Near-infrared (NIR) spectroscopy is a branch of vibrational spectroscopy that shares many of the principles that apply to other spectroscopic measurements. The NIR spectral region comprises two subranges associated with detectors used in the initial development of NIR instrumentation. The short-wavelength (Herschel or silicon region) extends from approximately 780 to 1100 nm (12,821–9000 cm–1); and longer wavelengths, between 1100 and 2500 nm, compose the traditional (lead sulfide) NIR region. Applications of NIR spectroscopy use spectra displayed in either wavelength or wavenumber units. As is the case with other spectroscopy measurements, interactions between NIR radiation and matter provide information that can be for both qualitative and quantitative assessment of the chemical composition of samples. In addition, qualitative and quantitative characterization of a sample's physical properties can be made because of the sample's influence on NIR spectra. Measurements can be made directly on samples in situ in addition to applications during standard sampling and testing procedures.

Applications of qualitative analysis include identification of raw material, in-process sample, or finished product. These applications often involve comparing an NIR spectrum from a sample to reference spectra and assessing similarities against acceptance criteria developed and validated for a specific application. In contrast, applications of quantitative analysis involve the development of a predictive relationship between NIR spectral attributes and sample properties. These applications typically use numerical models to quantitatively predict chemical and/or physical properties of the sample on the basis of NIR spectral attributes.

Vibrational spectroscopy in the NIR region is dominated by overtones and combinations that are much weaker than the fundamental mid-IR vibrations from which they originate. Because molar absorptivities in the NIR range are low, radiation can penetrate several millimeters into materials, including solids. Many materials, such as glass, are relatively transparent in this region. Fiber-optic technology is readily implemented in the NIR range, which allows monitoring of processes in environments that might otherwise be inaccessible.

The instrument qualification tests and acceptance criteria provided in this chapter may not be appropriate for all instrument configurations. In such cases, alternative instrument qualification and performance checks should be scientifically justified and documented. In addition, validation parameters discussed in this chapter may not be applicable for all applications of NIR spectroscopy. Validation parameters characterized for a specific NIR application should demonstrate suitability of the NIR application for its intended use.

Transmission and Reflection

Factors That Affect NIR Spectra

Sample Temperature— Sample temperature influences spectra obtained from aqueous solutions and other hydrogen-bonded liquids, and a difference of a few degrees may result in significant spectral changes. Temperature may also affect spectra obtained from less polar liquids, as well as solids that contain solvents and/or water.

Moisture and Solvent— Moisture and solvent present in the sample material and analytical system may change the spectrum of the sample. Both absorption by moisture and solvent and their influence on hydrogen bonding of the APIs and excipients can change the NIR spectrum.

Sample Thickness— Sample thickness is a known source of spectral variability and must be understood and/or controlled. The sample thickness in transmission mode is typically controlled by using a fixed optical path length for the sample. In diffuse reflection mode, the sample thickness is typically controlled by using samples that are “infinitely thick” relative to the detectable penetration depth of NIR light into a solid material. Here “infinite thickness” implies that the reflection spectrum does not change if the thickness of the sample is increased.

Sample Optical Properties— In solids, both surface and bulk scattering properties of calibration standards and analytical samples must be taken into account. Surface morphology and refractive index properties affect the scattering properties of solid materials. For powder materials, particle size and bulk density influence scattering properties and the NIR spectrum.

Polymorphism— Variation in crystalline structure (polymorphism) from materials with the same chemical composition can influence NIR spectral response. Different polymorphs and amorphous forms of solid material may be distinguished from one another on the basis of their NIR spectral properties. Similarly, different crystalline hydration or solvation states of the same material can display different NIR spectral properties.

Age of Samples— Samples may exhibit changes in their chemical, physical, or optical properties over time. Care must be taken to ensure that both samples and standards used for NIR analysis are suitable for the intended application.

Apparatus

Near-Infrared Reference Spectra

Qualification of NIR Instruments

Qualification— Qualification of an NIR instrument can be divided into three elements: Installation Qualification (IQ); Operational Qualification (OQ); and Performance Qualification (PQ). For further discussion, see the proposed general information chapter Analytical Instrument Qualification 1058.

Characterizing Instrument Performance— Specific procedures, acceptance criteria, and time intervals for characterizing NIR instrument performance depend on the instrument and intended application. Many NIR applications use previously validated models that relate NIR spectral response to a physical or chemical property of interest. Demonstrating stable instrument performance over extended periods of time provides some assurance that reliable measurements can be taken from sample spectra using previously validated NIR models.

Wavelength Uncertainty— NIR spectra from sample and/or reference standard materials can be used to demonstrate an instrument's suitable wavelength dispersion performance against target specifications. The USP Near IR System Suitability Reference Standard or the National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 2036 for reflectance measurement and NIST SRM 2035 for transmittance measurement can be used for wavelength verification. Suitable materials for demonstrating wavelength dispersion performance include polystyrene, mixtures of rare earth oxides, and absorption by water vapor for instruments that use an interferometer for wavelength dispersion. With appropriate justification, alternative standards may be used. Wavelength uncertainty typically is characterized from a single spectrum (collected with the same spectral resolution to obtain the standard value) using a minimum of three peaks that cover a suitable spectral range of the instrument. Typical tolerances for agreement with standard values are ±1.0 nm from approximately 700 to 2000 nm and ±1.5 nm above 2000 nm to approximately 2500 nm (±8 cm–1 below 5000 cm–1 and ±4 cm–1 from 5000 cm–1 to approximately 14,000 cm–1). Alternative tolerances may be used when justified for specific applications.

Photometric Linearity and Response Stability— NIR spectra from samples and/or reference standard materials with known relative transmittance or reflectance can be used to demonstrate a suitable relationship between NIR light attenuation (due to absorption) and instrument response. For reflectance measurements, commercially-available reflectance standards with known reflectance properties are often used. Spectra obtained from reflection standards are subject to variability as a result of the difference between the experimental conditions under which they were factory calibrated and those under which they are subsequently put to use. Hence, the reflectance values supplied with a set of calibration standards may not be useful in the attempt to establish an “absolute” calibration for a given instrument. Provided that (1) the standards do not change chemically or physically, (2) the same reference background is also used to obtain the standard values, and (3) the instrument measures each standard under identical conditions (including precise sample positioning), the reproducibility of the photometric scale will be established over the range of standards. Subsequent measurements on the identical set of standards give information on long-term stability. Photometric linearity is typically characterized using a minimum of four reference standards in the range from 10% to 90% reflection (or transmission). NIR applications based on measuring an absorbance larger than 1.0 may require standards with reflectivity properties between 2% and 5% reflection (or transmission) for characterizing instrument performance at low reflectance. The purpose is to demonstrate a linear relationship between NIR reflectance and/or transmittance and instrument response over the scanning range of the instrument. Typical tolerances for a linear relationship are 1.00 ± 0.05 for the slope and 0.00 ± 0.05 for the intercept of a plot of the measured photometric response versus standard photometric response. Alternative tolerances may occur when justified for specific applications.

Spectroscopic Noise— NIR instrument software may include built-in procedures to automatically determine system noise and to provide a statistical report of noise or S/N over the instrument's operating range. In addition, it may be desirable to supplement such checks with measurements that do not rely directly on manufacturer-supplied procedures. Typical procedures involve measuring spectra of traceable reference materials with high and low reflectance. Tolerances for these procedures should demonstrate suitable S/N for the intended application.

Introduction

Validation Parameters

Specificity— The extent of specificity testing depends on the intended application. Demonstration of specificity in NIR methods is typically accomplished by using the following approaches:

Linearity— Quantitative NIR methods generally attempt to demonstrate a linear relationship between NIR spectral response and the property of interest. Although demonstrating a linear response is not required for all NIR applications, the model chosen, whether linear or not, should properly represent the relationship.

Range— The specified range of an NIR method depends on the specific application. The range typically is established by confirming that the NIR method provides suitable measurement capability (accuracy and precision) when applied to samples within extreme limits of the NIR measurement. Controls must be used to ensure that results outside the validated range are not accepted. In certain circumstances, it may not be possible or desirable to extend the validated range to include sample variability outside the validated range. Extending the range of an NIR method requires demonstration of suitable measurement capability within the limits of the expanded range. Examples of situations in which only a limited sample range may be available are samples from a controlled manufacturing process and in-process samples. A limited method range does not preclude the use of an NIR method.

Accuracy— Accuracy in NIR methods is demonstrated by showing the closeness of agreement between the value that is accepted as either a conventional true value or an accepted reference value. Accuracy can be determined by direct comparison between NIR validation results and actual or accepted reference values. Suitable agreement between NIR and reference values is based on required measurement capability for a specific application. The purpose is to demonstrate a linear relationship between NIR results and actual values. Accuracy can be determined by agreement between the standard error of prediction (SEP) and the standard error of the reference method for validation. The error of the reference method may be known on the basis of historical data, through validation results specific to the reference method, or by calculating the standard error of the laboratory (SEL). Suitable agreement between SEP and SEL is based on required measurement capability for a specific application.

Precision— The precision of an NIR method expresses the closeness of agreement between a series of measurements under prescribed conditions. Two levels of precision should be considered: repeatability and intermediate precision. The precision of an NIR method typically is expressed as the relative standard deviation of a series of NIR method results and should be suitable for the intended application. Demonstration of precision in NIR methods may be accomplished using the following approaches:

Intermediate Precision— Intermediate precision can be shown by the following:

Robustness— NIR measurement parameters selected to demonstrate robustness will vary depending on the application and the sample's interface with the NIR instrument. Critical measurement parameters associated with robustness often are identified and characterized during method development. Typical measurement parameters include the following:

Ongoing Method Evaluation

• Addition of a new material to the spectral reference library
• Changes in the physical properties of the material
• Changes in the source of material supply
• Identification of previously unknown critical attribute(s) of material(s)

• Changes in the composition of the test sample or finished product
• Changes in the manufacturing process
• Changes in the sources or grades of raw materials
• Changes in the reference analytical method
• Major changes in instrument hardware

Outliers— Sample spectra that produce an NIR response that differs from the qualitative or quantitative calibration model may produce an outlier. This does not necessarily indicate an out-of-specification result; but rather an outlier indicates that further testing of the sample may be required and is dependent on the particular NIR method. If subsequent testing of the sample by an appropriate method indicates that the property of interest is within specifications, then the sample meets its specifications. Outlier samples may be incorporated into an updated calibration model subsequent to execution and documentation of suitable validation studies.