166 Analyst, February, 1983, VoE. 108, $9. 166-170 A Pneumatic Recirculating Nebuliser System for Small Sample Volumes Peter Hulmston Analytical Chemistry Branch, Chemistry Division, 11.300( PE) , Atomic Weapons Research Establishment, Aldermaston, Berkshire, RG7 4PR The nebuliser spray chamber described generates an aerosol, for use with an inductively coupled plasma, on as little as 1 ml of sample solution for over 10 min. The nebuliser has direct application to inductively coupled plasma optical emission spectroscopic analysis in those instances where economy of sample solution is important. The performance of the nebuliser spray chamber was tested and it was found that the sensitivity and precision are as good as those obtained with a conventional nebuliser system. Memory effects are readily overcome by including a system that allows rapid, thorough flushing and flooding of the spray chamber.Keywords : Inductively coupled plasma ; optical emission spectrometry ; pneumatic nebuliser The most common method of introducing sample solutions into an inductively coupled plasma (ICP) source is to generate a fine aerosol by pneumatic nebulisation. With con- ventional nebulisers, less than 5% of the solution actually enters the source, the remainder, which consists of a coarse spray, being trapped in the spray chamber and allowed to drain to waste. This poor efficiency in sample transfer to the plasma is normally of little importance as the amount of sample solution available is not limited, but there are occasions when the volume is limited by the amount of sample available or the amount that can be dissolved.Further, with the application of photographic recording1s2 and of sequential multi-element analysis3 there is a need to nebulise a sample into the plasma for long, uninterrupted periods of time. I t is obvious that a considerable gain in efficiency in sample use can be achieved by recovering the bulk of the solution that normally drains to waste and recycling it to the nebuliser. The use of pneumatic nebulisers where the waste solution is recovered is almost as old as spectroscopy itself4p5 and was particularly popular in flame photometry up to 30 years ago, after which their use declined, probably because the inconvenience of changing samples became more important as rapid monochromator and photoelectric detection systems took over from photographic recording.Many of the systems used metal components, most required a considerable volume of solution (10-20 ml) and all required a high gas flow-rate, which is not suitable with ICPs. Novak6 constructed a recirculating fixed cross-flow nebuliser for use with ICPs on the Rauterberg and Knippenberg design,’ but this requires more solution and its construction would require the service of a very high quality glassblower. No reference to a system using commercially available nebulisers using small volumes of sample for ICP could be identified and, to meet some of our current analytical requirements, a nebuliser design based on the Meinhard concentric system has been investigated. Design Requirements A number of factors were considered to be of importance in designing the iiebuliser spray chamber. Nebulisers for ICPs are made t o fine tolerance because of the limited volume of gas available for nebulisation and it could be expected that a consistent performance could be most easily achieved by using a commercial nebuliser such as the Meinhard glass concentric nebuliser.For our analytical requirements the maximum volume of sample available was 2 ml and to operate on this volume the hold-up of sample in the spray chamber should be small, as should be the drainage time. The spray chamber design should also allow for Crown Copyright.HULMSTON 167 rapid and effective flushing to eliminate contamination of a solution by a previous sample. Low-power argon - argon plasmas are adversely affected by the ingress of air and it would be necessary to prevent air from entering the spray chamber during sample changeover and to provide an uninterrupted flow of argon to the plasma to maintain its stability.Finally, the precision and sensitivity of the system should not be significantly worse than those of a con- ventional nebuliser spray chamber, and the system should be easy to operate routinely. Apparatus ICP Equipment The ICP source used was a Plasma-Therm system, Model HFP 2500D. The generator of this system was a crystal-controlled type operating a t 27.12 MHz with a maximum output of 2.5 kW. The plasma torch used was an integral Plasma-Thenn, Model T1.5. Solutions are introduced into the plasma using either the prototype recirculating nebuliser or the concentric nebuliser (taken from the recirculating nebuliser) in a conventional spray chamber.Radiation emitted from the plasma was focused on to the entrance slit of a 1-m Czerny- Turner monochromator (Spex) as a 1 : 1 image. The detection system employed consisted of an EHI 9862 photomultiplier in conjunction with a Spex DPC 2 digital photometer, which also supplied the EHT to the photomultiplier. Optimised conditions are listed in Table I. TABLE I OPTIMISED CONDITIONS FOR ICP R.F. power . . Viewing height Plasma gas . . Auxiliary gas Injector gas . . Entrance slit Exit slit . . DPC 2 range.. Integration time .. . . 1.0kW .. . . 23 mm above load coil .. . . 12lmin-l .. . . 1.2 lmin-1 .. . . 32 lb in-2 for both nebuliser systems .... 20pm .. .. 35pm .. . . 0.01-0.1 pA .. .. 5 s Prototype Recirculating Nebuliser Figs. 1 and 2 illustrate the design features of the prototype recirculating nebuliser. The return of waste solution to the nebuliser inlet is facilitated by moving the nebuliser and spray chamber vertically. The Meinhard concentric nebuliser operates efficiently in this position and is small enough to be readily accommodated within the spray chamber. The nebuliser sprays into a central inner tube, which traps the bulk of the coarse spray and the fine spray passes through a hole a t the top of this tube on its way to the plasma. Rapid drainage is assisted by trapping most of the coarse spray inside the central tube, by angling the bottom edge of this tube and by coating all the interior surfaces with silicone.Efficient cleaning of the system between sample nebulisation is achieved by filling the spray chamber completely with de-ionised water with argon passing through the nebuliser during the earlier stages of filling. One system, consisting of a drain valve and water inlet valve, is used for rapid flushing and flooding of the spray chamber. The second system is used for controlling the argon by-pass system, which allows the ICP source to be sustained during sample changeover. The following description of the operation of the system should be read in conjunction with Fig. 2. Sample introdzcction. A 1-ml volume of sample solution is injected by syringe into the nebuliser chamber. With only valve 1 open, the nebuliser will generate the required aerosol for the ICP source.Sample removal. After analysis, valves 3 and 4 are opened to operate the by-pass system and, by opening valve 5, the sample can be drained into a container and stored for future analysis if required. Flztshing andjooding. With the pump switched on and valve 2 open, de-ionised water enters the spray chamber and flushes out residual sample. Valve 5 is then closed to allow The purpose of the two sets of valve systems is as follows.168 HULMSTON : PNEUMATIC RECIRCULATING Analyst, Vol. 108 the chamber to fill with water. During the earlier stage of flooding the concentric nebuliser remains pressurised. This allows argon to bubble through the water thus improving the effectiveness of the washing process. Valve 1 is closed before the nebuliser chamber becomes completely flooded to avoid splashing solution into the plasma tube.With valve 2 closed and valve 5 open the system can be drained. The flushing, flooding and drainage procedure may be repeated to ensure total removal of the previous sample. With valve 5 closed and valve 1 open the next sample can then be injected into the nebuliser and by closing valves 3 and 4 normal nebulisation of the sample is restored. Fine aerosol p----- 1-ml syringe Screw-top and silicone IU septum lasses . ..I.......... 220 mm 28 mm diameter 35 mm diameter 83435 Quickfit joint concentric nebuliser 3-mm glass valve va Ive Fig. 1. Prototype recirculating nebuliser. Performance Nebulising Characteristics The nebulising characteristics of the recirculating and conventional nebulisers were com- pared by measuring the sensitivity, the limit of detection and the precision of each of the three elements, zirconium, titanium and copper.The sensitivity, i.e., the net signal above background for unit concentration of an impurity in solution, does not differ significantly between the two nebulisers, and the intensities of the background emissions are also the same. The limit of detection, defined as the concentration equivalent to twice the standard deviation of the background, and the precision, obtained by repetitive readings taken from one portion of standard, were calculated for each element and the results are shown in Table 11. I t is concluded that the amount and size of spray reaching the plasma from either nebuliser system are not sufficiently different to affect the performance of the ICP.Again, no significant difference between the two systems was observed.February, 1983 NEBULISER SYSTEM FOR SMALL SAMPLE VOLUMES To ICP source Valve 4 (fine adjustment valve) Argon in I J Fig. 2. Schematic representation systems used during operation. U showing valve 169 Memory Effects In the practical application of the recirculating nebuliser the efficiency with which all traces of a previous solution can be removed from the spray chamber is important. This can be tested by monitoring the level of sample present after each flushing and flooding cycle. This was carried out by recording the detector response of the 327.9-nm zirconium emission from 1 ml of a 100 pg ml-l zirconium solution.After a flooding and flushing cycle, the response from 1 ml of a blank solution injected into the nebuliser was also recorded. This latter procedure was repeated until a reading equivalent to a blank level was reached. The results show a decontamination factor of 5 x 10-3, i.e., the actual zirconium content of the l-ml blank solution recorded after the first wash was 0.02 pgml-1. Thus in most practical instances, two washing and flooding cycles would eliminate contamination. TABLE I1 COMPARISON OF DETECTION LIMITS AND STABILITY ACHIEVED BY CONVENTIONAL AND RECIRCULATING NEBULISERS Short-term precision/pg ml-1 (mean concentration 0.125 pg ml-l) Detection limitlpg ml-1 r \ I I Element Wavelength/nm Recirculating Conventional Recirculating Conventional A A Zr.. .... 343.8 0.02 0.02 0.002 6 0.0020 T i . . .. .. 338.4 0.01 0.02 0.007 8 0.007 7 cu .. .. 327.4 0.003 0.004 0.002 6 0.003 4 Precision The short-term precision described under Nebulising Characteristics does not take into consideration the error associated with sample changeover, in particular the recirculating system where injection of sample and efficiency in drainage can significantly contribute to precision. The precision of the nebuliser systems was compared by making ten replicate170 HULMSTON determinations of a 2.5 pg ml-l zirconium solution. The standard deviations obtained were 0.06 and 0.05 pg ml-l for the recirculating and conventional nebulisers, respectively, which are not significantly different. Conclusions The nebuliser described operates successfully on as little as 1 ml of sample for at least 10 min. The performance of this recirculating nebuliser, in terms of precision and sensitivity, is not significantly different from that of a conventional system. The working procedure devised has reduced the memory effects to insignificant levels. No deterioration in the quality of results for multi-elemental analysis has been observed. The success of this prototype justifies the construction of a further system employing improved control systems to simplify its operat ion. References 1. 2. 3. 4. 5. 6. 7. Karser, M . , Spectrochim. Ada, 1978, 33, 536. Witmer, A., Philips Tech. Rev., 1974, 34, 322. Boumans, P., Analyst, 1976, 101, 587. Arrhenius, S., Ann. Phys. Chem., 1891, 42, 18. Gouy, G., Ann. Phys. Chem., 1879, 18, 5. Novak, J., Anal. Chem., 1980, 52, 576. Rauterberg, E., and Knippenberg, E., Angew. Chem., 1940, 54, 477. Received July 14th, 1982 Accepted September 17th, 1982