Determination of Dopants and Impurities in Optical Crystals of p=Barium Borate by Inductively Coupled Plasma Atomic Emission Spectrometry and Flame Atomic Absorption Spectrometry Journal of Analytical I Atomic I Spectrometry 1 ELISAVETA IVANOVA NONKA DASKALOVA SERAFIM VELICHKOV PETRANKA SLAVOVA AND GALJA GENTSCHEVA Institute of General and Inorganic Chemistry Bulgarian Academy of Sciences BG-SoJia 11 13 Bulgaria Inductively coupled plasma atomic emission spectrometry and flame atomic absorption spectrometry were applied to the determination of dopants (Ce Nd Eu and Er) and impurities (Fe Na Mg and Al) in the optical crystals of 8-barium borate dissolved in hydrochloric acid. The matrix interferences occurring in analyses carried out by means of ICP-AES and FAAS were studied. The detection limits of the analytes in a 0.8% solution of pbarium borate in 2 mol I-' hydrochloric acid were determined.The relative standard deviation of the obtained results lies within 2-6%. Keywords Dopants; impurities; P-barium borate; single crystals; inductively coupled plasma atomic emission spectrometry; flame atomic absorption spectrometry P-Barium borate is a relatively new and important material for non-linear optical applications in the visible and ultraviolet regions.'.2 It has a number of physical properties which make it particularly attractive. It exhibits large effective coefficients of second harmonic generation a wide transparent waveband high damage thresholds and high optical homogeneity. These properties suggest that P-barium borate will have important applications especially in the ultraviolet region.' The concept of improving or modifying the characteristics of non-linear optical crystals is not new.For many crystals of the solid-solution type it is possible to effect changes in certain material properties such as the birefringence and transparency by selecting an appropriate composition of the various elements in the c r y ~ t a l . ~ Doping is a way of varying this composition. On the other hand the incorporation of impurities in the crystal also affects its proper tie^.^ The correct interpretation of the properties of non-linear optical crystals and the control of their synthesis requires a knowledge of the concentration of dopants and impurities. This knowledge is acquired by analysis.The purpose of the present work was to determine the concentration of dopants (Ce Nd Eu and Er) and impurities (Na Fe A1 and Mg) in the optical crystals of P-barium borate with a view to the analytical control of their synthesis. Inductively coupled plasma atomic emission spectrometry is a technique particularly appropriate to this purpose.' Comparative data for the concentration of the trace impurities were obtained by flame atomic absorption spectrometry. EXPERIMENTAL Instrumentation The Jobin-Yvon (Longjumeau France) equipment and the operating conditions used in the ICP-AES experiments are specified in Table 1. T, was measured by the Boltzmann plot method.6 The FAAS measurements were performed with a Pye Unicam SP 192 spectrometer using the most sensitive lines of the elements under the operating conditions recommended in the manuals including deuterium background correction.Reagents and reference solutions Reagents of the highest available purity and triply distilled water from a quartz apparatus were used. The stock solutions of the elements (1 mg ml-') were prepared from Merck Titrisols when available (E. Merck Darmstadt Germany). The stock solutions of Ce Nd Eu and Er were prepared by dissolving the corresponding oxide (Specpure Johnson and Matthey Royston Hertfordshire UK) in hydrochloric acid as described el~ewhere.~ The reference solutions were prepared in 2 mol 1-' hydrochloric acid. The solvent blank was 2 mol 1-' hydrochloric acid while the matrix blank was a 0.8% m/v solution of undoped P-barium borate.Plastic laboratory ware was used throughout. Dissolution procedure A 200 mg amount of the finely ground crystalline material was dissolved in 2 mol 1-' hydrochloric acid in a polypropylene cup at room temperature. The solution was transferred into a Table 1 conditions Speciation of the spectrometer ICP source and operating Monochromator Mounting Grating Wavelength range Dispersion Entrance slit Exit slit Resultant spectral slit Practical spectral bandwidth Photomultiplier rf Generator Frequency Oscillator Power output Nebulizer Pump Operating conditions Incident power Reflected power Outer argon flow rate Carrier flow rate Liquid uptake rate Excitation temperature Transport efficiency of ICP system JY 38 (Jobin-Yvon) Czerny-Turner focal length 1 m Holographic 2400 grooves mm - ' 170-700 nm (1st order) 0.38 nm mm-' 0.02 mm 0.04 mm 15.2 pm 15.6 pm Hamamatsu TV R 446 HA Plasma Them Model HFP 1500 D Crystal controlled at 13.56 MHz Meinhard concentric glass Peristaltic ten-roller Gilson Minipuls I1 (Gilson Medical Electronics France) 27.12 MHz (+0.050/,) 0.5- 1.5 k W 1.0 kW 10 w 15 1 min-' 0.5 1 min -' 1.3 1 min-' 7200 K 3 yo Journal of Analytical Atomic Spectrometry August 1996 Vol. 11 (567-570) 56725 ml polypropylene calibrated flask and was made up to the volume with the same acid.The concentration of P-barium borate in the solution obtained was 0.8% m/v. RESULTS AND DISCUSSION Dissolution of the Single Crystals of PBarium Borate The solubility of the crystalline material was tested in water nitric acid and hydrochloric acid.It was found that a-barium borate is not soluble in water and is hardly soluble in nitric acid. The most effective dissolution was achieved with hydro- chloric acid in the concentration range 2-3 moll-'. No heating was necessary. When hydrochloric acid of lower or higher concentration was used dissolution slowed down and even stopped. An optimum dissolution medium of 2 mol 1-1 HCl was chosen. Solutions of /?-barium borate with concentrations up to 2% m/v were prepared. The optimum concentration of the sample matrix in the solution was found to be 0.8% m/v which permitted a reliable determination of trace element concentrations at a relatively low level of matrix interference in both ICP-AES and FAAS. The solubility of the /?-barium borate samples doped with Ce Nd and Eu was the same as that of the undoped sample.Only the Er-doped sample exhibited lower solubility the 0.8% m/v solution of the Er-doped #?-barium borate was close to its saturated solution. Studies of the Determination of Ce Nd Eu Er Fe Mg Na and A1 in the Solution of #?-Barium Borate by Means of ICP-AES Spectral interferences and line selection Boumans discussed various aspects of spectral interferences in atomic emission spectros~opy.~~~ In the present study the spectral interferences around the prominent analysis lines of Ce Nd Eu Er Na Fe Mg and A1 were investigated in the presence of 5 mg ml-' of barium and 1 mg ml-' of boron as interferents in the solution these being their levels in the 0.8% m/v sample solution. Information on the interfering matrix lines in a A k300 pm spectral window for each of the interferents was derived from wavelength scans centred around the analysis lines Aa of the analytes.Fig. 1 shows an example. El~ewhere,'.~ we have accurately specified how the results were obtained and quantified. The interpretation of the results revealed that firstly all investigated analysis lines are free of line and wing interferences in the presence of 1 mg ml-' boron (the latter has a poor emission spectrum 94 identified emission lines are listed in the MIT Tables"); secondly barium has a richer emission spec- trum than boron (472 identified emission lines are listed in the MIT Tables"). With 5 mg ml-' of barium as interferent the analysis lines under investigation were influenced by wing interferences.Line interferences were established around two analysis lines Ce 11 413 380 pm was overlapped by Ba 413243pm and Eu I1 412970 pm by Ba I1 413066pm (see Fig. 1). The identified interfering barium lines in the spec- tral windows of cerium and europium analysis lines are not included in the Boumans Line Coincidence Tables" but are listed in the MIT Tables." These data provided the possibility of correct evaluation of the type of spectral interferences and optimum line selection.' To the latter purpose the true detection limit criterion of Boumans and Vrakking was applied.12 According to this approach Q-values for line interference [QI(Aa)] and Q-values for wing (background) interference [Q,(AA,)] for each of the above mentioned interferents were distinguished.The term QI(Aa) is expressed as the ratio SI(Aa)/SA where S,(A,) is the partial sensitivity of the interfering line defined as the signal -h Ba 41 3 243 412 970 WavelengWpm Fig. 1 Example of a spectral scan over a spectral region of & 300 pm around the analysis line. Central wavelength 412 970 pm (Eu line). Interferent barium. Analytes are 4 pg ml-' of europium in pure solvent or 5 mg ml-' of barium per unit interferent concentration produced by the interfering line at the peak wavelength of the analysis line A and SA is the sensitivity of the analysis line. The term Qw(Aha) is expressed as the ratio Sw(AA,)/SA where Sw(AAa) is the wing sensitivity of the interfering line in the spectral window AA and SA is as stated above. Table 2 lists Qw (Ah,) and Q,(h,) values in the presence of 5 mg ml-l of barium in solution (as was mentioned above boron does not contribute to the Q- values).Evidently the fi-barium borate matrix causes rather Table 2 Q,(AA,) and QI(Aa) values of the investigated analysis lines of Ce Nd Eu Er Na Fe A1 and Mg. Interferent 5 mg ml-' barium Analysis linefpm Ce I1 413 765 Ce I1 413 380* Ce I1 395254 Nd I1 401 225 Nd I1 430358 Nd I1 406 109 Eu I1 381 967 Eu I1 4129707 Eu I1 420505 Er I1 337271 Er I1 349910 Er I1 323058 Na I 588995 Fe I1 238204 A1 I 396 151 Mg I1 279 553 Q w ( W 6.0 x lo-' 7.4 x lo-' 8.2 x 10-5 1.3 x lo-' 1.1 x 9.6 x 1.9 x 3.3 x 1.6 x lop6 2.1 x lo-' 7.6 x 8.8 x 6.6 x lop6 3.5 x 10-5 5.9 x 1.2 x 10-6 QI (La ) 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8.9 x 10-5* 2.8 x 10-5+ * Overlapping of Ce I1 413 380 pm with Ba 413 243 prn.'' Overlapping of Eu I1 412970 pm with Ba I1 413066 pm." 568 Journal of Analytical Atomic Spectrometry August 1996 VoL 1 1low interferences.This permitted us to use the most sensitive lines of the analytes as analysis lines. Efect of the P-barium borate matrix on the net line intensities of the analytes The net line intensities decreased by about 5% in the presence of 0.8% m/v P-barium borate as a matrix in the solution compared with the net line signals in the pure solvent (2 mol 1-1 HCl). The effect is negligible as regards the detection limits but reflects on the accuracy of the analysis. The correct calibration requires precise matching of the matrix content. Detection limits By use of the Qr(Aa) and Qw(AAa) values the conventional detection limits (CL,conv) and the true detection limits (C,,,,,) were determined in the 0.8% m/v solution of P-barium borate (Table 3).For the sake of comparison the detection limits in the pure solvent (2 mol 1-1 HC1) are also listed (C,). As can be seen the true detection limits of the ICP-AES determination in the presence of the P-barium borate matrix do not signifi- cantly differ from those in a pure solution (column 1) only in the case of wing interference. With additional line interference substantially higher true detection limits are registered. The detection limits with respect to the dissolved solid are of interest rather than the detection limits in solution. Table 4 shows the true detection limits of the analytes with respect to the dissolved solid (0.8% m/v P-barium borate) using the 'best' analysis lines (column 1).The fictitious detection limits i.e. the detection limits in the presence of the same matrix assuming no contribution of an interfering signal at the peak wavelength of the analysis line are also shown (column 2). Study of the Matrix Effects on the Flame AAS Determination of Fe Na Mg and Al in a Solution of /?-Barium Borate The detection limits of the FAAS determination of Fe Na Mg and A1 in /I-barium borate (0.8% m/v solution of P-barium borate in 2 mol 1-' HCl) determined from the fluctuations of the blank signal (30 criterion) are presented in Table 5. Table 3 CL CL,conv and CL,true values of the prominent analysis lines of Ce Nd Eu Er Na Fe A1 and Mg in the presence of 0.8% m/v P-barium borate in solution Analysis line/pm Ce I1 413 765 Ce I1 413 380 Ce I1 395254 Nd I1 401 225 Nd I1 430358 Nd I1 406 109 Eu I1 381 967 Eu I1 412970 Eu I1 420505 Er I1 337271 Er I1 349 910 Er I1 323058 Na I 588995 Fe I1 238204 A1 I 396 151 Mg I1 279 553 CL/ ng ml-' 17.0 17.0 17.0 40.0 50.0 50.0 0.7 1.4 2.0 3.0 3.7 3.7 19.0 8.4 19.0 0.1 CL,conv( CL true /' ng ml- ng ml- 25.0 40.0 215.0 28.0 - - 42.0 - 52.0 51.0 - - 1 .o 6.0 55.0 2.2 - - 6.0 5.0 4.7 - - - 20.0 - 13.0 - 20.0 - 0.2 - Table4 True detection limits with respect to the dissolved solid (0.8% m/v P-barium borate) for the 'best' analysis lines of the elements (column 1).Fictitious detection limits in the presence of the same matrix supposing no matrix interference (column 2) Detection limit/pg g-' 'Best' analysis line/pm Ce I1 413 765 Nd I1 401 225 Eu I1 381 967 Er I 337271 Na I 588995 Fe I1 238204 A1 I 396 151 Mg I1 279 553 1 3.14 5.25 0.13 0.75 2.38 1.63 2.38 0.25 2 2.13 5.00 0.09 0.38 2.38 1.05 2.38 0.02 Table 5 Detection limits (30 criterion) of the FAAS determination of the impurity elements with respect to P-barium borate Element Na Fe A1 Mg Detection limit/pg g-' 1.7 10.0 150 0.35 The effect of the P-barium borate matrix on the analytical signal of the elements was studied.To this purpose the ratio between the slope of the curve of standard additions in the 0.8% m/v solution of P-barium borate and that of an aqueous calibration curve was determined (see Table 6). Since the content of A1 in P-barium borate was found to be below its detection limit in flame AAS it was spiked to the test solutions at a level of l o x the detection limit.As can be seen the matrix P-barium borate significantly suppresses the analytical signal the slope ratio is less than unity for all analytes examined. This imposes the use of standard additions for adequate calibration. Analysis of Single Crystals of /?-Barium Borate Single crystals of undoped P-barium borate and P-barium borate doped with Ce Nd Eu and Er respectively were analysed for their dopant and impurity content after dissolution according to the procedure. The concentration of dopants was determined by ICP-AES. The results are presented in Table 7. The mean of three separate dissolutions is shown. The relative standard deviation for Er is 2.6%. It follows from the results Table6 Slope ratio between the curve of standard additions in a 0.8% m/v solution of P-barium borate in 2 mol 1-' HCl and the calibration curve in 2 mol 1-' HCl Element Na Fe Mi3 A1 Slope ratio 0.85 0.70 0.77 0.86 Table 7 Content of dopants in the optical crystals of P -barium borate Added to the initial Found in the RSD Dopant solid solution (% m/m) crystal (% m/m) (%) Ce 0.62 - Nd 0.64 Eu 0.68 Er 0.75 0.39 2.6 - * - - - - ~~ * Below the corresponding detection limit of the ICP-AES determi- nation (see Table 4).Journal of Analytical Atomic Spectrometry August 1996 VoI. 1 1 569Table8 Content of impurities in the optical crystals of /&barium borate ICP-AES FAAS Content/ RSD Content/ RSD Impurity P8 g-' W) P8 g-' (%) Na 460.0 2.0 489.0 2.5 Fe 36.0 2.5 35.0 2.9 A1 59.0 4.2 * - - Mg 7.0 5.9 5.9 5.5 * Below the detection limit of the FAAS determination (see Table 5).obtained that the only successful doping of @-barium borate achieved was that with Er. No appreciable amounts of Ce Nd or Eu have been incorporated in the crystals. The content of impurities in the single crystals of /?-barium borate was determined by ICP-AES and FAAS. The results obtained for the undoped sample are shown in Table 8. Each value is a mean of three separate dissolutions. The relative standard deviation of all results obtained by the two methods lies within 2-6%. The content of impurities in the doped crystals of @-barium borate was found to be at the same level as that shown for the undoped sample (Table8). Hence the impurity content primarily depends on the purity of the starting substances and of the environment during the synthesis (e.g.furnace utensils ambient air). The financial support from the National Fund for Scientific Research of the Ministry of Science Education and Technology of Bulgaria under registration no. x-335 is gratefully acknowledged. REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 Feigelson R. S. Raymakers R. J. and Route R.K. Prog. Cryst. Growth Charact. 1990 20 115. Miyazaki K. Sakie H. and Sato T. Opt. Lett. 1986 11 797. Laudise R. A. in Crystal Growth and Characterization ed. Ueda R. and Mullin J. B. North-Holland Amsterdam 1975. Morris P. A. J. Crystal Growth 1990 106 76. Daskalova N. Velichkov S. Slavova P. Ivanova E. and Aleksieva L. Spectrochim. Acta Part B submitted for publication. Mermet J. M. in Inductively Coupled Plasma Emission Spectroscopy Part 2 Applications and Fundamentals ed. Boumans P. W. J. M. Wiley New York 1987 ch. 10 p. 353. Daskalova N. Velichkov S. Krasnobaeva N. Slavova P. Spectrochim. Acta Electronica included in Spectrochim. Acta Part B 1992 47 E1595. Boumans P. W . J . M. Fresenius' Z . Anal. Chem. 1986 324 397. Boumans P. W. J . M. in Inductively Coupled Plasma Emission Spectroscopy Part I Methodology Instrumentation and Performance ed. Boumans P. W. J. M. Wiley New York 1987 p. 358. Harrison G. R. MIT Wavelength Tables The MIT Press Cambridge MA USA 1969 p. 1939. Boumans P. W . J . M. Line Coincidence Table for Inductively Coupled Plasma Emission Spectrometry Pergamon Press Oxford 1980. Boumans P. W. J. M. and Vrakking J. J . A. M. Spectrochim. Acta Part B 1987 42 819. Paper 6/01 980C Received March 21 1996 Accepted May 22 1996 570 Journal of Analytical Atomic Spectrometry August 1996 Vol. 11