Colorless Near‐Infrared Absorbing Dyes Based on B‐N Fused Donor‐Acceptor‐Donor π‐Conjugated Molecules for Organic Phototransistors

Abstract

The introduction of a colorless function to organic electronic devices allows responses to light in the near‐infrared (NIR) region and is expected to broaden the applications of these devices. However, the development of a colorless NIR dye remains a challenge due to the lack of a rational molecular design for controlling electronic transitions. In this study, to suppress the π‐π* transitions in the visible region, polycyclic donor‐acceptor‐donor π‐conjugated molecules with boron bridges (Py‐FNTz‐B and IP‐FNTz‐B) are designed and synthesized, which contain pyrrole or indenopyrrole as donor units with fluorinated naphthobisthiadiazole (FNTz) as an acceptor unit. The pyrrole end‐capped Py‐FNTz‐B shows an absorption band in the NIR region without distinct visible‐light absorption, which has led to the establishment of colorless characteristics. The indenopyrrole end‐capped IP‐FNTz‐B shows a narrow optical energy gap of 0.87 eV in films. Time‐resolved microwave conductance and field‐effect transistors demonstrate the semiconducting characteristics of these molecules, and Py‐FNTz‐B‐based devices function as NIR phototransistors. Theoretical analyses indicate that the combination of a polyene‐like electronic structure with orbital symmetry is important to obtain NIR wavelength‐selective absorption. This study suggests that a molecular design based on electronic structures can be effective in the development of colorless NIR‐absorbing dyes for organic electronics.

Introduction

The development of high-performance organic semiconducting materials has enabled the embodiment of organic thin-film electronics such as organic light-emitting diodes, organic solar cells, and organic field-effect transistors (OFETs).^[1] As one of the next logical applications, organic electronic devices that could respond to light in the near-infrared (NIR) region would be in high demand due to distinct advantages in photodetection, bioimaging, chemotherapy, optical filters, and security marking.^[2] To realize such NIR-type devices, it is essential to develop novel organic semiconductors with a bandgap of less than 1.5 eV (equivalent to wavelengths longer than 800 nm) as well as charge-transport characteristics. Narrow bandgaps can be produced by using one of the donor (D)-acceptor (A) configurations,^[3] quinoidal conjugations,^[4] phthalocyanines,^[5] polyrylenes,^[6] cyanine dyes,^[7] antiaromatic conjugations,^[8] radical characteristics,^[9] and B-N fused structures.^[10] However, precedents whereby these compounds also possess semiconducting characteristics have been still limited.^{[3,4a–d,f,7a,b,d,e]}

Unlike the visible region, the NIR region is invisible, and the introduction of an additional function such as “colorless and transparent” has resulted in the development of novel NIR-response organic electronic devices.^{[3a,d,f,h,7b,d,e]} Although several NIR absorption dyes are known to furnish colorless characteristics,^{[6b–d,9c,10a–d,g,h,j,k,m–o]} the establishment of a rational design principle for molecular structure and application to electronic devices remains a challenge that has not been developed.

Obtaining colorless characteristics for NIR absorption dyes based on π-extended molecules requires 2 developments: 1) Narrowing the absorption bandgap to reach the NIR region; and, 2) Suppressing electronic transitions that correspond to the absorption of visible light (Figure 1a,b). Nakamura et al. reported that the L2B2 molecule, which is composed of A-D-A types of structures with B-N linkages, could achieve an absorption spectrum that reaches 1100 nm (Figure 1c).^[10e]

[IMAGE OMITTED. SEE PDF]

However, L2B2 projects a blue color when used in a solution due to the transitions that are allowed in the visible region. We hypothesized that a colorless NIR absorption dye could be achieved via a molecular design wherein the first excited state shows an allowed transition while the second and third excited states are forbidden transitions (Figure 1b). Focusing on the electronic structure of L2B2, the central benzodipyrrole unit would cause multiple π-π* transitions due to the orbital symmetries of the benzene ring, which would result in broad absorption bands in the vis-NIR region. We assumed that tuning the electronic structures could accomplish NIR wavelength-selective absorption and that the contribution of the polyene-like electronic structure of naphthobisthiadiazole (NTz) would suppress multiple π-π* transitions via a resonance effect (Figure 1d). In addition, the strong electron-accepting nature of the NTz unit has the advantage of acquiring a narrow bandgap within a D-A system. In this context, we developed a fluorinated NTz, defined as FNTz, as a strong electron-accepting unit that contains D-A type derivatives with high carrier-transport characteristics.^[11] Therefore, in this study, we designed and synthesized polycyclic D-A-D type π-conjugated molecules with boron bridges between the D and A units (Py-FNTz-B, IP-FNTz-B), as shown in Figure 1e. The pyrrole-capped Py-FNTz-B showed maximum absorption at 900 nm and colorless and transparent characteristics in thin films. IP-FNTz-B has terminal D units that are composed of indenopyrrole and show a maximum absorption wavelength of 1200 nm with an onset that reaches upward of 1500 nm. Furthermore, these molecules exhibited semiconductor characteristics in OFETs, and Py-FNTz-B has also been used to produce organic phototransistors (OPTs) with current amplification in response to NIR light.

Results and Discussion

Computational Studies

To clarify the electronic states and the energy levels of Py-FNTz-B and IP-FNTz-B as well as the corresponding π-conjugated framework without B-N linkages (structures of Py-FNT and IP-FNTz, as shown in Scheme 1), density functional theory (DFT) calculations were conducted at the B3LYP/6-31g(d,p) level (Figure 2). All the alkyl groups were replaced with either hydrogen atoms or methyl groups to reduce the computational cost. Thus, these were denoted as Py-FNTz-B’, IP-FNTz-B’, Py-FNTz’, and IP-FNTz’. As shown in Figure 2, the highest occupied molecular orbitals (HOMOs) of these molecules were distributed across the entire π-conjugated frameworks, while the lowest unoccupied molecular orbitals (LUMOs) were localized on the FNTz unit due to its strong electron-accepting characteristics. The HOMO energy levels of Py-FNTz-B’ and IP-FNTz-B’ were almost the same as those of Py-FNTz’ and IP-FNTz’. On the other hand, the LUMO energy levels of Py-FNTz-B’ (−3.72 eV) and IP-FNTz-B’ (−3.57 eV) were relatively low-lying compared with those of Py-FNTz’ (−3.00 eV) and IP-FNTz’ (−2.97 eV). As a result, Py-FNTz-B’ (1.36 eV) and IP-FNTz-B’ (1.05 eV) showed a narrow HOMO-LUMO energy gap, compared with those of Py-FNTz’ (2.01 eV) and IP-FNTz’ (1.68 eV). Focusing on the extension of π-conjugation showed that the HOMO-LUMO energy gap of IP-FNTz-B’ was narrower by 0.31 eV than that of Py-FNTz-B’, which indicated that utilization of the π-extended D unit is effective for narrowing the HOMO-LUMO energy gap.

[IMAGE OMITTED. SEE PDF]

[IMAGE OMITTED. SEE PDF]

We used UHF/6-31g(d) to estimate the diradical character (y₀)^[12] of these D-A-D compounds because some NIR absorption dyes show resonance between a closed-shell aromatic state and an open-shell diradical singlet ground state.^[13] In this estimation, the y₀ value of a completely closed-shell state was calculated as 0, while the open-shell single diradical state was 1. As shown in Figure 2, the y₀ values of Py-FNTz’ and IP-FNTz’ were determined to be 0.00. On the other hand, the narrow bandgap molecules of Py-FNTz-B’ and IP-FNTz-B’ showed moderate y₀ values of 0.31 and 0.40, respectively.

Synthesis and Thermal Properties

The processes for the syntheses of Py-FNTz-B and IP-FNTz-B appear in Scheme 1. First, we synthesized the pyrrole and indenopyrrole terminal units for these molecules. Pyrrole unit 3 was obtained from 3-hexylpyrrole (1) via protection from a tert-buthoxycarbonyl (Boc) group and a subsequent regioselective borylation. To construct the indenopyrrole unit 10, 1,4-dihydroindeno[1,2-b]pyrrole (7) was synthesized from 1-(p-toluenesulfonyl)pyrrole (5) through Friedel-Craft acylation with 2-iodobenzoyl chloride, intramolecular cyclization by Suzuki-Miyaura coupling, and a reduction of the carbonyl group using LiAlH₄. Via protection of the pyrrole N-H proton of 7 using a Boc group, alkylation, and the following borylation reactions were conducted to provide 10. Next, Suzuki-Miyaura coupling between FNTz-Br and 3 in the presence of Pd(dppf)Cl₂·CH₂Cl₂ afforded Py-FNTz in a 67% yield. Note that the deprotection of the Boc groups proceeded smoothly under heated conditions. Finally, intramolecular B-N linkages were formed by the reaction of Py-FNTz with triphenylborane (BPh₃) in toluene at 130 °C under a nitrogen atmosphere to provide Py-FNTz-B. IP-FNTz-B was also prepared via the same synthetic protocol. Notably, IP-FNTz-B showed sufficient solubility against common solvents such as toluene, chloroform, and chlorobenzene. All the new compounds were unambiguously identified and characterized via ¹H and ¹³C NMR spectroscopy as well as by high-resolution mass spectroscopy (HRMS).

To investigate the thermal stabilities, thermogravimetric analysis (TGA) was conducted for Py-FNTz-B and IP-FNTz-B (Figure S1, as shown in the Supporting Information). The TGA curves of Py-FNTz-B and IP-FNTz-B showed a 5% weight loss at 250 and 265 °C, respectively, which indicates that these π-conjugated frameworks possess thermal stability adequate for application to organic semiconducting materials.

X-ray Analyses

To investigate the influence that boron linkages exert on the molecular structures, single-crystal X-ray diffraction was conducted for IP-FNTz and IP-FNTz-B.^[14] The single crystals of IP-FNTz and IP-FNTz-B were slowly grown by vapor diffusion with chlorobenzene/methanol. The monomer structures appear in Figure 3a,b and Figure S2 (Supporting Information). IP-FNTz showed a highly planar structure with a dihedral angle of 0.17°, which was due to the intramolecular hydrogen bonding between N-H-N (2.14 Å) and C-H-F (2.31 Å). IP-FNTz-B also showed a planar structure with dihedral angles of 0.97° and 9.17°. The B atoms of IP-FNTz-B were located by a slight displacement of 0.05 and 0.31 Å against the π-plane of FNTz. Such a displacement has been reported in the structure of quadruply B-N-fused dibenzoazaacene.^[10i] Phenyl groups on the boron atoms and alkyl chains were perpendicularly oriented to the π-backbone.

[IMAGE OMITTED. SEE PDF]

We analyzed the influence that the electronic structure of the FNTz unit exerted on the chemical structures. Based on the carbon-carbon distances of C4-C5 (1.434 Å), C5-C6 (1.381 Å), C6-C7 (1.421 Å), C7-C8 (1.414 Å), C8-C9 (1.416 Å), C9-C10 (1.379 Å), and C10-C11 (1.428 Å) (Figure 3c), the intramolecular bond length alternative (BLA) of the FNTz unit was estimated using the following equation: 1 $$$\begin{equation}{\overline {BLA} _{x - y - z}} = {\mathrm{\;}}\frac{{\left| {{l_1} - {l_2}} \right| + \left| {{l_2} - {l_3}} \right| + \cdot \cdot \cdot \left| {{l_{n - 1}} - {l_n}} \right|}}{{\left( {n - 1} \right)}}\end{equation}$$$

In that equation, the $${\overline {BLA} _{x - y - z}}$$ parameter is defined as the average difference in the distance between the continuous C-C and C = C bonds in the conjugated path, n denotes the number of bonds related to BLA, and l_n denotes the corresponding bond length.^[15] The $${\overline {BLA} _{x - y - z}}$$ parameter of IP-FNTz-B from C6-C11 was determined to be 0.038 Å, which is shorter than that of IP-FNTz (0.052 Å). Distinct bond length alternations similar to the polyene structure were observed in both molecules. In particular, the shorter $${\overline {BLA} _{x - y - z}}$$ in IP-FNTz-B implies a more effective delocalization of π-electrons such as that in the band structure of doped polyacetylene. This is because the diradical characteristics of IP-FNTz-B were increased, as shown in Figure S3 (Supporting Information), which is in good agreement with the calculated results.

Focusing on the packing diagram, we found that IP-FNTz formed π-π stacking structures in the single crystal with the shortest C-C distance (3.368 Å), as shown in Figure S4 (Supporting Information). By contrast, IP-FNTz-B formed dimer structures with the central FNTz units placed in a face-to-face stacking order with twisted angles of 74.6°, as shown in Figure 4a. In addition, the shortest intermolecular C9-C9 distance was determined to be 3.196 Å (Figure 4b), which is significantly shorter than the sum of the van der Waals radii of carbon and carbon atoms (C-C: 3.40 Å). This result indicated the presence of a relatively strong π-π intermolecular interaction between the π-planes in the dimer structure. DFT calculations predicted a moderate diradical characteristic for IP-FNTz-B of 0.40 (Figure 2) and a spin density that is localized mainly on the central NTz unit, as shown in Figure S5 (Supporting Information). Therefore, we assumed that the short π-π stacked distance of IP-FNTz-B was caused by the radical-radical interaction.^[16] Such a short and strong π-π stacking of IP-FNTz-B has an advantage that allows the transport of charge carriers between the molecules. To investigate the influence of the stacking structures on the charge transport characteristics, Amsterdam Density Functional (ADF) calculations were conducted to estimate the transfer integral at the PW91/TZP level. The transfer integrals of the dimer fragment extracted from the crystal structure of IP-FNTz were calculated to be 5.3 meV for the holes (t_h) and 34.5 meV for the electrons (t_e) as shown in Figure S4 (Supporting Information). On the other hand, the corresponding transfer integrals of IP-FNTz-B were estimated to be 32.0 meV for t_h and 59.1 meV for t_e, which are larger than the corresponding values for IP-FNTz. This result indicates that the introduction of B-N linkages contributed to increasing the transfer integrals due to the formation of closed packing structures.

[IMAGE OMITTED. SEE PDF]

Photophysical Properties

The UV–Vis-NIR absorption spectra of Py-FNTz, IP-FNTz, Py-FNTz-B, and IP-FNTz-B in chloroform solutions appear in Figure 5a,b, and as well as in photophysical data summarized in Table 1. Py-FNTz showed an absorption band with a peak top of 583 nm in the visible region, which resulted in a solution with a blue color. On the other hand, upon the formation of the B-N linkage, the absorption band that corresponds to the π-π* transition was significantly red-shifted toward the NIR region for Py-FNTz-B, and its maximum absorption wavelength (λ_max) reached 900 nm. Based on the calculated results shown in Figure 2, we considered such a significant red-shift of Py-FNTz-B would have derived from the deepened LUMO energy levels. In addition, there were no apparent absorption bands (< 2000 M⁻¹ cm⁻¹) in the visible region of around 450–650 nm. Eventually, Py-FNTz-B showed an almost colorless solution. A similar tendency was also observed for IP-FNTz and IP-FNTz-B; namely, the λ_max value of IP-FNTz-B reached 1091 nm, which showed a bathochromic shift by ≈400 nm compared with that of IP-FNTz.

[IMAGE OMITTED. SEE PDF]

Table 1 Summary of physical properties.

Compd.	λmax (ε) [nm]a)	ΔEopt [eV]b)	Ered [V]c)	Eox [V]c)	IP [eV]d)	EA [eV]e)
Py-FNTz	583 (1.9 × 104)	–	−1.52	0.23	–	–
Py-FNTz-B	900 (1.7 × 104)	1.17	−0.88	0.36	5.62	3.93
IP-FNTz	681 (4.0 × 104)	–	−1.47	0.18	–	–
IP-FNTz-B	1134 (3.0 × 104)	0.87	−0.90	0.10	5.27	4.03

We also measured the absorption spectra in neat Py-FNTz-B and IP-FNTz-B thin films (Figure 5c). The absorption spectra of Py-FNTz-B and IP-FNTz-B appeared in an absorption band with peak tops at 827 and 1134 nm, respectively. Py-FNTz-B formed colorless films, which maintained the colors of the solution. Since the onset of IP-FNTz-B reached 1500 nm, this molecule showed significantly narrow optical energy gaps (ΔE_opt) of 0.87 eV in the film state. This result was attributed to the strong intermolecular interactions between the π-planes due to the concentration of molecules.

Physicochemical Properties

The electrochemical properties of Py-FNTz, Py-FNTz-B, IP-FNTz, and IP-FNTz-B were investigated via cyclic voltammetry (CV) and differential pulse voltammetry (DPV) measurements. All the measurements were conducted in a dichloromethane solution containing 0.1 M tetra-n-butylammonium perfluorophosphate (TBAPF₆) as a supporting electrolyte, and the oxidation/reduction potentials were referenced against ferrocene/ferrocenium (Fc/Fc⁺) as an internal standard. Oxidation and reduction potentials were determined by the top peak of DPV, as shown in Figure S6 (Supporting Information). Focusing on the anodic scan, Py-FNTz and Py-FNTz-B show irreversible oxidation waves in Figure 6. This irreversibility indicates that Py-FNTz would undergo polymerization under electrochemical oxidation because a dark-colored film was formed on the working electrode after repeated CV cycles. On the other hand, IP-FNTz and IP-FNTz-B showed 2 reversible oxidation waves, which indicates the stable formation of radical cationic and dicationic species. Since the α-positions of pyrroles are end-capped by the fused structure with phenyl groups, the decomposition of cationic species was suppressed. In the cathodic scan, Py-FNTz showed 1 irreversible and 2 reversible reduction waves. On the other hand, IP-FNTz and IP-FNTz-B showed 1 reversible reduction wave. The first reduction potentials, E_red, of both Py-FNTz-B and IP-FNTz-B appeared at −0.88 and −0.90 V, respectively, as these were anodically shifted compared with those of Py-FNTz (−1.52 V) and IP-FNTz (−1.47 V) (Table 1). Based on the assumption that Fc/Fc⁺ is below 4.8 eV from the vacuum level, the HOMO/LUMO energy levels were estimated to be −3.28/−5.03 eV for Py-FNTz, −3.92/−5.16 eV for Py-FNTz-B, −3.33/−4.98 eV for IP-FNTz, and −3.90/−4.90 eV for IP-FNTz-B. These results indicate that the introduction of boron bridges efficiently lowers the LUMO energy levels, which is consistent with the qualitative trend observed in the DFT calculations.

[IMAGE OMITTED. SEE PDF]

To gain information on the energy level regarding ionization potential (IP) and electron affinity (EA) in the films, photoelectron yield spectroscopy (PYS) and low-energy inverted photoemission spectroscopy (LEIPS) were used. As shown in Figure 7, IP/EA was determined to be 5.62/3.93 eV for Py-FNTz-B and 5.27/4.03 eV for IP-FNTz-B. The observed EA values indicated that these molecules have the potential to function as n-type organic semiconducting materials.

[IMAGE OMITTED. SEE PDF]

Semiconducting Characteristics

To investigate whether Py-FNTz-B and IP-FNTz-B possess intrinsic carrier-transport characteristics at the nanometer level, time-resolved microwave conductance (TRMC) measurements were performed for the films prepared on quartz substrates.^[17] The microwave frequency and its power were ≈9 GHz and ≈3 mW, respectively. A third harmonic generation (355 nm) of a Nd:YAG laser (Continuum Inc., Surelite II, 5−8 ns pulse duration, 10 Hz) was used for the excitation (incident photon density I₀ = 9.1 × 10¹⁵ photons cm^–2 pulses^–1). The photoconductivity (Δσ = A⁻¹ ΔP_r P_r^–1 where A is the sensitivity factor, P_r is the reflected microwave power, and ΔP_r is the change in P_r upon exposure to light) was converted into the product of the quantum yield (φ) and sum of the hole mobility (µ_h) and electron mobility (µ_e), Σµ ( = µ_h + µ_e), using the relationship φΣµ = Δσ(eI₀F_light)^–1, where e and F_Light are the electron charge and correction (or filling) factors, respectively.

The experiments were performed at room temperature under air. Thin films were prepared via spin-coating using chloroform solutions of Py-FNTz-B and IP-FNTz-B. As shown in Figure 8a,b, transient photoconductivity was observed in both compounds, indicating the possibility of semiconducting characteristics. The value of φΣµm_ax for Py-FNTz-B was determined to be 1.2 × 10⁻⁸ m² V⁻¹ s⁻¹, which is higher than that of IP-FNTz-B (2.5 × 10⁻⁹ m² V⁻¹ s⁻¹).

[IMAGE OMITTED. SEE PDF]

Bottom-gate, bottom-contact types of transistor devices were utilized to evaluate the semiconducting characteristics of thin films at the micrometer level, and the active layers were fabricated via spin-coating onto an octadecyltrichlorosilane (ODTS)-modified Si/SiO₂ substrate. The optimized result is listed in Table S1 (Supporting Information). The Py-FNTz-B film on an ODTS-modified SiO₂/Si substrate showed typical n-type characteristics with a maximum electron mobility (µ_e,max) of 2.4 × 10⁻⁵ cm² V⁻¹ s⁻¹, a threshold voltage (V_th) of 20 V, and a current on/off ratio (I_on/I_off) of 10⁴, as shown in Figure 8c and Figure S7 (Supporting Information). On the other hand, the IP-FNTz-B film showed both p- and n-type characteristics (Figure 8d; Figure S8, Supporting Information). The hole and electron mobilities were determined to be 2.0 × 10⁻⁴ and 9.5 × 10⁻⁵ cm² V⁻¹ s⁻¹, respectively (Table S2, Supporting Information). These mobilities of Py-FNTz-B and IP-FNTz-B are relatively higher than those of other reported π-conjugated derivatives incorporating a BPh₂ bridge (10⁻⁷–10⁻⁶ cm² V⁻¹ s⁻¹).^[18]

To clarify the differences in the mobilities between Py-FNTz-B and IP-FNTz-B, atomic force microscopy (AFM) measurement and X-ray diffraction were conducted. Based on analysis of the AFM measurement, the Py-FNTz-B film showed relatively rough morphologies with sub-micrometer grains (Figure S9a, Supporting Information) while the IP-FNTz-B film showed grains larger than 1 mm (Figure S10a, Supporting Information). The out-of-plane (XRD) characterization of the Py-FNTz-B film showed several peaks around 2θ = 2–5°, which indicated the formation of a polycrystalline film (Figure S9b, Supporting Information). On the other hand, the IP-FNTz-B film showed a peak at 5.72°, which corresponded to a d-spacing of 15.7 Å (Figure S10b, Supporting Information). These results indicated that IP-FNTz-B formed a higher orientation in the film state compared with that of Py-FNTz-B, which resulted in a higher level of carrier mobility.

By taking advantage of the NIR-selective absorption and n-type semiconducting characteristics of Py-FNTz-B, the OPT characteristics were evaluated in the dark and under NIR light illumination of 810 nm at an intensity of 143 mW cm⁻² using light-emitting diodes (LEDs). As shown in Figure 9a, the drain-source current (I_DS) was enhanced up to 10³-fold under irradiation. On the other hand, this device was almost inactive under the illumination of visible light using room-light LEDs, which indicates the NIR selective detection (Figure 9b). To reveal the current enhancement factor, the dependence of photo intensity on I_DS was investigated. As shown in Figure 9c and Figure S11 (Supporting Information), the photocurrent, defined as I_light−I_dark, was linearly increased against photo intensity under V_G = 0 V. This result indicates that the Py-FNTz-B-based device functioned as a phototransistor. Figure S12 (Supporting Information) shows the photosensitivity (P), photodetectivity (D*), and photoresponsivity (R) values as functions of the gate voltage V_G. Based on this result, the maximum photosensitivity P was estimated to be 10³, shown by the equation P = (I_light − I_dark)/I_dark. The specific detectivity (D*) is given by 2 $$$\begin{equation}{D^*}{\mathrm{\;}} = {\mathrm{\;}}\frac{{R\sqrt S }}{{{I_N}}}\end{equation}$$$where R is the photoresponsivity, S is the channel area, I_N is the noise spectral density.^[19] Since the shot noise of a photodetector is dominated by the dark current (I_dark) in the device, the noise spectral density can be approximately expressed as follows: 3 $$$\begin{equation}{I_N} = \;\sqrt {2q{I_{dark}}} \end{equation}$$$where q is the electron charge, I_dark is the dark current.^[19] According to Equations (2) and (3), D* value was determined to be 8.9 × 10⁹ Jones. It should be noted that the D* would potentially provide an overestimated value if the observed dark current were higher than the noises.^[20] Photodetectivity R was determined to be 4.2 mA W⁻¹, as shown in Figure S12c (Supporting Information). Time-dependent photoresponses were measured under NIR light at 810 nm with on/off switching at 30 s intervals under 0 V of V_G and 100 V of V_DS. As shown in Figure 9b, this device showed stable responses with repeated amplifications and suppression of I_DS. These results indicated that Py-FNTz-B based on the D-A-D structure with B-N linkage has the potential for application as OPTs selective for NIR light.

[IMAGE OMITTED. SEE PDF]

To measure the charge-transport characteristics of the IP-FNTz-B film against a sandwich-type device with a vertical carrier-transport direction, we also fabricated space-charge-limited current (SCLC) devices that consisted of ITO/ZnO/IP-FNTz-B/Ca/Al. The J–V plots appear in Figure S13a (Supporting Information), and the SCLC electron mobilities were determined to be 1.3 × 10⁻⁷ cm² V⁻¹ s⁻¹. The active layer in the inset of Figure S13a (Supporting Information) shows almost transparent properties. As shown in Figure S13b (Supporting Information), the transmittance spectra of the active layer on ZnO/ITO showed weak absorption bands at less than 500 nm and at more than 1000 nm and thus showed a high transmittance of more than 80% in the 500–1000 nm region. Although fabricating devices utilizing transparent metal electrodes falls beyond the scope of this study, these results indicate the potential for realizing transparent devices.

Analyses of Electronic Transitions

Both Py-FNTz-B’ and L2B2 possess the same symmetrical structure with B-N linkages, which belong to the point group as C_2h. However, there were distinct differences in visible light absorption. To unveil the origin of the significantly different photophysical properties, the details of the electronic transition states were theoretically analyzed using time-dependent (TD)-DFT calculations at the CAM-B3LYP/6-31g(d,p) level. As summarized in Table 2, the first excited state (S₁) of Py-FNTz-B’ was the main contribution to the HOMO-LUMO transition, and the corresponding wavelength was estimated to be 806 nm with an oscillator strength (f) of 0.417. The second (S₂) and third (S₃) excited states were attributed mainly to the HOMO-LUMO+1 and HOMO−1-LUMO transitions, respectively, and these wavelengths were estimated to be 538 and 469 nm, respectively. However, their f values were determined to be 0.000. These represent the prevention of transitions and indicate that there is little absorption in the visible region for Py-FNTz-B’, where the appearance of a colorless state was expected. These simulated transitions of Py-FNTz-B’ well reproduced the absorption spectra, as shown in Figure 10. On the other hand, irrespective of a similar structure such as Py-FNTz-B’, the L2B2 molecule showed the allowance of electronic transitions in both the S₁ and S₃ states (Table 2). This result indicates that L2B2 shows absorption bands in both the visible and NIR regions, which also is consistent with the experimental results.

Table 2 List of electronic transitions for Py-FNTz-B’ and L2B2.

Compd.	ESa)	λ [nm]	f	Major contributionsb)	Electronic transition
	S1	806	0.417	HOMO to LUMO (97%)	Au to Bg : Allowed
Py-FNTz-B’	S2	538	0.000	HOMO to LUMO+1 (68%)	Au to Au : Forbidden
	S3	469	0.000	HOMO−1 to LUMO (67%)	Bg to Bg : Forbidden
	S1	760	0.242	HOMO to LUMO (95%)	Bg to Au : Allowed
L2B2	S2	698	0.000	HOMO to LUMO+1 (94%)	Bg to Bg : Forbidden
	S3	569	0.323	HOMO−1 to LUMO (93%)	Bg to Au : Allowed

[IMAGE OMITTED. SEE PDF]

To determine the reason for such a difference in the electronic transitions between Py-FNTz-B’ and L2B2, we focused on the molecular symmetries at each of the energy levels of Py-FNTz-B’ and L2B2. Based on rules for the quantum mechanics of selection, the possibility of electronic transition is described in Equation (4).^[21]4 $$$\begin{equation}\smallint {\varphi _i}\mu {\varphi _f}d\tau \end{equation}$$$

In Equation (4), φ_i and φ_f denote the wave functions at the initial and final states, respectively, and µ is the electronic transition moment. When the integral value in Equation (4) becomes 0, the electronic transition is forbidden. On the other hand, a value other than 0 allows the transition. This difference can be determined by the change in the orbital symmetry when there is an electronic transition, which is referred to as the Laporte rule.^[22] According to this rule, parity of the quantum state is essential for transition. For example, orbital symmetries in the initial and final states in a symmetrical orientation are referred to as gerade (as subscript g) (an asymmetrical orientation is referred to as ungerade (as subscript u)), in which case the corresponding electronic transition is forbidden. On the other hand, when the orbital symmetry is inverted from g to u, the electronic transition should be allowed.

Since the point group of both Py-FNTz-B’ and L2B2 was assigned C_2h symmetry, the symmetry of the molecular skeleton was not related to an electronic transition. Instead, we assumed that it would be important to focus on the orbital symmetries of energy levels in order to establish the possibility of electronic transition. In the case of L2B2, the HOMO and HOMO−1 orbitals were localized mainly on the central benzodipyrrole (BDP) unit, as shown in Figure 11b and Figure S14 (Supporting Information). The degeneration of the HOMO energy level in the intrinsic benzene ring was divided into HOMO and HOMO−1 orbitals due to the perturbation that originated from the fused pyrrole rings. Hence, the point groups of both HOMO and HOMO−1 orbitals are considered to be B_g. On the other hand, the LUMO and LUMO+1 orbitals were dominant on the terminal benzothiaziazole units. Therefore, the LUMO orbital is considered to be A_u due to the in-phase orientation, while the LUMO+1 orbital is considered to be B_g due to an antiphase orientation. As a result, not only the S₁ state of the HOMO-LUMO transition but also the S₃ state of the HOMO−1-LUMO transition was allowed (Table 2), which resulted in a low level of selective absorption.

[IMAGE OMITTED. SEE PDF]

The HOMO−1 and HOMO of Py-FNTz-B’ were delocalized over the π-conjugated backbone (Figure 11a; Figure S14a, Supporting Information). Since these orbital arrangements are regarded as the 14−electron system based on a polyene structure, the HOMO−1 and HOMO possess 5 and 6 nodes, respectively (Figure 11a). Hence, the HOMO−1 and HOMO of Py-FNTz-B’ were considered to be B_g and A_u, respectively. In contrast, the LUMO and LUMO+1 were localized on the central FNTz unit. Hence, based on the orbital symmetry of the FNTz unit, the LUMO and LUMO+1 symmetries were considered to be B_g and A_u, respectively. As a result, the S₂ transition of HOMO-LUMO+1 and the S₃ transition of HOMO−1-LUMO are symmetrically forbidden, while the HOMO-LUMO transition is allowed (Table 2), which led to NIR-wavelength selective absorption and a colorless behavior.

Based on the tuning of orbital interactions and orbital symmetries, the incorporation of polyene structures enables the design of molecules possessing NIR selective absorption. As summarized in Figure S15 (Supporting Information), this molecular design also allows us to rationalize the transparency of the reported compounds. Therefore, we have considered that this concept is effective for designing NIR selective absorbing dyes with colorless and transparent features. Additionally, it is helpful for interpreting the relationship between absorption properties and molecular structures.

Conclusion

In summary, to develop colorless NIR-absorbing dyes for organic electronics, we focused on a molecular design based on controlling electronic transitions. To possess NIR absorption in the NIR region with suppressed π-π* transitions in the visible region, we designed and synthesized polycyclic D-A-D types of π-conjugated molecules (Py-FNTz-B, IP-FNTz-B, and IP-FNTz-B), which contain pyrrole or indenopyrrole as donor units and FNTz as an acceptor unit. Analyses of single-crystal X-ray diffraction for IP-FNTz and IP-FNTz-B were conducted, and their chemical structures revealed distinct bond length alternations similar to the polyene structure. Moreover, the introduction of a B-N linkage reduced the bond length alternation, which indicated effective delocalization of the π-electrons. The pyrrole end-capped Py-FNTz-B showed an absorption band in the NIR region at 888 nm without distinct absorption in the visible region, which led to colorless and transparent characteristics in the film state. The indenopyrrole end-capped IP-FNTz-B in films showed a narrow ΔE_opt of 0.87 eV, which is narrower by 0.30 eV than that of Py-FNTz-B. Transient photoconductivity in the TRMC measurements was observed in both compounds, indicating the possibility of semiconducting characteristics. Eventually, the OFET devices of Py-FNTz-B showed typical n-type characteristics and photodetection behaviors under NIR light irradiation at 810 nm. The diode device based on IP-FNTz-B showed an almost transparent film possessing electron transport characteristics. TD-DFT calculations showed that Py-FNTz-B has one allowed electronic transition in the NIR region and 2 forbidden electronic transitions in the visible region. From the analysis of orbital correlations, the orbital symmetries of Py-FNTz-B from HOMO−1 to LUMO+1 showed an alternating inverted configuration due to the contribution of the polyene-like structure and the presence of a strong FNTz unit, which led to NIR wavelength-selective absorption. This study demonstrates that orbital symmetries of the electronic structures in D-A-D systems are an important guideline for the molecular design of colorless NIR absorbing dyes, which could contribute to the development of colorless and transparent organic electronics with NIR-conducive characteristics.

Acknowledgements

This work was supported by JSPS KAKENHI (Grant Nos. 20H02814, 20H05836, 20H05841, 20KK0123, 20KAKK15352, 21K14602, 23K17947, 24H00482, and 24K08553), JST (Grant No. 22713577) and from The Ministry of Education, Culture, Sports, Science and Technology, Japan. This work was also supported by JST, JST-Mirai Program (Grant No. JPMJMI22I1), A-step (Grant No. JPMJSF23B3), CREST (Grant No. JPMJCR20R1), Japan, and the Mitsubishi Foundation (Grant No. 202310004). The authors are also thankful to the Comprehensive Analysis Center (CAC), SANKEN, for collecting the HRMS spectra.

Conflict of Interest

The authors declare no conflicts of interest.

Data Availability Statement

The data that support the findings of this study are available in the supplementary material of this article.

References

a) S. E. Root, S. Savagatrup, A. D. Printz, D. Rodriquez, D. J. Lipomi, Chem. Rev. 2017, 117, 6467;

b) A. Naibi Lakshminarayana, A. Ong, C. Chi, J. Mater. Chem. C 2018, 6, 3551;

c) J. Wang, X. Zhan, Acc. Chem. Res. 2021, 54, 132;

d) D.‐H. Lim, J.‐W. Ha, H. Choi, S. C. Yoon, B. R. Lee, S.‐J. Ko, Nanoscale Adv. 2021, 3, 4306;

e) Z.‐G. Zhang, Y. Li, Angew. Chem., Int. Ed. 2021, 60, 4422;

f) T. Zhang, Y. Xiao, H. Wang, S. Kong, R. Huang, V. K.‐M. Au, T. Yu, W. Huang, Angew. Chem., Int. Ed. 2023, 62, [eLocator: e202301896].

a) Y. He, Y. Cao, Y. Wang, Asian J. Org. Chem. 2018, 7, 2201;

b) Y. Ning, M. Zhu, J.‐L. Zhang, Coord. Chem. Rev. 2019, 399, [eLocator: 213028].

c) B. Li, M. Zhao, F. Zhang, ACS Mater. Lett. 2020, 2, 905;

d) V. K. Praveen, B. Vedhanarayanan, A. Mal, R. K. Mishra, A. Ajayaghosh, Acc. Chem. Res. 2020, 53, 496;

e) D. Meng, R. Zheng, Y. Zhao, E. Zhang, L. Dou, Y. Yang, Adv. Mater. 2022, 34, [eLocator: 2107330].

a) C.‐C. Chen, L. Dou, R. Zhu, C.‐H. Chung, T.‐B. Song, Y. B. Zheng, S. Hawks, G. Li, P. S. Weiss, Y. Yang, ACS Nano 2012, 6, 7185;

b) X. Shi, X. Liao, K. Gao, L. Zuo, J. Chen, J. Zhao, F. Liu, Y. Chen, A. K.‐Y. Jen, Adv. Funct. Mater. 2018, 28, [eLocator: 1802324];

c) S. Dai, T. Li, W. Wang, Y. Xiao, T. K. Lau, Z. Li, K. Liu, X. Lu, X. Zhan, Adv. Mater. 2018, 30, [eLocator: 1706571];

d) Y. Xie, R. Xia, T. Li, L. Ye, X. Zhan, H.‐L. Yip, Y. Sun, Small Methods 2019, 3, [eLocator: 1900424];

e) T. Li, L. Yang, Y. Xiao, K. Liu, J. Wang, X. Lu, X. Zhan, J. Mater. Chem. A 2019, 7, [eLocator: 20667];

f) J. Lee, H. Cha, H. Yao, J. Hou, Y.‐H. Suh, S. Jeong, K. Lee, J. R. Durrant, ACS Appl. Mater. Interfaces 2020, 12, [eLocator: 32764];

g) S. Kim, D. Lee, J. Lee, Y. Cho, S. H. Kang, W. Choi, J. H. Oh, C. Yang, Chem. Mater. 2021, 33, 7499.

h) P. Kafourou, M. I. Nugraha, A. Nikitaras, L. Tan, Y. Firdaus, F. Aniés, F. Eisner, B. Ding, J. Wenzel, M. Holicky, L. Tsetseris, T. D. Anthopoulos, M. Heeney, ACS Mater. Lett. 2022, 4, 165.

a) K. Kawabata, M. Saito, I. Osaka, K. Takimiya, J. Am. Chem. Soc. 2016, 138, 7725;

b) K. Yamamoto, Y. Ie, M. Nitani, N. Tohnai, F. Kakiuchi, K. Zhang, W. Pisula, K. Asadi, P. W. M. Blom, Y. Aso, J. Mater. Chem. C 2018, 6, 7493;

c) Z. Lin, L. Chen, Q. Xu, G. Shao, Z. Zeng, D. Wu, J. Xia, Org. Lett. 2020, 22, 2553;

d) W. Wang, L. Ge, G. Xue, F. Miao, P. Chen, H. Chen, Y. Lin, Y. Ni, J. Xiong, Y. Hu, J. Wu, Y. Zheng, Chem. Commun. 2020, 56, 1405;

e) R. Rausch, M. I. S. Röhr, D. Schmidt, I. Krummenacher, H. Braunschweig, F. Würthner, Chem. Sci. 2021, 12, 793;

f) K. Yamamoto, S. M. Quintero, S. Jinnai, E. Jeong, K. Matsuo, M. Suzuki, H. Yamada, J. Casado, Y. Ie, J. Mater. Chem. C 2022, 10, 4424.

a) H. Yamada, D. Kuzuhara, T. Takahashi, Y. Shimizu, K. Uota, T. Okujima, H. Uno, N. Ono, Org. Lett. 2008, 10, 2947;

b) N. Kobayashi, T. Furuyama, K. Satoh, J. Am. Chem. Soc. 2011, 133, [eLocator: 19642];

c) A. Abulikemu, Y. Sakagami, C. Heck, K. Kamada, H. Sotome, H. Miyasaka, D. Kuzuhara, H. Yamada, ACS Appl. Mater. Interfaces 2019, 11, [eLocator: 20812].

a) N. G. Pschirer, C. Kohl, F. Nolde, J. Qu, K. Müllen, Angew. Chem., Int. Ed. 2006, 45, 1401;

b) Y. Li, J. Gao, S. Di Motta, F. Negri, Z. Wang, J. Am. Chem. Soc. 2010, 132, 4208;

c) H. Langhals, D. Zgela, R. Lüling, J. Org. Chem. 2015, 80, [eLocator: 12146];

d) R. Fujita, S. Yoshida, H. Kano, K. Matsuo, H. Hayashi, H. Yamada, N. Aratani, Chin. J. Chem. 2023, 41, 1023.

a) G. M. Fischer, A. P. Ehlers, A. Zumbusch, E. Daltrozzo, Angew. Chem., Int. Ed. 2007, 46, 3750;

b) W. Naim, V. Novelli, I. Nikolinakos, N. Barbero, I. Dzeba, F. Grifoni, Y. Ren, T. Alnasser, A. Velardo, R. Borrelli, S. Haacke, S. M. Zakeeruddin, M. Graetzel, C. Barolo, F. Sauvage, JACS Au 2021, 1, 409;

c) W. Xu, E. Leary, S. Sangtarash, M. Jirasek, M. T. González, K. E. Christensen, L. Abellán Vicente, N. Agraït, S. J. Higgins, R. J. Nichols, C. J. Lambert, H. L. Anderson, J. Am. Chem. Soc. 2021, 143, [eLocator: 20472];

d) T. Baron, W. Naim, I. Nikolinakos, B. Andrin, Y. Pellegrin, D. Jacquemin, S. Haacke, F. Sauvage, F. Odobel, Angew. Chem., Int. Ed. 2022, 61, [eLocator: e202207459];

e) T. Baron, W. Naim, M. Kurucz, I. Nikolinakos, B. Andrin, Y. Pellegrin, D. Jacquemin, S. Haacke, F. Sauvage, F. Odobel, J. Mater. Chem. A 2023, 11, [eLocator: 16767].

a) A. Konishi, K. Horii, D. Shiomi, K. Sato, T. Takui, M. Yasuda, J. Am. Chem. Soc. 2019, 141, [eLocator: 10165];

b) J. Melidonie, E. Dmitrieva, K. Zhang, Y. Fu, A. A. Popov, W. Pisula, R. Berger, J. Liu, X. Feng, J. Org. Chem. 2020, 85, 215;

c) Y. Tanaka, N. Fukui, H. Shinokubo, Nat. Commun. 2020, 11, 3873;

d) M. Murai, T. Enoki, S. Yamaguchi, Angew. Chem., Int. Ed. 2023, [eLocator: e202311445].

a) S. L. Cai, Y.‐B. Zhang, A. B. Pun, B. He, J. Yang, F. M. Toma, I. D. Sharp, O. M. Yaghi, J. Fan, S. R. Zheng, W. G. Zhang, Y. Liu, Chem. Sci. 2014, 5, 4693;

b) Z. Mi, P. Yang, R. Wang, J. Unruangsri, W. Yang, C. Wang, J. Guo, J. Am. Chem. Soc. 2019, 141, [eLocator: 14433];

c) K. Okino, D. Sakamaki, S. Seki, ACS Mater. Lett. 2019, 1, 25.

a) M. Nakamura, H. Tahara, K. Tkahashi, T. Nagata, H. Uoyama, D. Kuzuhara, S. Mori, T. Okujima, H. Yamada, H. Uno, Org. Biomol. Chem. 2012, 10, 6840;

b) A. Wakamiya, T. Murakami, S. Yamaguchi, Chem. Sci. 2013, 4, 1002;

c) Y. Wu, C. Cheng, L. Jiao, C. Yu, S. Wang, Y. Wei, X. Mu, E. Hao, Org. Lett. 2014, 16, 748;

d) H. Yokoi, N. Wachi, S. Hiroto, H. Shinokubo, Chem. Commun. 2014, 50, 2715;

e) T. Nakamura, S. Furukawa, E. Nakamura, Chem. Asian J. 2016, 11, 2016;

f) W. Sheng, Y.‐Q. Zheng, Q. Wu, Y. Wu, C. Yu, L. Jiao, E. Hao, J. Y. Wang, J. Pei, Org. Lett. 2017, 19, 2893;

g) W. Sheng, Y. Wu, C. Yu, P. Bobadova‐Parvanova, E. Hao, L. Jiao, Org. Lett. 2018, 20, 2620;

h) H. Shimogawa, Y. Murata, A. Wakamiya, Org. Lett. 2018, 20, 5135;

i) Y. Min, C. Dou, D. Liu, H. Dong, J. Liu, J. Am. Chem. Soc. 2019, 141, [eLocator: 17015];

j) Q. Wu, G. Jia, B. Tang, X. Guo, H. Wu, C. Yu, E. Hao, L. Jiao, Org. Lett. 2020, 22, 9239;

k) Y. Min, X. Cao, H. Tian, J. Liu, L. Wang, Chem.‐Eur. J. 2021, 27, 2065;

l) J. Xu, J. Liu, L. Wang, Angew. Chem., Int. Ed. 2023, 62, [eLocator: e202303870];

m) L. Wang, Q. Wu, Z. Kang, X. Guo, W. Miao, Z. Li, H. Zuo, H. Wang, H. Si, L. Jiao, E. Hao, Org. Lett. 2023, 25, 5055;

n) W. Miao, R. Gan, W. Liu, X. Guo, Q. Wu, Y. Wei, C. Yu, E. Hao, L. Jiao, Chem.‐Eur. J. 2023, 29, [eLocator: e2023004].

a) S. Chatterjee, Y. Ie, T. Seo, T. Moriyama, G.‐J. A. H. Wetzelaer, P. W. M. Blom, Y. Aso, NPG Asia Mater 2018, 10, 1016;

b) S. Chatterjee, Y. Ie, Y. Aso, ACS Omega 2018, 3, 5814;

c) Y. Ie, K. Izuno, T. Moriyama, Y. Aso, J. Photopolym. Sci. Tech. 2019, 32, 721;

d) M. Saito, T. Fukuhara, S. Kamimura, H. Ichikawa, H. Yoshida, T. Koganezawa, Y. Ie, Y. Tamai, H. D. Kim, H. Ohkita, I. Osaka, Adv. Energy Mater. 2020, 10, [eLocator: 1903278];

e) S. Jinnai, A. Oi, T. Seo, T. Moriyama, R. Minami, S. Higashida, Y. Ie, Synthesis 2021, 53, 3390;

f) K. Iguchi, T. Mikie, M. Saito, K. Komeyama, T. Seo, Y. Ie, I. Osaka, Chem. Mater. 2021, 33, 2218.

R. Kishi, S. Ochi, S. Izumi, A. Makino, T. Nagami, J. Y. Fujiyoshi, N. Matsushita, M. Saito, M. Nakano, Chem.‐Eur. J. 2016, 22, 1493.

a) Z. Zeng, X. Shi, C. Chi, J. T. López Navarrete, J. Casado, J. Wu, Chem. Soc. Rev. 2015, 44, 6578;

b) Y. Li, Y. Li, L. Li, Y. Wu, J. Phys. Chem. C 2017, 121, 8579;

c) Z. Chen, W. Li, M. A. Sabuj, Y. Li, W. Zhu, M. Zeng, C. S. Sarap, M. M. Huda, X. Qiao, X. Peng, D. Ma, Y. Ma, N. Rai, F. Huang, Nat. Commun. 2021, 12, 1;

d) W. Zhu, Z. Chen, J. Huang, W. Liang, C. Liao, J. Wang, T. Du, Y. Deng, G. Li, R. Chen, X. Peng, J. Hou, Y. Li, J. Phys. Chem. C 2023, 127, 8894.

Deposition numbers 2308994 (for IP‐FNTz) and 2308995 (for IP‐FNTz‐B) contain the supplementary crystallographic data for this paper. These data are provided free of charge by the joint Cambridge Crystallographic Data Centre and Fachinformationszentrum Karlsruhe Access Structures service via www.ccdc.cam.ac.uk/data_request/cif.

a) J. L. Brédas, J. Chem. Phys. 1985, 82, 3808;

b) R. L. Gieseking, M. K. Ravva, V. Coropceanu, J. L. Brédas, J. Phys. Chem. C 2016, 120, 9975.

a) T. Kubo, A. Shimizu, M. Sakamoto, M. Uruichi, K. Yakushi, M. Nakano, D. Shiomi, K. Sato, T. Takui, Y. Morita, K. Nakasuji, Angew. Chem., Int. Ed. 2005, 44, 6564;

b) A. Shimizu, M. Uruichi, K. Yakushi, H. Matsuzaki, H. Okamoto, M. Nakano, Y. Hirao, K. Matsumoto, H. Kurata, T. Kubo, Angew. Chem., Int. Ed. 2009, 48, 5482.

a) A. Saeki, Y. Koizumi, T. Aida, S. Seki, Acc. Chem. Res. 2012, 45, 1193;

b) A. Saeki, Polym. J. 2020, 52, 1307.

M. R. Koli, A. Labiod, S. Chakraborty, M. Kumar, P. Lévêque, G. Ulrich, N. Leclerc, D. Jacquemin, S. Mula, ChemPhotoChem 2020, 4, 729.

N. Li, Z. Lan, L. Cai, F. Zhu, J. Mater. Chem. C 2019, 7, 3711.

P. C. Y. Chow, N. Matsuhisa, P. Zalar, M. Koizumi, T. Yokota, T. Someya, Nat. Commun. 2018, 9, 4546.

D. C. Harris, M. D. Bertolucci, Symmetry and Spectroscopy: an Introduction to Vibrational and Electronic Spectroscopy, Dover Books on Chemistry, New York, NY, USA 1989.

O. Laporte, W. F. Meggers, J. Opt. Soc. Am. 1925, 11, 459.