# Influence of grain size on the solid-state direct reduction of polycrystalline iron oxide

## TL;DR

This paper explores how the grain size of iron oxide affects its reduction process using hydrogen, impacting the efficiency and speed of steelmaking.

## Contribution

The study reveals how initial hematite grain size influences reduction kinetics and microstructure evolution during hydrogen-based direct reduction.

## Key findings

- Large-grained hematite samples reduce faster initially due to directional pore channels.
- Ultrafine-grained samples achieve more efficient reduction in later stages with a homogenous pore network.

## Abstract

Direct reduction of iron oxide using hydrogen offers a sustainable route to lower carbon emissions in steelmaking. Although iron oxide feedstocks consist of polycrystalline pellets, the influence of initial hematite grain size on direct reduction remains unexplored. Herein, the effect of grain size on reduction kinetics and microstructure evolution were uncovered using model polycrystalline hematite samples with large ( ~ 30 µm) and ultrafine ( ~ 1 µm) grains. Thermogravimetric analysis showed grain-size-dependent reduction behavior, while microstructural examination of partially reduced samples revealed that large-grained hematite forms finer directional pore channels due to fewer grain boundaries and orientation changes. Consequently, large-grained samples reduce faster initially as the pore network develops, while ultrafine-grained samples achieve more efficient reduction in later stages facilitated by a more homogenous pore network. These results demonstrate how grain size dictates porosity and texture evolution, providing fundamental insights relevant not only to hydrogen-based iron production but also to the design of porous materials by solid-state reduction processes.

Reduction of iron oxide using hydrogen offers a sustainable route to lower carbon emissions in steelmaking. Here, the effect of iron oxide grain size on reduction kinetics and microstructure evolution are uncovered, finding that large-grained samples reduce faster initially, while finer-grained samples achieve more complete and efficient reduction.