JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL. 9 1311 Characteristics of an Inductively Coupled Argon Plasma Operating with Organic Aerosols Part 1 I Spectral and Spatial Observations D. G. Weir and M. W. Blades* Department of Chemistry University of British Columbia Vancouver British Columbia Canada The effect of solvent and solvent load on the background spectra and visual features obtained from an inductively coupled argon plasma have been recorded. An experimental system for making these measure- ments is described. The solvents studied were water methanol and chloroform at solvent loads correspond- ing to the maximum and minimum obtainable. The most conspicuous features observed were emission atomic carbon diatomic carbon and cyanide. A thermal pinch created by the introduction of organic solvents and a recirculation eddy at the base of the torch have been characterized.Keywords Inductively coupled argon plasma; organic solvent; solvent plasma load; thermal pinch The purpose of this paper is to provide a detailed description of the effects of added organic aerosols on the spectra and appearance of the inductively coupled argon plasma (TCAP). Specific regard was paid to the effects that solvent plasma load had on both the discharge and the analytical performance of ICAP-atomic emission spectrometry (AES). Solvent plasma load may be defined simply as the amount of solvent delivered to the discharge per unit time.' It may be conveniently expressed in mg s-' or pmol s-'. In spite of its simple defi- nition solvent plasma load can complicate trace metal analysis by ICAP-AES considerably.In particular solvent plasma loading may cause spectral and non-spectral interferences and degrade analytical perform- ance by introducing noise. This becomes clear when the analyte and background signals are examined. Solvent plasma loading can drastically increase the intensity of both atomic and molecular background emis~ion.'~~ Solvent pyrolysis products principally C and CN in the boundary regions of the discharge and atomic carbon in the plasma region are the sources of background emission. Their complex spectra may overlap with analyte lines and thus interfere with background subtraction. Moreover intense background signals can degrade the detec- tion limits for ICAP-AES. In addition to spectral interferences non-spectral inter- ferences may result from the effect of solvent plasma load on the analyte signal.Solvent plasma load may lower the amount of energy available to the analyte. It is likely that the power required to desolvate aerosol droplets atomize the solvent molecules and then excite the solvent pyrolysis products is supplied at the expense of the power available to vaporize atomize and excite the analyte. Plasma powers of 10-100 W are typically required to dissociate solvent molecules while the other processes require far less. It is true that these powers are small in comparison to the total power dissipated in the discharge (500-1750 W) but they are quite significant in comparison to the small percentage of total power available to the sample (<lo%).Ripson and de Galan4 claimed that much of the power dissipated in the plasma is spent heating the plasma gas and is carried away by convection. For example at an r.f. power of 1.25 kW only 300 W are available to vaporize dissociate and excite the sample including the solvent. Of that 300 W perhaps 100 W are required to heat the carrier argon and another 100 W may be lost radiatively by the brightly emitting solvent pyrolysis products. Alternatively solvent plasma load may cause non-spectral interferences by increasing the amount of power available to * To whom correspondence should be addressed. the analyte. One way it can do this is by altering the geometry of the discharge. In general any molecular material entrained from the aerosol channel into the plasma gas will alter the temperature profile over the toroidal induction region.For example the plasma may shrink and grow hotter if solvent mixes with the argon outer (plasma) gas flowing into the plasma. This is because the molecular material increases the thermal conductivity of the plasma gas-a phenomenon called a thermal p i n ~ h . ~ . ~ Briefly the thermal pinch effect depends on the thermal and electrical conductivity of the plasma gas and their effects on energy loading into and energy dissipation out of the plasma. When a plasma gas containing molecular species (including solvents and diatomic gases) reaches a sufficient temperature (at a given pressure) the bulk of molecu- lar constituents dissociate and the enthalpy of their dis- sociation not only cools the plasma gas but increases its thermal conductivity (the thermal conductivity of nitrogen for example increases from approximately 10 times that of argon to 36 times that of argon when the temperature increases from 5000 to 7000 K).This increase in thermal conductivity acceler- ates heat conduction away from the plasma especially across the steep thermal gradients at its boundary. This accelerated heat loss rapidly cools the peripheral regions of the plasma volume. As the peripheral regions cool down they lose their electrical conductivity causing the plasma volume to shrink while maintaining constant the overall power loading into the plasma. In order to keep the total power loading constant the plasma reacts by increasing its power density while contracting its natural volume. This process eventually results in a more compact hotter plasma able to maintain a stable balance between energy dissipation and energy loading.The contrac- tion into a smaller hotter plasma is known as the thermal pinch effect. Apart from causing spectral and non-spectral interferences solvent plasma loading may also introduce noise. One can expect signal noise when the physical characteristics of the plasma fluctuate in response to the vaporization of incom- pletely desolvated droplets or when the overall solvent plasma load drifts. In short solvent plasma loading introduces noise to both the analyte and background components of the analytical signal. Four physical properties of the sample aerosol have received a great deal of attention (i) the solvent plasma load (ii) the distribution of solvent mass between the vapour and droplet phases of the aerosol (iii) the size distribution of the aerosol droplet^,^ and (iv) the physical properties and chemical com- position of the solvent.Maessen et aE.' described reliable methods for determining both the total solvent plasma load and the distribution of solvent between vapour and droplets1312 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994. VOL 9 phases. For aqueous aerosols Cicerone and Farnsworth' and Olesik et nl.' have demonstrated that droplet effects are significant in ICAPs loaded with aqueous aerosols. Ofesik and c o - w ~ r k e r s ~ - ~ ~ demonstrated that although the overwhelming majority of droplets in an aqueous aerosol are very small a few large droplets (statistically few yet significant in mass) may survive the traverse through the toroidal induction region and on up to the analytical viewing zone of the discharge.These droplets can create regions of localized cooling about 1.5 mm in diameter as they travel downstream. This was a startling revelation since it implies that the temperature profile of the ICAP fluctuates as droplet disturbances flow by. Tt also calls into question the conclusions of many previous investigations which assumed that the solvent loaded ICAP had a temporally stable temperature profile. There is another perspective largely ignored by the droplet investigators. They have mostly assumed that the water loading was confined to the aerosol channel a reasonable assumption if water enters the TCAP predominantly as droplets rather than vapour.Droplets will follow the argon stream along the axial channel. However for many organic solvents the vapour phase predominates. In contrast to droplets vapour mass can diffuse across streamlines and away from the axial channel. Consequently the distribution of solvent over the argon stream is an important solvent load parameter at least for volatile organic solvents. Pan et aL3 and Maessen and KreuninglJ recognized this parameter in their work with organic solvents. Pan et al. investigated it experimentally by varying the auxiliary argon flow rate? The effects of the distribution of solvent vapour over the argon stream is critical to the solvent load- ing process.Table 1 Summary of experimental equipment used Since the first use of the ICAP for spectrochemical analysis many research groups have contributed to the ever expanding body of ICAP-AES and ICAP-MS literature. However rela- tively few authors have reported how the discharge alters in appearance when the operating parameters for example gas flow r.f. power solvent and solvent load are varied. Truitt and Robinson made detailed observations of mixed gas ICAPs14 and introduced terminology for describing the spatial characteristics. They described their mixed gas ICP as consisting of three zones the brilliant blue white opaque core; the bright white transparent secondary or transition region (weaker in intensity than the core); and the faint blue trans- parent tailflame.Their core (also known by other investigators as the energy loading region resided within the load coils and assumed the shape of an annular cylinder or toroid. This region appeared opaque because its emission was so intense that only a very bright object indeed could be perceived behind it. The transition region appeared dimmer than the core and hence appeared transparent; objects behind it could be readily per- ceived. (This region is also known by other authors as the decay region under the assumption that energy dissipation from the plasma outweighs energy loading there.) Finally the tailflame capped the secondary region and was not strictly part of the plasma at all. It was actually a boundary region where air was entrained into the plasma gas resulting in molecular emission and weak atomic emission.It was faint transparent violet in appearance. Truitt and Robinson also conducted a spectro- scopic study of an TCP into which they had introduced organic compounds.14 They supplied a typical emission spectrum from such an ICP. but they did not go into extensive detail in reporting their visual observations of the discharge. ICP unit Power supply Impedance matcher Incident power Reflected power Induction coil Torch Argon flow rates Outer Intermediate Aerosol carrier Translation stage Sample introduction system Nebulizer Spray Chamber Desolvator Adgas fitting Sample uptake rate Solvent load calibration Solvents Imaging lens Spectrometer Monochromator Detectors Irradiance standard Data acquisition board Plasma Therm TCP2500 ( Plasma-Them.Kreeson NJ USA) Plasma Therm HFP2500F Plasma Therm AMN-PS-1 1.00-1.50 kW 0- 50 W 3 turn coif 1 in i.d. 1/16 in spacing betueen turns 1j8 in 0.d. copper tubing 14-15 "C cooling water UBC low Bow (ref. 18) 10.0 1 min- ' (needle valve control rotameter reading) 0.5 1 min-' (needle valve control rotameter reading) 0.61 h-1.01 1 min-' (0.61 I min-l nebufizer flow rate plus 0-0.40 1 min-' adgas controlled by mass flow Stepping motor driven (Superior Electric Slo-Syn type M062-TD03) variable scan distance in 0.0125 mm controller) steps MAK cross-flow (Sherritt-Gordon Alberta Canada} MAK double pass (Sherritt-Cordon) Variable temperature condenser (U shaped 1.3 in pyrex tube 11 mm i.d. cooled by 6 thermoelectric coolers (Melcor Model CP1.4-127-06L) Inner tube (aerosol) 4 mm id.outer tube (adgas) 11 mm i.d. 1.0 ml min- (controlled by Gilson Minipuls 2 peristaltic pump equiped with 2 mm i.d. isoversinic pump Organic solvents continuous weighing method and water glass wool filter followed by cold trap Merck analytical-reagent grade chloroform carbon tetiachloride. propan-2-01 m-xylene methanol Oriel (Stratford CT USA) Model 41775 fused silica plano-convex; centre thickness = 6.9 mm edge backed by water coofed plates (ref. 17) tubing Mandel Scientific) thickness =2.0 mm diameter = 50.8 mm; radius of curvature = 68 & 7 mm; focal length (589 nm) = 150 mm; back focal length = 145.2 mm 1m-Schoeffel-McPherson (Acton AM) :!061 Czerny-Turner equiped with a 120 x 140 mm 1200 g mm- holographic grating 0.833 nm mm - reciprocal linear dispersion (Schoeffel-McPherson Model AH-3264) Photodiode array detectors vertical spatial detector Reticon (Sunnyvale CA USA) RL4096j20 4096 pixels 7 pm wide on 15 pm centres; horizonlal spectral detector Reticon RS 2048; 2048 pixels 12 Ltni wide on 24 pm centres.Arrays cooled to - 15 "C by a Melcor (Trenton NJ USA) CP14-71-IOL backed by water cooled plates dry nitrogen purge prevented frosting Photomultiplier tube Hnmamatsu R95:'i; Kiethley Model 427 current amplifier; Kepco ABC I500 high voltage d.c power supply Electro-Optical Associates (Palto-Alto CA USA) Model QL-l 0 tungsten iodine standard lamp RC Electronics (Santa Barbara CA USA) ISC-16 board sampling rate up to 1 MHz for single channel operationJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL. 9 1313 Fig.1 Schematic diagram of the experimental setup 1-11 sample introduction system (see text for further details); 12 ICAP torch; 13 ICAP vent; 14 alignment laser; 15 lens; 16 removable aperture; 17-20 grating spectrometer; 21 folding mirror in the exit focal plane of the grating spectrometer; 22 spectral photodiode array; 23 spatial photodiode array; 24 PMT. The monochromator and ICAP were mounted on 25 an optical rail bed. The ICAP was mounted on a linear translation stage 26 Greenfield and co-workers have provided the most careful analysis of visual observations for mixed gas I C P S . ~ ' ~ They projected an image of the discharge onto graph paper in order to trace the boundaries of the plasma region. From such geometric records they produced a table of plasma dimensions for different flow rates of molecular gas added to the discharge. Boumans and Lux-Steiner16 described the general appear- ance of an ICAP loaded with methyl isobutyl ketone.Their sketches roughly portray the spatial relation between the boundary regions of the discharge where molecular emission predominates and the plasma region where atc . . jsion :s to the variation in operating parameters was descr This paper is intended to provide a backdrop 1 quent papers on the topic of operation of the ICAP with organic aerosols which will focus on the fundamental properties of such discharges. predominates. Moreover the response of these Experimental Fig. 1 depicts the overall experimental set-up used to study the effects of solvent and solvent plasma load on the ICAP.The set-up essentially consisted of a sample introduction system (1-11) an ICAP (12) light collection optics (15,16) and a grating spectrometer (17-24). Briefly a peristaltic pump 2 fed the nebulizer 5 with test solution 1 through the sample transfer line 4. The nebulizer generated an aerosol stream on argon which flowed through the spray chamber 7 and through tube 9 into a thermoelec- trically cooled desolvating condenser 10 (ref. 7) to the plasma torch 12 (ref. 18). The peristaltic pump drained the condenser through the waste line 8 and the spray chamber through waste line 6 into the waste flask 3. Because other avenues of solvent loss were insignificant the mass difference between the test solution and the waste solution determined the amount of solvent mass delivered to the plasma i.e.the solvent plasma load. Moreover the solvent plasma load could be calibrated accurately and reproducibly against the condenser temperature. The individual components used for the experiments are itemized in Table 1. The experimental system for controlling solvent plasma load has been previously described.17 In most experiments the primary aerosol gas flow was set at 0.61 1 min-l by the high pressure MAK cross flow nebulizer while any extra gas fed through the adgas adapter was con- trolled using a mass flow controller. Solvent plasma load inner gas flow rate and inner gas composition could be controlled reproducibly and accurately. For all of the organic solvents investigated the solvent plasma load was calibrated against the condenser temperature (Fig.2). The method is depicted schematically in Fig. 3. The - 20 ~ 10 0 10 20 Condenser temperature/ C Fig. 2 condenser temperature Calibration plot of chloroform solvent plasma load versus the1314 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL. 9 To torch Aerosol Waste I Balance Fig. 3 Schematic flow chart of the continuous weighing method used to calibrate the solvent plasma load against the condenser temperature solvent plasma load was calibrated against the condenser temperature using the method described previously,17 a con- tinuous weighing method devised by Maessen et al.' In this work the method was modified slightly by equipping the sample and waste flasks with rubber septa and hypodermic needles (Fig.3). Essentially the solvent plasma load was determined by monitoring the decrease in solvent mass at the balance. Maessen et al. provided further details including the appropriate intervals for sampling the decrease in solvent mass and the determination of the error in the solvent plasma load. In general the dependence of solvent load on condenser temperature could be fit with a cubic polynomial and the solvent load could be calibrated for a set of condenser tempera- tures and those condenser temperatures could be used to set the solvent plasma load with good reproducibility. The mini- mum and maximum achievable solvent load for all the solvents calibrated are listed in Table 2. The continuous weighing method was not suitable for measuring water plasma load because large water droplets tended to cling to the condenser and spray chamber walls.Rather than adding surfactant to make the water drain as freely as the organic solvents the water plasma load was determined by trapping the aerosol at the exit of the condenser. A cold trap condensed the vapour while glass wool trapped the droplets. In order to vary the total inner argon flow rate extra argon was added downstream from the condenser through an adgas adapter (11 in Fig. 1). In this way the total aerosol carrier argon flow rate could be varied without perturbing the solvent load or the analyte transport efficiency through the nebulizer. The extra gas surrounded the primary aerosol stream as an annular sheath. Observations of laser light (red helium-neon) scattered off the aerosol confirmed that the annular sheath did not mix with the central aerosol stream in agreement with the low Reynolds number for the flow stream.As a result the aerosol test species were probably concentrated towards the centre of the aerosol stream while freely diffusing solvent vapour was probably distributed more evenly. The size distributions of the aerosol droplets and the trans- port efficiency of the test analytes were not determined in this work. These deficiencies were compensated for by observations of the aerosol stream which indicated that the transport efficiency through the condenser was close to loo% except at very low condenser temperatures and that large droplets were not important in desolvated aerosols.A high transport efficiency through the condenser was indicated by observations of the aerosol stream within the condenser. The aerosol stream was clearly stratified and separated from the condenser wall. It appeared as though the aerosol had imparted a static charge on the condenser tube. Charged droplets had probably collided with the wall and stuck onto it. The charged wall then repelled all the following droplets of like charge. Hence electrostatic repulsion kept the droplets from colliding with the condenser wall and as a result the transport efficiency through the condenser was kept high. Another set of observations indicated the overall trends of the droplet size distribution. When the aerosol was sufficiently desolvated the visibility through the aerosol at the exit of the condenser increased presumably because the droplets had vaporized so that the vapour component of solvent plasma load predominated over droplet component. Scattered laser light (from a red helium-neon laser) helped in these obser- vations.In general the observations indicated that the vapour component predominated over droplets for most desolvated aerosols. A rigorous analysis of the droplet size was considered beyond the scope of this investigation. Details of the r.f. power supply impedance matching network and induction coil are summarized in Table 1. Briefly the r.f. power at 27.12 MHz was operated between 0.75 and 1.75 kW with most experiments conducted at 1.00 1.25 and 1.50 kW. Meassen et al.' described an elaborate ignition procedure for ICAPs loaded with organic solvents.Their procedure was designed to prevent carbon soot from forming on the walls of their torch before the plasma could be tuned (by manually adjusting the impedance matching network) for stable oper- ation. Their procedure involved cleaning out all of the solvent material from the sample introduction system. In this work it was found that the ICAP could be easily ignited by simply turning up the adgas to extremely high flow rates thereby diluting the aerosol stream with argon and preventing any solvent material from interacting with the plasma during ignition. Table 2 Maximum and minimum mass loading molar loading elemental loading and bond dissociation load for a variety of solvents Solvent Water m-Xylene Propan-2-01 Methanol Carbon Chloroform tetrachloride Mass load Q s d l mg s- 0.10 0.30 0.20 0.45 0.30 1 .o 0.20 1.3 1.3 6.3 3.0 10.0 Molar load Q s d pmol s-l 5.6 16.7 1.9 4.2 5.0 16.7 6.3 40.6 8.4 40.9 25.0 83.3 Carbon load Q c d pmol s-' 15.2 33.6 15.0 50.1 6.3 40.6 8.4 40.9 25.0 83.3 Hydrogen load QwJ pmol s - l 11.2 33.4 19 42 40.0 134.0 25.2 162.4 25.0 83.3 Oxygen load Q o d pmol s-' 5.6 16.7 ~ ~~ Chlorine load Q ~ P L I pmol s-' 5.0 16.7 6.3 40.6 33.6 164.0 100.0 250.0 Bond dissociation load 5.2 15.5 14.8 32.8 22.0 73.5 13.0 83.6 11.0 53.2 35.0 117.0 QdissocIWJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL.9 1315 Results and Discussion Spectral Characteristics The line of sight emission from a solvent-loaded ICAP was surveyed over visible wavelengths for several observation heights and for loading by three different solvents water methanol and chloroform.The results of this survey are presented in Fig.4-6. In each figure all of the spectra (each corresponding to a different height above the load coil) were scanned simultaneously using a 4096 pixel linear photodiode array mounted vertically to sample emission from a range of observation heights. The resulting emission survey made it possible to identify the conspicuous emission features and provided a survey of how these emission features depended on the observation height and solvent. The most conspicuous emission features are band emission from diatomic carbon (450-520 nm) and the cyanide radical (410-430 nm) in boundary regions of the discharge line emis- sion from atomic carbon (second order from C I at 248 nm) and argon; and the ubiquitous continuum emission from the atomic plasma. Of the three solvents surveyed chloroform loading resulted in the most intense diatomic carbon emission while both chloroform and methanol loading resulted in intense cyanide emission.The weak diatomic carbon emission from the methanol-loaded ICAP is not surprising when one con- siders the competition between carbon monoxide formation and diatomic carbon. It is likely that the concentration of carbon monoxide predominates over diatomic carbon because of its higher bond energy and therefore greater stability. Of course neither cyanide diatomic carbon or atomic carbon emission were observed for water loading. In general these emission features displayed four distinct trends for the dependence of their intensity on observation height.Atomic line emission from carbon and argon and the plasma continuum emission decreased monotonically with observation height; their emission originated from the plasma 0.6 11 (a) I 0.4 0.2 - - . I . I L\ 0. I / 1 0.6 I (a) 0.4 1 I I I I i l l ( b ) 0.6 0.4 u ~~ 400 440 480 520 560 600 640 680 720 Wavelengthinm Fig. 5 Spectral survey of the visible emission from the ICAP loaded with water for several observation heights (a) 21 mm; (b) 18 mm; (c) 15 mm; ( d ) 12 mm; (e) 9 mm; ( f ) 6 mm " 400 440 480 520 560 600 640 680 720 Wavelengthhm Fig. 4 Spectral survey of the visible emission from the ICAP loaded with methanol for several observation heights (a) 21 mm; (b) 18 mm; (c) 15 mm; (d) 12 mm; (e) 9 mm; (f) 6 mm 400 440 480 520 560 600 640 680 720 Wave len g t h/n m Fig.6 Spectral survey of the visible emission from the ICAP loaded with chloroform for several observation heights (a) 21 mm; (b) 18 mm; (c) 15 mm; (d) 12 mm; (e) 9 mm; ( f ) 6 mmJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL. 9 1315 Results and Discussion Spectral Characteristics The line of sight emission from a solvent-loaded ICAP was surveyed over visible wavelengths for several observation heights and for loading by three different solvents water methanol and chloroform. The results of this survey are presented in Fig.4-6. In each figure all of the spectra (each corresponding to a different height above the load coil) were scanned simultaneously using a 4096 pixel linear photodiode array mounted vertically to sample emission from a range of observation heights.The resulting emission survey made it possible to identify the conspicuous emission features and provided a survey of how these emission features depended on the observation height and solvent. The most conspicuous emission features are band emission from diatomic carbon (450-520 nm) and the cyanide radical (410-430 nm) in boundary regions of the discharge line emis- sion from atomic carbon (second order from C I at 248 nm) and argon; and the ubiquitous continuum emission from the atomic plasma. Of the three solvents surveyed chloroform loading resulted in the most intense diatomic carbon emission while both chloroform and methanol loading resulted in intense cyanide emission.The weak diatomic carbon emission from the methanol-loaded ICAP is not surprising when one con- siders the competition between carbon monoxide formation and diatomic carbon. It is likely that the concentration of carbon monoxide predominates over diatomic carbon because of its higher bond energy and therefore greater stability. Of course neither cyanide diatomic carbon or atomic carbon emission were observed for water loading. In general these emission features displayed four distinct trends for the dependence of their intensity on observation height. Atomic line emission from carbon and argon and the plasma continuum emission decreased monotonically with observation height; their emission originated from the plasma 0.6 11 (a) I 0.4 0.2 - - .I . I L\ 0. I / 1 0.6 I (a) 0.4 1 I I I I i l l ( b ) 0.6 0.4 u ~~ 400 440 480 520 560 600 640 680 720 Wavelengthinm Fig. 5 Spectral survey of the visible emission from the ICAP loaded with water for several observation heights (a) 21 mm; (b) 18 mm; (c) 15 mm; ( d ) 12 mm; (e) 9 mm; ( f ) 6 mm " 400 440 480 520 560 600 640 680 720 Wavelengthhm Fig. 4 Spectral survey of the visible emission from the ICAP loaded with methanol for several observation heights (a) 21 mm; (b) 18 mm; (c) 15 mm; (d) 12 mm; (e) 9 mm; (f) 6 mm 400 440 480 520 560 600 640 680 720 Wave len g t h/n m Fig. 6 Spectral survey of the visible emission from the ICAP loaded with chloroform for several observation heights (a) 21 mm; (b) 18 mm; (c) 15 mm; (d) 12 mm; (e) 9 mm; ( f ) 6 mmJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL.9 1317 Enveloped within the region of diatomic carbon emission sits the region of intense continuum and argon line emission. This region will hereafter be referred to as the atomic plasma for convenience of discussion. The atomic plasma consists of three discernible components the plasma core the bright secondary plasma and the dim secondary plasma from all of which one observes continuum emission and atomic argon line emission to the exclusion of molecular emission. These regions typically form an annulus or toroid within the load coils which appears to coalesce into a cone further downstream as shown in Fig. 9. Although these three regions of the atomic plasma may not always be readily discerned as they may all blend into one gradual transition they too have been introduced for the convenience of discussion.A cross section through the plasma core is represented by the two oval regions of lightest grey sitting side by side within the confinement tube of the torch. Enveloping the plasma core is the bright secondary plasma which in turn is enveloped or bounded by the dim secondary plasma. Distortions of these two secondary emission regions of the atomic plasma in response to variation of the operating parameters and the solvent loading are perhaps the most important observations to note in this paper. They qualitatively indicate changes in the physical properties of the ICAP changes which determine how energy is transferred to the analyte and hence changes critical to the analytical performance of ICAP-AES.One further feature of the atomic plasma worth introducing here is the channel along the axis through the centre of the toroid. This will be referred to as the central channel and does not necessarily coincide with the aerosol channel or distribution of analyte injected into the discharge. The region enveloping the downstream cone of the atomic plasma represents the entrainment region or tail flame of the discharge. Weak violet emission from the air entrainment region from species such as CN was clearly visible from this region. One additional emission structure may be observed under certain operating conditions; a hollow cone of incandescent emission possibly consisting of glowing carbonaceous soot particles is often found nested within the hollow axial plume of diatomic carbon emission.The solid curve nested underneath the discharge represents the hollow cone of incandescent radiation. This emission feature was usually observed for high levels of loading by solvents with excess carbon relative to their oxygen content. Significantly the bright orange cone was never observed for methanol loading or for ethanol-water mixtures nor was buildup of carbonaceous soot on the torch wall ever a problem for these solvents. In both cases carbon and oxygen were present in near stoichiometric proportions for the formation of CO. On the other hand build up of carbonaceous soot accompanied the appearance of the hollow incandescent cone for relative high loading of solvents such as xylene chloroform and hexane.The following figures illustrate how the emission structure depicted in Fig. 9 varies with solvent solvent load inner argon flow rate and forward power. The behaviour depicted in these figures includes the response of the plasma volume (owing to the thermal pinch effect) vertical translation and vertical contraction of the plasma the spatial structure of the diatomic carbon emission plume (including nested cone structures) and the behaviour of the normal analytical zone (at the apex of the plasma decay region). Comparison of Water Methanol and Chloroform In order to describe the behaviour and emission structure of an ICAP loaded with any of the solvents investigated in this work it is sufficient to consider the distinctive behaviour and emission structure resulting from loading by only three of them; water methanol and chloroform.Representative obser- vations for ICAPs loaded by these three solvents are depicted in Fig. lo@)-@); Figure 1O(a) depicts representative obser- vations for an ICAP without solvent loading i.e. for a pure argon ICAP flowing into air in order to provide a basis for comparison. In all cases intermediate settings for power gas flow and solvent load were applied. In all four discharges the atomic plasma assumed the geometry of a torus which coalesced downstream into a cone. Within this general geometry one may distinguish the atomic plasma regions of Fig. 10(b)-(d) by the extent that their central channel has dilated how far up through the load coil the plasma has translated and how much the downstream portion of the secondary plasma has ‘bloomed open’.For the ICAP loaded with water the atomic plasma region did not appear to have dilated or translated downstream to any great extent when compared with the pure argon ICAP (without solvent loading). However upon close inspection it could be noted that the central channel became less diffuse and that the atomic plasma appeared to be translated upwards approximately 0.5-1.5 mm when loaded with water. The same minor changes with respect to the pure argon ICAP were noted when chloroform loading was used. In contrast more Fig. 10 Representative observations of an ICAP discharge (a) without solvent load and loaded with (b) water; (c) methanol; and (d) chloroform1318 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL.9 pronounced changes were observed when methanol loading was used. In this case the atomic plasma clearly translated downstream far enough in fact (under light to medium solvent load) for its base (or upstream edge) to reside with the load coil. Also in response to methanol loading the central channel of the ICAP appeared to dilate strongly indicating that methanol loading causes a thermal pinch effect. Beyond the characteristics of the atomic plasma region no further distinguishing characteristics were noted for pure argon and water-loaded ICAPs. In contrast both the methanol- and chloroform-loaded ICAPs exhibited brilliant plasma boundary regions caused by emission from their solvent pyrolysis prod- ucts. The chloroform-loaded ICAP displayed brilliant green emission from a sharply defined spatial structure described previously as an outer annulus joined at the upstream end to a hollow inner plume.For an indication of how sharply defined this emission structure was the cylindrical wall of the inner plume was often <0.5 mm thick while the boundaries appeared perfectly sharp. The transition from intense green emission to no perceptible green emission appeared to follow a step function. The small barbs of diatomic carbon emission on the downstream end of the outer cup are not artifacts of Fig. 10 but were reproducibly observed and clearly visible owing to the sharply defined spatial structure of diatomic carbon emission. They probably indicate the presence of a back eddy in the outer gas stream just beyond the exit of the torch.The methanol loaded ICAP displayed relatively dim green emission from a diffusely defined spatial structure. The boundaries of this diffuse structure gradually faded over a distance of approximately one millimeter. However under operating conditions of high methanol loading or low power the methanol-loaded ICAP exhibited a brilliant green sharply defined structure similar to that of the chloroform-loaded ICAP. This sharply defined structure was invariably nested within the diffuse structure. The methanol- and chloroform- loaded ICAPs also differed from the pure argon and water- loaded ones by displaying weak violet emission from their tail flames. In general the appearance of the discharge depended on the relative amounts of oxygen and carbon in the aerosol stream.For example m-xylene- propan-2-01- and hexane-loaded ICAPs were similar in appearance to a chloroform-loaded ICAP. On the other hand an ICAP loaded with an ethanol- water mixture was similar in appearance to an ICAP loaded with methanol. Interestingly an ICAP loaded with xylene but with oxygen added to the aerosol stream in equimolar pro- portions to the solvent carbon was also similar in appearance to an ICAP loaded with methanol. Because of this general dependence on the relative amounts of carbon and oxygen load the following discussion will be confined to chloroform- methanol- and water-loaded ICAPs and their response to solvent load power and gas flow rates. The appearance of an ICAP loaded with any other solvent solvent mixture or combination of solvent loading and oxygen addition may be regarded an intermediate of the methanol- and chloroform- loaded ICAPs. Water loading The ICAP appeared to respond very slightly to water loading in comparison to its response to loading by other solvents.In fact the atomic plasma region appeared to translate down- stream through the load coil by only 1 mm when the water load was increased from its minimum attainable level (0.15 mg s-’) to its maximum attainable level (0.30 mg s-l). Whether or not this was a downstream translation or a contraction of the atomic plasma along the direction of flow owing to the thermal pinch effect was not clear from obser- vations alone. It has been suggested” that the plasma is lifted off the tulip of the torch by expanding water vapour and vaporizing droplets near the base of the plasma but in future publications arguments and experimental evidence against this conjecture and in support of the thermal pinch effect will be presented. Accompanying the downstream translation the central channel became just perceptibly darker or more clearly defined in space.On the whole the ICAP appeared to be relatively insensitive to water loading an insensitivity which may be explained by the characteristically low mass loading of water compared with other solvents when nebulized by conventional pneumatic nebulizers. Typically the maximum water load that pneumatic nebulizers are capable of delivering to the ICAP is between 0.5 and 0.8 mg s-’. One should note that ultrasonic nebulizers are usually capable of delivering much greater water loads to an ICAP than pneumatic nebuliz- Fig.11 Representative observations for an ICAP loaded with methanol at (a) minimum; (b) intermediate; and (c) maximum solvent loadJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL. 9 1319 ers. However they must be fitted with some sort of desolvation device such as a condenser in order to desolvate the aerosol and reduce the water load before the aerosol stream reaches the plasma. Methanol loading A far greater range of methanol loading was accessible to observation. The ICAP response to methanol loading is illus- trated in Fig. 11 (a)-@) which depicts typical observations for an ICAP loaded with the minimum obtainable intermediate and maximum levels of methanol loading respectively.The most obvious responses are the way the atomic plasma appar- ently translates downstream and the way both the diffuse and sharp diatomic carbon plumes extend along the central channel as the methanol load increases. In addition to its downstream translation the atomic plasma obviously contracts in the direction of flow with increasing methanol load a contraction which may result from the thermal pinch effect. The obvious extension translation and contraction noted above are accompanied by a more subtle response. The second- ary plasma appears to bloom open in response to an increase in the methanol load. In Fig. 11 (a) the bright secondary plasma (light grey) retains the characteristic shape of a toroid capped cone.Then as the methanol load increases from the lowest attainable to the intermediate level the cone of the bright secondary plasma is almost completely penetrated leaving only a thin arch near its apex as shown in Fig. ll(b). In response to higher methanol loading the apex of the bright secondary plasma blooms open. The dim secondary plasma appears to bloom open in a similar manner but a step behind the bright secondary plasma. This blooming holds important implications for the analyt- ical performance of ICAP-AES. It reveals that methanol load- ing drastically alters how much energy is available at the central channel from the energy dissipation region or plasma core energy required to desolvate vaporize atomize ionize and excite the analyte. At the extreme of maximum methanol load the discharge may be regarded as having folded com- pletely in itself effectively retracting from the analyte so that the plasma interacts incompletely with the analyte if at all.At the other extreme of minimum methanol loading one would expect the plasma to interact or supply energy to the analyte effectively yet also expect the higher background levels resulting from the higher continuum emission to degrade the signal-to-noise ratio. Directly linked with the blooming of the secondary atomic plasma is the behaviour of the diatomic carbon emission. As the secondary plasma blooms open the central plume extends downstream. In general the cup and plume grow downstream with increasing solvent load and the intense sharply defined component of green emission overtakes and predominates over the diffuse one.Chloroform loading Chloroform loading also caused the diatomic carbon cup and plume to grow downstream in a manner similar to methanol loading however several other characteristics distinguished the response of the ICAP to chloroform loading from its response to methanol loading. Fig. 12(a)-(c) illustrates typical observations of how the ICAP responded to variation of chloroform loading. In contrast to methanol loading chloro- form loading did not cause the atomic plasma to translate very far downstream and no obvious thermal pinch effect was observed. Moreover the diatomic carbon emission structure always appeared sharply defined with no conspicuous diffuse structure such as the one observed for methanol loading.Nested within the sharply defined central plume a bright orange hollow cone of incandescent soot particles formed at high chloroform loads something never observed for methanol loading. This hollow orange cone became brighter with greater chloroform loading and appeared to nest closely within the central plume of diatomic carbon emission as shown by the continuous solid curve in Fig. 12(c). Because the spatial structure of diatomic carbon emission was so sharply defined in a chloroform-loaded ICAP relative to the same structure in a methanol-loaded ICAP several subtle spatial responses to chloroform loading were noted and are illustrated in Figures 13(a)-(c). The most interesting response was how the central plume changed shape as it extended up through the central channel of the discharge.At minimum chloroform loading the hollow inner plume could be regarded as a hollow cone [Fig. 13(u)] of approximately (a) Fig. 12 Representative observations for an ICAP loaded with chloroform at (a) minimum; (b) intermediate; and (c) maximum solvent load1320 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL. 9 n Fig. 13 Observations of how the shape of the inner plume varied as the chloroform plasma load was increased (a) at lowest chloroform load (b) at an intermediate load; and (c) at the highest load triangular cross section. At intermediate solvent loads the base of the plume remained approximately conical but the tip of the plume extended downstream to form a cylindrical annulus capped by a bullet shaped region [Fig.13(6)]. At maximum chloroform loading the top of the cylinder dilated to give the plume a bulbous end as shown in Fig. 13(c). Accompanying this extension and expansion were changes in the thickness of the wall of the hollow plume. As the plume extended downstream the wall of its leading edge appeared to grow thicker i.e. thicker in the apparent direction of gas flow whereas the thickness of the walls in the radial direction appeared to remain constant. It’s likely that the steepness of thermal gradients across the wall of the plume determine its thickness. Diatomic carbon is probably only stable over a relatively narrow temperature range or a small distance along a steep thermal gradient. At temperatures above its range of stability diatomic carbon tends to dissociate or become unstable relative to atomic carbon while at temperatures below its range of stability it tends to associate into polyatomic carbon-containing species.It follows that the cross section through the diatomic carbon plume may be regarded as an isothermal contour in space thinner when the thermal gradient crossing it is steeper. This explains why the wall of the plume varies in thickness; the side walls are thinnest because the steepest temperature gradients in the discharge are the radial gradients extending from the discharge axis out across the side wall of the plume and into the toroidal plasma core. The tip of the plume is thickest because axial temperature gradients are characteristically gentler than radial ones in a solvent-loaded ICAP.The shape of the plume is probably determined by gas flow patterns and heat conduction. In Figure 13(a) the triangular cross section of the plume may be understood as a dissociation front receding radially towards the discharge axis. The blunt- ness (aspect ratio) of the cone is determined by the gas flow velocity and the competitive rates of heat consumption by the enthalpy of dissociation and radial heat conduction towards the axis from the toroidal core. In Figure 13(a) heat conduction overtakes heat consumption before the gas flows out of the torch. In Figure 13(b) they are nearly balanced. Fig. 13(c) is more difficult to explain. It appears as though heat consump- tion has overtaken heat conduction but one must remember that above the torch rim the discharge gas expands radially so that the expansion of the plume may simply be a manifes- tation of the radial expansion of the discharge gas.The outer or peripheral cup of diatomic carbon emission also displayed behaviour indicative of thermal gradients gas flow patterns and heat conduction. Moreover its behaviour indicated how solvent material had been distributed within the discharge. It was thickest around the base of the discharge and thinnest between the plasma torus and the torch wall presumably for reasons similar to those determining the thick- ness of the walls of the central plume. However it is question- able whether the outer cup can be regarded as a dissociation front similar to the inner plume. If it were then solvent material would have to be swept around the base of the plasma and enter through the outer periphery.Another more realistic possibility is that solvent material was folded into the outer argon stream by a recirculation eddy at the upstream edge of the discharge (in the wake of the intermediate tube). [n that case solvent material could have been either swept (around the base of the discharge or folded into the plasma then transported to the periphery by diffusion (by either path reaching the region of the outer cup and forming diatomic carbon via dissociation of solvent molecules or via association of carbon atoms diffusing out of the atomic plasma). The flow dynamics in the ICAP and similar discharges have been investigated with high speed anem- ometer probes (pitot t ~ b e s ) ~ ~ - ~ ~ particle tracking,24 and analy- sis of temporally resolved emission.* However most of the flowfield remains inaccessible to experiment for both funda- mental and non-fundamental reasons.Invasive probes disrupt the flow stream high temperatures of the plasma melt the probes and vaporize tracking particles and the intense emission complicates laser doppler anemometry. Fortunately computer simulations can offer complete access to the fl~wfield.~~ The simulation results relevant to solvent loading are the flow structure they predict within the confinement tube where solvent distributes over the argon stream. Beyond the torch exit however the complexity of the flowfield apparently defies simulation at least at the moment. Nevertheless insight into the flowfield beyond the confinement tube may be found in the literature on axisymmetric jets and f l a r n e ~ .~ ~ ~ ~ Both of these flow systems resemble the tail flame of the ICAP in many respects. A representative diagram of the flowfields in the ICAP is provided in Fig. 14. The recirculation eddy at the base of the discharge can mix solvent vapour from the inner aerosol stream into the outer coolant stream. Note that this eddy is not necessarily a turbulent phenomenon. Downstream from the eddy the flowfield develops a relatively flat velocity profile except for a central maximum. Particle tracking (of aerosol droplets) reveals that the central flow velocity is approximately 25 m s-l. Moreover Reynolds numbers <<2000 validate the // / \ \\Airentrainment \ Development into a unidirectional f lowf ield with both axial and tangential velocity components Recirculation eddy at the base of the discharge Tangential argon inlet port Fig.14 Schematic diagram of the flowfields in the ICAPJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL. 9 1321 assumption that the flowfield here is laminar rather than turbulent. The Reynolds number is defined as where L is the characteristic length of the structure confining the flow vo is the centerline velocity p is the density of the fluid and p its viscosity. In the confinement tube of the ICAP approximate values for these are 0.01 m 10 m s-l 0.2 kg m-3 and 2 x lop4 kg ms-l respectively,28 so Re = 100 is <<2000. The laminar flowfield has both an axial and tangential component the latter imparted by the tangential gas inlet for the coolant flow.The tangential component or swirl helps to stabilize the discharge. Simulations reveal that the swirl also concentrates the power density towards the axis.28 This hap- pens because the centrifugal moment of the swirl holds the bulk of the coolant stream against the confinement tube. As a result the outer boundary of the induction region is kept cool and both the electrical conductivity power dissipation are kept low beyond a certain radius so the plasma is confined to a smaller radius than if swirl were absent. All of these flowfield characteristics help us understand how solvent material can be transported through the discharge by convection. Beyond the exit of the torch the flow field becomes far more complex.When the plasma jet flows out of the torch into the quiescent room air the flowfield is no longer bounded by the torch wall but extends beyond into the argon stream. Where the flowfield crosses from the argon jet into the air there is a surface of discontinuity or sudden jump between flowing argon and the air at rest (for clarity the surrounding air being drawn into the argon stream is ignored). Varicose instabilities form in this cylindrical surface of discontinuity and modulate the diameter of argon stream. Instabilities of this sort are familiar to anyone who has seen the jumping orange flame of a Bunsen burner. As the varicose pulsations propagate downstream they roll up into ring vortices. Winge et uL21 provide high speed movies of these structures.Experimental evidence indicates that the varicose pulsations penetrate to the very axis of the discharge. Re = LPV,/P Effect of Sheath Gas Flow The observed response of a chloroform-loaded ICAP to vari- ation of the inner argon flow rate provided further insight into how solvent material might be distributed in the discharge. These observations are illustrated by Fig. IS@)-(c). All three frames depict an ICAP loaded with an intermediate amount of chloroform but at low moderate and high inner argon flow rates. As shown in Fig. 15 the length that the plume extended downstream was inversely related to the length that the outer cup extended downstream. The atomic plasma also responded to variation of the inner argon flow rate in a conspicuous manner. As the inner argon flow rate was increased the atomic plasma appeared to move upstream or sit down within the intermediate tube presumably because higher inner argon flow rates prevent the recirculation eddy at the base of the discharge from folding solvent material into the outer argon stream attenuating the downstream translation/thermal pinch effect.This attenuating effect that the inner argon flow rate had on the ICAP’s response to solvent loading was observed in varying degrees depending on the solvent and the level of solvent load. It wasn’t really an attenuation (the central channel became more sensitive as the toroidal region became less sensitive) distribution of solvent. The observed response of the solvent loaded ICAP to variation of power may be stated quite simply.At lower powers the discharge responded to all of the other parameters as described above only more sensitively. At higher powers the discharge responded less sensitively while at powers approaching 2.0 kW the discharge became relatively insensitive to variation of any other parameter including solvent load. Extrapolation to Other Solvents It was stated earlier that the observed response of the ICAP to loading by any solvent investigated in this work could be conveniently described as similar to a chloroform-loaded ICAP similar to a methanol-loaded ICAP or resembling a hybrid of the two depending on the relative content of carbon and oxygen in the solvent. If the carbon oxygen content of the solvent approached 1 1 then its appearance was similar to that of a methanol-loaded ICAP.If the carbon content greatly exceeded the oxygen content then its appearance was similar to a chloroform-loaded ICAP. This generalization may be extended further to solvent mixtures and to oxygen addition to the aerosol stream. For example the appearance of an (a) Fig. 15 Representative observations for an ICAP loaded with an intermediate level of chloroform at an inner argon flow rate of (a) 0.6 1 m-l; (h) 0.8 1 m-l; and (c) 1.01 m-’1322 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY DECEMBER 1994 VOL. 9 ethanol-loaded ICAP may be described as a hybrid of chloroform- and methanol-loaded ICAPs but loading by an equimolar mixture of ethanol and water results in a discharge which is virtually indistinguishable from a methanol-loaded ICAP.The same is true for an ICAP loaded with xylene when oxygen has been added to the sheath gas in equimolar pro- portion to the amount of carbon in the xylene. Conclusions The detailed observations reported in this paper point out how some of the macroscopic structure of the ICAP discharge changes in response to changes in solvent and solvent load. They also reveal that the appearance of the discharge depends on the relative proportions of oxygen and carbon in the aerosol stream irrespective of the chemical form in which the oxygen and carbon are introduced. Beyond that a number of physical phenomena are evident in the observations. These include a thermal pinch and convective distribution of solvent material over the argon stream. Moreover incandescent radiation was observed from a conical shell nested within a dissociation front indicating that solvent pyrolysis proceeds via macro- scopic soot particles.Overall the observations reported in this paper provide a valuable survey of the parametric behavior of the solvent-loaded ICAP. This survey will be followed up in future publications with more detailed physical measurements of spatial emission structure electron number density exci- tation temperature and ion atom emission intensity ratios. The authors acknowledge funding from the Natural Sciences and Engineering Research Council of Canada. References 1 Maessen F. J. M. J. Kreuning G. and Balke J. Spectrochim. Acta Part B 1986 41 3. 2 Pan C. Zhu G. and Browner R. F. J. Anal. At. Spectrorn. 1990 5 537. 3 Pan G. Zhu G. and Browner R. F. J. Anal. At. Spectrom. 1992 7 1231. 4 Ripson P. A. M. and de Galan L. Spectrochim. Acta Part B 1983 38 707. 5 NASA Contractor Report 1143 1968. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Greenfield S.. and McGeachin H. M. Anal. Chim. Acta 1978 100 101. Canals A. and Hernandis V. J. Anal. At. Spectrom. 1990 5 61. Cicerone M. T. and Farnsworth P. B. Spectrochim. Acta Part B 1989 44 897. Olesik J. W. Smith L J. and Williamsen E. J. Anal. Chem. 1989 61 2002. Olesik J. W. and Fister J. C. Spectrochirn. Acta Part B 1991 46 851. Olesik J. W. and Fister J. C. Spectrochirn. Acta Part B 1991 46 851. Olesik J. W. and Fister J. C. Spectrochim. Acta Part B 1991 46 869. Maessen F. J. M. J. and Kreuning G. Spectrochim. Acta Part B 1989 44 387. Truitt D. and Robinson J. W. Anal. Chim. Acta 1970 51 61. Greenfield S. Jones I. L. McGeachin H. M. and Smith P. B. Anal. Chim. Acta 1975 74 225. Boumans P. W. J. M. and Lux-Steiner M. C. Spectrochim. Acta Part B 1982 37 97. Weir D. G. J. and Blades M. W. Spectrochim. Acta Part B 1990 45 615. Burton L. L. and Blades M. W. Appl. Spectrosc. 1986 40 265. Caughlin B. L. Blades M. W. Spectrochim. Acta Part B 1987 42 353. Winge R. K. Eckels D. E. DeKalb E. L. and Fassel V. A. J. Anal. At. Spectrom. 1988 3 849. Winge R. K. Crain J. S. and Houk R. S. J. Anal. At. Spectrom. 1991 6 601. Barnes R. M. and Genna J. L. Spectrochim. Acta Part B 1981 36 299. Barnes R. M. and Schleicherr R. G. Spectrochim. Acta Part B 1981 36 81. Donskoi A. V. Goldfarb V. M. and Klubnikin V. S. Physics and Technology of Low- Temperature Plasmas English ed. Iowa State University Press 1977 p. 471. Patankar S. V. Numerical Heat Transfer and Heat Flow McGraw-Hill New York 1980. Becker H. A. and Massaro T. A. J. Fluid Mech. 1968 31 435. Dahm W. J. A. Frieller C. E. and Tryggvason G. J. Fluid Mech. 1992 241 371. Benoy D. A. van der Mullen J. A. M. van der Sijde B. and Schram D. C. J. Quant. Spectrosc. Radiat. Transfer 1991 46 195. Paper 4/01 259C Received March 1 1994 Accepted July 26 1994