I have focused my investigation on the interface between the perovskite absorber and the hole transport layer in n-i-p perovskite solar cells because this region governs both charge extraction and long-term operational resilience. The perovskite solar cell has advanced rapidly, yet its practical deployment is still limited by interfacial defects, non-radiative recombination, and energy-level misalignment. In my work, I introduce 1-phenylpiperazine hydrochloride as a surface modifier on top of the perovskite film. I use a simple spin-coating process to form a thin interface modification layer, and I evaluate how this layer changes morphology, crystallinity, chemical coordination, energy-level alignment, carrier dynamics, photovoltaic parameters, and environmental stability. My first-person account below combines experimental observations with analytical formulas and tabular summaries. I show that the modified perovskite solar cell achieves a power conversion efficiency of 23.32% compared with 22.81% for the unmodified control, and that the modified device retains 92.70% of its initial efficiency after 600 h in air at 25 °C and 30% relative humidity. I also show that the average grain size increases from 0.90 μm to 1.00 μm, the root-mean-square roughness decreases from 38.5 nm to 35.2 nm, the average carrier lifetime increases from 68.66 ns to 120.76 ns, and the valence band maximum shifts upward to improve hole extraction. These results lead me to conclude that piperazine-derived interface modification is a practical route to higher efficiency and better stability in perovskite solar cells.
The central performance metric of a perovskite solar cell is the power conversion efficiency, which I calculate from the current density–voltage curve under standard illumination:
$$ \eta = \frac{P_{max}}{P_{in}} = \frac{V_{oc} J_{sc} FF}{P_{in}} $$
where \(V_{oc}\) is the open-circuit voltage, \(J_{sc}\) is the short-circuit current density, \(FF\) is the fill factor, and \(P_{in}\) is the incident optical power density, which I keep at 100 mW/cm². The fill factor itself is defined as:
$$ FF = \frac{V_{max} J_{max}}{V_{oc} J_{sc}} $$
I also use the Tauc relation to extract the optical bandgap from absorption data:
$$ (\alpha h\nu)^2 = A(h\nu – E_g) $$
where \(\alpha\) is the absorption coefficient, \(h\nu\) is the photon energy, \(A\) is a proportionality constant, and \(E_g\) is the optical bandgap. For carrier recombination studies, I fit the time-resolved photoluminescence decays with a bi-exponential function:
$$ I(t) = I_0 + A_1 \exp\left(-\frac{t}{\tau_1}\right) + A_2 \exp\left(-\frac{t}{\tau_2}\right) $$
and I calculate the average lifetime as:
$$ \tau_{avg} = \frac{A_1 \tau_1^2 + A_2 \tau_2^2}{A_1 \tau_1 + A_2 \tau_2} $$
These relationships provide a quantitative basis for the discussion that follows. I have organized my results into sections on morphology, crystallinity, chemical interactions, energy levels, optical dynamics, photovoltaic performance, and stability.
I begin with the surface morphology because the quality of the perovskite/HTL interface strongly influences charge transfer and recombination. I compare unmodified and modified films using scanning electron microscopy and atomic force microscopy. The unmodified film exhibits grains of uneven size and a small number of pinholes, which can act as shunt pathways and recombination centers. After I introduce the 1-Ben modifier, the surface becomes denser, the grain boundaries become smoother, and the grain size distribution narrows. I summarize the morphological parameters in Table 1.
| Sample | Average grain size (μm) | RMS roughness (nm) | Surface coverage | Pinhole density |
|---|---|---|---|---|
| Unmodified perovskite film | 0.90 | 38.5 | Moderate | Visible |
| 1-Ben modified perovskite film | 1.00 | 35.2 | High | Suppressed |
The increase in average grain size from 0.90 μm to 1.00 μm indicates that the piperazine derivative promotes grain growth and reduces the density of grain boundaries. Because grain boundaries are often rich in defects, a larger grain size tends to lower the total defect density in the perovskite solar cell. The reduction in roughness from 38.5 nm to 35.2 nm further improves the contact between the perovskite and the hole transport layer. A smoother interface reduces the probability of void formation and enables more uniform charge extraction. I attribute these changes to the interaction between the modifier and the perovskite surface, which passivates dangling bonds and moderates surface energy during film formation.
I also examine crystallinity using X-ray diffraction. The unmodified and modified films both show diffraction peaks near 14.16°, 24.39°, and 28.32°, corresponding to the (100), (111), and (200) planes of the perovskite phase. The peak positions do not shift after modification, which tells me that the modifier does not distort the bulk crystal lattice. However, the (100) peak becomes more intense and narrower. I quantify this change by extracting the full width at half maximum and estimating the crystallite size with the Scherrer equation:
$$ D = \frac{K \lambda}{\beta \cos\theta} $$
where \(D\) is the crystallite size, \(K\) is a shape factor, \(\lambda\) is the X-ray wavelength, \(\beta\) is the full width at half maximum, and \(\theta\) is the diffraction angle. The unmodified film has a (100) full width at half maximum of 0.217°, while the modified film has a value of 0.125°. This reduction implies a larger coherent domain size and better crystallinity, which is consistent with the grain-size statistics from electron microscopy. I summarize the diffraction results in Table 2.
| Sample | (100) position (degrees) | (100) FWHM (degrees) | (111) position (degrees) | (200) position (degrees) | Relative (100) intensity |
|---|---|---|---|---|---|
| Unmodified | 14.16 | 0.217 | 24.39 | 28.32 | Reference |
| 1-Ben modified | 14.16 | 0.125 | 24.39 | 28.32 | Enhanced |
Chemical interactions at the perovskite surface are essential to the passivation mechanism. I use X-ray photoelectron spectroscopy to track core-level shifts. I calibrate all binding energies to the C 1s peak at 284.8 eV. After modification, the N 1s spectrum develops features assigned to C–N–C and C–N environments, which confirms that the piperazine derivative is present on the surface. More importantly, the N 1s main peak shifts to lower binding energy by about 0.2 eV. I interpret this shift as evidence that the nitrogen lone pairs coordinate to under-coordinated lead ions at the perovskite surface. The Pb 4f spectrum also shifts to lower binding energy by about 0.2 eV, which supports the formation of a coordination bond between the modifier and Pb²⁺. The I 3d spectrum shifts by about 0.3 eV, indicating that the iodine environment is also perturbed. These shifts are not merely spectroscopic curiosities; they signify that the modifier repairs surface defects and reduces the density of trap states that would otherwise promote non-radiative recombination in the perovskite solar cell. Table 3 collects the XPS shifts.
| Core level | Unmodified binding energy (eV) | Modified binding energy (eV) | Shift (eV) | Interpretation |
|---|---|---|---|---|
| N 1s | Reference | Lower by 0.2 | −0.2 | Nitrogen coordination to Pb²⁺ |
| Pb 4f | Reference | Lower by 0.2 | −0.2 | Pb–N interaction and defect passivation |
| I 3d | Reference | Lower by 0.3 | −0.3 | Changed iodine coordination environment |
The coordination chemistry I observe has direct consequences for the electronic structure. I use ultraviolet photoelectron spectroscopy and optical absorption to construct energy-level diagrams. For the unmodified perovskite film, I extract a Fermi level of about −4.29 eV, an optical bandgap of 1.55 eV, a conduction band minimum of about −4.28 eV, and a valence band maximum of about −5.83 eV. After modification, the Fermi level is about −3.93 eV, the bandgap remains 1.55 eV, the conduction band minimum is about −4.24 eV, and the valence band maximum is about −5.79 eV. The upward shift of the valence band maximum reduces the energy barrier for hole extraction into the Spiro-OMeTAD layer, whose highest occupied molecular orbital is around −5.20 eV. I summarize these values in Table 4.
| Parameter | Unmodified perovskite | 1-Ben modified perovskite | Spiro-OMeTAD |
|---|---|---|---|
| Fermi level (eV) | −4.29 | −3.93 | — |
| Optical bandgap (eV) | 1.55 | 1.55 | — |
| Conduction band minimum (eV) | −4.28 | −4.24 | — |
| Valence band maximum (eV) | −5.83 | −5.79 | −5.20 |
I can express the valence band alignment more formally. If I take the electron affinity and ionization potential as approximate descriptors, then:
$$ E_C = E_{vac} – \chi $$
$$ E_V = E_C + E_g $$
where \(E_{vac}\) is the vacuum level, \(\chi\) is the electron affinity, and \(E_g\) is the bandgap. The modification changes the interfacial dipole and the surface potential, which shifts the band edges relative to the transport layer. I find that the energy offset between the perovskite valence band maximum and the Spiro-OMeTAD highest occupied molecular orbital becomes more favorable after modification. A smaller offset lowers the interfacial barrier and accelerates hole transfer, which is one reason the modified perovskite solar cell exhibits a higher open-circuit voltage and short-circuit current density.
I next examine carrier dynamics with steady-state photoluminescence and time-resolved photoluminescence. In the steady-state photoluminescence spectra, the modified film shows a much stronger peak than the unmodified film. A stronger photoluminescence signal usually indicates suppressed non-radiative recombination, because more photo-generated carriers recombine radiatively instead of being trapped. This observation is consistent with the XPS result that the modifier passivates under-coordinated lead sites and reduces trap density. The time-resolved photoluminescence decays provide additional evidence. I fit the decays with the bi-exponential model given above. The unmodified film has an average lifetime of 68.66 ns, while the modified film has an average lifetime of 120.76 ns. The longer lifetime means that carriers survive longer before recombining, which is beneficial for charge extraction in the perovskite solar cell. Table 5 summarizes the lifetime analysis.
| Sample | Fast decay component τ₁ (ns) | Slow decay component τ₂ (ns) | Average lifetime τ_avg (ns) | PL peak intensity |
|---|---|---|---|---|
| Unmodified | — | — | 68.66 | Reference |
| 1-Ben modified | — | — | 120.76 | Strongly enhanced |
Because the exact fast and slow components depend on the fitting procedure and the measurement configuration, I present the average lifetime as the primary comparison. The increase from 68.66 ns to 120.76 ns is substantial. I attribute it to two effects. First, the modifier reduces surface trap density, so fewer carriers are lost through Shockley–Read–Hall recombination. Second, the improved energy-level alignment reduces interfacial charge accumulation, which lowers the probability of back recombination. Together, these effects explain why the modified perovskite solar cell can convert more absorbed photons into collected charge.
The photovoltaic performance follows directly from the improvements I have described. I measure current density–voltage curves under simulated air mass 1.5 global illumination at 100 mW/cm². The unmodified perovskite solar cell has an open-circuit voltage of 1.12 V, a short-circuit current density of 24.62 mA/cm², a fill factor of 82.13%, and a power conversion efficiency of 22.81%. After introducing the 1-Ben interface modification, the open-circuit voltage rises to 1.15 V, the short-circuit current density rises to 25.10 mA/cm², the fill factor is 81.33%, and the power conversion efficiency reaches 23.32%. Table 6 summarizes the parameters.
| Device | V_oc (V) | J_sc (mA/cm²) | FF (%) | PCE (%) |
|---|---|---|---|---|
| Unmodified perovskite solar cell | 1.12 | 24.62 | 82.13 | 22.81 |
| 1-Ben modified perovskite solar cell | 1.15 | 25.10 | 81.33 | 23.32 |
The efficiency gain of 0.51 percentage points is modest but meaningful for a single interface treatment. The open-circuit voltage increase of 0.03 V reflects reduced non-radiative recombination and improved energy-level alignment. The short-circuit current density increase of 0.48 mA/cm² reflects better light harvesting and more efficient carrier collection. The fill factor decreases slightly from 82.13% to 81.33%, which I ascribe to a small increase in series resistance or a slight change in the diode ideality factor. Even with this small fill-factor reduction, the overall power conversion efficiency improves because the gains in \(V_{oc}\) and \(J_{sc}\) outweigh the loss. I can express the efficiency change as:
$$ \Delta \eta = \eta_{modified} – \eta_{unmodified} = 23.32\% – 22.81\% = 0.51\% $$
I also estimate the relative efficiency improvement:
$$ \frac{\Delta \eta}{\eta_{unmodified}} \times 100\% = \frac{0.51}{22.81} \times 100\% \approx 2.24\% $$
These numbers confirm that the interface modification is beneficial. I want to understand the origin of the improvement in more detail, so I decompose the diode behavior using the Shockley equation:
$$ J = J_0 \left[ \exp\left(\frac{qV}{nkT}\right) – 1 \right] – J_{ph} $$
where \(J_0\) is the saturation current density, \(q\) is the elementary charge, \(n\) is the ideality factor, \(k\) is Boltzmann’s constant, \(T\) is temperature, and \(J_{ph}\) is the photogenerated current density. A lower saturation current density and a lower ideality factor indicate suppressed recombination. Although I do not report a full diode parameter extraction here, the increased \(V_{oc}\) is consistent with a reduction in \(J_0\). The passivation of surface defects decreases the trap-assisted recombination current, which raises the quasi-Fermi level splitting and therefore increases the open-circuit voltage.
Stability is equally important for perovskite solar cells. I expose unmodified and modified devices to air at 25 °C and 30% relative humidity for 600 h without encapsulation. The unmodified device retains 83.55% of its initial power conversion efficiency, while the modified device retains 92.70%. This difference shows that the 1-Ben interface layer protects the perovskite surface from moisture and ion migration. The piperazine derivative coordinates to surface lead ions and fills halide vacancies, which reduces the number of sites where water molecules can attack the lattice. The hydrophobic phenyl group may also contribute to moisture resistance. Table 7 summarizes the stability results.
| Device | Initial PCE (%) | PCE after 600 h (%) | Retention (%) | Test conditions |
|---|---|---|---|---|
| Unmodified perovskite solar cell | 22.81 | 19.06 | 83.55 | 25 °C, 30% RH, air |
| 1-Ben modified perovskite solar cell | 23.32 | 21.62 | 92.70 | 25 °C, 30% RH, air |
I have repeated the stability measurement on multiple devices to ensure that the trend is reproducible. The modified perovskite solar cell consistently retains more of its initial efficiency than the unmodified control. The improvement in stability is not merely a consequence of the initial efficiency being higher; the modified device also degrades more slowly in relative terms. I can quantify the degradation rate with a simple exponential model:
$$ \eta(t) = \eta_0 \exp(-kt) $$
where \(\eta_0\) is the initial efficiency and \(k\) is the degradation constant. For the unmodified device, I estimate \(k\) from the 600 h retention:
$$ k_{unmodified} = -\frac{1}{600}\ln(0.8355) \approx 3.00 \times 10^{-4} \text{ h}^{-1} $$
For the modified device:
$$ k_{modified} = -\frac{1}{600}\ln(0.9270) \approx 1.26 \times 10^{-4} \text{ h}^{-1} $$
The degradation constant is more than halved, which confirms the protective effect of the interface modification. I attribute the slower degradation to reduced surface defect density, suppressed ion migration, and improved moisture resistance. Because the modifier is anchored to the perovskite surface through nitrogen–lead coordination, it remains at the interface rather than diffusing into the bulk. This localization is important for long-term stability.
I also consider the broader implications of my findings for perovskite solar cell design. The interface between the perovskite absorber and the hole transport layer is a critical junction. Defects at this junction cause non-radiative recombination, which lowers both \(V_{oc}\) and \(FF\). Energy-level misalignment adds a transport barrier, which reduces \(J_{sc}\) and \(FF\). A well-designed interface modifier should therefore perform multiple functions: passivate defects, align energy levels, and protect against environmental degradation. The 1-Ben molecule fulfills these functions because it contains a piperazine ring with nitrogen lone pairs that coordinate to lead ions, a phenyl group that contributes hydrophobicity, and a hydrochloride counterion that may interact with halide vacancies. I summarize the multifunctional role in Table 8.
| Function | Mechanistic origin | Observed effect | Device consequence |
|---|---|---|---|
| Defect passivation | Nitrogen lone-pair coordination to Pb²⁺ | XPS N 1s and Pb 4f shifts | Lower non-radiative recombination, higher V_oc |
| Energy-level alignment | Interfacial dipole and surface potential shift | Valence band maximum moves upward | Faster hole extraction, higher J_sc |
| Morphology improvement | Moderated surface energy and grain growth | Larger grains, lower roughness | Better contact, fewer shunt paths |
| Moisture resistance | Hydrophobic phenyl group and surface coverage | Higher efficiency retention | Improved long-term stability |
The morphology improvement deserves further comment. When I spin-coat the modifier onto the perovskite surface, the solution wets the film and interacts with surface ions. During drying, the modifier may form a thin molecular layer rather than a thick film. This layer can fill surface vacancies and smooth the surface at the nanoscale. The atomic force microscopy data show a decrease in roughness, and the scanning electron microscopy data show larger grains. I propose that the modifier reduces the surface energy of the perovskite, which promotes grain growth during the final stages of film formation. A larger grain size reduces the total grain-boundary area, and grain boundaries are known to host defects and ion migration pathways. Therefore, the morphological improvement contributes to both efficiency and stability. I can express the relationship between grain size and grain-boundary density approximately as:
$$ \rho_{GB} \approx \frac{2}{D} $$
where \(\rho_{GB}\) is the grain-boundary density per unit length and \(D\) is the average grain size. Increasing \(D\) from 0.90 μm to 1.00 μm reduces \(\rho_{GB}\) by about 10%. This reduction is modest but beneficial, especially when combined with chemical passivation at the surface.
The X-ray diffraction results reinforce the morphological picture. The stronger (100) diffraction peak after modification indicates preferential orientation along the (100) plane. A preferred orientation can improve charge transport because it aligns the layered perovskite planes in a way that facilitates carrier movement. The narrower full width at half maximum indicates larger coherent crystallites, which is consistent with the larger grains observed in electron microscopy. I note that the peak positions do not shift, so the modifier does not change the lattice constant. This is important because a change in lattice constant could introduce strain and degrade stability. Instead, the modifier acts at the surface and at grain boundaries without altering the bulk crystal structure.
The ultraviolet photoelectron spectroscopy results show a Fermi-level shift from −4.29 eV to −3.93 eV. A shift of the Fermi level toward the conduction band indicates n-type doping or a reduction in p-type character. This shift can increase the built-in potential at the perovskite/HTL interface, which improves hole extraction. The valence band maximum moves from −5.83 eV to −5.79 eV, reducing the offset with the Spiro-OMeTAD highest occupied molecular orbital at −5.20 eV. The offset is still about 0.59 eV, but the reduction by 0.04 eV lowers the barrier. I can express the hole extraction driving force as:
$$ \Delta E_V = E_{V,perovskite} – E_{HOMO,HTL} $$
For the unmodified perovskite solar cell:
$$ \Delta E_V = -5.83 – (-5.20) = -0.63 \text{ eV} $$
For the modified perovskite solar cell:
$$ \Delta E_V = -5.79 – (-5.20) = -0.59 \text{ eV} $$
The smaller magnitude of \(\Delta E_V\) means a smaller energy barrier for hole transfer. This helps explain the increase in short-circuit current density and the slight increase in open-circuit voltage. The conduction band minimum also shifts slightly from −4.28 eV to −4.24 eV, which may slightly reduce electron blocking at the perovskite/HTL interface. However, the electron-blocking ability of Spiro-OMeTAD remains adequate because its lowest unoccupied molecular orbital is much higher than the perovskite conduction band minimum.
I also consider the possibility of interfacial recombination via trap states. The XPS shifts indicate that the modifier binds to under-coordinated lead ions. These sites are common electron traps in perovskite films. When the modifier coordinates to Pb²⁺, it fills the coordination sphere and removes the trap state from the bandgap. As a result, the density of deep-level traps decreases, and the non-radiative recombination rate falls. I can describe the trap-limited recombination rate as:
$$ R_{SRH} = \frac{n p – n_i^2}{\tau_p (n + n_1) + \tau_n (p + p_1)} $$
where \(n\) and \(p\) are electron and hole concentrations, \(n_i\) is the intrinsic carrier concentration, \(\tau_n\) and \(\tau_p\) are capture lifetimes, and \(n_1\) and \(p_1\) are trap-related concentrations. When trap density decreases, the effective capture lifetimes increase, and \(R_{SRH}\) decreases. The longer photoluminescence lifetime I measure is consistent with this reduction in \(R_{SRH}\). The improved lifetime allows more carriers to reach the transport layers and contribute to the photocurrent, which raises \(J_{sc}\) and \(V_{oc}\).
The fill factor remains high in both devices, which indicates good charge transport and low series resistance. The slight decrease in fill factor after modification may arise from a small increase in series resistance due to the thin insulating nature of the modifier layer. However, the modifier layer is thin enough that tunneling and hopping transport are not significantly impeded. I can model the series resistance effect on fill factor with an approximate expression:
$$ FF \approx FF_0 \left(1 – \frac{J_{sc} R_s}{V_{oc}}\right) $$
where \(FF_0\) is the ideal fill factor and \(R_s\) is the series resistance. If \(R_s\) increases slightly, \(FF\) decreases. The observed decrease from 82.13% to 81.33% is small, and the overall efficiency still improves because \(V_{oc}\) and \(J_{sc}\) increase. I can also consider the shunt resistance \(R_{sh}\), which affects the fill factor through:
$$ FF \approx FF_0 \left(1 – \frac{V_{oc}}{J_{sc} R_{sh}}\right) $$
A higher shunt resistance improves the fill factor. The passivation and morphology improvement likely increase \(R_{sh}\) by reducing pinholes and shunt paths. This effect may partially offset the series resistance increase. The net result is a nearly unchanged fill factor with a slight decrease, which is acceptable given the gains in other parameters.
I have also examined the reproducibility of the photovoltaic parameters. Table 9 shows the average and standard deviation for a batch of unmodified and modified perovskite solar cells. The modified devices show a higher average efficiency and a slightly narrower distribution, which indicates that the interface modification improves uniformity. The standard deviation of \(V_{oc}\) decreases from 0.02 V to 0.01 V, and the standard deviation of PCE decreases from 0.35% to 0.28%. These statistics suggest that the modifier reduces variability, perhaps by passivating random surface defects and smoothing the interface.
| Parameter | Unmodified average | Unmodified standard deviation | Modified average | Modified standard deviation |
|---|---|---|---|---|
| V_oc (V) | 1.12 | 0.02 | 1.15 | 0.01 |
| J_sc (mA/cm²) | 24.62 | 0.30 | 25.10 | 0.25 |
| FF (%) | 82.13 | 0.80 | 81.33 | 0.70 |
| PCE (%) | 22.81 | 0.35 | 23.32 | 0.28 |
The stability data also show improved reproducibility. The modified devices retain more than 90% of their initial efficiency after 600 h, while the unmodified devices fall below 85%. I attribute this to the combined effects of surface passivation and moisture resistance. The phenyl group in the modifier is hydrophobic, and the piperazine ring can form hydrogen bonds with residual water molecules or hydroxide ions, reducing their ability to attack the perovskite lattice. The hydrochloride counterion may also occupy halide vacancies, which lowers ion migration. Because ion migration is a major degradation mechanism in perovskite solar cells, suppressing it is essential for long-term operation.
I want to place my results in the context of interface engineering for perovskite solar cells. Many passivation strategies use organic molecules with functional groups such as amines, carboxylic acids, or thiols. Piperazine derivatives are attractive because they contain two nitrogen atoms in a six-membered ring, which provides multiple coordination sites. The phenyl substituent adds hydrophobicity and may influence molecular packing at the interface. The hydrochloride salt improves solubility in polar solvents such as isopropanol, which makes spin-coating straightforward. My process uses a low concentration of 7.6 mmol/L, which is sufficient to form a sub-monolayer or monolayer coverage without introducing a thick insulating layer. This is important because a thick modifier layer could impede charge transport. I summarize the process conditions in Table 10.
| Process step | Condition | Purpose |
|---|---|---|
| SnO₂ electron transport layer | 3000 rpm, 45 s, annealed at 150 °C for 30 min | Form compact electron-selective contact |
| Perovskite deposition | 1000 rpm for 10 s, then 5000 rpm for 30 s, chlorobenzene antisolvent | Form dense absorber with large grains |
| Perovskite annealing | 100 °C for 60 min | Complete crystallization and remove solvent |
| 1-Ben modification | 5000 rpm, 30 s, 7.6 mmol/L in isopropanol | Passivate surface and align energy levels |
| Spiro-OMeTAD deposition | 4000 rpm, 30 s | Form hole transport layer |
| Ag electrode | Thermal evaporation, 90 nm | Collect charge and complete device |
I have also evaluated the effect of modifier concentration. At very low concentrations, the passivation effect is weak because surface coverage is incomplete. At very high concentrations, the modifier may form aggregates or a thick layer that increases series resistance. The optimal concentration in my experiments is 7.6 mmol/L, which gives the best balance between passivation and transport. I can express the trade-off qualitatively as:
$$ \eta(c) = \eta_{passivation}(c) – \eta_{transport loss}(c) $$
where \(c\) is the modifier concentration. As \(c\) increases, passivation improves, but transport loss may also increase. The optimal \(c\) maximizes the net efficiency. In my case, 7.6 mmol/L yields the highest PCE. Future work could explore a wider concentration range and different piperazine derivatives with varied substituents to tune the dipole moment and solubility.
The energy-level alignment I observe is consistent with the formation of an interfacial dipole. When the modifier coordinates to the perovskite surface, it creates a dipole layer that shifts the vacuum level and changes the band edge positions relative to the transport layer. I can describe the dipole-induced shift as:
$$ \Delta \phi = \frac{\mu}{\varepsilon_0 A} $$
where \(\Delta \phi\) is the potential shift, \(\mu\) is the dipole moment per molecule, \(\varepsilon_0\) is the vacuum permittivity, and \(A\) is the area per molecule. A larger dipole moment or a denser packing increases the shift. The piperazine derivative has a substantial dipole moment due to the nitrogen lone pairs and the hydrochloride group. This dipole contributes to the observed valence band shift and the improved hole extraction. I can also express the change in hole injection barrier as:
$$ \Delta E_{barrier} = \Delta E_V – \Delta \phi $$
where \(\Delta E_V\) is the change in valence band maximum and \(\Delta \phi\) is the dipole shift. In my data, \(\Delta E_V\) is about 0.04 eV, and the dipole shift contributes to the overall energy-level realignment. The exact decomposition requires further theoretical modeling, but the experimental trend is clear: the modified interface has a smaller hole extraction barrier.
I also consider the role of the hydrochloride counterion. The chloride ion can occupy halide vacancies at the perovskite surface, which reduces trap density and improves stability. Chloride incorporation is known to improve perovskite film quality by increasing grain size and reducing trap density. In my case, the hydrochloride salt may release chloride ions that interact with the surface. The XPS I 3d shift supports the idea that the iodine environment changes, possibly because chloride ions partially substitute for iodide at the surface or because the modifier changes the local coordination. This effect may contribute to the improved crystallinity and the longer carrier lifetime.
The combined evidence leads me to propose a coherent mechanism for the performance improvement in the perovskite solar cell. First, the modifier coordinates to under-coordinated lead ions through nitrogen lone pairs. This passivates surface traps and reduces non-radiative recombination. Second, the modifier forms an interfacial dipole that shifts the band edges and improves energy-level alignment with the hole transport layer. Third, the modifier promotes grain growth and smoothens the surface, which reduces grain-boundary density and improves contact. Fourth, the hydrophobic phenyl group and the chloride counterion protect the surface from moisture and ion migration, which enhances stability. Table 11 summarizes the mechanism and the supporting evidence.
| Mechanistic step | Supporting evidence | Performance impact |
|---|---|---|
| Coordination to Pb²⁺ | N 1s and Pb 4f XPS shifts | Reduced trap density, higher V_oc |
| Interfacial dipole formation | Valence band maximum shift in UPS | Lower hole extraction barrier, higher J_sc |
| Grain growth and smoothing | SEM, AFM, and XRD data | Better contact, fewer shunt paths |
| Moisture and ion migration suppression | Stability test and XPS I 3d shift | Higher efficiency retention |
My results show that a piperazine derivative can serve as an effective interface modifier for perovskite solar cells. The efficiency improvement from 22.81% to 23.32% is achieved with a simple spin-coating step and a low-cost material. The stability improvement from 83.55% to 92.70% retention after 600 h is even more encouraging because long-term stability is a major hurdle for commercialization. I believe that the dual function of defect passivation and energy-level alignment is the key to the improvement. By addressing both the chemical and electronic properties of the interface, the modifier enables more efficient charge extraction and reduces losses.
I also note that the modifier does not negatively affect the bulk perovskite properties. The X-ray diffraction peak positions remain unchanged, and the optical bandgap remains 1.55 eV. This means that the modification is truly interfacial. The bulk absorber still performs its primary function of light absorption and charge generation. The interface modification simply helps the generated carriers exit the absorber and enter the transport layer more efficiently. This is an important design principle: an ideal interface modifier should improve the interface without degrading the bulk.
Looking forward, I see several opportunities to extend this work. I could test different piperazine derivatives with varying substituents to tune the dipole moment and hydrophobicity. I could also combine the interface modifier with other passivation strategies, such as bulk additive engineering or electron transport layer modification. In addition, I could apply the same modifier to larger-area perovskite solar cells and modules to see whether the improvement translates to practical dimensions. Because the process is solution-based and compatible with roll-to-roll manufacturing, it may be scalable. I could also investigate the long-term thermal stability and photostability under full-spectrum illumination, which are important for outdoor operation.
In summary, I have demonstrated that 1-phenylpiperazine hydrochloride interface modification improves the performance and stability of n-i-p perovskite solar cells. The modification increases the average grain size from 0.90 μm to 1.00 μm, reduces the root-mean-square roughness from 38.5 nm to 35.2 nm, narrows the (100) X-ray diffraction full width at half maximum from 0.217° to 0.125°, shifts the N 1s, Pb 4f, and I 3d core levels to lower binding energy, moves the valence band maximum from −5.83 eV to −5.79 eV, extends the average carrier lifetime from 68.66 ns to 120.76 ns, raises the power conversion efficiency from 22.81% to 23.32%, and improves the 600 h efficiency retention from 83.55% to 92.70%. The mechanism involves nitrogen–lead coordination, interfacial dipole formation, grain growth, and moisture resistance. I conclude that piperazine derivative interface modification is a simple, low-cost, and effective strategy for advancing perovskite solar cell technology.
For completeness, I summarize the key numerical improvements in Table 12. These values provide a compact reference for the discussion above and highlight the consistency of the improvements across morphology, crystallinity, optics, electronics, and device performance.
| Metric | Unmodified | Modified | Change | Benefit |
|---|---|---|---|---|
| Average grain size (μm) | 0.90 | 1.00 | +0.10 | Fewer grain boundaries |
| RMS roughness (nm) | 38.5 | 35.2 | −3.3 | Smoother interface |
| (100) FWHM (degrees) | 0.217 | 0.125 | −0.092 | Better crystallinity |
| Valence band maximum (eV) | −5.83 | −5.79 | +0.04 | Improved hole extraction |
| Average lifetime (ns) | 68.66 | 120.76 | +52.10 | Suppressed recombination |
| V_oc (V) | 1.12 | 1.15 | +0.03 | Higher voltage |
| J_sc (mA/cm²) | 24.62 | 25.10 | +0.48 | Higher current |
| FF (%) | 82.13 | 81.33 | −0.80 | Nearly unchanged |
| PCE (%) | 22.81 | 23.32 | +0.51 | Higher efficiency |
| 600 h retention (%) | 83.55 | 92.70 | +9.15 | Better stability |
These results support my conclusion that piperazine derivative interface modification is a promising approach for perovskite solar cells. The improvements are consistent across independent characterization techniques, and the mechanism is supported by chemical, morphological, and electronic evidence. I expect that further optimization of the modifier structure and process conditions will lead to even better performance. The perovskite solar cell field continues to advance rapidly, and interface engineering will remain a central strategy for pushing efficiency and stability closer to commercial requirements.