Lube-Tech PUBLISHED BY LUBE: THE EUROPEAN LUBRICANTS INDUSTRY MAGAZINE


The COAT system (Figure 1) is designed to deliver ASTM-identical results as separate methods, for either AN or BN at rates of up-to 100 samples/h. The evolution of this methodology has been extensively documented in the scientific literature (3) but its efficacy in a commercial setting has never been assessed or reported on. After substantive discussions with Thermal-Lube, Fluid Life determined that the COAT system and its methodology merited serious consideration and evaluation to determine whether quantitative FTIR analysis could be implemented and work in a commercial setting. Today, Fluid Life has four COAT systems in operation and has recently published a comprehensive assessment of the systems’ operational performance in the Journal of Laboratory Automation (4). This article provides a synopsis of that assessment and summarizes the principles and performance of the methodology and those interested in more detail are referred to the original JALA article.


The Basic Concepts The principle behind FTIRAN/BN


analysis is essentially identical


to the corresponding ASTM titrimetric methods; the use of stoichiometric acid-base reactions, but using infrared-active organic acids and bases so that the products can be measured spectrally rather than titrimetrically (2). In the FTIR methods, the acid or base is delivered in a solvent which acts both as a reaction medium and a diluent thus substantially lowering the viscosity of the sample. As a result the sample is readily pumped from an autosampler vial (~20 ml) into a ~200 µm KCl infrared cell using a micro-pump rather than the peristaltic or syringe pump usually required to handle viscous, neat oils (5). A primary calibration is developed by gravimetric addition of a pure acid (e.g., oleic acid) or base (e.g., 1-methyl imidazole) to additive-free mineral oil which is reacted with the solvated infrared active base or acid. The appropriate spectral changes induced are then measured to develop a calibration curve. One practical impediment to using weaker acids and bases is that their “titer” (mg KOH/g oil) will be less than that obtained using the ASTM methods (6). This used to be an issue as it required one to change one’s analytical frame of reference in terms of the AN or BN value used to condemn an oil, however, this problem has been resolved through the use of a unique mixed-mode chemometric calibration that ensures FTIR results are expressed in the identical terms as the corresponding ASTM reference procedure.


Calibration Basics Mixed-mode calibrations are devised through a combination of “ideal” and “real” oils, the latter selected to be representative of the type of in-service oils routinely analyzed. Chemically pure standards (ideal primary standards) are gravimetrically prepared in mineral oil and used to calibrate and validate the expected FTIR stoichiometric response in terms of mg KOH/g oil. Such ideal standards do not contain soot, oxidative products, glycol, water, various residual additives and/or their byproducts which may be present in “real” in-service oils. The presence of such constituents can perturb the stoichiometrically-induced changes in the infrared signals of interest, affecting the predictions obtained and need to be accounted for. This is achieved through the use of chemometrics (7), specifically Partial Least Squares (PLS), which relates the spectral changes in the sample spectra to the corresponding ASTM AN- or BN-derived values


No.98 page 2


obtained from the “real” calibration standards. Because the final calibration devised is directly related to data obtained from samples analyzed by ASTM procedures, the results are expressed in ASTM terms, avoiding the issue of the weaker infrared-active acids or bases producing lower “titer” values. By combining both ideal and real calibration standards, PLS is able to differentiate between spectral responses directly related to the ASTM reference method results and spectral changes that do not correlate. It is important to note that this chemometric approach is quite different from that advocated for the PLS- based FTIR direct BN method (1). The direct BN method is not anchored by well-defined stoichiometric reactions as the FTIRAN/BN


correlations which further deteriorate as the oil ages or is in extended service. Thus in the case of COAT FTIRAN/BN


chemometrics is predominantly used to refine and account for any spectral variability or interferences inherent to “dirty” in-service lubricants and to ensure that the “titer” reflects that of the ASTM reference method used in its development. As a consequence, the COAT FTIRAN/BN


methods do not


simply “estimate” AN or BN, but actually produce statistically equivalent ASTM results (4).


Calibration Performance Fluid Life invested significant effort in developing, assessing and validating its mixed-mode PLS calibrations and to validate their performance. As ASTM analyses are always ongoing, splitting samples and collecting the corresponding FTIR spectra and associated ASTM data was not that onerous a task or process, however, professional expertise was called upon for the chemometric calibration development. Prior to any calibration development, all sample handling, preparation and minimization of sample carryover required standardization, achieved largely through the assessment of calibration standards and analysis of standardized QC oils.


methods are and relies solely on poorly defined spectral methods,


Figure 2. FTIR leave-one-out cross-validation calibration plot for BN derived using mixed-mode ASTM BN calibration.


LUBE MAGAZINE NO.127 JUNE 2015 25


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44  |  Page 45  |  Page 46  |  Page 47  |  Page 48  |  Page 49  |  Page 50  |  Page 51  |  Page 52  |  Page 53