HSPiT Library

Hansen Solubility Parameters in Theory (HSPiT) — a Python library for modeling solvent/solute miscibility, mixture stability, and solvent selection among other features, using Hansen Solubility Parameters (HSPs).

Install

HSPiT is pip installable, simply run in a terminal:

pip install HSPiT

HSP based activity coefficient model

Computes the activity coefficients of species in a mixture based on diffirences in HSPs

from HSPiT import Regula

# Set system parameters
x_lst=[1/3,1/3,1/3]                         # NA
hsp_lst=[(10,10,10),(15,15,15),(20,20,20)]  # MPa^0.5
mvol_lst=[50,50,50]                         # cm^3/mol
T=298                                       #K

# Define mixture
mixture=Regula(x_lst, hsp_lst, mvol_lst,T)


# Get Hansen distance between components 0 and 1
mixture.ra(0,1)

# Get activity coefficient of all components in mixture
mixture.gamma_v()

Stability analysis

The HSP-based activity coefficient can be used to perform a stability analysis of a mixture

from HSPiT import Regula

# Set system parameters
x_lst=[1/3,1/3,1/3]                         # NA
hsp_lst=[(10,10,10),(15,15,15),(20,20,20)]  # MPa^0.5
mvol_lst=[50,50,50]                         # cm^3/mol
T=298                                       #K

# Define mixture
mixture=Regula(x_lst, hsp_lst, mvol_lst,T)

# Stability analysis
mixture.is_stable_local() # computes local stability (spinodal)

mixture.is_stable_global() # computes global stability (binodal)

mixture.is_stable()  # computes local stability first (cheap), and only if that is True does it proceed to calculate global stability (expensive)

Solvent optimizer for stable mixtures

Methods for obtaining the HSP parameters of an additive that is likely to stabilize a given mixture. Method 1 is based on the activity coefficient model, and method 2 is based purely on geometric arguments in HSP+mvol space.

from HSPiT import Regula
from HSPiT import Nexus

# Set system parameters
x_lst=[1/3,1/3,1/3]                         # NA
hsp_lst=[(10,10,10),(15,15,15),(20,20,20)]  # MPa^0.5
mvol_lst=[50,50,50]                         # cm^3/mol
T=298                                       #K
N_max =1                                    # number of additives
phi_max=0.05                                # total volume fraction for all additives

# Define systems
mixture1=Regula(x_lst, hsp_lst, mvol_lst,T)   # Method 1
mixture2=Nexus(x_lst, hsp_lst, mvol_lst)      # Method 2

# Find optimal additive parameters
dd_add1,dp_add1,dh_add1, phi_add1 =           mixture1.optimize_additive(phi_max, N_max) # Method 1
dd_add2,dp_add2,dh_add2,mvol_add2, phi_add2 = mixture2.optimize_additive(phi_max, N_max) # Method 2

print(f"method 1: {dd_add1,dp_add1,dh_add1, phi_add1}")           # Method 1
print(f"method 2: {dd_add2,dp_add2,dh_add2,mvol_add2, phi_add2}") # Method 2

Surrogate formulator

Forming surrogates with a reduced number of components representative of an original mixture. A factor k is used to indicate the degree of prioritization for representation of outlying or isolated components in solubility space, as these are likely to have the greatest influence on mixture stability. k=1 gives equal priority, and k>1 leads to increasingly higher priority in representing isolated components.

from HSPiT import surrogate

# Define system parameters
hsp_lst=[...]  # list with original HSP
mvol_lst=[...] # list with original molar volumes
phi_lst=[...]  # list with original volume fractions
k=...          # prioritization factor for isolated components
N_max=...      # number of surrogate representative points

# Returns parameters for surrogate formulation
hsp_lst,mvol_lst,phi_lst=surrogate(hsp_lst, mvol_lst, phi_lst, k, N_max)

HSP predictor

Group contribution method

This includes an implementation of the Stefanis-Panayiotou group contribution method. The functional group data is embedded in the function. One needs only pass a SMILES string, and the function will automatically dissect it into its constituent functional groups and access the parameter values for use in the model. Additionally, it will automatically trigger either the low or the high path discussed in the original publication.

from HSPiT import stefanis_gc

smiles="CCO"

dd,dp,dh=stefanis_gc(smiles)

Fitting based on activity coefficients

This method utilizes the HSP-based activity coefficient model along with UNIFAC-Dortmund to fit the HSP for a target compound. The HSP is the argument that minimizes the discrepancy between the two activity coefficient models, across a set of binary mixtures.

from HSPiT import hsp_fit

smiles = "CCO" # NA
mvol = 58.4    # cm^3/mol

# Returns the fitted HSP parameters
dd,dp,dh=hsp_fit(smiles,mvol)