Merge pull request #414 from JuliaPhysics/charge_trapping

Implement charge trapping models
JuliaPhysics · Oct 7, 2024 · 8bbf0e8 · 8bbf0e8
2 parents 43a78c2 + 8e4f3f3
commit 8bbf0e8
Show file tree

Hide file tree

Showing 15 changed files with 476 additions and 50 deletions.
diff --git a/.github/workflows/Downgrade.yml b/.github/workflows/Downgrade.yml
@@ -3,11 +3,13 @@ on:
   pull_request:
     branches:
       - main
+      - dev
     paths-ignore:
       - 'docs/**'
   push:
     branches:
       - main
+      - dev
     paths-ignore:
       - 'docs/**'
 jobs:
@@ -22,10 +24,9 @@ jobs:
         with:
           version: ${{ matrix.version }}
       - name: Add LegendJuliaRegistry
-        run: julia -e 'using Pkg; Pkg.Registry.add(Pkg.RegistrySpec(url = "https://github.com/JuliaRegistries/General")); Pkg.Registry.add(Pkg.RegistrySpec(url = "https://github.com/legend-exp/LegendJuliaRegistry"))'
+        run: julia -e 'using Pkg; Pkg.Registry.add("General"); Pkg.Registry.add(Pkg.RegistrySpec(url = "https://github.com/legend-exp/LegendJuliaRegistry"))'
         shell: bash
       - uses: julia-actions/julia-downgrade-compat@v1
-#        if: ${{ matrix.version == '1.6' }}
         with:
           skip: Pkg,TOML
       - uses: julia-actions/julia-buildpkg@v1

diff --git a/docs/src/man/charge_drift.md b/docs/src/man/charge_drift.md
@@ -144,7 +144,6 @@ The first step is to define a `struct` for the model which is a subtype of `Soli
 ```julia
 using SolidStateDetectors
 using SolidStateDetectors: SSDFloat, AbstractChargeDriftModel
-using StaticArrays
 
 struct CustomChargeDriftModel{T <: SSDFloat} <: AbstractChargeDriftModel{T} 
     # optional fields to parameterize the model
@@ -155,6 +154,7 @@ The second step is to define two methods (`getVe` for electrons and `getVh` for
 Note, that the vectors are in cartesian coordinates, independent of the coordinate system (cartesian or cylindrical) of the simulation. 
 
 ```julia
+using StaticArrays
 function SolidStateDetectors.getVe(fv::SVector{3, T}, cdm::CustomChargeDriftModel)::SVector{3, T} where {T <: SSDFloat}
     # arbitrary transformation of fv
     return -fv
@@ -173,6 +173,95 @@ charge_drift_model = CustomChargeDriftModel{T}()
 sim.detector = SolidStateDetector(sim.detector, charge_drift_model)
 ```
 
+
+
+## Charge Trapping Models
+
+Electrons and holes can be trapped on their way to the contacts, resulting in a decrease in height of the observed charge pulse. SolidStateDetectors.jl allows for incorporating and testing different charge trapping models.
+
+In order to implement a charge trapping model, the first step is to define a `struct` for the model which is a subtype of `SolidStateDetectors.AbstractChargeTrappingModel`:
+
+```julia
+using SolidStateDetectors
+using SolidStateDetectors: SSDFloat, AbstractChargeTrappingModel
+
+struct CustomChargeTrappingModel{T <: SSDFloat} <: AbstractChargeTrappingModel{T} 
+    # optional fields to parameterize the model
+end
+```
+
+The second step would be to define a method for `_calculate_signal`, which returns a charge signal based on a given charge drift `path` and its `pathtimestamps`, the `charge` value of the charge cloud, the interpolated weighting potential `wpot` and the coordinate system type `S` (`Cartesian` or `Cylindrical`):
+
+```julia
+using SolidStateDetectors: CoordinateSystemType, SSDFloat
+using Interpolations
+function SolidStateDetectors._calculate_signal( 
+        ctm::CustomChargeTrappingModel{T}, 
+        path::AbstractVector{CartesianPoint{T}}, 
+        pathtimestamps::AbstractVector{T}, 
+        charge::T,          
+        wpot::Interpolations.Extrapolation{T, 3}, 
+        S::CoordinateSystemType
+    )::Vector{T} where {T <: SSDFloat}
+
+    # Implement method here
+end
+```
+
+!!! note
+    So far, the charge trapping models do not influence the charge drift path simulation, but just the signal calculation afterwards.
+
+### `NoChargeTrappingModel`
+
+If no charge trapping model is selected, the default is the `NoChargeTrappingModel`.
+In this case, the signal is calculated using the Schockley-Ramo theorem, i.e. by evaluating the weighting potential `wpot` at every point `path[i]` of the charge drift path and multiplying it with the `charge` of the charge cloud.
+
+
+### `BoggsChargeTrappingModel`
+
+In SolidStateDetectors.jl, the only charge trapping model implemented so far is the one presented in [Boggs _et al._ (2023)](https://doi.org/10.1016/j.nima.2023.168756).
+In this model, the charge cloud loses part of its charge at every point `path[i]` of the charge drift, depending on its drift and thermal velocity, as well as the trapping product $[n\sigma_{E/h}]^{-1}$ for electrons and holes.
+The charge signal is then given by the charge-decreased charge cloud reaching the contacts and the charges trapped on the way.
+
+The `BoggsChargeTrappingModel` can be applied in the configuration file by adding a field `charge_trapping_model` to the `semiconductor` with `model: Boggs` and `parameters` defining the (inverse) trapping products and (optionally) the `temperature`:
+```yaml 
+detectors:
+  - semiconductor:
+      material: #...
+      geometry: #...
+      charge_trapping_model:
+        model: Boggs
+        parameters:
+          nσe-1: 1020cm
+          nσh-1: 2040cm
+          temperature: 78K
+
+# ... or ...
+
+detectors:
+  - semiconductor:
+      material: #...
+      geometry: #...
+      charge_trapping_model:
+        model: Boggs
+        parameters:
+          nσe: 0.001cm^-1
+          nσh: 0.0005cm^-1
+          temperature: 78K
+```
+
+It can also be applied to an already read-in `SolidStateDetector` using for example:
+```julia
+using SolidStateDetectors, Unitful
+T = Float32
+sim = Simulation{T}(SSD_examples[:InvertedCoax])
+
+parameters = Dict("parameters" => 
+  Dict("nσe-1" => 1000u"cm", "nσh-1" => 500u"cm", "temperature" => 78u"K"))
+sim.detector = SolidStateDetector(sim.detector, BoggsChargeTrappingModel{T}(parameters))
+```
+
+
 ## Group Effects
 
 The movement of electrons and holes is not only given by the forces resulting from external electric fields. In addition, diffusion and self-repulsion of the charge carriers can play a significant role when simulating drift paths. To simulate this, electron and holes cannot be described as single point-like charges anymore, but as charge clouds consisting of multiple point-like charges. SolidStateDetectors.jl offers different models for the [Initial Charge Cloud Distribution](@ref). Right now, [Diffusion](@ref) and [Self-Repulsion](@ref) are implemented as experimental features.

diff --git a/examples/example_config_files/public_ivc_trapping_config.yaml b/examples/example_config_files/public_ivc_trapping_config.yaml
@@ -0,0 +1,159 @@
+name: Public Inverted Coax
+units:
+  length: mm
+  angle: deg
+  potential: V
+  temperature: K
+grid:
+  coordinates: cylindrical
+  axes:
+    r:
+      to: 40
+      boundaries: inf
+    phi:
+      from: 0
+      to: 0
+      boundaries: periodic
+    z:
+      from: -10
+      to: 90
+      boundaries:
+        left: inf
+        right: inf
+# grid:
+#   coordinates: cartesian
+#   axes:
+#     x:
+#       from: -40
+#       to: 40
+#       boundaries:
+#         left: inf
+#         right: inf
+#     y:
+#       from: -40
+#       to: 40
+#       boundaries:
+#         left: inf
+#         right: inf
+#     z:
+#       from: -10
+#       to: 90
+#       boundaries:
+#         left: inf
+#         right: inf
+medium: vacuum
+detectors:
+  - semiconductor:
+      material: HPGe
+      temperature: 78
+      impurity_density:
+        name: cylindrical
+        r:
+          init: 0
+          gradient: 0
+        z:
+          init: -1e7
+          gradient: -1e5
+      charge_drift_model:
+        include: ADLChargeDriftModel/drift_velocity_config.yaml
+      charge_trapping_model:
+        model: Boggs
+        parameters:
+          nσe-1: 1020cm
+          nσh-1: 2040cm
+      geometry:
+        difference:
+          - tube:
+              r: 35
+              h: 80
+              origin:
+                z: 40
+          - cone:
+              r:
+                bottom:
+                  from: 35
+                  to: 36
+                top:
+                  from: 23.71
+                  to: 36
+              h: 64
+              origin:
+                z: 52
+          - tube:
+              r: 5
+              h: 80
+              origin:
+                z: 65
+    contacts:
+      - material: HPGe
+        id: 1
+        potential: 0
+        geometry:
+          tube:
+            r: 3
+            h: 2
+            origin:
+              z: 1
+      - material: HPGe
+        id: 2
+        potential: 3500
+        geometry:
+          union:
+            - tube:
+                r:
+                  from: 15
+                  to: 35
+                h: 0
+            - tube:
+                r:
+                  from: 35
+                  to: 35
+                h: 20
+                origin:
+                  z: 10
+            - cone:
+                r:
+                  bottom:
+                    from: 35
+                    to: 35
+                  top:
+                    from: 24.42
+                    to: 24.42
+                h: 60
+                origin:
+                  z: 50
+            - tube:
+                r:
+                  from: 5
+                  to: 24.42
+                h: 0
+                origin:
+                  z: 80
+            - tube:
+                r:
+                    from: 5
+                    to: 5
+                h: 55
+                origin:
+                  z: 52.5
+            - tube:
+                r: 5
+                h: 0
+                origin:
+                  z: 25
+surroundings:
+  - material: HPGe
+    name: Passivation Layer
+    geometry:
+      tube:
+        r: 
+          from: 4
+          to: 14
+        h: 0.1
+        origin:
+          - 0
+          - 0
+          - -0.05
+    charge_density:
+      name: constant
+      value: 0 # Reasonable density: +2e-11 # => 2*10⁻¹1 C/m⁻³