betaFeedback

WorksPhysics and Astronomy

Functionality Optimization for Singlet Fission Rate Screening in the Full-Dimensional Molecular and Intermolecular Coordinate Space

Abstract

In computational chemistry, accurately predicting molecular configurations that exhibit specific properties remains a critical challenge. Its intricacies become especially evident in the study of molecular aggregates, where the light-induced functionality is tied to highly structure-dependent electronic couplings between molecules. Here, we present an efficient strategy for the targeted screening of the structural space employing a "functionality optimization" technique, in which a chosen descriptor, constrained by the ground state energy expression, is optimized. The chosen algorithmic differentiation (AD) framework allows to automatically obtain gradients without its tedious implementation. We demonstrate the effectiveness of the approach by identifying Perylene Bisiimide (PBI) dimer motifs with enhanced SF rate. Our findings reveal that certain structural modifications of the PBI monomer, such as helical twisting and bending, as well as slipped-rotated packing arrangements, can significantly increase the SF rate.

Introduction

The prediction of molecular arrangements that facilitate or enhance a certain process represents a key research question in computational chemistry, spanning fields such as drug design,[1] (bio)catalysis[2, 3] as well as molecular electronics[4]. Its intricacies are particularly evident in the studies of molecular aggregates, in which highly structure-dependent electronic couplings between monomers dictate the pathway of light-induced processes, leading to functionalities such as Singlet Fission (SF).[5] The latter has garnered significant attention, as it holds the potential to exceed the theoretical efficiency limits of single-junction solar cells by yielding two triplets for the price of one photon.[6, 7]

So far, only very few materials facilitate efficient SF, spurring extensive research aimed at identifying new potential molecular candidates[8–12] as well as structural arrangements[13–21] that enable SF. However, an unbiased screening of the full structural space requires an enormous amount of electronic structure calculations. An illustrative example of the high computational demand consists in recent work by Zaykov et al.[22]: keeping the monomer structure rigid, they scanned the six-dimensional conformational space of the ethylene dimer, in order to identify the best packing motif for SF, carrying out several billion electronic structure calculations. Typically, such screening is conducted using a narrowed selection of intermolecular stacking configurations. Additionally, the rigidity assumed for the monomer structure precludes the consideration of intramolecular structural adjustments, although molecular contortion (such as twisting and bending) and the employment of nonplanar Poly­aromatic Hydrocarbons has been proven as an successful approach in all fields of organic electronics.[23–28] In particular with respect to Perylene Bisimides, numerous experimental studies have shown how contortion leads to unique optoelectronic characteristics and further serves as a rational tool to adjust orbital energies as well as the energies of electronic transitions.[29] In fact, bending of PBIs has been employed to align S1S_1 and T1T_1 energies for enhanced SF.[30] This aspect has been also theoretically addressed by Cruz et al., studying a series of Pyrenacenes.[31] However, the impact of contortion on SF diabatic couplings is far less explored. An alternative theoretical strategy that allows for an more efficient, targeted screening of molecular arrangements might consist in the development of property-based optimization procedures. It is important to note that an unrestricted geometry optimization of a chosen quantity attributed to the ”functionality” is not straightforwardly applicable, as the descriptors might not be constrained by the (ground state) potential energy of the system, potentially leading to ”unphysical” atomic configurations. Successful strategies consist e.g. in the development of metadynamics-based methodologies that allow for an automated extraction of structures with certain electronic properties, employing molecular dynamics simulations, in which iteratively added bias potentials steer the molecular system towards structural arrangements with high functionality, also taking into account the structural rearrangement of the individual monomers..[32–35] Tom et al. have recently employed genetic algorithms along with a fitness function tailored to simultaneously optimize the SF rate and the lattice energy for the inverse design of the tetracene crystal packing, generating structures predicted to have higher SF rates while delivering insight into packing motifs associated with improved SF performance. [36] A promising strategy has been followed in different context by Ciminelli et al. for the optimization of minimum energy Conical Intersection (CI) structures, for which they introduced a penalty function approach, in which a modified potential energy function is minimized. [37] These function contains two terms: the average of the state energies and a penalty function, that increases with the ground state energy. In that way, the optimization procedure targets structures with both minimized potential energy and minimized energy gap between considered states. In this work, we present a simple and computationally efficient functionality optimization strategy to identify molecular arrangements with enhanced SF rate by employing a penalty function approach, for which the objective function includes an expression for the SF rate based on Fermi’s golden rule augmented by an potential ground state energy dependent constraint.

Structure of PBI

Figure 1. Structure of PBI. The carbon atoms along the blue arrows serve as guidance atoms. The mean of the blue and orange arrows define the long(x) and short(y) axis of the planar molecule. Along the atoms marked with the blue arrow the curvature is calculated.

A general challenge in the optimization of chosen functionality descriptors consist in the calculation of their gradients, as the analytic implementation can be tedious. We overcome this barrier by employing automatic differentiation (AD), that delivers gradients up to machine precision by systematically applying the chain rule to traverse the computational graph implemented functions. [38] While AD serves as a standard tool in machine learning, e.g. to evaluate loss functions, the implementation of quantum chemistry in the frame of the AD is an emerging field, that opens possibilities not only for gaining access to energy gradients of arbitrary order, [39, 40] but also in the construction and parametrization of Hamiltonians [41, 42] as well as for inverse molecular design [43]. As a consequence, the implementation of “differentiable quantum chemistry” within the AD framework is a promising direction. [44] We have recently implemented model-CI Hamiltonian in the frame of AD to construct adiabatic wavefunctions in the basis of quasi-diabatic states. [16] The AD directly delivered

non-adiabatic couplings, which we screened as a proxy for the SF rate.

Methods

SF rate expression

We describe the SF rate in the frame of Fermi’s golden rule, which is expressed as follows:

W(SF)=2πℏTRP2ρ(E)(1)W(\mathrm{SF}) = \frac{2\pi}{\hbar} T_{RP}^{2}\rho(E) \tag*{(1)}

Where W(SF)W(\mathrm{SF}) represents the rate of singlet fission, proportional to the square of the SF electronic matrix element TRPT_{RP} and a Frank-Condon weighted density of states ρ(E)\rho(E), that depends on the energetics of the process. As our intention is not to predict total SF rates but to identify structural motifs most likely to be favorable for SF, we introduce several approximations: We consider only pairs of molecules, also it has been shown that polarizability and delocalization introduced by the bulk can have significant influence on the SF process. Furthermore, we focus on the geometry optimization for maximizing the rate of the formation of the singlet-paired triplets as a key intermediate. For the expression of TRPT_{RP}, we employ the diabatic frontier orbital approach, widely applied by the Michl group and others, that considers diabatic states formed for four electrons within the four frontier orbitals hAh_A, lAl_A, hBh_B, lBl_B of the two monomers AA and BB [45, 46, 47]. The orbitals are assumed to be real, normalized and orthogonal to each other. For each monomer, 3 electronic states are considered, these are: the ground state S0IS_0^I, the singlet excited state S1IS_1^I and a triplet state, T1IT_1^I. By removing or adding an electron to a monomer, the radical cation CIC^I or anion AIA^I is formed. From that, a subspace of the seven lowest electronic singlet states of the combined system can be constructed that can be classified as the ground state 1(SA0SB0){}^{1}(S_A^0S_B^0), two local excited (LE) states 1(SA1SB0){}^{1}(S_A^1S_B^0) and 1(SA0SB1){}^{1}(S_A^0S_B^1), two charge transfer states 1(CA+AB−){}^{1}(C_A^+A_B^-) and 1(AA−CB+){}^{1}(A_A^-C_B^+), and two doubly excited states 1(TA1TB1){}^{1}(T_A^1T_B^1). The last state is commonly excluded as the energy of two localized singlet excitations is much higher than that of two localized triplet ground states paired into a singlet. The singlet fission rate is than assumed to be proportional to the squared matrix element of the diabatic Hamiltonian linking the exciton and the biexciton triplet states. The direct coupling between these states is negligible, while the indirect coupling through higher-level charge transfer states that couple with both the singlet exciton and the spin-correlated triplet pair state is significant. The resulting matrix element reads as follows:

TRP=∣⟨LE∣H^∣CA⟩⟨CA∣H^∣TT⟩ΔE(CA)+⟨LE∣H^∣AC⟩⟨AC∣H^∣TT⟩ΔE(AC)∣(2)T_{RP}=\left|\frac{\langle LE|\hat{H}|CA\rangle\langle CA|\hat{H}|TT\rangle}{\Delta E(CA)}+\frac{\langle LE|\hat{H}|AC\rangle\langle AC|\hat{H}|TT\rangle}{\Delta E(AC)}\right| \tag*{(2)}

Here, ΔE(CA)\Delta E(CA) and ΔE(AC)\Delta E(AC) represent the energy differences between states CA and LE and between states AC and LE, respectively.

In the simple frontier orbital model combined with the zero differential overlap approximation, eqn. (2) reduces to

FockTRP=∣(lA∣F^∣lB)(lA∣F^∣hB)ΔE(CA)−(hA∣F^∣hB)(hA∣F∣lB)ΔE(AC)∣(3)FockT_{RP}=\left|\frac{(l_A|\hat{F}|l_B)(l_A|\hat{F}|h_B)}{\Delta E(CA)}-\frac{(h_A|\hat{F}|h_B)(h_A|F|l_B)}{\Delta E(AC)}\right| \tag*{(3)}

where F^\hat{F} is the Fock matrix. The rate expression (TRPT_{RP}) was approximated by transitioning from fock matrix elements to overlap matrix elements, represented as ⟨ϕ∣F^∣ψ⟩=ϵ⟨ϕ∣ψ⟩\langle\phi|\hat{F}|\psi\rangle=\epsilon\langle\phi|\psi\rangle. This approximation aligns with the further simplification of Michl’s model [15]. The value of epsilon is considered constant through the optimisation it can be omitted.

OverlapTRP=∣⟨lA∣lB⟩⟨lA∣hB⟩ΔE(CA)−⟨hA∣hB⟩⟨hA∣lB⟩ΔE(AC)∣(4)OverlapT_{RP}=\left|\frac{\langle l_A|l_B\rangle\langle l_A|h_B\rangle}{\Delta E(CA)}-\frac{\langle h_A|h_B\rangle\langle h_A|l_B\rangle}{\Delta E(AC)}\right| \tag*{(4)}

Also ΔE(CA)\Delta E(CA) and ΔE(AC)\Delta E(AC) are approximated to be the and are set to one. This leads to the final approximation for the singlet fission rate:

TRP=∣⟨lA∣lB⟩⟨lA∣hB⟩−⟨hA∣hB⟩⟨hA∣lB⟩∣(5)T_{RP}=\left|\langle l_A|l_B\rangle\langle l_A|h_B\rangle-\langle h_A|h_B\rangle\langle h_A|l_B\rangle\right| \tag*{(5)}

Computational Details

Function

A simple optimization of the matrix element TRPT_{RP} as derived in Sec. 1.1 with respect to the nuclear coordinates will lead to unphysical structures, specifically to implosion of the molecular system, as the overlap of orbitals without any boundary is favored. However, the introduction of suitable constraints is not straightforward. In this work, we aim to identify dimer structures with enhanced coupling, for which both the intermolecular packing motif as well as structural contortion of the monomer is allowed, while the rearrangement of atoms within molecules or formation of bonds between monomers should be unfavorable. We therefore construct a nuclear coordinate-dependent objective function, that consists of a term for the ground state potential energy E(x)E(x) augmented by the expression of the matrix element TRPT_{RP} for which we take the logarithm to smooth the ”peaked” behaviour with respect to changes in coordinates. This reads in total:

L(x)=E(x)−ωlog⁡(TRP(x)2)(6)L(x)=E(x)-\omega\log(T_{RP}(x)^2) \tag*{(6)}

To balance the optimisation between energy and SF rate the free parameter ω\omega is used. As we consider TRPT_{RP} to be proportional to a rate, which usually has an exponential dependence on some activation energy (k=AeEa/kTk=Ae^{E_a/kT}, log⁡(TRP(x)2)\log(T_{RP}(x)^2)) would be defined as an activation energy, however, its physical interpretation is not obvious.

Optimization Procedure

In the following, we outline the SF rate optimization procedure applied in this work: As a first step, the monomer structure is energetically relaxed employing the AM1[48] implementation in PYSEQМPYSEQМ [49] in conjunction with the python package geomeTRICgeomeTRIC [50]. We then generate an ensemble of initial dimer structures by randomly translating and rotating one monomer in relation to the other, until the lowest atom-atom distance between the monomers falls between 2 Å and 3 Å. For each dimer, we optimize the coordinates to minimize the defined objective function in equation 6 incorporating the implemented approximation for TRPT_{RP} (see equation (5)), and the ground state energy given by the AM1 dimer energy adjusted for dispersion effects[51]. The free parameter ω\omega in equation 6 has been chosen as 0.3 based on the evaluation of test calculations (see ESI for details Figure 1). The AD framework of the implementation allows us to automatically obtain gradients for both the SF rate and energy as well as for L(x)L(x). For the optimization the value of the objective function and its gradient are given to geomeTRICgeomeTRIC. The entire implementation is available on GitHubGitHub[52].

Dimer structure analysis

For the analysis of the optimized dimer structure, we define a set of parameters describing the dimer geometry. These are specifically the relative translations and rotations between monomers as well as the curvature and twisting within the monomers. For that, we define a coordinate frame for each monomer. Therefore a set of guidance atoms is used to generate a vector along the long(x)(x) and the short(y)(y) axis of the original planar molecule (blued and orange in Figure 1). Using the Gram-Schmidt orthogonalization process, the orthogonality of these vectors is ensured and the vector normal to the plane(z)(z) can be calculated.

The relative translations can be calculated by projecting the vector, which connects the center of mass of both monomers on the internal coordinate system of a monomer. Relative rotations are handled in a similar way. As an example, z-rotation can be determined by projection of the x-vector of one molecule on the xy-plane of the other molecule and calculation of the angle between the projected vector and the x-vector. curvature is calculated by projecting the guidance atoms of each longside (blue arrows in Figure 1) onto the x-z plane of the monomer’s coordinate frame and fitting these points to the parabola z(x)=ax2+bx+cz(x) = ax^{2} + bx + c. The curvature, defined as the second derivative of zz, is then 2a2a. Since both sides are fitted independently, the mean curvature of a monomer is given. The twisting angle of a monomer is defined as the angle between the two short axes vectors (orange arrows in Figure 1) of the molecule.

Quantum Chemical calculations

Energies of the ground state, S1S_{1} and T1T_{1} states of monomers were calculated using CASPT2 [53] employing the BAGEL software[54]. The active space comprised four orbitals: HOMO-1, HOMO, LUMO, and LUMO+1, accommodating four electrons. The cc-pVDZ basis set was utilized for these computations.

Results

We expect the ”SF rate landscape” of the PBI dimer to have several local extrema. In order to search for pronounced minima, we generate an ensemble of 500 dimer structures which we individually optimized, according to the described procedure. The resulting optimized structures are sorted after the best SF rate and analyzed in terms of the defined structural parameters. An extensive list of all structures is given in the SI. 14 out of the 500 optimizations did not converge within the chosen limit of 300 steps. Optimiziation convergence was reached in average within 88 steps. In Figure 3 the SF rate optimization results for the ensemble are visualized. The blue broad distribution located around 5.58×10−115.58 \times10^{-11} depicts the SF rates of the initially generated dimer ensemble, while the orange sharp distribution around 1.20×10−71.20 \times10^{-7} gives the optimized values, clearly illustrating the drastically enhanced SF rates of the final structures.

Histogram showing the distribution of SF rates for optimized and start structures

Figure 3. The red/orange histogram shows the distribution of the SF rate for the optimized structures and the blue one shows the distribution of the SF rate for the start structures. The numbers are the structure counts of the bars. The optimisation starts with a broad distribution of SF rates and can improve them to a sharp distribution of higher SF rates.

The structure with a overall highest SF rate of 4.06×10−74.06 \times10^{-7} (illustrated in Figure 2) converged within 51 optimization steps. The packing motif exhibits a z-translation of 3.073.07 Å paired with small x- and y-translations of −0.40-0.40 Å and −0.67-0.67 Å, respectively. Furthermore the monomers show a z-rotation of 53.8∘53.8^\circ with respect to each other, while the x- and y-rotations are small with 6.1∘6.1^\circ and 6.9∘6.9^\circ, respectively. Inspection of the individual monomer structures reveal, that

Top structure of the 500 trajectories with high SF rate

Figure 2. Top structure of the 500 trajectories with high SF rate of 4.06×10−74.06 \times10^{-7}.

both PBI molecules feature significant twisting angles of 24.3∘24.3^\circ (monomer AA) and 15.7∘15.7^\circ (monomer BB). Furthermore, monomer AA exhibit a noticeable curvature of 10.6×10−310.6 \times10^{-3} Å−2^{-2} whereas monomer BB is only very slightly bent of −0.7×10−3-0.7 \times10^{-3} Å−2^{-2}.

To gain a deeper understanding of the structural motifs that feature high SF rates, we conducted a detailed analysis on a subset of structures with TRPT_{RP} greater than 1×10−71 \times10^{-7}. This subset comprises 239 structures, which represent 47.8% of all optimized structures and is indicated by the orange bar in Figure 3. The SF rate ranges from 1.0×10−71.0 \times10^{-7} to 4.1×10−74.1 \times10^{-7} with an average rate of 2.0×10−72.0 \times10^{-7}. Given that the average rotations around the x- and y-axes are relatively low (7.1∘7.1^\circ and 6.2∘6.2^\circ, respectively) as well as the z-axis translation (averaging 2.812.81 Å) we have excluded these parameters for further analysis. Considering the symmetry of the PBI molecule, we adjusted the structural coordinates to ensure that the translations along the x- and y-axes are always positive, while the rotation about the z-axis is modified accordingly. We normalized the selected parameters for structural evaluation and applied a three-dimensional Principal Component Analysis (PCA)[55, 56] to organize and reduce the dimensionality of the data. This PCA approach captured 68 % of the variance (see Table I), indicating that the three PCA dimensions likely encompass the most significant structural features.

dimensioncontribution
133.05 %
221.72 %
313.26 %
Σ\Sigma68.03 %

Table I. Variance Contribution per dimension in PCA.

Within the explored three-dimensional space, we identified four distinct clusters through the application of the k-means algorithm[56, 57], as visualized Figures in Figure 4 and Figure 5 with SF rates visualized using a color map.

Three-dimensional PCA analysis of the seven attributes

Figure 4. Three-dimensional PCA analysis of the seven attributes: x- and y-translation, z-rotation and twist angle and curvature for both monomers. The inner white to red sphere indicates the SF rate for each geometry. The SF rate scale is shown in the color bar on the right. The outer transparent sphere indicates the corresponding cluster affiliation.

Two-dimensional projection onto the most significant PCA dimensions

Figure 5. Two-dimensional projection onto the most significant PCA dimensions of three-dimensional Figure 4. The inner white to red sphere indicates the SF rate for each geometry. The SF rate scale is shown in the color bar on the right. The outer transparent sphere indicates the corresponding cluster affiliation

We further categorized the individual clusters based on their distinct structural patterns. For each cluster, we determined the mean and standard deviation of the attributes in PCA space, subsequently converting these back to the real attribute space. This approach enabled the generation of a representative ”mean”-structure for each cluster, fixing the z-translation at 3.00 Å. These structures are depicted in Figure 6 from both top and side view.

Generated structures as mean of each cluster

Figure 6. Generated structures as mean of each cluster. (a) Cluster 1 PBI – C1 with color green: The dimer features a translation in x of 1.46 Å and in y of 3.30 Å, a z-rotation of −49.3°, curvatures of 3.8×10−33.8 \times10^{-3} Å−2^{-2} and 1.6×10−31.6 \times10^{-3} Å−2^{-2} and twisting angles of 10.0° and 5.4° on monomer AA and BB, respectively. (b) Cluster 2 PBI – C2 with color blue: The dimer features a translation in x of 1.46 Å and in y of 1.37 Å, a z-rotation of −24.1°, curvatures of 3.6×10−33.6 \times10^{-3} Å−2^{-2} and 4.7×10−34.7 \times10^{-3} Å−2^{-2} and twisting angles of 21.6° and 14.5° on monomer AA and BB, respectively. (c) Cluster 3 PBI – C3 with color chartreuse: The dimer features a translation in x of 3.06 Å and in y of 0.72 Å, a z-rotation of −6.2°, curvatures of −18.8×10−3-18.8 \times10^{-3} Å−2^{-2} and 19.3×10−319.3 \times10^{-3} Å−2^{-2} and twisting angles of 7.2° and 8.0° on monomer AA and BB, respectively. (d) Cluster 4 PBI – C4 with color cyan: The dimer features a translation in x of 1.44 Å and in y of 3.79 Å, a z-rotation of 43.0°, curvatures of −1.1×10−3-1.1 \times10^{-3} Å−2^{-2} and 5.5×10−35.5 \times10^{-3} Å−2^{-2} and twisting angles of 8.6° and 7.2° on monomer AA and BB, respectively.

The individual dimer and monomer structures are in the following named PBI-CnA/B, where nn denotes the index of the cluster, while subscript AA represents the

structure of the first monomer and subscript BB represents the structure of the second monomer. The first cluster (”Cluster 1”), represented in green, exhibits the lowest mean SF rate of 1.47×10−71.47 \times10^{-7}. Its significant attributes include a x-translation of (1.46±0.54)(1.46 \pm0.54) Å, a pronounced y-translation of (3.30±0.10)(3.30 \pm0.10) Å and a strong z-rotation of (−49.3±9.7)∘(-49.3 \pm9.7)^\circ. Curvature (3.8×10−3(3.8 \times10^{-3} Å−2^{-2} and 1.6×10−31.6 \times10^{-3} Å−2^{-2} as well as twisting (10.0∘(10.0^\circ and 5.4∘5.4^\circ for monomer AA and BB, respectively) are rather subtle.

The second cluster (”Cluster 2”), depicted in blue, displays the highest mean SF rate of 2.98×10−72.98 \times10^{-7}, distinguished by notable twist angles in both monomers ((21.6±5.3)∘(21.6 \pm5.3)^\circ for monomer AA and (14.5±4.1)∘(14.5 \pm4.1)^\circ for monomer BB). This cluster also features significant x and y-translations (1.46±0.39)(1.46 \pm0.39) Å and (1.37±0.05)(1.37 \pm0.05) Å, respectively). The significance of z-rotation −24.1∘-24.1^\circ is reduced due to the large standard deviation of 36.6∘36.6^\circ. Curvatures of both monomers 3.6×10−33.6 \times10^{-3} Å−2^{-2} and 4.7×10−34.7 \times10^{-3} Å−2^{-2} are again subtle.

The third cluster (”Cluster 3”), given in chartreuse, possesses the second-highest mean SF rate of 2.15×10−72.15 \times10^{-7} and is primarily characterized by pronounced curvatures in both monomers ((−18.8±4.5)×10−3-18.8 \pm4.5) \times10^{-3} Å−2^{-2} for monomer AA and (19.3±2.0)×10−3(19.3 \pm2.0) \times10^{-3} Å−2^{-2} for monomer BB). The fact that the signs of the curvatures are opposite, leads to an arrangement in which the molecules are bend in that way that the vertices of the parabola are closest (see Figure 6 (c)). The cluster features a distinct x-translation of (3.06±0.43)(3.06 \pm0.43) Å and slight y-translation of (0.72±0.16)(0.72 \pm0.16) Å, while again, relevance of the z-rotation is reduced by a large standard deviation of (−6.2±27.9)∘(-6.2 \pm27.9)^\circ. The twisting angles are subtile (7.2° and 8.0°).

The fourth cluster (PBI-C4), illustrated in cyan, has a lower mean SF rate of 1.67×10−71.67 \times10^{-7}. The defining characteristics for this cluster include a noticeable x-translation with (1.44±0.44)(1.44 \pm0.44) Å and a strong y-translation with (3.79±0.16)(3.79 \pm0.16) Å as well as a pronounced z-rotation of (43.0±10.8)∘(43.0 \pm10.8)^\circ. Monomer AA exhibits virtually no curvature (−1.1×10−3-1.1 \times10^{-3} Å−2^{-2}), while Monomer BB displays a slight bend along its length axis (5.5×10−35.5 \times10^{-3} Å−2^{-2}). Twisting is subtle (8.6° and 7.2° on monomer AA and BB, respectively).

Notice, that the motifs found for the green cluster and the cyan cluster appear very similar. This is reflected in the two-dimensional PCA plot in Figure 5, in which the two clusters are nearer to each other than to the others. However, a significant difference consists in the overlapping parts of the individual monomers. In the mean structure of Cluster 1, the outer sections of the aromatic system of both monomers are stacked, while for Cluster 4 we find, that the middle part of one molecule overlaps with the outer aromatic part of the other.

We further calculated adiabatic energies for the S1S_1 and T1T_1 state of the individual monomers in the discussed “mean” structures in order to evaluate, whether they match the energy condition for SF, that goes along with the requirement of pronounced electronic couplings. As an ideal case, the singlet state should be slightly higher in energy than twice the energy of the triplet ground state (2ΔET1−ΔES12\Delta E_{T_1} - \Delta E_{S_1}), so that the process S0+S1→T1+T1S_0 + S_1 \rightarrow T_1 + T_1 is (slightly) exoergic. As the applied optimization routine allows for the individual structural rearrangement of monomers within the dimer system, energies of the molecules within one stack can differ. The obtained results are presented in Table II. For Cluster 1 and 2, we find exceptionally good energies, in the sense that they clearly fulfill the required energy condition. The monomers of cluster 3, that both show significant contortion, exhibit almost the same energy values as the planar PBI. The energy values of the two monomers of the mean structure of Cluster 4 differ significantly. Deeper inspection of these findings indicate, that already slight twisting leads to drastically improved energy values, while the presence of curvature leads to comparable values as for the planar PBI. As Liu et al. [29] noticed, both types of contortions impact the LUMO transition energy, which can be attributed to the overlap of the π\pi-bond within the PBI’s (LUMO), which is amplified by bending, while twisting reduces it.

NameΔET1\Delta E_{T_1} in eVΔES1\Delta E_{S_1} in eV2ΔET1−ΔES12\Delta E_{T_1}-\Delta E_{S_1} in eV
PBI-C1A_A1.10382.2758-0.0682
PBI-C1B_B1.10682.2799-0.0663
PBI-C2A_A1.08712.2539-0.0795
PBI-C2B_B1.09892.2693-0.0714
PBI-C3A_A1.23312.16690.2993
PBI-C3B_B1.23322.16620.3002
PBI-C4A_A1.10482.2773-0.0676
PBI-C4B_B1.22742.16420.2905
PBI-Planar1.40222.48910.3154

Table II. ΔES1\Delta E_{S_1} and ΔET1\Delta E_{T_1} by CASPT2/cc-pVDZ(4,4)

The significance of our obtained results can be evaluated in the context of recent synthetic work on contorted PBIs. Numerous studies report on the synthesis of twisted PBIs as we have found in this work, that can be achieved through the introduction of ligands and heteroatoms at the bay positions.[58, 59, 60, 61] Based on the ligand size, the dihedral angle describing the twist can be modulated. Osswald et al. [62] reported on the synthesis of a series of twisted PBIs with a variety of halogen atoms attached to the bay positions, exhibiting dihedral angles ranging from 0∘0^\circ to 35∘35^\circ. This spectrum clearly captures the dihedral angle values found in our function-optimized structures. Also bending of the PBI molecules along the long axis has been reported in Ref. [29], with the degree of curvature fine-tuned by introduction of aromatic ligands of different sizes to the ortho sides of the molecule, acting as ”strings” to stabilize the bend. However, a review of the therein presented crystal structures shows significantly more pronounced bending than what we observed for the discussed mean structures, with values between 109×10−3 A˚−2109 \times10^{-3}\,\text{Å}^{-2} and 163×10−3 A˚−2163 \times10^{-3}\,\text{Å}^{-2}. Interestingly, Conrad-Burton et al. have experimentally shown that the marked contortion leads to an 2 orders of magnitude increase in the SF rate. They attributed this enhancement to the alignment of excited singlet and triplet state energies in the PBI, determined trough TDDFT calculations.[30] In contrast, the maximum curvature observed in our optimized structures is −23.4×10−3 A˚−2-23.4 \times10^{-3}\,\text{Å}^{-2}. To further explore this variance, we calculated the SF rate for the corresponding crystal structure, finding a relatively low SF rate of 2.95×10−112.95 \times10^{-11}. This suggests, that pronounced bending might be disadvantageous with respect to the electronic coupling strength which we evaluate as a proxy for the SF rate calculation. Finally it should be also mentioned, that the stabilization of two monomer structures bearing different contortion within one aggregate arrangement is synthetically hardly realizable.

Nonetheless, we believe that our findings provide a solid foundation for the development of novel PBI-based SF materials. Additionally, the computational methodology outlined herein offers ample opportunities for further refinement and customization of the functionality optimization procedure.

Conclusions

Within this work, we have introduced a computational strategy to identify structural motifs that exhibit desired properties through a ”functionality optimization” procedure. Utilizing a penalty function approach, we formulated an objective function that integrates the ground state energy as well as the inverse of the expression for the chosen ”functionality” descriptor. This ensures that the molecular conformation remains physically plausible while enhancing the functionality. The gradients for both the energy and the descriptor are automatically delivered by the algorithmic differentiation framework. We applied the outlined strategy to discover PBI dimer structures that increase the SF rate, determined by assessing their diabatic couplings. Through 500 optimizations and subsequent PCA, we identified 4 clusters, each displaying unique structural characteristics. The highest mean SF rate, has been found for Cluster 2, that features a pronounced monomer twist as well as a stacking motif that shows translation both in x- and y- direction as well as rotation around the z-axis. An almost equally high mean SF rate has been found for Cluster 3, that is characterized by pronounced monomer curvature and strong translation along the x axis. The structures of Cluster 1 and Cluster 4 show almost planar structures and share a significant rotation around the z-axis as well as very similar relative translations of the monomers in x- and y-direction. In conclusion, our research lays the groundwork for new PBI-based SF materials as well as establishes a versatile computational framework for the detection of structural motifs with tailored properties.

Author Contributions

J. G.: Writing – Original Draft, Writing – Review & Editing, Conceptualization, Investigation, Methodology, Software, Validation, Visualization A. S.: Investigation, Methodology, Software, Validation M. I. S. R.: Writing – Original Draft, Writing – Review & Editing, Conceptualization, Project Administration, Supervision, Funding Acquisition

CONFLICTS OF INTEREST

There are no conflicts to declare’.

ACKNOWLEDGEMENTS

A. S. and M. I. S. R. are thankful for funding by the Bavarian State Initiative ”Solar Technologies Go Hybrid”. J. G. acknowledges support from Max-Weber-Program scholarship

Functionality Optimization for Singlet Fission Rate Screening in the Full-Dimensional Molecular and Intermolecular Coordinate Space - Supporting Information

Johannes Greiner; Anurag Singh; Merle I. S. Röhr

Analysis of all 500 optimized PBI dimers

rankidtranslationrotationcurvaturetwistrateΔE\Delta Esteps
xyzxyzABAB
1114−0.40−0.673.076.16.953.810.6−0.724.315.74.06×10−74.06 \times10^{-7}0.1951
2222−0.40−0.673.077.77.953.512.5−0.826.014.93.99×10−73.99 \times10^{-7}0.1953
3252−0.480.343.40174.97.0178.7−22.623.59.51.83.95×10−73.95 \times10^{-7}0.3389
422−0.520.293.38174.66.4178.5−20.323.69.02.83.92×10−73.92 \times10^{-7}0.3277
5310.47−0.703.11172.08.0125.68.91.027.015.33.90×10−73.90 \times10^{-7}0.1258
64890.50−0.663.13172.310.6124.810.22.828.114.63.90×10−73.90 \times10^{-7}0.10113
7498−0.43−0.743.077.87.653.69.6−1.128.613.03.88×10−73.88 \times10^{-7}0.1892
8395−0.140.733.131.30.954.69.81.126.315.33.80×10−73.80 \times10^{-7}0.1859
9192−0.080.663.190.65.253.210.03.727.318.03.79×10−73.79 \times10^{-7}0.1661
10224−0.48−0.703.149.311.155.08.92.327.016.33.78×10−73.78 \times10^{-7}0.1565
113160.120.683.18179.94.7126.48.71.528.217.13.75×10−73.75 \times10^{-7}0.1052
123510.50−0.703.14170.711.2124.49.42.726.617.23.73×10−73.73 \times10^{-7}0.0954
134950.51−0.673.14170.411.6125.07.60.929.115.83.71×10−73.71 \times10^{-7}0.0954
1455−0.140.623.200.74.654.68.82.227.516.63.63×10−73.63 \times10^{-7}0.1576
15141−0.490.263.42174.07.7179.5−23.422.67.44.93.54×10−73.54 \times10^{-7}0.2570
163391.040.223.34171.78.7179.5−20.224.28.29.53.48×10−73.48 \times10^{-7}0.2743
rankidtranslation xxtranslation yytranslation zzrotation xxrotation yyrotation zzcurvature AAcurvature BBtwist AAtwist BBrateΔE\Delta Esteps
17256−0.90−0.153.366.86.50.7−19.925.45.08.73.46×10−73.46 \times10^{-7}0.3495
18841.040.193.36172.48.8179.7−20.926.87.19.83.44×10−73.44 \times10^{-7}0.2660
191771.000.103.34173.68.0178.7−19.525.73.911.63.40×10−73.40 \times10^{-7}0.26227
20451−0.42−1.022.84178.72.3121.818.11.420.84.13.34×10−73.34 \times10^{-7}0.30105
215−0.42−1.072.82175.00.4122.615.4−1.622.23.63.33×10−73.33 \times10^{-7}0.3066
22321.87−1.693.005.10.956.13.03.825.69.43.22×10−73.22 \times10^{-7}0.1352
2344−1.80−1.712.937.33.864.3−7.610.20.31.63.22×10−73.22 \times10^{-7}0.2869
24361−1.79−1.622.949.44.465.0−3.86.80.41.63.22×10−73.22 \times10^{-7}0.2865
25207−0.141.102.86175.47.3121.915.40.624.22.23.22×10−73.22 \times10^{-7}0.2761
264801.88−1.693.004.50.555.93.64.625.99.83.20×10−73.20 \times10^{-7}0.1359
27317−1.79−1.672.938.74.365.4−3.37.90.81.73.19×10−73.19 \times10^{-7}0.2863
283021.88−1.692.995.30.956.03.63.725.69.13.18×10−73.18 \times10^{-7}0.1373
294021.87−1.683.004.70.756.13.14.425.59.73.17×10−73.17 \times10^{-7}0.1356
302841.87−1.683.015.10.956.03.24.425.810.03.15×10−73.15 \times10^{-7}0.1362
311382.181.662.805.15.257.21.43.124.99.63.15×10−73.15 \times10^{-7}0.1347
32312−2.121.682.84177.13.7122.54.77.325.29.23.11×10−73.11 \times10^{-7}0.16138
332381.86−1.523.089.56.056.6−2.91.721.413.43.11×10−73.11 \times10^{-7}0.1486
34861.791.702.93172.44.0114.9−6.18.10.31.23.09×10−73.09 \times10^{-7}0.1997
35124−1.78−1.742.905.73.565.4−4.98.00.51.43.04×10−73.04 \times10^{-7}0.2753
363751.86−1.463.117.95.757.1−5.01.820.213.63.02×10−73.02 \times10^{-7}0.1376
37441−1.75−1.582.9814.05.364.5−3.28.64.83.62.96×10−72.96 \times10^{-7}0.2145
382502.201.652.830.01.656.1−1.33.025.914.62.96×10−72.96 \times10^{-7}0.1253
39233−1.83−1.573.10169.96.4124.02.24.025.613.62.93×10−72.93 \times10^{-7}0.1566
40328−1.92−1.662.976.00.454.32.64.429.114.92.89×10−72.89 \times10^{-7}0.3671
41381−1.83−1.583.10168.96.8123.5−0.22.123.513.52.89×10−72.89 \times10^{-7}0.1461
42641.86−1.513.118.96.057.2−5.32.819.914.22.89×10−72.89 \times10^{-7}0.1286
rankidtranslation xtranslation ytranslation zrotation xrotation yrotation zcurvature Acurvature Btwist Atwist BrateΔE\Delta Esteps
431461.91−0.48−3.08166.32.2115.5−1.6−2.69.913.02.86×10−72.86 \times10^{-7}0.4197
442302.161.582.880.30.956.6−1.34.024.113.52.86×10−72.86 \times10^{-7}0.1166
45531.791.692.97169.34.7115.0−5.87.50.64.32.84×10−72.84 \times10^{-7}0.1365
46194−1.76−1.573.0115.65.864.8−0.88.31.44.62.84×10−72.84 \times10^{-7}0.2093
47333−1.92−1.642.977.00.454.13.35.730.012.62.83×10−72.83 \times10^{-7}0.3673
4882−0.67−2.622.728.30.266.5−1.78.55.52.12.83×10−72.83 \times10^{-7}0.2150
49163−3.990.073.19179.22.2179.9−19.118.92.95.42.82×10−72.82 \times10^{-7}0.2942
50404−2.271.662.703.97.054.53.35.330.014.42.81×10−72.81 \times10^{-7}0.3456
5110.762.662.66166.22.2113.5−1.47.87.22.52.79×10−72.79 \times10^{-7}0.1449
52386−3.960.053.21178.72.3179.8−18.117.21.43.82.77×10−72.77 \times10^{-7}0.2974
531544.050.163.150.10.90.8−22.819.04.96.92.75×10−72.75 \times10^{-7}0.2365
543491.90−0.453.09165.02.6115.12.41.611.213.72.74×10−72.74 \times10^{-7}0.4062
55346−4.000.063.19176.02.1179.9−20.819.83.07.22.74×10−72.74 \times10^{-7}0.2865
563144.060.163.125.50.80.6−20.820.27.26.52.72×10−72.72 \times10^{-7}0.2353
57490−0.682.462.892.72.165.9−5.611.02.01.52.68×10−72.68 \times10^{-7}0.1739
583−4.010.073.17179.31.9179.4−19.018.43.95.42.65×10−72.65 \times10^{-7}0.2850
59418−3.980.033.21178.21.9179.4−20.619.22.23.32.65×10−72.65 \times10^{-7}0.2739
60376−2.291.682.712.16.754.72.96.729.913.52.65×10−72.65 \times10^{-7}0.3372
61283−3.96−0.073.23176.73.0179.3−20.519.30.68.42.64×10−72.64 \times10^{-7}0.2754
6283−3.960.063.22177.22.4179.8−18.717.71.63.22.62×10−72.62 \times10^{-7}0.2785
63178−3.97−0.043.23177.52.0179.3−19.918.22.03.12.59×10−72.59 \times10^{-7}0.27104
642144.010.153.180.61.50.1−22.018.53.56.72.59×10−72.59 \times10^{-7}0.2243
65301−2.241.712.722.86.655.93.27.328.512.12.58×10−72.58 \times10^{-7}0.3259
664810.752.622.77173.90.1113.8−2.99.83.71.42.57×10−72.57 \times10^{-7}0.1043
672664.070.143.131.10.00.8−23.318.86.42.62.57×10−72.57 \times10^{-7}0.2186
681182.281.742.68175.310.2124.02.35.129.512.82.56×10−72.56 \times10^{-7}0.2453
rankidtranslationrotationcurvaturetwistrateΔE\Delta Esteps
xyzxyzABAB
693881.95−0.533.11161.45.9115.2−1.84.211.714.72.54×10−72.54 \times10^{-7}0.3466
701112.301.762.67177.79.0125.17.74.830.912.62.53×10−72.53 \times10^{-7}0.2353
71151−2.261.722.712.88.556.32.25.628.311.32.52×10−72.52 \times10^{-7}0.2937
72524.060.003.203.71.31.8−23.022.51.94.52.50×10−72.50 \times10^{-7}0.2165
7329−1.800.633.138.71.964.5−0.81.46.914.22.50×10−72.50 \times10^{-7}0.4056
744461.94−1.692.98174.12.1125.85.06.328.012.62.49×10−72.49 \times10^{-7}0.2492
753431.97−0.483.11160.95.8115.50.01.612.714.92.47×10−72.47 \times10^{-7}0.3352
7694−3.96−0.093.24177.32.9178.9−20.520.11.06.82.46×10−72.46 \times10^{-7}0.25173
77476−1.92−0.533.1115.84.564.80.11.010.615.02.45×10−72.45 \times10^{-7}0.40111
78293−2.301.732.722.89.155.31.58.629.511.72.44×10−72.44 \times10^{-7}0.28105
79714.020.073.183.91.50.4−22.418.64.46.12.43×10−72.43 \times10^{-7}0.2070
80269−1.95−0.433.1216.54.464.42.13.69.713.42.43×10−72.43 \times10^{-7}0.4045
8142−1.91−0.56−3.1113.93.865.00.1−4.19.813.52.41×10−72.41 \times10^{-7}0.4060
82143−3.151.502.93179.00.8118.5−9.68.612.34.82.38×10−72.38 \times10^{-7}0.2997
833642.291.712.74175.29.7124.40.09.328.511.42.38×10−72.38 \times10^{-7}0.2160
84150−3.071.492.97178.40.8117.7−8.36.011.46.62.37×10−72.37 \times10^{-7}0.2991
851002.940.653.1710.50.56.9−17.036.119.07.02.35×10−72.35 \times10^{-7}0.2869
86298−1.92−0.573.1114.74.464.7−1.63.09.914.62.32×10−72.32 \times10^{-7}0.3961
87119−0.140.962.97166.611.3118.87.8−1.511.813.02.31×10−72.31 \times10^{-7}0.1746
88228−2.97−1.483.11174.03.2117.7−7.38.913.45.12.31×10−72.31 \times10^{-7}0.2850
892041.95−0.563.11162.45.6115.4−1.93.510.014.62.30×10−72.30 \times10^{-7}0.3165
9013−0.29−0.962.94172.07.9119.312.20.113.110.12.30×10−72.30 \times10^{-7}0.18114
913292.66−0.633.409.54.56.1−17.834.919.68.12.30×10−72.30 \times10^{-7}0.2846
9210−2.95−1.503.13173.03.6118.1−6.98.814.47.72.30×10−72.30 \times10^{-7}0.2840
931791.97−0.483.13162.65.2115.7−2.93.311.812.72.29×10−72.29 \times10^{-7}0.3168
94152.73−0.653.399.53.34.6−22.835.617.17.82.29×10−72.29 \times10^{-7}0.2891
rankidtranslation xtranslation ytranslation zrotation xrotation yrotation zcurvature Acurvature Btwist Atwist BrateΔE\Delta Esteps
95374−0.28−0.972.95173.37.5118.34.8−2.510.210.52.27×10−72.27 \times10^{-7}0.1846
96221−2.95−1.503.13173.23.4118.4−9.59.212.57.42.27×10−72.27 \times10^{-7}0.2858
97142−3.010.633.0012.76.78.1−15.026.422.01.32.27×10−72.27 \times10^{-7}0.5368
9857−3.121.482.99177.02.4118.5−9.010.611.89.32.25×10−72.25 \times10^{-7}0.2677
9921−1.531.653.01162.89.9114.7−6.012.711.21.62.25×10−72.25 \times10^{-7}0.3058
100672.72−0.563.403.93.65.6−19.932.916.62.62.22×10−72.22 \times10^{-7}0.2768
1014731.601.643.0120.310.966.3−10.44.75.05.22.21×10−72.21 \times10^{-7}0.2745
102265−1.521.623.04162.79.9114.8−5.913.812.12.72.14×10−72.14 \times10^{-7}0.2876
103367−2.920.573.22174.50.2173.8−16.534.519.11.92.13×10−72.13 \times10^{-7}0.2837
104275−4.15−0.013.070.65.30.0−25.518.90.61.72.13×10−72.13 \times10^{-7}0.4959
105168−2.920.543.24175.80.0174.6−18.033.917.72.82.12×10−72.12 \times10^{-7}0.2843
106416−4.130.04−3.070.34.70.325.2−16.70.62.52.06×10−72.06 \times10^{-7}0.4882
1073654.150.043.06179.25.7179.5−25.916.80.60.32.06×10−72.06 \times10^{-7}0.4168
108313−1.894.032.750.27.10.1−14.53.013.61.32.05×10−72.05 \times10^{-7}0.4368
1093261.934.052.67178.76.7178.9−13.42.115.02.12.04×10−72.04 \times10^{-7}0.4099
1102371.75−4.062.79178.93.5179.3−16.32.016.50.52.02×10−72.02 \times10^{-7}0.4154
1118−0.061.013.04163.913.4117.21.51.311.812.82.01×10−72.01 \times10^{-7}0.1382
112188−4.150.043.070.65.50.1−26.018.60.81.81.99×10−71.99 \times10^{-7}0.4772
1132704.150.043.03178.75.6179.5−24.018.44.12.71.98×10−71.98 \times10^{-7}0.4089
114474−4.140.03−3.090.44.80.524.3−18.51.31.21.98×10−71.98 \times10^{-7}0.47104
115345−1.874.042.740.16.70.3−14.02.613.20.71.96×10−71.96 \times10^{-7}0.3259
1162321.634.122.831.60.52.3−4.816.47.58.71.95×10−71.95 \times10^{-7}0.29108
117323−1.604.052.91179.53.3177.9−4.114.16.59.81.95×10−71.95 \times10^{-7}0.2063
11838−4.14−0.013.080.15.50.3−24.718.50.82.81.95×10−71.95 \times10^{-7}0.4655
119213−1.584.102.79179.61.3175.60.611.411.810.31.91×10−71.91 \times10^{-7}0.2072
120492−1.564.192.70176.70.5175.6−2.012.310.26.41.91×10−71.91 \times10^{-7}0.29121
rankidtranslation xxtranslation yytranslation zzrotation xxrotation yyrotation zzcurvature Acurvature Btwist Atwist BrateΔE\Delta Esteps
1212411.564.15−2.782.60.83.62.5−16.17.98.31.90×10−71.90 \times10^{-7}0.2973
122305−1.594.172.73177.90.7175.8−0.213.411.08.91.89×10−71.89 \times10^{-7}0.2886
123488−1.544.172.78177.71.2176.4−1.616.27.310.51.86×10−71.86 \times10^{-7}0.2895
124400−1.674.072.89179.21.4177.4−4.114.011.110.21.83×10−71.83 \times10^{-7}0.1863
1254861.564.162.762.90.24.6−1.415.410.07.51.83×10−71.83 \times10^{-7}0.2840
126417−2.530.943.11166.89.6116.9−0.719.20.610.81.83×10−71.83 \times10^{-7}0.2770
1273633.10−1.552.98177.31.1119.0−11.95.814.76.91.82×10−71.82 \times10^{-7}0.4682
128272−2.55−0.993.08167.69.3118.4−3.016.92.311.41.82×10−71.82 \times10^{-7}0.2883
1294321.584.122.802.13.03.0−1.515.68.89.41.81×10−71.81 \times10^{-7}0.1778
130107−3.251.552.821.73.060.8−12.06.814.46.31.68×10−71.68 \times10^{-7}0.4964
1313382.581.013.1213.711.562.7−4.915.60.312.91.62×10−71.62 \times10^{-7}0.1456
132287−2.47−0.913.20162.811.1118.9−4.213.30.613.11.60×10−71.60 \times10^{-7}0.2447
133390−2.691.082.928.36.060.10.34.75.910.11.58×10−71.58 \times10^{-7}0.3258
134456−2.701.132.919.26.859.2−1.53.84.712.01.57×10−71.57 \times10^{-7}0.3265
1352621.751.803.023.92.364.1−7.418.02.26.11.55×10−71.55 \times10^{-7}0.1550
13633−2.490.943.22161.013.6117.7−6.113.80.412.71.55×10−71.55 \times10^{-7}0.1725
137258−2.470.943.21160.813.8118.1−5.612.64.112.01.53×10−71.53 \times10^{-7}0.1754
138436−2.61−1.133.0213.69.159.3−1.75.25.113.21.53×10−71.53 \times10^{-7}0.3155
139229−2.59−1.113.0815.49.558.2−7.00.24.310.01.51×10−71.51 \times10^{-7}0.3188
140246−0.872.582.67170.08.1127.015.0−1.622.46.21.51×10−71.51 \times10^{-7}0.1346
14189−0.62−2.492.82175.52.5127.716.70.422.36.01.49×10−71.49 \times10^{-7}0.1341
1423922.61−1.033.16157.311.7122.0−8.8−1.40.32.11.48×10−71.48 \times10^{-7}0.1871
143477−3.872.183.38178.72.6111.1−13.216.31.05.51.43×10−71.43 \times10^{-7}0.3450
144358−0.602.922.71165.81.1114.23.217.76.24.91.43×10−71.43 \times10^{-7}0.1965
1453402.62−1.123.21157.210.1125.0−15.0−7.11.03.21.41×10−71.41 \times10^{-7}0.1778
146122−0.662.972.66152.82.9114.76.317.81.06.51.40×10−71.40 \times10^{-7}0.1443
rankidtranslationrotationcurvaturetwistrateΔE\Delta Esteps
xxyyzzxxyyzzAABBAABB
147431−1.97-1.97−5.03-5.032.162.16167.9167.99.59.5119.2119.28.18.11.31.311.911.915.815.81.40×10−71.40 \times10^{-7}0.150.1543
1484210.660.66−2.72-2.72−2.73-2.732.32.33.83.853.853.8−9.5-9.5−1.7-1.717.917.95.05.01.40×10−71.40 \times10^{-7}0.190.19150
149105−0.86-0.862.602.602.682.68169.7169.78.08.0126.9126.915.215.2−2.1-2.120.520.57.47.41.40×10−71.40 \times10^{-7}0.110.1161
150347−2.59-2.590.930.933.153.1519.419.410.810.858.758.7−7.2-7.21.61.64.34.34.14.11.39×10−71.39 \times10^{-7}0.220.2290
1511403.993.99−2.18-2.183.353.355.45.43.83.867.067.0−16.3-16.318.418.40.30.310.510.51.39×10−71.39 \times10^{-7}0.250.2569
152923.923.92−2.27-2.273.293.291.31.31.81.867.967.9−14.2-14.214.014.02.82.85.55.51.39×10−71.39 \times10^{-7}0.250.25101
153350−3.90-3.902.102.103.423.42170.8170.86.56.5111.6111.6−13.8-13.814.414.41.41.411.611.61.39×10−71.39 \times10^{-7}0.300.3055
1542610.230.236.546.541.691.698.48.40.40.486.686.6−8.5-8.5−4.2-4.20.50.50.30.31.38×10−71.38 \times10^{-7}0.230.2358
155201−0.97-0.972.482.482.782.789.19.112.012.053.753.75.75.711.411.424.524.51.31.31.37×10−71.37 \times10^{-7}0.350.35174
156413−2.57-2.571.061.063.163.1619.919.911.311.357.257.2−7.7-7.7−2.9-2.90.70.74.34.31.36×10−71.36 \times10^{-7}0.220.2267
1572470.310.316.526.521.741.7418.218.22.72.785.685.6−10.5-10.5−4.4-4.41.21.23.83.81.36×10−71.36 \times10^{-7}0.220.2299
1581650.310.31−6.51-6.511.781.7850.450.43.63.685.385.3−9.6-9.6−4.0-4.01.91.96.36.31.35×10−71.35 \times10^{-7}0.240.2492
159108−0.30-0.306.476.471.741.74150.1150.11.51.595.595.5−8.1-8.1−3.9-3.93.13.10.70.71.34×10−71.34 \times10^{-7}0.220.2240
16050−2.08-2.085.065.062.052.05169.2169.210.010.0119.4119.46.96.92.42.48.88.813.013.01.33×10−71.33 \times10^{-7}0.120.12100
161482.152.155.075.072.002.0013.613.611.711.760.360.37.77.70.70.712.312.315.615.61.33×10−71.33 \times10^{-7}0.100.10125
162380−1.96-1.96−5.05-5.052.142.14167.0167.09.99.9118.8118.87.57.51.71.711.211.216.816.81.32×10−71.32 \times10^{-7}0.140.1459
1632431.321.325.745.742.052.052.32.312.112.136.736.7−9.7-9.72.12.19.19.12.22.21.32×10−71.32 \times10^{-7}0.200.2098
1643151.451.455.725.722.042.041.41.411.511.534.234.2−8.3-8.32.42.48.98.90.30.31.31×10−71.31 \times10^{-7}0.200.20119
1654831.031.035.735.732.232.234.44.411.611.641.341.3−10.9-10.90.90.98.28.22.32.31.31×10−71.31 \times10^{-7}0.200.2070
166202.082.085.055.052.022.0214.814.813.913.961.761.77.37.30.60.611.111.115.015.01.31×10−71.31 \times10^{-7}0.110.1153
1674571.431.435.665.66−2.03-2.03176.9176.915.015.0144.9144.915.115.13.13.18.18.14.94.91.30×10−71.30 \times10^{-7}0.480.4896
1682162.072.075.065.061.991.9914.614.613.813.862.162.17.47.41.01.011.511.514.214.21.30×10−71.30 \times10^{-7}0.110.1199
1691661.501.505.615.612.162.164.14.112.512.533.833.8−9.8-9.80.20.26.86.80.90.91.29×10−71.29 \times10^{-7}0.150.1579
1703191.511.515.655.652.002.002.12.113.413.433.733.7−7.2-7.2−0.7-0.79.79.73.13.11.29×10−71.29 \times10^{-7}0.150.1553
1713221.591.595.635.631.921.92178.7178.715.115.1147.6147.6−12.7-12.7−1.1-1.110.810.86.16.11.29×10−71.29 \times10^{-7}0.480.48120
1723721.401.405.745.741.891.891.11.110.910.934.634.6−5.9-5.91.71.710.010.02.52.51.28×10−71.28 \times10^{-7}0.200.2096

Table 1. 500 optimized PBI dimer configurations, ordered from highest to lowest SF Rate. id: index of the original trajectory. translation (x,y,z): Relative intermolecular translation in Å. rotation (x,y,z): Relative intermolecular rotation in degree (due to symmetry the angle is always positive and lower than 180°. curvature (A,B): mean curvature measured along two parallel lines along the long axis of PBI in 1×10−31 \times10^{-3} Å−2^{-2}. twist (A,B): dihedral angle between outer carbons along the long axis of PBI in degrees. rate: SF rate calculated with the overlap approximation. ΔE\Delta E: AM1 stabilization energy of the dimer with respect to two non-interacting monomers. steps: Optimisation steps until convergence. Maximal step number has been set to 300.

Continued on next page

rankidtranslationrotationcurvaturetwistrateΔE\Delta Esteps
xxyyzzxxyyzzABAB
173632.195.051.9911.511.559.45.71.211.310.01.26×10−71.26 \times10^{-7}0.0872
174103−0.663.012.65153.12.2114.56.417.52.67.21.26×10−71.26 \times10^{-7}0.1148
175752.145.072.0412.210.860.27.71.910.814.31.25×10−71.25 \times10^{-7}0.0848
176159−1.97−0.593.4925.97.723.51.4−6.811.58.11.25×10−71.25 \times10^{-7}0.2762
1774531.82−0.533.405.52.811.5−23.325.85.221.11.23×10−71.23 \times10^{-7}0.29179
1782731.475.62−2.03176.515.8145.614.91.311.25.21.22×10−71.22 \times10^{-7}0.4685
1793411.585.661.86179.115.4147.7−12.6−1.711.29.31.22×10−71.22 \times10^{-7}0.46103
1804202.205.082.0414.112.059.94.62.512.113.81.21×10−71.21 \times10^{-7}0.0781
181180.965.112.13179.52.6112.513.04.28.210.31.21×10−71.21 \times10^{-7}0.0879
1821761.545.57−2.04176.616.8146.111.82.811.26.11.21×10−71.21 \times10^{-7}0.3769
1831260.885.052.27176.13.8112.912.14.63.110.81.20×10−71.20 \times10^{-7}0.0872
1841481.43−5.662.06179.512.3146.8−13.1−0.99.98.51.20×10−71.20 \times10^{-7}0.47105
1852101.525.69−1.92178.016.1146.211.10.510.45.41.20×10−71.20 \times10^{-7}0.3762
186113−2.040.633.3628.18.924.05.7−11.89.78.31.20×10−71.20 \times10^{-7}0.2969
187387−2.210.593.3726.710.622.93.8−4.114.57.11.20×10−71.20 \times10^{-7}0.2563
1881025.340.863.115.41.75.2−25.016.36.816.61.19×10−71.19 \times10^{-7}0.2490
1893570.935.072.20179.53.4113.111.24.66.49.61.19×10−71.19 \times10^{-7}0.1181
1901230.925.07−2.20179.33.3113.1−11.1−4.46.19.41.19×10−71.19 \times10^{-7}0.1174
191690.925.062.18177.12.6113.311.04.06.18.21.19×10−71.19 \times10^{-7}0.1194
1921720.925.062.21178.93.2113.110.94.65.89.51.18×10−71.18 \times10^{-7}0.1056
1931560.935.072.20180.03.6112.910.54.25.79.91.17×10−71.17 \times10^{-7}0.1082
19435−0.825.042.226.55.267.612.51.31.211.51.17×10−71.17 \times10^{-7}0.12117
1951285.340.833.135.21.35.9−25.817.66.813.51.17×10−71.17 \times10^{-7}0.2398
1964400.915.062.20175.61.1113.511.15.67.48.21.17×10−71.17 \times10^{-7}0.0798
19773−0.885.072.200.54.067.011.94.23.910.11.16×10−71.16 \times10^{-7}0.15126
198132−0.895.082.190.13.967.011.94.34.110.01.16×10−71.16 \times10^{-7}0.1580
rankidtranslation xtranslation ytranslation zrotation xrotation yrotation zcurvature Acurvature Btwist Atwist BrateΔE\Delta Esteps
1992571.144.462.73173.27.9170.3−10.824.213.46.51.16×10−71.16 \times10^{-7}0.3074
200160.975.082.20177.53.5112.411.34.15.611.81.16×10−71.16 \times10^{-7}0.0758
201111.395.681.96177.815.1144.4−13.7−1.210.48.21.16×10−71.16 \times10^{-7}0.45100
202470−0.855.042.300.82.766.511.35.14.810.51.16×10−71.16 \times10^{-7}0.1282
20320.683.43−2.438.61.755.3−17.5−12.95.05.71.15×10−71.15 \times10^{-7}0.0948
2043852.080.903.24179.74.5168.7−29.618.38.523.01.15×10−71.15 \times10^{-7}0.32113
205911.575.721.75178.415.1148.1−9.51.312.05.91.14×10−71.14 \times10^{-7}0.3598
2063421.555.58−2.09175.617.1146.513.6−1.710.06.01.14×10−71.14 \times10^{-7}0.3592
207487−2.99−1.733.0415.02.561.7−15.09.47.93.71.14×10−71.14 \times10^{-7}0.3860
208771.555.701.80179.916.7147.1−8.6−0.610.78.71.14×10−71.14 \times10^{-7}0.3562
2093732.76−0.773.02176.00.4121.2−4.615.711.617.11.14×10−71.14 \times10^{-7}0.1869
210344−0.825.042.260.22.666.112.96.43.28.91.13×10−71.13 \times10^{-7}0.1172
211203−1.525.66−1.911.617.133.412.73.39.77.71.13×10−71.13 \times10^{-7}0.36158
2122531.074.372.88175.210.9172.3−12.027.710.58.81.13×10−71.13 \times10^{-7}0.22112
2134291.645.721.78179.815.6148.6−10.30.313.54.01.13×10−71.13 \times10^{-7}0.3591
2144270.83−5.072.20176.52.6113.911.13.310.44.81.13×10−71.13 \times10^{-7}0.1142
215174−1.044.38−2.903.813.05.59.6−26.212.612.01.12×10−71.12 \times10^{-7}0.2563
216297−0.984.432.864.412.24.8−9.326.111.310.91.12×10−71.12 \times10^{-7}0.2567
217244−1.064.442.844.910.15.7−11.824.813.812.01.12×10−71.12 \times10^{-7}0.32123
2184542.120.863.29179.95.4168.8−34.921.67.222.81.12×10−71.12 \times10^{-7}0.3280
2194241.204.442.73173.55.9166.6−13.424.29.28.21.12×10−71.12 \times10^{-7}0.2969
2204791.074.372.90175.810.6171.7−12.527.29.89.91.11×10−71.11 \times10^{-7}0.2246
2214392.090.883.26179.34.9167.9−33.117.38.324.01.11×10−71.11 \times10^{-7}0.31219
2221911.625.701.86178.316.5147.7−10.80.613.34.11.11×10−71.11 \times10^{-7}0.3488
2239−1.41−3.563.14174.213.5135.7−20.624.315.88.51.11×10−71.11 \times10^{-7}0.31156
224870.693.482.399.12.356.916.014.54.01.21.10×10−71.10 \times10^{-7}0.1247
rankidtranslation xtranslation ytranslation zrotation xrotation yrotation zcurvature Acurvature Btwist Atwist BrateΔE\Delta Esteps
225353−1.535.662.043.115.633.5−14.10.27.73.71.10×10−71.10 \times10^{-7}0.3574
226162−1.044.372.923.112.56.2−12.125.212.512.81.10×10−71.10 \times10^{-7}0.2453
227591.625.691.85179.316.1148.0−11.62.013.25.61.09×10−71.09 \times10^{-7}0.34100
228455−1.104.432.844.79.86.9−10.025.313.410.61.09×10−71.09 \times10^{-7}0.31140
229260.563.552.3615.80.653.722.615.61.64.01.09×10−71.09 \times10^{-7}0.1256
230239−1.014.39−2.865.811.46.912.1−28.09.710.71.07×10−71.07 \times10^{-7}0.2379
231289−1.98−0.783.360.38.911.7−30.119.86.825.91.07×10−71.07 \times10^{-7}0.3580
232444−0.603.502.41166.30.8125.119.616.31.14.71.04×10−71.04 \times10^{-7}0.0530
23393−1.28−3.873.13176.86.5148.6−26.630.112.20.81.04×10−71.04 \times10^{-7}0.35111
23499−1.263.623.10173.215.5132.1−16.623.916.26.91.03×10−71.03 \times10^{-7}0.32300
23562−1.96−0.743.370.88.311.4−28.921.75.824.91.02×10−71.02 \times10^{-7}0.3455
236496−0.543.572.38164.71.3125.120.014.83.36.41.02×10−71.02 \times10^{-7}0.0492
237468−0.58−3.572.34163.80.1125.221.213.63.65.71.01×10−71.01 \times10^{-7}0.0984
238320−1.55−4.952.8612.16.744.44.20.46.64.91.01×10−71.01 \times10^{-7}0.2595
2392781.645.062.64175.15.3133.9−0.23.25.56.31.00×10−71.00 \times10^{-7}0.22135
2403101.655.192.33179.83.9135.03.2−2.93.45.99.93×10−89.93 \times10^{-8}0.2258
241472−1.655.162.521.21.344.0−1.12.94.66.49.91×10−89.91 \times10^{-8}0.2778
24227−1.745.16−2.480.02.946.21.7−1.81.07.19.91×10−89.91 \times10^{-8}0.2772
243200−1.595.17−2.411.14.944.3−2.7−1.31.24.69.89×10−89.89 \times10^{-8}0.2274
244202−1.475.012.8112.08.444.53.4−0.86.03.19.87×10−89.87 \times10^{-8}0.2249
2451952.05−5.081.98175.710.2119.85.45.211.76.89.86×10−89.86 \times10^{-8}0.44121
2461571.575.06−2.71178.32.5136.84.2−6.40.73.59.84×10−89.84 \times10^{-8}0.17126
247423−1.533.502.9410.111.753.7−9.48.69.11.09.83×10−89.83 \times10^{-8}0.2671
24865−1.535.132.553.95.843.94.6−0.50.25.09.75×10−89.75 \times10^{-8}0.2165
249336−1.645.152.462.35.545.11.3−1.92.14.89.72×10−89.72 \times10^{-8}0.2191
250330−0.634.652.63168.02.8175.9−13.726.83.29.99.71×10−89.71 \times10^{-8}0.27116
rankidtranslation xxtranslation yytranslation zzrotation xxrotation yyrotation zzcurvature AAcurvature BBtwist AAtwist BBrateΔE\Delta Esteps
25141−4.56−1.013.2425.816.063.7−2.6−3.913.61.69.60 × 10−810^{-8}0.3857
25256−0.905.632.13175.67.8119.8−4.34.60.411.29.59 × 10−810^{-8}0.1677
253130−4.16−0.673.18163.88.7119.5−11.3−3.43.67.59.55 × 10−810^{-8}0.2773
254494−0.714.582.69172.26.7177.1−8.620.23.310.49.54 × 10−810^{-8}0.26107
2557−1.763.422.832.73.653.10.6−1.83.86.89.51 × 10−810^{-8}0.2463
2562741.755.242.35173.11.7136.0−8.54.88.47.99.48 × 10−810^{-8}0.21139
257186−4.741.002.9921.313.062.5−1.3−3.612.70.59.46 × 10−810^{-8}0.2784
2584060.734.582.6912.21.05.3−15.227.60.39.49.45 × 10−810^{-8}0.3473
259379−0.65−4.602.65168.72.8177.5−10.122.52.012.89.45 × 10−810^{-8}0.3599
2604931.625.202.37177.62.5136.9−1.50.66.84.69.45 × 10−810^{-8}0.16140
26117−3.702.203.3319.98.351.7−15.5−5.78.314.69.43 × 10−810^{-8}0.36102
2623780.763.522.36164.72.9125.516.318.19.25.39.37 × 10−810^{-8}0.2186
263109−4.16−0.673.17164.48.4118.2−11.8−3.96.110.49.33 × 10−810^{-8}0.2639
264215−4.14−0.713.19162.69.0119.0−11.2−3.22.99.29.32 × 10−810^{-8}0.2645
265271−1.743.332.976.33.152.8−1.60.24.98.29.31 × 10−810^{-8}0.1752
26640−2.105.08−1.8612.813.861.3−5.3−4.613.413.19.31 × 10−810^{-8}0.3785
26724−1.615.17−2.410.95.244.5−3.51.32.56.89.27 × 10−810^{-8}0.2062
26874−0.674.62−2.69171.75.5177.79.4−20.52.19.99.26 × 10−810^{-8}0.2560
269434−4.54−0.983.27154.916.5116.51.0−2.013.54.19.26 × 10−810^{-8}0.2588
2703090.833.222.56170.54.7127.112.217.913.04.79.24 × 10−810^{-8}0.2194
271321−4.731.023.0020.612.261.7−1.9−5.311.82.59.12 × 10−810^{-8}0.2649
2723972.085.101.80166.416.2117.45.22.613.913.19.11 × 10−810^{-8}0.4091
27346−3.652.143.3822.710.252.7−11.4−5.511.412.09.11 × 10−810^{-8}0.3441
2743562.145.081.77170.513.6119.03.33.714.48.19.09 × 10−810^{-8}0.3469
275411−1.723.362.955.72.952.5−0.51.24.18.89.08 × 10−810^{-8}0.1798
276149−1.733.342.966.03.052.9−1.1−0.24.48.89.08 × 10−810^{-8}0.1637
rankidtranslation xxtranslation yytranslation zzrotation xxrotation yyrotation zzcurvature Acurvature Btwist Atwist BrateΔE\Delta Esteps
2774620.905.622.053.73.358.32.02.23.68.19.07×10−89.07 \times10^{-8}0.1379
278235−4.71-4.710.993.0423.414.062.3−1.7-1.7−4.9-4.910.80.69.07×10−89.07 \times10^{-8}0.2546
279263−4.17-4.17−0.66-0.663.17165.18.2118.7−11.6-11.6−3.7-3.74.19.59.05×10−89.05 \times10^{-8}0.2555
280167−0.53-0.534.612.66170.73.8177.4−5.6-5.620.66.711.89.05×10−89.05 \times10^{-8}0.3292
281129−2.09-2.095.081.8314.915.161.95.82.715.013.99.03×10−89.03 \times10^{-8}0.3660
282120−4.59-4.59−0.99-0.993.2426.215.962.2−3.2-3.2−4.9-4.99.11.89.03×10−89.03 \times10^{-8}0.3641
28345−2.06-2.065.111.7911.915.462.55.83.013.212.69.01×10−89.01 \times10^{-8}0.4148
284158−4.52-4.52−0.99-0.993.29154.317.1115.9−0.2-0.20.512.46.48.99×10−88.99 \times10^{-8}0.2438
285438−1.62-1.625.19−2.39-2.391.43.644.0−1.5-1.50.24.26.28.98×10−88.98 \times10^{-8}0.1969
286308−1.74-1.74−3.43-3.432.815.55.655.53.2−3.9-3.96.84.28.91×10−88.91 \times10^{-8}0.2393
287219−0.84-0.845.672.04176.56.6119.5−3.8-3.81.40.411.08.91×10−88.91 \times10^{-8}0.1253
288442−4.57-4.57−1.01-1.013.2525.815.962.1−1.7-1.7−3.4-3.410.51.48.86×10−88.86 \times10^{-8}0.2668
289112−4.55-4.55−1.03-1.033.29155.316.9116.8−1.7-1.71.211.46.38.85×10−88.85 \times10^{-8}0.2470
290199−0.94-0.945.671.94177.98.0121.1−1.0-1.0−0.2-0.22.010.18.84×10−88.84 \times10^{-8}0.1474
2914913.702.173.34160.85.8129.1−15.3-15.3−2.5-2.58.512.68.82×10−88.82 \times10^{-8}0.2095
2922125.30−0.93-0.933.14179.53.4175.5−26.5-26.514.811.613.68.79×10−88.79 \times10^{-8}0.5761
293296−4.49-4.49−0.98-0.983.33151.119.9114.91.2−1.4-1.411.09.58.77×10−88.77 \times10^{-8}0.1671
2944150.905.671.994.24.858.5−2.3-2.32.02.69.38.76×10−88.76 \times10^{-8}0.1469
295145−3.76-3.76−2.17-2.173.2520.26.049.8−9.9-9.9−3.1-3.110.312.78.74×10−88.74 \times10^{-8}0.2764
2962590.875.632.064.08.561.1−6.6-6.62.40.210.68.70×10−88.70 \times10^{-8}0.1484
2973993.722.213.32162.37.0128.5−16.9-16.9−4.4-4.48.711.88.69×10−88.69 \times10^{-8}0.2694
298173−0.94-0.945.622.06174.33.6121.8−4.8-4.80.82.44.88.66×10−88.66 \times10^{-8}0.1354
299251−0.70-0.703.70−2.22-2.2218.92.156.0−14.9-14.9−15.5-15.52.46.48.64×10−88.64 \times10^{-8}0.21110
300482−0.87-0.875.661.96177.15.7120.0−3.6-3.6−1.5-1.50.76.58.62×10−88.62 \times10^{-8}0.1346
30170−4.65-4.651.023.11158.215.7117.30.3−0.4-0.49.78.28.53×10−88.53 \times10^{-8}0.1474
3021753.752.203.32162.66.6128.2−17.8-17.8−3.8-3.87.912.38.52×10−88.52 \times10^{-8}0.25106
rankidtranslationrotationcurvaturetwistrateΔE\Delta Esteps
xxyyzzxxyyzzAABBAABB
303220−4.14−0.563.0914.47.261.4−12.8−8.34.72.18.49×10−88.49 \times10^{-8}0.5257
304255−4.51−1.063.36152.819.9116.4−0.40.211.79.88.48×10−88.48 \times10^{-8}0.1554
305448−4.54−1.093.31154.518.6117.1−0.6−0.912.06.98.41×10−88.41 \times10^{-8}0.1567
306600.975.691.821.16.659.3−3.60.22.87.58.32×10−88.32 \times10^{-8}0.11172
307396−1.205.25−2.42179.06.9141.01.40.94.66.08.23×10−88.23 \times10^{-8}0.5376
30836−3.990.753.2327.012.062.4−16.7−14.64.81.28.21×10−88.21 \times10^{-8}0.4252
30951−5.31−0.863.140.43.44.2−24.515.311.012.28.09×10−88.09 \times10^{-8}0.6148
310121−1.215.212.57172.810.3138.9−1.5−1.07.76.57.94×10−87.94 \times10^{-8}0.5258
311337−3.542.473.50159.113.9133.6−12.98.611.65.17.79×10−87.79 \times10^{-8}0.2881
3124610.763.492.43162.83.1125.513.317.96.96.17.65×10−87.65 \times10^{-8}0.1567
3133320.703.502.42161.01.8126.816.819.45.06.07.62×10−87.62 \times10^{-8}0.1565
314810.773.452.46163.23.2126.012.418.67.77.27.60×10−87.60 \times10^{-8}0.1566
315428−4.030.643.1621.39.662.3−13.9−12.22.82.07.57×10−87.57 \times10^{-8}0.4954
31672−3.502.673.48163.213.7136.3−10.314.812.22.77.53×10−87.53 \times10^{-8}0.27105
3173830.825.761.72169.34.9119.8−8.82.40.57.57.10×10−87.10 \times10^{-8}0.4367
31842.583.802.196.43.563.717.08.511.30.56.94×10−86.94 \times10^{-8}0.1034
319260−0.875.741.8112.44.059.2−10.92.50.36.36.91×10−86.91 \times10^{-8}0.4276
3204261.82−0.173.48159.47.9169.7−13.319.616.111.06.91×10−86.91 \times10^{-8}0.2087
3211870.885.751.72172.56.5119.7−8.12.42.46.46.88×10−86.88 \times10^{-8}0.4174
3221710.885.761.72172.46.4119.7−8.12.22.46.36.84×10−86.84 \times10^{-8}0.4173
3231350.945.741.72176.87.4119.6−6.02.95.07.26.81×10−86.81 \times10^{-8}0.3935
324131−0.835.781.6313.54.260.1−7.21.00.54.06.79×10−86.79 \times10^{-8}0.4234
325409−1.45−1.253.39173.98.2178.9−11.219.717.312.66.70×10−86.70 \times10^{-8}0.2557
326211−0.835.831.810.28.862.7−9.47.84.411.26.63×10−86.63 \times10^{-8}0.4173
32795−1.671.353.24174.63.6178.5−10.117.118.811.36.57×10−86.57 \times10^{-8}0.2551
32825−3.922.733.0319.14.167.1−16.19.43.81.96.38×10−86.38 \times10^{-8}0.46239
rankidtranslation xtranslation ytranslation zrotation xrotation yrotation zcurvature Acurvature Btwist Atwist BrateΔE\Delta Esteps
329217−1.420.233.41157.00.8165.95.213.214.822.46.34 × 10−8^{-8}0.1355
330437−1.330.293.44155.41.7164.06.79.113.121.75.83 × 10−8^{-8}0.1057
331133−1.83−5.002.70165.415.1127.0−2.7−1.69.47.95.75 × 10−8^{-8}0.36124
3323042.533.892.02165.32.9115.917.89.111.90.55.51 × 10−8^{-8}0.2961
333801.861.922.84168.07.8177.93.28.225.514.15.35 × 10−8^{-8}0.2765
334966.53−1.112.91179.10.3117.9−14.43.38.75.85.18 × 10−8^{-8}0.58110
33568−2.463.972.0216.22.665.019.411.310.63.75.09 × 10−8^{-8}0.3142
336499−6.42−1.513.01176.72.3115.3−10.17.43.30.95.02 × 10−8^{-8}0.3867
337464−6.401.293.18175.93.5117.5−10.66.87.11.14.99 × 10−8^{-8}0.3458
338311−6.381.513.14174.63.5115.5−11.28.74.11.34.97 × 10−8^{-8}0.3772
339101−6.381.31−3.27175.04.5116.310.4−9.19.21.94.94 × 10−8^{-8}0.37152
340226−6.49−1.352.967.32.061.4−14.13.20.46.44.93 × 10−8^{-8}0.6140
341152−6.45−1.403.01178.62.2115.9−10.57.56.90.54.92 × 10−8^{-8}0.3450
342466−6.331.313.30177.53.9115.7−12.47.710.70.74.89 × 10−8^{-8}0.3753
343407−6.51−1.242.954.21.261.4−13.32.34.37.74.87 × 10−8^{-8}0.6164
3441646.46−1.302.984.30.664.3−11.27.610.92.84.86 × 10−8^{-8}0.3066
345190−6.551.432.897.01.161.0−15.63.20.96.34.85 × 10−8^{-8}0.6153
346277−6.51−1.27−2.98176.81.0116.213.0−7.310.41.24.83 × 10−8^{-8}0.3371
347198−6.291.423.29175.04.4114.3−9.98.89.11.04.80 × 10−8^{-8}0.3697
348450−6.411.343.124.93.062.3−12.53.58.57.24.73 × 10−8^{-8}0.70155
349443−6.371.18−3.235.13.863.313.9−4.510.42.34.59 × 10−8^{-8}0.6067
3502645.342.492.80172.63.1117.1−3.48.06.29.63.93 × 10−8^{-8}0.21218
3513825.322.672.62176.42.9118.11.16.011.611.03.91 × 10−8^{-8}0.21121
352181−5.392.562.703.83.664.0−2.02.77.57.53.87 × 10−8^{-8}0.31152
3532945.282.223.09165.24.8116.2−4.54.13.48.13.86 × 10−8^{-8}0.21157
354497−5.34−2.083.0915.83.362.9−3.74.69.310.03.85 × 10−8^{-8}0.2860
rankidtranslation xxtranslation yytranslation zzrotation xxrotation yyrotation zzcurvature Acurvature Btwist Atwist BrateΔE\Delta Esteps
35507.680.14−2.400.45.61.122.8−2.67.88.83.84×10−83.84 \times10^{-8}0.3793
356394−5.232.103.2717.85.063.2−3.76.18.88.93.83×10−83.83 \times10^{-8}0.2746
3571695.362.562.76178.72.0118.4−3.75.56.19.03.80×10−83.80 \times10^{-8}0.25205
358391−7.65−0.042.54178.55.3179.6−22.72.90.13.63.80×10−83.80 \times10^{-8}0.44184
359369−7.66−0.142.53174.94.7177.9−22.42.81.111.03.80×10−83.80 \times10^{-8}0.4481
36097−7.650.032.54177.25.1179.7−22.72.80.35.43.79×10−83.79 \times10^{-8}0.44236
361306−7.650.032.55176.64.9179.5−22.62.70.76.53.79×10−83.79 \times10^{-8}0.44118
362134−5.25−2.592.908.75.163.2−1.25.59.48.83.78×10−83.78 \times10^{-8}0.26128
363292−5.26−2.582.908.65.263.7−1.36.49.28.23.76×10−83.76 \times10^{-8}0.2696
3643035.402.60−2.62178.82.9115.92.0−1.910.08.33.75×10−83.75 \times10^{-8}0.24300
36579−5.412.662.572.11.963.2−1.23.010.88.43.73×10−83.73 \times10^{-8}0.30136
366485−5.372.442.843.51.062.6−5.45.54.79.83.73×10−83.73 \times10^{-8}0.2668
367419−5.402.55−2.714.22.263.56.3−4.97.410.53.69×10−83.69 \times10^{-8}0.25171
368267−5.352.51−2.760.90.562.64.6−7.57.110.43.66×10−83.66 \times10^{-8}0.25114
369227−5.312.233.099.93.363.8−5.53.34.57.33.66×10−83.66 \times10^{-8}0.3080
370475.322.722.58175.03.9117.84.73.112.811.73.65×10−83.65 \times10^{-8}0.19123
3714657.640.102.585.17.12.6−23.22.71.611.33.60×10−83.60 \times10^{-8}0.35195
37266−5.322.682.77176.64.8120.2−3.713.28.72.83.54×10−83.54 \times10^{-8}0.30150
3733005.312.642.72172.24.0118.7−0.73.68.912.63.53×10−83.53 \times10^{-8}0.18127
374478−5.322.64−2.751.91.560.60.3−8.68.18.83.52×10−83.52 \times10^{-8}0.2488
375170−5.392.642.660.20.661.6−3.47.410.19.33.50×10−83.50 \times10^{-8}0.2482
3763355.322.662.773.35.160.7−3.413.38.32.53.46×10−83.46 \times10^{-8}0.23118
377290−5.252.44−3.06168.87.6116.81.9−13.20.30.83.41×10−83.41 \times10^{-8}0.29203
3781155.302.552.886.36.162.6−3.513.75.60.93.38×10−83.38 \times10^{-8}0.23117
379182−5.302.77−2.74179.05.1120.7−1.2−13.19.00.63.31×10−83.31 \times10^{-8}0.2863
38088−6.643.92−2.204.95.14.46.4−2.49.43.63.30×10−83.30 \times10^{-8}0.65180
rankidtranslationrotationcurvaturetwistrateΔE\Delta Esteps
xyzxyzABAB
381325−5.262.522.97169.88.0116.1−1.113.23.32.53.29×10−83.29 \times10^{-8}0.2855
3822546.603.662.59178.93.5174.0−6.03.71.91.63.27×10−83.27 \times10^{-8}0.6298
38314−5.252.762.77175.86.3119.71.711.58.23.63.24×10−83.24 \times10^{-8}0.2763
384136−6.653.302.982.70.410.2−4.52.94.59.83.21×10−83.21 \times10^{-8}0.6455
385231−6.56−3.792.57174.42.7171.8−4.35.514.09.33.18×10−83.18 \times10^{-8}0.3582
386127−6.68−3.282.904.72.08.9−6.65.98.67.73.17×10−83.17 \times10^{-8}0.6595
38730−6.673.522.650.14.47.3−7.63.30.81.73.17×10−83.17 \times10^{-8}0.6479
388459−6.474.072.31174.12.2178.3−2.15.69.24.03.17×10−83.17 \times10^{-8}0.3685
389280−6.793.82−1.89169.85.1171.06.7−6.417.48.73.15×10−83.15 \times10^{-8}0.36121
3904086.623.87−2.253.72.65.05.6−3.910.95.43.10×10−83.10 \times10^{-8}0.2898
391286−6.663.942.06171.14.1174.5−4.85.712.81.73.10×10−83.10 \times10^{-8}0.3588
3921896.573.992.225.63.53.4−3.54.711.65.23.08×10−83.08 \times10^{-8}0.3184
393285−6.523.472.954.32.85.3−5.20.75.50.23.06×10−83.06 \times10^{-8}0.5497
39461−6.63−3.692.563.20.08.4−7.41.912.48.03.03×10−83.03 \times10^{-8}0.55176
3952056.553.492.87177.14.7175.3−9.4−2.25.51.53.03×10−83.03 \times10^{-8}0.5198
396331−6.514.19−2.06175.13.0179.24.2−5.114.08.33.02×10−83.02 \times10^{-8}0.3152
397307−6.663.892.234.22.76.1−8.02.59.56.53.02×10−83.02 \times10^{-8}0.53295
398242−6.56−3.752.55176.41.5172.9−4.00.813.68.22.98×10−82.98 \times10^{-8}0.36153
39985−6.624.00−2.154.53.54.34.5−2.711.96.42.96×10−82.96 \times10^{-8}0.5359
400249−6.464.162.29177.03.9179.7−3.03.99.76.52.96×10−82.96 \times10^{-8}0.3479
401185−6.504.152.13175.62.9179.2−3.23.812.27.72.96×10−82.96 \times10^{-8}0.3154
40258−6.564.012.15174.94.5177.2−4.11.111.92.32.96×10−82.96 \times10^{-8}0.3137
40398−6.454.33−2.10175.75.1178.11.8−4.613.35.52.96×10−82.96 \times10^{-8}0.34135
40428−6.643.992.09172.43.4174.7−2.25.613.34.62.95×10−82.95 \times10^{-8}0.3451
405236−7.650.502.482.26.83.5−22.20.11.211.02.88×10−82.88 \times10^{-8}0.8482
406318−7.660.612.370.97.32.8−22.3−0.74.713.02.87×10−82.87 \times10^{-8}0.8479
rankidtranslationrotationcurvaturetwistrateΔE\Delta Esteps
xxyyzzxxyyzzAABBAABB
4073545.262.682.836.31.761.0−3.2-3.215.24.27.62.87×10−82.87 \times10^{-8}0.27106
4083525.25−2.63-2.632.944.62.562.5−4.3-4.315.33.16.42.80×10−82.80 \times10^{-8}0.26139
409218−5.31-5.312.892.63169.71.3120.7−0.5-0.513.99.19.72.77×10−82.77 \times10^{-8}0.29107
410209−6.23-6.235.26−1.45-1.45165.91.5131.32.90.013.013.42.77×10−82.77 \times10^{-8}0.37215
411139−6.11-6.115.491.13166.75.1128.5−3.1-3.10.218.312.72.73×10−82.73 \times10^{-8}0.3645
412384−5.27-5.272.882.69170.91.9120.6−0.6-0.615.07.69.72.71×10−82.71 \times10^{-8}0.2876
41378−6.20-6.205.331.34166.12.1130.3−2.6-2.6−0.5-0.515.013.32.68×10−82.68 \times10^{-8}0.3690
414452−5.28-5.282.892.68170.11.7120.4−0.5-0.514.18.59.62.65×10−82.65 \times10^{-8}0.2870
415449−6.30-6.301.722.60174.73.3175.92.013.422.716.92.62×10−82.62 \times10^{-8}0.2063
416370−6.08-6.08−1.73-1.733.16176.66.3176.6−0.5-0.512.121.519.22.52×10−82.52 \times10^{-8}0.1655
4174356.411.902.29171.43.7177.6−6.0-6.08.625.619.52.35×10−82.35 \times10^{-8}0.4162
418401−6.39-6.39−1.14-1.143.1316.45.363.7−3.9-3.913.67.811.02.32×10−82.32 \times10^{-8}0.3984
4191557.540.062.58173.88.4174.4−18.5-18.5−2.8-2.81.50.82.30×10−82.30 \times10^{-8}0.7175
420126.371.80−2.38-2.38170.72.8176.03.7−10.2-10.224.817.82.29×10−82.29 \times10^{-8}0.40160
421463−6.31-6.311.782.487.91.44.2−2.4-2.411.424.019.62.28×10−82.28 \times10^{-8}0.4580
422184−6.26-6.261.103.2118.85.867.7−1.4-1.412.71.812.72.27×10−82.27 \times10^{-8}0.3151
423206−3.87-3.875.671.902.93.158.80.8−2.9-2.911.63.92.26×10−82.26 \times10^{-8}0.2398
42437−4.02-4.025.631.921.35.562.51.6−0.8-0.89.93.62.25×10−82.25 \times10^{-8}0.2765
425161−6.24-6.241.00−3.32-3.3222.96.565.81.5−9.7-9.77.212.32.24×10−82.24 \times10^{-8}0.31117
426193−3.65-3.655.652.042.63.855.82.5−3.3-3.37.10.22.23×10−82.23 \times10^{-8}0.2282
427279−6.25-6.250.993.3019.86.667.6−3.7-3.711.77.211.62.23×10−82.23 \times10^{-8}0.3689
428475−7.39-7.39−0.24-0.242.7612.19.510.4−15.7-15.7−0.3-0.35.99.92.21×10−82.21 \times10^{-8}0.76111
429291−3.93-3.935.682.004.16.560.52.91.010.18.82.20×10−82.20 \times10^{-8}0.2686
430248−6.36-6.361.83−2.37-2.3710.13.53.74.3−9.5-9.525.016.22.19×10−82.19 \times10^{-8}0.4440
431368−6.38-6.381.91−2.35-2.358.82.92.82.9−9.8-9.826.719.12.18×10−82.18 \times10^{-8}0.44141
432414−7.50-7.500.002.637.38.57.4−16.8-16.8−2.5-2.51.61.92.17×10−82.17 \times10^{-8}0.7472
rankidtranslation xxtranslation yytranslation zzrotation xxrotation yyrotation zzcurvature AAcurvature BBtwist AAtwist BBrateΔE\Delta Esteps
433348−7.510.042.5413.89.28.5−16.9−2.93.65.92.16×10−82.16 \times10^{-8}0.7345
43449−7.40−0.152.6914.610.610.8−15.1−1.62.39.32.16×10−82.16 \times10^{-8}0.7591
435223−7.530.082.569.59.07.0−17.4−3.01.60.32.15×10−82.15 \times10^{-8}0.73105
436334−6.241.243.36155.410.6117.5−1.913.05.16.62.14×10−82.14 \times10^{-8}0.2572
437110−3.775.442.544.85.259.11.9−3.61.20.32.14×10−82.14 \times10^{-8}0.2583
438460−6.26−1.533.28157.912.1119.3−2.411.61.61.22.10×10−82.10 \times10^{-8}0.2685
4393276.391.85−2.36170.47.9175.83.4−9.824.919.12.06×10−82.06 \times10^{-8}0.4285
440362−7.45−0.05−2.668.98.78.916.23.11.51.12.05×10−82.05 \times10^{-8}0.7272
441240−3.765.642.150.13.457.62.4−2.37.65.92.02×10−82.02 \times10^{-8}0.1944
44223−6.321.663.04162.411.6116.71.910.88.98.92.00×10−82.00 \times10^{-8}0.31127
443405−6.12−1.92−3.009.02.22.93.5−7.625.719.21.98×10−81.98 \times10^{-8}0.4795
444116−4.244.25−3.4225.114.956.2−2.1−6.49.86.21.96×10−81.96 \times10^{-8}0.34300
445208−6.33−4.901.6732.22.335.8−8.31.315.813.11.93×10−81.93 \times10^{-8}0.63285
446467−6.351.912.399.67.83.9−2.07.123.620.31.93×10−81.93 \times10^{-8}0.4439
447469−6.351.982.3210.58.83.4−3.65.224.120.71.85×10−81.85 \times10^{-8}0.4381
448389−6.321.912.3910.88.14.5−2.26.324.020.41.77×10−81.77 \times10^{-8}0.4187
449360−6.504.561.8027.42.933.0−10.62.811.511.01.59×10−81.59 \times10^{-8}0.56300
450106−5.35−3.923.09149.69.1106.41.110.811.49.01.58×10−81.58 \times10^{-8}0.52300
451447−7.831.862.6733.411.964.63.44.811.09.04.47×10−94.47 \times10^{-9}0.7173
452125−7.73−1.793.0333.813.565.23.84.37.512.14.41×10−94.41 \times10^{-9}0.7069
453117−7.861.672.76153.610.2117.94.24.87.910.34.34×10−94.34 \times10^{-9}0.4049
454422−7.831.503.00151.812.9118.04.85.71.814.24.32×10−94.32 \times10^{-9}0.4095
455147−7.80−1.512.99155.610.8117.34.54.31.27.84.30×10−94.30 \times10^{-9}0.40131
4561977.84−1.322.9416.88.962.65.64.16.75.94.26×10−94.26 \times10^{-9}0.35111
457412−7.82−1.493.0127.612.865.31.77.74.910.84.13×10−94.13 \times10^{-9}0.69230
458288−7.84−1.452.98158.211.1117.34.03.00.78.34.13×10−94.13 \times10^{-9}0.37285
rankidtranslation xtranslation ytranslation zrotation xrotation yrotation zcurvature Acurvature Btwist Atwist BrateΔE\Delta Esteps
459410−7.75−1.48−3.1029.014.965.3−5.3−5.33.510.84.12 × 10−910^{-9}0.59246
460160−8.92−0.032.69176.30.5180.0−1.21.52.95.43.08 × 10−910^{-9}0.4271
461234−8.970.202.53179.51.3179.00.22.45.55.83.07 × 10−910^{-9}0.4249
462180−9.120.29−2.145.67.51.21.9−2.812.39.13.01 × 10−910^{-9}0.82300
463196−9.040.222.362.33.50.2−0.31.65.78.83.00 × 10−910^{-9}0.82245
464144−8.970.20−2.43179.33.1178.8−0.83.25.14.62.98 × 10−910^{-9}0.41139
465268−9.060.15−2.321.36.52.12.5−0.72.08.12.96 × 10−910^{-9}0.82187
466137−9.070.012.364.44.82.2−1.31.82.04.12.89 × 10−910^{-9}0.8177
467484−8.86−0.272.89171.71.5179.7−0.41.49.95.42.88 × 10−910^{-9}0.4088
468445−8.98−0.052.61178.82.2179.0−1.01.80.33.92.86 × 10−910^{-9}0.4087
469324−9.10−0.022.336.15.94.1−2.44.67.86.72.70 × 10−910^{-9}0.79118
4703930.763.213.8217.817.915.0−2.6−2.50.91.09.87 × 10−1010^{-10}0.55300
471398−1.186.012.2943.076.78.56.417.45.43.72.03 × 10−1010^{-10}0.82300
472355−8.243.21−3.3559.028.366.9−5.24.48.513.61.88 × 10−1010^{-10}0.74300
47319−9.671.913.323.78.526.67.4−3.35.56.98.50 × 10−1110^{-11}0.71122
474153−9.621.773.34178.58.5152.07.6−3.53.33.88.29 × 10−1110^{-11}0.5151
475299−9.74−1.553.05177.35.3152.27.1−3.312.511.58.29 × 10−1110^{-11}0.5383
47676−9.81−1.872.86173.93.8154.17.6−4.94.811.68.22 × 10−1110^{-11}0.52160
4772811.187.834.86102.479.897.422.9−4.05.02.58.04 × 10−1110^{-11}0.62139
47834−9.115.281.67149.28.2132.28.2−4.08.510.17.72 × 10−1110^{-11}0.55225
47943−9.264.961.99152.47.1134.67.2−4.51.08.57.65 × 10−1110^{-11}0.55162
480377−9.304.90−2.09149.28.8135.9−6.75.51.02.47.24 × 10−1110^{-11}0.54213
481359−9.711.813.30177.09.1151.74.3−3.51.81.47.05 × 10−1110^{-11}0.46113
4826−8.625.003.8731.120.151.28.5−1.816.28.66.87 × 10−1110^{-11}0.74150
483458−5.748.43−2.199.05.437.6−5.53.310.815.13.69 × 10−1110^{-11}0.5944
484433−9.282.451.17142.621.4121.77.71.86.711.93.50 × 10−1110^{-11}0.6175
rankidtranslationrotationcurvaturetwistrateΔE\Delta Esteps
xyzxyzABAB
485425−9.27-9.272.742.74−0.95-0.95145.1145.119.419.4125.9125.9−5.9-5.9−1.9-1.90.50.55.65.63.41×10−113.41 \times10^{-11}0.600.60207
486366−5.39-5.398.338.332.862.86161.9161.93.93.9142.2142.29.29.2−4.7-4.74.14.12.92.93.32×10−113.32 \times10^{-11}0.430.4392
487282−5.31-5.318.418.412.752.75157.0157.09.59.5139.0139.07.07.0−3.8-3.86.66.63.43.43.31×10−113.31 \times10^{-11}0.430.4382
48854−7.18-7.187.387.381.731.73155.1155.17.37.3116.1116.114.314.3−2.6-2.69.19.12.62.63.17×10−113.17 \times10^{-11}0.630.63225
48939−5.39-5.398.448.442.502.50163.3163.36.06.0140.0140.07.07.0−3.1-3.19.89.810.810.83.16×10−113.16 \times10^{-11}0.410.4156
490430−5.67-5.678.128.123.093.09162.3162.31.21.2145.2145.24.94.9−3.3-3.35.65.64.24.23.16×10−113.16 \times10^{-11}0.410.41158
491104−5.51-5.518.478.472.382.3817.417.47.67.639.639.63.83.8−4.0-4.010.210.210.710.73.13×10−113.13 \times10^{-11}0.570.5756
492295−5.51-5.518.458.452.342.34161.0161.08.88.8140.3140.32.62.6−3.0-3.012.012.07.97.93.03×10−113.03 \times10^{-11}0.400.4079
493471−10.04-10.040.300.305.075.07154.2154.233.233.2139.4139.47.87.8−4.5-4.51.31.35.65.61.46×10−111.46 \times10^{-11}0.800.80107
494403−5.86-5.866.366.361.921.9225.225.220.620.642.642.69.09.00.60.65.85.816.016.01.21×10−111.21 \times10^{-11}0.740.74300
495225−5.66-5.667.817.811.781.78115.8115.838.838.898.498.48.28.2−2.6-2.64.94.98.38.31.18×10−111.18 \times10^{-11}0.830.83224
496371−5.41-5.417.907.901.691.6961.461.438.338.379.279.26.86.8−1.4-1.46.76.79.59.51.14×10−111.14 \times10^{-11}0.650.65142
497276−6.70-6.70−0.83-0.833.863.8637.637.634.234.211.811.82.62.6−1.3-1.31.11.18.58.54.60×10−124.60 \times10^{-12}1.021.02300
498902.462.469.649.642.952.95154.1154.13.43.4141.4141.41.01.0−8.3-8.37.27.21.21.21.00×10−131.00 \times10^{-13}1.001.00300
499183−2.75-2.755.235.233.663.6690.790.752.552.513.713.7−0.7-0.7−0.1-0.10.40.40.20.24.88×10−154.88 \times10^{-15}−0.14-0.14300
5002451.361.3610.9110.91−1.16-1.1650.050.027.527.572.972.9−0.8-0.8−0.3-0.30.60.60.20.22.69×10−172.69 \times10^{-17}0.100.10300

Table 1 (continued). from previous page

Determining an Optimal Value for ω\omega

In our optimization process, the objective function L(x)=E(x)−ωlog⁡(TRP⁡(x)2)L(x) = E(x) - \omega\log(\operatorname{TRP}(x)^2), where E(x)E(x) represents the energy and ∣TRP⁡(x)∣2\lvert\operatorname{TRP}(x)\rvert^2 signifies the SF rate for the nuclear configuration x, necessitates the careful selection of the free parameter ω\omega to achieve a balance between optimizing energy and SF rate. To this end, we adjust ω\omega in increments of 0.1, ranging from 0.1 to 10. This adjustment is performed across the same set of 20 randomly translated and rotated ethene dimers, each optimized to minimize L(x)L(x).

The average and standard deviation of the convergent steps, along with the SF rate (SFR) and energy for each ω\omega, are depicted in Figure 1. The graphical representation in Figure 1b illustrates a strong positive correlation between SF rate and energy, complicating the choice of an ideal ω\omega. Since our goal is to minimize energy while maximizing SF rate, we select the value of ω\omega that maximizes the hysteresis between these two parameters as optimal. Consequently, ω=0.3\omega= 0.3 has been identified as the preferred choice. Moreover, values exceeding this threshold tend to exhibit greater instability in terms of convergence steps and energy, as indicated by their standard deviations.

Optimization procedure with varying omega

Figure 1 (PDF p. 31). Optimization procedure with varying ω\omega from 0.1 to 1.0 in steps of 0.1 for the same 20 random ethene dimers each.

References

  1. [38]M. Bartholomew-Biggs, S. Brown, B. Christianson, and L. Dixon, Journal of Computational and Applied Mathematics 124, 171 (2000).
  2. [39]F. Pavošević and S. Hammes-Schiffer, Automatic differentiation for coupled cluster methods (2020), arXiv:2011.11690 [physics].arxiv.org/abs/2011.11690
  3. [40]A. S. Abbott, B. Z. Abbott, J. M. Turney, and H. F. I. Schaefer, J. Phys. Chem. Lett. 12, 3232 (2021), publisher: American Chemical Society.
  4. [41]T. Tamayo-Mendoza, C. Kreisbeck, R. Lindh, and A. Aspuru-Guzik, ACS Cent. Sci. 4, 559 (2018), publisher: American Chemical Society.
  5. [42]K. Inui and Y. Motome, Commun Phys 6, 37 (2023).
  6. [43]R. A. Vargas-Hernández, K. Jornér, R. Pollice, and A. Aspuru–Guzik, The Journal of Chemical Physics 158, 104801 (2023).
  7. [44]X. Zhang and G. K.-L. Chan, The Journal of Chemical Physics 157, 204801 (2022).
  8. [45]J. Michl, A. J. Nozik, X. Chen, J. C. Johnson, G. Rana, A. Akdag, and A. F. Scherer, in SPIE Proceedings, edited by Z. H. Kafafi and P. A. Lane (SPIE, 2007).
  9. [46]M. B. Smith and J. Michl, Chemical reviews 110, 6891 (2010).
  10. [47]E. A. Buchanan, Z. Havlas, and J. Michl, in Advances in Quantum Chemistry (Elsevier, 2017) pp. 175–227.
  11. [48]M. J. S. Dewar, E. G. Zoebisch, E. F. Healy, and J. J. P. Stewart, Journal of the American Chemical Society 107, 3902–3909 (1985).
  12. [49]G. Zhou, N. Lubbers, K. Barros, S. Tretiak, and B. Nebgen, Proceedings of the National Academy of Sciences 119, e2120333119 (2022).
  13. [50]L.-P. Wang and C. Song, The Journal of Chemical Physics 144, 214108 (2016).
  14. [51]S. Grimme, Journal of computational chemistry 27, 1787 (2006).
  15. [52]J. Greiner, GitHub - greinerjo-watt/singlet-fission-optimization, https://github.com/greinerjo-watt/singlet-fission-optimization (2024), [Accessed 22-03-2024].
  16. [53]K. Anderson, P.-Å. Malmqvist, and B. O. Roos, The Journal of Chemical Physics 96, 1218 (1992).
  17. [54]T. Shiozaki, WIREs Computational Molecular Science 8, e1331 (2018).
  18. [55]T. Minka, in Advances in Neural Information Processing Systems, Vol. 13, edited by T. Leen, T. Dietterich, and V. Tresp (MIT Press, 2000).
  19. [56]F. Pedregosa, G. Varoquaux, A. Gramfort, V. Michel, B. Thirion, O. Grisel, M. Blondel, P. Prettenhofer, R. Weiss, V. Dubourg, J. Vanderplas, A. Passos, D. Cournapeau, M. Brucher, M. Perrot, and E. Duchesnay, Journal of Machine Learning Research 12, 2825 (2011).
  20. [57]S. Lloyd, IEEE transactions on information theory 28, 129 (1982).
  21. [58]M. Queste, C. Cadiou, B. Pagoaga, L. Giraudet, and N. Hoffmann, New J. Chem. 34, 2537 (2010).
  22. [59]W. Qiu, S. Chen, X. Sun, Y. Liu, and D. Zhu, Org. Lett. 8, 867 (2006).
  23. [60]A. Nowak-Król and F. Würthner, Org. Chem. Front. 6, 1272 (2019).
  24. [61]A. Nowak-Król, M. I. S. Röhr, D. Schmidt, and F. Würthner, Angew Chem Int Ed 56, 11774 (2017).
  25. [62]P. Osswald and F. Würthner, J. Am. Chem. Soc. 129, 14319 (2007).

Paper details

vlnc:W_m779v323
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents