Literature/202212281435 polydisperse samples with inta

<ul> <li>Source: [[<a class="litnote tooltip" href="/literature/@kashkanova2022">@kashkanova2022<span class="tooltiptext">Precision size and refractive index analysis of weakly scattering nanoparticles in polydispersions</span></a>]]</li> <li>Tags: <a href="/tags/Refractive-Index">#Refractive-Index</a> <a href="/tags/polydisperse-samples">#polydisperse-samples</a></li> </ul> <h2>Polydisperse Samples</h2> <p><img alt="images/Pasted image 20221228143222.png" class="wikiimage" src="/images/pasted image 20221228143222.png" /></p> <p>The paper makes a very clear case that for studying polydisperse samples, access to an accurate measurement of refractive index (or in the case of iNTA, <em>contrast</em> $C$) gives access to a better resolving power in two dimensions. It is clear that in some cases, like 15, 20, and 30nm gold particles, size alone is not enough to separate subpopulations, while the <em>contrast</em> is able to clearly distinguish between each other. </p> <p>In order to create clusters of particles, they have implemented a <a class="wikilink" href="/gaussian_mixture_model/">Gaussian Mixture Model</a>, which is an unsupervised machine learning model tool that fundamentally fits gaussian distributions of univariate samples. </p> <p>The paper also points out that there's a difference between the hydrodynamic radius and the nominal radius of the sample. For the gold particles it is $l_H=1.8\pm 0.3 nm$ for all particle sizes, which becomes a 10% discrepancy for the smallest particles, and there's a chance that more complex samples will have larger hydration effects.</p>