JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 199 1 VOL. 6 54 1 Laser Ablation in a Liquid Medium as a Technique for Solid Sampling Yasuo lida Akira Tsuge Yoshinori Uwamino Hisashi Morikawa and Toshio lshizuka Government Industrial Research Institute Nagoya 7 Hirate-cho Kita Nagoya 462 Japan A new technique for solid sampling laser ablation in a liquid medium has been developed and evaluated. A solid sample was held in a liquid medium contained in a flat-bottomed beaker. Laser pulses from a Q-switched Nd:YAG laser were introduced from the bottom of the beaker and were focused on the surface of the solid sample. Both the vapour and the particles produced by the laser ablation were trapped directly by the surrounding liquid and a suspension was formed. The technique offers both time and spatial separation of the sampling and introduction and excitation processes in the laser ablation.The morphological observation of trapped particles the evaluation of trapping efficiency and fractional ablation and the direct introduction of the suspension into an inductively coupled argon plasma were carried out. Nearly 100% trapping efficiency and the absence of fractional ablation were obtained. Keywords Laser ablation; solid sampling; suspension; metal and ceramic samples; liquid medium Laser ablation has been widely applied to the direct sampling and introduction of solid samples into inductively coupled plasmas (ICPS),'-~ microwave-induced plasma^,^^^ glow discharges8v9 and various other plasma sources for atomic spectrometry. In most instances an inert gas stream has been used as the carrier for the sample vapour and particles.However loss of analyte during transport between the ablation site and the excitation source has been discussed as a cause for the decrease in the accuracy and precision of this t e ~ h n i q u e . ~ ~ ~ ~ ~ ~ ~ Arrowsmith and Hughes'O have investigated the entrainment and transport of ablated particles in the gas flow to a secondary excitation source. Furthermore from the pulse-like evolution of the same vapour either a high-speed scanning or multichannel detection system is inevitably needed for multi-element quantification and background correction. In this study a mode of laser ablation for solid sampling is proposed i.e. laser ablation in a liquid medium (LALM). The technique provides separation between the sampling and introduction processes and the following merits emerge (1) temporal separation ie.the necessity of high- speed scanning or a multichannel detection system can be precluded; (2) spatial separation i.e. the laser can be positioned separately from the massive excitation and detection instruments; (3) solid standard samples become unnecessary if the direct introduction of a laser-generated suspension into an excitation plasma or the dissolution of the suspension is quantitatively achieved; (4) the loss of analyte on transport between the ablation cell and the plasma source can be eliminated; and ( 5 ) some insights concerning the laser ablation process such as fractional ablation (vaporization) can be obtained.Supposed demerits of LALM are the dilution effects by a liquid medium and the possibility of contamination in the sampling process as compared with the direct sampling and introduction technique. Two-step methods where the laser sampling and the sample introduction are separated by using graphite collectors have been reported by other groups.11.12 However in these approaches the trapping efficiency of the laser-ablated particles has not been sufficient and the graphite can be a source of contamination in ultratrace analysis. The merits of the proposed technique LALM have been ascertained in combination with ICP atomic emission spectrometry (ICP-AES) and some of the results viz. observation of the trapped particles by scanning electron microscopy (SEM) trapping efficiency and the direct introduction of the suspension into the ICP will be discussed later from an analytical point of view.Experimental Apparatus The specifications of the apparatus and the operating conditions are summarized in Table 1. The assembly for LALM is schematically represented in Fig. 1. A sample was positioned in a liquid medium (10 ml) contained in a flat-bottomed beaker (25 ml). Two methods of sample holding were employed firstly sticking the sample to the flat top of a quartz rod with double-sided adhesive tape; and secondly holding the sample at the top of a stainless- steel rod by suction through a bore with an O-ring seal. A Quick Connects vacuum-tight connector (Swagelok Part No. SS-QC4-B-400) was assembled at the other side of the rod in order to maintain a reduced pressure without suction.The latter holding mode was used in the sample Table 1 Apparatus and operating conditions Q-switched Nd:YAG laser- Supplier Model Energy Pulse width Repetition rate Wavelength Electronic rn icro balance- Quantel International 150 mJ per pulse 10 ns 10 Hz 1064 nm Y G-580A Supplier Mettler ME 30 Model Weighing range (electrical) 0-30 mg Capacity 1020 mg Reproducibility +. 1 Pg ICP atomic emission spectrometer- Supplier Model Monochrometer R.f. incident power Ar flow rate Nebulizer Scanning electron microscope- Seiko Instruments 100 cm focal length JY-38P I1 3600 grooves mm-l holographic grating 1.3 kW Coolant gas 16 1 min-l auxiliary gas 0.2 1 min-' carrier gas 0.4 1 min-' Glass concentric Supplier JEOL Accelerating voltage 7 kV Working distance 1 5 mm Model JSM-F7542 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 199 1 VOL.6 Fig. 1 Schematic diagram of the laser ablation assembly. A Flat- bottomed beaker; B sample; C quartz (or stainless-steel) rod for holding the sample which can be rotated by the action of a stepping motor; D single lens of 150 mm focal length; and E 45" prism mass measurement. The sample could be rotated at a speed of 50 rev min-l by a stepping motor to prevent the creation of a deep crater on the sample surface. The laser beam was focused with a single lens of 150 mm focal length on to the sample surface through the base of the flat-bottomed beaker. The spot size of the laser beam was adjusted to 1 mm. The introduction of the laser beam from the bottom of the beaker prevents the deleterious effects of surface waves which stem from the shock of laser ablation.Samples A brass sheet (Nilaco 62.3% Cu; and 37.7% Zn) of 1 mm thickness was cut into squares of 10 x 10 mm and polished with No. 800 silicon carbide paper (Refine Tec). Sintered zirconia (Tosoh Zr02 with 4% Y) was obtained as sheets of 10 x 10 x 1 mm. The zirconia sheets were washed with dilute hydrochloric acid solution for 1 h at room tempera- ture. Both samples were washed with distilled water and ethanol and dried with hot air. Elemental Analysis The elemental contents of the samples and suspensions were determined by ICP-AES. The brass sample (0.3 g) was dissolved in 10 ml of nitric acid (1 + 1) by heating on a hot-plate and then diluted.The brass suspension formed by the 2000 laser pulses taken into 10 ml of distilled water was dissolved by the addition of 1 ml of nitric acid (1 + l) on a hot-plate. The zirconia suspension formed in 10 ml of distilled water was trans- ferred into a poly(tetrafluoroethy1ene) vessel and evapor- ated to near dryness. After the addition of 10 ml of sulphuric acid (1 + 2) the vessel was set in a pressure bomb (San-ai Model NT-25) and heated at 230 "C for 24 h.13 SEM Samples A 5 pl volume of settled suspension was dried on a glass plate (5 x 5 x 1 mm) at room temperature. The plate was fixed on to a sample stage using a silver paste (Fujikura Chemical Dotite) and was then coated with gold. Results and Discussion SEM Observations Fig. 2 shows the trapped products of LALM of a brass sample in water.These are fine spherical particles typically less than 1 ,urn together with a small amount of amorphous aggregates. Thompson et al. l4 observed the substances ablated in a glass chamber with gas flow and subsequently collected on a filter with a 0.4 pm pore size. They reported that the particles collected on the filter represented 20-30% (by mass) of the ablated material and that there were a large number of spherical particles ranging in size from 10 pm to less than 1 pm. On comparing the results the smaller size of the particles and the large amount of amorphous material obtained by LALM is significant. It suggests that in LALM the ablated materials in the form of liquid or vapour have cooled more rapidly and been trapped by the surrounding liquid medium.The ablated material from zirconia placed in water is shown in Fig. 3. Distorted spheres of around 1 pm or less and crooked needles are observed. Less amorphous ma- terial is found in the ablated material from zirconia than that from brass. A spherical particle with a tail has also been observed by Thompson et all4 The morphological differ- ence between the ablated material from the metal brass and that from the ceramic material zirconia stems from the difference in the melting-p~intsl~ (brass 932; Cu 1083; Zn 420; and Zr02 about 2700 "C) and b~iling-pointsl~ (Cu 2567; Zn 907; and ZrOz about 5000 "C). Namely the amorphous materials are condensed directly from the sample vapour and the needles are solidified from the splashing of viscous drops of ablated material.* Fig. 2 Scanning electron microscope photograph of spherical particles and amorphous material produced by the laser ablation of brass in water 1 prn H Fig. 3 produced by the laser ablation of sintered zirconia in water Scanning electron microscope photograph of materialJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 1991 VOL. 6 543 Table 2 Trapping efficiency of brass samples ablated in water Amount trapped/mg Amount Sample ablated*/mg Cu Zn Total 1 0.273 0.163 0.099 0.267 2 0.243 0.151 0.090 0.241 3 0.264 0.162 0.098 0.260 4 0.257 0.158 0.095 0.253 5 0.253 0.155 0.093 0.248 6 0.267 0.164 0.099 0.263 Average 0.260 0.160 0.096 0.255 SD$ 0.01 1 0.006 0.004 0.010 * 2000 laser shots. t Cu:Zn in target brass samples were 1.65 f 0.02. $ Standard deviation.Recovery (O/o) 97.8 99.2 98.5 98.4 98.0 98.5 98.4 0.5 Cu:Znt 1.65 1.68 1.65 1.66 1.67 1.66 1.66 0.0 1 Trapping Efficiency and Fractional Ablation The trapping efficiency of LALM was tested with the brass samples. The efficiency was calculated from the mass lost on ablation by weighing the sample directly with the electronic microbalance and the mass trapped in the suspension by analysing the elemental content with ICP- AES after the dissolbtion with nitric acid. The results are shown in Table 2. A trapping efficiency of nearly 100% was achieved for LALM. Although the formation of fine bubbles were occasionally observed by video monitoring of the LALM process the results indicate that sufficient cooling and condensation of vapour were attained.The fractional vaporization occurring in the laser abla- tion process has been discussed previously.2J6 The Cu:Zn ratios in the suspensions were also determined and are shown in Table 2. The Cu:Zn ratio in the suspension agreed with that of the target brass sample and the existence of fractional ablation can be denied. Baldwin16 has collected the laser-ablated material from a brass sample fixed on a film and suspended 0.5 mm above the sample surface. The Cu:Zn ratio in the material collected (2.0) was significantly smaller than that in the target material (2.7). Baldwin16 also used Q-switched lasers hence the discrepancy between the results might indicate that elemental redistribution occurs among the vapour and particles according to the particle sizes and the vapour pressures of the elements.However the nearly 100% collection efficiency of ablated materials achieved with LALM offers a reliable method of sampling the target material whether an elemental redistribution exists or not. Direct Introduction of the Suspension into the ICP As the suspension obtained by using LALM consists of fine particles of around 1 pm or less and is stable over several hours it can be introduced directly into the ICP via the usual concentric nebulizer without any additional operations. The brass suspension used was dark brown and could be instantaneously changed into a clear solution by the addition of 2 p1 of nitric acid. The emission intensities of the Cu I (324.75 nm) and Cu I1 (224.70 nm) and Zn I (213.86 nm) and Zn I1 (206.19 nm) lines were compared between the suspensions and the acidified solutions.All of the lines showed a similar tendency in that the intensities of the suspensions were decreased to 60-70% of those of the solution. Because replacement of the nebulized sample of a suspension by dilute nitric acid solutions brought about a rapid rise and fall of the emission intensity the causes for the difference in the intensities were considered to be (i) absorption of suspended materials on the way to the nebulizer; and (ii) the difference in uptake efficiency in the spray chamber and plasma between the droplets formed from a solution and the particles from a suspension. The use of a specified nebulization system for slurry introduc- tion17J8 would minimize the difference in the intensities.Ebdon et al.17 reported that an alumina slurry of very fine particle size 100% of the particles less than 5 pm which can be easily obtained by LALM gave the same emission intensity as compared with solution introduction. The direct introduction of zirconia suspensions obtained by LALM was also investigated. The emission intensities of the Zr I1 line at 343.82 nm of the suspension were approximately 35% of those of the dissolved solution of the suspension. This decrease is more striking than that of the brass suspension. Similar results have been reported in the direct slurry introduction study. Long and Brennerl* showed that the refractory materials were difficult to atomize directly even in an ICP. Unlike the brass suspen- sions the zirconia suspensions were difficult to dissolve.The dissolved portion of the zirconia which was measured by filtering the suspension with a disposable syringe filter of 0.45 pm pore size were t l % both for the suspension obtained by LALM in a sulphuric acid (1 +2) medium and for the suspension with the addition of sulphuric acid after LALM in water. By using LALM in sulphuric acid (1 + 2) a decrease of about 40% in the amounts ablated versus that obtained by the LALM in water were brought about. This could be because of the difference in the viscosities of the media which suppress the expansion of the laser induced plasma into the liquid medium. From a practical point of view there exist some limita- tions for the combination of LALM and the direct introduc- tion of a suspension for ICP-AES because of the insufficient efficiency of the nebulization introduction step in an ICP which is usually 1% efficient or less.Compared with the gas- phase transfer in the usual laser ablation process aqueous sample introduction might cause solvent loading in the ICP and in ICP mass spectrometry higher intensities of molecular oxide species. Therefore other types of sample introduction techniques such as electrothermal vaporiza- tion or glow discharge sputtering,19 should be more advan- tageous for trace element determinations by LALM. Conclusions The merits of LALM have been experimentally ascertained i.e. the temporal and spatial separation of the sampling and introduction and excitation processes and the almost 100% trapping efficiency. The LALM process produces a suspen- sion consisting of fine particles of around 1 pm or less,544 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY OCTOBER 1991 VOL.6 which can be directly introduced into an ICP. Once the suspension is transformed into a solution the usual concentration or separation techniques can be adopted such as extraction ion exchange precipitation etc. which are impossible for gas-phase laser ablation. The demerit of LALM i.e. the dilution by the liquid medium can be overcome by the above mentioned tech- niques and/or by the use of highly sensitive instruments e.g. an ICP mass spectrometer. However at this stage the advantage of the spatial resolution in normal laser ablation is lost by the accumulated sampling in LALM which is needed in practice owing to the lower concentration level of the suspension obtained. The quantitative analysis of solid samples without solid standards is one of the merits of LALM but this objective has not been fully accomplished yet especially for ceramic materials.The techniques and devices developed for the slurry injection will improve the situation for ICP-AES with direct introduction of the suspension. Also the dissolution procedure if required should be easier for a suspension consisting of fine particles than for solid sample materials. References 1 Thompson M. Goulter J. E. and Sieper F. Analyst 1981 106 32. 2 Kawaguchi H. Xu J. Tanaka T. and Mizuike A. Bunseki Kagaku 1982,31 E185. 3 Ishizuka T. and Uwamino Y. Spectrochim. Acta Part B 1983 38 519. 4 5 6 7 8 9 10 11 16 17 18 19 Gray A. L. Analyst 1985 110 551. Arrowsmith P. Anal. Chem. 1987 59 1437. Leis F. and Laqua K. Spectrochim. Acta Part B 1978 33 727. Ishizuka T. and Uwamino Y. Anal. Chem. 1980 52 125. Iida Y. Spectrochim. Acta Part B 1990 45 427. Barshick C. M. and Harrison W. W. Microkim. Acta 1989 111 169. Arrowsmith P. and Hughes S. K. Appl. Spectrosc. 1988,42 1231. Rudnevsky N. K. Tumanova A. N. and Maximova E. V. Spectrochim. Acta Part B 1984 39 5. Wennrich R. and Dittrich K. Spectrochim. Acta Part B 1987 42 995. Ishizuka T. Uwamino Y. and Tsuge A. Bunseki Kagaku 1985 34 487. Thompson M. Chenery S. and Brett L. J. Anal. At. Spectrom. 1990 5 49. CRC Handbook of Chemistry and Physics ed. Weast R. C. CRC Press Boco Raton 1983. Baldwin J. M. Appl. Spectrosc. 1970 24 429 Ebdon L. Foulkes M. E. and Hill S. J. Anal. At. Spectrum. 1990 5 67. Long G. L. and Brenner I. B. J. Anal. At. Spectrom. 1990,5 495. Kitagawa K. Kanoh S. Ohta K. and Yanagisawa M. Anal. Sci. 1988 4 153. Paper 1/00 784J Received February 19th 1991 Accepted May 29th 1991