Using neural networks for dynamic light scattering time series processing

Abstract

A basic experiment to record dynamic light scattering (DLS) time series was assembled using basic components. The DLS time series processing using the Lorentzian function fit was considered as reference. A Neural Network was designed and trained using simulated frequency spectra for spherical particles in the range 0–350 nm, assumed to be scattering centers, and the neural network design and training procedure are described in detail. The neural network output accuracy was tested both on simulated and on experimental time series. The match with the DLS results, considered as reference, was good serving as a proof of concept for using neural networks in fast DLS time series processing.

A Study of Milk Particles Size Variation with pH Change using Dynamic Light Scattering

“…1 and consists of a laser source, which was a Laser diode, 15 mW, with the wavelength of 633 nm, working in continuous regime, a circular glass tube for the liquid sample, a detector, a preamplifier with a linear response in the audio frequency range 20 -20000 Hz, a PC for recording and later on processing the time series. The scattering angle was chosen to be 90 o , which is typical for DLS experiments [6], [7], [9], [11][12][13]. The data acquisition rate was 16 KHz, which was sufficient for the particles range covered in this work.…”

“…( ) ( ) ( ) ( ) (1) If the suspended particles are considered to be monodispersed, the autocorrelation of the time series (ACR hereafter) has a simplified form [6], [11][12][13], [14]:…”

“…Firstly, constant B is subtracted, which makes ACR to be zero-based. Secondly, an experimental adjustment is made in such a manner that the detector will cover a single speckle, as described in detail in [11][12][13] and [14], making the spatial coherence factor, included in A for this particular case, to be equal to 1. After normalization, the ACR has a simple form depending on a single parameter, which is D:…”

“…A Matlab code was written and used to record the DLS time series [11][12][13]. The time length of the time series and the delay between consecutive recordings are adjustable parameters.…”

Multiple-Penalty-Weighted Regularization Inversion for Dynamic Light Scattering

“…A variety of the inversion methods have been proposed, including the cumulants method [6], the Laplace transform method [7], the non-negative least-squares method (NNLS) [8], the constrained regularization method (CONTIN) [9,10], singular value decomposition method [11], and the regularization method [12][13][14]. In addition, some improved techniques have been proposed containing a modified cumulants method [15][16][17], a regularized NNLS [18], a modified regularization algorithm [19], a modified truncated singular value decomposition [20], and many intelligent optimization-based algorithms [21][22][23][24][25] used in DLS inversion. Each of these methods has its own characteristics and limitations, and the inversion of the bimodal or multimodal distribution particles has always been a difficult problem.…”