Module 03: Why We Bake
← Academy Why do we bake vacuum systems?
Outgassing & Temperature showed why heat works. This module is what we do with it: the recipe, the race, the sim, the "what comes out when" you actually operate a bake — and, at the far end, what Vent Physics takes back the moment you open the chamber again.
The mechanism, in one line: a molecule sits in an energy well of depth Ea, and its residence time τ collapses exponentially with temperature (r = ν · exp(−Ea/kBT), full derivation in the previous module). The consequence, which is where this module lives: a surface at 150 °C releases chemisorbed water ~3×104× faster than at room temperature (Ea 0.9–1.05 eV; this module's charts use 0.9). That factor is why bake-outs exist, and it is what the operational cards below explore.
The mechanism, in one line: a molecule sits in an energy well of depth Ea, and its residence time τ collapses exponentially with temperature (r = ν · exp(−Ea/kBT), full derivation in the previous module). The consequence, which is where this module lives: a surface at 150 °C releases chemisorbed water ~3×104× faster than at room temperature (Ea 0.9–1.05 eV; this module's charts use 0.9). That factor is why bake-outs exist, and it is what the operational cards below explore.
The wall during the bake, live
Same stratified wall you saw in Outgassing & Temperature, where the exponential is derived, and that Vent Physics plays in reverse when the vent breaks it. Here it runs forward: multilayers peel first (Ea 0.4–0.5 eV), the bound-water monolayer follows (Ea 0.9–1.05 eV; this canvas uses 0.9), the surface hydroxyls stay (the floor the bake never touches, above the bake range), and the bulk hydrogen drains only its skin, √(D·t) ≈ 0.24 mm after 48 h at 150 °C. Slide the temperature.
room to vacuum-furnace regime · slider re-rates in flight, no reset
bake clock: 0 h / 48 h
left the wall: 0 physi · 0 bound · 0 H
PHYSISORBED multilayers (~0.45 eV) — first to peel
BOUND WATER (0.9–1.05 eV) — leaves once exposed
OH BASE — the floor the bake never touches
H in bulk (0.55 eV diffusion) — drains a √(D·t) skin
Illustrative drain model: the temperature RATIO between layers and between bake settings is exact Arrhenius on the shared Ea, the absolute clock is anchored (bound water: decades at 20 °C, ~15 h time constant at 150 °C), not measured. Same honesty as the vent-physics inventory model.
Molecular Surface Simulation
Watch molecules desorb from a metal surface as temperature rises. Each dot is a molecule bound with a specific energy. Drag the temperature slider below — at 150°C the water flies off; at 250°C+ only stubborn H₂ remains.
ON SURFACE 40 / 40
DESORPTION RATE 0.0 /s
THERMAL ENERGY 0.026 eV
H₂O physi (0.45 eV)
H₂O chemi (0.90 eV)
HC (0.60–1.0 eV)
H recomb. (1.0 eV)
CO (1.10 eV)
Bake Race — 100°C vs 150°C
Two identical chambers. Same contamination. Different temperatures. Which one cleans first? Start the race and watch Arrhenius in action.
100°C
150°C
Chamber A — 100°C
30 / 300 h
Race Time
0 h
Chamber B — 150°C
30 / 300 h
Chamber A
Chamber B
The Arrhenius equation is why vacuum scientists obsess over bake temperature. A 50°C increase doesn’t give you 50 % more — it gives you ≈30× more (Ea ≈ 0.9–1.0 eV). The energy cost of running the heaters hotter is trivial compared to the days of pumping time you save.
Temperature Simulator
Drag the slider to see how temperature affects outgassing. The bars show relative desorption rates for 8 species commonly found in accelerator vacuum systems.
Desorption Rate (H₂O chemi)
1.0s−1
Multiplier vs 25°C
1×
Mean Residence Time
--
Regime
Ambient
Arrhenius Rate vs Temperature
Logarithmic desorption rate for chemisorbed H₂O (Ea=0.9 eV). The vertical marker follows the slider. Notice how the rate changes by orders of magnitude over the bake temperature range.
Activation Energies & Parameters
| Species | Ea (eV) | ν (s−1) | Binding | Typical Source |
|---|---|---|---|---|
| H₂O physisorbed | 0.45 | 1013 | Van der Waals | Ambient humidity |
| H₂O chemisorbed | 0.90 | 1013 | Hydroxyl (OH) | Surface oxide layer |
| CO₂ | 0.50 | 1013 | Weak chemisorption | Air exposure, decomposition |
| CO | 1.10 | 1013 | Metal-carbonyl | Oxide reduction, beam-induced |
| Light HC (C1–C3) | 0.60 | 1013 | Physisorbed | Pump oil, solvents |
| Heavy HC (C6+) | 1.00 | 1013 | Chemisorbed | Pump oil backstreaming |
| H recombinative (chemisorbed) | 0.9–1.1 | 1013 | Chemisorbed | Rate-limiting for UHV H₂ floor |
| H₂ bulk (diffusion) | 0.50 | 1010 | Interstitial | Dissolved in bulk steel |
Key relationships:
Mean residence time: τ = 1/r(T) = (1/ν) · exp(Ea / kBT)
Outgassing rate: q(T) ∝ n0 · r(T) (monolayer model)
Diffusion-limited: q(T) ∝ D0 · exp(−Ed/kBT) / √t (bulk H₂)
Mean residence time: τ = 1/r(T) = (1/ν) · exp(Ea / kBT)
Outgassing rate: q(T) ∝ n0 · r(T) (monolayer model)
Diffusion-limited: q(T) ∝ D0 · exp(−Ed/kBT) / √t (bulk H₂)
Numbers to Carry
- +50°C ≈ ×30 desorption rate (Eₐ 0.9–1.0 eV)
Sources · 8
- P.A. Redhead, Recommended practices for measuring and reporting outgassing data, JVST A 20, 1667 (2002).
- P. Chiggiato, Outgassing properties of vacuum materials for particle accelerators, CERN Accelerator School : Vacuum for Particle Accelerators, Glumslov 2017, arXiv:2006.07124.
- M.R. Louthan & R.G. Derrick, Hydrogen transport in austenitic stainless steel, Corrosion Science 15, 565 (1975).
- P.A. Redhead, J.P. Hobson, E.V. Kornelsen, The Physical Basis of Ultrahigh Vacuum, AIP, 1968.
- K. Jousten (ed.), Handbook of Vacuum Technology, 2nd ed., Wiley-VCH, 2016.
- M. Li & H.F. Dylla, Model for the outgassing of water from metal surfaces, JVST A 11, 1702 (1993).
- C. Benvenuti, Molecular surface pumping : the getter pump, CAS Vacuum Technology, CERN, 1999.
- SLAC-TN-23-003, Vacuum bake-out procedures for LCLS-II cryomodules, 2023.