Comparative Study of Several Nebulizers in Inductively Coupled Plasma Atomic Emission Spectrometry: Low-pressure versus High-pressure Nebulization JUAN MORA, JOSE� L. TODOLI�, ANTONIO CANALS* AND VICENTE HERNANDIS Departamento de Quý�mica Analý�tica, Universidad de Alicante, 03071 Alicante, Spain Five nebulizers for use in ICP-AES were compared. Two of Thus, pneumatic nebulizers are those in which aerosol is generated by the strip and subsequent break-up of the liga- them work at low pressure, a Meinhard and a V-groove nebulizer (VGN), and three at high pressure, a single-bore ments generated from a liquid bulk when it is exposed to a high-velocity gas stream.2,3 Although these nebulizers are very high-pressure pneumatic nebulizer (SBHPPN), a hydraulic high-pressure nebulizer and a thermospray (TN).The common, in particular the concentric type, they suffer from some drawbacks that limit their use as liquid sample introduc- comparison was made using three solvents, water, ethanol and butan-1-ol, using the sample uptake rate (Q l ) as a variable and tion systems (e.g., low analyte transport rates, tendency to become clogged).1 Because of this, nebulizers have been devel- studying its influence on drop size distribution, analyte transport rate and analytical behaviour, i.e., emission intensity oped in which liquid and gas streams interact more efficiently.For some of them, the nebulization principle is not pneumatic, and limits of detection (LODs).The sample introduction system includes a desolvation unit. The Sauter mean diameters whereas for the rest the interaction between the liquid and gas streams has been improved by reducing the cross-sectional of the primary aerosols generated by the high-pressure nebulizers (HPNs) are between 1.5 and 5.8 times lower than area of the gas outlet4–7 and/or the width of the liquid conduction walls.6 Nevertheless, a lower gas section implies a those generated by the low-pressure nebulizers (LPNs), this reduction being more noticeable at high liquid flow rates. In higher gas pressure to keep the gas flow constant.Therefore, gas has to be supplied at a pressure higher than usual, and addition, at high liquid flow rates, HPNs achieve higher analyte transport rates (between 2.4 and 19 times higher), connecting lines and apparatus should withstand this pressure. These requirements increase the cost of the nebulizer and make higher emission signals (up to 1.8 times for methanol and up to 4.5 times for water, using the Mn II 257.610 nm line) and it more difficult to use.Recently, a pneumatic nebulizer which works at high gas lower LODs for nine elements than the LPNs. Among HPNs, the SBHPPN gives rise to the best results at low Q l (i.e., and liquid pressures [single-bore high-pressure pneumatic nebulizer (SBHPPN)] was developed in our laboratories.8 The 0.6 ml min-1), whereas at high Q l (i.e., 1.2 ml min-1) the results are similar for all three HPNs when using methanol SBHPPN affords finer primary aerosols, higher analyte transport rates to the atomization cell, higher sensitivities and lower and butan-1-ol.With water, at high Q l, the TN gives the best results. For all the nebulizers tested, organic solvents limits of detection (LODs) than a conventional pneumatic concentric nebulizer (Meinhard type). So far the SBHPPN has (methanol and butan-1-ol ) provide better results than water, the relative improvement being more important for LPNs been applied in FAAS,8 ICP-AES9 and ICP-MS.10 The so-called thermospray (TN) is a kind of nebuli- (e.g., with VGN at 1.2 ml min-1, the improvement with methanol over water for Mn II is around sixfold) than for zer that offers many advantages, although some limitations, in comparison with pneumatic nebulizers.11,12 This nebulizer can HPNs (e.g., when SBHPPN is used at 1.2 ml min-1 for Mn II this improvement is 4.5-fold).be considered pneumatic in nature, since nebulization takes place by interaction between the liquid stream and a gas Keywords: Inductively coupled plasma atomic emission stream generated through the evaporation of a fraction of the spectrometry ; pneumatic nebulizers; thermospray; hydraulic solvent.13 Recently, the fundamental processes of thermal nebu- nebulizer ; high-pressure nebulization ; drop size distribution ; lization have been studied experimentally.14 As with SBHPPN, analyte transport rate; desolvation the gas and liquid streams emerge from the nebulizer through a single bore.The TN has been widely applied in FAAS,14,15 In recent years, much effort has been dedicated to liquid ICP-AES11,13,16 and ICP-MS.17 The LODs achieved with the sample introduction in atomic spectrometry as a means of thermospray are similar to those obtained with an ultrasonic improving the analytical response. Most of the research has nebulizer.16 been devoted to the development of new and more efficient The hydraulic high-pressure nebulizer (HHPN) is another systems of aerosol generation and transport.1 highly efficient nebulizer which generates the aerosol by making The aerosol generation process (i.e., nebulization) requires a high-speed liquid stream impact against a solid surface (cloud the supply of energy to a liquid bulk by means of a nebulizer.converter). Its analytical behaviour has been tested in FAAS,18 The first step is the generation of some wave-like perturbances ICP-AES19 and ICP-MS,20 and compares favourably with that on the liquid’s surface.The growth and subsequent break-up of a pneumatic concentric nebulizer. of these waves give rise to the droplets of the final liquid Operation of the last three nebulizers (SBHPPN, TN and aerosol.2 The characteristics of these aerosols are very depen- HHPN) requires the use of HPLC pumps and pressuredent on the amount of available energy and on the efficiency resistant transfer lines.For this reason, they will be referred of the energy transfer, whatever the kind of energy employed to as high-pressure nebulizers (HPNs) throughout the rest (kinetic, thermal, acoustic, etc.). of the paper. Accordingly, the conventional pneumatic Usually, nebulizers have been classified on the basis of the nebulizers used in this study will be collectively referred to as low-pressure nebulizers (LPNs). type of energy employed in the break-up of the liquid stream.Journal of Analytical Atomic Spectrometry, April 1997, Vol. 12 (445–451) 445Table 1 Main dimensions of the nebulizers and working conditions In general terms, the analytical performance of HPNs is superior to that of LPNs. However, HPNs also suffer from Nebulizer — some drawbacks: (i) they are more expensive; (ii) they are MN Meinhard TR-30-A3 more difficult to use; and (iii) they require a desolvation unit Gas outlet cross-sectional area: to avoid the negative effects of a too high solvent load to the 2.83×10-2 mm2 Liquid outlet cross-sectional atomizer.The last problem is more important when working area: 12×10-2 mm2 with organic solvents.9,14 VGN Gas outlet cross-sectional area: So far, most HPN evaluation studies have been carried out 5.66×10-2 mm2 by comparing them one by one with conventional pneumatic TN Outlet cross-sectional area: nebulizers, such as concentric, cross-flow and V-groove 1.27×10-2 mm2 types.9,10,16,19,20 Fraction of solvent vaporized: ~90%14 This study was aimed at evaluating the behaviour in ICPHHPN Outlet cross-sectional area: AES of the three above-mentioned HPNs and two LPNs with 0.31×10-3 mm2 water, methanol and butan-1-ol as solvents.The sample uptake Nebulizer tip to cloud converter rate (Ql) was used as an independent variable and the mean gap: 15mm drop size of the primary aerosol, the analyte transport rate to SBHPPN Outlet cross-sectional area: the atomizer, the emission signal and the LOD were used as 5.92×10-3 mm2 experimental parameters for comparison.To the authors’ Desolvation system — knowledge, this is the first attempt to compare the HPNs with Temperature of the heating 180 each other, working under the same set of experimental unit/°C conditions. Temperatures of the condensation units/°C First step 20 EXPERIMENTAL Second step 0 Five nebulie used, a Meinhard-type pneumatic concen- Plasma operating conditions — tric nebulizer (Meinhard, Santa Ana, CA, USA) (MN) and a Incident power/kW 1.0 V-groove nebulizer (Varian, N.Springvale, Australia) (VGN) Reflected power/W <5 as LPNs, and a laboratory-made TN, an HHPN (Knauer, Integration time/s 0.2 Berlin, Germany) and a laboratory-made SBHPPN as HPNs. Outer gas flow rate/l min-1 16.0 Intermediate gas flow rate/ 1.7 Table 1 shows the most relevant dimensions of each nebulizer. l min-1 The solution was fed at sample uptake rates between 0.6 and Nebulizer – carrier gas flow 0.33 1.2 ml min-1 by means of an Iso-Chrom HPLC pump rate/l min-1 (Spectra-Physics, San Jose, CA, USA) equipped with a pulse Observation height (mm 7 damper (a stainless-steel capillary, 0.50 mm id×1.59 mm above load coil)* od×1.50 m) placed at the outlet of the pump.The nebulizer– Torch Fassel type (4 mm id) Sample uptake rate Variable carrier gas flow was kept constant throughout at 0.33 l min-1 by means of a Model FC260 mass flow controller * Optimized for Mn II line and for all the nebulizers and solvents (Tylan, Torrance, CA, USA).This gas flow was the optimum tested at Qg=0.33 l min-1. in terms of emission signal for all the conditions studied. All the reagents and solvents were of analytical-reagent grade. Table 2 lists the most significant physical properties of Table 2 Physical properties of the solvents (20 °C) the solvents used. Solvent s*/dyn cm-1 g†/cP a‡ w§ All measurements were made in triplicate.Water 70.4 1.00 0.08 1700 Drop size distributions (DSDs) of the primary aerosols were Methanol 22.7 0.60 1.00 693 measured by means of a Model 2600c laser Fraunhofer diffrac- Butan-1-ol 22.8 2.38 0.11 502 tion system (Malvern Instruments, Malvern, Worcestershire, UK). The measurement of the DSDs was made at a down- * Surface tension. stream distance of 11 mm from the nebulizer tip, or from the † Viscosity. back of the cloud converter in the case of the HHPN. In this ‡ Relative volatility, defined as the ratio of liquid volume of solvent to the liquid volume of methanol necessary to saturate a given case, the measurement position was varied vertically so that empty volume.21 the laser beam would intercept the maximum aerosol concen- § Expansion factor, defined as the volume of gas produced by the tration zone.A lens with a focal length of 63 mm was used, evaporation of a unit volume of liquid solvent at its boiling which enabled the system to measure droplets with diameters temperature.14 between 1.2 and 118 mm.The software employed was version B.0D.22 The calculations to transform the energy distribution into size distribution were made using a model-independent temperatures. A Model F3-K thermostated bath (Haake, Karlsruhe, Germany) was used to control the temperature of algorithm that does not preclude any particular distribution function. the second condenser. The desolvation conditions are given in Table 1.The fraction of solvent vaporized inside the TN (Fv) was estimated by using the volume concentration (VC) value of the The analyte transport rate (Wtot) was measured by nebulizing a solution of Mn (100 mg ml-1) and trapping the aerosol at primary aerosol22 as described in a previous paper.14 All the nebulizers were coupled to a desolvation system the exit of the desolvation unit with glass-fibre filters (Type A/E, 47 mm diameter) of 0.3 mm pore size (Gelman, Ann Arbor, which consisted of a heating unit and a condensation unit.Fig. 1 shows a schematic diagram of the desolvation system MI, USA). Collected aerosols were washed out from the filters into calibrated flasks with 1.0% (m/m) hot nitric acid. The used. The first unit was a heating tape (265 W and 15.0 cm length; J. P. Selecta, Barcelona, Spain) coiled around a single- analyte content in each calibrated flask was determined by ICP-AES. One might question whether the analyte is com- pass spray chamber.Temperature was controlled by means of a contact thermometer. The vapour condensation unit featured pletely collected in a filter such as this, placed at the exit of the desolvation unit. Obviously, dry particles smaller than two Liebig condensers (30 cm×1.5 cm id) kept at different 446 Journal of Analytical Atomic Spectrometry, April 1997, Vol. 12driven and water cooled, with an operating frequency of 40.68 MHz. The rf power was computer controlled with an automatic matching network.The spectrometer was controlled by an IBM PC. RESULTS AND DISCUSSION Characteristics of the Primary Aerosols Aerosol characteristics determine, to a large extent, transport rate and analytical behaviour in atomic spectrometry. This section describes the influence of the sample uptake rate on the properties of the aerosols generated by the different nebulizers. The nebulizers tested have very different aerosol generation mechanisms. Some of the variables are specific (e.g., the fraction of solvent evaporated for the TN and the nozzle–cloud converter gap for the HHPN).In this study, the specific variables were kept constant for each nebulizer at values close to the optimum. The sample uptake rate is the only variable which is common for all nebulizers. Fig. 1 Schematic diagram of the desolvation system. The Sauter mean diameter (D3,2) was chosen to describe the central tendency of the DSD since it is the most commonly 0.3 mm would pass across the filter.However, this situation used diameter. In addition, other parameters such as aerosol does not apply here, where the aerosol leaving the condensation velocity, cone angle and nebulization yield were observed unit is never completely dry, since owing to the nucleation visually. processes that take place on cooling, the particles do contain liquid. This nucleation effect has two important consequences. Sauter mean diameter of primary drop size distributions First, particles that would be completely dry at the exit of the Fig. 2 shows the variation of D3,2 with Ql for each nebulizer heating unit, and the size of which would be smaller than and solvent.Two well differentiated behaviours appear. First, 0.3 mm at this point, would increase their size in the conden- for TN and HHPN, D3,2 decreases as Ql is increased. This sation unit and be mostly retained by the filter. Second, the effect can be accounted for by the concomitant increase in the filter is wetted when these drops are retained, thus reducing available energy for nebulization.Thus, when working with the pore size. Therefore, the effective pore size would be smaller the TN, an increase in Ql leads to a similar increase in the than the nominal value of 0.3 mm. In addition, it has been nebulization gas flow rate, since the fraction of solvent evapor- checked experimentally, by measuring the DSDs of aerosols ated, Fv, is almost independent of Ql.For instance, when leaving the desolvation system, that the most important frac- nebulizing water under the experimental conditions employed, tion of the aerosol volume is contained in particles between the volumetric gas flow rate increases from 0.92 l min-1 at 1.2 and 10 mm, so that the analyte mass fraction contained in Q1=0.6 ml min-1 to 1.84 l min-1 at Q1=1.2 ml min-1.14 For droplets the diameter of which is smaller than 0.3 mm would the HHPN, a similar increase in Ql multiplies by a factor of be negligible in comparison with the analyte contained in four the kinetic energy available for the nebulization process.droplets larger than 0.3 mm. For the other nebulizers, D3,2 is almost independent of Ql. The experiments on the analytical behaviour in ICP-AES Working with the SBHPPN requires the gas pressure to be were performed with a solution containing 1 mg ml-1 of a total slightly increased as the sample uptake rate is increased, so as of nine elements, prepared from a 1000 mg ml-1 reference to keep gas flow constant.Therefore, the energy of the gas solution (ICP multielemental standard solution IV; Merck, stream also increases, as Ql is increased.8,9 However, this gain Darmstadt, Germany). Table 3 lists the elements, wavelengths in energy is much lower for SBHPPN than for TN and HHPN, and slit widths employed. A Model 2070 ICP-AES instrument so that D3,2 scarcely varies with Ql for the former.For LPNs, made by Baird (Bedford, MA, USA) was used and operated since the energy of the gas stream does not depend on Ql , the under the conditions given in Table 1. An air-vacuum path, kinetic energy per unit mass of liquid decreases as Ql is 1 mfocal length Czerny–Turner monochromator with a grating increased, so that the primary aerosol is slightly coarser. of 1800 grooves mm-1 blazed at 400 nm was employed. The In addition, Fig. 2 shows that the extent of the variation of scanning speed was 400 nm s-1 and the bandpass was 0.01 nm D3,2 with Ql for a given nebulizer is a function of the solvent full width at half-maximum.Detection was effected by means used. Thus, for TN and water, D3,2 decreases from 8.27 to of two photomultipliers, one for the 160–290 and the other for 1.68 mm when Ql is increased from 0.6 to 1.2 ml min-1 (i.e., a the 290–900 nm range. The viewing height was controlled 79.7% decrease). However, with methanol or butan-1-ol, manually.The rf generator was crystal controlled, solid-state D3,2 decreases by about 42%. Further, for TN, at Ql = 1.2 ml min-1, D3,2 is about the same for all three solvents. Table 3 Elements, wavelengths and slit widths used This result can be accounted for by taking into account that, in thermal nebulization, the energy available for nebulization Element line Wavelength/nm Slit width*/nm depends not only on Fv and Ql, but also on the so-called Mn II 257.610 0.2 expansion factor (w) of the solvent.14 Water is the solvent with Ag I 328.068 0.3 the highest w value (Table 2), hence its volumetric gas flow Cd I 228.802 0.1 Co II 228.616 0.2 rate is much higher than that of methanol and butan-1-ol. Cr II 283.563 0.2 This higher energy of water vapour counterbalances the nega- Cu I 324.754 0.4 tive effect of its higher surface tension.14 HHPN gives rise to Fe II 259.940 0.2 similar D3,2 decreases for all three solvents (from 44 to 54%) Zn I 213.856 0.3 when Ql is increased from 0.6 to 1.2 ml min-1.It is also worth Ni II 221.647 0.1 noting that butan-1-ol affords, with the HHPN, higher D3,2 values than water and methanol. This can be explained in * Optimized for each element. Journal of Analytical Atomic Spectrometry, April 1997, Vol. 12 447Fig. 2 Variation of the Sauter mean diameter, D3,2, with the sample uptake rate, Ql, for all the nebulizers tested. A, SBHPPN; B, TN; C, HHPN; D, VGN; and E, MN. (a) Water; (b) methanol; and (c) butan-1-ol.Error bars represent the range between the lower and upper values. terms of the high viscosity of butan-1-ol, since in hydraulic a dense cloud moving slowly and carried by the gas stream, so that coalescence losses and gravitational settling will appear. nebulization viscosity contributes to damping the disturbances that appear on the liquid’s surface before and when the liquid In addition, one should take into account that the nebulization yield is not 100% for the HHPN (i.e., not all the liquid flow stream strikes the cloud converter.2 With the pneumatic nebulizers, SBHPPN and LPNs, the role of the viscosity is less is nebulized). A significant fraction (usually 20–50%) of this liquid flow goes to drain at the cloud converter.23 Although significant.Therefore, methanol and butan-1-ol afford lower D3,2 values than water, owing to their lower surface tension the aerosol generated by the HHPN is fine and slow, the analyte transport rate is mainly determined by the aerosol yield.(Table 2).5,9 It is apparent from Fig. 2 that HPNs generate finer aerosols The TN requires some specific comments, since most of the liquid is vaporized (in our case, 90% of the initial liquid than LPNs. Among HPNs, the SBHPPN provides the smallest D3,2 values at low flow rates (e.g., 0.6 ml min-1), whereas the volume is converted into vapour). This causes the analyte concentration to increase with respect to the initial level.TN performs the best at high Ql values (e.g., 1.2 ml min-1). These results can be explained by the energy available for Moreover, as the aerosols leaving the TN are at high temperature, evaporation will take place fairly rapidly. Nevertheless, nebulization in each case. As regards the solvent, Fig. 2 shows that when water is used, at Ql=1.2 ml min-1, the SBHPPN the fact that solvent condensation will be favoured since the solvent vapour expansion and evaporation itself are endo- gives higher D3,2 than the TN, whereas when using methanol or butan-1-ol the D3,2 values obtained with the SBHPPN are thermicprocesses should be taken into account.As evaporation depends on drop size, the smallest droplets will evaporate closer to those of the TN.From these results, it can be concluded that the SBHPPN more quickly than the coarsest droplets, so that an irregular distribution of the analyte into the aerosol is to be expected. summarizes the characteristics of high-pressure nebulization (low D3,2 values) and pneumatic nebulization (finer aerosols for organic solvents than for water).Analyte Transport Rate Fig. 3 shows the influence of Ql on the analyte transport rate, Other dynamic characteristics of the aerosols Wtot. The general trend is that Wtot increases as Ql is increased. This behaviour is inversely related to the variation of the Some other characteristics are also important in order to describe the aerosol (e.g., aerosol velocity, cone angle and aerosol mean drop size as Ql is increased (Fig. 2). Thus, in general, a marked decrease in D3,2 corresponds to a noticeable nebulization yield) since they have a noticeable influence on the transport variables. This section includes a brief description increase in Wtot, whereas slight variations in D3.2 give rise to slight variations in Wtot . In addition to the mean drop size, of these characteristics, based on visual inspection of the aerosols.the nebulization yield must also be taken into consideration, for the HHPN, since an increase in Ql causes the nebulization The aerosol velocity determines, to a large extent, the analyte transport rate. If the aerosol velocity is too high, an important yield to increase (e.g., for water it increases from 24.8 to 46.0% when Ql is increased from 0.6 to 1.2 ml min-1). fraction of the finest droplets will be lost by inertial impact against the bottom walls of the spray chamber.If this velocity As expected from the drop size results (Fig. 2), Fig. 3 shows that under most conditions, HPNs provide higher Wtot values is too low, droplets will be lost by gravitational settling.4 Moreover, when a desolvation unit is used, as in our case, than LPNs. At low sample uptake rates, the SBHPPN gives the highest Wtot values with all the solvents studied, whereas aerosol heating is more efficient if the aerosol velocity is low. With the SBHPPN and the TN, the aerosol velocity is higher with water and butan-1-ol, at high Ql, the TN is the nebulizer that performs the best.For the SBHPPN at 0.6 ml min-1, than with the LPNs, whereas the HHPN generates the slowest aerosols. analyte transport efficiencies are between 21 and 32% for water and methanol, respectively, whereas for LPNs these The cone angle of the aerosol is related to its velocity. The higher the velocity of the aerosol, the sharper is its cone angle. values are 7 and 20%.On the other hand, with the TN at 1.2 ml min-1, the analyte transport efficiencies are between 20 A large cone angle contributes to an increase in the amount of aerosol lost by impact against the side walls of the spray and 25% for water and methanol, respectively, and with the SBHPPN 11 and 30%. These values are between 1.5 and chamber.4 In agreement with their aerosol velocity, the SBHPPN and the TN generate aerosols the cones of which 12.5% for LPNs. As regards the solvents, it can be seen from Fig. 3 that, in are sharper than those corresponding to the LPNs. In the case of the HHPN, aerosol is not conical in shape but is rather like general, organic solvents afford higher Wtot than water.9,14,21 448 Journal of Analytical Atomic Spectrometry, April 1997, Vol. 12Fig. 3 Variation of the analyte transport rate, Wtot, with the sample uptake rate, Ql, for all the nebulizers tested. A, SBHPPN; B, TN; C, HHPN; D, VGN; and E, MN. (a) Water; (b) methanol; and (c) butan-1-ol.Error bars represent the range between the lower and upper values. Methanol is the solvent which shows the highest Wtot values and HHPN, whereas at Ql=1.2 ml min-1 the improvements in Wtot are more important for LPNs than for HPNs. in all the cases. Table 4 shows the enhancement ofWtot achieved on switching from water to methanol or butan-1-ol, (Wtot)rel. It can be seen that (Wtot)rel decreases when Ql increases for HHPN and TN whereas it increases for the other nebulizers Analytical Behaviour (i.e., pneumatic).This reveals the different aerosol generation Emission intensity mechanisms. For the five nebulizers, methanol provides higher (Wtot)rel Fig. 4 shows the variation of the net emission intensity (Inet) of Mn in ICP-AES versus Ql for each nebulizer using (a) water, values than does butan-1-ol. This is accounted for by the lower D3,2 values obtained with methanol and its higher relative (b) methanol and (c) butan-1-ol as solvents. It can be seen that the variations of Inet are concomitant with those of Wtot volatility (Table 2).As regards the nebulizer, it appears that at a Ql of 0.6 ml min-1 the largest improvements in (Wtot)rel (Fig. 3). Hence, in most cases, HPNs provide higher Inet values than do LPNs. The SBHPPN gives the highest Inet values at caused by switching from water to alcohols correspond to TN Table 4 Analyte transport rates for methanol and butan-1-ol relative to water for the different nebulizers and sample uptake rates tested (Wtot)rel* Ql/ml min-1 Solvent MN VGN TN HHPN SBHPPN 0.6 Methanol 2.89 3.00 3.44 6.18 1.54 0.8 Methanol 4.69 5.33 2.15 2.21 2.33 1.0 Methanol 5.73 6.70 1.74 1.89 2.61 1.2 Methanol 9.09 12.65 1.22 1.31 2.63 0.6 Butan-1-ol 1.35 1.32 1.86 2.10 1.41 0.8 Butan-1-ol 2.38 2.43 1.26 1.17 1.64 1.0 Butan-1-ol 3.89 4.36 0.88 1.16 1.87 1.2 Butan-1-ol 6.32 7.60 0.64 0.78 2.05 * (Wtot )rel=(Wtot)solvent i/(Wtot)water. Fig. 4 Variation of the net emission intensity for manganese, Inet , with the sample uptake rate, Ql, for all the nebulizers tested.A, SBHPPN; B, TN; C, HHPN; D, VGN; and E, MN. (a)Water; (b) methanol; and (c) butan-1-ol. Error bars represent the range betweenthe lower and upper values. Journal of Analytical Atomic Spectrometry, April 1997, Vol. 12 449low Ql for all the solvents. However, at higher sample uptake obtained with the three HPNs are similar, perhaps the LODs for water with the TN [Fig. 6(a)] are slightly lower than with rates the relative behaviour of the HPNs depends mainly on the solvent. Thus, the TN gives the highest Inet values for the HHPN and the SBHPPN, and lower than those obtained with LPNs. water [Fig. 4(a)] and butan-1-ol [Fig. 4(c)], and the SBHPPN for methanol [Fig. 4(b)]. As regards the solvent, again in good agreement with the transport results (Fig. 3), the Inet obtained with methanol are CONCLUSIONS higher than those obtained with butan-1-ol and water.As High-pressure nebulization is a good means for liquid sample occurs with Wtot, switching from water to alcohols leads to an introduction in ICP-AES. The analytical figures of merit are increase in Inet that is relatively more important for LPNs usually better than those obtained using the conventional low- than for HPNs. Thus, the Inet values obtained with methanol pressure pneumatic nebulizers. HPNs generate finer aerosols are around six times higher than those for water when MN than do LPNs, thus giving rise to higher Wtot, higher Inet and and VGN are used.In the case of HPNs this factor is up to lower LODs. However, HPNs are more difficult to use and 4.5 for the SBHPPN. more demanding in terms of requirements. The relative behaviour shown by the HPNs depends on the L imits of detection (L ODs) solvent and Ql employed. At low flow rates, the SBHPPN provides the highest sensitivity and lowest LODs, whereas at Figs. 5 and 6 show the LODs for nine elements at Ql values high Ql with water, the TN performs the best.The results of 0.6 ml min-1 (Fig. 5) and 1.2 ml min-1 (Fig. 6) for each obtained with HPNs could be improved by reducing the outlet solvent and nebulizer tested. As expected from the results for section area of the nebulizer tip. Inet (Fig. 4), the LODs follow the order water>butan- Provided that a desolvation system is employed, switching 1-ol>methanol. This behaviour is clearer for 1.2 ml min-1 from water to alcohols improves the drop size distribution of (Fig. 6) than for 0.6 ml min-1 (Fig. 5). As regards the nebulizer the aerosol, the analyte transport and the emission intensity tested, Figs. 5 and 6 show that the improvement in LOD for all the nebulizers tested. introduced by HPNs depends on the value of Ql. Thus, at Ql=0.6 ml min-1 (Fig. 5) the SBHPPN gives the lowest LOD, in agreement with the corresponding Inet values (Fig. 4), especi- The authors thank the DGICyT (Spain) for financial support (Project PB92-0336).ally for water [Fig. 5(a)]. At 1.2 ml min-1 (Fig. 6), the LODs Fig. 5 Limits of detection in ICP-AES with the different nebulizers evaluated, calculated according to the 3sb criterion, sb being the standard deviation obtained from 20 replicates of the blank. (a) Water; (b) methanol; and (c) butan-1-ol. Ql=0.6 ml min-1. 450 Journal of Analytical Atomic Spectrometry, April 1997, Vol. 12Fig. 6 Limits of detection obtained in ICP-AES with the different nebulizers evaluated, calculated according to the 3sb criterion, sb being the standard deviation obtained from 20 replicates of the blank.(a) Water; (b) methanol; and (c) butan-1-ol. Ql=1.2 ml min-1. 13 Koropchak, J. A., and Winn,D. H., Appl. Spectrosc., 1987, 41, 1311. REFERENCES 14 Mora, J., Canals, A., and Hernandis, V., Spectrochim. Acta, Part 1 Sample Introduction in Atomic Spectroscopy, ed. Sneddon, J., B, 1996, 51, 1535. Elsevier, New York, 1990. 15 Robinson, J. W., and Choi, D. S., Spectrosc. L ett., 1987, 20, 375. 2 Atomization and Sprays, ed. Lefebvre, A. H., Hemisphere, New 16 Vermeiren, K. A., Taylor, P. D. P., and Dams, R., J. Anal. At. York, 1989. Spectrom., 1988, 3, 571. 3 Sharp, B. L., J. Anal. At. Spectrom., 1988, 3, 613. 17 Thomas, C., Jakubowski, N., Stu�wer, D., and Broekaert, J. A. C., 4 Canals, A., Hernandis, V., and Browner, R. F., Spectrochim. Acta, J. Anal. At. Spectrom., 1995, 10, 583. Part B, 1990, 45, 591. 18 Berndt, H., Fresenius’ Z. Anal. Chem., 1988, 331, 321. 5 Canals, A., Hernandis, V., and Browner, R. F., J. Anal. At. 19 Luo, S. K., andBerndt, H., Spectrochim. Acta, Part B, 1994, 49, 485. Spectrom., 1990, 5, 61. 20 Jakubowski, N., Feldmann, I., Stuewer, D., and Berndt, H., 6 Nixon, D. E., Spectrochim. Acta, Part B, 1993, 48, 447. Spectrochim. Acta, Part B, 1992, 47, 119. 7 Olesik, J. W., Kinzer, J. A., and Harkleroad, B., Anal. Chem., 21 Browner, R. F., Canals, A., and Hernandis, V., Spectrochim. Acta, 1994, 66, 2022. Part B, 1992, 47, 659. 8 Todolý�, J. L., Canals, A., and Hernandis, V., Spectrochim. Acta, 22 Instruction Manual, Issue 2.2, Version B.0D, Malvern Instruments, Part B, 1993, 48, 373. Malvern, 1991. 9 Todolý�, J. L., Canals, A., and Hernandis, V., J. Anal. At. Spectrom., 23 Cano, J. M., Todolý�, J. L., Canals, A., and Hernandis, V., paper 1996, 11, 949. presented at the XIV Reunio�n Nacional de Espectroscopý�a, Baeza 10 Hernandis, V., Todolý�, J. L., Canals, A., and Sala, J. V., (Spain), September 1994. Spectrochim. Acta, Part B, 1995, 50, 985. 11 Meyer, G. A., Roeck, J. S., and Vestal, M. L., ICP Inf. Newsl., Paper 6/06781F 1985, 10, 955. Received October 3, 1996 12 Koropchak, J. A., and Veber, M., Crit. Rev. Anal. Chem., 1992, 23, 113. Accepted January 2, 1997 Journal of Analytical Atomic Spectrometry, April 199