RELATED APPLICATIONS)

[0001] This application claims the benefit of U.S. Provisional Application No.

62/450,372, filed on January 25, 2017. The entire teachings of the above application(s) are incorporated herein by reference.

STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH

[0002] This invention was made with U.S. government support under DE-SC0015959 awarded by the U.S Department of Energy. The U.S. government has certain rights in the invention.

BACKGROUND OF THE INVENTION

[0003] The combustion of fossil fuels releases vast amounts of CO2 into the atmosphere, which has been shown to contribute to global increases in temperatures. Solar energy represents a possible solution to the global dependence on fossil fuels for energy. Today's commercial photovoltaic devices are based on metal or silicon, because those materials offer relatively high solar energy conversion efficiencies. While these efficiencies are desirable, they carry with them high costs, short lifetimes, and brittle devices. Organic photovoltaics (OPV) are comprised of small-molecule or polymer carbon-based materials. Devices constructed from these materials have substantial advantages over their inorganic

counterparts in flexibility, lifetime, and cost. However, they are generally less efficient than metal or silicon devices.

[0004] OPVs include an electron donor and electron acceptor. Both of these materials typically feature extended π-conjugation with opposite electronic properties. Electron acceptor materials have lowest unoccupied molecular orbitals (LUMOs) that are relatively low-lying. Electron donor materials highest occupied molecular orbitals (HOMOs) that are relatively high-lying. Over the last twenty years, fullerene (C60) derivatives have been used in OPV devices because of their LUMO energies and charge transport properties. However, fullerenes have drawbacks including that they have extremely limited absorbance in the solar spectrum and are expensive to manufacture. SUMMARY OF THE INVENTION

[0005] Thus, a need exists for OPVs that can provide higher conversion efficiencies without the difficulties associated with fullerenes. The compounds described herein are capable of providing high conversion efficiencies in a solar panel.

[0006] Accordingly, in one aspect, the present invention is a molecule represented by structural formula (I):

T^-C-S^T ² (I)

[0007] In a second aspect, the present invention is a solar panel or photovoltaic device comprising at least one of the molecules of the first aspect.

[0008] The OPVs that employ the compounds disclosed herein possess a number of advantages over the existing OPVs, such as facile synthesis that permits energy-level tuning. Furthermore, the OPVs of the present invention show improved absorbance in the visible range without the high price tag of [6,6]-Phenyl-C7i-butyric acid methyl ester, the most common fullerene OPV component.

BRIEF DESCRIPTION OF THE DRAWINGS

[0009] The foregoing will be apparent from the following more particular description of example embodiments of the invention, as illustrated in the accompanying drawings in which like reference characters refer to the same parts throughout the different views. The drawings are not necessarily to scale, emphasis instead being placed upon illustrating embodiments of the present invention.

[0010] FIG. 1 shows a bar graph of Z-scores of the core fragments pertaining to the top OPV candidates, having PCEs greater than or equal to 8.0%. The overrepresented (positive) Z-scores are shown on the right and the underrepresented (negative) Z-scores are shown on the left.

[0011] FIG. 2 shows a bar graph of Z-scores of the spacer fragments pertaining to the top OPV candidates, having PCEs greater than or equal to 8.0%. The overrepresented (positive) Z-scores are shown on the right and the underrepresented (negative) Z-scores are shown on the left.

[0012] FIG. 3 shows a bar graph of Z-scores of the terminal fragments pertaining to the top OPV candidates, having PCEs greater than or equal to 8.0%. The overrepresented (positive) Z-scores are shown on the right and the underrepresented (negative) Z-scores are shown on the left. DETAILED DESCRIPTION OF THE INVENTION

[0013] A description of example embodiments of the invention follows.

[0014] It has now been discovered that a group of molecular fragments— core (referred herein as C), spacers (or linkers) (referred herein as S ¹ or S ² ), and terminal fragments (referred herein as T ¹ or T ² )— can be linked together to form molecules suitable for use in OPVs (such molecules are also referred to herein as "OPVs") of the general structural formula (I):

T^S^C-S^T ² (I).

As seen from formula (I), the OPVs of the present invention include a divalent core fragment linked to divalent spacer fragments, capped with a monovalent terminal fragment.

[0015] A schematic diagram of a computer-implemented method of generating the OPVs of the present invention is exemplified in Scheme 1 :

Scheme 1

Glossary

[0016] As used herein, percent conversion efficiencies (PCEs) refers to a value that quantifies the portion of solar energy that is converted into electricity by a photovoltaic device. The method of computing PCE values employed to evaluate the compounds of the present invention is described hereinbelow.

[0017] The term "alkyl," as used herein, refers to a saturated aliphatic branched or straight-chain monovalent hydrocarbon radical having the specified number of carbon atoms. Thus, "Ci-C ₆ alkyl" means a radical having from 1-6 carbon atoms in a linear or branched arrangement. Examples of "C1-G5 alkyl" include, ^-propyl, / ^' -propyl, «-butyl, / ^' -butyl, sec- butyl, t-butyl, «-pentyl, «-hexyl, 2-methylbutyl, 2-methylpentyl, 2-ethylbutyl, 3- methylpentyl, and 4-methylpentyl. An alkyl can be optionally substituted with

halogen, -OH, Ci-Ce alkyl, Ci-Ce alkoxy, -NO2, -CN, and -N^XR ² ) wherein R ¹ and R ² are each independently selected from -H and C1-C3 alkyl.

[0018] The term "alkenyl," as used herein, refers to a straight-chain or branched alkyl group having one or more carbon-carbon double bonds. Thus, "C2-C6 alkenyl" means a radical having 2-6 carbon atoms in a linear or branched arrangement having one or more double bonds. Examples of "C2-G5 alkenyl" include ethenyl, propenyl, butenyl, pentenyl, hexenyl, butadienyl, pentadienyl, and hexadienyl. An alkenyl can be optionally substituted with the substituents listed above with respect to alkyl.

[0019] The term "alkynyl," as used herein, refers to a straight-chain or branched alkyl group having one or more carbon-carbon triple bonds. Thus, "C2-C6 alkynyl" means a radical having 2-6 carbon atoms in a linear or branched arrangement having one or more triple bonds. Examples of C2-C6 "alkynyl" include ethynyl, propynyl, butynyl, pentynyl, and hexynyl. An alkynyl can be optionally substituted with the substituents listed above with respect to alkyl.

[0020] The term "cycloalkyl," as used herein, refers to a saturated monocyclic or fused poly cyclic ring system containing from 3-12 carbon ring atoms. Saturated monocyclic cycloalkyl rings include, for example, cyclopropyl, cyclobutyl, cyclopentyl, cyclohexyl, and cyclooctyl. Saturated bicyclic and poly cyclic cycloalkyl rings include, for example, norbornane, [2.2.2]bicyclooctane, decahydronaphthalene and adamantane. A cycloalkyl can be optionally substituted with the substituents listed above with respect to alkyl.

[0021] The term "amino," as used herein, means an "- H2," an " HRp," or an " RpRq," group, wherein R _p and Rq can be alkyl, alkenyl, alkynyl, alkoxy, cycloalkyl, aryl, and heteroaryl. Amino may be primary ( H2), secondary (NHR _P ) or tertiary (NRpRq).

[0022] The term "alkylamino," as used herein, refers to an "NHRp," or an "NRpRq" group, wherein R _p and R _q can be alkyl, alkenyl, alkynyl, alkoxy, or cycloalkyl. The term "dialkylamino," as used herein, refers to an "NRpRq" group, wherein R _p and Rq can be alkyl, alkenyl, alkynyl, alkoxy, or cycloalkyl.

[0023] The term "alkoxy", as used herein, refers to an "alkyl-O-" group, wherein alkyl is defined above. Examples of alkoxy group include methoxy or ethoxy groups. The "alkyl" portion of alkoxy can be optionally substituted as described above with respect to alkyl.

[0024] The term "aryl," as used herein, refers to an aromatic monocyclic or polycyclic ring system consisting of carbon atoms. Thus, "C ₆ -Ci8 aryl" is a monocylic or polycyclic ring system containing from 6 to 18 carbon atoms. Examples of aryl groups include phenyl, indenyl, naphthyl, azulenyl, heptalenyl, biphenyl, indacenyl, acenaphthylenyl, fluorenyl, phenalenyl, phenanthrenyl, anthracenyl, cyclopentacyclooctenyl or benzocyclooctenyl. An aryl can be optionally substituted with halogen, -OH, Ci-C ₆ alkyl, C2-C6 alkenyl, C2-C6 alkynyl, Ci-C ₆ haloalkyl, Ci-C ₆ alkoxy, C ₆ -Ci8 aryl, C ₆ -Ci8 haloaryl, (5-20 atom) heteroaryl, -C(0)Ci-C ₃ haloalkyl, -S(0) ₂ -, -NO2, -CN, and oxo. [0025] The terms "halogen," or "halo," as used herein, refer to fluorine, chlorine, bromine, or iodine.

[0026] The term "heteroaryl," as used herein, refers a monocyclic or fused polycyclic aromatic ring containing one or more heteroatoms, such as oxygen, nitrogen, or sulfur. For example, a heteroaryl can be a "5-20 atom heteroaryl," which means a 5 to 20 membered monocyclic or fused polycyclic aromatic ring containing at least one heteroatom. Examples of heteroaryl groups include pyridinyl, pyridazinyl, imidazolyl, pyrimidinyl, pyrazolyl, triazolyl, pyrazinyl, quinolyl, isoquinolyl, tetrazolyl, furyl, thienyl, isoxazolyl, thiazolyl, oxazolyl, isothiazolyl, pyrrolyl, quinolinyl, isoquinolinyl, indolyl, benzimidazolyl, benzofuranyl, cinnolinyl, indazolyl, indolizinyl, phthalazinyl, pyridazinyl, triazinyl, isoindolyl, purinyl, oxadiazolyl, thiazolyl, thiadiazolyl, furazanyl, benzofurazanyl, benzothiophenyl, benzotriazolyl, benzothiazolyl, benzoxazolyl, quinazolinyl, quinoxalinyl, naphthyridinyl, dihydroquinolyl, tetrahydroquinolyl, dihydroisoquinolyl,

tetrahydroisoquinolyl, benzofuryl, furopyridinyl, pyrolopyrimidinyl, and azaindolyl. A heteroaryl can be optionally substituted with the same substituents listed above with respect to aryl.

[0027] In other embodiments, a "5-20 member heteroaryl" refers to a fused polycyclic ring system wherein aromatic rings are fused to a heterocycle. Examples of these heteroaryls include:

[0028] The term "haloalkyl," as used herein, includes an alkyl substituted with one or more of F, CI, Br, or I, wherein alkyl is defined above. The "alkyl" portion of haloalkyl can be optionally substituted as described above with respect to alkyl.

[0029] The term "haloaryl," as used herein, includes an aryl substituted with one or more of F, CI, Br, or I, wherein aryl is defined above. The "aryl" portion of haloaryl can be optionally substituted as described above with respect to aryl.

[0030] The term "oxo," as used herein, refers to =0.

[0031] The term "nitro," as used herein, refers to -NO2.

[0032] The term "symmetrical molecule," as used herein, refers to molecules that are group symmetric or synthetic symmetric. The term "group symmetric," as used herein, refers to molecules that have symmetry according to the group theory of molecular symmetry. The term "synthetic symmetric," as used herein, refers to molecules that are selected such that no regioselective synthetic strategy is required.

[0033] The term "multivalent," as used herein, refers to a molecular fragment that is connected to at least two other molecular fragments. For example, a bridge moiety, is multivalent.

[0034] " " as used herein, refers to a point of attachment between two atoms.

[0035] The terms "solar panel" and "photovoltaic device", as used herein, are known in the art and refer to a device for converting light into electricity. Generally speaking, a solar panel or photovoltaic device comprises first and second portions. The first portion is capable of absorbing light and upon absorption of light creates an exciton. An exciton is an excited state quasiparticle consisting of a coulomb-bound electron-hole pair that diffuses to a donor- acceptor interface. The exciton then separates into free charges (positive and negative) across the interface between the first and second materials, resulting in a voltage across electrodes connected to the first and second portions, respectively.

Compounds of the Invention

[0036] In a first aspect, the molecules of the present invention are compounds represented by structural formula (I):

T^S^C-S^T ² (I).

In certain embodiments of the first aspect, C is selected from the fragments represented by structural formulas in List 1 : List 1

-Ί-

S ¹ and S ² , each independently, are absent or are selected from the fragments represented by the structural formulas in List 2:

List 2

Tl and T2, each independently, are selected from the fragments represented by the structural formulas in List 3 :

List 3

, _s X-COOH, X-CN, X-CF3, or X-H; and

further wherein:

X represents the point of attachment between fragments;

each substitutable atom (i.e., any position where a hydrogen atom is attached) is, independently, unsubstituted or substituted by a substituent R selected from a Ci-6 alkyl, - SO2R', -CN, -OR', -SR', halo, R'2, a Ce-u aryl, a C3-12 cycloalkyl, or a 5-20 atom heteroaryl; and

R' is selected from H, Ci-6 alkyl, C3-12 cycloalkyl, C ₆ -i8 aryl, or a 5-20 atom heteroaryl. Preferably, substitutable positions are unsubstituted or substituted with Ci-C ₆ alkyl or phe [0037] In certain embodiments of the first aspect, C is selected from the structural formulas in List 2. According to these embodiments, values for T ¹ , T ² , S ¹ , S ² , the substitution patterns on those fragments, R, and R' may be selected as described above and below with respect to the first aspect.

[0038] In certain embodiments of the first aspect, T ¹ and T ² are independently selected from the structural formulas in List 1. According to these embodiments, values for C, S ¹ , S ² , the substitution patterns on those fragments, R, and R' may be selected as described above and below with respect to the first aspect.

[0039] In certain embodiments of the first aspect, C is selected from the fragments represented by the structural formulas in List 4:

List 4

S ¹ and S ² , each independently, are absent or are selected from the fragments represented by the structural formulas in List 5:

List 5

Tl and T2, each independently, are selected from the fragments represented by the structural formulas in List 6:

List 6

-16-

H; and further wherein:

each atom indicated by * is, independently, unsubstituted or substituted by a substituent R selected from a Ci-e alkyl, -SO2R', -CN, -OR', -SR', halo, R'2, a Ce-u aryl, a C3-12 cycloalkyl, or a 5-20 atom heteroaryl; and

R' is selected from H, Ci-6 alkyl, C3-12 cycloalkyl, C ₆ -i8 aryl, or a 5-20 atom heteroaryl.

[0040] In certain embodiments of the first aspect, C is selected from the structural formulas in List 4. According to these embodiments, values for T ¹ , T ² , S ¹ , S ² , the

substitution patterns on those fragments, R, and R' may be selected as described above and below with respect to the first aspect.

[0041] In certain embodiments of the first aspect, T ¹ and T ² are independently selected from the structural formulas in List 5. According to these embodiments, values for C, S ¹ , S ² , the substitution patterns on those fragments, R, and R' may be selected as described above and below with respect to the first aspect.

[0042] In certain embodiments of the first aspect, R is selected from a Ci-6 alkyl, -SO2R', -CN, -OR', -SR', halo, NR'2, or a C3-12 cycloalkyl. In certain embodiments, R is selected from a Ci-6 alkyl, -CN, halo, or a C3-12 cycloalkyl. In certain embodiments, R is selected from a Ci-6 alkyl or a C3-12 cycloalkyl. In certain embodiments, R is selected from -CN or halo. According to these embodiments, values for C, T ¹ , T ² , S ¹ , S ² , the substitution patterns on those fragments, and R' may be selected as described above and below with respect to the first aspect.

[0043] In certain embodiments of the first aspect, R' is selected from H, Ci-6 alkyl, or C3- 12 cycloalkyl. In certain embodiments, R' is H. In certain embodiments, R' is Ci-6 alkyl, or C3-12 cycloalkyl. According to these embodiments, values for C, T ¹ , T ² , S ¹ , S ² , the substitution patterns on those fragments, and R may be selected as described above and below with respect to the first aspect.

[0044] In certain embodiments of the first aspect, each substitutable atom is

unsubstituted. In certain embodiments, at least one substitutable atom is substituted. In certain embodiments, at least one of the atoms indicated by * is substituted. According to these embodiments, values for C, T ¹ , T ² , S ¹ , S ² , R, and R' may be selected as described above and below with respect to the first aspect. 0045] In certain embodiments of the first aspect, C is selected from:

ents, C is

According to these embodiments, values for T , T , S , S , the substitution patterns on the fragments, R, and R' may be selected as described above and below with respect to the first aspect.

[0046] In certain embodiments of the first aspect, S ¹ and S ² , each independently, are absent or are selected from:

certain embodiments, S ¹ and S ² are . According to these embodiments, values for C, T ¹ , T ² , the substitution patterns on the fragments, R, and R' may be selected as described above and below with respect to the first aspect. [0047] In certain embodiments of the first aspect, T ¹ and T ² , each independently, are selected from:

, X-COOH, X-CN, X-CF3, or According to these

embodiments, values for C, S ¹ , S ² , the substitution patterns on the fragments, R, and R' may be selected as described above and below with respect to the first aspect.

0048] In certain embodiments of the first aspect, C is selected from:

S ¹ and S ² each independently, are absent or are selected from:

T ¹ and T ² each independently, are selected from: , X-COOH, X-CN, X-CF ₃ , or

X-H. According to these embodiments, the substitution patterns on the fragments, R, and R' may be selected as described above and below with respect to the first aspect. [0049] In certain embodiments of the first aspect, S ¹ and S ² are identical. In certain embodiments, T ¹ and T ² are identical. In certain embodiments, S ¹ and S ² are not identical. In certain embodiments, T ¹ and T ² are not identical. In certain embodiments, S ¹ and S ² are identical and T ¹ and T ² are identical. According to these embodiments, values for C, S ¹ , S ² , T ¹ , T ² , the substitution patterns on the fragments, R, and R' may be selected as described above and below with respect to the first aspect.

[0050] In certain embodiments of the first aspect, the compound is selected from:

NgN

[0051] In a second aspect, the present invention is a solar panel or photovoltaic device comprising at least one of the molecules of the first aspect.

EXEMPLIFICATION

[0052] The invention now being generally described, it will be more readily understood by reference to the following examples, which are included merely for purposes of illustration of certain aspects and embodiments of the present invention, and are not intended to limit the invention.

Example 1 : OPV Efficiency Calculations

[0053] The electronic properties of the OPV donors of the present invention are calculated as follows. For each molecule combinatorially generated from the fragment libraries, a set of initial 1500 guesses of the 3D molecular conformation are generated via the RDKit (Open source cheminformatics software used to generate conformers) conformer generator. The proposed 3D structures are minimized using the MMFF force field, described in Halgren, T. A. Merck molecular force field. "I. Basis, form, scope, parameterization, and performance of MMFF94." J. Comput. Chem. 17, 490-519 (1996), and duplicate structures are removed using the obfit functionality implemented in the Open Babel software package, which is a chemical tookit designed to speak the many languages of chemical data, described in O'Boyle, N. M. et al. "Open Babel: An open chemical toolbox." J. Cheminformatics

2011, 3, 33. Up to twenty lowest-energy conformer clusters within 5 kcal mol-1 represent energetically feasible conformations for the molecule.

[0054] The molecular geometry for each of these conformers is then optimized with the BP86/def2-SVP density functional theory (DFT) model chemistry, as described in Perdew, J. P., "Density-functional approximation for the correlation energy of the inhomogeneous electron gas" Phys. Rev. B. 1986, 33, 8822-8824; Becke, A. D. "Density-functional exchange energy approximation with correct asymptotic behavior" Phys. Rev. A 1988, 38, 3098-3100; and Weigend, F.; Ahlrichs, R. "Balanced basis sets of split valence, triple zeta valence and quadruple zeta valence quality for H to Rn: Design and assessment of accuracy." Phys. Chem. Chem. Phys. 2005, 7, 3297-3305. BP86 is a generalized gradient

approximation (GGA) functional used to optimize molecular geometries. For each unique conformation, single point energies were then calculated with B3LYP. B3LYP is a hybrid functional containing elements from density functional theory and hartree-fock theory. The double-ζ def2-SVP level of theory was used to improve the accuracy of the HOMO and LUMO energies. These frontier molecular orbital energies were then calibrated to correct for the systematic errors in the quantum chemical calculations. The calibration was performed using Gaussian Processes (GP) regression as described in Seeger, M. "Gaussian Processes for Machine Learning" Int. J. Neural Syst. 2004, 14, 69-106, based on a set of experimental HOMO and LUMO energies for 316 molecules including those in the recently published Harvard Organic Photovoltaic Set and fifty other acceptor materials used in OPVs, summarized in the supporting information. A subset of the top molecules is depicted in Scheme 2 with the PCEs and frontier molecular orbital energies.

[0055] To improve agreement between experiment and calculated frontier molecular orbitals, a Gaussian Process (GP) order is trained to predict the gap between theory and experiment in molecular orbital energies, as described in Pyzer-Knapp, E. O.; Simm, G. N.; Aspuru-Guzik, A. "A Bayesian Approach to Calibrating High-Throughput Virtual Screening Results and Application to Organic Photovoltaic Materials" Mater. Horiz. 2016, 3, 226-233.. A GP defines a distribution over functions, as weighted averages of the training data set. The weights are non-linear to a distribution of functions optimizing marginal likelihood. The marginal likelihood represents the probability of the data, given model assumptions, enhancing predictive power and model complexity.

[0056] Each molecule is encoded as a Morgan molecular fingerprint, a bit vector where each bit encodes the local connectivity of atoms with respect to its neighbors, within a given bond distance. 1024 bit vectors with a connectivity radius of 4 bonds were utilized. The GP works optimally when the model is trained with a representative dataset to learn its statistical features, optimize hyperparameters, and validate the model via cross validation to prevent overfitting. Training was done with a set 94 molecules for which experimental and theoretical results were known using a 5-fold cross validation (CV). In 5-fold CV, data is separated in 5 parts, and each part is used for testing while the rest is used for training, model performance is the average over 5-CV. Hyperparameters in this context, are the GP parameters that cannot be directly learned from training data. In this example, the parameters of the covariance matrix (also known as a kernel), were optimized according to in equation (1).

K{x ₁ , x ₂ ) = A _kernel e Λ fva ) + A ² _noise I (1)

[0057] In equation (1), A _kernel is the kernel amplitude, Tanimoto similarity, fp _a is the distribution and weighting of each data point and A _noise is the noise amplitude which is added to identical elements, adding intrinsic uncertainty to each measurement.

Hyperparameter optimization was initially carried out using Bayesian optimization to localize a starting point and numerical gradient descent. Validation was done using leave- one-out (LOO) cross validation. During LOO-CV, all data except one is used to train a model and the GP is used to predict the missing data. This is repeated for all datapoints. This procedure was done with the HOMO and LUMO orbital energies and repeated for other properties including HOMO-LUMO gap. These calibrated HOMO, LUMO, and gap values enter the modified Scharber model, as described below, for computing open-circuit voltage for electron acceptor materials.

[0058] PCEs are calculated by a modified Scharber model. First, rubrene, a benchmark donor material, is set as the reference donor material. The measured frontier molecular orbital energies of rubrene are HOMO: -5.40 eV, LUMO: -3.00 eV, gap 2.40 eV. The open-circuit voltage (Voc) is the ΕΌΑ (energy difference between the D ^H0M0 and A ^LUM0 ) and the short circuit current density (Jsc) assumes that all photons in the solar spectrum with energy greater than the HOMO-LUMO gap of the acceptor material are absorbed. To give a non-zero predicted PCE, the D ^H0M0 - A ^H0M0 offset is required to be larger than 0.3 eV to promote interfacial charge separation. An empirical loss parameter of 0.3 eV was incorporated in the computational methods described herein, which accounts for typical voltage loss in experimental devices. The Scharber model stipulates that the fill factor (FF) and external quantum efficiency (EQE) are set to 65%. PCEs are computed using Eq (3).

Voc-FF-Jsc

PCE = 100 (3)

[0059] Using this model, PCEs were calculated for the exemplary molecules depicted in Scheme 2. Under each molecule is shown, in the first row, the PCE based on calibrated HOMO and LUMO energies, and in the second row, the calibrated HOMO (left) and LUMO (right) energies in eV.

10.9% 10.6% 10.1 % -5.63, -3.90 -5.66, -3.84 -5.65, -3.74

Scheme 2

[0060] Using results generated by these techniques, a statistical analysis of the fragments was performed to determine which fragments overperform relative to their population in the initial fragment library. First a statistical analysis is performed of the frequency of the molecular fragments used to build the screening set. A hypergeometric geometric analysis was used to assess the frequency of the fragments in the top candidates. This hypergeometric distribution gives the probability of finding k instances in a subpopulation n, given K observations in the entire population N, see J. A. Rice, Mathematical Statistics and Data

Analysis, Duxbury Press, Pacific Grove, CA, 3rd edn, 2006. A Z-score is computed for each fragment based on its relative composition in top molecules compared to a random distribution. The Z-score is computed according to Eq. 4.

(k-<k>)

a(k) >

[0061] Here, k is the frequency of a fragment in the top set (PCE>8.0%); <k> is the expectation value given by <k>=nK/N. Positive Z-scores indicate that a particular fragment is overrepresented in the top candidates; negative Z-scores indicate that a particular fragment is underrepresented amongst the top candidates. Z-scores for cores, spacers, and terminals were computed to elucidate design principles for top non-fullerene acceptor materials. Figure 1 shows the Z-score distribution for exemplary core fragments. Molecules that contain higher Z scored fragments also have on average, higher PCE values. Z-scoring combines average PCE and variability into a single measure. Figure 1 shows that the

diketopyrrolopyrrole (DPP), naphthalenetetracarboxylic dianhydride ( DI), and quinoidal oligothiophene core fragments are statistically overrepresented in top candidates.

[0062] This analysis was also applied to 'spacer' fragments, as shown in Figure 2. The fused-thienobenzothiadiazole spacer has the highest Z-score of any fragment in this study. This fragment and the others with positive Z-scores are less electrophilic than the NDI and DPP cores. These cores are subjected to a 'push-pull' effect that decreases the HOMO- LUMO gap and improves JSC and in turn, PCE. Those with negative Z-scores are typically more electrophilic (carbonyl-based and perfluorinated aryl rings). The electron-poor nature of these fragments decrease both HOMO and LUMO energies. The HOMO-LUMO gap remains approximately the same or increases with two adjacent electrophilic fragments.

[0063] Finally, the terminal fragments were similarly analyzed to determine how best to cap OPV molecules. Figure 3 shows the distribution of Z-scores for the terminal fragments. As with the spacers, the best terminal fragments are electron-donating (p-aniline) or weakly electron-withdrawing. Other mercanocyanine derivatives have positive values, the extended π-system likely helps to decrease HOMO-LUMO gaps, which increases JSC and PCE. Carbonyl-based and electron-poor thiophene terminal groups deactivate the activities of terminal groups. Bulky terminal groups (i.e. indoles) cause substantial unfavorable twisting along the backbone and decreased π-conjugation. The main design principles that can be inferred from this analysis are as follows: Use an electron-deficient core and build outwards with ambiphilic or electron-donating fragments. This will result in materials with small HOMO-LUMO gaps and substantial Voc.

Example 2: Preparation of Exemplary Compounds of the Invention

[0064] Compounds of the present invention can be prepared from the fragments via well- understood cross-coupling reactions, as described, for example, in Miyaura, N., Suzuki, A. Chem. Rev. "Palladium-catalyzed Cross-Coupling Reactions of Organoboron Compounds" 1995, 95, 2457-2483; Cheng, Y-J.; Yang, S-H.; Hsu, C-S. "Synthesis of Conjugated

Polymers for Organic Solar Cell Applications" Chem. Rev. 2009, 109, 5868-5923; and Cherney, A. H.; Kadunce, N. T.; Reisman, S. E. "Enantioselective and Enantiospecific Traniti on-Metal -Catalyzed Cross-Coupling Reactions of Organometallic Reagents to Construct C-C bonds" Chem. Rev. 2015, 115, 9587-9652.

[0065] An exemplary synthesis of a diketopyrrolopyrrole (a DPP), depicted in Scheme 3, is provided in Sparke, J. M.; Watson, K. D. J. Chem. Soc, Perkin Trans. 1976, 1, 5.

Scheme 3

[0066] An exemplary synthesis of a dicyano-fused thiophene, depicted in Scheme 4, is provided in Takahashi, K.; Fujita, S.; Akiyama, K.; Miki, M.; Yanagi, K. Angew. Chem. Int. Ed. 1998, 37, 2484.

Scheme 4

[0067] An exemplary synthesis of a quinoidal thiophene, depicted in Scheme 5, is provided in Komatsu, M.; Nakazaki, J.; Uchida, S.; Kubo, T.; Segawa, H. Phys. Chem.

Chem. Phys. 2013, 1 Scheme 5

[0068] The teachings of all patents, published applications and references cited herein are incorporated by reference in their entirety.

[0069] While this invention has been particularly shown and described with references to example embodiments thereof, it will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the scope of the invention encompassed by the appended claims.