Orthogonal colloidal quantum dot inks enable efficient multilayer optoelectronic devices

The study addresses a key bottleneck in multilayer colloidal quantum dot (CQD) device fabrication: conventional high-quality CQD inks require highly polar solvents, which redissolve and damage underlying CQD layers, preventing multifunctional multilayer stacks. Layer-by-layer solid-state ligand exchange has been used but is incompatible with scalable manufacturing and introduces energetic disorder and film cracking. The authors hypothesize that using aromatic ligands with appropriately chosen anchoring groups can induce ligand dipoles that strengthen interactions with weakly polar solvents, stabilizing CQDs without bulky insulating ligands. They propose benzoic acid (BA) as a ligand to produce weakly polar, orthogonal CQD inks that enable smooth, crack-free hole transport layers (HTLs) and preserve underlying light-absorbing layers, targeting improved infrared photovoltaics and photodetectors. They further combine experimental and computational analysis to elucidate the role of ligand dipoles in ink solubility and device performance, aiming for efficient fully ink-processed CQD devices with high EQE in the IR.

Prior CQD films used long insulating ligands (e.g., oleic acid), necessitating solid-state ligand exchange to shorter ligands for carrier transport. Early layer-by-layer methods led to inhomogeneous energy states and are not scalable. Solution-phase exchanges using lead halides, acetate salts, and thiols produced high-quality CQD inks but relied on highly polar solvents that redissolve underlying CQD layers, forcing a final solid-state exchange step for HTLs. Attempts to stabilize CQDs in weakly polar solvents used aromatic thiols like 4-methylbenzenethiol, but required bulky pendent groups and yielded limited charge transport and low PCE (≈1.34%). Cracking and trap formation associated with solid-state EDT exchanges are known issues. This work builds on these observations by seeking short aromatic ligands with anchoring groups that polarize the ring, enhancing CQD–solvent interactions in weakly polar media and enabling orthogonal processing.

• Ligand and ink design: OA-capped CQDs were subjected to solution-phase ligand exchange with aromatic ligands: benzoic acid (BA), 4-methylbenzoic acid (4-CH3-BA), benzenethiol (BT), and 4-methylbenzenethiol (4-CH3-BT). Ligands were dissolved (0.3 M) in appropriate solvents; for BT-type ligands, triethylamine was added to deprotonate. Ligand solution (1 mL) was added dropwise to CQDs in toluene (1 mL, 20 mg mL−1), stirred 10 min, then exchanged CQDs were precipitated with hexane, centrifuged, and purified by redispersion in chlorobenzene and reprecipitation with hexane (three cycles). BA-exchanged CQDs were dispersed in weakly polar solvents (chlorobenzene, toluene, dichloromethane, chloroform, anisole). FTIR confirmed removal of OA (loss of alkane/alkene C–H stretches).
• Extension to other CQDs: Optimized exchange conditions achieved BA-exchanged InP (core/shell) and InAs CQDs stably dispersed in chlorobenzene.
• Computational analysis: DFT (CP2K Quickstep; GGA-PBE, GTH pseudopotentials, D3 dispersion, DZVP basis, 600 Ry grid cutoff) on ~1 nm PbS QD models capped with ligands; four chlorobenzene molecules placed in ligand shell, fully relaxed; solvent interaction energies averaged. Electrostatic potentials mapped for ligand-capped QDs and isolated ligands to assess dipole-induced polarization.
• Energy level and optical characterization: UPS to determine EF and VBM; UV–vis absorption and steady-state/time-resolved PL to probe band-edge and band-tail states and exciton dynamics.
• Electrical characterization: Hole-only devices (ITO/NiOx/PbS CQDs/MoO3/Au) measured in dark; SCLC analysis yielded hole mobility; trap-filled limit gave trap state densities.
• Morphology: SEM and AFM to evaluate film continuity and roughness.
• Orthogonality tests: Underlying halide-passivated PbS CQD light-absorbing layers (in butylamine:DMF 4:1) were subjected to chlorobenzene spin-coating (BA HTL protocol) versus ACN soaking/washing (EDT protocol). Effects were probed by absorption/PL, logarithmic band-tail absorption, and XPS (iodide:Pb ratios).
• Device fabrication and testing: PV architecture ITO/ZnO/IR-absorbing PbS CQDs (Eg ≈ 1.0 eV)/HTL (Eg ≈ 1.3 eV)/Au. HTLs: BA- or 4-CH3-BA-functionalized CQD inks (single-step spin-coating) vs EDT via solid-state exchange (OA-capped CQDs, soak in 0.01 vol% EDT/ACN, ACN washes; two layers to mitigate cracking). J–V under AM 1.5 with 1100 nm Si cutoff filter; EQE beyond 1100 nm; HDR EQE for band-tail DOS. Operational stability at MPP under 1 sun, N2.
• Tandem characterization: 4T perovskite/CQD tandems with semitransparent perovskite front cell (1.63 eV) stacked over BA-based CQD back cell; combined J–V and EQE.
• Photodetectors: Dark J–V, noise equivalent power, specific detectivity D* at 1 kHz, zero bias; TPC with femtosecond excitation.

• Ligand dipole controls solubility in weakly polar solvents: DFT showed higher chlorobenzene interaction energies for BA- vs BT-functionalized CQDs; para-methyl enhances interaction for both, but BA without methyl still surpasses BT, consistent with stronger polarization by the carboxyl anchor. Electrostatic potential maps revealed larger positive potential pockets for BA-capped CQDs.
• BA-exchanged CQDs form stable inks in weakly polar solvents (chlorobenzene, toluene, dichloromethane, chloroform, anisole) without pendent methyl groups; remained colloidally stable for at least 6 months in air. Strategy generalized to InP and InAs CQDs.
• BA-based HTLs yield favorable energy alignment: UPS/UV–vis showed similar energy levels among BA, 4-CH3-BA, and EDT HTLs; VBMs (~4.86–4.94 eV) shallower than the active layer VBM (5.2 eV) for efficient hole extraction; CBs (~3.67–3.74 eV) block electrons from the active layer (~4.2 eV).
• Reduced trap states and improved film morphology: HOD SCLC mobilities were ~1×10−4 (BA), 7×10−5 (4-CH3-BA), and 2×10−4 cm2 V−1 s−1 (EDT). Trap densities: 5×10^16 cm−3 for BA and 4-CH3-BA vs 2×10^17 cm−3 for EDT (≈3.5× higher in EDT). BA-ink films were smooth and crack-free (RMS <1 nm), while EDT films exhibited roughness and cracks.
• Orthogonality preserves underlying layers: Chlorobenzene spin-coating (BA) caused negligible changes in absorption/PL and band-tail absorption. ACN soaking (EDT protocol) induced red-shifted, broadened, and quenched PL with increased band-tail absorption; XPS showed iodide:Pb ratio drop from 53% to 47%, indicating ligand loss and underpassivation.
• Device performance (IR PVs with 1100-nm Si cutoff): BA-based devices achieved EQE of 84% at 1210 nm, higher VOC than EDT, and best Si-filtered PCE of 1.43% (VOC 0.438 V, JSC 5.34 mA cm−2, FF 61.3%). Statistics (n=20 per HTL) showed BA: 1.40 ± 0.02% vs 1.28 ± 0.02% for both 4-CH3-BA and EDT. Operational stability: BA devices retained >80% initial PCE after 60 h at MPP under 1 sun (N2), outperforming EDT.
• Tandems: 4T perovskite (1.63 eV) front + BA-based CQD back delivered additional JSC of 12.6 mA cm−2 and added PCE of 3.57% from the CQD rear cell; combined JSC 34.7 mA cm−2 and overall PCE 22%.
• Photodetectors: BA-based devices showed lower dark current and NEP; higher peak D* of 1.4×10^12 Jones at 1230 nm vs 1.0×10^12 Jones (EDT), attributed to high EQE and reduced noise.

The results validate the hypothesis that engineering ligand dipoles via aromatic carboxylate anchoring enhances CQD–solvent interactions in weakly polar media, enabling stable, short-ligand CQD inks without bulky insulating ligands. BA-functionalized CQDs yield orthogonal processing: HTLs can be deposited from chlorobenzene without redissolving or depassivating underlying halide-passivated CQD layers, unlike ACN-based solid-state EDT exchanges that remove iodide, introduce traps, and degrade optoelectronic quality. This orthogonality and improved film quality reduce trap density, mitigate morphological defects, and enhance device metrics including EQE, VOC, stability, and detectivity. The approach generalizes across CQD chemistries (PbS, InP, InAs), indicating broader applicability to multilayer CQD optoelectronics and tandem photovoltaics. The combined experimental and DFT insights establish ligand dipole strength as a predictive parameter for ink solubility and process compatibility.

Aromatic ligand engineering—specifically using benzoic acid anchors—enables weakly polar, process-orthogonal CQD inks that form smooth, crack-free HTLs with low trap densities. Orthogonal deposition preserves the integrity of underlying CQD light-absorbing layers. Devices fabricated entirely from inks attain high EQE (84% at 1210 nm), improved stability, and strong tandem performance, while photodetectors exhibit enhanced detectivity. The findings identify ligand dipole-induced CQD–solvent interactions as central to achieving stable nonpolar inks and pave the way for efficient multilayer CQD optoelectronic devices across PVs and detectors.