Figure 7 . (a) Predicted solvation diameter of BF4- varying with the pore width of nanoslit. (b) Capacitance with and without ion desolvation effect varying with the pore size of nanoslit. The blue dotted line is drawn to guide the eye.
Overall, the predicted capacitance in terms of pore width displays an oscillating feature, and besides, the capacitances in both cases become almost identical as the pore width exceeds 2.0 nm. This is because the solvation diameter of BF4- gradually recovers to its bulk value when increasing the pore width. Nevertheless, two important different features can be found between the capacitances considering ion-desolvation effect and that without ion-desolvation effect. Firstly, owing to the pore size limit, the capacitance with uniform ion size vanishes when the pore width is less than 0.73 nm; while for the case considering ion-desolvation, as the ion size can decrease down to 0.33 nm, the capacitance sharply increases when the pore width decreases from 0.73 to 0.33 nm, generally according with the experimental observation.12 Secondly, an obvious deviation in the capacitance can be observed when the pore width varies from 1.0 nm to 2.0 nm. This is likely due to the mismatch between and resulting from the oscillation of solvation diameter in confined MeCN.
The overall capacitance, , is calculated with the help of eq.(10) and the PSD extracted from the measurement of TiC-CDC microporous material.56Two types of overall capacitances are calculated, i.e ., the ones with and without ion desolvation. The calculated overall capacitances versus the average pore size are plotted in Figure 8 in comparison with the experimental measurement from Chmiola et al.12The predicted overall capacitance involving ion desolvation effect nicely agrees well with the experimental observation in both the magnitude and the variation trend along with the average pore size. Particularly, the sharp increase of capacitance is well captured as decreasing the average pore size. However, the overall capacitances without ion desolvation are overall much smaller than the experimental values, and in addition, the sharp increase of capacitance in nanopore is not captured. These differences highlight the importance of ion-desolvation effect in determining the capacitance of microporous electrode. Furthermore, the agreement between our predictions and experimental measurements shows that the capacitances of practical microporous electrodes involving liquid electrolytes can be quantitatively predicted by the proposed multiscale molecular approach.