|
|
|
|
ATOMIC ABSORPTION
INSTRUMENTATION
THE BASIC COMPONENTS |
| |
To understand the workings of the atomic absorption spectrometer,let us build
one,piece by piece.Every absorption spectrometer must have components which
fulfill the three basic requirements .There must be:(1)a light
source;(2)a sample cell;and (3)a means of specific light measurement.
In atomic absorption,these functional areas are implemented by the components
illustrated in Figure 2-2.A light source which emits the sharp atomic lines of the
element to be determined is required.The most widely used source is the hollow
cathode lamp.These lamps are designed to emit the atomic spectrum of a particu-lar
element,and specific lamps are selected for use depending on the element to
be determined.
..It is also required that the source radiation be modulated (switched on and off rap-idly)
to provide a means of selectively amplifying light emitted from the source
lamp and ignoring emission from the sample cell.Source modulation can be ac-complished
with a rotating chopper located between the source and the sample
cell,or by pulsing the power to the source.
Special considerations are also required for a sample cell for atomic absorption.
An atomic vapor must be generated in the light beam from the source.This is gen-erally
accomplished by introducing the sample into a burner system or electrically
heated furnace aligned in the optical path of the spectrophotometer.
Several components are required for specific light measurement.A monochroma-tor
is used to disperse the various wavelengths of light which are emitted from the
source and to isolate the particular line of interest.The selection of a specific
source and a particular wavelength in that source is what allows the determination
of a selected element to be made in the presence of others.
The wavelength of light which is isolated by the monochromator is directed onto
the detector,which serves as the ''eye''of the instrument.This is normally a
photomultiplier tube,which produces an electrical current dependent on the light
intensity.The electrical current from the photomultiplier is then amplified and
processed by the instrument electronics to produce a signal which is a measure of
the light attenuation occurring in the sample cell.This signal can be further proc-essed
to produce an instrument readout directly in concentration units.
LIGHT SOURCES
An atom absorbs light at discrete wavelengths.In order to measure this narrow
light absorption with maximum sensitivity,it is necessary to use a line source,
which emits the specific wavelengths which can be absorbed by the atom.Narrow
line sources not only provide high sensitivity,but also make atomic absorption a
very specific analytical technique with few spectral interferences.The two most
common line sources used in atomic absorption are the ''hollow cathode lamp''
and the ''electrodeless discharge lamp.''
The Hollow Cathode Lamp
The hollow cathode lamp is an excellent,bright line source for most of the ele-ments
determinable by atomic absorption.Figure 2-3 shows how a hollow cathode
lamp is constructed.The cathode of the lamp frequently is a hollowed-out cylinder
of the metal whose spectrum is to be produced.The anode and cathode are sealed
in a glass cylinder normally filled with either neon or argon at low pressure.At
the end of the glass cylinder is a window transparent to the emitted radiation.
The emission process is illustrated in Figure 2-4.When an electrical potential is
applied between the anode and cathode,some of the fill gas atoms are ionized.The
positively charged fill gas ions accelerate through the electrical field to collide
with the negatively charged cathode and dislodge individual metal atoms in a proc-ess
called ''sputtering''.Sputtered metal atoms are then excited to an emission
state through a kinetic energy transfer by impact with fill gas ions.
Hollow cathode lamps have a finite lifetime.Adsorption of fill gas atoms onto the
inner surfaces of the lamp is the primary cause for lamp failure.As fill gas pressure
decreases,the efficiency of sputtering and the excitation of sputtered metal atoms
also decreases,reducing the intensity of the lamp emission.To prolong hollow
cathode lamp life,some manufacturers produce lamps with larger internal vol-umes
so that a greater supply of fill gas at optimum pressure is available.
Figure 2-3.Hollow cathode lamp.
Figure 2-4.Hollow cathode lamp process,where Ar +is a positively-charged ar-gon
ion,M o is a sputtered,ground-state metal atom,M*is an excited-state metal
atom,and l is emitted radiation at a wavelength characteristic for the sputtered
metal.
.The sputtering process may remove some of the metal atoms from the vicinity of
the cathode to be deposited elsewhere.Lamps for volatile metals such as arsenic,
selenium,and cadmium are more prone to rapid vaporization of the cathode during
use.While the loss of metal from the cathode at normal operating currents (typi-cally
5-25 milliamperes)usually does not affect lamp performance,fill gas atoms
can be entrapped during the metal deposition process which does affect lamp life.
Lamps which are operated at highly elevated currents may suffer reduced lamp
life due to depletion of the analyte element from the cathode.
Some cathode materials can slowly evolve hydrogen when heated.As the concen-tration
of hydrogen in the fill gas increases,a background continuum emission
contaminates the purity of the line spectrum of the element,resulting in a reduction
of atomic absorption sensitivity and poor calibration linearity.To eliminate such
problems,most modern hollow cathode lamps have a tantalum ''getter''on the an-ode
which irreversibly adsorbs evolved hydrogen as the lamp is operated.
The cathode of the hollow cathode lamp is usually constructed from a highly pure
metal resulting in a very pure emission spectrum of the cathode material.It is
sometimes possible,however,to construct a cathode or cathode insert from several
metals.The resulting ''multi-element''lamp may provide superior performance
for a single element or,with some combinations,may be used as a source for all
of the elements contained in the cathode alloy.However,not all metals may be
used in combination with others because of metallurgical or spectral limitations.
Special consideration should be given before using a multi-element lamp as ana-lytical
complications may result.Often the intensity of emission for an element
in a multi-element lamp is not as great as that which is observed for the element
in a single-element lamp.This loss of intensity could be a disadvantage in appli-cations
where high precision or low detection limits are required.The increased
spectral complexity of multi-element lamps may require that alternate wave-lengths
or narrower slits be used,which may also adversely affect sensitivity or
baseline noise.
Each hollow cathode lamp will have a particular current for optimum perform-ance.
In general,higher currents will produce brighter emission and less baseline
noise.As the current continues to increase,however,lamp life may shorten and
spectral line broadening may occur,resulting in a reduction in sensitivity and lin-ear
working range.The recommended current specified for each lamp will usually
provide the best combination of lamp life and performance.For demanding analyses requiring the best possible signal-to-noise characteristics,higher currents can
be used for the lamp,up to the maximum rated value.Lower lamp currents can
be used with less demanding analyses to prolong lamp life.
Confusion over exactly what current is being used for a hollow cathode lamp may
occur due to the method used for lamp modulation.As explained earlier,the source
for atomic absorption must be modulated in order to accomplish selective ampli-fication
of the lamp emission signal.This can be accomplished mechanically by
using a rotating chopper or electronically by pulsing the current supplied to the
lamp,as illustrated in Figure 2-5.Both methods produce similar results;however,
in some instruments the use of electronic modulation may create the impression
that a lower lamp current is being applied than is actually the case.
The cause for the apparent difference in measured currents with mechanically and
electronically modulated systems is also shown in Figure 2-5.For mechanical
modulation,the lamp is run at a constant current.Under these conditions,an am-meter
reading of lamp current will indicate the actual current flow.For electronic
modulation,the current is switched on and off at a rapid rate.An ammeter nor-mally
will indicate the time-averaged current rather than the actual peak current
which is being applied.
While some instruments are designed to apply a correction factor automatically
to electronically modulated lamp current readings to provide true peak current val-ues,
many do not.For electronically modulated systems without such correction,
the actual peak current can be approximated from the measured current by dividing it by the ''duty cycle'',the fraction of time that current is applied to the lamp.
For example,for a duty cycle of 40%and a measured lamp current of 10 milliam-peres,
the actual peak operating current for an electronically modulated system is:
10 milliamperes/0.4 =25 milliamperes
Specified lamp current settings may appear to be lower for atomic absorption in-struments
which modulate the source electronically and do not apply correction.
The only valid basis of comparison between the current settings used by two dif-ferent
systems is one which includes compensation for the duty cycle,as shown
above.
The Electrodeless Discharge Lamp
For most elements,the hollow cathode lamp is a completely satisfactory source
for atomic absorption.In a few cases,however,the quality of the analysis is im-paired
by limitations of the hollow cathode lamp.The primary cases involve the
more volatile elements where low intensity and short lamp life are a problem.The
atomic absorption determination of these elements can often be dramatically im-proved
with the use of brighter,more stable sources such as the ''electrodeless dis-charge
lamp''.
Figure 2-6 shows the design of the Perkin-Elmer System 2 electrodeless discharge
lamp (EDL).A small amount of the metal or salt of the element for which the
source is to be used is sealed inside a quartz bulb.This bulb is placed inside a small,
self-contained RF generator or ''driver''.When power is applied to the driver,an
RF field is created.The coupled energy will vaporize and excite the atoms inside
the bulb,causing them to emit their characteristic spectrum.An accessory power
supply is required to operate an EDL.
Figure 2-6.Electrodeless discharge lamp.
.Electrodeless discharge lamps are typically much more intense and,in some cases,
more sensitive than comparable hollow cathode lamps.They therefore offer the
analytical advantages of better precision and lower detection limits where an
analysis is intensity limited.In addition to providing superior performance,the
useful lifetime of an EDL is typically much greater than that of a hollow cathode
lamp for the same element.It should be noted,however,that the optical image for
the EDL is considerably larger than that in a hollow cathode lamp.As a result,the
performance benefits of the EDL can only be observed in instruments with optical
systems designed to be compatible with the larger image.
Electrodeless discharge lamps are available for a wide variety of elements,includ-ing
antimony,arsenic,bismuth,cadmium,cesium,germanium,lead,mercury,
phosphorus,potassium,rubidium,selenium,tellurium,thallium,tin and zinc.
OPTICAL CONSIDERATIONS
Photometers
The portion of an atomic absorption spectrometer's optical system which conveys
the light from the source to the monochromator is referred to as the photometer.
Three types of photometers are typically used in atomic absorption instruments:
single-beam,double-beam and what might be called compensated single-beam or
pseudo double-beam.
Single-Beam Photometers
The instrument diagrammed in Figure 2-7 represents a fully functional ''single-beam''
atomic absorption spectrometer.It is called ''single-beam''because all
measurements are based on the varying intensity of a single beam of light in a sin-gle
optical path.
The primary advantage of a single-beam configuration is that it has fewer com-ponents
and is less complicated than alternative designs.It is therefore easier to
construct and less expensive than other types of photometers.With a single light
path and a minimum number of optical components,single-beam systems typi-cally
provide very high light throughput.The primary limitation of the single-beam
photometer is that it provides no means to compensate for instrumental
variations during an analysis,such as changes in source intensity.The resulting
signal variability can limit the performance capabilities of a single-beam system.
Double-Beam Photometers
An alternate photometer configuration,known as ''double-beam''(Figure 2-8)
uses additional optics to divide the light from the lamp into a ''sample beam''(di-rected
through the sample cell)and a ''reference beam''(directed around the sam-ple
cell).In the double-beam system,the reference beam serves as a monitor of
lamp intensity and the response characteristics of common electronic circuitry.
Therefore,the observed absorbance,determined from a ratio of sample beam and
reference beam readings,is more free of effects due to drifting lamp intensities
and other electronic anomalies which similarly affect both sample and reference
beams.
Modern atomic absorption spectrometers are frequently highly automated.They
can automatically change lamps,reset instrument parameters,and introduce sam-ples
for high throughput multielement analysis.Double-beam technology,which
automatically compensates for source and common electronics drift,allows these
instruments to change lamps and begin an analysis immediately with little or no
lamp warm-up for most elements.This not only reduces analysis time but also pro-longs
lamp life,since lamp warm-up time is eliminated.Even with manual analy-ses,
the ability to install a lamp or turn on the instrument and start an analysis
almost immediately is a decided advantage for double-beam systems.
Double-beam photometers do divert some source energy from the sample beam
to create the reference beam.Since it is the signal:noise ratio of the sample beam
which determines analytical performance,modern double-beam instruments typi-cally
devote a much higher percentage of the source emission to the sample beam
than to the reference beam.For example,a modern double-beam system which
uses a beam splitter to generate sample and reference beams may use 75%of the
source emission for the sample measurement and only 25%for the reference meas-urement.
Using such techniques,modern double-beam instruments offer virtually
the same signal-to-noise ratio as single-beam systems while enjoying the high-speed
automation benefits and operational simplicity of double-beam operation.
Alternative Photometer Designs
There are several alternative system designs which provide advantages similar to
those of double-beam optical systems and the light throughput characteristic of
single-beam systems.Such systems can be described as compensated single-beam
or pseudo double-beam systems.One such design uses two mechanically-adjusted
mirrors to alternately direct the entire output of the source through either the sam-ple
path (during sample measurements)or through a reference path (Figures 2-9
and 2-10).
These alternative photometer designs provide light throughput comparable to that
provided by single-beam photometer systems.They also compensate for system
variations in a manner similar to that of double-beam photometers----similar,but
not the same.This type of photometer performs compensation for drift much less
frequently than do double-beam systems,typically only once per analytical read-ing.
Double-beam systems typically provide drift compensation at rates in excess
of 50 times per second.The lower compensation frequency limits the ability of
alternative photometer systems to compensate for large,quickly changing vari-ations
in source intensity such as those that frequently occur when a source is first
lighted.
As previously discussed,an important factor in determining the baseline noise in
an atomic absorption instrument is the amount of light energy reaching the
photomultiplier (PMT).Lamp intensity is optimized to be as bright as possible
while avoiding line broadening problems.The impact of single-beam and double-beam
photometer systems has been discussed above.But the impact of other com-ponents
must also be considered to determine the capabilities of the complete
optical system.
Figure 2-9.A compensated single-beam system with source light directed
through the sample path.
Figure 2-10.A compensated single-beam system with source light directed
through the reference path.
.Light from the source must be focused on the sample cell and directed to the mono-chromator,
where the wavelengths of light are dispersed and the analytical line of
interest is focused onto the detector.Some energy is lost at each optical surface
along the way.Front-surfaced,highly reflective,mirrors can be used to control the
focus of the source lamp and the field of view of the light detector precisely and
with minimal light loss.Alternately,focusing can be accomplished by refraction
instead of reflection by using a lens system.Since the focal length of a lens varies
with wavelength,additional optics (which may further reduce energy throughput)
or complex optical adjustments must be used to obtain proper focus over the full
spectral range for atomic absorption.
Particular care must be taken in the monochromator to avoid excessive light loss.
A typical monochromator is diagrammed in Figure 2-11.Wavelength dispersion
is accomplished with a grating,a reflective surface ruled with many fine parallel
lines very close together.Reflection from this ruled surface generates an interfer-ence
phenomenon known as diffraction,in which different wavelengths of light
diverge from the grating at different angles.Light from the source enters the mono-chromator
at the entrance slit and is directed to the grating where dispersion takes
place.The diverging wavelengths of light are directed toward the exit slit.By ad-justing
the angle of the grating,a selected emission line from the source can be
allowed to pass through the exit slit and fall onto the detector.All other lines are
blocked from exiting.
.The angle of dispersion at the grating can be controlled by the density of lines on
the grating.Higher dispersion will result from greater line density,i.e.,more
lines/mm.High dispersion is important to good energy efficiency of the mono-chromator,
as illustrated in Figure 2-12.
The image of the source focused on the entrance slit and dispersed emission lines
at the exit slit are shown for both a low-dispersion and a high-dispersion grating.
In order to isolate a desired line from nearby lines,it is necessary to use a narrower
exit slit in the low-dispersion example than is required in the high-dispersion case.
Good optical design practices dictate that the entrance and exit slits be similarly
sized.The use of a larger entrance slit will overfill the grating with the source im-age,
while the use of a smaller entrance slit restricts the amount of light entering
the monochromator.Both reduce the amount of energy available at the exit slit.
For a low dispersion grating,this means that the size of the monochromator en-trance
slit is limited to the narrow size demanded of the exit slit to exclude nearby
lines.Thus,much of the available light energy is prevented from ever entering the
monochromator.In contrast,the greater wavelength separation provided by a
high-dispersion grating allows the use of wider slits,which make use of more of
the available light without any sacrifice in resolution.
To a first approximation,the energy throughput of a monochromator is propor-tional
to the illuminated ruled grating area and inversely proportional to the re-ciprocal
linear dispersion.To obtain the full energy benefit of high dispersion,it
is necessary to use a grating with a ruled surface area large enough to capture all
of the light from the magnified slit image.Large,quality gratings of high disper-sion
are difficult and expensive to make.Therefore,the incentive is great to accept
smaller gratings with lesser line densities and poorer dispersion for atomic absorp-tion
instrumentation.However,better instruments take advantage of the superior
energy throughput afforded by larger gratings.
.Another factor affecting the optical efficiency of the monochromator is the blaze
angle of the grating,whether it is mechanically ruled or holographically generated.
An illustration of a mechanically-ruled blaze angle appears in Figure 2-13.
Mechanical grating rulings are in the form of V-shaped grooves carved into the
surface of the grating.As discussed earlier,an interference phenomenon causes
light of different wavelengths to diverge from the grating at different angles.The
particular wavelength which diverges from the blazed surface at an angle corre-sponding
to specular reflectance (i.e.,angle of reflection equals angle of inci-dence)
will suffer the least loss in intensity as a result of the diffraction process.
A grating can be constructed for a blaze at any desired wavelength by controlling
the angle of cut during ruling.The farther removed a given wavelength of light is
from the wavelength for which a grating is blazed,the greater will be the extent
of monochromator light loss at that wavelength.
The useful atomic absorption wavelength range runs from 189 to 851 nanometers.
With one grating blazed somewhere in the middle of this range,significant energy
fall-off occurs at the wavelength extremities due to energy inefficiencies in the dif-fraction
process.One technique used to overcome this problem and to provide en-hanced
energy throughput at the wavelength extremities is to equip the instrument
with two gratings,one blazed in the ultraviolet and the other blazed in the visible
region of the spectrum.Then by choosing the grating blazed nearest the working
.wavelength,the optimum energy throughput can be achieved.Alternately,a single
''dual-blazed''grating can be used,with two regions blazed for the two spectral
regions.As the dual blazed grating rotates from one wavelength extreme to an-other,
the region blazed for the current working wavelength is brought into align-ment
with the optical beam,thereby offering improved efficiency compared with
a single grating blazed at one wavelength.
THE ATOMIC ABSORPTION ATOMIZER
Pre-Mix Burner System
The sample cell,or atomizer,of the spectrometer must produce the ground state
atoms necessary for atomic absorption to occur.This involves the application of
thermal energy to break the bonds that hold atoms together as molecules.While
there are several alternatives,the most routine and widely applied sample atomizer
is the flame.
Figure 2-14 shows an exploded view of an atomic absorption burner system.In
this ''premix''design,sample solution is aspirated through a nebulizer and sprayed
as a fine aerosol into the mixing chamber.Here the sample aerosol is mixed with
fuel and oxidant gases and carried to the burner head,where combustion and sam-ple
atomization occur.
.Fuel gas is introduced into the mixing chamber through the fuel inlet,and oxidant
enters through the nebulizer sidearm.Mixing of the fuel and oxidant in the burner
chamber eliminates the need to have combustible fuel/oxidant in the gas lines,a
potential safety hazard.In addition to the separate fuel and oxidant lines,it is ad-vantageous
to have an auxiliary oxidant inlet directly into the mixing chamber.
This allows the oxidant flow adjustments to be made through the auxiliary line
while the flow through the nebulizer remains constant.Thus,for a burner system
with an auxiliary oxidant line,the sample uptake rate is independent of flame con-dition,
and the need to readjust the nebulizer after every oxidant flow adjustment
is eliminated.
Only a portion of the sample solution introduced into the burner chamber by the
nebulizer is used for analysis.The finest droplets of sample mist,or aerosol,are
carried with the combustion gases to the burner head,where atomization takes
place.The excess sample is removed from the premix chamber through a drain.
The drain uses a liquid trap to prevent combustion gases from escaping through
the drain line.The inside of the burner chamber is coated with a wettable inert plas-tic
material to provide free drainage of excess sample and prevent burner chamber
''memory.''A free draining burner chamber rapidly reaches equilibrium,usually
requiring less than two seconds for the absorbance to respond fully to sample
changes.
Impact Devices
The sample aerosol is composed of variously sized droplets as it is sprayed into
the mixing chamber.Upon entering the flame,the water in these droplets is va-porized.
The remaining solid material must likewise be vaporized,and chemical
bonds must be broken to create free ground state atoms.Where the initial droplet
size is large,the sample vaporization and atomization process is more difficult to
complete in the short time in which the sample is exposed to the flame.Incomplete
sample vaporization and atomization will lead to increased susceptibility to ana-lytical
interferences.
Impact devices are used to reduce droplet size further and to cause remaining
larger droplets to be deflected from the gas stream and removed from the burner
through the drain.Two types of impact device are used typically,impact beads and
flow spoilers.
.Impact bead systems are normally used to improve nebulization efficiency,the
percentage of sample solution converted to smaller droplets.The impact bead is
normally a spherical bead made of glass,silica or ceramic.Glass or quartz impact
beads may be less corrosion resistant and may cause more contamination problems
than more chemically inert ceramic beads.
The impact bead is positioned directly in the nebulizer spray as it exits the nebu-lizer.
The sample spray contacts the impact bead at high speed,causing some of
the larger droplets to be broken up into smaller droplets.The design and position-ing
of the impact bead are critical in determining how well it will work.Properly
designed impact bead systems will improve nebulization efficiency and remove
many of the remaining large droplets from the spray.However,poorly designed
or positioned impact beads may have little or no effect on nebulization efficiency
and may be very inefficient at removing larger droplets from the spray.The in-creased
population of large droplets in the aerosol may create undesirable effects,
including poorer precision and increased interferences.Additionally,burner sys-tems
using an impact bead may exhibit memory problems with high concentration
solutions or solutions with high dissolved solids content.
The quality of an impact bead system can frequently be determined by the increase
in sensitivity it provides for selected elements.A poorly designed system will pro-vide
improved sensitivity for easily atomized elements simply because more sam-ple
is transported to the flame and less to the drain.However,there normally will
be little or no improvement in sensitivity for the less easily atomized elements.A
properly designed impact bead system will provide improved nebulization effi-ciency
and improved sensitivity for all elements.
Flow spoilers normally do not improve nebulization efficiency.The primary use
of a flow spoiler is to remove the remaining large droplets from the sample aerosol.
The flow spoilers used in atomic absorption burner systems normally are placed
between the nebulizer and the burner head.They typically have three or more large
vanes constructed from or coated with a corrosion resistant material.Smaller drop-lets
are transported through the open areas between the vanes while larger droplets
contact the vanes and are removed from the aerosol.
For routine atomic absorption analyses where maximum sensitivity is not re-quired,
use of an efficient flow spoiler alone will provide the required analytical
stability and freedom from interference.A burner system optimized for maximum
sensitivity and performance should include both a high nebulization efficiency ce-ramic
impact bead and an efficient flow spoiler.
Several important factors enter into the nebulizer portion of the burner system.In
order to provide efficient nebulization for all types of sample solution,the nebu-lizer
should be adjustable.Stainless steel has been the most common material used
for construction of the nebulizer.Stainless steel has the advantage of durability
and low cost but has the disadvantage of being susceptible to corrosion from sam-ples
with a high content of acid or other corrosive reagents.For such cases,nebu-lizers
constructed of a corrosion resistant material,such as an inert plastic,
platinum alloys or tantalum should be used.
Burner heads typically are constructed of stainless steel or titanium.All-titanium
heads are preferred as they provide extreme resistance to heat and corrosion.
Different burner head geometries are required for various flame or sample condi-tions.
A ten-centimeter single-slot burner head is recommended for air-acetylene
flames.A special five-centimeter burner head with a narrower slot is required
when a nitrous oxide-acetylene flame is to be used.Burner heads also are available
for special purposes,such as use with solutions that have exceptionally high dis-solved
solids contents.
In addition to the flame,there are several options for atomic absorption atomizers.
These options are discussed in detail in Chapter 4.Most of these options require
removal of the premix burner system and replacement by an alternate atomizer in
the spectrometer sample compartment.Since these alternate atomizers offer com-plementary
and extended analytical capabilities,it is likely that the analyst will
want to alternate between the use of flame AA and one or more of the other sys-tems.
A ''quick change''atomizer mount is an important item to facilitate conven-ient
changeover from one device to another without the use of tools.In addition
to convenience,a ''quick change''mount may reduce or eliminate entirely the
need for realignment of the atomizer when it is replaced in the sample compart-ment.
ELECTRONICS
Precision in Atomic Absorption Measurements
We have already discussed the effects of light energy on the precision of an atomic
absorption measurement.The analyst will have little control over these optical fac-tors,
as they are an inherent part of the instrument design.However,the analyst
can exercise some degree of control over precision by proper selection of integra-tion
time with flame atomic absorption.
.Observed precision will improve with the period of time over which each sample
is read.Where analyte concentrations are not approaching detection limits,inte-gration
times of one to three seconds will usually provide acceptable precision.
When approaching instrument detection limits where repeatability is poor,preci-sion
can be improved by using even longer integration times,up to 10 seconds.
In most instances;however,there is little advantage to using integration times
longer than 10 seconds.(To a first approximation,improvement in signal:noise ra-tio
is proportional to the square root of the ratio of integration times.)
Since the detection limit is defined based on the observed precision,the detection
limit also can be improved by increasing the integration time.The analyst,there-fore,
has control over the priorities for a particular analysis,maximum speed or
optimum precision and detection limits.
Current instruments offer statistical functions for averaging and calculating stand-ard
deviation and relative standard deviation or coefficient of variation of replicate
measurements.These functions can be used to determine the precision under vari-ous
experimental conditions,allowing the analyst to optimize method parameters
for each individual requirement.
Calibration of the Spectrometer
Most modern atomic absorption instruments include microcomputer-based elec-tronics.
The microcomputer provides atomic absorption instruments with ad-vanced
calculation capabilities,including the ability to calibrate and compute
concentrations from absorbance data conveniently and accurately,even for non-linear
calibration curves.In the linear region,data on as little as one standard and
a blank may be sufficient for defining the relationship between concentration and
absorbance.However,additional standards are usually used to verify calibration
accuracy.Where the relationship becomes nonlinear,however,more standards are
required.The accuracy of a calibration computed for a nonlinear relationship de-pends
on the number of standards and the equations used for calibration.
For the equation format which optimally fits atomic absorption data,it has been
experimentally shown that accurate calibration can be achieved with a minimum
of three standards plus a blank,even in cases of severe curvature.Figure 2-15 il-lustrates
the accuracy of microcomputer-calculated results for cobalt with single
standard ''linear''and three-standard ''nonlinear''calibrations.After the instru-ment
was calibrated using the specified procedure,a series of standards were ana-lyzed.
Figure 2-15 shows plots of the actual concentrations for those standards
versus the measured values for both calibration procedures.The results obtained
.with ''linear''calibration are accurate only where the absorbance:concentration
relationship is linear,up to about 5 mg/mL.The results obtained with three-stand-ard
''nonlinear''calibration are still accurate at 30 mg/mL,significantly extending
the useful working range.For versatility,current instruments allow fitting of more
than three standards to these same basic equations.
AUTOMATION OF ATOMIC ABSORPTION
Automated Instruments and Sample Changers
One of the greatest contributions to the efficiency of the analytical laboratory is
the automated atomic absorption spectrometer.Automatic samplers were the first
step in freeing the analyst from the monotonous task of manually introducing each
and every sample.
.However,the real advancement in analysis automation came in the late 1970's,
when automated multielement atomic absorption was introduced.In addition to
automatic sample introduction,these instruments offer automatic setup of instru-ment
parameters to preprogrammed values.These instrument ''programs''can be
accessed sequentially,making it possible to analyze a tray full of samples for mul-tiple
elements,without any operator intervention.
Automated Sample Preparation
While automated instrumentation has meant considerable time savings to the ana-lyst,
analytical throughput (i.e.,the number of samples which can be analyzed in
a given time)frequently is limited by the time required to prepare the sample.Even
when the sample is available in a suitable solution form,there often are pretreat-ment
steps which must be performed prior to analysis.The introduction of com-mercial
systems based on techniques such as flow injection have directly
addressed the need for automated sample preparation capabilities.Flow injection
techniques can be used to automate relatively simple procedures such as dilution
or reagent addition.They can also be used to automate complex chemical pretreat-ments,
including analyte preconcentration and cold vapor mercury and hydride
generation procedures.
The Stand-alone Computer and Atomic Absorption
Stand-alone and personal computers have extended the automation and data han-dling
capabilities of atomic absorption even further.These computers can inter-face
directly to instrument output ports to receive,manipulate,and store data and
print reports in user selectable formats.Also,data files stored in personal com-puters
can be read into supplemental software supplied with the system or third
party software such as word processor,spreadsheet and database programs for
open-ended customization of data treatment and reporting.
|
|
