Principal frequency component analysis based on modulate chopper technique used in diffuse reflectance spectroscopy measurement

Abstract

Diffuse reflectance spectroscopy (DRS) is significantly affected from the interference of the ambient light and dark current of the instrument. Optical choppers, together with lock-in/synchronous amplification, can overcome these interferences. However, in spectral measurement, the sampling rate of the spectrometer is different from the Δ-pulse sampling, which is not high enough because of the integration time. In addition, the energy distribution is not perfectly concentrated as expected in modulate chopper technology. Therefore, in this study, based on the modulate chopper technique, we proposed a principal frequency component analysis (PFCA) method for DRS. This technique not only effectively eliminated the interference and dark current of the instrument but also improved the measurement precision using the energy of different frequencies. First, experiments were designed to successfully verify the function of optical choppers, eliminating the interference of the ambient light. Second, a set of 64 mixture solutions was designed and measured by DRS using the PFCA method to prove the feasibility of the proposed method. The solution was mixed with intralipid-20% suspension, India ink, and rhodamine B. These samples were analyzed by DRS under different conditions: no-chopper with overlapping and averaging, chopper demodulated by Fourier transform, and chopper demodulated by PFCA. The partial least square regression analysis was implemented to predict the concentration. Compared to the result of three methods, DRS equipped with chopper using the PFCA method showed the best results. The results of this study showed that the PFCA method not only satisfactorily eliminated the interference signals but also extracted useful information as much as possible, improving the analysis accuracy.

Optical choppers: novel approaches for higher chop frequencies

“…In addition, disks choppers with windows with linear margins (what we called "classical" choppers) have a limited possibility of generating different profiles of laser impulses, although we have demonstrated that, using certain relationships between its geometrical parameters and those of the laser beam, one can obtain a range of such profiles [14,15], even for the simpler case (from an analytical point of view) of top-hat incident laser beams. Therefore, we have first proposed and patented [17] a configuration of chopper disk with non-linear margins, outward or inward oriented.…”

“…Therefore, the transmission functions of the latter can be obtained as particular cases, when ρ → ∞ (thus, the windows margins would become linear). We have named these novel devices "eclipse" choppers, because of the way the obscuration of the section of the beam in the plane of the disk occurs, similarly to a planetary eclipse [15][16][17], and somehow similar to the way vignetting is produced in optical systems [35].…”

“…Choppers with rotational elements are one of the most utilized optomechanical modulators [1], together with optical scanners of different types [2][3][4][5][6]. Most common choppers are built with rotational disks [7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24], but we have introduced, patented [25], and developed [26][27][28][29][30] a novel type of choppers, with rotational shafts of different shapes (for example cylindrical, conical, or spherical), with slits of different profiles (i.e., with linear margins or as cylindrical holes). Such devices have the potential of surpassing the well-known 10 kHz limit of the chop frequency fc of choppers with disks.…”

“…Also, macro-choppers are a more robust solution from a mechanical point of view. Many of their applications (for example for radiometers and pyrometers [7][8][9], telescopes [10,11], laser manufacturing [12,13], polarimetry [14], spectroscopy [15], light attenuation [24] (as an alternative to other devices [33]) and biomedical systems [16]) are in fixed laboratory-based systems, therefore their (higher) weight and dimensions are not major issues.…”

Modeling with Multiple Correlated Spectral Data Based on Approximating the Nonlinear Spectrum Induced by Scattering

“…Song et al. 38 improved the S/N by reconstructing the signal with the fundamental frequency coefficient, while eliminating the background light interference. The spectrum was collected by short-time multiple sampling and obtained multiple frequency information.…”