Monolayer transition-metal dichalcogenides (TMDs) have been extensively investigated due to their unique optoelectronic properties,1^–5 including tunable bandgap, strong luminescence, atomic thickness,6 and high mobility.5 Stacking two atomically thin van der Waals (vdW) layers has provided an effective approach for creating heterostructures, with interlayer couplings becoming a new design parameter for engineering their electronic and optoelectronic properties.1^–5^,7^–10 Twist angle (θ) can influence interlayer couplings, triggering novel quantum phenomena, including localized bright excitons,11 superconductivity in magic-angle graphene,12 and Mott insulators in twisted bilayer TMDs.13 Nevertheless, the relatively weak vdW forces in layered graphene and TMDs limit the modulation of interlayer couplings through twist angles, resulting in small energy modulation (∼tens of meV).7^,11 Consequently, these interlayer couplings are susceptible to thermal fluctuations and disorders,11 thereby necessitating investigations primarily at low temperatures.

On the other hand, two-dimensional (2D) organic crystals exhibit high exciton binding energies, low-symmetry structure, near-unity quantum yield,14 super-transport excitons, etc.15 Combining TMDs with 2D organic single crystals enables significant control over interlayer couplings, including band structures,15 interlayer relaxations,8 charge redistributions, and exciton hybridizations.6^,16^–19 More importantly, the lower interfacial spacing (<3Å) in 2D organic–inorganic (O-I) systems enhances interlayer couplings through vdW bonding.17 In addition, the inherently low structural symmetry of organic single crystals allows for the manipulation of interlayer couplings through twisting methods that alter structural symmetry.15^,20 External stimuli including temperature and electrical gating can further modulate interlayer couplings and interlayer relaxations.8^,15 Therefore, fabricating twisted 2D O-I systems holds significant promise for improving modulation amplitudes of interlayer couplings. However, key challenges remain for advancing research in this field: (1) synthesizing 2D organic single crystals with controlled thickness,6^,15^,21 (2) achieving accurate twist angle control in 2D organic–inorganic (O-I) heterostructures, and (3) developing a continuous and reversible tuning method.7^,22^,23

In this work, we choose type II 2D pentacene-MoS2 heterostructures (HSs) to explore twisted-dependent interlayer couplings using density functional theory (DFT) calculations. The theoretical results reveal highly tunable interlayer spacing and interlayer charge transfers through varying twist angles, with a minimum interlayer spacing of ∼2.70Å at the twist angle of ∼30deg. Experimental results also substantiate that twist angle significantly influences interlayer couplings in these 2D O-I HSs at room temperature. The peak charge transfer efficiency (γ) occurs at the twist angle of ∼32deg, leading to a charge transfer induced Fermi level (EF) shift of ∼28.01meV. This pronounced n-doping effect in MoS2 is ascribed to the strong interlayer couplings at this specific angle, facilitating electron migration from pentacene to MoS2 and hole migration in the opposite direction. Temperature-dependent photoluminescence (PL) measurements reveal the enhanced couplings at lower temperatures, attributed to reduced interlayer spacing. The applied back gate voltage (VG) further modulates interlayer couplings by altering the relative band offset between pentacene and MoS2, substantiated by DFT calculations with bias voltages. This phenomenon contributes to untangling the convolution of excitons and trions, where the diffusion length (LD) of isolated excitons and trions is measured to be ∼1.95 and 0.93μm, respectively. In addition, horizontal electrical fields are found to increase exciton dissociations, thereby enhancing charge transfers and providing additional evidence for electrical coupling. These findings open new avenues for engineering interlayer couplings in 2D O-I HSs, holding promise for the development of high-performance optoelectronic devices such as light-emitting diodes and photovoltaic devices.

DFT calculations reveal that HSs exhibit a type II band alignment [Figs. 1(a) and S1, S2 in the Supplementary Material]. In this band alignment, the conduction band minimum and valence band maximum reside in distinct constituent layers, facilitating efficient charge transfers.17^,18 Notably, pentacene exhibits J-type molecular aggregation with a low symmetric structure,24 and MoS2 displays high lattice symmetry25 [Fig. 1(b)]. The interlayer couplings and charge transfers in HSs can be controlled by adjusting the distribution of wave functions for charges through varying twist angles.11 The twist angle θ refers to the deviated angle between the b-axis direction of pentacene and the armchair direction of MoS2 [Fig. 1(b)]. DFT calculations further illustrate that twist angles affect the interfacial spacing and interlayer charges, possibly due to steric effects.26 Specifically, the 2D O-I heterostructure shows a minimum interfacing spacing (∼2.70Å) at the twisted angle of 30 deg, implying the highest interlayer γ and strongest coupling at this twist angle [Figs. 1(c) and 1(d)].

Twisted HSs were constructed by transferring 2D pentacene crystals over CVD-grown monolayer MoS2 [detailed in Figs. S3–S5 in the Supplementary Material]. Selected-area electron diffraction (SAED) illustrates a set of diffraction patterns, confirming the high crystallinity of 2D pentacene [Fig. 1(e)]. By utilizing the diffraction spots, we determine the lattice constants of 2D pentacene as ∼6.25±0.19 and 7.74±0.11Å along the a- and b-axes, respectively, and the angle between them is ∼91.9deg. Figure 1(f) demonstrates the PL spectra of isolated 1L MoS2, HS, and 2D pentacene single crystals (∼10nm). The HS clearly shows PL quenching compared with isolated MoS2. Considering the clean interface between 2D pentacene and MoS2 [Fig. 1(f) and Fig. S6 in the Supplementary Material], this quenching is attributed to the type II heterostructure facilitating electrons transfer from pentacene to MoS2 and holes from MoS2 to pentacene, which enhances the n-doping level of MoS2 and p-doping of pentacene [Fig. 1(g)].27^,28 For quantitative analysis, we define the PL quenching factor (η) as the ratio of PL intensity for monolayer MoS2 to HS (IMoS2/IHS). In addition, the less pronounced quenching of pentacene may be attributed to the competing mechanisms of charge transfer and energy transfer. As the twist angle increases from 6 to 53 deg, η shows non-monotonic changes [Fig. 1(h)]. η reaches the maximum (∼2.57) at ∼32deg, primarily due to the strong coupling at this angle.29^–31 The proposed charge transfer efficiency (γ=1−1/η) quantitatively describes this process, showing a similar trend to η, with a maximum value of ∼61%. Efficient charge transfers lead to the conversion of excitons to trions.16^,32 To confirm this claim, we fit the PL spectra of HS at different twist angles by Lorentz functions (Fig. S7 in the Supplementary Material). The energy difference (EA−ET) approaches the peak value (∼34.7meV) at the twist angle of ∼32deg16^,17 (Fig. S8 in the Supplementary Material). The consistent trend of blue shifts, full width at half maximum (FWHM),33 and EA−ET validate the significant tuning of interlayer couplings in this O-I system through adjusting twist angles [Fig. 1(i) and Fig. S8 in the Supplementary Material].34^,35

We utilize a three-level model to quantitatively analyze how twist angles influence this interlayer relaxation process and coupling in the pentacene-MoS2 heterostructure [Fig. 2(a)].17^,36 The PL intensity ratio of trion to exciton (IT/IA)37 that reflects the doping level of MoS2 from the HS mirrors the γ trend [Fig. 2(b)]. The carrier density of MoS2 from the HS follows the relation17^,36NAneNT=βkBTexp(−EbkBT),(1)where NA and NT denote the trion and exciton population, respectively; kB, ne, and β are the Boltzmann constant, the doped electron density, and the coefficient, respectively; Eb is the trion binding energy (detailed in Supplementary Note 4 in the Supplementary Material). At ∼32deg, the doping level reaches a maximum (∼9.68×1012cm−2), correlating with the highest observed γ [Fig. 2(b)]. We also converted doping density to Fermi energy (EF=ℏ2πne/2meq2) using an electron band mass of me=0.35m0.32 To exclude the impact of intrinsic charging effects, we determined the neutral point of EF=0meV at VG=−107V by measuring the transfer curves of isolated 1L MoS2 [Fig. 2(c) and detailed in Supplementary Note 5 in the Supplementary Material]. Consequently, the twist angle induces a maximum EF shift of ∼28.01meV, corresponding to VG=28.33V [Fig. 2(d)].32 These pronounced changes underscore the effectiveness of twist angles in finely tuning interlayer relaxation processes8 and couplings within this 2D O-I system.

Temperature exerts a critical influence on interlayer couplings by altering interlayer spacing.37^,38 Here, we conducted the temperature-dependent PL measurements on 1L MoS2 and HS, exhibiting enhanced PL intensity at lower temperatures (Fig. S9 in the Supplementary Material). Both the PL quenching factor (η: 2.57→4.4) and charge transfer efficiency (γ: 61.1%→77.3%) increase as temperatures decrease [Fig. 2(e) and Fig. S10 in the Supplementary Material], coinciding well with the predicted reduction in interlayer spacing.37^,38 According to the quantum tunneling model, interlayer couplings follow an exponential relationship with interlayer spacings.38 To gain insight into interlayer couplings, we employ the modified Varshni relationship to fit the PL peak positions across different temperatures,39E(T)=E0−S⟨ℏω⟩[coth⟨ℏω⟩2kBT−1],(2)where E0 and S represent the emission energy at 0 K and interlayer coupling, and ℏ and ⟨ℏω⟩ represent Planck’s constant and the average phonon energy, respectively. As the twist angle evolves, S generally shares the same trend as η and EA−ET, reaching a maximum of ∼2.72 at ∼32deg [Fig. 2(f), Fig. S11 and Table S1 in the Supplementary Material]. This similarity underscores the significant role of twist angle in tuning interlayer couplings (detailed in Supplementary Note 6 in the Supplementary Material).

In addition, electrical band alignment plays a crucial role in influencing interlayer couplings in 2D heterostructures.17^,40^,41 Using the DFT method, we first examine the electrical band structure of HS under different bias voltages. The position of EF is close to the conduction band of MoS2, indicating n-doping feature under VG>0.42 Conversely, when VG<0, EF shifts toward the valence band of the MoS2 bandgap, displaying p-doping effects [Figs. 3(a) and 3(b)]. In contrast, pentacene exhibits insensitivity to applied voltages due to its low carrier mobility.15^,43

To further evaluate the doping effects, the partial density of states (PDOS) of HS is calculated under different bias voltages, demonstrating the same variations [Fig. 3(c)]. To experimentally confirm the tunability of interlayer relaxations and couplings, we fabricate a metal–oxide–semiconductor (MOS) structure using HS with a twist angle of ∼32deg [Fig. 3(d), detailed in the experimental section]. At 83 K, applying a vertical bias voltage reveals highly voltage-sensitive PL spectra for both the heterostructure and MoS2 (Fig. S12 in the Supplementary Material). As VG sweeps from −50 to 50 V, η increases from 1.88 to 5.15, and γ increases from 46.9% to 80.6% [Fig. 3(e)]. These variations are attributed to the changed relative band offsets between pentacene and MoS2.41^,44 Specifically, a decrease in band offset at high negative voltages reduces the driving force for charge transfer, thereby suppressing efficiency. Notably, the out-of-plane electric field from a single back-gate setup is very small and has a negligible effect on our heterostructure. Conversely, higher positive voltages increase the band offset and enhance γ. These results clarify how vertical bias voltage tunes interlayer relaxation and couplings by modifying electrical band alignment.

Although the exciton LD is the key to determining the exciton transport24^,45 and the performance of exciton devices, its measurements are convoluted by neutral excitons and trions attributed to natural background dopings.22 The highly gate-tunable interlayer relaxations help to disentangle the diffusion of neutral excitons and trions, crucial for exploring the fundamental limits and potential of new optoelectronic devices.22Figure 4(a) presents the PL images of HS excited by a diffraction-limited CW laser at different VG. The extracted LD generally decreases as VG ranges from −50 to 50 V (detailed in Supplementary Note 7 in the Supplementary Material). At high negative voltages, p-doping effects dominate, enhancing the presence of neutral excitons.46^–48 Conversely, positive voltages lead to a conversion of excitons into trions, providing a platform to explore trion transport properties (Fig. S13 in the Supplementary Material).32^,49Figure 4(b) illustrates that the observed LD for isolated neutral excitons (at VG=−50V) and trions (at VG=50V) are ∼1.95 and 0.93μm, respectively. The LD is defined as the average distance that excitons/trions travel between generation and recombination. The increased lifetime leads to a larger diffusion length for excitons. Given the longer lifetime of neutral excitons (∼10ns) compared with trions (∼50ps),22^,23^,45 the former exhibit longer diffusion distances [Fig. 4(c)]. This direct observation of isolated neutral exciton diffusion in the 2D O-I HS is not only essential for understanding exciton transport physics but also offers valuable insights for designing and characterizing systems reliant on neutral excitons and trions.50^,51

Furthermore, horizontal electrical fields promote the dissociation of excitons due to the non-zero in-plane dipole moment for intralayer excitons, facilitating interlayer relaxations.52^,53 The combination of 2D pentacene and monolayer MoS2 forms a 2D O-I PN junction, ensuring quick and efficient charge transfer due to no exciton (or minority carrier) diffusion.52 To explore the dynamic process, we conduct high-resolution spatial mapping of photocurrents on the heterostructure (the setup is shown in Fig. S14 in the Supplementary Material). As shown in Figs. 4(d)–4(f), photocurrents are predominantly generated at the PN configuration from the conversion of excitons into free carriers, which is primarily due to the built-in electric field. In contrast, the monolayer MoS2 device shows a uniform photocurrent across the entire channel due to the lack of built-in electric fields (Fig. S15 in the Supplementary Material). These differences verify a faster exciton dissociation process (∼6.7ps)27 compared with other decay channels within the heterostructures and isolated MoS2. In particular, the strongest photocurrent occurs at negative voltages (VD=−0.5V) [Figs. 4(g)–4(i)]. This phenomenon is attributed to the in-plane electric field increasing the band offset between pentacene and MoS2, facilitating efficient charge transfers.41^,44 Under forward bias voltage (VD=0.2V), reduced relative band offsets lead to majority carrier recombination through interlayer tunneling.27^,53Figure 4(j) provides a comprehensive view of the dynamic process, shedding light on the charge transfer mechanism under an in-plane electric field. The photocurrent characteristics depend significantly on the competition between interlayer relaxations (e.g., exciton dissociation and free-carrier drift) and loss pathways.

We demonstrate highly tunable interlayer relaxations and couplings within twisted HS. At a twist angle of 32 deg, we achieve a maximum γ of ∼61% and a peak EA−ET value of ∼34.7meV, corresponding to the charge transfer-induced Fermi level of ∼28.01meV. Temperature-dependent PL measurements reveal an increasing trend of interlayer couplings with the decrease in temperatures, reaching a maximum coupling strength of ∼2.72 at 32 deg. Using the DFT method, both electrical band alignment and PDOS calculations illustrate enhanced n-doping at VG>0V, due to the increased band offsets that facilitate efficient charge transfers. These theoretical results are corroborated by gate-tuned exciton emissions of the heterostructure, showing the highest γ of ∼80.6% at VG=50V. Furthermore, we obtain the LD of the isolated neutral exciton (∼1.95μm) and trions (∼0.93μm) in the HS by tuning the interlayer relaxations. Finally, the photocurrent mapping indicates that the photocurrent is primarily generated in the heterostructure regions, suggesting efficient exciton dissociations and charge redistributions driven by built-in electric fields, horizontal electric fields, and strong interlayer couplings. In the future, combining twist and gate effects could further enhance the tunability of interlayer couplings in 2D O-I heterostructures, enabling the exploration of exciton phases, such as type II interlayer trions, hybrid excitons, and quadrupolar excitons. This approach also holds the potential for developing high-performance next-generation optoelectronic devices, including polarization-sensitive photodetectors, quantum emitters, optical routers, exciton transistors, and light emitting diodes (LEDs).

Prior to spin-coating the liquid phase precursor onto the SiO2 (300 nm)/Si substrate, the substrate underwent a cleaning process involving deionized water and isopropyl alcohol to remove surface impurities, followed by argon plasma treatment. A mixed solution of 20 mmol/L (NH4)2MoO4 and 20 mmol/L potassium iodide was then spin-coated onto the substrate to synthesize MoS2 monolayers. The liquid phase precursor-coated substrate and sulfur powder were strategically positioned within the CVD furnace, which was then heated to 800°C at a ramping rate of 40°Cmin−1 and maintained at this temperature for 10 min under an argon gas flow of 60 sccm. Upon completion of the growth process, the furnace was allowed to naturally cool down to room temperature. Pentacene crystals were grown using a microspacing in-air sublimation method.54

HS samples with different twist angles were fabricated using the dry transfer method.55 The pentacene crystal was transferred onto MoS2 monolayers that were grown on SiO2 (300 nm)/Si using the CVD method. The samples were subsequently heated on a hot stage at 60°C for ∼6min.

Monolayer MoS2 was mechanically exfoliated using the Scotch tape method and transferred onto the SiO2/Si substrate via the dry transfer method. The monolayer TMD films were characterized by optical microscopy and Raman spectroscopy using a 532-nm excitation laser and a 50× objective lens, enabling the identification of samples. The 2D pentacene crystal of suitable size was placed onto the TMD film with a micro-aligner stage. Subsequently, two gold electrodes were mechanically transferred onto the TMD film and the HS, serving as the contact pads for the heterostructure photodetectors. To ensure strong coupling, all heterostructures underwent an annealing process in a high-vacuum environment.

All the measurements (PL and Raman) were performed using a home-built-in PL system equipped with a confocal microscope, imaging spectrographs, and scanning monochromator (SpectraPro HRS 500). A 532-nm continuous-wave laser, cleaned by a bandpass filter, serves as the excitation source, and it was synchronized and subsequently focused to a diffraction-limited spot using a 50×0.55 NA objective lens. Temperature-dependent measurements were performed in the range of 83 to 300 K. The sample was placed in a microscope-compatible chamber with a low-temperature controller [using liquid nitrogen (N2) as the coolant agent]. Multiple samples are measured for each structure with repeatable results. The electrical bias voltage was applied using a Keithley 4200 semiconductor analyzer.

For exciton LD measurements, we set a PL imaging configuration in our PL system. A 532-nm laser was used as the excitation source, which was synchronized and subsequently focused to a diffraction-limited spot using a 50×0.55 NA objective lens. A long-pass filter was placed in the detection path allowing only the PL image signal to be measured by an optical detector (SONY Exmor CMOS, sensor size 14.4mm×9.9mm, 1600×1100pixels) with a pixel size of 0.184μm.

Transmission electron microscopy (TEM) and SAED images were collected using a Hitachi HF5000 environmental aberration-corrected TEM/STEM/SE under 200 kV.

All measurements were performed using a Keithley 4200 parameter analyzer and Keithley 6482 at 300 K. Photocurrent mapping was conducted under a 532-nm laser with modulation from a square-wave generator. Incident light intensity was measured using a Thorlabs PM100D power meter. Fast temporal photoresponses were captured using a home-built setup employing a high-frequency oscilloscope and a low-noise current preamplifier (Stanford Research SR570). Responsivity spectra were obtained using a Newport xenon lamp source and a spectrophotometer.

We employed first-principle calculations to simulate the atom-resolved band structures and partial electron densities of HS using the Quantum ATK software with the LCAOCalculator. The exchange functional and correlation functionals were described by Perdew–Burke–Ernzerhof functions. The GGASG15 pseudopotential is applied to describe the ionic cores, whereas the density mesh cutoff was 370 Ry, and the k-point sampling grid for 2D crystal slabs is 4×4×1. The Grimme DFT-D2 is used to correct the energy dispersion of van der Waals interactions. A 2D pentacene-MoS2 slab was constructed using a 2×5 supercell of MoS2 along [11¯0] and [110] directions, and a 2×2 supercell of pentacene along [100] and [010] directions, maintaining a lattice mismatch below 5%. All atomic sites of the 2D slabs were fully relaxed before computing energy dispersion curves and density of states. The simulations of the external basis were conducted by applying 0.003e (n-doping) and −0.003e (p-doping) on Mo and S atoms, respectively.

The twisting model of MoS2-pentacene heterostructures was built by rotating the MoS2 monolayer along the b-axis with the angles of 10, 15, 20, 30, 40, 45, 50, and 60 deg. The calculations were performed using the Quantum ATK software with the LCAOCalculator, applying nonperiodic boundary conditions during the optimization process. The initial interlayer spacing was set as 2.6 Å. To reduce the calculation consumption, the MoS2 and pentacene monolayer were treated as rigid bodies, focusing solely on the influence of interlayer spacing on interlayer interaction.

Acknowledgment. We acknowledge the Center for Microscopy and Analysis at Nanjing University of Aeronautics and Astronautics for optical characterizations and data analysis. L. L. Z. acknowledges the support from the National Natural Science Foundation of China (NSFC) (Grant Nos. 62204117 and 62004086), the Jiangsu Province Science Foundation for Youths (Grant No. BK20210275), the Science and Technology Innovation Foundation for Youths (Grant No. NS2022099), the Postgraduate Research & Practice Innovation Program of Jiangsu Province (Grant No. KYCX22 0325), the Research Plan for Short Visit Program, Nanjing University of Aeronautics and Astronautics (NUAA) (Grant No. 250101DF08), and the Visiting Scholar Foundation of Key Laboratory of Optoelectronic Technology & Systems (Chongqing University), Ministry of Education. S. C. Q acknowledges support from the Guangyue Young Scholar Innovation Team of Liaocheng University (Grant No. LUGYTD2023-01). F. Z. acknowledges the support from the Natural Science Foundation of Southwest University of Science and Technology (Grant No. 22zx7130).

Shafqat Hussain is a PhD student at Nanjing University of Aeronautics and Astronautics (NUAA), China. His research interests include 2D material synthesis, exciton physics, and high-performance optoelectronic devices.

Linglong Zhang is an associate professor at Nanjing University of Aeronautics and Astronautics (NUAA), China. He was awarded PhD by Nanjing University, China. He worked as a postdoctoral researcher at the Australian National University. His research interests include the synthesis of novel nanomaterials, exciton physics, and high-performance optoelectronic devices.

Biographies of the other authors are not available.