ASTM D6342-22

This document is not an ASTM standard and is intended only to provide the user of an ASTM standard an indication of what changes have been made to the previous version. Because
it may not be technically possible to adequately depict all changes accurately, ASTM recommends that users consult prior editions as appropriate. In all cases only the current version
of the standard as published by ASTM is to be considered the official document.
´1
Designation: D6342 − 12 (Reapproved 2017) D6342 − 22
Standard Practice for
Polyurethane Raw Materials: Determining Hydroxyl Number
of Polyols by Near Infrared (NIR) Spectroscopy
This standard is issued under the fixed designation D6342; 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 (´) indicates an editorial change since the last revision or reapproval.
ε NOTE—Reapproved with editorial changes in August 2017.
1. Scope Scope*
1.1 This standard covers a practice for the determination of hydroxyl numbers of polyols using NIR spectroscopy.
1.2 Definitions, terms, and calibration techniques are described. Procedures for selecting samples, and collecting and treating data
for developing NIR calibrations are outlined. Criteria for building, evaluating, and validating the NIR calibration model are also
described. Finally, the procedure for sample handling, data gathering and evaluation are described.
1.3 The implementation of this standard requires that the NIR spectrometer has been installed in compliance with the
manufacturer’s specifications.
1.4 The values stated in SI units are to be regarded as standard. No other units of measurement are included in this standard.
1.5 This standard does not purport to address all of the safety concerns, if any, associated with its use. It is the responsibility
of the user of this standard to establish appropriate safety safety, health, and healthenvironmental practices and determine the
applicability of regulatory limitations prior to use.
NOTE 1—This standard is equivalent ISO 15063.
1.6 This international standard was developed in accordance with internationally recognized principles on standardization
established in the Decision on Principles for the Development of International Standards, Guides and Recommendations issued
by the World Trade Organization Technical Barriers to Trade (TBT) Committee.
2. Referenced Documents
2.1 ASTM Standards:
D883 Terminology Relating to Plastics
D4274 Test Methods for Testing Polyurethane Raw Materials: Determination of Hydroxyl Numbers of Polyols
D4855 Practice for Comparing Test Methods (Withdrawn 2008)
E131 Terminology Relating to Molecular Spectroscopy
E168 Practices for General Techniques of Infrared Quantitative Analysis
This practice is under the jurisdiction of ASTM Committee D20 on Plastics and is the direct responsibility of Subcommittee D20.22 on Cellular Materials - Plastics and
Elastomers.
Current edition approved Aug. 1, 2017July 1, 2022. Published August 2017July 2022. Originally approved in 1998. Last previous edition approved in 20122017 as
ɛ1
D6342 - 12.D6342 - 12(2017) . DOI: 10.1520/D6342-12R17E01.10.1520/D6342-22.
For referenced ASTM standards, visit the ASTM website, www.astm.org, or contact ASTM Customer Service at service@astm.org. For Annual Book of ASTM Standards
volume information, refer to the standard’s Document Summary page on the ASTM website.
*A Summary of Changes section appears at the end of this standard
Copyright © ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA 19428-2959. United States
D6342 − 22
E222 Test Methods for Hydroxyl Groups Using Acetic Anhydride Acetylation
E275 Practice for Describing and Measuring Performance of Ultraviolet and Visible Spectrophotometers
E456 Terminology Relating to Quality and Statistics
E1655 Practices for Infrared Multivariate Quantitative Analysis
E1899 Test Method for Hydroxyl Groups Using Reaction with p-Toluenesulfonyl Isocyanate (TSI) and Potentiometric Titration
with Tetrabutylammonium Hydroxide
2.2 ISO Standard:
ISO 15063 Plastics—Polyols for use in the production of polyurethanes—Determination of hydroxyl number by NIR
spectroscopy
3. Terminology
3.1 Definitions—Terminology used in this practice follows that defined in Terminology D883. For terminology related to
molecular spectroscopy methods, refer to Terminology E131. For terms relating to multivariate analysis, refer to Practice E1655.
3.2 Definitions of Terms Specific to This Standard:
3.2.1 hydroxyl number—the milligrams of potassium hydroxide equivalent to the hydroxyl content of 1 g of sample.
4. Summary of Practice
4.1 Multivariate mathematics is applied to correlate the NIR absorbance values for a set of calibration samples to the respective
reference hydroxyl number for each sample. The resultant multivariate calibration model is then applied to the analysis of unknown
samples to provide an estimate of their hydroxyl numbers.
4.2 Multilinear regression (MLR), principal components regression (PCR), and partial least squares regression (PLS) are the
mathematical techniques used for the development of the calibration model.
4.3 Statistical tests are used to detect outliers during the development of the calibration model. Outliers can include high leverage
samples and samples whose hydroxyl numbers are inconsistent with the model.
4.4 Validation of the calibration model is performed by using the model to analyze a set of validation samples. The hydroxyl
number estimates for the validation set are statistically compared to the reference hydroxyl number for this set to test for agreement
of the model with the reference method.
4.5 Statistical expressions are given for calculating the precision and bias of the NIR method relative to the reference method.
5. Significance and Use
5.1 General Utility:
5.1.1 It is necessary to know the hydroxyl number of polyols in order to formulate polyurethane systems.
5.1.2 This practice is suitable for research, quality control, specification testing, and process control.
5.2 Limitations:
5.2.1 Factors affecting the NIR spectra of the analyte polyols need to be determined before a calibration procedure is started.
Chemical structure, interferences, any nonlinearities, the effect of temperature, and the interaction of the analyte with other sample
components such as catalyst, water and other polyols needs to be understood in order to properly select samples that will model
those effects which cannot be adequately controlled.
5.2.2 Calibrations are generally considered valid only for the specific NIR instrument used to generate the calibration. Using
different instruments (even when made by the same manufacturer) for calibration and analysis can seriously affect the accuracy
and precision of the measured hydroxyl number. Procedures used for transferring calibrations between instruments are problematic
and are to be utilized with caution following the guidelines in Section 16. These procedures generally require a completely new
validation and statistical analysis of errors on the new instrument.
D6342 − 22
5.2.3 The analytical results are statistically valid only for the range of hydroxyl numbers used in the calibration. Extrapolation to
lower or higher hydroxyl values can increase the errors and degrade precision. Likewise, the analytical results are only valid for
the same chemical composition as used for the calibration set. A significant change in composition or contaminants can also affect
the results. Outlier detection, as discussed in Practices E1655, is a tool that can be used to detect the possibility of problems such
as those mentioned above.
6. Instrumentation
6.1 Introduction—A complete description of all applicable types of NIR instrumentation is beyond the scope of this standard. Only
a general outline is given here. A diagram of a typical NIR spectrometer is shown in Fig. 1.
6.2 Light Source and Detector—Tungsten-halogen lamps with quartz envelopes usually serve as the energy sources for NIR
instruments. Most of the detectors used for NIR are solid-state semiconductors. PbS, PbSe, and InGaAs detectors are most
commonly used.
6.3 Light Dispersion—Spectrophotometers can be classified based on the procedure by which the instrument accomplishes
wavelength selection.
6.3.1 Monochromator Instrument—Grating monochromator instruments, often called “dispersive” instruments, are commonly
used in the laboratory and for process applications. In a holographic grating system, the grating is rotated so that only a narrow
band of wavelengths is transmitted to a single detector at a given time.
6.3.2 Filter-Wheel Instrument—In this type of NIR instrument, one or several narrow band filters are mounted on a turret wheel
so that the individual wavelengths are presented to a single detector sequentially.
6.3.3 Acoustic Optic Tunable Filter (AOTF) Instrument—The AOTF is a continuous variant of the fixed-filter photometer with no
moving optical parts for wavelength selection. A birefringent TeO crystal is used in a noncollinear configuration in which acoustic
and optical waves move through the crystal at different angles. Variations in the acoustic frequency cause the crystal lattice spacing
to change. That in turn causes the crystal to act as a variable transmission diffraction grating for one wavelength. The main
advantage of using AOTF instruments is the speed. A wavelength or an assembly of wavelengths can be changed hundreds of times
per second under computer control.
6.3.4 Light-Emitting Diode (LED) Instrument—Each wavelength band is produced by a different diode. The major advantages of
the system are its small size and compactness, stability of construction with no moving parts, and low power consumption.
6.3.5 Fourier Transfer (FT) Instrument—In FT-NIR instruments, the light is divided into two beams whose relative paths are
varied by use of a moving optical element. The beams are recombined to produce an interference pattern that contains all of the
wavelengths of interest. The interference pattern is mathematically converted into spectral data using Fourier transform. FT
interferometer optics provide complete spectra with very high wavelength resolution. FT signal averaging also provides higher
signal-to-noise ratios in general than can be achieved with other types of instruments.
6.4 Sampling System—Depending upon the applications, several different sampling systems can be used in the laboratory or for
on-line instruments, or both.
6.4.1 Cuvette—Quartz or glass cuvettes with fixed or adjustable pathlengths can be used in the laboratory.
FIG. 1 Schematic of a Near-IR System
D6342 − 22
6.4.2 Flow-Through Cell—This type of cell can be used for continuous or intermittent monitoring of liquid sample.
6.4.3 Probes:
6.4.3.1 Transmission Probe—Transmission probes combined with optic fibers are ideal for analyzing clear liquids, slurries,
suspensions, and other high viscosity samples. Low absorptivity in the NIR region permits sampling pathlengths of up to 10 cm.
6.4.3.2 Immersion Probe—The immersion system uses a bi-directional optic fiber bundle and variable pathlength probe for sample
measurements. Radiation from the source is transmitted to the sample by the inner ring of fibers, and diffuse transmitted radiation
is collected by the outer ring of fibers for detection.
6.4.3.3 Attenuated Total Reflection (ATR) Probe—Attenuated total reflection occurs when an absorbing medium (the sample) is
in close contact with the surface of a crystal material of higher refractive index. At an optimized angle, the NIR beam reflects
internally along the crystal faces, penetrating a few microns into the sample surface, where selective absorption occurs. The
resulting spectrum is very close to the conventional transmission spectrum for the sample. There are many designs of ATR plates
and rods for specific applications. Single or multiple reflection units are available. ATR sampling accessories are available for the
laboratory and, in the form of fiber optic probes, can be used for on-line analysis. This is an advantage when handling viscous
liquids and highly absorbing materials.
6.5 Software—The ideal software has the following capabilities:
6.5.1 The capability to record all sample identification and spectral data accurately and to access the reference data,
6.5.2 The capability to record the date and time of day that all spectra and files were recorded or created,
6.5.3 The capability to move or copy spectra, or both, from file to file,
6.5.4 The capability to add or subtract spectral data, and to average spectra,
6.5.5 The capability to perform transformations of log l/R optical data into derivatives, or other forms of mathematical treatment,
and to reverse the transformation,
6.5.6 The capability to compute multiple linear regression (MLR), principal component regression (PCR), and partial least squares
regression (PLS),
6.5.7 The capability to store PCR or PLS loading, weights, scores or other desirable data, and to display these data for subsequent
examination and interpretation,
6.5.8 The capability to enable the operator to evaluate the calibration model by computing the standard error of validation (SEV),
coefficient of regression, and the root mean square deviation (RMSD), and to display various plots,
6.5.9 The capability to perform cross-validation automatically,
6.5.10 The capability to identify an outlier(s), and
6.5.11 The capability to develop and save regression equations and analyze a sample to calculate a hydroxyl number.
6.6 Software Packages—Most NIR instruments provide necessary software for collecting and modeling data. Several non-
instrumental companies also supply chemometric software packages that can be used to analyze NIR data.
7. Near-IR Spectral Measurements
7.1 NIR spectral measurements are based on Beer’s law, namely, the absorbance of a homogeneous sample containing an
absorbing substance is linearly proportional to the concentration of the absorbing species. The absorbance of a sample is defined
as the logarithm to the base ten of the reciprocal of the Transmittance (T):
A 5 log 1/T (1)
~ !
D6342 − 22
where:
T = the ratio of radiant power transmitted by the sample to the radiant power incident on the sample.
7.1.1 For most types of instrumentation, the radiant power incident on the sample cannot be measured directly. Instead, a reference
(background) measurement of the radiant power is made without the sample being present in the light beam.
7.1.2 A measurement is then conducted with the sample present, and the ratio, T, is calculated. The background measurement can
be conducted in a variety of ways depending on the application and instrumentation. The sample and its holder can be physically
removed from the light beam and a background measurement made on the “empty beam”. The sample holder (cell) can be emptied,
and a background measurement taken for the empty cell. The cell can be filled with a material that has minimal absorption in the
spectral range of interest, and the background measurement taken. Alternatively, split the light beam or alternately pass the light
beam through the sample and through an empty beam, and empty cell, or a background material in the cell.
7.1.3 The particular background referencing scheme that is used can vary among instruments, and among applications. The same
sample background referencing scheme must be employed for the measurement of all spectra of calibration samples, validation
samples, and unknown samples to be analyzed. Any differences between instrument conditions used for referencing and
measurement are to be minimized.
7.2 Traditionally, a sample is manually brought to the instrument and placed in a suitable optical container (a cell, vial, or cuvette
with windows that transmit in the region of interest). Alternatively, transfer pipes can continuously flow liquid through an optical
cell in the instrument for continuous analysis. With optical fibers, the sample can be analyzed remotely from the instrument. Light
is sent to the sample through an optical fiber or fibers and returned to the instrument by means of another fiber or group of fibers.
Instruments have been developed that use a single fiber to transmit and receive the light, as well as use bundles of fibers for this
purpose. Detectors and light sources external to the instrument can also be used, in which case only one fiber or bundle is needed.
The appropriate grade of optical fibers for use in the NIR range needs to be specified. Generally, these are fibers with low water
content (Low-OH). Total fiber length is not to exceed manufacturer’s recommendations.
7.3 For most NIR instrumentation, a variety of adjustable parameters is available to control the collection and computation of the
spectral data. These parameters control the optical and digital resolution and the rate of data acquisition (scan speed). Other
important program parameters include the number of wavelengths, number of scans, and number of data points. Additional
instrumental considerations for multivariate calibrations include temperature control and compensation, cell pathlength uniformity,
and wavelength stability. It is essential that all adjustable parameters and other factors not included in the model that control the
collection and computation of spectral data be maintained constant while collecting spectra of calibration samples, validation
samples, and samples for analysis.
7.4 For definitions and further description of general infrared techniques, refer to Practice E168.
8. Procedure to Develop a Feasibility Calibration
8.1 For each type of polyol or new process to produce the polyol, it is necessary to perform a feasibility calibration. The
relationship between NIR spectra and the hydroxyl number is generally evaluated during a feasibility study which will identify the
possible interferences and determine whether an adequate model can be constructed for the desired precision. Following a
successful feasibility study, the calibration can be expanded and validated.
8.2 A sample set having all of the characteristics of the samples of interest is to be identified. The samples chosen are to include
the expected hydroxyl number ranges and all the possible interferences in the sample matrix. In addition, interrelations between
components in the samples are to be avoided unless these interactions are expected in the routine samples being analyzed. The
number of samples is to be large (preferably 30 to 50 samples, especially if PLS regression is used to evaluate the calibration
model) and is to be evenly distributed throughout the hydroxyl number range. The range of the sample set is to cover at least three
times, but preferably at least five times the standard deviation of the reference method. An independent set of samples, known as
the validation sample set, is to be identified and set aside to evaluate the calibration model for feasibility. The size of the validation
set will depend on the number of samples used to evaluate the calibration model, generally one sample for every four samples used
in the calibration.
8.2.1 If samples with a wide range of hydroxyl number are not available, it can be necessary to perform spiking experiments to
D6342 − 22
expand the range and to optimize the regression line. Blending samples to achieve varying hydroxyl numbers is preferred over
spiking. If spiking is utilized, care must be taken to avoid changes that can affect the matrix and ultimately compatibility with the
spectra of the calibration set.
8.3 Samples are to be collected in a manner which reflects the actual process conditions and sample handling techniques which
are expected to be used during routine collection for analysis. Sample spectra are to be collected in a manner which reflects the
actual conditions, techniques, and sample handling procedures which are expected to be employed. If these and other such
variations cannot be controlled, the range of variation is to be included in the sample set of calibration.
8.4 The calibration sample set is to be analyzed at least in duplicate by the primary or reference method. If the range of samples
is less than five times the standard deviation of the reference method, then r replicate analyses are to be performed such that (r) ⁄2
times the range of the calibration set is greater than three times (preferably five times) the standard deviation of the reference
method.
8.5 A calibration model is developed using one of several available methods, for example, MLR, PCR, and PLS. The quality of
the calibration curve can be tested by several statistical tools described in Section 12. The calibration model is tested using
cross-validation methods (see 12.8.4). Other statistics can also be used to judge the overall quality of the calibration.
8.6 If the statistical analysis of the calibration and validation sets suggests the method is capable of providing adequate precision,
the model can be tuned by adding additional samples to assure a boxcar distribution (an even distribution of values along a defined
interval of the hydroxyl number range). A final model can be developed and validated as described in Section 12.
8.7 If the calibration set consists of a range of different types of polyols, and reliable calibration cannot be obtained, it is necessary
to group samples by chemistry, and to develop a separate calibration model for each chemical grouping. Examples of possible
groups are polyether, polyester, segregation based on the manufacturing technique (that is, the EO/PO ratio), or the functionality.
9. Selection of Calibration Samples
9.1 Samples selected for the calibration set will ideally comply with the following guidelines as well as those established in 8.2:
9.1.1 The samples chosen are to include all components which are expected to be present in the samples of interest,
9.1.2 The samples chosen are to include and ideally exceed the expected hydroxyl number range,
9.1.3 The sample hydroxyl numbers are to be evenly distributed throughout the calibration range as to provide a “boxcar”
distribution of samples (evenly distributed throughout the range of interest),
9.1.4 The number of samples chosen are to be large enough to statistically define the relationship between the spectral variables
and the hydroxyl numbers to be modeled, and
9.1.5 The spectra of all samples are to be similar to avoid erroneous modeling. For example, the same path length is to be used
for all samples, and the baseline, peak maxima, and peak minima are to be similar. See 8.7.
9.2 The model is to exclude all potential sources of variation that can be excluded in the actual applications. If these sources cannot
be eliminated they must be included in the sample set, if possible. Sources of variation can include the following:
9.2.1 Chemical composition:
9.2.2 Physical characteristics, and
9.2.3 Sample handling, temperature, and humidity.
9.3 The number of samples required to calibrate the NIR model is dependent on the complexity of the samples being analyzed.
Simple models which contain only a few components that vary in concentration will have only a small number of spectral variables
D6342 − 22
and typically do not require a large sample set to define relationships. On the other hand, complex systems containing several
components which vary in concentration will require a large number of samples to define the relationships and to assure the model
development is adequate.
9.3.1 If a multivariate model is developed with 5 or fewer variables (wavelengths in MLR or factors in PCR or PLS), the
calibration must contain a minimum of 30 samples after elimination of outliers.
9.3.2 If a multivariate model is developed with k (>5) variables (wavelengths in MLR or factors in PCR or PLS), the calibration
set must include a minimum of 6 k samples after elimination of outliers.
10. Collecting NIR Spectra
10.1 Before developing a calibration model, it is necessary to determine the optimal pathlength at which to perform the analysis.
The optimal pathlen
...