1: \begin{abstract}
2: Iron oxides and oxyhydroxides are challenging to model computationally as
3: competing phases may differ in formation energies by only several
4: kJ\,mol$^{-1}$, they undergo magnetization transitions with temperature, their
5: structures may contain partially occupied sites or long-range ordering of
6: vacancies, and some loose structures require proper description of weak
7: interactions such as hydrogen bonding and dispersive forces. If structures and
8: transformations are to be reliably predicted under different chemical
9: conditions, each of these challenges must be overcome simultaneously, while
10: preserving a high level of numerical accuracy and physical sophistication. Here
11: we present comparative studies of structure, magnetization, and elasticity
12: properties of iron oxides and oxyhydroxides using density functional
13: theory calculations with plane-wave and locally-confined-atomic-orbital
14: basis sets, which are implemented in VASP and SIESTA packages, respectively. We
15: have selected hematite ($\alpha$-Fe$_{2}$O$_{3}$), maghemite
16: ($\gamma$-Fe$_{2}$O$_{3}$), goethite ($\alpha$-FeOOH), lepidocrocite
17: ($\gamma$-FeOOH), and magnetite (Fe$_{3}$O$_{4}$) as model systems from a total
18: of 13 known iron oxides and oxyhydroxides; and use same convergence criteria and
19: almost equivalent settings in order to make consistent comparisons.
20: Our results show both basis sets can reproduce the energetic stability and
21: magnetic ordering, and are in agreement with experimental observations.
22: There are advantages to choosing one basis set over the other, depending on the
23: intended focus. In our case, we find the method using PW basis set most
24: appropriate, and combine our results to construct the first phase diagram of
25: iron oxides and oxyhydroxides in the space of competing chemical potentials,
26: generated entirely from first principles.
27:
28: keywords: iron oxides and oxyhydroxides; phase diagram; density functional
29: theory; modeling and simulation.
30: \end{abstract}
31: