Master instrument

To minimize calibration development costs, it would be ideal to produce a calibration using data obtained on only one of the analyzers (the master instrument), and simply transfer this calibration to all of the other analyzers (the slave instruments). 

This method can be considered a calibration transfer method that involves a simple instrument-specific postprocessing of the calibration model outputs [108,113]. It requires the analysis of a subset of the calibration standards on the master and all of the slave instmments. A multivariate calibration model built using the data from the complete calibration set obtained from the master instrument is then applied to the data of the subset of samples obtained on the slave instruments. Optimal multiplicative and offset adjustments for each instrument are then calculated using linear regression of the predicted y values obtained from the slave instrument spectra versus the known y values. 

The PDS method [108,109,114,115] can be very effective for spectroscopic analyzers, and other analyzers that generate data on a continuous variable axis. In PDS, the responses of a set of transfer standards are obtained on both the master and the slave instrument (thus producing and X, respectively). It is then desired to obtain a transformation matrix F such that the spectra obtained on the slave instrument can be transformed into spectra that would have been obtained on the master instrument ... 

Equation 12.56 indicates that the response that would have been obtained on the master instrument is simply a linear combination of the responses obtained on the slave instrument. However, in spectroscopy, it is very unlikely that the response at a specific wavelength on the master instrument should depend on any responses on the slave instrument that are at wavelengths far from the specified wavelength. As a result, the PDS method imposes a structure on the F matrix such that only wavelengths on the slave instrument that are within a specified range of the target wavelength on the master instrument can be used for the transformation. 

Figure 7.9. A possible metrological traceability chain for the result of a measurement of protein in a sample of grain. aTris = 2-amino-2-hydroxymethyl-1,3-propanediol b Dumas apparatus is calibrated using tris CRM and grain samples are certified in an interlaboratory study cthe master instruments measure grain samples to act as the grower s calibrator for field measurements.

This method is probably the simplest of the software-based standardization approaches.73,74 It is applied to each X-variable separately, and requires the analysis of a calibration set of samples on both master and slave instruments. A multivariate calibration model is built using the spectra obtained from the master instrument, and then this model is applied to the spectra of the same samples obtained from the slave instrument. Then, a linear regression of the predicted Y-values obtained from the slave instrument spectra and the known Y-values is performed, and the parameters obtained from this linear regression fit are used to calculate slope and intercept correction factors. In this... 

The wavelength uncertainty test verifies the accuracy and precision of the spectrophotometer x-axis. Typically, the x-axis will be in nanometers for a dispersion instrument and cm for a FT instrument. The use of cm for the spectral axis of an FT instrument is due to the mathematics of the interference term (Atkins 1996). The wavelength standards have stable isolated peaks usually based on a mixture of rare-earth oxides. The center of mass of the peaks is compared to standard values established on master instruments at National Institute of Standards and Technology (NIST). The typical tolerance values for the peak accuracy are 1 nm [19]. The observed precision values are usually much smaller than 1 nm due to the high reproducibility of modern spectrophotometers. The photometric linearity verifies that the y-axis of the spectrophotometer is linear over a typical refiectance range. The linearity is verified by scanning a series of standards of known reflectance (absorbance) values. The measured absorbance is plotted versus the standard values. The USP chapter specifies that the slope of this curve is equal to 1.0 0.05 with an intercept of 0.0 0.05. Photometric standards are available from instrument vendors and third party suppliers. 

For the standardization of different instruments, it should be noted that the stability of the standardization samples is not as critical as for the standardization of a single instrument over hme. When two NIR instruments must be standardized, the standardization samples need only be stable between the moment at which they are measured on the master instrument and the moment at which they are measured on the slave instrument. This period of time does not exceed a few minutes, if both instruments are located at the same place, and is not longer than a few days, if both instruments are far from each other. Therefore, less stable but more representative standardization samples can be used. For instance, Shenk and Westerhaus provide a set of 30 standardization cups containing different agricultural products that are representahve of most of the agricultural samples currently analyzed by NIR spectrophotometry [32,33]. These agricultural samples are probably not perfectly stable over several years, but because they can be considered as very stable over the space of a few days, they can therefore be used as standardization samples. 

This standardization approach consists of transferring the calibration model from the calibration step to the prediction step. This transferred model can be applied to new spectra collected in the prediction step in order to compute reliable predictions. An important remark is that the standardization parameters used to transfer calibration models are exactly the same as the ones used to transfer NIR spectra. Some standardization methods based on transferring spectra yield a set of transfer parameters. For instance, the two-block PLS algorithm yields a transfer matrix, and each new spectrum collected in the prediction step is transferred by simply multiplying it by the transfer matrix. For these standardization methods, the calibration model can be transferred from the calibration step to the prediction step using the same transfer matrix. It should be pointed out that all standardization methods yielding a transfer matrix (direct standardization, PDS, etc.) could be used in order to transfer the model from the calibration to the prediction step. For instrument standardization, the transfer of a calibration model from the master instrument to the slave instruments enables each slave instrument to compute its own predictions without systematically transferring the data back to the master instrument. 

This concept involves two methodologies transformation of the master monochromator spectra to look like spectra from another instrument model, and standardization of spectra to look like spectra from the master instrument of the same model. Transformed spectra are said to come from a virtual instrument because the instrument from which the spectra appear to come did not produce the spectra. A monochromator can be configured as a virtual tilting or fixed filter instrument. These virtual filter instruments can be used as master instruments to which actual filter instruments are standardized. Both methods rely on the spectra of the standardization samples to characterize the instruments. 

In addition, the GH and NH outlier tests should have the same sensitivity on the host instruments as they do on the master instrument. Our experience over the past 20 years with standardization of NIRS instruments, monochromators manufactured by others, filter instruments, and now the Infractec instrument has shown that this goal can be achieved using the methods described in this chapter. They have worked for ground and unground samples, dry and high moisture samples, solids, liquids and mixtures of the two, filter and monochromator instruments, and in reflectance and transmission. 

---