A python package with the purpose of assessing the most reliable isotherm models (i.e., Langmuir, Freundlich, Temkin, Toth, Sips, DR), kinetic models (i.e., PFO, PSO, Weber-Morris, Avrami, Bangham, Elovich) and Arrhenius parameters (i.e., Ea, A), adsorption dynamic models (i.e., Thomas, Yoon-Nelson, Adams-Bohart), adsorption enthalpy and entropy as well as isosteric heat of adsorption. The package also offers the opportunity of calculating the amount of adsorbent needed in a scaled-up adsorption unit.
Install / Usage / Isotherms / Kinetics / Arrhenius / AdsorptionDynamics / AdsorptionEnthalpy / IsostericHeat / ScaleUP / References / Contact
First, make sure you have a Python 3 environment installed. To install from github:
pip install -e git+https://github.com/amvro23/adsorption/#egg=adsorptionNote: It might be useful to write "git+https://github.com/amvro23/adsorption/#egg=adsorption" if installing directly from a Python interpreter as # can be interpreted as a comment.
from adsorption.models import (Isotherms, Kinetics, ModifiedArrhenius,
AdsorptionDynamics, AdsorptionEnthalpy, IsostericHeat, Adsorbent_ScaleUp)
import numpy as np
import pandas as pdCreate an instance. Default optional values are P = 1 atm, Mr = 44.01 g/mol for CO2 adsorption, T = 298.15 K, R = 8.205e-5 atm.m3/mol/K).
isotherm = Isotherms()Set inlet values for adsorption isotherm equations. Optional parameter x represents the dimensionless equilibrium concentration [%] and optional parameter y represents the equilibrium capacity [mg/g].
isotherm.set_inlet()Adjust the values of x and y parameters according to your experimental results (the default values are the following).
x = np.array([0.002, 0.005, 0.01, 0.02, 0.05, 0.1, 0.2, 0.5, 1])
y = np.array([52.64, 72.59, 99.67, 141.79, 182.48, 203.68, 203.56, 204.33, 204.90])
isotherm.set_inlet(x, y)You can obtain either a single isotherm model plot (e.g., Langmuir model, Freundlich model),
isotherm.plot_langmuir_fit()
isotherm.plot_freundlich_fit()
or you can obtain all isotherm models.
isotherm.plot_all_models()
You can assess the fit of isotherm models.
isotherm.assess_fit()Out
{'Langmuir R2': 0.9817163923417866,
'Freundlich R2': 0.7463363496394846,
'Temkin R2': 0.8590282631439854,
'Toth R2': 0.9786970833618412,
'Sips R2': 0.9794054428262444,
'DR R2': 0.7707376307818574}
The package can directly return the best isotherm model.
isotherm.best_fit()Out
The best model is that of Langmuir R2 = 0.9817163923417866
You can also obtain a dataframe with all the calculated parameters of isotherm equations with their units,
isotherm.all_params()Out
Parameters Values
0 K_Langmuir [L/mg] 0.05971152127312045
1 qmax_Langmuir [mg/g] 211.83876099132755
2 K_Freundlich [L^(1/n_Freundlich)·mg^(1-1/n_Fre... 71.6650763022144
3 n_Freundlich 6.254943331258903
4 A_Temkin [L/mg] 2.9827960331000063
5 B_Temkin [J/mol] 90.1390756026659
6 K_Toth [mg^z·L^-n_toth] 15.637639450829793
7 n_Toth 0.9832070920173336
8 qmax_Toth [mg/g] 212.27301879029434
9 K_Sips [L^n_Sips·mg^-n_Sips] 0.0685096207219515
10 n_Sips 1.056671824783758
11 qmax_Sips [mg/g] 213.6471311521534
12 E_DR [J/mol] 147.67076438840127
13 qmax_DR [mg/g] 192.6633619518013
or you can have access in isotherm parameters individually in order to create your own plots.
isotherm.dubinin_radushkevich_params()Out
{'E_DR [J/mol]': 147.67076438840127,
'qmax_DR [mg/g]': 192.6633619518013}
isotherm.sips_curve(isotherm.x)Out
array([ 0.04084583, 0.09719205, 0.18721156, 0.36046663, 0.85595811,
1.64334054, 3.14432104, 7.33485598, 13.69841098])
You can get an excel file with the dataframe with all the calculated parameters.
isotherm.to_excel("Isotherms")Create an instance.
kinetic = Kinetics()Set inlet values for adsorption kinetic equations. Optional parameter x represents the adsorption time [min] and optional parameter y represents the accumulative adsorption capacity [mg/g].
kinetic.set_inlet()Adjust the values of x and y parameters according to your experimental results (the default values are the following).
df_kin = pd.read_csv('adsorption_kinetics.csv')
x = df_kin.loc[:, 'minutes'].values
y = df_kin.loc[:, 'qt'].values
kinetic.set_inlet(x, y)You can obtain either a single kinetic model plot (e.g., Bangham model),
kinetic.plot_bangham_fit()
or you can obtain all kinetic models.
kinetic.plot_all_models()
You can assess the fit of kinetic models.
kinetic.assess_fit()Out
{'PFO R2': 0.9835043733020796,
'PSO R2': 0.9762472082928675,
'WEBER-MORRIS R2': 0.9280676006898012,
'AVRAMI R2': 0.9980776118158866,
'BANGHAM R2': 0.9980776118158825,
'ELOVICH R2': 0.874909009634501}
The package can directly return the best kinetic model.
kinetic.best_fit()Out
The best model is that of AVRAMI R2 = 0.9980776118158866
You can also obtain a dataframe with all the calculated parameters of kinetic equations with their units,
kinetic.all_params()Out
Parameters Values
0 k_pfo [1/min] 0.018620160753212878
1 qmax_PFO [mg/g] 74.01352655075553
2 k_pso [g mg^-1 min^-1] 0.00012514880623542925
3 qmax_PSO [mg/g] 109.00232132690607
4 k_wm [mg g^-1 min^-0.5] 6.184044312110834
5 C 6.836063566443158e-18
6 k_avrami [1/min] 0.024543231342313475
7 qmax_avrami [mg/g] 63.298324486155906
8 n_avrami 1.4416620072834452
9 k_bangham [1/min^n] 0.0047733638932455965
10 qmax_bangham [mg/g] 63.29831524175902
11 n_bangham 1.4416628511124052
12 a_elovich [mg/g/min] 4.602398573502221
13 b_elovich [g/mg] 0.05477589170964753
or you can have access in kinetic parameters individually in order to create your own plots.
kinetic.avrami_params()Out
{'k_avrami [1/min]': 0.024543231342313475,
'qmax_avrami [mg/g]': 63.298324486155906,
'n_avrami': 1.4416620072834452}
kinetic.pso_curve(kinetic.x)Out
array([ 0., 0.08073146, 0.16136812, ..., 67.24288195,
67.25473634, 67.266584 ])
You can get an excel file with the dataframe with all the calculated parameters.
kinetic.to_excel("Kinetics")After finding the best kinetic model, the equation is applied to the data obtained at 3 to 4 different temperatures (at least) in order to create the following instance. In this example Bangham model is regarded as the best to describe the process of CO2 adsorption.
arrh = ModifiedArrhenius()Set inlet values for modified Arrhenius equation. Optional parameter x represents the temperatures [K] in which adsorption tests were performed and y represents the k value of the best kinetic model, which also carries the units (e.g., if Bangham model is more suitable k carries the units of k in bangham model) at the corresponding temperatures.
arrh.set_inlet()Adjust the values of x and y parameters according to your experimental results (the default values are the following).
x = np.array([298.15, 308.15, 323.15, 373.15])
y = np.array([0.00478, 0.00583, 0.00728, 0.01956])
arrh.set_inlet(x, y)You can obtain the values of Arrhenius parameters A and Ea. Given that bangham model is more suitable to describe the process, A adopts the units from k_bangham [1/min^n].
arrh.arrhenius_params()Out
{'Ea [J/mol]': 18300.09180621676,
'A [same as k]': 0.008242440438229868}
You can also obtain the plot and assess the fitting.
arrh.plot_arrhenius_fit()
arrh.assess_fit()Out
{'Arrhenius R2': 0.9949596964116147}
Create an instance. Default optional values are C=0.1, Mr=44.01 g/mol, T=298.15 K, P=1 atm, h=2 cm, r=0.45 cm, Q=100 ml/min, W=1 g, U=0.1, R=8.205e-5 atm.m3/mol/K) where C represents the initial concentration of the adsorbed molecule (CO2: 10%).
ads_dyn = AdsorptionDynamics()Set inlet values for adsorption dynamic equations. Optional parameter x represents the adsorption time [min] and optional parameter y represents the dimensionless concentration Ct/C0.
ads_dyn.set_inlet()Adjust the values of x and y parameters according to your experimental results (the default values are the following).
df_dyn = pd.read_csv('Co_10%.csv')
x = df_dyn.loc[:, 'x'].values
y = df_dyn.loc[:, 'y'].values
ads_dyn.set_inlet(x, y)You can obtain a single adsorption dynamic model plot (e.g., Yoon-nelson model).
ads_dyn.plot_yoon_nelson_fit()
Note that Adams-Bohart fit is used for the description of the initial part of the breakthrough curve ranging from 10-15% (default value is U=0.1).
ads_dyn.plot_adams_bohart_fit()
You can assess the fit of dynamic models.
ads_dyn.assess_fit()Out
{'THOMAS R2': 0.9958327653626917,
'YOON-NELSON R2': 0.9958327653619077,
'ADAMS-BOHART R2': 0.9983541738775606}
You can also obtain a dataframe with all the calculated parameters of dynamic adsorption equations with their units.
ads_dyn.all_params()Out
Parameters Values
0 k_thomas [ml/mg/min] 2.678369108434965
1 qmax_thomas [mg/g] 202.16211953564525
2 k_yoon_nelson [1/min] 0.4818450777710938
3 tau_yoon_nelson [min] 11.237299942413042
4 k_adams_bohart [ml/mg/min] 5.913583287817446
5 N0_adams_bohart [mg/ml] 137.6058956768907
You can get an excel file with the dataframe with all the calculated parameters.
ads_dyn.to_excel("AdsDynamics")Create an instance. Default optional values are C=0.01, Mr=44.01 g/mol, T=298.15 K, P=1 atm, R=8.205e-5 atm.m3/mol/K.
ads_H = AdsorptionEnthalpy()Set inlet values for adsorption enthalpy. Optional parameter x represents the adsorption temperature [K] and optional parameter y represents the equilibrium adsorption capacity [mg/g].
ads_H.set_inlet()Adjust the values of x and y parameters according to your experimental results (the default values are the following).
x = np.array([298.15, 308.15, 323.15, 348.15, 373.15])
y = np.array([203.6870035, 162.2365645, 116.2852302, 65.14332759, 34.46486588])
ads_H.set_inlet(x, y)You can obtain the values of Vant Hoff parameters enthalpy and entropy,
ads_H.vant_hoff_params()Out
{'enthalpy [J/mol]': -21754.33170685983,
'entropy [J/mol/K]': -52.31720971406195,
'R2': 0.9923449042417911,
'slope': 2616.5902943059696,
'intercept': -6.292664146507331}
Tha package returns directly the vant Hoff plot,
ads_H.plot_vant_hoff()
or your can obtain the numerical values to create your own plot.
ads_H.lnKdOut
array([2.4267528, 2.19922383, 1.86621433, 1.28675816, 0.65010871])
ads_H.vant_hoff_line(ads_H.x)Out
array([2.4834227 , 2.19862352, 1.80447432, 1.22303396, 0.71950333])
Create an instance. Default optional values are T1=273.15 K, T2=293.15 K.
iso_heat = IsostericHeat()Set inlet values for adsorption isosteric heat. Optional parameter x1 represents the absolute pressure at kPa for adsorption isotherm at T1 and y1 represents the equilibrium adsorption capacity [mmol/g] for adsorption isotherm at T1. The same stands for x2 and y2 but for adsorption isotherm at T2.
iso_heat.set_inlet()Adjust the values of x and y parameters according to your experimental results (the default values are the following).
df_iheat1 = pd.read_csv('iso_heat1.csv')
x_iheat1 = df_iheat1.loc[:, 'x'].values
y_iheat1 = df_iheat1.loc[:, 'y'].values
df_iheat2 = pd.read_csv('iso_heat2.csv')
x_iheat2 = df_iheat2.loc[:, 'x'].values
y_iheat2 = df_iheat2.loc[:, 'y'].values
iso_heat.set_inlet(x1=x_iheat1, y1=y_iheat1, x2=x_iheat2, y2=y_iheat2)Tha package returns instantly the Freundlich-Langmuir plot,
iso_heat.plot_freundlich_langmuir()
You can assess the fit of Freundlich-Langmuir model applied at both T1 and T2,
iso_heat.assess_fit()Out
{'Freundlich-Langmuir R2 for T1': 0.9998116123272973,
'Freundlich-Langmuir R2 for T2': 0.9999895124385948}
while you can also get access to the calculated parameters of the freundlich-langmuir model at both T1 and T2.
iso_heat.all_params()Out
Parameters Values
0 a at T1 [mmol/g] 5.931481472276362
1 b at T1 [1/kPa^c] 0.05028631551093073
2 c at T1 [dimensionless] 1.0559760862112229
3 a at T2 [mmol/g] 5.892741001110274
4 b at T2 [1/kPa^c] 0.016824160548525938
5 c at T2 [dimensionless] 1.0951835091030404
You can also get the dataframe of the desired parameters to create your own plots if necessary,
iso_heat.get_dataframe()and save them into an excel file.
iso_heat.to_excel("IsoHeat")The package directly returns the plot of isosteric heat (calculated via the Clausius-Clapeyron equation) vs adsorbed quantity.
iso_heat.plot_isoHeat_vs_mmol()
The package offers the opportunity of calculating the amount of adsorbent needed in a scaled-up adsorption unit.
Create an instance. Default optional values are Mr=34.1 as H2S is the targe molecule, and molar fraction is y=0.001.
scale_up = Adsorbent_ScaleUp()Set inlet values for the experimental and pilot unit. Default values are given herein. Adsorption capacity in mg/g for y=0.001, adsorption time in min for y=0.001, and flow rate in mL/min.
scale_up.exp_unit(ads_capacity=125, exp_ads_time=500)
scale_up.pilot_unit(total_flow_rate=2500, pilot_ads_time=2000)The amount of the adsorbent needed for the pilot unit is automaticaly printed in grams, however you can also have access to it as an attribute.
scale_up.quantity_ads_in_pilot_unitOut
60.89285714285715