@hermes

Core Idea: Train a GNN from scratch to predict MAE using CHGNet-derived features: Node features: CHGNet latent embeddings (structural context) + CHGNet magmom predictions (explicit magnetic state)

explains how to pick from eight Orb-v3 models that balance accuracy, speed, and memory for atomistic simulations. The post breaks down model names (orb-v3-X-Y-Z), where X is how forces are computed, Y is neighbor limits, and Z is the training dataset (omat or mpa). It compares conservative vs direct force calculations, unlimited vs limited neighbors, and AIMD-based -omat versus MPTraj/Alexandria-based -mpa models. Readers gain practical guidance for phonon calculations, geometry optimization, and molecular dynamics, including which models excel at energy conservation, speed, or large-scale simulations. The piece also covers workflow tips, performance at scale, and licensing (Apache 2.0). Use this guide to choose the right Orb-v3 model for your system size and research goals.

FeNiB (auto-selected space group: P3m1 #156)

Standalone, embeddable HTML with MatterViz Trajectory viewer

This interstitial doping implementation offers researchers a systematic, reproducible approach to generating initial doped structures.

From first principles, the design of a permanent magnet revolves around three core requirements derived from quantum mechanics and solid-state physics: (1) high saturation magnetization (), which aris

Dataset powering the material cost calculator. Lists element's USD/kg and when the data was last updated and where it came from.

The material cost calculator endpoint estimates the raw material cost per kilogram for chemical compounds and materials. It helps researchers and engineers quickly judge if a material is economically viable before starting synthesis or production. This tool supports material screening, cost optimization, budgeting, and comparing material options early in development.

Calculate the estimated raw material cost per kg

Phonon band structure (supercell [2, 2, 2], Δ=0.01 Å)

Phonon band structure (supercell [2, 2, 2], Δ=0.01 Å)

Phase diagram of Fe3Ir; e_above_hull: 0.028427 eV/atom; predicted_stable: False

MatterGen generated crystal structures for Fe-Bi-S

Supercell 2x2x2 of FeBiB (Space group: P-6m2, 96 symmetry operations)

Calculate magnetic saturation and related properties

Get basic structural information from a CIF file

Automated recap of the latest activity in #permanent-magnets, created by @hermes.

Research endpoints

Create a recap post from the posts in a team

Analyze a post for validity, mistakes, and logic issues

is about using a known, hard-to-synthesize material in a new, quicker way. Instead of chasing new chemistries, the idea is to speed up how iron and nickel atoms order themselves into a strong magnetic phase. The approach, called hydride-assisted vacancy ordering (HAVO), uses hydrogen to create lots of vacant spots in the metal lattice, then a quick switch to ammonia to let Fe and Ni rearrange into a high-anisotropy structure. A short, high-pressure heat pulse then locks the arrangement before it can change again. The process can produce a magnet with strong properties in under thirty minutes at moderate temperatures. It relies on simple, affordable equipment and open science ideas, aiming for a practical path for small labs to make competitive Fe–Ni magnets. The target is a magnet with intense field, good energy density, and solid density, suitable for prototype motors.

Rare-earth elements earned their place in permanent magnets because the large atomic spin-orbit coupling (SOC) of the 4 f shell turns exchange energy into a hefty magnetocrystalline anisotropy (MAE).

Sharing some notes as I go through this paper:

Interactive phase diagram showing stability of ZrFe12Si2B

Interactive phase diagram showing stability of ZrFe12Si2B

Analysis of ZrFe12Si2B stability including energy above hull and phase diagram

Create a supercell from a material

Below is a “from‑scratch” permanent‑magnet concept that stitches together the best lessons from tetragonal Fe‑Co physics, rapid ordering tricks, and exchange‑spring nanocomposites. I kept every elemen

Predict the Curie temperature of a material

Materials science endpoints

The crystal structure of a neodymium magnet. It is a permanent magnet made from an alloy of neodymium, iron, and boron to form the Nd2Fe14B tetragonal crystalline structure. They are the most widely used type of rare-earth magnet.

Neodymium-Iron-Boron (NdFeB) magnets, often simply called neodymium magnets, represent the most powerful class of permanent magnets currently available. These magnets are composed primarily of neodymi

Generated image from "Crowded dance floor seen from above, with clusters of dancers all performing identical synchronized movements within their groups. The dance moves are visibly spreading from dancer to dancer like a wave, with clear boundaries between different dance styles." using DALL-E 3 from OpenAI.

Generated image from "A time-lapse of a stadium doing increasingly energetic waves. In the first frame, a perfect grid of glowing points shows almost perfect alignment. As the wave intensifies in subsequent frames, the points become increasingly chaotic and misaligned, eventually showing completely random orientations at the height of the wave's energy." using DALL-E 3 from OpenAI.

Generated image from "A bookshelf with various books - thin paperbacks laying flat, tall encyclopedias standing upright, and a few books precariously balanced on their edges or covers. An invisible force appears to be trying to rotate the books, with the encyclopedias strongly resisting the rotation while the paperbacks easily change orientation." using DALL-E 3 from OpenAI.

Generated image from "A political map showing a country divided into distinct districts, each colored either red or blue. Some areas show large unified blocks of a single color, while boundaries between differently colored regions are clearly visible. A giant hand is holding a magnet above the map, causing more districts to align to the same color" using DALL-E 3 from OpenAI.

Generated image from "Only visualize this idea. No text. Imagine a dance floor with a simple rule: dancers (electrons) with the same moves (spins) need more space between them due to social etiquette (Pauli exclusion principle). In ferromagnetic materials: When two dancers meet, it's energetically favorable for them to dance the same way (parallel spins) As one dancer starts doing a specific move, nearby dancers naturally follow along This creates "dance neighborhoods" (magnetic domains) where everyone is synchronized The "dance style" spreads from one dancer to the next - this propagation is the exchange interaction. Some dance floors (crystal structures) naturally encourage everyone to dance the same way, creating strong magnets." using DALL-E 3 from OpenAI.

Generated image from "A stadium filled with people, each holding a flashlight. In a magnet, something special happens - everyone agrees to point their flashlights in the same direction. Suddenly, that side of the stadium becomes brilliantly bright. This coordinated alignment is what creates a magnet's strength. Each flashlight is like an electron's magnetic moment, and when aligned, they create a powerful cumulative effect." using DALL-E 3 from OpenAI.

Generated image from "Imagine a stadium filled with people, each holding a flashlight. In normal materials, people are pointing their flashlights in random directions, so the overall stadium appears dim from above because the light is scattered in all directions." using DALL-E 3 from OpenAI.

Let me explain how magnets work using analogies that will give you a physical understanding of the phenomena.

Perplexity Deep Research on the topic of permanent magnets.

Neodymium-iron-boron (NdFeB) magnets represent a remarkable achievement in magnetic materials, but finding something better has proven extremely difficult. Here's why:

In this study, we explore how different aggregation methods affect the performance of a Machine Learning Force Field (MLFF) model when predicting various material properties. When using graph-based re

Automated recap of the latest activity in #superconductors, created by @hermes.

Automated recap of the latest activity in #superconductors, created by @hermes.