Sexta problemática: Análisis de los hábitos alimentarios generales analizados

SPECTROSCOPYlI

Atomic absorption spectroscopy finds con-siderably wider use than either of the other two atomic spectroscopic methods because it is best suited to routine analyses in the hands of relatively unskilled operators.

Sources for

Atomic Absorption Methods

Analytical methods based on atomic absorp-tion are potentially highly specific because atomic absorption lines are remarkably narrow and because electronic transition energies are unique for each element. On the other hand, the limited line widths create a measurement problem not encountered in molecular ab-sorption. Recall that Beer's law applies only for monochromatic radiation; a linear relation-ship between absorbance and concentration, however, can be expected only if the bandwidth of the source is narrow with respect to the width of the absorption peak (p. 155). No ordinary monochromator is capable of yielding a band' of radiation as narrow as the peak width of an atomic absorption line (0.002 to 0.005 nm). Thus, when a continuous source is employed with a monochromator for atomic absorption, only a minute fraction of the radi-ation is of a wavelength that is absorbed,

• Reference books on atomic absorption spectroscopy includl: G. F. Kirkbright and M. Sargent, Atomic Absorption aIIII Fluorescence Spectroscopy. New York:

Academic Press, 1974; J.W. Robinson, Atomic Absor~

lion S~rrroscopy, 2d ed. New York: Marcel Dekker, 1975; and W. Slavin. Alomic Absorption S~crroseopy, 2d ed. New York: Interscience, 1978.

and the relative change in intensity of the emergent band is small in comparison to the change suffered by radiation that actually cor-responds to the absorption peale. Under these conditions, Beer's law is not followed; in ad-dition, the sensitivity of the method is lessened significantly.

This problem has been overcome by em-ploying a source of radiation that emits a line of the same wavelength as the one to be used for the absorption analysis. For example, if the S89.6-nm line of sodium is chosen for the absorption analysis of that element, a sodium vapor lamp is useful as a source. Gaseous sodium atoms are excited by electrical dis-charge in such a lamp; the excited atoms then emit characteristic radiation as they return to lower energy levels. An emitted line will thus have the same wavelength as the resonance absorption line. With a properly designed source (one that operates at a lower tempera-ture than the flame to minimize Doppler broadening1 the emission lines will have band-widths significantly narrower than the absorp-tion bandwidths. It is only necessary, then, for the monochromator to have the capability of isolating a suitable emission line for the ab-sorption measurement (see Figure 11-11).The radiation employed in the analysis is thus sufficiently limited in bandwidth to permit measurements at the absorption peak. Greater sensitivity and better adherence to Beer's law result.

A separate lamp source is needed for each element (or sometimes, group of elements).

To avoid this inconvenience, attempts have been made to employ a continuous source with a very high resolution monochromator or, alternatively, to produce a line source by introducing a compound of the element to be determined into a high-temperature flame.

Neither of these alternatives is as satisfactory as individual lamps.

Hollow Cathode Lamps. The most common source for atomic absorption measurements is the hollow cathode lamp, which consists of a tungsten anode and a cylindrical cathode

Emiaion spectrum

of source

and produce II.I1 atomic cloud; this process is called sputtering. A portion of the sputtered metal atoms are in excited states and thus emit their characteristic radiation in the usual way. Eventually, the metal atoms'!iift'use back to the cathode surface or the glass walls of the tube and are redeposited.

The cylindrical configuration of the cath-ode tends to concentrate tlie radiation in a limited region of the tube; this design also enhances the probability that redeposition will occur at the cathode rather than on the glass walls.

The efficiency of the hollow cathode lamp depends upon its geometry and the operating potential. High potentials, and thus high cur-rents, lead to greater intensities. This advan-tage is 01Tset somewhat by an increase in Doppler broadening of the emission lines.

Furthermore, the greater currents result in an increase in the number of unexcited atoms in the cloud; the unexcited atoms, in turn, are capable of absorbing the radiation emitted by the excited ones. This se!{-absorption leads to lowered intensities, particularly at the center of the emission band.

A variety of hollow cathode tubes are available commercially. The cathodes of some consist of a mixture of several metals; such lamps permit the analysis of more than a

single element. .

II

i

Emission spectrum I efter passage

Absorption of a resonance

sealed in a glass tube that is filled with neon or argon at a pressure of 1 to 5 torr (see Figure 11-12).The cathode is constructed of the metal whose spectrum is desired or serves to support a layer of that metal.

Ionization of the gas occurs when a poten-tial is applied across the electrodes, and a cur-rent of about 5to 10mA is generated as the ions migrate to the electrodes. If the potential is sufficiently large, the gaseous cations ac-quire enough kinetic energy to dislodge some of the metal atoms from the cathode surface

Gila NeorAr

shield at 1·5 torr

FIGURE 11-12 Schematic cross section of a hollow cathode lamp.

GlISeOlB l>isc:hlrge Lamps. Gas discharge lamps produce a line spectrum as a con-sequence of the passage of an electrical cur-rent through a vapor of metal atoms; the familiar sodium and mercury vapor lamps are examples. Sources of this kind are partic-ularly useful for producing spectra of the alkali metals.

Source ModuiatiolL In the typical atomic absorption instrument, it is necessary to elim-inate interferences caused by;emission of ation by the flame. Most of the emitted radi-ation can be removed by locating the mono-chromator between the flame and the detector;

nevertheless, this arrangement does not remove the flame radiation correspoBding to the wave-length selected for the analysis. The flame will contain such radiation, since excitation and radiant emission by some atoms of the analyte can occur. This difficulty is overcome by modulating the output ofihe source so that its intensity fluctuates at a atnstant frequency.

The detector then receives two types of signal, an alternating one from the source' and a continuous one from the flame. These signals are converted to the corresponding types of electrical response. A simple high-pass, RC filter (p. 23) is then employed to remove the unmodulated dc signal and pass the ac signal for amplification.

A simple and entirely satisfactory way of modulating radiation from the source is to interpose a circular disk in the beam between the source and the flame. Alternate quadrants of this disk are removed to permit passage of light. Rotation of the disk at constant speed provides a beam that is. chopped to the desired frequency. As all; alternative, the power supply for the source can be designed for intermittent or ac operafion.

Instruments for Atomic Absorption Spectroscopy

Instruments for atomic absorption work are offered by numerous manufacturers. The range of sophistication and cost (upward

PRINCIPLES OF INSTR~NTAL ANALYSIS

Czemey-Turner monochromator

~--!;:---~

I I

I I

t t

I Flame, ~

in the several reference works listed in foot-note 5 (p. 315).

SelBitil'ity •••• Detection Limits. Two terms are employed in characterizing atomic absorption methods. The sensitivity is defined as the concentration of an element in/lg!ml (or ppm) which produces a transmittance signal of 0.99 or a corresponding absorbance signal of 0.0044. Modern atomic instruments have adequate precision to discriminate be-tween absorbance signals that differ by less than 0.0044. For this reason, the term detec-tion limit has been introduced, which is defined as the concentration of the element that pro-duces an analytical signal equal to twice the standard deviation of the background signal.

(For flame atomimtion, the standard deviation of the background signal is obtained byobserv-ing the signal variation when a blank is sprayed FIG U RE 11-13 A typical double-beam atomic absorption

spectro-photometer.

from a few thousand dollars) is substantial; as always, the potential user ~ust select the most appropriate design for the intended use.

S •..••••• Spec:t~ometers. A typical single-beam instrument for multielement anal-yses consists of several hollow cathode sources, a chopper, an atomizer, and a simple grating spectrophotometer with a photomultiplier transducer. It is used in the same way as a singlc-beam instrument for molecular absorp-tion work. Thus, the dark current is nulled with a shutter in front of the transducer. The 100% T adjustment is then made while a blank is aspirated into the flame (or ignited in

• nonflame atomizer). Finally, the transmit-tance is obtained with the sample replacing the blank.

Single-beam atomic absorption instruments have the same advantageS and disadvantages as their molecular absorption counterparts, which were discussed on page 185.

Double-... SpectroPbotometers. Figure 11-13 is a schematic diagram of a typica1 double-beam instrument. The beam from the hollow cathode source is split by a mirrored chopper, one half passing through the flame and the other half around it. The two beams are then recombined by a half-silvered mirror

and passed into a Czerney-Turner grating monochromator; a photomultiplier tube serves as the transducer. The output from the latter is fed to a lock-in amplifier which is synchron-ized with the chopper drive. The ratio between the reference and sample signal is then ampli-fied and fed to the readout which may be a mete~,digital device, or recorder. Alternatil'ely, the amplified signal from the reference beam may be attenuated to match the sample signal by means of a potentiometer; the transmittance or absorbance is then read from the position of the slide wire contact.

It should be noted that the reference beam in atomic absorption instruments does not pass through the flame and thus does not cor-rect for loss of radiant power due to absorp-tion or scattering by the flame itself.

into the flame.) Both the sensitivity and the detection limiU vary widely with such variables as flame temperature, spectral bandwi.dth, detector sensitivity, and type of signal pro-cessing. Small differences among quoted values of the two parameters are not significant; for example, whereas an order of magnitude differ-ence is certainly meaningful, a factor of 2 or 3 is probably not.

Detection limits ranging from about 3 x 10-4 ppm to 20 ppm are observed for the various metallic elements when flame atomi-zation is employed in atomic absorption. Non-flame atomization often enhances this limit bya factor of 10 to 1000.

Columns three and four of Table 11-3 list detection limits for several common elements by flame and nonflame atomimtion proce-dures.

TABLE 11-3 DETECTION LIMITS FOR THE ANALYSIS OF

SELECTED ELEMENTS BY ATOMIC ABSORPTION AND

FLAME EMISSION SPECTROMETRY

Detection Limit, Ilgfml

Wavelength, Noa8ame Flame Flame

Element

•••

Absorptioa" AlJsorptionb EmisskMf

Aluminum 396.2 0.03 0.005 (NzO)

309.3 0.1 (N O~

Calcium 422.7 0.0003 0.002 (air 0.005 gir)

Cadmium 326.1 0.0001 2 (Nz )

228.8 0.005 (air)

Chromium 425.4 0.005 0.005 (NzO)

357.9 0.005 (air)

Iron 372.0 0.003 0.05 (NzO)

248.3 0.005 lairl

Lithium 670.8 0.005 0.005 air 0.ססOO3(N~O)

Magnesium 285.2 0.00006 0.ססOO3(air) 0.005 (NzO

Potassium 766.5 0.0009 0.005lair~ 0.0005 lairl

Sodium 589.0 0.0001 0.002 air 0.0005 air

Applications of Atomic Absorption Spectroscopy

Atomic absorption spectroscopy is a sensitive means for the determination of more than 60 elements. Details concerning the methods for sample preparation and quantitative deter-mination of individual elements are available

• Data from J.W. Robinson and P.J.Slevin. Ammcan Laboratory,4 (8~ 14 (1972~ With permissioD, International Sc:ientific Communications. Inc. Fairfield, crC1972; waveJensth not reported. Copy-riPt 1972 by International Scientific Communications. Inc.

• Taken from data compiled by E. E. Pickett and S. R. KoirtyohanD, Anal.Chem.41 (14~ 28A (1969) and reprinted with permission. Copyright by the American Chemical Society. Data are for an acetylene flame with the oxidant shown in parentheses.

Accurecy

Under usual conditions, the relative err~r as-sociated with a flame absorption analysis is of the order of 1 to 2%. With special precau-tions, this figure can be lowered to a few tenths of 1

%.

Spectral Interferences

Interferences of two types are encountered in atomic absorption methods. Spectral inter-ferences arise when the absorption of an interfering species either overlaps or lies so close to the analyte absorption that resolu-tion by the monochromator becomes impos-sible.Chemical interferencesresult from various chemical processes occurring during atomiza-tion that alter the absorpatomiza-tion characteristics of the analyte. A brief discussion of spectral interferences follows; sources of chemical inter-ference are considered in the next section ..

Because the emission lines of hollow cath-ode sources are so very narrow, interference due to overlap of atomic spectral lines is rare.

For such an interference to occur, the separa-tion between the two lines would have to be less than perhaps 0.1

A.

For example, a vana-dium line at 3082.11

A

interferes in an anal-ysis b8sed upon the aluminum absorption line at 3082.lS

A.

The interference is readily avoided, however, by employing the aluminum line at 3092.7

A

instead.

Spectral interferences also result from the presence of combustion products that exhibit broad band absorption or particulate products that scatter radiation. Both diminish the power of the transmitted beam and lead to positive analytical errors. Where the source of these products is the fuel and oxidant mixture alone, corrections are readily obtained from ab-sorbance measurements while a blank is aspi-rated into the flame. Note that this correction must be employed with a double-beam as well as a single-beam instrument because the refer-ence beam of the former does not pass through the flame (see Figure 11-13).

A much more troublesome problem is en-countered when the source of absorption or scattering originates in the sample matrix;

here, the power of the transmitted beam, P, is reduced by the nonanalyte components of the sample matrix, but the incident beam power, Po is not; a positive error in absorbance and thus concentration results. An example of a potential matrix interference due to absorp-tion occurs in the determinaabsorp-tion of barium in alkaline-earth mixtures. :As shown by the dotted line in Figure 11-5, the wavelength of the barium line used for; atomic absorption analysis appears in the center of an absorp-tion band for CaOH; clearly, interference of calcium in a barium arlaIysis is to be ex-pected. In this particular situation, the elfeet is readily eliminated by substituting nitrous oxide for air as the oxidant for the acetylene;

the higher temperature decomposes the CaOH and eliminates the: absorption band.

Spectral interference ctpe to scattering by products of atomization, is most often en-countered when concentrated solutions con-taining elements such as Ti, Zr, and W-which form refractory oxides-are aspirated into the flame. Metal oxide particles with diameters greater than the wavelength of light appear to be formed; significant scattering of the incident beam results.

Fortunately, spectral interferences by matrix products are not widely encountered and often can be avoided by variations in the analytical parameters, such as temperature and fuel-to-oxidant ratio. Alternatively, if the source of interference is known, an excess of the inter-fering substance can be ad.ded to both sample and standards; provided the excess is large with respect to the concentration from the sample matrix, the contribution iof the latter will become insignificant. The

1

added substance is sometimes ca1Ied aradiation buffer.

The foregoing technique is not effective with complex samples because the source of the interference may not be known; a back-ground correction must, therefore, be

em-ployed.6 A description of some of the most commonly used techniques follows.

The Two-Liae Correctioa Method. The two-line correction procedure requires the presence of a reference line from the source; this line should lie as close as possible to the anaIyte line but must not be absorbed by the analyte.

If these conditions are met, it is assumed that any decrease in power of the reference line from that observed during calibration arises from absorption or scattering by the matrix products of the sample; this decrease is then used to correct the power of the analyte line.

The reference line may be from an impur-ity in the hollow cathode lamp, a neon or argon line from the gas contained in the lamp, or a nonresonant emission line of the element being determined. .

Unfortunately, a suitable reference line is often not available.

The ContiDuous-Soan Correction Metholl.

A second method for background correction is available with suitably equipped double-beam instruments. Here, a hydrogen or deute-rium lamp is employed as a source of continu-ous radiation throughout the ultraviolet re-gion. The configuration of the chopper is then rearranged so that the reference beam shown in Figure 11-13 is dispensed with, andinstcad, radiation from the continuous source and the hollow cathode lamp are passed alternatively , through the flame. The power of the beams

from the two sources is then compared rather than the power of a sample and reference beam from the hollow cathode source. The slit width is kept sufficiently wide so that the fraction of the continuous source that is ab-sorbed by the atoms ofthe sample is negligible.

Thus, the attenuation of its power during passage through the flame reflects only the

• For a critical discussion of the various methods for bacltsround correction, see: A. T. Zander,Amer.Lab. 8 (11~ 11 (1976).

broad band absorption or scattering by the flame components. A background correction is thus achieved.

The Zeeman Elect Correction MetbocL One commercial atomic absorption instrument achieves a background correction by taking advantage of the Zeeman effect to split the anaIyte line into two components, one of which is displaced from the other by a small wave-length increment (about O.ol nm). The two components are also polarized at 90 deg to one another and can be monitored alterna-tively by inserting a rotating polarizer into the beam path. Here, the displaced compo-nent is sufficiently separated from the absorp-tion peak to provide a means of correcabsorp-tion for background absorption or scattering.

Zeeman splitting can be accomplished by exposing the atomizer device or the light source to a strong magnetic field.7

Chemical Interferences

Chemical interferences are more common than spectral ones. Their effects can frequently be minimized by suitable choice of conditions.

Both theoretical and experimental evidence suggest that many of the processes occurring in the mantle of a flame are in approximate equilibrium. As a consequence, it becomes possible to regard the burned gases ofthe flame as a solvent medium to which thermodynamic calculations can be applied. The processes of principal interest include formation of com-pounds onow volatility, dissociation reactions, and ionizations.

Formatioa of Compounds of Low Volatility.

Perhaps the most common type of inter-ference is by anions which form compounds

, For a detailed discussion of the application of the Zeeman elfect to atomic absorption, see: T. Hadeilhi and R. D. McLaupwn, SCWlta, 17'" <lO4(1974); T. Hadeilbi and R. D. McLaulh\in, A••••I.

C"-.

48, 1009 (1976); H.

Koizuma and K. Yasuda,A ••••I.

C"-.

48, 1178 (t976);

and S. D. Brown,A ••••I.

C"-.'"

(14~ 1269A(1977~

of low volatility with the analyte and thus reduce the rate at which it is atomized. Low results are the consequence. An example is the decrease in calcium absorbance that is ob-served with increasing concentrations of sul-fate or phosphate. For a fixed calcium con-centration, the absorbance is found to fan off nearly linearly with increasing sulfate or

of low volatility with the analyte and thus reduce the rate at which it is atomized. Low results are the consequence. An example is the decrease in calcium absorbance that is ob-served with increasing concentrations of sul-fate or phosphate. For a fixed calcium con-centration, the absorbance is found to fan off nearly linearly with increasing sulfate or

In document COMPLUTENSE DE MADRID UNIVERSIDAD (página 182-191)