Included here is a more easily accessible version of the database used for regressions in our publication “Isoreticular Curves: A Theory of Capillary Condensation To Model Water Sorption within Microporous Sorbents” (https://pubs.acs.org/doi/10.1021/jacs.4c02743)
To make a plot of RH vs capacity, please include the data from both tables
The regression can be performed in Mathematica with the following code:
nlm = NonlinearModelFit[
data, {E^(2*0.714/(lcd/ 2)({a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12, a13,
a14, a15, a16, a17, a18, a19, a20, a21, a22, a23, a24, a25,
a26, a27, a28, a29, a30, a31, a32} . {lbenz, larom, lthio,
lfura, lpyra, lpyri, lalke, lazo, fnh3cl, fnh2, fso3h, foh,
fcooh, fome, fno2, fnhme, fnhcyp, fch3, fc2h5, fcf3, fc2f5,
xome*nor, mzrnor, mcr*nor, mal*nor, mti*nor, mhf*nor,
mnicar*nor, mzncar*nor, cucar*nor, my*nor,
mnipyr*nor})/(Total[{lbenz, larom, lthio, lfura, lpyra, lpyri, lalke, lazo, mzr*nor, mcr*nor, mal*nor, mti*nor, mhf*nor, mnicar*nor, mzncar*nor, cucar*nor, my*nor,
mnipyr*nor}]))}, {a1, a2, a3, a4, a5, a6, a7, a8, a9, a10,
a11, a12, a13, a14, a15, a16, a17, a18, a19, a20, a21, a22, a23,
a24, a25, a26, a27, a28, a29, a30, a31, a32}, {lcd, lbenz, larom,
lthio, lfura, lpyra, lpyri, lalke, lazo, fnh3cl, fnh2, fso3h, foh,
fcooh, fome, fno2, fnhme, fnhcyp, fch3, fc2h5, fcf3, fc2f5, xome,
mzr, mcr, mal, mti, mhf, mnicar, mzncar, cucar, my, mnipyr, nor},
MaxIterations -> 10000]
nlm["RSquared"]
->0.974133
nlm["AdjustedRSquared"]
->0.9612
nlm["BestFitParameters"]
->{a1 -> -2.10906, a2 -> 0.0805185, a3 -> -0.172253, a4 -> -1.61392,
a5 -> -4.85542, a6 -> -6.71165, a7 -> -0.185092, a8 -> 4.88506,
a9 -> -3.70125, a10 -> -1.10911, a11 -> -8.24841, a12 -> -2.08043,
a13 -> -1.02591, a14 -> 0.60088, a15 -> 1.27376, a16 -> -2.21946,
a17 -> 0.680827, a18 -> 1.84476, a19 -> 2.55759, a20 -> 1.71299,
a21 -> 6.55719, a22 -> 4.20624, a23 -> -7.42209, a24 -> -18.3026,
a25 -> -5.27035, a26 -> -10.9464, a27 -> -5.81818, a28 -> -17.5136,
a29 -> -14.103, a30 -> -7.28204, a31 -> -0.942191, a32 -> -6.79632}
nlm[“ParameterErrors”]
->{0.731827, 0.560999, 0.932505, 1.26811, 1.30109, 2.28484, 0.731331, \
1.7948, 3.33006, 1.38964, 5.70549, 0.849904, 1.60553, 0.862944, \
1.21459, 4.8193, 4.1845, 0.467562, 0.583968, 0.600266, 1.00153, \
0.805048, 0.798352, 2.13928, 0.706503, 1.88895, 1.22613, 3.08362, \
3.78418, 1.47402, 1.70368, 1.06559}
Rasterize[
MatrixPlot[nlm["CorrelationMatrix"], PlotLegends -> Automatic]]
->
MOF water sorption database
COF water sorption database
All the data is bounded by the curve Qmax = −0.25(1/ln α)2 – 1.75(1/ln α)