Strain engineering of the transition metal dichalcogenide chalcogen-alloy WSSe

The interest for two-dimensional (2D) materials, and, in particular, for monolayers (MLs) of transition metal dichalcogenides (TMDs), has skyrocketed in the past decade. Thanks to their sizable bandgap and strong spin–orbit interactions, 2D semiconducting TMDs have prompted many opportunities for quantum optoelectronics¹ and valleytronic² applications as well as for the observation of interesting physical phenomena like those reported in TMD heterostructures.³ The all-surface nature of TMD MLs make them especially responsive to mechanical deformations. The latter are an ubiquitous and very often inevitable presence in 2D materials and, at the same time, a precious tool to modify the sample’s physical properties.²⁴ Alloying is also a widely employed and efficacious method to modulate the electronic properties of bulk and nanostructured semiconductors in general⁴ and of 2D semiconductors in particular.⁵ Thanks to the recent progress in the synthesis of TMD alloys,^6–9 these materials are gaining increasing interest since they provide a virtually seamless variation in the properties of an alloyed $MC 2 x A C 2 ( 1 − x ) B$ ML between the end point compounds $MC 2 A$ and $MC 2 B$⁠, where M is a metal, such as Mo and W, and $C A , B$ are chalcogen atoms, such as S, Se, and Te.¹⁰ Metal-alloys are also possible, where the varying atomic species is that of the metal atom.^11,12 In $WS 2 x Se 2 ( 1 − x )$¹³ and $MoS 2 x Se 2 ( 1 − x )$⁷ alloys, the compositional dependence of the optical gap energy (derived by photoluminescence, PL, measurements) and the vibrational mode frequencies (derived by Raman spectroscopy) show a smooth dependence on x, with a correspondingly small bowing parameter (i.e., small deviation from a linear behavior).^6,9,14 A sizable bowing was instead found for metal alloys.^6,11,12 While the bandgap and exciton states of TMD alloy MLs (along with information on dark excitons and on the formation of localized excitons acting as quantum emitters) and their vibrational properties have been investigated,^6–14 much less is known about the effects of strain on the materials’ electronic properties and—on strictly related grounds—about the influence of alloying on the crystal’s adhesion energy,¹⁵ which is a quantity of uttermost relevance for the processing of TMDs and the fabrication of TMD-based heterostructures.

In this work, we focus on a chalcogen alloy $WS 2 x Se 2 ( 1 − x )$ ML with $x≈0.5$ (WSSe), subjected to high strain fields. Strain is achieved by exposing WSSe bulk flakes to a low-energy beam of ionized hydrogen that leads to the formation of ML micro-/nano-domes on the flake surface, as previously demonstrated for several TMDs and for hBN.^16,17 The quantitative analysis of the WSSe dome shape permits to show that the elastic parameters and adhesion energy of the WSSe alloy can be evaluated as an average between those of $WS 2$ and $WSe 2$⁠. Micro-photoluminescence (⁠ $μ$-PL) measurements, performed also as a function of temperature, reveal the presence of a direct (A)-to-indirect (I) exciton crossover caused by the strain-induced transition of the valence band (VB) maximum from the K to the $Γ$ point of the first Brillouin zone.¹⁸ The energies of the $A$ and $I$ excitons are close to the average values of the $A$ and $I$ excitons observed in $WS 2$ and $WSe 2$ domes.

WSSe thick flakes and MLs were mechanically exfoliated from bulk crystals grown by the flux zone method and purchased from 2D semiconductors. The WSSe alloy composition was controlled by energy dispersive x-ray analysis (EDX) using a ZEISS-Sigma300 scanning electron microscope (SEM) equipped with an Oxford Instruments X-Act 100 mm energy-dispersive spectrometer. The data were acquired with an acceleration voltage of 28 kV and analyzed by the INCA software. The alloy was found to be homogeneous over the whole investigated crystal with a x value of 0.57, see supplementary note 1 in the supplementary material. $WS 2$ and $WSe 2$ were purchased from 2D semiconductors and HQ graphene.

Strained MLs in the shape of domes were created by H-ion low-energy (⁠ $∼$10 eV) irradiation. Thick (tens to hundreds of layers) flakes of the material were first mechanically exfoliated by the scotch tape method and deposited on $SiO 2$/Si substrates. The substrates with the flakes were then mounted in a vacuum chamber, heated to 150  $°$C, and exposed to H-ions. As detailed in Ref. 16, protons penetrate through the topmost layer of bulk TMD crystals, leading to the production and accumulation of molecular hydrogen in the first interlayer region. The trapped gas coalesces, leading to the formation of micro/nano-domes made of an ML directly lifted from the underlying bulk crystal and filled with highly pressurized $H 2$ (tens to hundreds of atm).

The dome morphology was studied by atomic force microscopy (AFM). AFM measurements were performed in tapping mode. We used a Veeco Digital Instruments Dimension D3100 microscope. The microscope uses monolithic silicon probes with a nominal tip curvature radius of 5–10 nm and a force constant of 40 N/m. All the data were analyzed with the Gwyddion software, and the convolution of the tip used was duly taken into account.

Temperature-dependent $μ$-PL measurements were performed using a closed-circuit He cryostat by Montana Instruments equipped with x–y nano-positiones by Attocube. The excitation laser was provided by a single frequency Nd:YVO $4$ laser (DPSS series by Lasos) emitting at 532 nm. The luminescence signal was spectrally dispersed by a 20.3 cm focal length Isoplane 160 monochromator (Princeton Instruments) equipped with a 150 grooves/mm and a 300 grooves/mm grating and detected by a back-illuminated $N 2$-cooled Si CCD camera (100BRX by Princeton Instruments). The laser light was filtered out by a very sharp long-pass Razor edge filter (Semrock). A 100 $×$ long-working-distance Zeiss objective with NA = 0.75 was employed to excite and collect the light, in a backscattering configuration and using a confocal setup.

Figure 2(b) shows the dependence of the A and I exciton energy on the dome composition (the energy of the direct $A$ exciton of the unstrained ML for the corresponding materials is also displayed as a reference). The values shown in the figure are obtained by an average over different domes and measurements. Like for the elastic parameters discussed before, one may wonder up to what extent the compositional disorder influences the effect of strain on the electronic properties. The data shown in Fig. 2(b) indicate that the energy of the A exciton in the domes varies from $WSe 2$ to $WS 2$ showing a trend close to that of the direct exciton energy of the strain-free MLs. Furthermore, both the direct and indirect excitons exhibit a similar dependence on the sample composition. These findings indicate that the presence of strain does not alter appreciably the way the electronic properties change upon alloying, thus echoing the results discussed in Fig. 1.

The emission properties of the alloyed domes were also studied as a function of temperature. In particular, we investigated the temperature dependence of the $μ$-PL spectra of two WSSe domes, each characterized by different prevailing exciton transitions. Figure 3(a) shows the spectra of the two domes recorded with the same exciting laser power of 20  $μ$W and temperature ranging from T = 5 K to room temperature (290 K). In one case, the indirect I exciton is the dominant band (blue line), and in the other case, only the direct A exciton can be observed (red line). This is likely due to different strain values acting on the dome-shaped WSSe ML. Since the direct-to-indirect transition occurs rather sharply as a function of strain,¹⁸ domes with different aspect ratio (hence, strain value¹⁵) may show predominantly one exciton type or the other. Indeed, the dome ensemble is characterized by a distribution of aspect ratios as displayed in Fig. 1(b). Interestingly, Fig. 3(a) shows that in the range from T = 5 K to 40 K, the spectra of the two domes overlap. As a matter of fact, in this temperature interval, the domes deflate due to the liquid phase of $H 2$ and the PL arises likely from wrinkled WSSe MLs. However, when T increases, liquid $H 2$ turns gas and strain establishes again quickly. The center of mass of the spectrum is shown in Fig. 3(b) for both domes. In the dome displaying the indirect transition, one can notice a faster thermal shift of the emission band as compared to that of the other dome, where the direct exciton dominates. This behavior arises from the nearly double (negative) shift rate induced by the strain in the indirect bandgap energy with respect to the strain shift rate of the direct bandgap, as previously reported.^18,26–28 Indeed, as the temperature increases, $H 2$ expands and transfers an increasingly larger amount of strain to the ML dome.²⁹ In turn, this extra strain adds to the thermal shift of the bandgap of the I exciton in a more relevant manner than for the A exciton. Eventually, at room temperature, the two exciton bands differ by about 150 meV, similar to the data shown in Fig. 2(b).

In this work, we investigated the combined action of alloying and strain in WSSe and compared the effects observed therein with those of the end compounds $WSe 2$ and $WS 2$⁠. To this end, we studied the elastic and electronic properties of ML domes fabricated by H-ion irradiation. The domes turn out to be a practical and effective means to address up to what extent the mechanical deformations alter the way the elastic and electronic properties of TMD chalcogen-alloy MLs change with composition. Our results indicate that strain and composition combine in a “linear” manner and together widen considerably the range of tunability of the mechanical and electronic properties of TMD alloys. Such tunability could be exploited, e.g., for the design of TMD-based mechanical resonators³⁰ with tailored characteristics or for the achievement of strain- and composition-engineered quantum emitters^10,29,31 emitting at specific wavelengths.

See the supplementary material for the elemental analysis of the WSSe crystal investigated (supplementary Note 1) and time-resolved μ-PL measurements of a WSSe dome (supplementary note 2).

A.P. and M.F. acknowledge financial support from the PNRR MUR project No. PE0000023-NQSTI. This project was partly funded within the QuantERA II Programme that has received funding from the European Union’s Horizon 2020 research and innovation programme under Grant Agreement No. 101017733 and with funding organizations Ministero dell’Universitá e della Ricerca (A.P and M.F.) and by Consiglio Nazionale delle Ricerche (G.P. and D.C.).

The authors have no conflicts to disclose.

Salvatore Cianci: Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Validation (equal). Elena Blundo: Conceptualization (lead); Data curation (equal); Formal analysis (lead); Investigation (equal); Methodology (equal); Supervision (lead); Validation (lead); Visualization (lead); Writing – original draft (equal). Federico Tuzi: Investigation (equal); Methodology (equal). Daniele Cecchetti: Data curation (supporting); Investigation (supporting); Methodology (supporting); Validation (supporting). Giorgio Pettinari: Data curation (supporting); Funding acquisition (supporting); Investigation (supporting); Methodology (supporting); Supervision (supporting); Validation (supporting). Marco Felici: Funding acquisition (equal); Supervision (supporting). Antonio Polimeni: Conceptualization (supporting); Funding acquisition (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.