10.2: Emission Spectroscopy Based on Arc and Spark Sources
- Page ID
- 74333
\( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \)
\( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)
\( \newcommand{\id}{\mathrm{id}}\) \( \newcommand{\Span}{\mathrm{span}}\)
( \newcommand{\kernel}{\mathrm{null}\,}\) \( \newcommand{\range}{\mathrm{range}\,}\)
\( \newcommand{\RealPart}{\mathrm{Re}}\) \( \newcommand{\ImaginaryPart}{\mathrm{Im}}\)
\( \newcommand{\Argument}{\mathrm{Arg}}\) \( \newcommand{\norm}[1]{\| #1 \|}\)
\( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\)
\( \newcommand{\Span}{\mathrm{span}}\)
\( \newcommand{\id}{\mathrm{id}}\)
\( \newcommand{\Span}{\mathrm{span}}\)
\( \newcommand{\kernel}{\mathrm{null}\,}\)
\( \newcommand{\range}{\mathrm{range}\,}\)
\( \newcommand{\RealPart}{\mathrm{Re}}\)
\( \newcommand{\ImaginaryPart}{\mathrm{Im}}\)
\( \newcommand{\Argument}{\mathrm{Arg}}\)
\( \newcommand{\norm}[1]{\| #1 \|}\)
\( \newcommand{\inner}[2]{\langle #1, #2 \rangle}\)
\( \newcommand{\Span}{\mathrm{span}}\) \( \newcommand{\AA}{\unicode[.8,0]{x212B}}\)
\( \newcommand{\vectorA}[1]{\vec{#1}} % arrow\)
\( \newcommand{\vectorAt}[1]{\vec{\text{#1}}} % arrow\)
\( \newcommand{\vectorB}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \)
\( \newcommand{\vectorC}[1]{\textbf{#1}} \)
\( \newcommand{\vectorD}[1]{\overrightarrow{#1}} \)
\( \newcommand{\vectorDt}[1]{\overrightarrow{\text{#1}}} \)
\( \newcommand{\vectE}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash{\mathbf {#1}}}} \)
\( \newcommand{\vecs}[1]{\overset { \scriptstyle \rightharpoonup} {\mathbf{#1}} } \)
\( \newcommand{\vecd}[1]{\overset{-\!-\!\rightharpoonup}{\vphantom{a}\smash {#1}}} \)
\(\newcommand{\avec}{\mathbf a}\) \(\newcommand{\bvec}{\mathbf b}\) \(\newcommand{\cvec}{\mathbf c}\) \(\newcommand{\dvec}{\mathbf d}\) \(\newcommand{\dtil}{\widetilde{\mathbf d}}\) \(\newcommand{\evec}{\mathbf e}\) \(\newcommand{\fvec}{\mathbf f}\) \(\newcommand{\nvec}{\mathbf n}\) \(\newcommand{\pvec}{\mathbf p}\) \(\newcommand{\qvec}{\mathbf q}\) \(\newcommand{\svec}{\mathbf s}\) \(\newcommand{\tvec}{\mathbf t}\) \(\newcommand{\uvec}{\mathbf u}\) \(\newcommand{\vvec}{\mathbf v}\) \(\newcommand{\wvec}{\mathbf w}\) \(\newcommand{\xvec}{\mathbf x}\) \(\newcommand{\yvec}{\mathbf y}\) \(\newcommand{\zvec}{\mathbf z}\) \(\newcommand{\rvec}{\mathbf r}\) \(\newcommand{\mvec}{\mathbf m}\) \(\newcommand{\zerovec}{\mathbf 0}\) \(\newcommand{\onevec}{\mathbf 1}\) \(\newcommand{\real}{\mathbb R}\) \(\newcommand{\twovec}[2]{\left[\begin{array}{r}#1 \\ #2 \end{array}\right]}\) \(\newcommand{\ctwovec}[2]{\left[\begin{array}{c}#1 \\ #2 \end{array}\right]}\) \(\newcommand{\threevec}[3]{\left[\begin{array}{r}#1 \\ #2 \\ #3 \end{array}\right]}\) \(\newcommand{\cthreevec}[3]{\left[\begin{array}{c}#1 \\ #2 \\ #3 \end{array}\right]}\) \(\newcommand{\fourvec}[4]{\left[\begin{array}{r}#1 \\ #2 \\ #3 \\ #4 \end{array}\right]}\) \(\newcommand{\cfourvec}[4]{\left[\begin{array}{c}#1 \\ #2 \\ #3 \\ #4 \end{array}\right]}\) \(\newcommand{\fivevec}[5]{\left[\begin{array}{r}#1 \\ #2 \\ #3 \\ #4 \\ #5 \\ \end{array}\right]}\) \(\newcommand{\cfivevec}[5]{\left[\begin{array}{c}#1 \\ #2 \\ #3 \\ #4 \\ #5 \\ \end{array}\right]}\) \(\newcommand{\mattwo}[4]{\left[\begin{array}{rr}#1 \amp #2 \\ #3 \amp #4 \\ \end{array}\right]}\) \(\newcommand{\laspan}[1]{\text{Span}\{#1\}}\) \(\newcommand{\bcal}{\cal B}\) \(\newcommand{\ccal}{\cal C}\) \(\newcommand{\scal}{\cal S}\) \(\newcommand{\wcal}{\cal W}\) \(\newcommand{\ecal}{\cal E}\) \(\newcommand{\coords}[2]{\left\{#1\right\}_{#2}}\) \(\newcommand{\gray}[1]{\color{gray}{#1}}\) \(\newcommand{\lgray}[1]{\color{lightgray}{#1}}\) \(\newcommand{\rank}{\operatorname{rank}}\) \(\newcommand{\row}{\text{Row}}\) \(\newcommand{\col}{\text{Col}}\) \(\renewcommand{\row}{\text{Row}}\) \(\newcommand{\nul}{\text{Nul}}\) \(\newcommand{\var}{\text{Var}}\) \(\newcommand{\corr}{\text{corr}}\) \(\newcommand{\len}[1]{\left|#1\right|}\) \(\newcommand{\bbar}{\overline{\bvec}}\) \(\newcommand{\bhat}{\widehat{\bvec}}\) \(\newcommand{\bperp}{\bvec^\perp}\) \(\newcommand{\xhat}{\widehat{\xvec}}\) \(\newcommand{\vhat}{\widehat{\vvec}}\) \(\newcommand{\uhat}{\widehat{\uvec}}\) \(\newcommand{\what}{\widehat{\wvec}}\) \(\newcommand{\Sighat}{\widehat{\Sigma}}\) \(\newcommand{\lt}{<}\) \(\newcommand{\gt}{>}\) \(\newcommand{\amp}{&}\) \(\definecolor{fillinmathshade}{gray}{0.9}\)ZAP! What cartoon aficionado hasn't seen a lightning bolt blasting a chunk out of some material or character? Even sub-human life forms understand that transient, high-current discharges can destructively ablate materials. When an electrical current is passed through a boundary between dissimilar materials, cations travel in a direction opposite to electrons. If the electrons are generated in the gas phase, that means that atoms from the electrode into which the electrons travel will be blasted into the gas. Collisions between those atoms and either electrons, neutral atoms, atomic ions, or small molecules will excite or ionize the sampled atoms. Typically, such sampling and excitation is carried out at atmospheric pressure. Commonly used gases used include Ar, N2, or air. This section describes the common atmospheric pressure electrical discharges used for elemental analysis.
Below is a sketch of a plasma formed between two conducting electrodes (perhaps two sharpened thoriated tungsten rods, as seen in gas tungsten arc welding). The blue region is a plasma, created by putting a sufficiently high electrical potential between the electrodes that the intervening gas is ionized. If the predominant behavior over the course of an experiment is to have current passing between the electrodes, with only brief (or no) interruptions, the discharge is an arc. If current passes only intermittently, and most of the time the gas is unionized (that's un-ion-ized, not union-ized; the AFL-CIO isn't involved!), then the discharge is a spark. The ratio of current on-time to total experiment time is the duty cycle. If the duty cycle is <0.1, the system is certainly a spark.
Two electrodes with plasma
Cations (positive ions) are attracted to negatively-biased electrodes, cathodes. Sure enough, electrons are attracted to anodes. Which carries more momentum -- an electron with some amount of energy or an argon ion with the same amount of energy? Think about it, then click here for the answer. Ions hitting a surface can heat the surface (the electrical energy gets turned into heat) or sputter it (momentum knocks atoms off the surface just like a cue ball breaking a rack of balls in billiards). On which electrode is sputtering most pronounced? Take a guess, then click here for the answer.
Because arcs are on almost continuously, they are more nearly at thermal equilibrium than sparks. One can use arcs to carry out fractional distillation, where low-boiling elements are compounds are sampled before higher-boiling substances. While hardly free of matrix effects, sparks are more likely to sputter all elements at once than arcs.
Discharge Sources
Discharge initiation, control of sampling, and control of excitation requires control of discharge initiation timing and voltage, together with control of current delivery to the electrodes once the discharge has been ignited. A DC arc may be initiated by a single spark, with current maintained at a level high enough to avoid collapse of the discharge channel. An AC arc is reignited (with opposite polarity) every time the polarity of the power line reverses sign (100 Hz in most of the world, 120 Hz in North America). Sparks may be triggered either in synchrony with the power line, by monitoring the potential build-up across the spark gap so that the initiation conditions are precisely sustained, or by triggering at a controlled frequency. All of these strategies are in common use.
With the exception of the AC arc, the first portion of the source connected to the electrical power grid is a full-wave rectifier, converting AC potential to DC.
Both components are supplied with DC voltage. Once the arc is lit, current is typically kept constant for the duration of observation.
In contrast, a spark combines trigger and waveform control into a single circuit. The high voltages and rapid transients constrain the types of control that are viable. The rapid voltage change when the discharge gap transitions from insulating to conducting in a matter of nanoseconds (dropping the spark gap voltage from several kilovolts to less than 100 V) can send a radio frequency shock propagating through electronics, wreaking havoc with unprotected circuitry. Triggers in modern equipment include quarter-wave line RF standing wave devices, hydrogen thyratrons shunted by diodes, and thyristor stacks. Current wave-shaping may be passive (resistors, capacitors, and inductors only), semi-passive (the passive elements plus diodes), or active (transistor-controlled current injection). Regardless of the hardware, the critical issues are maintanence of discharge gap polarity and control of current/time waveform. Only the cathode is significantly sampled, so a unidirectional waveform is commonly preferred (early sources using only passive components frequently had damped oscillatory discharges that sampled both electrodes). While laser ablation has many characteristics in common with a unidirectional spark, the source cost is usually significantly higher. For a thorough discussion of one variety of spark source, see D. M. Coleman and J. P. Walters, "An Electronic, Adjustable-Waveform Spark Source for Basic and Applied Emission Spectrometry," Spectrochim. Acta Part B: Atomic Spectrosc. 31B, 547-587 (1976).
Sample Vaporization
Fractional distillation, dislodgement followed by vaporization, or sputtering? These are the mechanisms that have been put forth for sampling of conductive solids by arcs and sparks. The three mechanisms are illustrated in the following figure:
Inset A shows heat flux from the hot plasma bathing the cathode. The lowest boiling substances are removed and atomized, later to be excited and ionized. In this model, the temperature of the cathode gradually rises, so that inclusions with different melting points gradually are vaporized in sequence. The DC arc matches this behavior most closely. While one can distinguish NaCl (low boiling) from ZrO2 (high boiling), there is enough sputtering and temperature inhomogeneity that a strict quantitation by speciation is not feasible.
Inset B shows the ideal that has been pursued by atomic spectroscopists for 2 centuries: a magic bullet excavates a highly controlled volume of the solid electrode, independent of composition. The atoms are separated into a spatially isolated cloud of free atoms. The isolated atoms then are excited and emit light, independent of their history. While femtosecond pulse laser ablation comes close to this ideal, no conductive plasma source can achieve this behavior. Consequently, solid sampling results in matrix effects, where elements are differentially sampled, based on thermal history, concommitant elements, and specimen microstructure.
Inset C shows sputtering, where cations bombard the cathode, removing the top layer or two of atoms per the billiards analogy already given. Glow discharges at a fraction of atmospheric pressure come close to this behavior, as do sparks during the first few nanoseconds following argon ionization (when ion energies are higher than 100 V and bulk heating of the electrode is as yet negligible).
A number of embellishments have broadened the ranges of samples that can be vaporized and excited by sparks and arcs. These include spraying solutions into a free-running spark (an approach largely supplanted by ICP emission), flowing solutions over a conducting electrode for "liquid layer" sampling or calibration (R. M. Barnes and H. V. Malmstadt, "Liquid-Layer-on-Solid-Sample Technique for Emission Spectrochemical Analysis," Anal. Chem. 46(1), 66-72 (1974).), and use of sparks to elutriate powders into e.g. flames (the shock wave as the spark ignites blowing powders away from the point where the spark attacks the electrodes). Of greater importance has been variations that allow sampling of insulating or refractory materials. Samples such as glass, ceramics, or dried, powdered biological materials are mixed with "fluxes," either LiCO3 or graphite. Ionic compounds and graphitic carbon have sufficient conductivity (which increases as temperature rises) that ionic bombardment and sputtering can occur around the insulating specimen powders. Graphite melts at a sufficiently high temperature that all elements can be vaporized and excited. LiCO3 is most useful for low-boiling elements that do not form refractory oxides (examples: Ni, Cu, Sn). While graphite furnace absorption and slurry or powder aspiration into the ICP have become common successors to arc sampling of refractories, use of modern detectors with arc sampling is still a viable means of accurately quantifying elemental composition of high-melting samples.
Direct
External standard and internal standard working curves are the primary approaches employed to elemental quantification. Because many elements, particularly the transition elements, rare earths, and actinides, have rich spectra, even high-resolution spectra must account for line interferences (lines from more than one element impinging on a particular detector). Continuum emission almost always is observable, so background subtraction is required. For any given wavelength:
\[I_\lambda = \sum_{i=1}^{92}I_{i,\lambda}C_i + I_{\lambda,continuum}\]
Ii,λ is a coefficient expressing the contribution of element i at the wavelength in question. Element i is at concentration Ci. The observable, Ii can only be interpreted in terms of concentration if the Ii,λ coefficients have been measured from known samples and Iλ, continuum can be estimated from interpolation of emission at wavelengths where no line emission occurs.
Complicating the situation is variance of each of the fitting parameters, so that a large quantity of some element, j, that has a large Ij,λ at the strongest emission wavelength of element i that is present at low concentration may make precise estimation of Ci difficult. Nonlinearities from self-reversal, self-absorption, or influence of sample on plasma temperature are assumed to be negligible.
A variation on the internal standard, composition normalization, has shown some success at overcoming sample uptake variability. One starts with an external working curve to obtain a first approximation to sample composition. The amounts of each element are estimated. From this estimate, one determines what interferences are expected (especially emission line overlaps), includes the effect of interferences, then normalizes the total sample composition to 100%. An obvious problem is that oxygen or nitrogen may be part of a sample and the amount of these elements must be accurately determined in a world where contamination from atmospheric oxygen and nitrogen is difficult to avoid. One must also determine the concentrations of elements with few emission lines in the UV or visible parts of the spectrum, F, Cl, Br, and I.
Complicating things still further is the spatiotemporal heterogeneity of arcs and, especially, sparks. In the first 50 ns of spark emission, the species primarily seen are atomic ions. For easily ionized elements, one may see multiply charged ions (for example, Al2+ is readily observed when sparking to aluminum alloys). By 1 µs into a spark, neutral species or singly-charged ionic emission dominates. The signal-to-continuum ratio maximizes several microseconds after discharge initiation. If a detector has sufficient temporal resolution, boxcar integration thus improves signal-to-background and signal-to-noise ratios. Additionally, the contact point of arcs and sparks tend to wander over the surface of an electrode, spreading emitting atoms over a significant volume of space and varying the precise alignment of atomic emission with respect to the optical axis of the observation system. Heroic efforts in the research lab have allowed sparks to be stabilized so that the most intense continuum emission, centered on the axis of the spark, may be optically masked (see e.g. US patent 4,393,327, July 12, 1983). Even so, the heterogeneity of most specimens leads to significant spark-to-spark intensity variations. A number of spectrometer vendors allow measurement of the statistical distribution of spark-to-spark variation and correlation of such variance with sample properties.
Indirect
Separation of sampling and excitation has been a popular research area since the late 1970's. Sparks and arcs can be used to generate sample aerosols which can be wafted into other sources including other sparks, inductively-coupled plasmas, or flames. The advantage of separating sampling from excitation is that source parameters may be optimized for sampling alone without worrying about continuum generation or complete sample vaporization. All that is needed is to have the aerosol represent the composition of the bulk solid. The excitation source need only vaporize the aerosol and generate linear working curves; it need not sample. The operational assets of separate optimization are counterbalanced by the extra cost of dual sources and memory effects of particles depositing in various parts of the instrument and later elutriated during observation of a successive sample. See for example: D. M. Coleman, M. A. Sainz, and H. T. Butler, Anal. Chem. 52, 746-753 (1980).