Amine‐Functionalized Carbon Dots as PEDOT:PSS Dopants for Organic Solar Cells

The hole transport layer (HTL) is an important component in organic solar cells (OSCs). As a commonly used HTL, poly(3,4‐ethylenedioxythiophene):poly(styrene‐sulfonate) (PEDOT:PSS) in OSCs generally exhibits relatively low conductivity, hindering device performance improvement. Herein, amine‐functionalized carbon dots (CDs‐N) are introduced as dopants of PEDOT:PSS, which can effectively enhance the HTL conductivity while maintaining a similar work function. The PEDOT:PSS+CDs‐N devices exhibit higher power conversion efficiencies due to the improved charge extraction ability. It is found that the larger conjugated backbones of CDs‐N and the amine functional groups interacting with PSS are keys for such optimization. This work illustrates the potential of CDs for the applications on interface layers in OSCs.


Introduction
Organic solar cells (OSCs) are a promising photovoltaic technology due to their unique features of flexibility, lightweight, and DOI: 10.1002/admi.202300502semitransparency. [1]Owing to the development of active materials and device technologies, the power conversion efficiencies (PCEs) of OSCs have approached 20%. [2]5] Poly (3,4-ethylenedioxythiophene):poly (styrene-sulfonate) (PEDOT:PSS) is a widely used HTL, which presents appropriate work function (WF), good transparency, and solution processability. [6,7]owever, the PEDOT:PSS material used for OSCs generally exhibits moderate conductivity due to the core-shield structure of conducting PEDOT and hydrophilic insulating PSS. [8,9]OSCs with PEDOT:PSS HTL may show relatively large series resistance and charge recombination, resulting in unsatisfied device performance with limited short-circuit current density (J sc ) and fill factor (FF). [10,11] Some additives (e.g., graphene oxide, [12] dopamine hydrochloride, [13] and graphitic carbon nitride [14] ) have been reported as dopants of PEDOT:PSS to optimize its conductivity, revealing that doping is an effective method to enhance PEDOT:PSS conductivity for achieving higher device performance.
Carbon dots (CDs) are a kind of emerging nanomaterials with the advantages of easy synthesis, high conductivity, good compatibility, and rich surface functional groups. [15,16]CDs were used as doping materials to optimize HTL properties in perovskite solar cells and light-emitting diodes. [17]In addition, OSCs with fewlayered graphene CDs HTL have been reported with a PCE of 7.9%. [18]n this work, we developed an amine-functionalized CDs (CDs-N) doping approach to improve the PEDOT:PSS HTL for OSCs.CDs-N doped devices show higher J sc and PCEs than the undoped devices.In contrast, the doping of CDs without amine groups or 4-Aminopyridine on PEDOT:PSS do not show enhanced performance.It is found that CDs-N is unique as it has larger conjugated backbones and the amine functional groups can interact with PSS to weaken its shielding effect on PEDOT, thus leading to higher conductivity and better charge extraction ability of PEDOT:PSS+CDs-N.Meanwhile, such doping can still maintain appropriate WF.

Results and Discussion
Two kinds of CDs (CDs-N [19] and CDs-O) [20] with different functional groups were synthesized by the hydrothermal method according to the literature as shown in Figure 1a and Figures S1-S3 (Supporting Information).The precursor of CDs-N was synthesized by reacting citric acid with ethylenediamine, while the precursor of CDs-O was obtained by the nitration of pyrene.Xray photoelectron spectroscopy (XPS) was measured to study the chemical structure of CDs.The CDs-O spectrum affords obvious C 1s and O 1s peaks with no N 1s peak, suggesting there are no amine functional groups on its surface (Figure 1b).As presented in Figure S4 (Supporting Information), two peaks of C═C (288.2 eV) and C─O (284.8 eV) appears in high-resolution C 1s XPS spectrum, meaning CDs-O have hydroxyl as the literature reported. [20]Different from CDs-O, CDs-N exhibits an obvious N 1s (12.8%) peak at ≈400 eV in the XPS spectrum (Figure 1b).The N 1s spectrum is exhibited to further study the functional groups (Figure 1c).N 1s analysis reveals three types of N atoms: C─N─C, N─H, and N─(C) 3 .N─H is the dominant species of N atom in CDs-N, which may be originated from the amine functional groups.The result indicates the existence of amine functional groups on the surface of CDs-N.Fourier transform infrared (FT-IR) spectra were also measured to confirm their functional groups (Figure S5, Supporting Information).For CDs-O, the strong peak at ≈3400 cm −1 is ascribed to O─H bonds.The peak appears at ≈3200 cm −1 of CDs-N represents N─H bonds.These results were consistent with the XPS measurements, demonstrating the presence of hydroxyl groups in CDs-O and amine functional groups in CDs-N.
To evaluate HTL potential in OSCs, devices with the structure of ITO/HTL/PM6:Y6/PDINO/Ag was fabricated (Figure 2a; Figure S6, Supporting Information).The HTL layers were prepared by spin-coating of the corresponding solutions.CDs-N has a small size and can be easily dispersed in aqueous solutions.The PEDOT:PSS+CDs-N blending solution with 10 wt.% of CNs-N (when compared with PEDOT:PSS) exhibits good stability, and no precipitate was observed when stored at 5 °C for over one year.CDs-O and another amine functionalized organic molecule 4-Aminopyridine (NH 2 -Py) (Figure S6, Supporting Information) were also doped to form PEDOT:PSS+additive HTLs as control groups.As shown in Table 1, PEDOT:PSS+CDs-N afford higher PCEs than the pristine PEDOT:PSS, PEDOT:PSS+CDs-O, and PEDOT:PSS+ NH 2 -Py OSCs.In PEDOT:PSS+CDs-N OSCs, the PCE improves and then declines as the CDs-N percent rises from 5 to 20 wt.% (Table S1, Supporting Information).The optimal CDs-N doping proportion is 10 wt.%, and it was used for further studies.Compared to the pristine PEDOT devices (15.4%), the devices with PEDOT:PSS+CDs-N could achieve a higher average PCE of 16.2%.The champion PEDOT:PSS+CDs-N OSC presents a PCE up to 16.5%, with an open-circuit voltage (V oc ) of 0.841 V, J sc of 27.03 mA cm −2 , and FF of 72.7%  (Figure 2b).Compared to PEDOT:PSS OSCs (25.80 mA cm −2 ), the improved PCE of PEDOT:PSS+CDs-N devices is attributed to superior J sc , which can be further verified by external quantum efficiency (EQE) measurements (Figure S7, Supporting Information).
To explore the effect of CDs-N doping behavior in improving device performance, charge transport and exciton dissociation of OSCs have been measured.Space-charge-limited current (SCLC) method was characterized to evaluate charge mobilities (Figure S8, Supporting Information).The hole mobility (μ h ) was measured by a hole-only device with the structure of ITO/HTLs/PM6:Y6/MoO 3 /Ag.The measured μ h of PEDOT:PSS+CDs-N OSCs (1.6 × 10 −4 cm 2 V −1 s −1 ) is slightly higher than that of the PEDOT:PSS devices (1.4 × 10 −4 cm 2 V −1 s −1 ).The relationship between photocurrent density (J ph ) and effective voltage (V eff ) was studied in Figure S9 (Supporting Information), where V eff = V 0 -V app (V 0 is the voltage at a photocurrent of 0 mA cm −2 , V app is the applied bias voltage), J ph = J L -J D (J L and J D represent the current density under AM 1.5 G light and the dark current density, respectively). [21]J sat is the saturation current density when V eff is >2 V.The J ph /J sat ratio can be applied to define the probabilities of exciton dissociation. [22]The J ph /J sat ratio of PEDOT:PSS and PEDOT:PSS+CDs-N OSCs are 94.4% and 95.4%, respectively.It suggests that PEDOT:PSS+CDs-N HTL processes slightly better exciton dissociation ability, resulting in higher J sc . [23]lectrochemical impedance spectroscopy (EIS) was performed to investigate the interfacial resistances of the devices as shown in Figure S10 (Supporting Information).Based on the Nyquist plots and the fitted equivalent-circuit model, [24,25] the PEDOT:PSS+CDs-N OSC delivers smaller surface resistance than the PEDOT:PSS OSC.It suggests fewer interfacial defects in PEDOT:PSS+CDs-N OSC, which could depress the charge recombination. [26]To evaluate the charge recombination behavior of OSCs, the relationship between V oc , J sc , and light intensity (P light ) was studied.V oc as a function of P light is shown in Figure S11a (Supporting Information).The PEDOT:PSS+CDs-N device affords a lower slope (1.07 kT e −1 ) than the PE-DOT:PSS device (1.09 kT e −1 ), where k, T, and e are the Boltzmann constant, the absolute temperature of 300 K and the elementary charge, respectively.This suggests that the PEDOT:PSS+CDs-N OSC presents lower monomolecular recombination than the PEDOT:PSS device. [27]The relationship between J sc and P light is shown in Figure S11b (Supporting Information).According to the law of J sc ∝P light , OSCs with PE-DOT:PSS and PEDOT:PSS+CDs-N HTLs present the fitted slope of 0.974 and 0.985, respectively.It is noted that the slope of the PEDOT:PSS+CDs-N device is closer to 1, suggesting that bimolecular recombination is less in the PEDOT:PSS+CDs-N OSC than the PEDOT:PSS device. [28]These results indicate that CDs-N doped PEDOT:PSS can inhibit the charge carrier recombination of OSCs, which is conductive to higher J sc .
WF of HTLs was tested by ultraviolet photoemission spectroscopy (UPS) as shown in Figure S12 (Supporting Information).It is noted that PEDOT:PSS+CDs-N presents a slightly higher WF (5.17 eV) than PEDOT:PSS (5.11 eV), which is beneficial for enhancing hole collection at the anode, increasing V oc of OSCs. [29]The transmittance spectra of HTLs are shown in Figure S13 (Supporting Information).PEDOT:PSS+CDs-N film affords good transmittance over 80% from 300-1000 nm, which is almost the same as PEDOT:PSS film.It suggests that CDs-N additive nearly plays no effect on the light incidence into active layer.Atomic force microscopy (AFM) and water contact angle characterization were employed to reflect the HTL surface properties (Figures S14 and S15, Supporting Information).PEDOT:PSS and PEDOT:PSS+CDs-N display similar  root-mean-square roughness values of 0.897 and 0.947 nm, respectively.The water contact angles of PEDOT:PSS and PEDOT:PSS+CDs-N are 16.3°and 15.1°, respectively, which are basically the same.These results prove that CDs-N does not change the surface properties of PEDOT:PSS.
In an effort to further understand the doping mechanism of CDs-N, the conductivity of PEDOT:PSS, PEDOT:PSS+CDs-N, and control groups were investigated as shown in Figure 3a.PEDOT:PSS+CDs-O (6.6 × 10 −5 S cm −1 ) and PEDOT:PSS+NH 2 -Py (5.0 × 10 −5 S cm −1 ) HTL show similar conductivity with PE-DOT:PSS HTL (6.0 × 10 −5 S cm −1 ), while PEDOT:PSS+CDs-N delivers a higher conductivity (8.2 × 10 −5 S cm −1 ) than the others.To study the improvement of PEDOT:PSS+CDs-N HTL conductivity, the S 2p XPS spectra of HTLs were measured as shown in Figure 3b.In PEDOT:PSS, PSS ─SO 3 H interacts with PEDOT via Coulomb interaction, wrapping, and shielding PEDOT. [30]The two peaks at 163 and 167 eV in S (2p) XPS spectrum of PEDOT:PSS correspond to PEDOT and PSS, respectively. [31,32]After normalizing, the spectrum of PEDOT:PSS+CDs-O HTL is almost identical to the PEDOT:PSS, while PEDOT:PSS+CDs-N and PEDOT:PSS+NH 2 -Py HTLs afford higher peak at 163 eV (PEDOT), indicting more PE-DOT exposed after doping CNs-N or NH 2 -Py.It may be attributed to the protonated and positively charged amine functional groups of CDs-N and NH 2 -Py, which interacts with insulating PSS in PEDOT:PSS, weakening the shielding effect of PSS on PEDOT. [14,33,34]Different from NH 2 -Py with a small conjugated backbone, the larger conjugated backbone of CDs-N is beneficial for improving conductivity.These two effects combined in PEDOT:PSS+CDs-N can afford enhanced charge extraction and transportation, resulting in higher J sc and PCE in PEDOT:PSS+CDs-N device.
OSCs with PMT50:Y6 and PM6:BTP-eC9 (Figure S16, Supporting Information) as active layers were fabricated to estimate the universality of the PEDOT:PSS+CDs-N HTL.The PMT50:Y6 OSCs with PEDOT:PSS+CDs-N can afford higher PCEs than the PEDOT:PSS devices (Figure S17a and Table S2, Supporting Information).When PM6:BTP-eC9 was used as the active layer, the PEDOT:PSS+CDs-N OSCs also show higher PCEs up to 16.8% with slightly increased J sc values (Figure S17b and Table S2, Supporting Information).These results evident that the PEDOT:PSS+CDs-N HTL can function well with various active layers.

Conclusion
In conclusion, we have doped 10 wt.% CDs-N into PEDOT:PSS film to achieve an enhanced HTL for OSCs.CDs-N are optimized dopants due to their combined advantages of easy synthesis, good dispersion, large conjugated backbone, and amine functional groups for interacting with PSS.The PEDOT:PSS+CDs-N HTL exhibits higher conductivity and suitable work function.Compared to the pristine PEDOT:PSS devices (15.4%),OSCs with PEDOT:PSS+CDs N HTL can deliver an enhanced average PCE of 16.2% due to the better charge extraction and transport ability in devices.This study provides a new strategy to optimize PE-DOT:PSS for high-performance OSC devices and demonstrates the potential applications of functional CDs in OSCs.

Experimental Section
Devices were prepared based on ITO/HTL/Active layer/Electron transport layer (ETL)/Ag.The ITO substrates were subsequently sonicated with deionized water, acetone, and isopropanol for 30 min, and then dried in the oven.Next, the substrate was treated with UVO for 15 min in a UVozone chamber.Then, PEDOT:PSS and doping HTLs were spin-coated on top to form a thin layer (≈25 nm).All the HTLs were annealed for 15 min at 150 °C.The PM6:Y6 (1:1.2, w/w) blend was spin-coated on HTLs at 3000 rpm for 30 s to form the active layer (≈100 nm).The active layer was annealed at 110 °C for 10 min, followed by placing it in a vacuum oven below 1 × 10 −4 Pa for 4 h to remove additive.After that, PDINO dissolved in ethanol (1.0 mg mL −1 ) was coated onto the PM6:Y6 film at 3000 rpm for 30 s. Finally, 110 nm Ag was deposited on top.The area of OSCs was 0.045 cm 2 .
The J-V curves of the OSCs were characterized under AM1.5G solar simulator illumination (SS-F5-3A, Enlitech) on a computerized Keithley 2400 Source Measure Unit.EQE was measured by QE-R3011, Enlitech.The mobility was determined by Mott-Gurney rule of J = 9 r  0 μV 2 /8d 3 , where J is the current density,  0 is the permittivity of free space,  r is the relative permittivity of the material, μ is the hole/electron mobility, V is the voltage drop across the device, and d is the thickness of the active layers.Electrochemical impedance spectroscopy measurements were performed with PARSTAT 4000 at frequencies from 100 to 1 m Hz in dark at a bias voltage equal to V oc .XPS and UPS measurements were performed on the Escalab Xi+ system.Work function (Φ) could be calculated by the equation of Φ = h − E cutoff , where h is the incident ultraviolet photon energy (typically 21.2 eV) and E cutoff is the secondary-electron cutoff energy.The surface morphology of HTLs was characterized by Atomic force microscopy at tapping mode condition (XE7, Park System Inc).The water contact angle test was characterized by VCA Optima/VCA 3000S.The transmittance spectra of HTLs were measured by a Shimadzu UV3600 spectrophotometer at room temperature.

Figure 1 .
Figure 1.a) The structures and morphology of CDs-N and CDs-O.b) XPS spectra of CDs-O and CDs-N.c) N 1s XPS fine spectrum of CDs-N.

Figure 2 .
Figure 2. a) The device structure and PEDOT:PSS+CDs-N solution (10 wt.%) stored over one year.b) J-V curves of best OSCs under AM 1.5G.

Figure 3 .
Figure 3. a) Conductivity of HTLs.b) S(2p) XPS spectra of HTL films.c) A schematic illustration of PEDOT:PSS compositional distribution before and after CDs-N doping.

Table 1 .
Photovoltaic parameters of OSCs based on PEDOT:PSS or different HTLs with 10 wt.