There have been major advances in the science to predict the likely environmental concentrations of nanomaterials, which is a key component of exposure and subsequent risk assessment. Considerable progress has been since the first Material Flow Analyses (MFAs) in 2008, which were based on very limited information, to more refined current tools that take into account engineered nanoparticle (ENP) size distribution, form, dynamic release, and better-informed release factors. These MFAs provide input for all environmental fate models (EFMs), that generate estimates of particle flows and concentrations in various environmental compartments. While MFA models provide valuable information on the magnitude of ENP release, they do not account for fate processes, such as homo- and heteroaggregation, transformations, dissolution, or corona formation. EFMs account for these processes in differing degrees. EFMs can be divided into multimedia compartment models (e.g., atmosphere, waterbodies and their sediments, soils in various landuses), of which there are currently a handful with varying degrees of complexity and process representation, and spatially-resolved watershed models which focus on the water and sediment compartments. Multimedia models have particular applications for considering predicted environmental concentrations (PECs) in particular regions, or for developing generic “fate factors” (i.e., overall persistence in a given compartment) for life-cycle assessment. Watershed models can track transport and eventual fate of emissions into a flowing river, from multiple sources along the waterway course, providing spatially and temporally resolved PECs. Both types of EFMs can be run with either continuous sources of emissions and environmental conditions, or with dynamic emissions (e.g., temporally varying for example as a new nanomaterial is introduced to the market, or with seasonal applications), to better understand the situations that may lead to peak PECs that are more likely to result in exceedance of a toxicological threshold. In addition, bioaccumulation models have been developed to predict the internal concentrations that may accumulate in exposed organisms, based on the PECs from EFMs. The main challenge for MFA and EFMs is a full validation against observed data. To date there have been no field studies that can provide the kind of dataset(s) needed for a true validation of the PECs. While EFMs have been evaluated against a few observations in a small number of locations, with results that indicate they are in the right order of magnitude, there is a great need for field data. Another major challenge is the input data for the MFAs, which depend on market data to estimate the production of ENPs. The current information has major gaps and large uncertainties. There is also a lack of robust analytical techniques for quantifying ENP properties in complex matrices; machine learning may be able to fill this gap. Nevertheless, there has been major progress in the tools for generating PECs. With the emergence of nano- and microplastics as a leading environmental concern, some EFMs have been adapted to these materials. However, caution is needed, since most nano- and microplastics are not engineered, therefore their characteristics are difficult to generalize, and there are new fate and transport processes to consider.

预测纳米材料可能的环境浓度的科学取得了重大进展，这是暴露和后续风险评估的关键组成部分。自 2008 年首次基于非常有限的信息的材料流分析 (MFA) 以来，我们已经取得了长足的进步，目前的工具考虑了工程纳米粒子 (ENP) 的尺寸分布、形态、动态释放和更全面的信息释放因素。这些 MFA 为所有环境归宿模型 (EFM) 提供输入，生成各种环境分区中颗粒流量和浓度的估计值。虽然 MFA 模型提供了有关 ENP 释放量的宝贵信息，但它们没有考虑命运过程，例如同质和异质聚集、转化、溶解或电晕形成。 EFM 在不同程度上解释了这些过程。 EFM 可分为多媒体隔室模型（例如，大气、水体及其沉积物、各种土地利用中的土壤），其中目前有一些具有不同程度的复杂性和过程表示，以及空间分辨流域模型，重点关注水和沉积物隔室。多媒体模型在考虑特定区域的预测环境浓度（PEC）或开发通用“命运因素”（即给定隔间中的总体持久性）以进行生命周期评估方面具有特殊的应用。流域模型可以跟踪水道沿线多个来源的排放到流动河流的运输和最终命运，提供空间和时间解析的 PEC。两种类型的 EFM 都可以在连续排放源和环境条件下运行，也可以在动态排放下运行（例如，随着新纳米材料引入市场或季节性应用而暂时变化），以更好地了解以下情况：可能会导致 PEC 峰值，更有可能导致超过毒理学阈值。此外，已经开发出生物累积模型，以根据 EFM 的 PEC 来预测可能在暴露的生物体中累积的内部浓度。 MFA 和 EFM 面临的主要挑战是针对观察到的数据进行全面验证。迄今为止，还没有任何实地研究可以提供真正验证 PEC 所需的数据集。虽然已经根据少数地点的一些观测结果对 EFM 进行了评估，结果表明它们处于正确的数量级，但仍然非常需要现场数据。另一个主要挑战是 MFA 的输入数据，它依赖于市场数据来估计 ENP 的产量。目前信息存在较大差距，不确定性较大。还缺乏稳健的分析技术来量化复杂基质中的 ENP 特性；机器学习或许能够填补这一空白。尽管如此，生成 PEC 的工具已经取得了重大进展。随着纳米塑料和微塑料成为主要的环境问题，一些 EFM 已适应这些材料。然而，需要谨慎，因为大多数纳米和微塑料都不是经过工程设计的，因此它们的特性很难概括，并且需要考虑新的命运和运输过程。