Jaejin Kim

Impact of pellet thickness on quantitative terahertz spectroscopy of solid samples in a polyethylene matrix.

Pellets composed of different weight-percent (wt-%) of lactose within a polyethylene (PE) matrix are used to examine how the physical thickness of solid samples impact analytical measurements performed over terahertz (THz) frequencies when using time-domain THz spectroscopy. Results indicate that the thickness of each pellet depends on the mass and physical properties of the individual components that comprise the pellet. Thickness of mixture pellets depends on the porosity of the individual pellet components. Porosity measurements presented here for PE and lactose give values of 25.6 ± 0.3 and 14.5 ± 0.1, respectively, which indicate that more air is trapped within the compressed PE matrix compared to that for lactose. This difference in porosity creates different pellet thicknesses for pellets of the same nominal mass but with different relative amounts of PE and lactose. For this binary matrix, the thickness of each pellet is found to be a linear combination of the compressed densities of the individual components. Analysis of the time-domain THz spectra reveals that thinner samples are confounded by a fringe pattern observed in the frequency-domain spectra. This fringe pattern is created by an etalon corresponding to the air/pellet interfaces for the sample in the optical path. Spectra collected from thicker pellets are confounded by a sloping baseline caused by scattering effects within the pellet matrix. The quantitative impact of pellet thickness is determined by comparing the mean standard error of calibration (MSEC) and mean standard error of prediction (MSEP) for a set of leave-three-out cross validation multivariate calibration models based on the partial least-squares (PLS) algorithm. Results indicate that PLS models are capable of analytical measurements with MSEC and MSEP values between 0.04 and 0.20 wt-%. Analysis of spectral variance captured within the corresponding spectral loadings for each model indicates that spectral variance is lowest for the 300 mg samples where the impact of scattering is minimal under conditions when the sample etalon is nonexistent.

Direct on-line Raman measurement of flying solid samples: Determination of polyethylene pellet density

We demonstrated an on-line Raman measurement of polyethylene (PE) pellet density when it is flying in a sample line. While in flight, pellets are sparsely populated at spectral collection, a spectral collection strategy covering a large spatial volume (larger number of pellets simultaneously) is necessary to acquire reasonable Raman intensity. In addition, the Raman measurement must be less sensitive to pellet position, because position and distribution are uncontrollable in a flying condition. To fulfill these requirements, a wide area illumination (WAI) scheme capable of covering a large sample volume (illumination volume: 0.7 cm(3)) was used when the pellets were flying in a 2.5-cm-diameter sample line. In addition, a long focal length (250 mm) was used so that minor changes in pellet position would not significantly affect the resulting Raman spectral feature. Although Raman intensity substantially decreased due to the large void space among flying pellets, a correct spectral feature representing PE was successfully obtained without any significant spectral distortion. Using partial least squares (PLS) regression, the prediction error under flying conditions was 0.0009 g cm(-3), which was comparable to that acquired when the pellets were packed (0.0008 g cm(-3)). When a conventional Raman scheme covering a smaller sample volume with a short focal length was used, the PE intensity decreased dramatically, and the resulting signal-to-noise ratio was not proper for quantitative analysis.

Comparison of near-infrared and Raman spectroscopy for on-line monitoring of etchant solutions directly through a Teflon tube.

Both near-infrared (NIR) and Raman spectroscopy have been studied for the quantitative measurement of components (H(3)PO(4), HNO(3), and CH(3)COOH) in an etchant solution and the corresponding prediction robustness has been evaluated. Both measurements were accomplished by illuminating radiation directly through a Teflon tube. Raman spectral features of each component were much clearer and more selective than those observed in the NIR spectrum. Especially, NIR spectral variation pertinent to H(3)PO(4) and HNO(3) were mostly based on the displacement and perturbation of water bands rather than due solely to NIR absorption. Therefore, the resulting spectral variations were not highly specific. When the validation set contained minor spectral variations resulting from a slight instrumental change, NIR prediction performance for all three components degraded substantially by showing obvious prediction bias. However, the accuracies of Raman predictions were maintained. Since partial least squares (PLS) models for each component were built using NIR spectra of poor specificity with broadly overlapping features, even minor spectral differences introduced by instrumental variations sensitively influenced the prediction performance of the NIR models. Overall, the selectivity (specificity) of a targeting spectroscopic method should be considered critically to secure prediction robustness for monitoring components in an etchant solution.

Simple transmission Raman measurements using a single multivariate model for analysis of pharmaceutical samples contained in capsules of different colors.

Direct transmission Raman measurements for analysis of pharmaceuticals in capsules are advantageous since they can be used to determine active pharmaceutical ingredient (API) concentrations in a non-destructive manner and with much less fluorescence background interference from the capsules themselves compared to conventional back-scattering measurements. If a single calibration model such as developed from spectra simply collected in glass vials could be used to determine API concentrations of samples contained in capsules of different colors rather than constructing individual models for each capsule color, the utility of transmission measurements would be further enhanced. To evaluate the feasibility, transmission Raman spectra of binary mixtures of ambroxol and lactose were collected in a glass vial and a partial least squares (PLS) model for the determination of ambroxol concentration was developed. Then, the model was directly applied to determine ambroxol concentrations of samples contained in capsules of 4 different colors (blue, green, white and yellow). Although the prediction performance was slightly degraded when the samples were placed in blue or green capsules, due to the presence of weak fluorescence, accurate determination of ambroxol was generally achieved in all cases. The prediction accuracy was also investigated when the thickness of the capsule was varied.

Feasibility of a Wide Area Illumination Scheme for Reliable Raman Measurement of Petroleum Products

A newly developed Raman collection scheme, a wide area illumination (WAI) scheme, was employed to demonstrate its utility for the analysis of petroleum products. For this purpose, the compositional analysis of simulated naphtha samples was attempted. The WAI scheme utilized a laser beam that illuminated a sample in a circular fashion with a diameter of 6 mm and a focal length of 250 mm. The reproducibility of the Raman measurement can be improved due to decreased sensitivity of the sample position as well as orientation with regard to the focal plane, as shown in a previous study.1 Near-infrared (NIR) spectroscopy, widely adopted in the field of petroleum refining, was also employed to compare with the prediction results obtained using the WAI scheme. Since the Raman spectral feature is more distinct and selective, the resulting calibration accuracy could be improved as long as reproducible Raman spectra could be collected. Overall prediction results using Raman spectroscopy were superior to those from NIR spectroscopy. The feasibility of the WAI scheme for reliable Raman analysis of petroleum products such as naphtha was demonstrated in this paper.

Investigation on Raman spectral features of a coated tablet under variation of its orientation respective to laser illumination and measurement of nominal coating thickness of packed tablets

To investigate Raman spectral features of a coated biconvex tablet under variation of its orientation respective to laser illumination, spectra of the tablet were collected by illuminating laser on 12 different locations on the tablet with 3 different illumination angles of 45, 75 and 90°. The spectral variations were more substantial when the tablet faces with engraved letters and greater surface curvature were measured, since the sampled volume of coating relative to that of a core tablet changed significantly under these circumstances as the illumination angle varied. The preliminary examination confirmed that the acquisition of tablet-representative spectra was the requisite for reliable measurement of coating thickness. Then, to mimic real monitoring of coating process, Raman spectra were directly collected on a packing of 30 tablets with repetition of random tablet packing up to 15 times and univariate models utilizing the intensity of coating peak at 638cm-1 were developed using the cumulatively averaged spectra with an average weight of the 30 tablets as a reference. To acquire less tablet orientation-sensitive spectra, a wide area illumination (WAI) scheme providing a large sampling area (28.3mm2) on a tablet with a long focal length (∼25cm) was employed. The averaging of the first to seventh spectra, equivalently utilizing more packing-representative spectra for quantitative analysis, made the measurement of nominal coating thickness of packed tablets accurate.

Diffuser-incorporated transmission NIR measurement for reliable analysis of packed granular samples

A diffuser-incorporated transmission near-infrared (NIR) scheme that enables direct spectral collection of packed granular samples with reliable sample representation and reproducibility has been demonstrated. The analytical utility of this method has been evaluated for the determination of polyethylene (PE) pellet density and the discrimination of the geographical origin of rice samples. Based on the preliminary observation of transmission spectral features acquired from spherical polyoxymethylene (POM) packings composed of different particle sizes as well as packing thickness, a portion of the radiation was propagated through the void space in the packing without fully interacting with the POM pellets. This type of radiation, so-called non-fully interacted radiation (NFIR), adversely affected the sample representation as well as the reproducibility of transmission measurements. To maximize the interaction of NIR radiation with granular samples, a polytetrafluoroethylene (PTFE) diffuser was positioned in front of the sample packing to introduce isotropically diffused radiation into the sample. This diffuser-incorporated scheme resulted in highly reproducible transmission spectra for both packed granular samples. Consequently, the density determination of PE pellets as well as discrimination of rice samples according to geographical origin was more accurate using the proposed scheme.

Rapid quantitation of atorvastatin in process pharmaceutical powder sample using Raman spectroscopy and evaluation of parameters related to accuracy of analysis

The purpose of this study was to determine the atorvastatin (ATV) content in process pharmaceutical powder sample using Raman spectroscopy. To establish the analysis method, the influence of the type of Raman measurements (back-scattering or transmission mode), preparation of calibration sample (simple admixing or granulation), sample pre-treatment (pelletization), and spectral pretreatment on the Raman spectra was investigated. The characteristic peak of the active compound was more distinctively detected in transmission Raman mode with a laser spot size of 4mm than in the back-scattering method. Preparation of calibration samples by wet granulation, identical to the actual manufacturing process, provided unchanged spectral patterns for the in process sample, with no changes and/or shifts in the spectrum. Pelletization before Raman analysis remarkably improved spectral reproducibility by decreasing the difference in density between the samples. Probabilistic quotient normalization led to accurate and consistent quantification of the ATV content in the calibration samples (standard error of cross validation: 1.21%). Moreover, the drug content in the granules obtained from five commercial batches were reliably quantified, with no statistical difference (p=0.09) with that obtained by HPLC assay. From these findings, we suggest that transmission Raman analysis may be a fast and non-invasive method for the quantification of ATV in actual manufacturing processes.