High-performance near-infrared OLEDs maximized at 925 nm and 1022 nm through interfacial energy transfer

The study targets the longstanding challenge of low emission efficiency in organic near-infrared (NIR) fluorescent dyes, particularly beyond 900 nm, where photoluminescence quantum yields (PLQY) are typically very low due to the emission energy gap law (EEGL). As the emission energy decreases, exciton–vibration coupling increases, promoting non-radiative decay. A strategy to mitigate EEGL is to reduce the effective reorganization energy through orderly molecular assembly and exciton delocalization. The authors leverage self-assembling square-planar Pt(II) complexes that exhibit strong metal–metal-to-ligand charge transfer (MMLCT) to both red-shift emission into the NIR and suppress exciton–vibration coupling. The research question is whether an interfacial energy transfer architecture—using a self-assembled phosphorescent Pt(II) donor and a NIR fluorescent acceptor (BTP-eC9)—combined with transfer printing (stamping) to preserve morphology, can produce highly efficient fluorescence NIR OLEDs (>900 nm) with substantially improved EQE and radiance. The work also explores the generality of the approach to NIR(II) emission (~1000 nm).

Previous NIR fluorescence OLEDs beyond 900 nm have achieved limited efficiencies, with recent V-shaped A-DA-D-A dyes (face-on packing) reaching EQEs of 0.15–0.33% (Xie et al.). While NIR dyes can show aggregation-enhanced emission in bioimaging (low background beyond 900 nm), PLQYs in solids often remain <<1%, insufficient for OLEDs. Outcoupling considerations set theoretical EQE limits, and advances in QD-LEDs and PeLEDs have exceeded 25% EQE without outcoupling enhancements. Earlier work by the authors and others established highly efficient NIR phosphorescent OLEDs using self-assembled Pt(II) complexes with MMLCT, achieving EQE up to 4.2% and beyond. However, translating such efficiency to fluorescence hyperfluorescent NIR OLEDs requires efficient energy transfer from triplet or TADF donors to NIR dyes and preserving self-assembled structures, which is difficult with co-deposition or direct spin-coating. Transfer printing has been used in OPV and some OLED contexts mainly for transport layers, but not typically for emissive interlayers. This study builds on these foundations to implement interfacial FRET from Pt(II) donors to NIR fluorescent acceptors.

• Materials and design: Pt(fprpz)2 chosen as donor for high PLQY (>80%), chemical robustness, insolubility, and reduced bandgap (~2.16 eV) facilitating overlap with NIR dye absorption. BTP-eC9 selected as acceptor due to strong absorption overlap with Pt(fprpz)2 emission and favorable Y-family optoelectronic properties and type-I band alignment that mitigates back-transfer and charge transfer.
• Interfacial energy transfer concept: Leverages strong spin–orbit coupling in Pt(fprpz)2 for fast ISC (kISC > 10^12 s−1) to populate T1 and allow radiative T1→S0 phosphorescence with large transition dipole; enables FRET from donor T1 emission to acceptor S1 absorption across the interface within Förster radius.
• Film formation approach: Two methods compared: (i) spin-coating BTP-eC9 from CHCl3 on vapor-deposited Pt(fprpz)2; (ii) stamping (transfer printing) of preformed solid BTP-eC9 from PDMS onto Pt(fprpz)2 to avoid solvent-induced disruption and preserve self-assembly.
• Optical characterization: Time-resolved photoluminescence (TrPL) with fs excitation (505 nm) to probe donor lifetime quenching and acceptor lifetime changes; PLQY measurements; steady-state spectra; determination of FRET efficiency and pre-exponential factors (a1, a2) under front/rear excitation geometries. Förster radius R0 estimated (7.8 nm).
• Morphology and interfacial analyses: PL mapping, AFM, KPFM (surface potentials), GIWAXS for molecular packing and crystallinity, angle-dependent NEXAFS for interfacial electronic structure and quadrupole indications.
• Carrier dynamics: Picosecond transient absorption spectroscopy (ps-TAS) to monitor charge-transfer (CT) and charge separation (CS) signatures (780 nm positive ΔOD) and compare spin-coated vs stamped interfaces; qualitative correlation to interfacial density and smoothness.
• Electrical/device measurements: Space-charge-limited current (SCLC) devices to extract mobilities of Pt(fprpz)2 and BTP-eC9 (hole/electron). OLED fabrication of single-layer emissive devices, bilayer stamped devices with varying Pt(fprpz)2 thickness (10 nm vs 5 nm), and a sandwiched structure Pt/BTP-eC9/Pt to confine recombination near interfaces. Simulation with Setfos to model recombination zone positioning relative to interface and optimize thickness.
• Device fabrication details: • Pt(fprpz)2 OLED: ITO/HATCN (10 nm)/NPB (50 nm)/mCP (15 nm)/Pt(fprpz)2 (10 nm)/TPBi (55 nm)/LiF (1 nm)/Al (120 nm), deposited under ~1×10−6 Torr. • BTP-eC9 OLED: ITO/PEDOT:PSS (15 nm)/PVK (30 nm)/BTP-eC9 or Y11 (~100 nm)/C60 (5 nm)/BCP (2 nm)/Ag (120 nm), spin-coating under N2 glovebox. • Stamped bilayer and sandwiched OLEDs: ITO/HATCN (10 nm)/NPB (50 nm)/mCP (15 nm)/Pt(fprpz)2 (5–10 nm)/[stamped BTP-eC9 (~35 nm)]/optional Pt(fprpz)2 (5 nm)/PO-T2T (30 nm)/LiF (1.5 nm)/Al (150 nm). PDMS stamps prepared (10:1 base:curing agent, degassed, baked 90 °C 90 min; UV-ozone treated for wettability) and used to transfer BTP-eC9 films. • Additional NIR(II) case: BTPV-eC9 with Pt(II) No. 2 donor: ITO/HATCN (10 nm)/NPB (50 nm)/mCP (15 nm)/Pt(II) No. 2 (4 nm)/BTPV-eC9/PO-T2T (25 nm)/LiF (1 nm)/Al (120 nm).
• Stability testing: Un-encapsulated sandwiched devices measured at 25 °C, 50±5% RH; initial radiance 11.86 W sr−1 m−2 at 13 V, T50 ~103 min.

• Efficient interfacial energy transfer via stamping: • Pure Pt(fprpz)2 film: emission ~750 nm, lifetime 322.0 ± 1.2 ns, PLQY ~80%. • Pure BTP-eC9 film: emission >900 nm (940–952 nm range), lifetime 0.65 ± 0.04 ns at 1000 nm, PLQY 5.57% (5.42 ± 0.17%). • Spin-coating BTP-eC9 onto Pt(fprpz)2 destroyed donor morphology: Pt emission blue-shifted 750→735 nm; PLQY dropped 78.6% → 0.92%; no effective energy transfer observed; TrPL at 750 nm reduced to ~50 ns due to morphology quench, not transfer. • Stamped bilayers preserved order and enabled energy transfer: Pt(fprpz)2 lifetime reduced 322.0 → 100.5 ns, implying 68.8% transfer component; pre-exponential a1 (transfer fraction) front 0.76 ± 0.04 vs rear 0.11 ± 0.02, localizing transfer at the interface. BTP-eC9 lifetime increased to 0.87 ± 0.02 ns and PLQY to 8.85% (8.61 ± 0.21%), indicating improved packing (shorter d-spacing, better face-on orientation). • Förster radius R0 estimated at 7.8 nm; simulations showed recombination near the interface enhances transfer; larger distances reduce transfer and leave residual donor emission.
• Morphology/interface evidence: • PL mapping and AFM: spin-coated films show comet-like tails and erosion; stamped films maintain granular morphology. • KPFM surface potentials: Pt(fprpz)2 1.03 eV, BTP-eC9 0.77 eV, stamped 0.98 eV, spin-coated 1.17 eV (indicating multiple interfaces/defects and charge separation in spin-coated). • GIWAXS: Pt(fprpz)2 exhibits ordered edge-on π–π stacking; BTP-eC9 lamellar face-on; spin-coating deteriorates Pt crystallinity; stamped maintains characteristic peaks. NEXAFS edge-on peak blue-shifted by 0.11 eV in stamped film, suggesting interfacial T-shaped dimers and enhanced quadrupole moments. • ps-TAS: 780 nm CS signal absent in pure BTP-eC9; appears at 85 ps and stronger for spin-coated Pt/BTP-eC9 indicating many interfaces; delayed to 355 ps and weaker in stamped films, indicating fewer CS-prone interfaces.
• Device performance (EL): • Pure Pt(fprpz)2 (10 nm active emitter device): V_on 4.2 V, EL 720 nm, radiance 51.95 W sr−1 m−2, EQE 11.14%. • Pure BTP-eC9 device (~100 nm): V_on 1.2–1.3 V, EL 952 nm, radiance 18.81 W sr−1 m−2, EQE 0.18% (0.14 ± 0.04). • Stamped Pt(fprpz)2 (10 nm)/BTP-eC9: V_on 6.0 V, EL 918–920 nm dominant with 11.2% residual Pt emission at ~692 nm in EL spectrum, radiance 31.73 W sr−1 m−2, EQE 2.00% (1.80 ± 0.14%). • Reducing Pt thickness to 5 nm: V_on 6.0 V, residual Pt emission reduced to 4.3%, radiance 34.14 W sr−1 m−2, EQE 2.07% (1.83 ± 0.15%). • Sandwiched Pt(fprpz)2 (5 nm)/BTP-eC9/Pt(fprpz)2 (5 nm): V_on 6.2 V, peak EL 925 nm (96.1% at 925 nm; ~3.9% residual donor emission at ~682 nm), maximum radiance 39.97 W sr−1 m−2, EQE 2.24% (1.94 ± 0.18%). Horizontal dipole ratio O for BTP-eC9 emission ~67% (unchanged by transfer), Pt(fprpz)2 ~78%. • Stability: Unencapsulated sandwiched device at 13 V, initial radiance 11.86 W sr−1 m−2; T50 ~103 min in ambient (25 °C, RH 50±5%).
• Generality and NIR(II) extension: • Y11 dye: despite higher single-layer EQE and radiance vs BTP-eC9, its LUMO (−3.87 eV) nearly aligns with Pt(fprpz)2 (−3.85 eV), leading to incomplete charge dispersion and inferior sandwiched-device performance relative to BTP-eC9. • BTPV-eC9 (extended conjugation): neat-film PLQY 2.53% (2.45 ± 0.08%); with Pt(II) No. 2 donor (800 nm emission), double-layer device achieved EL 1022 nm with EQE 0.66% (0.55 ± 0.10%) and maximum radiance 18.67 W sr−1 cm−2; single-emitter device EQE 0.08% (0.06 ± 0.02%) and radiance 9.69 W sr−1 cm−2.

The results confirm that preserving the self-assembled order of the Pt(II) donor and NIR dye acceptor is crucial for enabling efficient interfacial FRET and suppressing deleterious charge separation. Stamping allows the formation of a smooth, planar interface with minimal defects, maintaining donor crystallinity and improving acceptor packing. Consequently, energy transfer from the donor triplet emission to the acceptor singlet increases NIR hyperfluorescence beyond 900 nm, markedly boosting EQE and radiance compared to prior fluorescence NIR OLEDs. Simulations and experiments show that positioning the recombination zone within the Förster distance of the interface (by minimizing donor thickness and adopting a sandwiched architecture) maximizes transfer probability and minimizes residual donor emission. The study also identifies critical design criteria: strong spectral overlap and high donor PLQY with high acceptor absorption; sufficient donor–acceptor energy-level offsets to avoid even charge distribution and mitigate back-transfer or unwanted electron transfer; and device architectures that confine recombination near the interface. Extension to BTPV-eC9 at 1022 nm validates the generality of the interfacial energy transfer concept, while the Y11 case underscores the importance of energy-level alignment. Overall, the findings directly address the initial hypothesis by demonstrating record EQE and radiance for fluorescence NIR OLEDs via interfacial energy transfer.

This work demonstrates a transfer-printed interfacial energy transfer strategy using self-assembled Pt(fprpz)2 as a phosphorescent donor and BTP-eC9 as a fluorescent acceptor to realize high-performance fluorescence NIR OLEDs. By employing a stamped bilayer and an optimized sandwiched architecture, the devices achieve a peak at 925 nm with EQE 2.24% (1.94 ± 0.18%) and maximum radiance 39.97 W sr−1 m−2, setting a new benchmark for fluorescence NIR OLEDs >900 nm. Morphological (GIWAXS, AFM), electronic (KPFM, NEXAFS), and dynamical (TrPL, ps-TAS) analyses corroborate efficient interfacial FRET and minimized charge separation. The approach generalizes to NIR(II) emission (1022 nm) using BTPV-eC9 with an appropriate Pt(II) donor. Future work could focus on: optimizing donor/acceptor energy-level offsets and interfacial dipole engineering; enhancing outcoupling; improving stability and encapsulation; exploring broader dye libraries via the sandwiched platform; and conducting electrical-pumping ultrafast spectroscopies to further elucidate recombination and transfer dynamics.

• Time-resolved interfacial dynamics under electrical excitation could not be directly measured; ps-TAS relied on optical pumping, offering only indirect insights into OLED-relevant interfaces.
• Residual donor emission persists unless donor thickness and recombination zone are carefully optimized; thicker layers shift recombination away from the interface, reducing transfer efficiency.
• Device stability for unencapsulated sandwiched structures shows a T50 of ~103 min at 13 V and 11.86 W sr−1 m−2 in ambient; longevity under practical operating conditions requires encapsulation and further engineering.
• While EQE is record-setting for fluorescence NIR OLEDs beyond 900 nm, absolute values remain modest compared to visible OLEDs; no outcoupling enhancements were implemented.
• Generalization to NIR(II) (BTPV-eC9) yields lower efficiencies (EQE 0.66%), indicating further material-level optimization is needed; energy-level proximity (e.g., Y11 LUMO near donor LUMO) can impair charge distribution and performance.