Designation: E 1866 – 97 (Reapproved 2002)

Standard Guide for

Establishing Spectrophotometer Performance Tests1
This standard is issued under the fixed designation E 1866; the number immediately following the designation indicates the year of
original adoption or, in the case of revision, the year of last revision. A number in parentheses indicates the year of last reapproval. A
superscript epsilon (e) indicates an editorial change since the last revision or reapproval.

1. Scope E 387 Test Method for Estimating Stray Radiant Power

1.1 This guide describes basic procedures that can be used Ratio of Spectrophotometers by the Opaque Filter Method2
to develop spectrophotometer performance tests. The guide is E 925 Practice for the Periodic Calibration of Narrow Band-
intended to be applicable to spectrophotometers operating in Pass Spectrophotometers2
the ultraviolet, visible, near-infrared and mid-infrared regions. E 932 Practice for Describing and Measuring Performance
1.2 This guide is not intended as a replacement for specific of Dispersive Infrared Spectrometers2
practices such as Practices E 275, E 925, E 932, E 958, E 1421, E 958 Practice for Measuring Practical Spectral Bandwidth
or E 1683 that exist for measuring performance of specific of Ultraviolet-Visible Spectrophotometers2
types of spectrophotometers. Instead, this guide is intended to E 1421 Practice for Describing and Measuring Performance
provide guidelines in how similar practices should be devel- of Fourier Transform Infrared (FT-IR) Spectrometers:
oped when specific practices do not exist for a particular Level Zero and Level One Tests2
spectrophotometer type, or when specific practices are not E 1655 Practice for Infrared, Multivariate, Quantitative
applicable due to sampling or safety concerns. This guide can Analysis2
be used to develop performance tests for on-line process E 1683 Practice for Testing the Performance of Scanning
spectrophotometers. Raman Spectrometers2
1.3 This guide describes univariate level zero and level one 3. Terminology
tests, and multivariate level A and level B tests which can be
implemented to measure spectrophotometer performance. 3.1 Definitions—For terminology relating to molecular
These tests are designed to be used as rapid, routine checks of spectroscopic methods, refer to Terminology E 131.
spectrophotometer performance. They are designed to uncover 3.2 Definitions of Terms Specific to This Standard:
malfunctions or other changes in instrument operation, but do 3.2.1 action limit, n—the limiting value from an instrument
not specifically diagnose or quantitatively assess the malfunc- performance test, beyond which the spectrophotometer is
tion or change. The tests are not intended for the comparison of expected to produce potentially invalid results.
spectrophotometers of different manufacture. 3.2.2 check sample, n—a single pure compound, or a
1.4 This standard does not purport to address all of the known, reproducible mixture of compounds whose spectrum is
safety concerns, if any, associated with its use. It is the constant over time such that it can be used in a performance
responsibility of the user of this standard to establish appro- test.
priate safety and health practices and determine the applica- 3.2.3 level A test, n—a pass/fail spectrophotometer perfor-
bility of regulatory limitations prior to use. mance test in which the spectrum of a check or test sample is
compared against historical spectra of the same sample via a
2. Referenced Documents multivariate analysis.
2.1 ASTM Standards: 3.2.4 level B test, n—a pass/fail spectrophotometer perfor-
E 131 Terminology Relating to Molecular Spectroscopy2 mance test in which the spectrum of a check or test sample is
E 275 Practice for Describing and Measuring Performance analyzed using a multivariate model, and the results of the
of Ultraviolet, Visible, and Near-Infrared Spectrophotom- analysis are compared to historical results for prior analyses of
eters2 the same sample.
3.2.5 level one (1) test, n—a simple series of measurements
designed to provide quantitative data on various aspects of
spectrophotometer performance and information on which to
1
This guide is under the jurisdiction of ASTM Committee E-13 on Molecular base the diagnosis of problems.
Spectroscopy and is the direct responsibility of Subcommittee E13.11 on Chemo-
metrics.
3.2.6 level zero (0) test, n—a routine check of spectropho-
Current edition approved March 10, 1997. Published May 1997. tometer performance, which can be done in a few minutes,
2
Annual Book of ASTM Standards, Vol 03.06. designed to visually detect significant changes in instrument

Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959, United States.

1
 E 1866 – 97 (2002)
performance and provide a database to determine instrument Spectral data used in performance tests should be date and time
performance over time. stamped, and the results of the tests should be stored in a
3.2.7 optical reference filter, n—an optical filter or other historical database.
device which can be inserted into the optical path in the
spectrophotometer or probe producing an absorption spectrum 6. Samples Used for Performance Testing
which is known to be constant over time such that it can be 6.1 The sample used for performance testing is chosen to be
used in place of a check or test sample in a performance test. compatible with the spectrophotometer configuration, and to
3.2.8 test sample, n—a process or product sample, or a provide spectral features which are adequate for the tests being
mixture of process or product samples which has a constant performed.
spectrum for a finite time period and which can be used in a 6.1.1 The sample used for performance testing should
performance test. Test samples and their spectra are generally generally be in the same physical state (gas, liquid, or solid) as
not reproducible in the long term. the samples to be analyzed during normal operation of the
spectrophotometer.
4. Significance and Use 6.1.2 The sample used for performance testing should be
4.1 If ASTM Committee E-13 has not specified an appro- physically and chemically compatible with the samples ana-
priate test procedure for a specific type of spectrophotometer, lyzed during normal operation.
or if the sample specified by a Committee E-13 procedure is 6.1.3 The sample used for performance is chosen such that
incompatible with the intended spectrophotometer operation, its spectrum is similar to the spectra which will be collected
then this guide can be used to develop practical performance during normal operation.
tests. 6.1.4 The sample used for performance testing should have
4.1.1 For spectrophotometers which are equipped with per- several significant absorbances (0.3 < absorbance < 1.0) across
manent or semi-permanent sampling accessories, the test the spectral range used for normal operation of the spectropho-
sample specified in a Committee E-13 practice may not be tometer.
compatible with the spectrophotometer configuration. For ex- 6.1.5 In order to adequately determine the photometric
ample, for FT-MIR instruments equipped with transmittance or linearity of the instrument, the peak absorbance for at least one
IRS flow cells, tests based on polystyrene films are impractical. absorption band of the sample should be similar to and
In such cases, these guidelines suggest means by which the preferably slightly greater than the largest absorbance expected
recommended test procedures can be modified so as to be for samples measured during normal operation.
performed on a compatible test material. 6.2 Check Samples—Check samples are generally used for
4.1.2 For spectrophotometers used in process measure- conducting performance tests. Check samples are single pure
ments, the choice of test materials may be limited due to compounds or mixtures of compounds of definite composition.
process contamination and safety considerations. These guide- 6.2.1 If mixtures are utilized as check samples, they must be
lines suggest means of developing performance tests based on prepared in a repeatable manner and, if stored, stored such that
materials which are compatible with the intended use of the the mixture is stable over long periods of time. In preparing
spectrophotometer. mixtures, components should be accurately pipetted or
4.2 Tests developed using these guidelines are intended to weighed at ambient temperature. It is recommended that
allow the user to compare the performance of a spectropho- mixtures be independently verified for composition prior to
tometer on any given day with prior performance. The tests are use.
intended to uncover malfunctions or other changes in instru- 6.2.2 While mixtures can be used as check samples, their
ment operation, but they are not designed to diagnose or spectra may be adversely affected by temperature sensitive
quantitatively assess the malfunction or change. The tests are interactions that may manifest themselves by wavelength
not intended for the comparison of spectrophotometers of (frequency) and absorbance changes.
different manufacture. 6.3 Test Samples—A test sample is a process or product
sample or a mixture of process or product samples whose
5. Test Conditions spectrum is expected to be constant for the time period it is
5.1 When conducting the performance tests, the spectropho- used in performance testing. The test sample must be stored in
tometer should be operated under the same conditions as will bulk quantities in controlled conditions such that the material is
be in effect during its intended use. Sufficient warm-up time stable over time.
should be allowed before the commencement of any measure- 6.3.1 Since test samples are often complex mixtures which
ments. cannot be synthetically reproduced, they can only be used for
5.1.1 If possible, the optical configuration used for measure- performance testing for limited time periods. If test samples are
ments of test and check samples should be identical to that used used for this purpose, collection of historical data on a new test
for normal operations. If identical optical configurations are sample should be initiated before previous test samples are
not possible, the user should recognize that the performance depleted. It is recommended that new test samples be analyzed
tests may not measure the performance of the entire instru- sequentially with old test samples at least 15 times before they
ment. are used to replace the old test sample. The 15 analyses must
5.1.2 Data collection and computation conditions should be performed over a time period that does not exceed one
generally be identical to those used in normal operation. month in duration.

2
 E 1866 – 97 (2002)
6.4 Optical Filters—An optical reference filter is an optical ric noise tests may be conducted on the spectrum of a check or
filter or other optical device located in the spectrophotometer test sample at regions where the spectrum is relatively flat and
or in a fiber optic sample probe which produces an absorption the sample absorbance is minimal (<0.1).
spectrum which is known to be constant over time. This filter 7.2.1 For single beam spectrophotometers where back-
may be automatically inserted into the optical path to allow ground and sample spectra are measured separately at different
instrument performance tests to be performed. times, a 100 % line spectrum is obtained by ratioing two
6.4.1 Optical filters are used principally with on-line pro- successive background measurements to obtain a transmittance
cess spectrophotometers equipped with fiber optic probes when spectrum. If, during normal operation of the spectrophotom-
removal of the probe is inconvenient, precluding the use of eter, backgrounds are collected with a reference material in the
check or test samples for routine instrument performance optical path, then this same configuration should be used for
testing. performance testing. Photometric noise calculations are pref-
6.4.2 If an optical filter is used routinely to check or correct erably done directly on the transmittance spectrum. Alterna-
the spectral data collection or computation, then the same filter tively, the transmittance spectrum may be converted to an
is preferably not used for instrument performance testing. If the absorption spectrum by taking the negative log10 before the
same filter is used, then the part of the filter spectrum used in photometric noise calculations.
the performance testing should preferably differ from that part
7.2.2 For double beam spectrophotometers, a 100 % line
used to check or correct the instrument. For example, polysty-
spectrum is measured when the two beams are both empty,
rene filters are used to standardize (continuously check and
both contain empty matched cells, or both contain reference
correct) the wavelength scale of some dispersive NIR spectro-
samples in matched cells.
photometers. For such systems, polystyrene filters are prefer-
ably not to be employed for wavelength stability performance 7.2.3 Photometric noise is measured by fitting a line to the
testing. If polystyrene filters are used, then the peaks used for spectrum over a short spectral region centered on the test
wavelength stability testing should be different from those used frequency (wavelength). The region should contain at least 11
for standardizing the wavelength scale. data points, preferably contains 101 data points, and should not
exceed 2 % of the spectral range. The line is subtracted from
7. Univariate Measures of Spectrophotometer the spectral data, and the RMS noise is calculated as the square
Performance root of the mean square residual.
7.1 Energy Level Tests—Energy level tests are intended to 7.2.3.1 If Ti is the transmittance at the frequency vi, then the
detect changes in the radiant power in the spectrophotometer slope, m, and intercept, b, of a line through the n data points
beam. Decreases in energy levels may be associated with centered at test frequency v0 are given by the following:
deterioration of the spectrophotometer source, with contami- n(iTi 2 (Ti(i
nation or misalignment of optical surfaces in the light path, or m5 (1)
n(i2 2 ~ (i!2
with malfunctions of the detector.
7.1.1 For single beam spectrophotometers where back- (i2(Ti 2 (i(iTi
b5 (2)
ground and sample spectra are measured separately at different n(i2 2 ~ (i!2
times, energy level tests are generally conducted on a back-
The photometric noise is calculated as follows:
ground spectrum. For double beam spectrophotometers where
the ratio of background and sample beam intensities is mea-
sured directly, energy levels can be measured if it is possible to NoiseRMS 5 Π(~Ti 2 ~mi 1 b!!2
n22 (3)
block the sample beam. The index i in Eq 1-3 runs from − (n − 1)/2 to (n − 1)/2 (n
7.1.2 Energy levels should be measured at at least three
must be odd). The intercept represents the transmittance at test
fixed frequencies (wavelengths), one each in the upper, middle
frequency v0.
and lower third of the spectral range. The frequencies (wave-
lengths) at which energy levels are measured should be chosen 7.2.3.2 If photometric noise is calculated on absorbance
to avoid interferences due to atmospheric components (for spectra, the absorbance values, Ai, are used in place of the
example, absorptions of water vapor and carbon dioxide) and transmittance values, Ti, in Eq 1-3. If the abscissa for the
from interferences due to optical components (for example, spectral data is wavelength, then wavelength values, li, are
OH absorptions in SiO2 cells and fibers). Preferably, regions used in place of the frequency values, vi, in Eq 1-3. Calcula-
where the background spectrum is relatively flat and slowly tions should be consistently performed on the same data types.
varying should be used for this test. 7.2.4 Increases in the photometric noise can indicate a
7.1.3 To minimize the effects of photometric noise on the misalignment of optical components, a source malfunction, or
energy level measurement, it is preferable to average the a malfunction in the detector or electronics.
energy over a narrow frequency (wavelength) window. Typi- 7.3 Short Term Baseline Stability Test—The transmittance is
cally, the intensity at five points centered on the test frequency monitored at each of the test frequencies (wavelengths) used in
are averaged. the energy level and photometric noise tests. The intercept
7.2 Photometric Noise Tests—Photometric noise is mea- calculated in Eq 2 represents the transmittance averaged over
sured at the same frequencies (wavelengths) used for the the n points around test frequency v0. Deviation from 100 %
energy level tests. Preferably, photometric noise tests are transmittance is an indication of short term baseline instability
conducted on a 100 % line spectrum. Alternatively, photomet- and may indicate a malfunction of the spectrophotometer.

3
 E 1866 – 97 (2002)
7.3.1 If the tests are conducted on absorbance spectra, (1)3 with corrections by Steiner, Termonia, and Deltour (2),
deviations from zero absorbance is used as an indication of with application criteria discussed by Willson and Polo (3). The
baseline instability. latter reference discusses optimum filter parameters based upon
7.3.2 If photometric noise tests are conducted on the spec- the relationship between spectral bandwidth and digitization
trum of a check or test sample, then variations in the absor- interval. A cubic filter is recommended. The number of points
bance spectrum at the test frequencies are taken as an indica- used in the filter should be the quotient of the full-width-at-
tion of short term baseline instability. half-maximum of the peak being measured divided by the
digital resolution, and rounded up to the nearest odd integer.
7.4 Optical Contamination Tests—The single beam back-
7.6.3.2 Identify the zero crossing associated with the peak
ground scan which was used for the energy tests is examined
absorbance and compute its location by linear interpolation
for absorptions which might indicate contamination of optical between the two adjacent points straddling the zero crossing.
surfaces in the beam path. The zero crossing is taken as a measure of the peak position.
7.4.1 Failure to clean cell or probe windows, IRS surfaces,
NOTE 1—Other peak finding algorithms can be used provided that they
etc., are the most common source of optical contamination.
accurately track peak position. The procedure described in Annex A1
Frequencies (wavelengths) at which typical samples exhibit should be used to test peak finding algorithms to determine if they are
maximum absorbance should generally be examined. For appropriate for this application. It is the users responsibility to demon-
example, for IR systems used in hydrocarbon analysis, the strate that the peak finding algorithm is appropriate for monitoring
regions where the C-H stretching vibrations occur should be spectrophotometer frequency (wavelength) stability.
examined. Significant increases above a nominal background 7.7 Resolution Stability Tests—The resolution stability of
level may indicate contamination of windows and surfaces. the spectrophotometer is monitored by measuring the band-
7.4.2 Spectrophotometer optical surfaces can be contami- widths of several absorption peaks in the absorption spectrum
nated by impurities in purge gases. For systems equipped with of the check/test sample or optical filter. At least three peaks
flow cells or circulating liquid temperature control, leaks in are used for the test. If possible, the peaks should be in the
connecting lines can expose an optical surface to contamina- upper, middle and lower third of the spectral range. Variations
tion. Users should consider possible sources of contamination in the measured bandwidths are taken as an indication that the
and determine appropriate frequencies at which absorptions optical resolution of the spectrophotometer is varying, suggest-
would result. ing a malfunction.
7.5 Purge Contamination Tests—For spectrophotometers 7.7.1 The absorption for peaks used in this test are prefer-
which are purged to minimize absorptions due to atmospheric ably in the range from 0.37 to 0.75. For peak absorptions
components, the single beam spectrum used for energy tests outside this range, the resolution stability measurement may
should be checked for variations in purge quality. Frequencies show increased sensitivity to photometric noise.
(wavelengths) at which potential contaminants absorb should 7.7.2 Peaks used for the resolution stability test are prefer-
be identified, as should baseline points where contaminant ably symmetric in shape and well resolved from neighboring
absorption would be minimal. The absorbance for contami- peaks. If such peaks are not available in the spectrum of the
nants is calculated as the negative log10 of the ratio of the peak check/test sample or optical filter, the results of the resolution
intensity to the baseline intensity. stability test may be variable.
7.7.3 It is recommended that the peak bandwidth be deter-
7.6 Frequency (Wavelength) Stability Tests—Frequency
mined by the following steps:
(wavelength) stability tests are conducted by monitoring the
7.7.3.1 Compute the second derivative of the spectrum by
peak positions of several peaks across the absorption spectrum
applying the appropriate digital filter to the spectrum. A
of the check or test sample or optical filter. At least three peaks
commonly used filter has been defined by Savitzki and Golay
are used for the test. If possible, the peaks should be in the
(1) with corrections by Steiner, Termonia, and Deltour (2), with
upper, middle and lower third of the spectral range. application criteria discussed by Willson and Polo (3). The
7.6.1 The absorption for peaks used in this test are prefer- latter reference discusses optimum filter parameters based upon
ably in the range from 0.37 to 0.75. For peak absorptions the relationship between spectral bandwidth and digitization
outside this range, the wavelength stability measurement may interval. A cubic filter is recommended. The number of points
show greater sensitivity to photometric noise. used in the filter should be the quotient of the full-width-at-
7.6.2 Peaks used for the frequency stability test are prefer- half-maximum of the peak being measured divided by the
ably symmetric in shape and well resolved from neighboring digital resolution, and rounded up to the nearest odd integer.
peaks. If such peaks are not available in the spectrum of the 7.7.3.2 Identify the zero crossing on each side of the peak
check/test sample or optical filter, the user should be aware that absorbance and compute their locations by linear interpolation
changes in spectrophotometer resolution will affect the mea- between the two adjacent points straddling the zero crossings.
sured peak position. The difference in the frequencies of the interpolated zero
7.6.3 It is recommended that the peak position be deter- crossings is taken as a measure of the peak bandwidth.
mined by the following steps:
7.6.3.1 Compute the first derivative of the spectrum by
applying the appropriate digital filter to the spectrum. A 3
The boldface numbers in parentheses refer to the list of references at the end of
commonly used filter has been defined by Savitzki and Golay this guide.

4
 E 1866 – 97 (2002)
7.8 Photometric Linearity Tests—Linearity of the spectrom- g 5 ahh 1 all 1 a11 (4)
eter response is important for quantitative applications. Unfor-
where:
tunately, the absolute photometric linearity cannot be checked
g = column vector con-
in a quick performance test. To do so would generally require
taining the current
the use of multiple standards of known absorbance. The test
spectrum of the
described here is intended only to measure changes in the
check sample, the test
photometric linearity of a spectrophotometer.
sample or the optical
7.8.1 Photometric linearity is tested using the ratio of the filter,
absorbances of two or more peaks in the absorbance spectrum. h = column vector con-
One peak should have an absorbance at or near the maximum taining the historical
absorbance that will be used for normal operations. The other spectrum of the
peaks are preferably less intense than this maximum. If only check sample, the test
two peaks are used, the second peak should be approximately sample or the optical
half the intensity of the first peak. filter,
7.8.2 Linear baselines for each peak are calculated from l (v for frequency based spectra) = column vector of the
points of minimal absorbance on opposite sides of the peaks. wavelength axis val-
The maximum absorbance for each peak is corrected for the ues for spectra g and
baseline, and the ratio of the absorbances for the two peaks is h,
calculated. The ratio is used to track changes in the photometric 1 = column vector of
linearity. ones,
ah = coefficient for scaling
8. Multivariate Measures of Spectrophotometer the historical spec-
Performance trum to match the
8.1 Level A Tests—A Level A performance test is a pass/fail current spectrum,
test that is sensitive to many of the univariate performance al = coefficient which
parameters discussed in Section 7. Level A tests do not identify scales l to provide a
specific failure modes, but merely indicate if the instrument baseline correction
performance is within historical bounds. In this test, the which is linear in
spectrum of a check sample, a test sample or an optical filter is wavelength (or fre-
compared to a historical spectra of the check sample, the test quency), and
sample or the optical filter by multivariate methods (least a1 = coefficient for a base-
squares fitting or a PCR/PLS model; see Practice E 1655 for line offset.
descriptions of PCR and PLS). This procedure can provide The column vectors h, l and 1 are combined into a matrix H.
some information about specific instrument parameters, but
essentially looks for deviations in the residual spectrum as ah
g = [h l 1] al = Ha
compared to the historical residual spectra. (5)
al
8.1.1 Level A tests are generally applied on spectrophotom-
eters which are in use for multivariate, quantitative analysis.
The spectral range used in Level A tests should be comparable 8.1.2.3 The estimated coefficients â are first determined by
to that used in the calibration model for the analysis being linear least squares.
performed. If the spectrum of the check sample, the test sample â 5 H1 g 5 ~HtH!21 H2t g (6)
or the optical filter used in the Level A test contains absorptions
that are significantly higher than those of the typical samples 8.1.2.4 The coefficients are then used to estimate the spec-
being analyzed, then these peaks can be excluded from the trum of the current sample ĝ.
Level A fit. ĝ 5 Hâ (7)
8.1.2 Level A Tests Using a Least Squares Method: 8.1.2.5 The residuals from the fit are the difference between
8.1.2.1 In this Level A test, a least squares fit of the current the measured and estimated values for the data points, g − ĝ.
spectrum of the check sample, test sample or optical filter is The residuals from the fit are squared, and summed. The
conducted against a historical spectrum of the same material. resulting measure, herein referred to as the spectral residual, is
Baseline terms may be included in the fit to compensate for used as a measure of changes in the instrument performance.
variations in baseline, and scaling may be applied to compen- This spectral residual should be plotted on control charts.
sate for pathlength variations. The types of compensations
(baseline or pathlength) used in the fit should be similar to NOTE 2—Any function of the sum of the squares of the residuals can be
those employed in the multivariate model used for the actual used, for example, the square root, or the square root of the sum divided
by the number of points.
analyzer measurement. Methodology for calculating the least
square fit is discussed by Blackburn (4) and by Antoon, 8.1.2.6 Additionally, the scaling and baseline coefficients
Koenig, and Koenig (5). can be monitored as an additional measure of instrument
8.1.2.2 A typical least squares model could be as follows: performance.

5
 E 1866 – 97 (2002)
8.1.3 Level A Tests Using a PCR or PLS Method—To the analyzer is serviced. For example, the energy output of an
perform a Level A test using PCR or PLS, one must first infrared source is expected to decrease continuously as the
develop an appropriate model. A series of historical spectra for source ages, until such time as the source is cleaned or
the check/test sample, or the optical filter are analyzed by a replaced. The decreased energy may be observed as an increase
PCR or PLS regression algorithm using 100 % for the compo- in the Level 0 photometric noise, or as an increase in the Level
sitional value to generate the Level A model. Generally, only A spectral residual. The daily change in energy, noise or
one variable should be retained in the model since all the residual may be large relative to the precision with which these
spectra are of the same material. The type of pre- or postpro- values can be measured, but have tolerable effect on the
cessing done in the Level A test model should be comparable spectrophotometer results for the intended application. For
to that done in the multivariate calibration models being used such tests, control charts and statistically derived limits may be
on the spectrophotometer. inappropriate. An action limit for such tests needs to be
NOTE 3—Chemometricians might refer to the analysis described in
determined from historical data or simulations as discussed in
8.1.4 as principal component analysis rather than principal components Section 10.
regression. However, the object here is to allow the Level A test to be 9.2.2 For some performance tests, the test results are ex-
developed and applied using the same chemometric software employed in pected to vary randomly about a fixed point. For example, for
the development and application of the multivariate calibration model. a properly operating instrument, the Level 0 wavelength value
8.1.3.1 The PCR or PLS model is used to analyze the might be expected to vary randomly about some average value.
current spectrum of the check/test sample or optical filter. The For such tests, the statistical control charts and control limits
estimate for the current spectrum is generated. The spectral can be usefully employed to set initial action limits in the
residuals are calculated as the difference between the current absence of historical data. Such initial action limits may be
spectrum and its estimate. The spectral residual can be charted loosened if statistically significant performance changes de-
to determine if the instrument is operating within historical tected by the control charts are not found to have significant
specifications. effect on the spectrophotometer results for the intended appli-
8.1.3.2 PCR and PLS models can also provide information cation.
on the scaling of the current spectrum. For simple models, 9.2.3 Since, for spectrophotometers used in multivariate
variation in the estimated composition from 100 % is an analyses, Level B composition or property results for check or
indication of scaling variation. test samples are most directly comparable to actual analyzer
8.2 Level B Tests—Level B performance tests analyze the results, the Level B composition or property estimates are most
spectrum of a check sample, a test sample or an optical filter amenable to statistical control charting. Action limits for Level
against multivariate models that are employed during the B composition or property estimates can be set to statistically
normal use of the spectrophotometer system. As such, Level B determined control limits.
tests can not be performed during the multivariate calibration
10. Performance Test Action Limits
of a spectrophotometer. Level B tests monitor the instrument
performance for deviations to which a calibration model is 10.1 Spectroscopic analyses differ greatly in their sensitivity
sensitive. Tests on a limited number of samples are not to various aspects of spectrophotometer performance, and each
rigorous, but failures in these tests are indicative that the application differs in what constitutes an acceptable tolerance
spectrophotometer operation has changed. to changes in the results caused by variations in spectropho-
8.2.1 The spectrum of the check sample, the test sample or tometer performance. Although spectrophotometer perfor-
the optical filter is analyzed using the multivariate model used mance tests are useful in their own right, the user must be
during normal spectrometer operation. The predicted value concerned with how changes in the spectrophotometer perfor-
(property or component concentration), the Mahalanobis dis- mance affect results during normal operation. Historical data-
tance, and the spectral residuals are again compared to histori- bases or simulations that define acceptable performance for one
cal values to detect any change in the analyzer performance. application may not be appropriate for another application. In
addition, the level of performance required by quantitative
9. Performance Test Charts applications may be changed by the updating of the calibration.
9.1 Performance test results should be plotted on charts and 10.2 Setting Action Limits Based on Historical Data for
examined for trends. Such trend analysis may provide early Performance Tests:
warnings of possible analyzer problems. 10.2.1 Performance tests provide quantitative measures of
9.2 Statistical quality control charting methods (for ex- spectrophotometer performance. These measures can be com-
ample, individual value control charts, exponentially weighted pared to historical data for the same tests in order to recognize
moving average control charts, and moving range of two changes of spectrophotometer performance. If historical data
control charts) can be used to detect statistically significant exist, limits for each test can be set and the performance can be
changes in instrument performance. However, the statistical judged against these limits. If historical data does not exist, it
control limits associated with these charts will not necessarily will be necessary to collect them as a standard part of the
be useful in judging the performance test results. Instead, some spectrophotometer operation, and such collection will eventu-
performance test results are typically compared to action limits ally allow performance limits to be established. The collection
as described in Section 10. of the historical database for performance tests should be an
9.2.1 For some performance tests, the test results are ex- integral part of the spectrophotometer operation and be con-
pected to trend continuously in one direction until such time as tinued for the life of the spectrophotometer.

6
 E 1866 – 97 (2002)
10.2.2 For spectrophotometers used in qualitative analysis, method on control charts, conditions that lead to invalid
what constitutes acceptable spectrophotometer performance quantitative results can be identified, and action limits for each
may be difficult to quantify. The expert analyst may only know Level 0 test can eventually be established.
that the spectra “look wrong”. If a change in instrument 10.2.4.7 For spectrophotometers where normal operation
performance is detected, spectra of known (check or test involves use of a multivariate calibration model, increases in
sample) materials can be collected, and the analyst can the spectral residuals that are detected by Level A tests will
examine the spectra to determine if the observed change in generally reflect some change in the quantitative results pro-
performance would distort the spectra sufficiently to cause duced by the spectrophotometer. Even if the analysis result
misidentifications to occur. does not change, the spectral residuals measured as part of the
10.2.3 If the normal operation of the spectrophotometer outlier testing will generally be expected to increase. The level
involves the use of spectral library searching or discriminative of increase that can be tolerated can be determined by plotting
analysis for materials identification or classification, then the Level A test spectral residuals against analyzer results and
known (check or test sample) materials should be reanalyzed determining the maximum level at which valid analysis results
whenever a change in performance is detected to determine if are produced.
the change reduces the discrimination power of the analysis. 10.2.4.8 For spectrophotometers where normal operation
10.2.4 If the normal operation of the spectrophotometers involves use of a multivariate calibration model, changes in the
involves quantitative analysis, then quality assurance for the values produced by a Level B test are the most straightforward
quantitative analysis should be part of the normal operation. to interpret since the values are directly comparable to the
The significance of changes in spectrophotometer performance analysis results. If the analysis of the spectrum of the check
are judged by their effects on the quality control results for the sample, the test sample, or the optical filter is an interpolation
normal analysis. of the model, then limits can be set directly based on the
10.2.4.1 If the quantitative method used during normal desired performance of the analyzer. If the analysis of the
operation is a primary method (for example, univariate cali- spectrum of the check sample, the test sample or the optical
bration against quantitatively prepared or certified standards), filter is an extrapolation of the model, then some care must be
then the quality control for the quantitative method will exercised in setting limits since the extrapolated result may be
generally involve measurement of a standard and comparison more sensitive to small changes in instrument performance
of the results against the known value. than analyses that are interpolations of the model. This is
10.2.4.2 If the quantitative method used during normal known as leverage. In this case, initial limits should be
operation is a secondary method (for example, univariate or confirmed by plotting the Level B results against analysis
multivariate calibration against results obtained by another results and determining the levels at which valid analysis
primary analytical method), then the quality control for the results are produced.
quantitative method will generally involve periodic analyses of
an unknown by both methods and statistical comparison of the NOTE 4—Any one test sample, check sample, or optical filter only tests
results. a small portion of the multivariate model space, and may not be sensitive
to all aspects of spectrophotometer performance. The Level 0, A, and B
10.2.4.3 If the quantitative method used during normal tests are intended to detect possible spectrophotometer failure modes.
operation is within statistical quality control, then the results Acceptable performance as measured by Level 0, A and B tests is
for the performance tests conducted during the same time necessary but not sufficient by themselves for demonstrating valid
period should be considered an example of acceptable instru- spectrophotometer performance. Comparison of analysis results to in
ment performance and added to the historical database. control, primary method laboratory values is also necessary to demon-
strate the validity of analysis results.
10.2.4.4 If the quantitative method used during normal
operation is not within statistical quality control, the results 10.3 Determining Performance Action Limits by Simulating
from the performance tests may be examples of unacceptable Instrument Response Changes:
instrument performance, particularly if the results from the 10.3.1 For spectrophotometers where normal operation in-
performance tests are inconsistent with the historical database. volves use of a multivariate calibration model, an alternative
Examples of unacceptable instrument performance can be used procedure for determining action limits for spectrophotometer
to set action limits for future performance tests. performance tests is to take actual, diverse, but representative
10.2.4.5 It is strongly recommended that, at the time the spectra that are predicted well by the model, and to mathemati-
quantitative method is developed, spectra of the check sample, cally modify these spectra to simulate the expected variations
the test sample or the optical filter be collected along with in the spectrophotometer performance. The model sensitivity,
spectra used for calibration. Performance tests can be applied for example, the change in the results per unit change in a
to this data to determine the level of performance at the time of performance parameter, can be estimated, and used to establish
calibration. action limits for each performance parameter based on the error
10.2.4.6 Changes in analyzer performance that are detected tolerance for the application. Spectrophotometer performance
by Level 0 tests may or may not produce a significant change parameters which can be modeled include wavelength (fre-
in the results during normal operation. Different types of quency) shifts, baseline shifts, changes in photometric noise,
quantitative analyses differ significantly in their sensitivity to resolution changes, and detector linearity changes. The impor-
various aspects of analyzer performance. By plotting the Level tance of different performance parameters is both application
0 test results with the quality control results for the quantitative and instrument type dependent. Historical data for Level 0

7
 E 1866 – 97 (2002)
performance tests are the best guide to the type response spectra, care must be exercised that the interpolation function
changes that should be modeled for a given instrument type. does not smooth or deresolve the spectra.
10.3.1.1 For example, the sensitivity of an analyzer to 10.3.2 Changes in spectrophotometer performance seldom
baseline drift can be simulated by adding various baselines to effect only one aspect of that performance. If simulations are
a set of representative spectra, analyzing these spectra with the used to set action limits for performance tests, it is essential
calibration model, and determining the change in the results as that multiple performance parameters by varied simulta-
a function of the added baseline. The added baseline can, for neously. The magnitude of the changes to the performance
example, be parameterized in terms of offset, slope and parameters that should be simulated are best obtained from
curvature so that the effects of each can be determined. examination of historical data on Level 0 performance tests
10.3.1.2 For example, the sensitivity of an analyzer to
conducted on the type of spectrophotometer being tested.
wavelength (frequency) shift can be simulated by shifting the
wavelength (frequency) of a set of representative spectra (see
11. Keywords
Annex A1), analyzing these spectra with the calibration model,
and determining the change in the results as a function of the 11.1 instrument performance; level 0; molecular spectro-
shift. If the shift is accomplished via interpolation of the photometer; spectrophotometer tests

ANNEX

(Mandatory Information)

A1. PROCEDURE FOR TESTING PEAK FINDING ALGORITHMS

A1.1 Interpolation of Test Spectrum points at each end) is treated as point 0 once. The values Ai(i
A1.1.1 To test a peak finding algorithm, it is necessary to = 0.2, 0.4, 0.6 and 0.8) calculated for the repetitive application
interpolate and shift a test spectrum by amounts equivalent to of the filter are collected into 4 spectral vectors. These vectors
a fraction of the digital resolution of the spectrum. The are assigned the same frequency scales as the original spec-
following procedure can be used to interpolate and shift the trum. If D is the digital resolution (data point spacing), then the
spectrum. vectors A0.8, A0.6, A0.4, and A0.2 generated by repetitive
A1.1.1.1 Four adjacent points in the spectrum are selected. application of the above procedure represent spectra which are
The points are numbered −1, 0, 1, and 2. The absorbance at the shifted by − 0.2D, − 0.4D, − 0.6D, and − 0.8D, respectively.
four points is fit to a cubic equation.
NOTE A1.1—If the data points are numbered −1, 0, 1, and 2 from left to
right, the interpolation estimates a value for the spectrum between point 0
A2 1 2 4 8 c0
and point 1. The original point at 0 is replaced with this estimate. The
A1 1 1 1 1 c1 ↔ A 5 XC
= effect is to shift the spectrum toward the left (negative index) direction.
A0 1 0 0 0 c2 (A1.1)
A–1 1 –1 1 –1 c3
A1.2 The peak finding algorithm is applied to the original
and shifted spectra. The peak positions for suitable peaks are
c0 0 0 6 0 A2 calculated.
c1 –1 6 –3 –2 A1
C = X–1 A ↔ (A1.2) A1.2.1 If the peak positions calculated by the peak finding
c2 0 3 –6 3 A0
= 1⁄6
c3 1 –3 3 –1 A–1 algorithm for the shifted spectra are shifted
by −0.2D, −0.4D, −0.6D, and −0.8D, respectively, relative to
A1.1.2 The cubic equation is evaluated at four intermedi- the peaks for the original spectrum, then the peak finding
ate points algorithm can be used for tracking frequency (wavelength)
stability.
A0.8 1 0.8 0.64 0.512 c0 A1.2.2 If the peak positions calculated by the peak finding
A0.6
=
1 0.6 0.36 0.216 c1
(A1.3)
algorithm for the shifted spectra are not shifted
A0.4 1 0.4 0.16 0.064 c2
A0.2 1 0.2 0.04 0.008 c3
by −0.2D, −0.4D, −0.6D, and −0.8D, respectively, relative to
the peaks for the original spectrum, then the peak finding
algorithm cannot be used for tracking frequency (wavelength)
A1.1.3 The procedure is repeated, moving the filter side- stability.
ways so that each point in the spectrum (other than the two

8
 E 1866 – 97 (2002)
REFERENCES

(1) Savitzky, A. and Golay, M. J. E., “Smoothing and Differentiation of (3) Willson, P. D. and Polo, S. R., “Polynomial Filters of Any Degree,”
Data by Simplified Least Squares Procedures,” Analytical Chemistry, Journal of the Optical Society of America, Vol 71, 1981, pp. 599–603.
Vol 36, pp. 1627–1639, 1964. (4) Blackburn, J. A., Analytical Chemistry, Vol 37, 1965, p. 1000.
(2) Steiner, J., Termonia, Y., and Deltour, J., “Comments on Smoothing
(5) Antoon, M. K., Koenig, J. H., and Koenig, J. L., Applied Spectros-
and Differentiation of Data by Simplified Least Squares Procedures,”
Analytical Chemistry, Vol 44, 1972, p. 1906. copy, Vol 31, 1977, p. 578.

ASTM International takes no position respecting the validity of any patent rights asserted in connection with any item mentioned
in this standard. Users of this standard are expressly advised that determination of the validity of any such patent rights, and the risk
of infringement of such rights, are entirely their own responsibility.

This standard is subject to revision at any time by the responsible technical committee and must be reviewed every five years and
if not revised, either reapproved or withdrawn. Your comments are invited either for revision of this standard or for additional standards
and should be addressed to ASTM International Headquarters. Your comments will receive careful consideration at a meeting of the
responsible technical committee, which you may attend. If you feel that your comments have not received a fair hearing you should
make your views known to the ASTM Committee on Standards, at the address shown below.

This standard is copyrighted by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959,
United States. Individual reprints (single or multiple copies) of this standard may be obtained by contacting ASTM at the above
address or at 610-832-9585 (phone), 610-832-9555 (fax), or service@astm.org (e-mail); or through the ASTM website
(www.astm.org).