Inverted perovskite solar cells lose their efficiency and their lifetime in the same place: the buried interface, where the hole transport layer meets the substrate on one side and the perovskite on the other. It is built first, inspected least, and asked to do the most — bond to the electrode, set the energy level for hole extraction, and sit against a defect-riddled perovskite surface, all while the rest of the device is coated on top of it. Most transport layers are designed to do one of those jobs well. A paper published November 5 in ACS Applied Materials & Interfaces reports a molecule designed to do two of them at once: TPCOOH, a hole transport layer whose dicarboxylic acid groups anchor it to the indium tin oxide (ITO) substrate while simultaneously passivating defects in the perovskite above. Inverted cells built on it reach 24.68% power conversion efficiency, against 23.72% for a control molecule that differs only in the anchoring group, and retain 75% of their initial efficiency after 1,200 hours of thermal aging, against 65% for the control.

The efficiency is solid but not a record — this desk covered a 27.4% inverted cell only two weeks ago — and that is precisely the point. The value is the design rule: one molecule, a named chemical mechanism for each of two interface jobs, and a control experiment that isolates the anchoring group’s contribution. That is the kind of interface-design evidence the contact-engineering arc of this desk keeps rewarding, because it is the kind that transfers.

The buried interface, and why it costs both efficiency and lifetime

An inverted, or p–i–n, cell is built in the opposite order to a conventional cell: the hole transport layer goes down first, on the transparent electrode, and the perovskite is coated on top of it. The interface between the two is therefore a buried interface — formed early, sealed in by everything deposited after it, and effectively inaccessible to any post-treatment once the device is built. Whatever defects exist at that junction, and whatever the quality of the contact, are locked in for the life of the cell.

The abstract states the stakes plainly: the hole transport layer “plays a pivotal role in the performance of inverted perovskite solar cells, especially in carrier transport, anchoring to the substrate as a buried interface, and defect passivation.” Three jobs, named in one sentence — and the paper’s opening claim is that “novel HTLs that simultaneously achieve indium tin oxide (ITO) substrate anchoring, passivation of buried perovskite defects, and favorable interfacial energy alignment are seldom addressed.” Most molecules are optimized for one of the three.

Each failure mode costs something different, and the costs compound. Poor anchoring means poor film coverage — thin spots and pinholes where the perovskite meets bare electrode or where charge has no clean path out. Poor energy alignment means holes linger at the interface instead of crossing promptly, and lingering carriers recombine. Unpassivated defects mean trap states at the buried perovskite surface — undercoordinated lead ions and halide vacancies that capture carriers and let them recombine without producing current, eating voltage. That is why the buried interface is where inverted cells lose both efficiency and lifetime at once — and why a weakly bonded one is also a delamination risk under thermal cycling.

The molecule: a dicarboxylic acid core, a methylthio triphenylamine periphery

TPCOOH has three named parts, each mapping onto one of the three jobs. The central phenyl ring carries dicarboxylic acid groups — two –COOH groups on the same core. The periphery couples methylthio-substituted triphenylamine and thiophene units. The abstract describes the division of labor directly: the dicarboxylic acid groups “can anchor onto ITO substrates and passivate the top perovskite,” while the methylthio-substituted triphenylamine and thiophene units “generate suitable energy alignment and transfer the hole.”

The anchoring chemistry is the familiar one from the self-assembled monolayer playbook this desk has covered before. A carboxylic acid group binds to the hydroxylated surface of a metal oxide — here, ITO — through coordination of the carboxylate oxygen to surface metal sites, often with hydrogen bonding as well. It is the same anchoring logic that phosphonic acid groups perform in MeO-2PACz, the SAM this desk covered last month for carbon-based cells. What is different here is that the anchoring group is doing double duty: the abstract attributes the passivation of perovskite defects to the same carboxyl groups, which coordinate to undercoordinated lead sites on the perovskite surface — the same Lewis acid–base interaction behind the TPAC passivator (21.9% against 19.7% for an untreated control).

The control molecule is what makes the claim testable rather than merely plausible. TPHex is the same scaffold with dihexyl substitution in the center — alkyl chains where TPCOOH carries the dicarboxylic acid groups, chemically inert at this interface: they cannot coordinate to ITO and cannot passivate lead defects. Everything else — the triphenylamine arms, the thiophene units, the energy levels — is held as close to constant as a single substitution allows. So the difference between the two devices is, to a first approximation, the difference the carboxyl groups make. That is a clean experiment, and it is the reason this paper reads as a design rule rather than a one-off efficiency report.

The methylthio substitution on the triphenylamine does the third job. Methylthio groups — –SCH₃ — are electron-donating, tuning the triphenylamine’s energy level, and sulfur’s lone pair can coordinate weakly to lead as well. The abstract credits these units with generating “suitable energy alignment and transfer the hole” — getting the hole out of the perovskite at minimal energy cost.

The device numbers, handled in order

The numbers that can be stated from the paper’s abstract are these: inverted perovskite devices built on TPCOOH and TPHex as hole transport layers, at 0.1 cm² area, achieved power conversion efficiencies of 24.68% and 23.72%, respectively. After 1,200 hours of thermal aging at 65 ± 5 °C in a nitrogen atmosphere, the TPCOOH and TPHex devices retained 75% and 65% of their initial efficiencies.

The 0.96-percentage-point gap is the headline, and it should be read as the size of the carboxyl effect on this stack, not as a claim about the molecule’s absolute ceiling. A near-one-point gain from changing only the central substituent is meaningful for a single functional-group swap, and the rest of the stack is identical, so the improvement is attributable to the interface they share.

The abstract attributes the improvement to three mechanisms: “suppressed nonradiative recombination, rapid interfacial charge transfer, and minimal energy loss due to well-matched energy levels.” Each maps onto one of the three jobs — passivation means fewer trap states, anchoring means better film coverage and a better-bonded contact, matched levels mean little voltage sacrificed in extraction. The abstract adds the causal chain explicitly: the carboxyl groups “anchor to ITO, thus ensuring better film coverage and improved charge transport, while the passivation functionality simultaneously mitigates perovskite defects.”

The stability numbers need the same careful handling. A 75% retention after 1,200 hours at 65 °C is not a commercial stability result — the field’s benchmark for meaningful durability is typically retention above 80% or 90% under standardized stress, and 75% means the device has lost a quarter of its output. What matters is the comparison: the control retained 65% under identical aging, so the carboxyl groups buy ten points of retention on the same stack. The test conditions carry their own caveats: thermal aging in nitrogen at 65 ± 5 °C is a storage test, not an operating test — the devices were not held at maximum power under illumination, and there is no humidity in a nitrogen atmosphere, so nothing here speaks to the moisture-driven degradation that has historically been perovskites’ hardest failure mode.

The area is the other honesty constraint, and it is stated plainly in the abstract: 0.1 cm². This is a small laboratory cell, not a module, and every number in this article belongs to that scale. Whether the anchoring holds at module scale — uniform coating over a square meter rather than a tenth of a square centimeter, by slot-die or blade rather than spin coating — is entirely untested in the material consulted. The abstract gives no deposition method, no large-area data, and no yield figures. The anchoring advantage is a film-quality claim, and film quality is the property that most often fails to scale.

Why one molecule doing two jobs matters for manufacturing

The reason to care about a 24.68% cell that is not a record is the process logic underneath it. Perovskite manufacturing, insofar as it has a consensus direction, is a story of fewer steps, fewer layers, and fewer materials — every additional deposition step is a yield risk, a uniformity risk, and a cost. A hole transport layer that also passivates is one fewer treatment step than a transport layer plus a separate passivation step, and one that anchors itself firmly is one fewer adhesion-promotion step. The paper’s framing — that HTLs combining anchoring, passivation, and alignment “are seldom addressed” — is a manufacturing claim disguised as a chemistry claim: the field has been solving the three interface jobs with three separate interventions, and each intervention is a step on a production line.

That logic is the same one this desk has tracked through October: the MeO-2PACz post-treatment that took a SAM from the record-chasers’ playbook into printable carbon cells, and the co-deposition work from Nature Energy in which phosphonic and carboxylic acids self-assemble on ITO while the perovskite crystallizes above them. Each is, at bottom, a step-count argument: do more of the interface chemistry in fewer depositions, with molecules whose mechanism you can name. TPCOOH belongs in that lineage: a single deposition that anchors, passivates, and aligns, with a control experiment proving the anchoring and passivation are doing work rather than decorating a molecule that would have worked anyway.

What this is not: a record, a module, or a weather test

The limits of the result are specific, and they should travel with the headline wherever it goes.

First, 24.68% is a solid but not record inverted-cell figure. The Northwestern–DGIST orthogonal-passivation cell this desk covered on October 27 reached 27.4%, with a certified quasi-steady-state 25.8%, on a 0.049 cm² aperture. The TPCOOH result sits nearly three points below that champion on a comparable laboratory scale. The two results optimize different things: one is a defect-chemistry record, the other is a multifunctional interface design. The comparison that matters for TPCOOH is TPCOOH against TPHex — 24.68% against 23.72% — because that is the comparison that isolates the design rule.

Second, the demonstration is a 0.1 cm² laboratory cell, with no module data, no large-area coating data, and no yield figures. The anchoring advantage is a film-coverage claim, and film coverage is the property most sensitive to scale-up. Until a slot-die or blade-coated version exists, the manufacturing case rests on the form of the fix — one deposition, one molecule — not on manufacturing evidence.

Third, the stability evidence is a single-axis storage test: 1,200 hours at 65 ± 5 °C in nitrogen, encapsulation status unstated, with no illumination, no maximum-power-point tracking, and no humidity. It is not a damp-heat test at 85 °C and 85% relative humidity, not a thermal-cycling test, and not an outdoor test. The ten-point retention advantage is real and mechanistically consistent, but 75% retention is a long way from a commercial durability claim, and the test does not address the moisture axis at all.

Fourth, the mechanism evidence available here is abstract-level. The abstract reports the efficiency attribution — suppressed nonradiative recombination, rapid charge transfer, matched energy levels — as conclusions, not as shown data. The supporting measurements that would corroborate each half of the claim sit behind the paywall, in the full text and supporting information. Treat the mechanism as the authors’ reported interpretation, not as independently inspected evidence. The same applies to the details the abstract does not state at all: the perovskite composition, the device stack, the deposition method, the open-circuit voltage, short-circuit current density, and fill factor, and the research team’s identity and affiliation. None of those appear in the abstract, and none is claimed here.

What to watch

The most informative follow-up would be a second dicarboxylic acid molecule on the same architecture. If TPCOOH’s advantage is genuinely the carboxyl chemistry rather than a property of this one scaffold, then other dicarboxylic-acid-functionalized transport layers should produce the same directional improvement, with the magnitude tracking their binding strength and energy levels. A second molecule would convert a demonstration into a design rule.

A damp-heat result would turn the stability question from a storage claim into a commercial one. A TPCOOH device held at 85 °C and 85% relative humidity, tracked at maximum power, would answer the moisture question the nitrogen storage test leaves open — and whether the carboxyl anchoring survives the condition that has historically broken perovskite interfaces first.

The bottom line

A team reporting in ACS Applied Materials & Interfaces on November 5 has designed TPCOOH, a hole transport layer whose dicarboxylic acid groups anchor it to the ITO substrate and passivate buried perovskite defects at the same time, with methylthio-substituted triphenylamine and thiophene units setting the energy alignment and transferring the hole. Inverted cells on TPCOOH reach 24.68% at 0.1 cm², against 23.72% for the dihexyl-substituted control TPHex, and retain 75% of initial efficiency after 1,200 hours at 65 ± 5 °C in nitrogen, against 65% for the control.

It is not a record, not a module, and not a weather test. It is a single molecule doing two interface jobs with a named chemical mechanism for each, proven against a control that differs only in the group that does the work. The number to watch is not 24.68%. It is the second molecule — the first additional dicarboxylic-acid transport layer that produces the same directional gain, proving anchoring-plus-passivation is a design rule for the buried interface rather than a property of one molecule.

No investment recommendation is made or implied anywhere in this article.

Sources and further reading

  • “Carboxylic Acid-Based Hole Transport Layer with Both Anchoring and Passivation in Inverted Perovskite Solar Cells.” ACS Applied Materials & Interfaces (2026), published November 5, 2026. DOI: 10.1021/acsami.6c18188 — the primary source for the mechanism (dicarboxylic acid groups on the central phenyl ring anchoring to ITO and passivating the perovskite; methylthio-substituted triphenylamine and thiophene units for energy alignment and hole transfer), the control design (TPHex with dihexyl substitution), and all device figures quoted (24.68% vs. 23.72% at 0.1 cm²; 75% vs. 65% retention after 1,200 hours at 65 ± 5 °C in nitrogen), as well as the efficiency attribution (suppressed nonradiative recombination, rapid interfacial charge transfer, minimal energy loss from well-matched levels). The full text is paywalled; this article works from the abstract and claims no figure the abstract does not state — no open-circuit voltage, short-circuit current density, fill factor, perovskite composition, device stack, deposition method, author names, or affiliations are quoted, because none appears in the abstract as consulted.
  • Excitonic.com, “Two Defect-Fixers That Don’t Fight Each Other Take Perovskite Cells to 27.4%” (October 27, 2026) — the Northwestern–DGIST orthogonal-passivation result cited for the record-cell comparison (27.4% champion, certified quasi-steady-state 25.8%, 0.049 cm² aperture).
  • Excitonic.com, “A Molecular Carpet Lifts Printable Carbon Perovskite Cells to 14.7%” (October 28, 2026) — the MeO-2PACz SAM post-treatment on carbon cells, cited for the phosphonic-acid anchoring precedent and the SAM-transfer theme.
  • Mäkinen, P., Han, Y., Koch, G., et al. “Synergistic Halogen-Bonding Passivation and Carboxylic Acid Anchoring in Carbazole-Based Hole-Transport Materials Enhance Performance and Stability of p–i–n Perovskite Solar Cells.” Small (2025), DOI: 10.1002/smll.202502236 — the Tampere TC-ICA result (halogen-bond passivation plus carboxylic-acid anchoring to ITO; 15.4% vs. 14.7% and 10.7% for single-function controls), cited as the closest prior dual-function precedent.
  • “Synergistic Interface Engineering via Carboxylic-Acid-Functionalized Triphenylamine Derivative for High-Performance and Stable Perovskite Solar Cells.” ACS Applied Energy Materials — the TPAC result (carboxylic-acid-functionalized triphenylamine passivating undercoordinated Pb²⁺; 21.9% vs. 19.7% for pristine MAPbI₃), cited for the carboxyl-to-lead passivation mechanism.
  • Zheng, X., et al. “Co-deposition of hole-selective contact and absorber for improving the processability of perovskite solar cells.” Nature Energy 8, 462–472 (2023), DOI: 10.1038/s41560-023-01227-6 — the co-deposition precedent in which carboxylic and phosphonic acids self-assemble on ITO during perovskite processing, cited for the step-count manufacturing theme.

Background on perovskite defect chemistry, Lewis acid–base passivation, and inverted-cell device physics is drawn from the standard perovskite literature. No market-size figure is quoted, because none is needed for this interface result, and no module, damp-heat, or outdoor figure is stated, because none appears in the sources consulted.