Page 65 - Read Online

Page 10 of 21 Wen et al. J. Mater. Inf. 2025, 5, 30 https://dx.doi.org/10.20517/jmi.2024.102

Next, the remaining molecules were ranked according to their comprehensive properties based on four
metrics: hole reorganization energy, solvation free energy, LogP, and SAScore. Given the significant
variations in value ranges for these metrics, Min-Max normalization was applied to scale all indicators to a
range of [0, 1]. For metrics where lower values indicate better properties (hole reorganization energy,
solvation free energy, and SAScore), reverse normalization was applied. Furthermore, to address the
limitations of spiro-OMeTAD, such as low hole mobility and high synthesis difficulty, higher weights were
assigned to these properties. Specifically, weights of 0.3 were assigned to hole reorganization energy and
SAScore, while solvation free energy and LogP were assigned weights of 0.2. The preliminarily screened
molecules were scored and ranked based on these weighted criteria. A flowchart of the process is shown in
Figure 3A, and the scores of the top 20 molecules are shown in Supplementary Table 2. The 6 highest-
scoring candidate molecules are screened out, denoted as M1-M6, which are presented in Figure 3B. To
assess the robustness of the weighting scheme, we systematically varied the weights of the four properties
(hole reorganization energy, solvation free energy, LogP, and SAScore) within a range of 0.1 to 0.4 using a
step size of 0.01 (maintaining the total sum of weights equal to 1), generating a total of 124 unique weighting
combinations. For each combination, the weighted scores of all candidate molecules were calculated, and
the distribution of the top 10 molecules with the highest scores is shown in Supplementary Figure 2A. The
molecular IDs and their occurrence frequencies are presented in Supplementary Table 3 and Supplementary
Figure 2B. Remarkably, 50% of the most frequently occurring top ten molecules across all weighting
scenarios (highlighted as red bars in Supplementary Figure 2B) overlap with our final selection,
demonstrating the stability of our weighted screening strategy.

The molecular properties of the six screened SM-HTM molecules M1-M6 are listed in Supplementary Table
4. Figure 4 compares the performance index data of the filtered M1-M6 molecules with that of several
common SM-HTMs, such as spiro-OMeTAD, DTPC8-ThDTPA, DTPC13-ThTPA, DTP-C6Th, YZ18,
YZ22, and TPA-TVT-TPA. The performance of the M1-M6 molecules remains promising when compared
to spiro-OMeTAD and several other HTMs. The HOMO levels of the six screened molecules are very close
to that of spiro-OMeTAD and are all higher than the common valence band energy level of perovskite
materials. This theoretically ensures that the perovskite material absorbs incident photons to form electron-
hole pairs, with the holes being able to enter the HTL more readily. The maximum light absorption peaks of
these molecules are all below 400 nm, avoiding overlap with the visible light absorption range of perovskite
materials. This ensures that M1-M6 molecules will not compete for light with the perovskite layer, allowing
perovskite materials to absorb and utilize sunlight more efficiently. In terms of hole reorganization energy,
as depicted in Figure 4A, the 6 molecules are similar to spiro-OMeTAD overall and show significant
improvement over YZ18 and YZ22, indicating that their hole mobility remains substantially excellent.
Figure 4B reveals that the solvation free energies of the 6 molecules are comparable to those of commonly
used HTMs, indicating their favorable solubility in chlorobenzene solution. As illustrated in Figure 4C, the
LogP values of the screened molecules are maintained around 10, significantly higher than that of YZ22,
indicating their superior hydrophobic properties. Furthermore, the SAScores of these molecules
demonstrate a significant improvement over YZ18, YZ22, and TPA-TVT-TPA, suggesting their promising
synthetic prospects, as shown in Figure 4D.

Although the screened SM-HTMs have not yet been experimentally synthesized and validated, their
synthetic accessibility is strongly supported by the design principles of previously reported and successfully
synthesized SM-HTMs [20-22,38] , combined with the use of the SAScore parameter in our screening process, a
metric specifically designed to evaluate synthetic feasibility. Furthermore, the stringent screening based on
critical performance parameters, including HOMO energy levels, hole reorganization energy, solvation free
energy, maximum absorption wavelength, and hydrophobicity, ensures their practical applicability. This