CP2K learning pathways
MODEL · CONVERGE · VALIDATE · INTERPRET
Build a consistent Quickstep model, control basis and grid errors, and connect molecules, periodic materials and dynamics to the observable you need. Keep boundary conditions, data files and statistical sampling explicit.
Inputs are unexecuted teaching examples. Record executable versions, data provenance and units; establish convergence and scientific validity for your own system. External tools are identified separately.
Choose a calculation question
1 Input, GPW and GAPW foundations
Understand the input tree, boundary conditions and mixed representation.
1.1 A reproducible executable and input tree
Intuition and prerequisites.
Study the workflow →1.2 Read energy and force together
Intuition.
Study the workflow →1.3 Gaussian basis and GTH potential are separate decisions
Intuition and prerequisites.
Study the workflow →1.4 GPW and all-electron GAPW
Intuition.
Study the workflow →1.5 Environment, GPW and the input tree
Choose a consistent executable/data environment and understand input nesting.
Study the workflow →1.6 Shared inputs, conventions and evidence
Complete parent inputs, unit conventions and reproducibility evidence for this course.
Study the workflow →2 Molecules and crystals
Build consistent molecular and periodic reference models.
2.1 Isolated boundary conditions and box convergence
Intuition and prerequisites.
Study the workflow →2.2 Spin states and atomic reference energies
Intuition.
Study the workflow →2.3 A molecular energy difference with a controlled scan
Intuition and prerequisites.
Study the workflow →2.4 Build and audit crystalline silicon
Intuition.
Study the workflow →2.5 Complete isolated-water calculation
Read energy and force evidence for a defined nonperiodic molecule.
Study the workflow →3 Numerical reliability
Test grids, basis choices, SCF algorithms and restart consistency.
3.1 A two-dimensional grid-convergence experiment
Intuition.
Study the workflow →3.2 Orbital transformation as a controlled solver choice
Intuition.
Study the workflow →3.3 Diagonalization and finite-temperature occupations
Intuition.
Study the workflow →3.4 Brillouin-zone sampling and supercell logic
Intuition.
Study the workflow →3.5 SCF, grids and basis: separate tests
Build an observable-specific error budget and cross-check grid controls.
Study the workflow →3.6 Analysis, restarts and HPC
Audit saved state, units and parallel performance with defensible evidence.
Study the workflow →4 Structures and interactions
Validate forces, relaxed cells and energy differences.
4.1 Optimize geometry and prove what converged
Intuition.
Study the workflow →4.2 Cell optimization and an independent equation of state
Intuition.
Study the workflow →4.3 Noncovalent interaction and dispersion choices
Intuition.
Study the workflow →4.4 Counterpoise and basis-set superposition error
Intuition.
Study the workflow →4.5 Geometry optimization and boundary conditions
Verify forces independently and distinguish molecules, crystals and slabs.
Study the workflow →5 Interfaces and solutions
Control boundary conditions, finite sizes and structural statistics.
5.1 A neutral slab and its electrostatic boundary
Intuition.
Study the workflow →5.2 Prepare liquid water without calling a packing a liquid
Intuition.
Study the workflow →5.3 Explicit solvation and the difference between interaction and free energy
Intuition.
Study the workflow →5.4 Adsorption energies as a reference bookkeeping problem
Intuition.
Study the workflow →6 Molecular dynamics
Check integration, ensemble sampling and trajectory observables.
6.1 NVE energy conservation separates timestep and SCF error
Intuition.
Study the workflow →6.2 NVT and NpT answer different ensemble questions
Intuition.
Study the workflow →6.3 Restart continuity is a state-recovery experiment
Intuition.
Study the workflow →6.4 RDF, mean-square displacement and honest uncertainty
Intuition.
Study the workflow →6.5 Short MD with honest diagnostics
Compare integration and ensemble diagnostics without overstating a tiny trajectory.
Study the workflow →7 Molecular response and excitations
Connect a perturbation and observable with units and convergence tests.
7.1 Is this optimized molecule really a minimum?
Question and intuition.
Study the workflow →7.2 From a dipole trajectory to an IR spectrum
Question and intuition.
Study the workflow →7.3 A vertical excitation is a response calculation
Question and intuition.
Study the workflow →7.4 PBE0 with ADMM: separate model change from acceleration
Question and intuition.
Study the workflow →8 Free energy, QM/MM and advanced workflows
Validate collective variables, embedding boundaries and hybrid approximations.
8.1 Distance metadynamics and the missing coordinates
Intuition and prerequisite.
Study the workflow →8.2 Umbrella windows versus exact constraints
Intuition and prerequisite.
Study the workflow →8.3 Design and audit a QM/MM boundary
Intuition and prerequisite.
Study the workflow →8.4 Performance is useful only with reproducible science
Intuition and prerequisite.
Study the workflow →Reference guides
- 1.5 Environment, GPW and the input tree — Choose a consistent executable/data environment and understand input nesting.
- 2.5 Complete isolated-water calculation — Read energy and force evidence for a defined nonperiodic molecule.
- 3.5 SCF, grids and basis: separate tests — Build an observable-specific error budget and cross-check grid controls.
- 4.5 Geometry optimization and boundary conditions — Verify forces independently and distinguish molecules, crystals and slabs.
- 6.5 Short MD with honest diagnostics — Compare integration and ensemble diagnostics without overstating a tiny trajectory.
- 3.6 Analysis, restarts and HPC — Audit saved state, units and parallel performance with defensible evidence.
What counts as a completed calculation?
Explain the model, input choices, units, parent/restart data, convergence of the target quantity and acceptance criteria. Preserve raw inputs and outputs with version and dataset provenance. A successful process exit, numerical convergence and physical interpretation require separate evidence.