High Performance Laser Transmission Spectroscopy : a powerful technique to investigate colloidal suspensions

Abstract

In recent years, the study of nanoparticle suspensions has become fundamental in different fields, from medical biotechnologies (e.g. development of innovative vectors for the intracellular drug delivery), to the food industry (e.g. study of new active probiotical agents) and, in general, for the development of new materials. For such applications, the knowledge of the physical properties of the suspended nanoparticles, in particular size, shape, polydispersity, concentration, and volume fraction, is of paramount importance. Electron and Scanning Probe Microscopies have been widely used for the morphological characterisation, providing essential information on the structure, shape and size of single nanoparticles. However, these techniques usually provide a non-statistically meaningful information on size particle distributions, since the study is based on a limited number of images. The statistically significant information of a particle suspension is typically obtained from measurements of Dynamic Light Scattering (DLS). However, DLS has three important limitations: it does not provide absolute concentration values but relative to the different components only; size determination is provided in terms of hydrodynamic radius that generally overestimate the real one, as strongly depends on suspending fluid properties (i.e., temperature and viscosity) and fluid particle interactions; scattered light intensity shows a high power dependence on the particle radius, thus, a limited number of large particles (often even just impurities or dust particles) can mask the contribution of a large number of small particles. To overcome these drawbacks, alternative Laser Transmission Spectroscopy (LTS) techniques, based on the measurement of turbidity of the suspension, have been proposed by Li et al. LTS allows to determine the density distribution of colloidal suspensions measuring the transmittance of a laser beam as a function of the wavelength. The particle density distribution, as a function of their radius, can be calculated through the Beer-Lambert law and the Mie scattering theory. The transmission data can be analysed and inverted by a mean square root-based algorithm that outputs the particle density distribution as a function of size. During my PhD program, I have designed and assembled an innovative experimental apparatus for LTS, which allows to obtain much better performances with respect to that one proposed by Li et al. Furthermore, I have developed a Fortran code for data analysis. The optical source emplyed in this apparatus is a pulsed laser equipped with a Nd: YAG pump laser (pulse duration 3-5 ns), tuned through an optical paramentric oscillator over a wide range of wavelengths (from about 410 to about 2600 nm). This source is a complex optical system which, using a pump laser and the technology based on non-linear optics (parametric optical oscillators, generators of second and third harmonics), is able to produce a laser beam that can be tuned on a large wavelength range. The whole optical system is able to produce two identical beams: the first impinges on the sample, i.e. particles + solvent, and the second on the solvent only (reference). Following the method proposed by Li et al., in order to cancel out the wavelength dependence of the optical components and detectors, the currents generated by the photodetectors are collected two times, by swapping the sample and reference positions. The transmission coefficient is then calculated as the double ratio of these measured signals. With this configuration, LTS performances appear significantly better than ones obtained with DLS and electron microscopy. LTS capability to provide size and absolute concentration of particles has revealed fundamental for some interesting cases of study in different field as pharmaceutics, biophysics, and cultural-heritage. In particular, we studied liposomes for drug delivery, microvesicles derived from Microglia cells in brain mice cancer, and nanocapsules as biocide container for stone manufacts preservation. Although the absolute concentration estimate results to be accurate and rather robust, a close inspection of the results obtained from test measurements on standard polystyrene nanoparticles showed not negligible inaccuracies. In particular, the width of the size distributions was partly uncontrolled, due to the strong dependence of the analysis on the experimental errors. To overcome this drawback, in collaboration with the Department of Industrial and Information Engineering of the University of L’Aquila, we developed an innovative lock-in amplifier for the detection of the signal. Standard techniques for small-signal detection indeed typically employ lock in amplifiers whose sensitivity and resolution are limited by the selected full-scale. These characteristics maximise the apparatus sensitivity and resolution respect to the small amplitude variations of the input signal. In particular, we designed a new, highly performant, variable gain, lock-in amplifier that, optimized for repetition rate and pulse duration of our laser, allows to detect small changes of the input signal. Thanks to this custom variable gain lock-in amplifier, we proposed an alternative method based on a calibration procedure, in which the gain of the new custom lock in is tuned, for each wavelength, to balance the transmitted intensity in both the experimental optical paths. After the calibration run, it is possible to obtain the transmittance, and the related extinction coefficient, as the single ratio between the measured intensities. The performances of the new balanced method, which exploits the custom lock-in amplifier, result to be better than those ones obtained from the double ratio procedure. In particular, the width of the size distribution is remarkably reduced, providing a more reliable system characterisation. In view of a future LTS analysis, a strong effort has been also devoted to the study of Outer Membrane Vesicles (OMVs) produced by Escherichia Coli bacteria (i.e. the analogous of exosomes for bacteria) by DLS and Small Angle X-rays Scattering (SAXS). DLS results showed interesting membrane properties, such as temperature driven phase transition, which could allow to relate OMVs with the progenitor cells. Furthermore, SAXS analysis allowed to determine the membrane thickness variation across the transition.