Vapor Phase Synthesis of SnS Facilitated by Ligand-Driven “Launch Vehicle” Effect in Tin Precursors

Extraordinary low-temperature vapor-phase synthesis of SnS thin films from single molecular precursors is attractive over conventional high-temperature solid-state methods. Molecular-level processing of functional materials is accompanied by several intrinsic advantages such as precise control over stoichiometry, phase selective synthesis, and uniform substrate coverage. We report here on the synthesis of a new heteroleptic molecular precursor containing (i) a thiolate ligand forming a direct Sn-S bond, and (ii) a chelating O^N^N-donor ligand introducing a “launch vehicle”-effect into the synthesized compound, thus remarkably increasing its volatility. The newly synthesized tin compound [Sn(SBut)(tfb-dmeda)] 1 was characterized by single-crystal X-ray diffraction analysis that verified the desired Sn:S ratio in the molecule, which was demonstrated in the direct conversion of the molecular complex into SnS thin films. The multi-nuclei (1H, 13C, 19F, and 119Sn) and variable-temperature 1D and 2D NMR studies indicate retention of the overall solid-state structure of 1 in the solution and suggest the presence of a dynamic conformational equilibrium. The fragmentation behavior of 1 was analyzed by mass spectrometry and compared with those of homoleptic tin tertiary butyl thiolates [Sn(SBut)2] and [Sn(SBut)4]. The precursor 1 was then used to deposit SnS thin films on different substrates (FTO, Mo-coated soda-lime glass) by CVD and film growth rates at different temperatures (300–450 °C) and times (15–60 min), film thickness, crystalline quality, and surface morphology were investigated.


Introduction
The monochalcogenide material SnS, which is isostructural to black phosphorous [1], has recently gained much attention due to its unique semiconductor and optoelectronic properties, as well as its low toxicity and the earth-abundant character of tin and sulfur [2][3][4][5]. Two-dimensional (2D) layered monochalcogenides like SnS exhibit a puckered lattice structure with reduced crystal symmetry (orthorhombic) [6] and display structural resemblance to the classical 2D van der Waals material graphene and most of the transition metal dichalcogenides (TMDCs) [7,8]. Their layer-dependent tunable bandgap, high carrier mobility, and strong in-plane anisotropy make these materials promising for emerging next-generation electronic [9] and photonic applications [10,11]. SnS has been actively studied due to its p-type semiconductor characteristics and remarkable optoelectronic properties [12] due to the reported direct bandgap of about 1.3 and indirect bandgap of about 1.0 eV [13].
In many of the above-mentioned compounds, the presence of various element combinations introduces higher ionicity in the bonds, which often leads to higher sublimation temperatures that affect the thermal decomposition profile, leading to multi-step decomposition processes and concomitant formation of secondary phases. Further approaches to circumvent limited volatility include the application of a carrier gas, which is challenging for process engineering and reproducibility. Difficulties may also arise from a mismatch in oxidation states between the metal center (Sn(II) vs. Sn(IV) in the precursor and the final material. Herein, we present the direct synthesis of phase-pure SnS by CVD experiments performed at extraordinary low precursor temperatures (T s ) using the SSP approach. By introducing a tridentate OˆNˆN ligand Htfb-dmeda [(3Z)-4-{[2-(dimethylamino)ethyl]amino}-1,1,1-trifluorobut-3-en-2-one] in pre-synthesized [Sn(SBu t ) 2 ], a monomeric and highly volatile tin(II) complex [Sn(SBu t )(tfb-dmeda)] 1 was obtained, which shows a clean and well-defined decomposition profile and sublimes without leaving any residue. The oxidation state of the Sn II metal center remains unaltered and the element ratio Sn:S 1:1 is the same in the precursor molecule and in the desired final material SnS. The neutral role of the tridentate ligand tfb-dmeda − in the steric stabilization of the complex and its non-participating nature in the thermal decomposition process (no incorporation of N or O in the final products) makes it an ideal "launch vehicle" to deliver Sn and S atoms on the substrate surface during the CVD process.

Synthesis and Characterization of New Molecular Precursor
The motivation of this work was to generate single-phase, nanostructured tin monosulfide SnS by CVD method using an extraordinary volatile molecular precursor. The monoanionic, tridentate coordinating Htfb-dmeda ligand was synthesized following the literature [34] Crystals of 1 suitable for single-crystal X-ray diffraction were obtained by sublimation. Compound 1 crystallized in the triclinic space group P1 with four identical molecules [Sn(SBu t )(tfb-dmeda)] present in the asymmetric unit. Since there are no significant interactions present among these four independent molecules to form a pseudotetramer, only one such molecule is shown in Figure 1. The central Sn(II) atom in 1 is coordinated by one monodentate {SBu t } -and one tridentate {tfb-dmeda} -unit, resulting in a distorted (pseudo)trigonal-bipyramidal coordination geometry for it ( Figure 1). Three equatorial positions are occupied by sulfur S(1), amine nitrogen N(11) and the lone pair of the Sn II metal center. The axial positions are occupied by the carbonyl oxygen O(1) and the nitrogen atoms N(12) of the terminal -NMe2 group. Some selected bond lengths and angles are listed in Tables S1 and S2, respectively, in the Supporting Information.

Single Crystal X-ray Diffraction Analysis of 1
Crystals of 1 suitable for single-crystal X-ray diffraction were obtained by sublimation. Compound 1 crystallized in the triclinic space group P1 with four identical molecules [Sn(SBu t )(tfb-dmeda)] present in the asymmetric unit. Since there are no significant interactions present among these four independent molecules to form a pseudotetramer, only one such molecule is shown in Figure 1. The central Sn(II) atom in 1 is coordinated by one monodentate {SBu t } − and one tridentate {tfb-dmeda} − unit, resulting in a distorted (pseudo)trigonal-bipyramidal coordination geometry for it ( Figure 1). Three equatorial positions are occupied by sulfur S(1), amine nitrogen N(11) and the lone pair of the Sn II metal center. The axial positions are occupied by the carbonyl oxygen O(1) and the nitrogen atoms N(12) of the terminal -NMe 2 group. Some selected bond lengths and angles are listed in Tables S1 and S2, respectively, in the Supporting Information. Crystals of 1 suitable for single-crystal X-ray diffraction were obtained by sublimation. Compound 1 crystallized in the triclinic space group P1 with four identical molecules [Sn(SBu t )(tfb-dmeda)] present in the asymmetric unit. Since there are no significant interactions present among these four independent molecules to form a pseudotetramer, only one such molecule is shown in Figure 1. The central Sn(II) atom in 1 is coordinated by one monodentate {SBu t } -and one tridentate {tfb-dmeda} -unit, resulting in a distorted (pseudo)trigonal-bipyramidal coordination geometry for it ( Figure 1). Three equatorial positions are occupied by sulfur S(1), amine nitrogen N(11) and the lone pair of the Sn II metal center. The axial positions are occupied by the carbonyl oxygen O(1) and the nitrogen atoms N(12) of the terminal -NMe2 group. Some selected bond lengths and angles are listed in Tables S1 and S2, respectively, in the Supporting Information.  The structure of 1 in solution was analyzed by a combination of 1D and 2D NMR experiments of 1 H, 13 C, 19 F, and 119 Sn nuclei. The 1 H-NMR spectrum of 1 in C 6 D 6 solvent recorded at 400 MHz showed ( Figure 2) the formation of a single reaction product that was also supported by the high yield of 1 obtained after sublimation, which excluded the simultaneous presence of other possible species such as fully substituted Sn atoms or oxidative addition product. As a result of deprotonation, the resonance of the (4-H) proton at δ = 6.56 becomes a doublet by vicinal coupling to (H-3). The associated dd (doublet of doublets) satellite signal showed the coupling between these protons and the 119 Sn-nuclei ( 3 J 1H-119Sn = 25 Hz).
Molecules 2021, 26, x FOR PEER REVIEW 5 of 16 observations above reveal the shortest nature of the Sn-S bond length in 1 among all the tin(II) alkylthiolates Sn(SR)2 reported so far.

Spectroscopic Characterization of 1
The structure of 1 in solution was analyzed by a combination of 1D and 2D NMR experiments of 1 H, 13 C, 19 F, and 119 Sn nuclei. The 1 H NMR spectrum of 1 in C6D6 solvent recorded at 400 MHz showed ( Figure 2) the formation of a single reaction product that was also supported by the high yield of 1 obtained after sublimation, which excluded the simultaneous presence of other possible species such as fully substituted Sn atoms or oxidative addition product. As a result of deprotonation, the resonance of the (4-H) proton at δ = 6.56 becomes a doublet by vicinal coupling to (H-3). The associated dd (doublet of doublets) satellite signal showed the coupling between these protons and the 119 Sn-nuclei ( 3 J1H-119Sn = 25 Hz). The signal splitting of the methylene proton resonances in 1 from the saturated region (C-5) and (C-6) of the ligand molecule leads to four distinct chemical shifts δ 1 H. This is interpreted as a result of these protons being part of a more or less rigid N-C5-C6-N(CH3)2 framework, which formed a 5-membered ring by the coordination of the nitrogen atoms to the Sn metal center ( Figure 1). Although clearly under the influence of second-order effects, and probably also by dynamic broadening, the not so fully-resolved multiplet of the (H-5a) and (H-5b) methylene protons (centered at δ = 2.73 and δ = 2.30, respectively) indicated a typical spin-system, that would arise from the relation between axial (endo) and equatorial (exo) protons of a ring structure. The resonance(s) of the methyl protons from the -N(CH3)2 group at δ = 1.83 displayed an increased linewidth (Δ1/2 = 25 Hz), as compared to the free protonated ligand ( Figure S1), pointing towards solution dynamics occurring at rates far below than that of free rotation. The sharp signal at δ = 1.78 is attributed to the dynamic nature of -N(CH3)2 group, since both signal integrals add up to six protons. However, in the case of broad signals, this is not a very reliable argument, and its origin may not be conclusively explained within this study. The signal splitting of the methylene proton resonances in 1 from the saturated region (C-5) and (C-6) of the ligand molecule leads to four distinct chemical shifts δ 1 H. This is interpreted as a result of these protons being part of a more or less rigid N-C 5 -C 6 -N(CH 3 ) 2 framework, which formed a 5-membered ring by the coordination of the nitrogen atoms to the Sn metal center ( Figure 1). Although clearly under the influence of second-order effects, and probably also by dynamic broadening, the not so fully-resolved multiplet of the (H-5 a ) and (H-5 b ) methylene protons (centered at δ = 2.73 and δ = 2.30, respectively) indicated a typical spin-system, that would arise from the relation between axial (endo) and equatorial (exo) protons of a ring structure. The resonance(s) of the methyl protons from the -N(CH 3 ) 2 group at δ = 1.83 displayed an increased linewidth (∆ 1/2 = 25 Hz), as compared to the free protonated ligand ( Figure S1), pointing towards solution dynamics occurring at rates far below than that of free rotation. The sharp signal at δ = 1.78 is attributed to the dynamic nature of -N(CH 3 ) 2 group, since both signal integrals add up to six protons. However, in the case of broad signals, this is not a very reliable argument, and its origin may not be conclusively explained within this study.
Upon cooling the solution of 1 in toluene-d 8 , the resonance at δ = 1.83 (H-7) underwent further broadening and flattened out at T = 268 K; thus the coalescence temperature was noted as T c = 278 K. Below this temperature, separate signals for the exchanging conformers are expected to emerge (further cooling was intended but due to instruments conditions of the VT-unit the full capabilities of the spectrometers could not be fully exploited). In the observed temperature range, no significant shift or change in linewidth was observed ( Figures S2 and S3).
In summary, the proton resonances from the unsaturated carbon moiety are not affected by a dynamic process, and coupling is observed between (H-4) and the central 119 Sn nuclei. The methylene protons at the saturated (C-5) and (C-6) carbon atoms are chemically and magnetically not equivalent, and are part of a conformational equilibrium and by variable temperature (VT) NMR, we could determine the coalescence temperature T c = 278 K of the -N(CH 3 ) 2 group, that shows fluxional behavior in solution.
The 119 Sn{ 1 H}-NMR spectrum at room temperature displayed a single broad resonance at δ = −157 ( Figure S4). The value lies well within the chemical shift region (roughly from +150 to −400 ppm) identified with dimers of tin(II) present in a threefold coordination by sulfur, oxygen, and nitrogen donor atoms in varying element combinations [50]. Since these compounds become four coordinated in the dimer, there is a high resemblance to the bonding situation in 1. In the above-mentioned chemical shift range, an identical set of donor atoms with identical coordinating geometry around the Sn II metal center may be found separated by more than a hundred ppm, which shows the nature of the ligand to be also highly influential [50]. The chemical shift of [Sn(SBu t ) 2 ] is reported at δ = 77 (CH 2 Cl 2 ) and 63 (THF) [51], which could not be observed, probably due to the fast solution dynamics observed in the reported signal's linewidth of ∆ 1/2 = 1 kHz.
The broad linewidth of the δ 119 Sn resonance (∆ 1/2 = 300 Hz) in compound 1 indicated a dynamic exchange present at room temperature ( Figure 3), supporting the observations made by the 1 H-NMR investigations. Upon cooling the solution down to T = 268 K, the chemical shift was observed at δ = −167 (∆ 1/2 = 200 Hz) with narrower linewidth (as compared to the one obtained at room temperature). This slight dependence of the chemical shift on temperature suggested a conformational equilibrium that does not involve changes in the coordination number or geometry of the 119 Sn-nuclei [52]. The observation that in the presence of a nucleophilic solvent such as THF, the chemical shift δ 119 Sn remained constant, showed a stable and saturated coordination sphere at the central Sn II atom. Additionally, in the absence of a concentration-dependence of the δ 119 Sn chemical shift, self-association processes such as aggregation or dimerization become unlikely [52][53][54]. Finally, the coupling between the central 119  Upon cooling the solution of 1 in toluene-d8, the resonance at δ = 1.83 (H-7) underwent further broadening and flattened out at T = 268 K; thus the coalescence temperature was noted as Tc = 278 K. Below this temperature, separate signals for the exchanging conformers are expected to emerge (further cooling was intended but due to instruments conditions of the VT-unit the full capabilities of the spectrometers could not be fully exploited). In the observed temperature range, no significant shift or change in linewidth was observed ( Figure S2 and S3).
In summary, the proton resonances from the unsaturated carbon moiety are not affected by a dynamic process, and coupling is observed between (H-4) and the central 119 Sn nuclei. The methylene protons at the saturated (C-5) and (C-6) carbon atoms are chemically and magnetically not equivalent, and are part of a conformational equilibrium and by variable temperature (VT) NMR, we could determine the coalescence temperature Tc = 278 K of the -N(CH3)2 group, that shows fluxional behavior in solution.
The 119 Sn{ 1 H}-NMR spectrum at room temperature displayed a single broad resonance at δ = −157 ( Figure S4). The value lies well within the chemical shift region (roughly from +150 to −400 ppm) identified with dimers of tin(II) present in a threefold coordination by sulfur, oxygen, and nitrogen donor atoms in varying element combinations [50]. Since these compounds become four coordinated in the dimer, there is a high resemblance to the bonding situation in 1. In the above-mentioned chemical shift range, an identical set of donor atoms with identical coordinating geometry around the Sn II metal center may be found separated by more than a hundred ppm, which shows the nature of the ligand to be also highly influential [50]. The chemical shift of [Sn(SBu t )2] is reported at δ = 77 (CH2Cl2) and 63 (THF) [51], which could not be observed, probably due to the fast solution dynamics observed in the reported signal's linewidth of Δ1/2 = 1 kHz.
The broad linewidth of the δ 119 Sn resonance (Δ1/2 = 300 Hz) in compound 1 indicated a dynamic exchange present at room temperature ( Figure 3), supporting the observations made by the 1 H-NMR investigations. Upon cooling the solution down to T = 268 K, the chemical shift was observed at δ = −167 (Δ1/2 = 200 Hz) with narrower linewidth (as compared to the one obtained at room temperature). This slight dependence of the chemical shift on temperature suggested a conformational equilibrium that does not involve changes in the coordination number or geometry of the 119 Sn-nuclei [52]. The observation that in the presence of a nucleophilic solvent such as THF, the chemical shift δ 119 Sn remained constant, showed a stable and saturated coordination sphere at the central Sn II atom. Additionally, in the absence of a concentration-dependence of the δ 119 Sn chemical shift, self-association processes such as aggregation or dimerization become unlikely [52][53][54]. Finally, the coupling between the central 119   To summarize the results from the NMR experiments, we showed that the structure of 1 observed in the solid state is maintained in solution and suggest the presence of To summarize the results from the NMR experiments, we showed that the structure of 1 observed in the solid state is maintained in solution and suggest the presence of dynamic conformational equilibrium in the temperature range between T = 298-268 K as displayed in Figure 3. If 1 is kept in solution for a prolonged duration, gradual formation of the oxidized product [Sn(SBu t ) 4 ] along with disulfide is observed. However, if stored under inert conditions, 1 remained stable without any detectable signs of degradation for several months. The fragmentation behavior of 1 was analyzed by mass spectrometry. For simplicity, the OˆNˆN-donor ligand moiety in compound 1 is abbreviated as L (L= tfb-dmeda − ) and, therefore, 1 and its fragment notation as molecular peak are written as (L)-Sn-(SBu t ) and [L-Sn-SC 4  There is no fragment ion with a tin atom bound to oxygen or nitrogen, unlike the complex [Re(tfb-dmeda)(CO) 3 ], where partial decomposition of the tfb-dmeda ligand while being bonded to the metal center has been observed by mass spectrometry [34]. dynamic conformational equilibrium in the temperature range between T = 298-268 K as displayed in Figure 3. If 1 is kept in solution for a prolonged duration, gradual formation of the oxidized product [Sn(SBu t )4] along with disulfide is observed. However, if stored under inert conditions, 1 remained stable without any detectable signs of degradation for several months.
The fragmentation behavior of 1 was analyzed by mass spectrometry. For simplicity, the O^N^N-donor ligand moiety in compound 1 is abbreviated as L (L= tfb-dmeda -) and, therefore, 1 and its fragment notation as molecular peak are written as  . For the latter compound, no literature data was available, therefore, it was analyzed for comparison purposes in this work. All three compounds share at least partially a common decomposition pathway.
For [Sn(SBu t )4], the consecutive loss of alkyl groups in the decomposition pattern has been reported earlier [55]. The loss of alkyl groups in this context is equivalent to the preferred retainment of the Sn-S bond. From Scheme 2, this becomes obvious by the sheer number of fragments with a Sn-S bond present in the gas phase. Our mass spectrometric analysis showed that the fragmentation pattern of [Sn(SBu t )4] could partially be found also in the fragmentation pattern of [Sn(SBu t )2] ( Figure S7). [Sn(SBu t )2] is, as [Sn(SR)2] (R = alkyl) in general, prone to oxidation upon heating. The experimental conditions, namely the sublimation step prior to fragmentation in the mass spectrometer, may explain the observation of tin(IV) species in the gas phase.
The observation was different for 1, which is sublimable without decomposition or any other undesired side-reaction. This was confirmed by repeated resublimation, where 1 was retrieved quantitatively after each sublimation step, and its identity was confirmed by NMR spectroscopy of the sublimate. To exclude the presence of possible contamination . For the latter compound, no literature data was available, therefore, it was analyzed for comparison purposes in this work. All three compounds share at least partially a common decomposition pathway.
For [Sn(SBu t ) 4 ], the consecutive loss of alkyl groups in the decomposition pattern has been reported earlier [55]. The loss of alkyl groups in this context is equivalent to the preferred retainment of the Sn-S bond. From Scheme 2, this becomes obvious by the sheer number of fragments with a Sn-S bond present in the gas phase. Our mass spectrometric analysis showed that the fragmentation pattern of [Sn(SBu t ) 4 ] could partially be found also in the fragmentation pattern of [Sn(SBu t ) 2 ] ( Figure S7). [Sn(SBu t ) 2 ] is, as [Sn(SR) 2 ] (R = alkyl) in general, prone to oxidation upon heating. The experimental conditions, namely the sublimation step prior to fragmentation in the mass spectrometer, may explain the observation of tin(IV) species in the gas phase.
The observation was different for 1, which is sublimable without decomposition or any other undesired side-reaction. This was confirmed by repeated resublimation, where 1 was retrieved quantitatively after each sublimation step, and its identity was confirmed by NMR spectroscopy of the sublimate. To exclude the presence of possible contamination by oxidized species, such as Sn(SBu t ) 4 , the purity of 1 and Sn(SBu t ) 2 was confirmed by 1 Hand 119 Sn-NMR spectroscopy prior to the measurements.
An investigation on the exact mechanisms that proceed under EI conditions in the gas phase is beyond the scope of this work and was not further looked into. [Sn(SPh) 4 ] and [Sn(SBu t ) 4 ] have been reported to yield Sn 2 O 3 films by CVD [21,45]. Given the similarities observed here in the fragments of 1, [Sn(SBu t ) 4 ] and [Sn(SBu t ) 2 ], the reported tin oxide formation by gas-phase deposition may likely not be an inherent property of the molecules, but rather due to experimental conditions such as the usage and nature of solvent during the aerosol-assisted CVD process. However, the application of [Sn(SR) 2 ] and [Sn(SR) 4 ] (R = alkyl, phenyl) for CVD remains limited since these compounds lack the desired sublimation properties.
by oxidized species, such as Sn(SBu t )4, the purity of 1 and Sn(SBu t )2 was confirmed by 1 Hand 119 Sn-NMR spectroscopy prior to the measurements.
An investigation on the exact mechanisms that proceed under EI conditions in the gas phase is beyond the scope of this work and was not further looked into. [Sn(SPh)4] and [Sn(SBu t )4] have been reported to yield Sn2O3 films by CVD [21,45]. Given the similarities observed here in the fragments of 1, [Sn(SBu t )4] and [Sn(SBu t )2], the reported tin oxide formation by gas-phase deposition may likely not be an inherent property of the molecules, but rather due to experimental conditions such as the usage and nature of solvent during the aerosol-assisted CVD process. However, the application of [Sn(SR)2] and [Sn(SR)4] (R = alkyl, phenyl) for CVD remains limited since these compounds lack the desired sublimation properties.

Thermogravimetric analysis of 1
Thermogravimetric analysis was performed under continuous N2(g) flow to further investigate the decomposition behavior of 1 in relation to the temperature ( Figure 5). Interestingly, the thermal decomposition of 1 occurred in one well-defined single step, as indicated by a steep mass loss in the TGA curve in the narrow temperature range between T = 250-275 °C. No significant mass difference or thermodynamic event was registered beyond 300°C that indicated the formation of a solid of definite composition. The experimental mass loss determined by TG-DSC is ∆mexp. = 57%, which is below the theoretical value ∆mth. = 64%, assuming the formation of SnS as the final product. Onestep decomposition of 1 at remarkably low temperatures underlines its potential as a promising molecular precursor in CVD experiments.

Thermogravimetric Analysis of 1
Thermogravimetric analysis was performed under continuous N 2 (g) flow to further investigate the decomposition behavior of 1 in relation to the temperature ( Figure 5). Interestingly, the thermal decomposition of 1 occurred in one well-defined single step, as indicated by a steep mass loss in the TGA curve in the narrow temperature range between T = 250-275 • C. No significant mass difference or thermodynamic event was registered beyond 300 • C that indicated the formation of a solid of definite composition. The experimental mass loss determined by TG-DSC is ∆m exp. = 57%, which is below the theoretical value ∆m th. = 64%, assuming the formation of SnS as the final product. One-step decomposition of 1 at remarkably low temperatures underlines its potential as a promising molecular precursor in CVD experiments.
by oxidized species, such as Sn(SBu t )4, the purity of 1 and Sn(SBu t )2 was confirmed by 1 Hand 119 Sn-NMR spectroscopy prior to the measurements.
An investigation on the exact mechanisms that proceed under EI conditions in the gas phase is beyond the scope of this work and was not further looked into. [Sn(SPh)4] and [Sn(SBu t )4] have been reported to yield Sn2O3 films by CVD [21,45]. Given the similarities observed here in the fragments of 1, [Sn(SBu t )4] and [Sn(SBu t )2], the reported tin oxide formation by gas-phase deposition may likely not be an inherent property of the molecules, but rather due to experimental conditions such as the usage and nature of solvent during the aerosol-assisted CVD process. However, the application of [Sn(SR)2] and [Sn(SR)4] (R = alkyl, phenyl) for CVD remains limited since these compounds lack the desired sublimation properties.

Thermogravimetric analysis of 1
Thermogravimetric analysis was performed under continuous N2(g) flow to further investigate the decomposition behavior of 1 in relation to the temperature ( Figure 5). Interestingly, the thermal decomposition of 1 occurred in one well-defined single step, as indicated by a steep mass loss in the TGA curve in the narrow temperature range between T = 250-275 °C. No significant mass difference or thermodynamic event was registered beyond 300°C that indicated the formation of a solid of definite composition. The experimental mass loss determined by TG-DSC is ∆mexp. = 57%, which is below the theoretical value ∆mth. = 64%, assuming the formation of SnS as the final product. Onestep decomposition of 1 at remarkably low temperatures underlines its potential as a promising molecular precursor in CVD experiments.

CVD Experiments
The CVD experiments were conducted in a quartz glass tube fitted with an inductively heated graphite holder, where the substrates were mounted on. The precursor temperature (T p ) for 1 was maintained between T p = 90-130 • C and substrate temperatures chosen in this study were T s = 300-450 • C that resulted in crystalline deposits of homogeneous SnS films.
At precursor temperature of T p = 130 • C and substrates maintained at 350 • C, the FTO substrate was uniformly covered by a black layer within 30 min of the processing time. The deposit was identified by XRD analysis to be SnS (Figure 6a) that exhibited diffraction peaks corresponding to (040) plane, along with a minor peak due to the (111)-plane. This showed the preferred orientation of grains in SnS films, namely along (040)-planes, parallel to the substrate surface. The SEM images (Figure 6b,c) showed the uniform coverage of the substrate surface by flake-like nanostructures. From the cross-sectional SEM (Figure 6d), the film thickness was measured to be~80 nm. The (040) surface in SnS is the lowest energy surface [56] and, therefore, favored by kinetically limited growth conditions, such as lower temperatures and sufficiently long diffusion lengths. These low energy surfaces are bonded by van der Waals forces, and an orientation parallel to the substrate surface is favored by corresponding substrates, as previously reported [57]. On transition metals like molybdenum, covalent edge bonding is favored, resulting in (040) planes perpendicular and (101) surface parallel to the substrate surface, which results in the formation of plates growing out of the substrate plane [58]. The strong adhesion of the deposited SnS on FTO is possibly due to a different bonding situation than the aforementioned Van der Waals interaction, for instance, the formation of a mixed tin oxide-chalcogenide SnO 1−x S x phase at the interface between the substrate and the film. The mechanism that leads to the orientational growth on FTO is not yet fully understood and is subject to further investigations. temperature (Tp) for 1 was maintained between Tp = 90-130 °C and substrate temperatures chosen in this study were Ts = 300-450 °C that resulted in crystalline deposits of homogeneous SnS films.
At precursor temperature of Tp = 130 °C and substrates maintained at 350 °C, the FTO substrate was uniformly covered by a black layer within 30 min of the processing time. The deposit was identified by XRD analysis to be SnS (Figure 6a) that exhibited diffraction peaks corresponding to (040) plane, along with a minor peak due to the (111)-plane. This showed the preferred orientation of grains in SnS films, namely along (040)-planes, parallel to the substrate surface. The SEM images (Figure 6 b,c) showed the uniform coverage of the substrate surface by flake-like nanostructures. From the cross-sectional SEM (Figure 6d), the film thickness was measured to be ~ 80 nm. The (040) surface in SnS is the lowest energy surface [56] and, therefore, favored by kinetically limited growth conditions, such as lower temperatures and sufficiently long diffusion lengths. These low energy surfaces are bonded by van der Waals forces, and an orientation parallel to the substrate surface is favored by corresponding substrates, as previously reported [57]. On transition metals like molybdenum, covalent edge bonding is favored, resulting in (040) planes perpendicular and (101) surface parallel to the substrate surface, which results in the formation of plates growing out of the substrate plane [58]. The strong adhesion of the deposited SnS on FTO is possibly due to a different bonding situation than the aforementioned Van der Waals interaction, for instance, the formation of a mixed tin oxide-chalcogenide SnO1-xSx phase at the interface between the substrate and the film. The mechanism that leads to the orientational growth on FTO is not yet fully understood and is subject to further investigations. In view of the application of SnS in photovoltaic devices and suitable band alignment, we chose to investigate the CVD process on molybdenum substrates. Herein, we use molybdenum layers of a uniform thickness that were sputtered on soda-lime glass (Mo-SLG). Under similar deposition conditions as described for the SnS deposition onto FTO In view of the application of SnS in photovoltaic devices and suitable band alignment, we chose to investigate the CVD process on molybdenum substrates. Herein, we use molybdenum layers of a uniform thickness that were sputtered on soda-lime glass (Mo-SLG). Under similar deposition conditions as described for the SnS deposition onto FTO (i.e., T p = 130 • C, T s = 350 • C), successful deposition of the crystalline SnS was observed onto Mo-SLG substrates, as confirmed by XRD measurements (Figure 7a). The SEM investigation of as-deposited films showed full coverage of the substrate (Figure 7 b,c). (i.e., Tp = 130 °C, Ts = 350 °C), successful deposition of the crystalline SnS was observed onto Mo-SLG substrates, as confirmed by XRD measurements (Figure 7a). The SEM investigation of as-deposited films showed full coverage of the substrate (Figure 7 b,c). The film thickness was controlled by the process time (Figure 8a-c). By varying deposition times (15,30, and 60 min), a linear relationship between the growth rate and deposition time was established (Figure 8d). The thicknesses were measured as 440 nm, 1 µm and 2 µm, thus, in the given process conditions the layers grew around 33 nm per minute, which is in good agreement with the film growth reported in the literature before [57]. The different SnS films, thicknesses were measured using SEM cross-sectional analyses (Figure 8a-c). The low dimensional platelets of SnS were found to be vertically oriented to the substrate surface, as clearly seen in Figure 8c. This morphological attribute offers a huge number of active sites of the layered SnS phase that consequently increases the catalytic behavior of as-deposit material potentially interesting for the application in photovoltaic systems [59]. To analyze the films further by high-resolution transmission electron microscopy (HRTEM) and selected area electron diffraction (SAED), parts of the film deposited on molybdenum substrate for 15 min at T = 350 °C were scratched off the substrate surface. The HRTEM image shows clear lattice fringes, therefore, further confirms the high crystallinity of the material (Figure 9). The diffraction spots in the SAED pattern are indexed consistently with the d-spacings of orthorhombic SnS (COD # 9008785). The indexed planes correspond to the [0 -1 2] zone axis presented as a ring diffraction pattern.  The film thickness was controlled by the process time (Figure 8a-c). By varying deposition times (15,30, and 60 min), a linear relationship between the growth rate and deposition time was established (Figure 8d). The thicknesses were measured as 440 nm, 1 µm and 2 µm, thus, in the given process conditions the layers grew around 33 nm per minute, which is in good agreement with the film growth reported in the literature before [57]. The different SnS films, thicknesses were measured using SEM cross-sectional analyses (Figure 8a-c). The low dimensional platelets of SnS were found to be vertically oriented to the substrate surface, as clearly seen in Figure 8c. This morphological attribute offers a huge number of active sites of the layered SnS phase that consequently increases the catalytic behavior of as-deposit material potentially interesting for the application in photovoltaic systems [59]. To analyze the films further by high-resolution transmission electron microscopy (HRTEM) and selected area electron diffraction (SAED), parts of the film deposited on molybdenum substrate for 15 min at T = 350 • C were scratched off the substrate surface. The HRTEM image shows clear lattice fringes, therefore, further confirms the high crystallinity of the material (Figure 9). The diffraction spots in the SAED pattern are indexed consistently with the d-spacings of orthorhombic SnS (COD # 9008785). The indexed planes correspond to the [0 −1 2] zone axis presented as a ring diffraction pattern.
(i.e., Tp = 130 °C, Ts = 350 °C), successful deposition of the crystalline SnS was observed onto Mo-SLG substrates, as confirmed by XRD measurements (Figure 7a). The SEM investigation of as-deposited films showed full coverage of the substrate (Figure 7 b,c). The film thickness was controlled by the process time (Figure 8a-c). By varying deposition times (15,30, and 60 min), a linear relationship between the growth rate and deposition time was established (Figure 8d). The thicknesses were measured as 440 nm, 1 µm and 2 µm, thus, in the given process conditions the layers grew around 33 nm per minute, which is in good agreement with the film growth reported in the literature before [57]. The different SnS films, thicknesses were measured using SEM cross-sectional analyses (Figure 8a-c). The low dimensional platelets of SnS were found to be vertically oriented to the substrate surface, as clearly seen in Figure 8c. This morphological attribute offers a huge number of active sites of the layered SnS phase that consequently increases the catalytic behavior of as-deposit material potentially interesting for the application in photovoltaic systems [59]. To analyze the films further by high-resolution transmission electron microscopy (HRTEM) and selected area electron diffraction (SAED), parts of the film deposited on molybdenum substrate for 15 min at T = 350 °C were scratched off the substrate surface. The HRTEM image shows clear lattice fringes, therefore, further confirms the high crystallinity of the material (Figure 9). The diffraction spots in the SAED pattern are indexed consistently with the d-spacings of orthorhombic SnS (COD # 9008785). The indexed planes correspond to the [0 -1 2] zone axis presented as a ring diffraction pattern.

Materials and Methods
All experiments were carried out under strict exclusion of moisture and oxygen under an atmosphere of dry N2 (g) using stock high vacuum apparatus. Solvents were dried by refluxing over sodium and benzophenone and freshly distilled prior to use and stored over sodium wire. HSBu t (2-methyl-2-propanethiol) was purchased from Sigma-Aldrich (99%) and used without further purification. [Sn{N(SiMe3)2}2] was synthesized as reported earlier [35]. [Sn(SBu t )2] was synthesized via the thiolysis of the stannylene [Sn{N(SiMe3)}2] in toluene at T = 0 °C under exclusion of ambient light. After removing all volatile components, bright yellow needle-like crystals were obtained and used without further purification. The ligand Htfb-dmeda (3Z)-4-{[2-(dimethylamino)ethyl]amino}-1,1,1trifluorobut-3-en-2-one, was provided by the working group members and checked by 19 F-, 13 C-and 1 H-NMR prior to use. A detailed synthetic procedure for ligand is described in the reference [34].
NMR spectra were recorded on a Bruker (Billerica, MA, USA) Model Avance II 300 and Avance 400 spectrometer. 1 H-NMR and 13 C-NMR chemical shifts are reported in parts per million relative to external tetramethylsilane (TMS) and are referenced internally to residual proton impurity or 13 C signal of the deuterated lock solvent. Assignment of all 13 C-NMR signals was made on the basis of 1 H, 13 C-HMQC, 1 H, 13 C-HMBC, and 19 F, 13 C-HMBC experiments. The 119 Sn{ 1 H} spectra were calibrated externally to [Sn{N(SiMe3)2}]2 [35] in C6D6 (δ 119 Sn = 775) and are reported relative to Sn(CH3)4 in C6D6 (δ = 0 ).
Elemental analyses were carried out on a HEKAtech CHNS Euro EA 3000 (Wegberg, Germany). TG-DSC measurements were performed on a TGA/DSC1 (Mettler-Toledo GmbH, Gießen, Germany) apparatus. Mass spectrometric data were recorded on a Finnigan MAT 95 (EI, 20 eV) and are reported as m/z (rel. intensity in %). High-resolution TEM (HRTEM) and SAED measurements were carried out on a JEOL JEM-2200FS microscope (Peabody, London, UK).
Data collection for X-ray structure elucidation was performed on a STOE IPDS 2T diffractometer (Darmstadt, Germany) using graphite-monochromated Mo-Kα radiation (0.71071 Å). The data were corrected for Lorentz and polarization effects. A numerical absorption correction based on crystal-shape optimization was applied for all data. The

Materials and Methods
All experiments were carried out under strict exclusion of moisture and oxygen under an atmosphere of dry N 2 (g) using stock high vacuum apparatus. Solvents were dried by refluxing over sodium and benzophenone and freshly distilled prior to use and stored over sodium wire. HSBu t (2-methyl-2-propanethiol) was purchased from Sigma-Aldrich (99%) and used without further purification. [Sn{N(SiMe 3 ) 2 } 2 ] was synthesized as reported earlier [35]. [Sn(SBu t ) 2 ] was synthesized via the thiolysis of the stannylene [Sn{N(SiMe 3 )} 2 ] in toluene at T = 0 • C under exclusion of ambient light. After removing all volatile components, bright yellow needle-like crystals were obtained and used without further purification. The ligand Htfb-dmeda (3Z)-4-{[2-(dimethylamino)ethyl]amino}-1,1,1trifluorobut-3-en-2-one, was provided by the working group members and checked by 19 F-, 13 C-and 1 H-NMR prior to use. A detailed synthetic procedure for ligand is described in the reference [34].
NMR spectra were recorded on a Bruker (Billerica, MA, USA) Model Avance II 300 and Avance 400 spectrometer. 1 H-NMR and 13 C-NMR chemical shifts are reported in parts per million relative to external tetramethylsilane (TMS) and are referenced internally to residual proton impurity or 13 C signal of the deuterated lock solvent. Assignment of all 13 C-NMR signals was made on the basis of 1 H, 13 C-HMQC, 1 H, 13 C-HMBC, and 19 F, 13 C-HMBC experiments. The 119 Sn{ 1 H} spectra were calibrated externally to [Sn{N(SiMe 3 ) 2 }] 2 [35] in C 6 D 6 (δ 119 Sn = 775) and are reported relative to Sn(CH 3 ) 4 in C 6 D 6 (δ = 0 ).
Elemental analyses were carried out on a HEKAtech CHNS Euro EA 3000 (Wegberg, Germany). TG-DSC measurements were performed on a TGA/DSC1 (Mettler-Toledo GmbH, Gießen, Germany) apparatus. Mass spectrometric data were recorded on a Finnigan MAT 95 (EI, 20 eV) and are reported as m/z (rel. intensity in %). High-resolution TEM (HRTEM) and SAED measurements were carried out on a JEOL JEM-2200FS microscope (Peabody, London, UK).
Data collection for X-ray structure elucidation was performed on a STOE IPDS 2T diffractometer (Darmstadt, Germany) using graphite-monochromated Mo-K α radiation (0.71071 Å). The data were corrected for Lorentz and polarization effects. A numerical absorption correction based on crystal-shape optimization was applied for all data. The software used in this work are STOE's X-Area, SHELXL [60,61], and ShelXle [62] (Version 93, Darmstadt, Germany). CCDC 2102248 contains the supplementary crystallographic data for this article. These data can be obtained free of charge from The Cambridge Crystallographic Data Centre, via www.ccdc.cam.ac.uk/data_request/cif, accessed on 10 August 2021.
The MOCVD experiments were conducted in a cold-wall CVD reactor under lowpressure conditions (p < 10 −6 bar), equipped with cooling traps, an internal pressure sensor and external temperature sensors. More detailed information can be found in previous publications from our working group [63][64][65].

Conclusions
We report here a new heteroleptic precursor [Sn(SBu t )(tfb-dmeda)] 1, which was characterized thorougly in the solution-and solid-phase by multi-nuclei NMR and single crystal X-ray structure and studied for thermal and fragmentation behaviors by thermogravimetric and mass spectrometry, respectively. As compared to corresponding homoleptic tin(IV) alkylthiolates, which could not be successfully applied in the CVD process, the heteroleptic configuration of 1 and the resulting increase in the Sn-S bond strength ensures a predictable decomposition pattern for it. So while the tertiarybutyl thiolate ligand decomposes in a neat and clean manner due to the presence of an inherent decomposition mechanism [66], the monoanionic tridentate chelate ligand departs in a clean manner without changing the electronic structure of the Sn(II) center or contaminating the growing SnS film, for example, through probable incorporation of heteroatoms (N, O). This, along with the low precursor temperature, enables deposition of high-quality SnS thin films by CVD under mild conditions. While the use of FTO as substrate generated a preferred orientation of the deposited SnS films along the (040) plane, the films deposited on Mo-coated soda-lime substrate showed no such preference. In both cases, the deposited films were well-crystalline, dense, and of uniform thickness. This work showing the "launch vehicle" role of the auxiliary ligand underlines the potential of design and subtle choice of ligands and coligands in synthesizing efficient molecular precursors for chemical vapor deposition.
Supplementary Materials: The following are available online: Figure S1: Comparison of the 1 H-NMR spectra of Htfb-dmeda and 1a in C 6 D 6 and recorded at 300 MHz, Figure S2: 1 H-NMR variable temperature spectra of 1 in toluene-d 8 . The bottom spectrum at T = 298 K was recorded after cooling when the NMR tube was warmed up to room temperature again, showing the full reversibility of the observed equilibria, Figure S3: Enlarged part of the 1 H-NMR variable temperature spectra of 1 ( Figure S2) for better visualization, Figure S4: Variable temperature 119 Sn-NMR spectra of 1. The bottom spectrum at T = 298 K was recorded at room temperature, after cooling, showing the initial chemical shift is restituted, Figure S5: 119 Sn, 1 H-HMBC spectrum of 1, recorded at 300 MHz, Figure S6: Experimental EI-MS fragmentation pattern of 1, Figure S7: Experimental EI-MS fragmentation pattern of Sn(SBu t ) 2 , Table S1: Selected bond lengths of 1, Table S2  Data Availability Statement: All data presented is available in this manuscript.