Department of Chemistry, Iowa State University, Ames, Iowa 50011, United States.
Environ Sci Technol. 2023 Sep 19;57(37):14058-14070. doi: 10.1021/acs.est.3c04473. Epub 2023 Sep 7.
Titanium-containing nanoparticles (NPs) and submicrometer particles (μPs) in the environment can come from natural or anthropogenic sources. In this study, we investigate the use of single-particle inductively coupled plasma time-of-flight mass spectrometry (spICP-TOFMS) to measure and classify individual Ti-containing particles as either engineered (Ti-eng) or naturally occurring (Ti-nat) based on elemental composition and multielement mass ratios. We analyze mixtures of four Ti-containing particle types: anthropogenic food-grade TiO particles and particles from rutile, ilmenite, and biotite mineral samples. Through characterization of neat particle suspensions, we develop a decision-tree-based classification scheme to distinguish Ti-eng from Ti-nat particles and to classify individual Ti-nat particles by mineral type. Engineered TiO and rutile particles have the same major-element composition. To distinguish Ti-eng particles from rutile, we developed particle-type detection limits based on the average crustal abundance ratio of titanium to niobium. For our measurements, the average Ti mass needed to classify Ti-eng particles is 9.3 fg, which corresponds to a diameter of 211 nm for TiO. From neat suspensions, we demonstrate classification rates of 55%, 32%, 75%, and 72% for Ti-eng, rutile, ilmenite, and biotite particles, respectively. Our classification approach minimizes false-positive classifications, with rates below 5% for all particle types. Individual Ti-eng particles can be accurately classified at the submicron size range, while the Ti-nat particles are classified in the nanoregime (diameter < 100 nm). Efficacy of our classification approach is demonstrated through the analysis of controlled mixtures of Ti-eng and Ti-nat and the analysis of natural streamwater spiked with Ti-eng particles. In control mixtures, Ti-eng particles can be measured and classified at particle-number concentrations (PNCs) 60-times lower than that of Ti-nat particles and across a PNC range of at least 3 orders of magnitude. In the streamwater sample, Ti-eng particles are classified at environmentally relevant PNCs that are 44-times lower than the background Ti-nat PNC and 2850-times lower than the total PNC.
环境中的含钛纳米颗粒（NPs）和亚微米颗粒（μPs）可能来自自然或人为来源。在本研究中，我们使用单颗粒电感耦合等离子体质谱飞行时间（spICP-TOFMS）根据元素组成和多元素质量比来测量和分类含钛颗粒，以确定其是工程（Ti-eng）还是自然存在（Ti-nat）的颗粒。我们分析了四种含钛颗粒类型的混合物：人为的食品级 TiO 颗粒和来自金红石、钛铁矿和黑云母矿物样本的颗粒。通过对纯颗粒悬浮液的特性进行表征，我们开发了一种基于决策树的分类方案，以区分 Ti-eng 与 Ti-nat 颗粒，并根据矿物类型对单个 Ti-nat 颗粒进行分类。工程 TiO 和金红石颗粒具有相同的主要元素组成。为了区分 Ti-eng 颗粒和金红石，我们根据钛与铌的平均地壳丰度比开发了颗粒类型检测限。对于我们的测量，将 Ti-eng 颗粒分类所需的平均 Ti 质量为 9.3 fg，相当于 TiO 的直径为 211nm。从纯悬浮液中，我们分别证明了 Ti-eng、金红石、钛铁矿和黑云母颗粒的分类率为 55%、32%、75%和 72%。我们的分类方法最大限度地减少了假阳性分类，所有颗粒类型的分类率均低于 5%。Ti-eng 颗粒可以在亚微米尺寸范围内准确分类，而 Ti-nat 颗粒则在纳米范围内分类（直径<100nm）。通过分析 Ti-eng 和 Ti-nat 的受控混合物以及分析含有 Ti-eng 颗粒的天然河水，证明了我们的分类方法的有效性。在对照混合物中，Ti-eng 颗粒可以在 Ti-nat 颗粒浓度低 60 倍的情况下进行测量和分类，且浓度范围至少跨越三个数量级。在河水样本中，Ti-eng 颗粒在环境相关的 PNC 下进行分类，该浓度比背景 Ti-nat PNC 低 44 倍，比总 PNC 低 2850 倍。
