Sushko Maria L, Alexandrov Vitaly, Schreiber Daniel K, Rosso Kevin M, Bruemmer Stephen M
Pacific Northwest National Laboratory, Richland, Washington 99352, USA.
J Chem Phys. 2015 Jun 7;142(21):214114. doi: 10.1063/1.4921940.
High temperature intergranular oxidation and corrosion of metal alloys is one of the primary causes of materials degradation in nuclear systems. In order to gain insights into grain boundary oxidation processes, a mesoscale metal alloy oxidation model is established by combining quantum Density Functional Theory (DFT) and mesoscopic Poisson-Nernst-Planck/classical DFT with predictions focused on Ni alloyed with either Cr or Al. Analysis of species and fluxes at steady-state conditions indicates that the oxidation process involves vacancy-mediated transport of Ni and the minor alloying element to the oxidation front and the formation of stable metal oxides. The simulations further demonstrate that the mechanism of oxidation for Ni-5Cr and Ni-4Al is qualitatively different. Intergranular oxidation of Ni-5Cr involves the selective oxidation of the minor element and not matrix Ni, due to slower diffusion of Ni relative to Cr in the alloy and due to the significantly smaller energy gain upon the formation of nickel oxide compared to that of Cr2O3. This essentially one-component oxidation process results in continuous oxide formation and a monotonic Cr vacancy distribution ahead of the oxidation front, peaking at alloy/oxide interface. In contrast, Ni and Al are both oxidized in Ni-4Al forming a mixed spinel NiAl2O4. Different diffusivities of Ni and Al give rise to a complex elemental distribution in the vicinity of the oxidation front. Slower diffusing Ni accumulates in the oxide and metal within 3 nm of the interface, while Al penetrates deeper into the oxide phase. Ni and Al are both depleted from the region 3-10 nm ahead of the oxidation front creating voids. The oxide microstructure is also different. Cr2O3 has a plate-like structure with 1.2-1.7 nm wide pores running along the grain boundary, while NiAl2O4 has 1.5 nm wide pores in the direction parallel to the grain boundary and 0.6 nm pores in the perpendicular direction providing an additional pathway for oxygen diffusion through the oxide. The proposed theoretical methodology provides a framework for modeling metal alloy oxidation processes from first principles and on the experimentally relevant length scales.
金属合金的高温晶界氧化和腐蚀是核系统中材料降解的主要原因之一。为了深入了解晶界氧化过程,通过将量子密度泛函理论(DFT)与介观泊松-能斯特-普朗克/经典DFT相结合,建立了一个中尺度金属合金氧化模型,其预测重点是与Cr或Al合金化的Ni。对稳态条件下的物种和通量分析表明,氧化过程涉及空位介导的Ni和微量合金元素向氧化前沿的传输以及稳定金属氧化物的形成。模拟进一步表明,Ni-5Cr和Ni-4Al的氧化机制在定性上是不同的。Ni-5Cr的晶界氧化涉及微量元素的选择性氧化,而不是基体Ni的氧化,这是由于Ni在合金中的扩散速度相对于Cr较慢,并且与Cr2O3形成相比,氧化镍形成时的能量增益明显较小。这种基本上是单组分的氧化过程导致连续的氧化物形成以及氧化前沿前方单调的Cr空位分布,在合金/氧化物界面处达到峰值。相比之下,Ni和Al在Ni-4Al中都被氧化,形成混合尖晶石NiAl2O4。Ni和Al不同的扩散率导致氧化前沿附近复杂的元素分布。扩散较慢的Ni在界面3nm范围内的氧化物和金属中积累,而Al则更深地渗透到氧化物相中。Ni和Al在氧化前沿前方3-10nm的区域都被耗尽,形成空隙。氧化物的微观结构也不同。Cr2O3具有板状结构,沿晶界有1.2-1.7nm宽的孔隙,而NiAl2O4在平行于晶界的方向上有1.5nm宽的孔隙,在垂直方向上有0.6nm宽的孔隙,为氧通过氧化物的扩散提供了额外的途径。所提出的理论方法为从第一原理和实验相关长度尺度上对金属合金氧化过程进行建模提供了一个框架。