jablonski 0.3 release

[olistarosta]

Jablonski, dyscolab's photochemical systems simulation library, just got an update including new features such as stochastic simulation, a revamped excitations API, and helper functions for excitations. Units in save_at for simulation are also now handled properly.

If you're new to jablonski you can see the getting started with jablonski or the original release announcement for a quick primer.

Stochastic simulation

Version 0.3 brings stochastic simulations with Gillespie's algorithm via an interface with the rebop library. The implementation is very similar to simbio's rebop interface, but support for piecewise simulation was added to support time dependent excitations of systems. To give an example, we can first declare a model

import pint
from jablonski import *
from jablonski.transitions import *
u = pint.get_application_registry()

class Model(SpectroscopicSystem):
    # Define states with energies, mid is a triplet state
    low: SingletState = initial(0 * u.eV, "singlet", default=100)
    mid: TripletState = initial(2 * u.eV, "triplet", default=0) 
    high: SingletState = initial(3 * u.eV, "singlet", default=0)

    # Define absorptions and emissions 
    absorption_1 = Absorption(ground=low, excited=high, rate=1e-15 * u.cm**2) 
    emission_1 = Phosphorescence(ground=low, excited=mid, rate=2e8 / u.s)
    emission_2 = IntersystemCrossing(source=high, target=mid, rate=0.5e8 / u.s)
    emission_3 = transitions.Fluorescence(ground=low, excited=high, rate=1e8 / u.s)

In this case we create a RebopSimulator instead of a normal one

from jablonski.rebop import RebopSimulator

rsim = RebopSimulator(Model)

then use the rebop_piecewise function to simulate

from jablonski.simulation import *

pulse = pulse_excitation(excitation=excitation, width=5*u.ns, start= 5*u.ns)
result = rebop_piecewise(rsim, events=pulse, upto_t=30 * u.ns, n_points=200)
result.pint.dequantify().to_dataframe().plot()

Revamped excitations API and helpers

The way excitations are handled in simulation functions was changed: instead of taking two separate parameters for excitation_transition and height, they now all take a dictionary with {excitation_transition: height}. So

pulse_excitation(Model.absorption_1, height=1e25 / (u.cm**2 * u.s), width=5 * u.ns)

becomes

pulse_excitation({Model.absorption_1: 1e25 / (u.cm**2 * u.s)}, width=5 * u.ns)

This makes it easier to excite multiple transitions, makes excitations reusable, and allows for the creation of helper functionsto create pumps easily. Version 0.3 includes pump_from_laser to calculate the effect of a laser of a certain wavelength, power, beam width, and linewidth.

pump_from_laser(Model, wavelength=413 * u.nm, power=1 * u.W, width=5 * u.um, linewidth=1 * u.nm)

Simulation time units

Previously, units in simulation times would just be stripped. Now simulation functions are aware of the units and time is scaled accordingly, so that
save_at=np.linspace(0,30e-9,100) * u.s and save_at=np.linspace(0,30,100) * u.ns have the same effect. The use of appropriate units in save_at is now required; using the wrong units will raise a warning, and in future versions it will raise an error.

Roadmap

Work on jablonski is ongoing. The next step is adding more analysis tools such as more sweeps, functions for fitting models to experimental results, and automatic lifetime calculation. A longer term goal is adding the concept of inewidths and line profiles to jablonski so that proper continuous emission and absorption spectra can be calculated.

For more information on jablonski see the documentation, or for more information about dyscolab packages see the dyscolab homepage.