasiliquid" state of cationic sublattice ions of alkali metals can migrate freely via interblock clefts and substitute each other in the spinel cluster (Uksche, 1977).Aluminates of alkaline and alkali-earth metals in water are hydrolyzed to А1(OH)3 and hydroxyaluminates (Hanamirova, 1983;Remi, 1973).Apparently, in aqueous suspensions of А1 2 3 and Y 3 Al 5 O 12 all these factors may take place and may have a different contribution depending on the dispersity and "prehistory" of material.It was found that the optimum sample mass/water volume ratio is 0.1 g per 10 ml of water.The time of obtaining of transparent extracts was 10 and 12 hours for α-A1 2 O 3 and γ-А1 2 3 (Y 3 Al 5 O 12 ), respectively.It is a compromise between the sensitivity of analysis and sample preparation time.Optimum conditions of the flame spectrometry determination of Na, K and Ca in aqueous extracts of Al 2 O 3 and Y 3 Al 5 O 12 have been studied.It has been found that Na, K and Ca increase analytical signals of each other.Al and Y up to concentration of 10 mg L -1 do not affect analytical signals of Na and Ca but increase the analytical signal of K. Using of CsCl as a spectroscopic buffer (2 g L -1 ) has allowed us to eliminate interference effects completely.The results of analysis of aluminium based oxide materials obtained using the proposed method of sample preparation and digestion in condensed phosphoric acid are shown in Table 8.One can see that aqueous extraction technique provides lower limits of determination (LOD) of Na, K and Ca and does not yield to more complicated techniques described in (Krasil ' shchik et al., 1989; Slovak et al., 1981) •nMgO:Co(Ni) was mixed with 10 ml of water and 0.2 ml of HNO 3 and then ultrasonic treatment was applied.Solutions obtained after the spontaneous sedimentation of solid phase were used for analysis.It was found that recovery of Co and Ni is quite similar and increases with augmentation of intensity of ultrasonic irradiation and processing time (Fig. 8 and 9).The maximum in Fig. 8 corresponds to optimum cavitation conditions.However, the maximum recovery at the optimum conditions (I = 78 W, τ =15 min) was only 50 %.It was assumed that only those atoms located on grains boundary (not entered into crystal lattice owing to various ionic radiuses of cobalt (nickel) and aluminium (Dobrovinskaya et al., 2007) have been extracted.In this work calibration standards were prepared by using melt fusion technique.It was found that for the determination 100…2000 ppm of Co(Ni) it is necessary to have not less than five calibration standards with concentrations of an analyte increased by a factor of 2-2.5.Calibration standards were prepared as follows.Extra pure ammonia alum was melted in platinum crucible on a hot plate and then cobalt (or nickel) sulphate preliminary calcined at 1200 °C was added.The melt was thoroughly stirred and dried under an infrared lamp.
The solids was placed into a muffle furnace and heated slowly during 4 hours up to 1100 °C with subsequent holding during 3 hours with subsequent calcination at this temperature during 3 hours.After cooling aluminium oxide was ground to powder in a teflon mortal during an hour and sifted through a 40 µm sieve.Content of dopant in the prepared samples was approved by chelatometry.All subsequent procedures were the same as those described above.Determined results were consistent with those obtained by analysis of samples digested in condensed phosphoric acid (Table 10).
As can be seen from

Conclusion
The results presented in this chapter show that flame atomic spectrometry can be successfully used for the express control of stoichiometric and impurity composition of refractory oxide single crystals such us Al 2 O 3 , yttrium-aluminium garnet, cadmium tungstate, strontium titanate.Precision and accuracy of analytical procedure as well as metrological characteristics depend strongly on the technique of sample preparation.Use of condensed phosphoric acid as a solvent allows in comparison with traditional techniques of decomposition to unify a sample preparation procedure and to reduce the time of analysis by 5-10 times.Optimum conditions of digestion of refractory oxide materials in condensed phosphoric acid are established.Influence of solvent, matrix and chemical interference effects on analytical signals of analysed elements with use of different mixed gas flames are also studied.It is shown that in some cases use of a propane-butane-nitrous oxide flame which is preferable and can significantly depress effects of macrocomponents and solvent.Determination of stoichiometric composition of single crystals requires paying much more attention to interference effects as well as matching of analysed materials to calibration standards as compared to impurities determination procedures.
The developed analytical procedures allow to evaluate major elements content such us Mg, Al and Y in different oxide compounds with relative standard deviations (%) 0.8…1.2,1.0…1.5 and 1.5…2.5, respectively.
The proposed sample preparation methods for water extracts of aluminium oxide powders and yttrium-aluminium garnet allows to carry out an express determination of alkaline metals, calcium, nickel and cobalt in raw materials for single crystals growing.These sample preparation methods ensure low blank signals and better reproducibility of analytical signals as compared to acid digestion technique.Limits of determination of impurities in all investigated materials after phosphoric acid digestion were 1…5 ppm.Na, K and Ca can be determined in aqueous extracts of Al 2 O 3 with the limit of determination of 0.2 ppm.It was shown that Co and Ni dopants can be determined by FAAS directly from aqueous suspensions of aluminium oxide based materials.

Fig. 3 .
Fig. 3. Dependence of absorption signals of Al (1) and Y (2) and an emission signal of Al (3) in phosphate solutions of Y 3 Al 5 O 12 on the inner cone of the flame (l, mm)

Fig. 7 .
Fig. 7. Effect of α-A1 2 O 3 particle size on the recovery of Na, K and Ca

Table 2 .
Optimum operating conditions of the atomizer Generally, change of optimum fuel/oxidant ratio as well as the height of the burner can lead to the considerable decrease in sensitivity and increase uncertainty in FAES and FAAS analyses of phosphate solutions.

Table 3 .
Results of investigation of effect of major elements (Al, W, Ti) of high-melting compounds on analytical signals of analysed elements are presented in Table3.Concentration of the phosphoric acid was 12.4 mas.%.Measurement conditions matched those given in Table2.One can see that in comparison with effect of a phosphoric acid, Al increments emission signals of Li and Na.With increase of the Na concentration the rate of this effect decreases.Al shows depressing effect on the emission signal of K, and the rate of this effect depends on redox characteristics of a flame and is incremented with decrease of the content of a combustible gas in a gas mixture.Increase in analytical signals of Na in the yttrium-aluminium garnet matrix is observed, whereas analytical signals of K do not change substantially.Effect of matrix components on analytical signals of elements An absorption signal of Mg in the presence of Al decreases slightly and with increase of the Mg concentration the rate of effect of Al increases.Use of a propane-butane-nitrous oxide flame allows analyst to eliminate matrix effect completely.Aluminium based compounds (Al 2 O 3 , magnesium aluminium spinel, yttrium-aluminium garnet) show depressing effect on analytical signals of Ca, even in a high-temperature acetylene-nitrous oxide flame.The rate of depressing effect increases in the order yttriumaluminium garnet < magnesium aluminium spinel < Al 2 O 3 which is correlated with aluminium content in these compounds.Increase in analytical signals of Ca in SrTiO 3 matrix can be explained by the fact that Sr acts as an ionization buffer(Havezov,1983).Results of our investigations have shown that absorption signals of hardly volatile elements such as Ti and V, and emission signals of Nd in the presence of Al 2 O 3 are increased.
Table 3, a cadmium tungstate matrix does not affect analytical signals of alkali metals in acetylene-air and propanebutane-nitrous oxide flames.Absorption signals of Mg in a cadmium tungstate matrix in acetylene-air and propanebutane-nitrous oxide flames are reduced by 80 and 30 %, respectively.This matrix has a weak depressing effect on emission signals of alkali-earth elements in an acetylene-nitrous oxide flame.

Table 6 .
Results of the determination of dopants in aluminium oxide based materials (n=3; P=0.95)Some metrological characteristics of developed procedures are presented in the Table7.Thus, phosphoric acid digestion is a very effective and universal method of sample preparation of refractory oxide single crystals and raw materials.In some cases this is the only method suitable for the successive determination of alkali metals in single crystals since sample solutions obtained by alkali fusion may have extremely high blank signals.

Flame spectrometry determination of dopants Co and Ni in subacid aqueous extracts of α-and γ-Al 2 O 3
. The developed technique can be used for the express quality control of fusion mixtures of А1 2 3 and Y 3 Al 5 O 12 of different dispersity and for the determination of alkali metals dopants in А1 2 3 charge.The possibility of application of subacid aqueous extraction technique combined with ultrasonic and microwave treatment for the determination of Co and Ni in -and -Al 2 O 3 fusion mixtures was studied.An ultrasonic dispenser UZDN-A (Selmi, Ukraine) with operating frequency of 22 kHz a microwave digestion system MDS-2000 (CEM, USA) were used for the treatment of subacid aqueous suspensions of Al 2 O 3 powders.Fractions with particle size less than 40 µm were used throughout.0.1 g of -Al 2 O 3 :Co(Ni) or mAl 2 O 3

Direct flame atomic absorption determination of Co and Ni dopants from aqueous suspensions of aluminium oxide based materials
Thus, ultrasonic treatment of aqueous suspensions of Al 2 O 3 p o w d e r s i s f o u n d t o b e unsuitable for quantitative analysis of extracts.However, as have shown our further investigations, it is possible to use successfully suspensions of -Al 2 O 3 for direct determination of Co and Ni dopants (see section 4).Extraction of Co and Ni from subacid aqueous suspensions of Al 2 O 3 powders using microwave heating in closed vessels was found to be more effective.Dependence of recovery of Co and Ni on time of exposure (at pressure of 100 psi) and acidity of aqueous suspensions was studied.0.05, 0.2, 0.5 and 1 ml of HNO 3 were added to 10 ml of aqueous suspensions of Al 2 O 3 in order to obtain the samples with different acidity.As can be seen from Fig.10, the most principal parameter is exposure time.Obviously, the transfer of dopants into solution under microwave heating is caused by deterioration of surface layers and activation of diffusion-controlled transfer processes.Accuracy of the developed procedure for the determination of dopants in aluminium oxide based materials was proved by comparison with the results obtained by FAAS after digestion of samples in condensed phosphoric acid (Table9).The obtained results are in good agreement with each other.The proposed sample preparation technique has allowed us to improve limits of determination of Co and Ni owing to exclusion of acid solvents from analytical procedure and use of aqueous calibration standards.Whereas Co and Ni dopants cannot be extracted quantitatively to aqueous phase during ultrasonic treatment of Al 2 O 3 suspensions, it is possible to use directly such suspensions in analysis by flame atomic absorption spectrometry.Under ultrasonic irradiation the average size of suspended particles of a powder of Al 2 O 3 decreases and suspension of very good stability and uniformity is formed.This is an important factor which affects degree of atomization and detection limits.For example, analytical signals of Co and Ni in such suspensions are two times higher in comparison with those obtained in suspensions prepared at ambient temperature by mechanical agitation.If there are certified reference materials available, this analysis technique has a number of benefits such us speed and low labor expenditures. www.intechopen.com

Table 10 .
Table 10, the proposed method has lower relative standard deviations as compared to the acid digestion procedure.Comparison of results of the determination of Co and Ni with use of two different sample preparation techniques (n = 6; P = 0.95)