Two new polymorphs of niobium trisulfide are established by single crystal x-ray diffraction. NbS3-iv crystallizes in the monoclinic space group P21/c with lattice parameters a = 6.7515(5) Å, b = 4.9736(4) Å, c = 18.1315(13) Å, and β = 90.116(2)°. Its structure is based on chains of [NbS6] trigonal prisms containing Nb–Nb pairs with a bond length of 3.0448(8) Å; this pairing causes the chains to corrugate slightly along their axis, a feature also present in triclinic NbS3-i that leads to semiconductor properties. The stacking arrangement of chains is different in these polymorphs, however, with NbS3-i having an ABCDE repeating sequence of chain bilayers and NbS3-iv having an AB repeating sequence. HRTEM studies show the presence of topotactically-oriented intergrown zones and numerous dislocations, which result in mosaic structuring. A second new polymorph, NbS3-v, crystallizes in the monoclinic space group P21/m with lattice parameters a = 4.950(5) Å, b = 3.358(4) Å, c = 9.079(10) Å, β = 97.35(2)°. In contrast to NbS3-iv, NbS3-v maintains fixed a Nb–Nb bond distance of 3.358(4) Å along the chains, and it has an ABCDE repeating sequence of chain bilayers similar to NbS3-i. High resolution scanning transmission electron microscopy (HR-STEM) imaging of an exfoliated NbS3-v nanoribbon shows the continuous [NbS6] chains oriented along the b-axis. These results provide the first firmly established structural data for monoclinic NbS3. In addition, SEM images show the formation of NbS3 rings and cylinders, and a combination of powder x-ray diffraction and Raman spectroscopy provides a way to distinguish between NbS3 polymorphs.

Many classes of materials have been proposed as replacements for semiconductor-based electronics technology as it progresses in complexity and miniaturization.1 Transition metal chalcogenides are leading candidates in the current “beyond-graphene” milieu.2 Metal dichalcogenide (MX2) and trichalcogenide (MX3) compositions, together with their polymorphs and alloys, provide a diverse set of materials exhibiting low dimensionality,3–5 wide ranging electrical properties,6–9 and reasonable processing options.10,11 In recent work, we and others have investigated nanostructured MX3 materials for electronic device applications.6,7,9,12,13 The unique properties of these materials result from their dual low-dimensional nature, i.e., quasi-one-dimensional and two-dimensional (Q1D-2D). Their primary structural unit is the [MX6] trigonal prism, which is fused end-to-end into continuous chains; this motif leads to quasi-one-dimensional electronic character as well as distinct anisotropy. Additional metal-chalcogenide interactions between neighboring chains result in bilayer sheets that can be flat (e.g., ZrSe3) or corrugated (e.g., TaSe3). These two-dimensional sheets, stacked and held together by van der Waals bonding, then make up the three-dimensional structure. Depending on the composition, structure, and morphology of any particular MX3 sample, either one- or two-dimensional character may dominate. Much like MX2 materials, MX3 materials are amenable to mechanical or chemical exfoliation along the van der Waals gap,6,14 and recent studies show that their properties can be maintained or even enhanced by nanostructuring.12,15

Here we consider NbS3, arguably the least-well-understand metal trichalcogenide. The niobium–sulfur system contains a rich cornucopia of phases, including an unknown number of NbS3 polymorphs, varying compositions from NbS to Nb21S8, and the possibility of non-stoichiometric phases.16–19 Offsetting this complexity and piquing our interest are intriguing reports about the electrical and charge density wave (CDW) properties of certain NbS3 samples, including multiple CDW transitions and room temperature charge transport.3,20,21 However, further progress has been impeded by the absence of well-established structural information, spectroscopic characterization, or even reproducible synthetic details. Therefore, in this contribution, we aim to clarify the polymorphism of NbS3 as part of the foundation required to understand its electronic properties.

To date, NbS3 has been reported to crystallize in four forms: types I, II, III, and a high-pressure phase (Table I). Jellinek and co-workers first proposed a monoclinic ZrSe3-type NbS3 structure in 1960.16 They reported a definitive structure in 1978, using single crystal x-ray diffraction data to refine the NbS3 structure with a triclinic unit cell (notably having α = γ = 90°).22 This polymorph is known as NbS3-i in the literature. Compared to other MX3 structures, NbS3-i has several unique features. The most remarkable feature is Nb–Nb pairing along the NbS3 chain axis that manifests as alternating Nb–Nb bonds (∼3.0 Å) and gaps (∼3.7 Å). This linear Peierls distortion at room temperature causes NbS3-i to be semiconducting, with an experimentally determined bandgap in the range of 0.66–1 eV (based on transport, optical transmittance, and photoconduction measurements), rather than metallic.23–27 Another consequence is the doubling of the b-axis (6.730 Å) compared to other MX3 structures (typically ∼3.3 Å), which is needed to accommodate Nb–Nb pairs in the unit cell (Fig. 1). A recent report by Fedorov and co-workers demonstrated the solution-based exfoliation of NbS3-i into nanoribbons.11 

TABLE I.

Unit cell and synthesis parameters for previously reported NbS3 polymorphs as well as new polymorphs NbS3-iv and NbS3-v.

Unit cell parameters
PolymorphSynthesis conditionsSpace groupa, b, c, (Å)α, β, γ (deg)ZNbNb (Å)References
NbS3-i NbS2Cl2; 588 → 569 °C,  4.963 90  3.045 22  
 48 h; slow cool P1¯ 6.730 97.17 3.702  
   9.144 90    
NbS3-HP Nb + S; 700 °C at  9.68 90    
 2 GPa, 0.5 h P21/m 3.37 109.9 3.370 29  
   14.83 90    
NbS3-ii Nb + S; 600 →  9.9 90    
(proposed) 580 °C, 15 days P21/m 3.4 97 … … 26 and 27  
   18.3 90    
NbS3-ii Nb + S; 500 °C … 9.19.6 90    
(proposed)   18.719.9 9798 … 25  
   3.4 90    
NbS3-iii NbS2 + S; 550 °C, 21 days,  ∼5 90    
(proposed) 400 °C 48 h, air quench … … 9899 … … 20  
   ∼9 90    
NbS3-iv Nb + S, I2 transport; 670 →  6.7515(5) 90  3.0448(8)  
 570 °C, 10 days P21/c 4.9736(4) 90.116(2) 4 (8) 3.7087(8) This work 
   18.1315(13) 90    
NbS3-v Nb + S, 10% S transport;  4.950(5) 90    
 700 → 670 °C, 10 days P21/m 3.358(4) 97.35(2) 3.358(4) This work 
   9.079(10) 90    
Unit cell parameters
PolymorphSynthesis conditionsSpace groupa, b, c, (Å)α, β, γ (deg)ZNbNb (Å)References
NbS3-i NbS2Cl2; 588 → 569 °C,  4.963 90  3.045 22  
 48 h; slow cool P1¯ 6.730 97.17 3.702  
   9.144 90    
NbS3-HP Nb + S; 700 °C at  9.68 90    
 2 GPa, 0.5 h P21/m 3.37 109.9 3.370 29  
   14.83 90    
NbS3-ii Nb + S; 600 →  9.9 90    
(proposed) 580 °C, 15 days P21/m 3.4 97 … … 26 and 27  
   18.3 90    
NbS3-ii Nb + S; 500 °C … 9.19.6 90    
(proposed)   18.719.9 9798 … 25  
   3.4 90    
NbS3-iii NbS2 + S; 550 °C, 21 days,  ∼5 90    
(proposed) 400 °C 48 h, air quench … … 9899 … … 20  
   ∼9 90    
NbS3-iv Nb + S, I2 transport; 670 →  6.7515(5) 90  3.0448(8)  
 570 °C, 10 days P21/c 4.9736(4) 90.116(2) 4 (8) 3.7087(8) This work 
   18.1315(13) 90    
NbS3-v Nb + S, 10% S transport;  4.950(5) 90    
 700 → 670 °C, 10 days P21/m 3.358(4) 97.35(2) 3.358(4) This work 
   9.079(10) 90    
FIG. 1.

Views of NbS3-i, iv, and v: from left to right including unit cell, chain cross section, layer, and perspective views. The NbS3 bilayers are differentiated by color to illustrate the van der Waals gaps, which is further illustrated with a red plane in each perspective view. The structures of these three polymorphs have been established by single crystal x-ray diffraction.

FIG. 1.

Views of NbS3-i, iv, and v: from left to right including unit cell, chain cross section, layer, and perspective views. The NbS3 bilayers are differentiated by color to illustrate the van der Waals gaps, which is further illustrated with a red plane in each perspective view. The structures of these three polymorphs have been established by single crystal x-ray diffraction.

Close modal

A second polymorph, NbS3-ii, was identified in 1978 based on electron diffraction studies, which showed inter-row pairs of satellite diffraction lines that resolved into rows of spots with varying temperature.28 These rows occurred at irrational positions with similar separations as NbS3-i, leading to the proposal that NbS3-ii was a superstructure of NbS3-i with different chain stacking. Several years later, Prodan and co-workers introduced a model monoclinic NbS3-ii unit cell with lattice parameters 2a,b/2,2c, β′ based on the NbS3-i unit cell (lattice parameters a, b, c, β) (Table I).29,30 In parallel, Roucau and co-workers proposed a closely related model monoclinic NbS3-ii unit cell with lattice parameters 2a,2c,b/2, β′′ based on a different multiple of the NbS3-i unit cell (Table I).26,31,32 In the almost forty years since NbS3-ii was identified, however, its atomic structure has remained unsolved. During this time, continued interest in this polymorph has been motivated by its three CDWs: a low temperature transition at 150 K, a second transition at 330-370 K, and a high temperature transition at 620-650 K.21,26 The transition at 330-370 K, just above room temperature, is an unusual one for CDW transport, and according to Zybtsev and co-workers, this CDW exhibits high transport velocities and sliding coherence.21 

A high-pressure modification of NbS3 prepared at 700 °C with 2 GPa pressure was reported by Kikkawa and co-workers in 1982.33 The monoclinic unit cell deduced from powder x-ray diffraction data is similar to that of monoclinic NbSe3 (Table I), but atomic positions for NbS3-HP could not be determined.33 Yet another polymorph, designated NbS3-iii, was reported by Zettl and co-workers in 1982. In this case, powder x-ray diffraction data provided only partial lattice parameters. However, NbS3-iii can be distinguished by a characteristic CDW transition at 155 K.20 Subsequent electrical measurements by Zybtsev and co-workers on “low-ohmic” and “high-ohmic” NbS3 samples showed that low-ohmic NbS3 exhibits CDW transitions at both 150 and 360 K, whereas high-ohmic NbS3 exhibits a single CDW transition at 150 K.3 On this basis, it has been postulated that low-ohmic and high-ohmic NbS3 correspond to NbS3-ii and NbS3-iii, respectively; an alternate interpretation is that low- and high-ohmic NbS3 as well as NbS3-iii are sub-phases of NbS3-ii.21 

This letter reports the growth and characterization of two new monoclinic polymorphs of NbS3, designated NbS3-iv and NbS3-v following the nomenclature currently in place. Interestingly, we isolated NbS3-iv under synthetic conditions previously reported to produce NbS3-i and NbS3-v under synthetic conditions previously reported to produce NbS3-ii.3,34 Single crystal x-ray diffraction studies of both NbS3-iv and NbS3-v provide unambiguous structural information about these polymorphs. Such definitive characterization is necessary for accurate identification, especially because NbS3 polymorphs form under similar conditions and often more than one can be found in the same crystal growth batch.28 Furthermore, electron microscopy analysis provides information about structural defects in NbS3-iv and atomic order in NbS3-v, and Raman spectroscopy of NbS3-iv and NbS3-v provides data that can be used to distinguish between polymorphs.

Niobium trisulfide samples were synthesized using niobium powder (99.99%, STREM) and sublimed sulfur powder (Baker, >99%) using a chemical vapor transport method. Iodine (Baker, 98%) at 4–6 mg/cm3 was used for the growth of the NbS3-iv phase, whereas excess sulfur (∼10%) was utilized as the transport agent for the growth of NbS3-v crystals. For all syntheses, niobium and sulfur were lightly ground by mortar and pestle before being placed in a quartz ampule (∼13 cm3). The ampules were then evacuated and backfilled with argon three times before flame sealing under vacuum (∼10−2 Torr). Reactions were placed in a two-zone furnace with a fixed gradient for 10–14 days; upon completion, the ampule/furnace was allowed to cool to room temperature naturally.

NbS3-iv crystals used for x-ray diffraction analysis were prepared at 670 °C with a 100 °C gradient (670 °C source zone → 570 °C growth zone). This synthesis temperature is comparable to temperatures used to prepare NbS3-i,11,27,34 and the only difference here is the use of I2 as a transport agent. This NbS3 polymorph crystallizes in a monoclinic unit cell, and the structure was solved using space group P21/c. Additional details are provided in Tables S1 and S2 of the supplementary material, and crystallographic data have been deposited with the Cambridge Crystallographic Data Centre (structure number 1574381).

The unit cell of NbS3-iv is larger than that of NbS3-i because the c-axis is doubled, resulting in twice as many chains per unit cell (Fig. 1). The asymmetric unit of NbS3-iv is Nb2S6, making Z formally 4, which becomes 8 when normalized to the formula NbS3. Also, in contrast to other MX3 structures, the chain axis in NbS3-iv is the a-axis rather than the b-axis (Fig. 1). Looking down the a-axis, the structure of NbS3-iv in cross section is similar to the model ZrSe3-type structure, specifically the “A-variant.”35 However, a difference between these structures is their layer-stacking order. This is clearly seen when examining multiple unit cells of each polymorph (Fig. 1). The chains in NbS3-i realign every five bilayers, where a bilayer is defined as a pair of chains bounded by van der Waals gaps (Fig. 1 where pairs of blue or green chains designate the bilayers). Thus, the repeat sequence is ABCDE. However, NbS3-iv experiences a realignment of the chain after only two bilayers, making the repeat sequence AB.

In the chain cross section of NbS3-iv, three sulfur atoms surround a niobium atom to form acute, isosceles triangles. The bases of the triangles are made of a 2.015 Å S–S bond; the presence of this bond is consistent with the formulation of NbS3 as Nb(S2)S.22 The sides of the triangles are separated by longer SS distances of 3.404-3.418 Å or 3.787-3.828 Å. Two interchain Nb–S bonds of 2.623 and 2.654 Å exist between each chain, which hold the bilayer together. Each bilayer is separated from the next by a distance of ∼2.8 Å, the van der Waals gap.

Looking down the b-axis (Fig. 1), we see that NbS3-iv contains Nb–Nb pairing with a bond distance of 3.0448 Å, much like NbS3-i. Thus, NbS3-iv demonstrates that it is possible to have Nb–Nb pairing within a space group with P21/c symmetry, i.e., it is not essential to reduce the symmetry to triclinic as in NbS3-i. It is curious that Jellinek and co-workers published two papers prior to their report of NbS3-i’s triclinic structure in which they reported monoclinic unit cells extremely similar to what we have found for NbS3-iv.16,17 It is possible that they actually made NbS3-iv in earlier work, which was based on the synthesis of NbS3 from the elements, whereas their structural study was conducted on NbS3 crystals generated through the thermal decomposition of NbS2Cl2.22 The differences in preparative conditions may have led to the formation of NbS3-i versus NbS3-iv. In addition, van Smaalen has re-interpreted the triclinic space group P1¯ of NbS3-i as a monoclinic superspace group, proposing that this is an uncommon example of a modulated structure.36 

The relatively high wR value (11.38%) associated with the refinement of NbS3-iv reflects the presence of structural disorder, possibly including sulfur vacancies, faulting or slippage along the van der Waals planes, and twinning, all of which have been observed in various MX3 materials.28,35,37 To identify the specific disorder in NbS3-iv, the TEM analysis was performed on a Thermo Fisher Scientific Themis 300 instrument, fitted with a X-FEG electron gun and CETA 4K × 4K CMOS camera, at an accelerating voltage of 300 kV. Specimens were prepared by dispersing a small amount of as-synthesized NbS3-iv powder in distilled water without grinding and depositing a droplet of the resulting suspension onto lacey carbon grids. The TEM study revealed that the sample consists of three types of particles: (1) long thin whiskers with aspect ratios >100 and thicknesses <100 nm that are elongated parallel to the a-axis [Fig. 2(a)]; (2) prismatic crystals with aspect ratios <20, with a long side parallel to the a-axis, a short side parallel to the b-axis, and thicknesses <100 nm along the c-axis [Fig. 2(b)]; (3) small aggregates consisting of isometric particles <50 nm in diameter.

FIG. 2.

(a) Bright field TEM image of a NbS3-iv fiber. Insets show (i) HRTEM image (lower left) of the crystal along the b-axis and (ii) the corresponding indexed FFT pattern (upper right). The location of the HRTEM image is indicated by a black arrow. In the HRTEM inset, multiple disruptions of the lattice fringes parallel to {001} are visible; some dislocations are indicated by white arrows. (b) Bright field high magnification TEM image of a prismatic NbS3-iv particle. Insets shown are (i) low magnification BF TEM image of the whole particle (upper) and (ii) the corresponding selected area electron diffraction pattern along the c-axis, which confirms the single crystal nature of the whole particle. The topotactically oriented individual intergrown zones are clearly visualized, with some of the coherent interfaces between zones indicated by black arrows. (c) HRTEM image of the particle in (b). The inset shows a magnified view of the area marked by a white box. The complex lattice contrast and the mosaic structure of low-angle domains is revealed here; the slight tilt of the domains causes considerable change in the lattice fringe contrast, which prevents the reliable interpretation of any stacking disorder that might be present parallel to the c-axis.

FIG. 2.

(a) Bright field TEM image of a NbS3-iv fiber. Insets show (i) HRTEM image (lower left) of the crystal along the b-axis and (ii) the corresponding indexed FFT pattern (upper right). The location of the HRTEM image is indicated by a black arrow. In the HRTEM inset, multiple disruptions of the lattice fringes parallel to {001} are visible; some dislocations are indicated by white arrows. (b) Bright field high magnification TEM image of a prismatic NbS3-iv particle. Insets shown are (i) low magnification BF TEM image of the whole particle (upper) and (ii) the corresponding selected area electron diffraction pattern along the c-axis, which confirms the single crystal nature of the whole particle. The topotactically oriented individual intergrown zones are clearly visualized, with some of the coherent interfaces between zones indicated by black arrows. (c) HRTEM image of the particle in (b). The inset shows a magnified view of the area marked by a white box. The complex lattice contrast and the mosaic structure of low-angle domains is revealed here; the slight tilt of the domains causes considerable change in the lattice fringe contrast, which prevents the reliable interpretation of any stacking disorder that might be present parallel to the c-axis.

Close modal

The fibers of NbS3-iv are twisted around their long axis, resulting in the development of numerous dislocations that can be identified through disruptions of the lattice fringes parallel to {001} [Fig. 2(a)]. The prismatic particles consist of topotactically intergrown zones 5–50 nm wide parallel to the b-axis and several microns long parallel to the a-axis [Fig. 2(b)]. They are coherently intergrown with interfaces parallel to the {010} planes. The individual crystals exhibit a mosaic structure with fine domains about 10 nm in diameter [Fig. 2(c)] that are coherently intergrown with small angular mis-orientation around the c-axis, which cause changes in the contrast and appearance of the lattice fringes resolved by HRTEM imaging. It is possible that the observed variations in the lattice images along the c-axis are caused by stacking defects; it is possible to resolve such features by imaging down the a-axis, but this has not been done here due to the unfavorable particle shapes.

Further exploration of the NbS3 system provided yet another monoclinic NbS3 polymorph, NbS3-v. It was prepared at 700 °C with a 30 °C gradient (700 °C source zone → 670 °C growth zone) using 10% excess sulfur for transport. This polymorph also crystallizes in the P21/m space group. Additional details are provided in Tables S1 and S2 of the supplementary material, and crystallographic data have been deposited with the Cambridge Crystallographic Data Centre (structure number 1574382). As illustrated in Fig. 1, the unit cell of NbS3-v is similar to that of NbS3-i but with a halved b-axis; the lattice parameters can be expressed as a, b/2, c, β′ based on the NbS3-i unit cell (Table I). These dimensions result in two chains per unit cell, albeit the unit cell is only one [MX6] trigonal prism wide (Z = 2). This unit cell is smaller than those proposed for NbS3-ii.

In the chain cross section of NbS3-v, three sulfur atoms surround a niobium atom to form acute isosceles triangles with bases composed of 2.020 Å S–S bonds and sides formed by SS distances of 3.576-3.612 Å. Interestingly, the interchain Nb–S interactions alternate in pairs of 2.619 Å and 2.635 Å Nb–S bonds, which results in a small but measurable chain pairing within the bilayer sheet. Although NbS3-v features the same ABCDE stacking that occurs in NbS3-i (Fig. 1), it contains a single, uniform Nb–Nb distance of 2.358 Å, leading to ideal (non-corrugated) [MX6] chains. Theoretically, NbS3 with such an undistorted structure should display metallic conductivity due to a single electron remaining on each Nb4+ after creation of the [NbS6] chains.23,38

NbS3-iv and NbS3-v can be exfoliated in a similar fashion as other MX3 materials.5,6,13 Figure 3 shows HAADF-STEM images of a ∼40 nm wide NbS3-v nanoribbon exfoliated by sonication in ethanol. The high-resolution image in panel (b) reveals the [NbS6] chains along the b-axis. Both the Nb–Nb bonds with 3.37 Å spacing and the van der Waals gap between chains are clearly visible. There are no defects visible at this magnification in this particular area of the sample or others imaged along this nanoribbon, although the high wR value (21.99%) associated with the refinement of NbS3-v points to the presence of significant structural disorder in the larger crystal used for x-ray diffraction analysis.

FIG. 3.

HAADF-STEM images of an exfoliated NbS3-v nanoribbon along the b-axis at low (a) and high magnifications (b), together with a structural model for the observed atomic spacings (c).

FIG. 3.

HAADF-STEM images of an exfoliated NbS3-v nanoribbon along the b-axis at low (a) and high magnifications (b), together with a structural model for the observed atomic spacings (c).

Close modal

In experimental material research, powder x-ray diffraction (PXRD) is usually the initial characterization method used to identify crystalline phases and evaluate phase purity. The patterns for NbS3-i, iv, and v are compared in Fig. 4(a). Although the theoretical patterns (calculated from single-crystal XRD structures) of NbS3-i and NbS3-iv are similar, each has distinct peaks in the range 10°-40° 2θ that should make polymorph identification unambiguous. In practice, however, the experimental peak intensities of NbS3-iv (powder mount) are altered by orientation effects, such that the resulting pattern can be matched to either NbS3-i or NbS3-iv. In the case of NbS3-v, its theoretical pattern is essentially identical to that of NbS3-i, but experimentally NbS3-v powder generates a distinct pair of peaks at ∼22° 2θ. Thus, we caution that PXRD alone is not sufficient to differentiate between all NbS3 polymorphs.

FIG. 4.

(a) Powder x-ray diffraction patterns of NbS3-i, NbS3-iv, and NbS3-v; drop lines for NbS3-i are from JCPDS card no. 04-007-1130. Powder x-ray diffraction data for NbS3-iv, and NbS3-v were collected on a Bruker Advance D8 diffractometer with Co-Kα radiation at 35 kV and 40 mA from 5° to 80° 2θ with a 0.01 step size at 0.4 s/step. (b) Raman spectroscopy of NbS3-iv and NbS3-v. Raman data were collected on a Renishaw InVIA microscope at 785 nm with a 50× objective lens. SEM images of NbS3-iv [(c) and (d)] and NbS3-v [(e) and (f)]. SEM was conducted on a FEI Teneo FE-SEM instrument equipped with an Oxford Instruments X-MAXN detector at 5–20 keV.

FIG. 4.

(a) Powder x-ray diffraction patterns of NbS3-i, NbS3-iv, and NbS3-v; drop lines for NbS3-i are from JCPDS card no. 04-007-1130. Powder x-ray diffraction data for NbS3-iv, and NbS3-v were collected on a Bruker Advance D8 diffractometer with Co-Kα radiation at 35 kV and 40 mA from 5° to 80° 2θ with a 0.01 step size at 0.4 s/step. (b) Raman spectroscopy of NbS3-iv and NbS3-v. Raman data were collected on a Renishaw InVIA microscope at 785 nm with a 50× objective lens. SEM images of NbS3-iv [(c) and (d)] and NbS3-v [(e) and (f)]. SEM was conducted on a FEI Teneo FE-SEM instrument equipped with an Oxford Instruments X-MAXN detector at 5–20 keV.

Close modal

Raman spectroscopy provides additional data to differentiate NbS3-iv and NbS3-v. As shown in Fig. 4(b), the spectra of these polymorphs are clearly different. However, the spectrum of NbS3-v is not as clearly resolved as that of NbS3-iv, which may be due to the structural disorder present. The spectrum of NbS3-iv is similar but not identical to data published for NbS3-i.11,39–41 The specific differences are (1) the relative intensities of the two peaks at 145-155 nm are switched and (2) the literature spectra for NbS3-i show a single peak at ∼190 nm [assigned as ν(NbS2) in Ref. 38], whereas the spectrum of NbS3-iv in Fig. 4(b) has two peaks at this position. This latter difference is reasonable based on the asymmetric unit of Nb2S6 in NbS3-iv, which should lead to two distinct ν(NbS2) modes.

SEM images [Figs. 4(c)–4(f)] show that the morphology of NbS3-iv and NbS3-v is fibrous. NbS3-iv crystals generally appear as wider, straight fibers, whereas NbS3-v crystals look more like flexible wires. These morphological observations are similar to descriptions of NbS3-i (larger, straighter needles) and NbS3-ii (finer, bent whiskers) in the literature.22,30,34 In addition to the fibers, both products contain numerous cylinders and rings in various stages of winding/unwinding [Figs. 4(d) and 4(e)]. Moreover, energy dispersive spectroscopy (EDS) analysis indicates that these microstructures have a NbS3 stoichiometry identical to the fibers. Such “topological crystals,” previously observed for TaS3, NbSe3, and TaSe3, apparently form when droplets of molten elemental chalcogen template crystal growth.42–46 Indeed, Enyashin and Ivanovskiĭ have provided a theoretical basis indicating that NbS3 ring structures are energetically favorable.47 Continuing work will be required to understand the formation mechanism of such morphologies in the NbS3 system and characterize the distortions that accompany crystal bending.

In summary, the crystal structures of two new niobium trisulfide polymorphs, NbS3-iv and NbS3-v, greatly enrich our understanding of this system. Both polymorphs crystallize in monoclinic unit cells, which fills a long-standing gap of knowledge about monoclinic NbS3. We find that the structure of NbS3-iv is related to triclinic NbS3-i with respect to Nb–Nb pairing along the chain axis. In comparison, the structure of NbS3-v is undistorted along the chain axis but contains subtle chain pairing. This structural information will be extremely helpful for the interpretation of spectroscopic data and modeling the electronic properties of NbS3-iv and NbS3-v.

See supplementary material for additional information about the x-ray diffraction studies of NbS3-iv and NbS3-v, including atomic coordinates.

This work was supported by a U.S. National Science Foundation’s Emerging Frontiers of Research Initiative 2-DARE Grant (No. NSF EFRI-1433395).

1.
A.
Kuc
,
Chemical Modelling: Volume 11
(
The Royal Society of Chemistry
,
2015
), Vol. 11, p.
1
.
2.
X.
Song
,
Z.
Guo
,
Q.
Zhang
,
P.
Zhou
,
W.
Bao
, and
D. W.
Zhang
,
Small
13
,
1700098
(
2017
).
3.
S. G.
Zybtsev
,
V. Y.
Pokrovskii
,
V. F.
Nasretdinova
, and
S. V.
Zaitsev-Zotov
,
Phys. B
407
(
11
),
1696
(
2012
).
4.
Y. S.
Hor
,
Z. L.
Xiao
,
U.
Welp
,
Y.
Ito
,
J. F.
Mitchell
,
R. E.
Cook
,
W. K.
Kwok
, and
G. W.
Crabtree
,
Nano Lett.
5
(
2
),
397
(
2005
).
5.
Y.
Jin
,
X.
Li
, and
J.
Yang
,
Phys. Chem. Chem. Phys.
17
(
28
),
18665
(
2015
).
6.
M. A.
Stolyarov
,
G.
Liu
,
M. A.
Bloodgood
,
E.
Aytan
,
C.
Jiang
,
R.
Samnakay
,
T. T.
Salguero
,
D. L.
Nika
,
S. L.
Rumyantsev
,
M. S.
Shur
,
K. N.
Bozhilov
, and
A. A.
Balandin
,
Nanoscale
8
(
34
),
15774
(
2016
).
7.
G.
Liu
,
S.
Rumyantsev
,
M. A.
Bloodgood
,
T. T.
Salguero
,
M.
Shur
, and
A. A.
Balandin
,
Nano Lett.
17
(
1
),
377
(
2017
).
8.
A.
Grisel
,
F.
Lévy
, and
T. J.
Wieting
,
Physica B+C
99
(
1
),
365
(
1980
).
9.
J. O.
Island
,
M.
Barawi
,
R.
Biele
,
A.
Almazán
,
J. M.
Clamagirand
,
J. R.
Ares
,
C.
Sánchez
,
H. S. J.
van der Zant
,
J. V.
Álvarez
,
R.
D’Agosta
,
I. J.
Ferrer
, and
A.
Castellanos-Gomez
,
Adv. Mater.
27
(
16
),
2595
(
2015
).
10.
P. A.
Poltarak
,
S. B.
Artemkina
,
A. I.
Bulavchenko
,
T. Yu.
Podlipskaya
, and
V. E.
Fedorov
,
Russ. Chem. Bull.
64
(
8
),
1850
(
2015
).
11.
V. E.
Fedorov
,
S. B.
Artemkina
,
E. D.
Grayfer
,
N. G.
Naumov
,
Y. V.
Mironov
,
A. I.
Bulavchenko
,
V. I.
Zaikovskii
,
I. V.
Antonova
,
A. I.
Komonov
, and
M. V.
Medvedev
,
J. Mater. Chem. C
2
(
28
),
5479
(
2014
).
12.
J. O.
Island
,
A. J.
Molina-Mendoza
,
M.
Barawi
,
R.
Biele
,
E.
Flores
,
J. M.
Clamagirand
,
J. R.
Ares
,
C.
Sánchez
,
H. S. J.
van der Zant
,
R.
D’Agosta
,
I. J.
Ferrer
, and
A.
Castellanos-Gomez
,
2D Mater.
4
(
2
),
022003
(
2017
).
13.
A.
Lipatov
,
P. M.
Wilson
,
M.
Shekhirev
,
J. D.
Teeter
,
R.
Netusil
, and
A.
Sinitskii
,
Nanoscale
7
(
29
),
12291
(
2015
).
14.
Z.
Yan
,
C.
Jiang
,
T. R.
Pope
,
C. F.
Tsang
,
J. L.
Stickney
,
P.
Goli
,
J.
Renteria
,
T. T.
Salguero
, and
A. A.
Balandin
,
J. Appl. Phys.
114
(
20
),
204301
(
2013
).
15.
J.
Wu
,
D.
Wang
,
H.
Liu
,
W. M.
Lau
, and
L.
Liu
,
RSC Adv.
5
,
21455
(
2015
).
16.
F.
Jellinek
,
G.
Brauer
, and
H.
Muller
,
Nature
185
(
4710
),
376
(
1960
).
17.
F.
Kadijk
and
F.
Jellinek
,
J. Less-Common Met.
19
(
4
),
421
(
1969
).
18.
M.
Kurdi
,
A.-M.
Marie
, and
M.
Danot
,
Solid State Commun.
64
(
4
),
395
(
1987
).
19.
A. V.
Mishchenko
,
I. V.
Yushina
, and
V. E.
Fedorov
,
Zh. Neorg. Khim.
33
(
2
),
437
(
1988
).
20.
A.
Zettl
,
C. M.
Jackson
,
A.
Janossy
,
G.
Grüner
,
A.
Jacobsen
, and
A. H.
Thompson
,
Solid State Commun.
43
(
5
),
345
(
1982
).
21.
S. G.
Zybtsev
,
V. Y.
Pokrovskii
,
V. F.
Nasretdinova
,
S. V.
Zaitsev-Zotov
,
V. V.
Pavlovskiy
,
A. B.
Odobesco
,
W.
Wu Pai
,
M. W.
Chu
,
Y. G.
Lin
,
E.
Zupanič
,
H. J. P.
van Midden
,
S.
Šturm
,
E.
Tchernychova
,
A.
Prodan
,
J. C.
Bennett
,
I. R.
Mukhamedshin
,
O. V.
Chernysheva
,
A. P.
Menushenkov
,
V. B.
Loginov
,
B. A.
Loginov
,
A. N.
Titov
, and
M.
Abdel-Hafiez
,
Phys. Rev. B
95
(
3
),
035110
(
2017
).
22.
J.
Rijnsdorp
and
F.
Jellinek
,
J. Solid State Chem.
25
(
4
),
325
(
1978
).
23.
D. W.
Bullett
,
J. Solid State Chem.
33
(
1
),
13
(
1980
).
24.
M. E.
Itkis
,
F. Ya.
Nad
,
S. V.
Zaitsev-Zotov
, and
F.
Lévy
,
Solid State Commun.
71
(
11
),
895
(
1989
).
25.
M. E.
Itkis
,
F. Ya.
Nad
, and
F.
Levy
,
Synth. Met.
43
(
3
),
3969
(
1991
).
26.
Z. Z.
Wang
,
P.
Monceau
,
H.
Salva
,
C.
Roucau
,
L.
Guemas
, and
A.
Meerschaut
,
Phys. Rev. B
40
(
17
),
11589
(
1989
).
27.
V.
Nasretdinova
and
S.
Zaitsev-Zotov
,
Phys. B
407
(
11
),
1874
(
2012
).
28.
T.
Cornelissens
,
G.
Van Tendeloo
,
J.
Van Landuyt
, and
S.
Amelinckx
,
Phys. Status Solidi A
48
(
1
),
K5
(
1978
).
29.
F. W.
Boswell
and
A.
Prodan
,
Physica B+C
99
(
1
),
361
(
1980
).
30.
A.
Prodan
,
A.
Budkowski
,
F. W.
Boswell
,
V.
Marinkov
,
J. C.
Bennett
, and
J. M.
Corbett
,
J. Phys. C: Solid State Phys.
21
(
23
),
4171
(
1988
).
31.
C.
Roucau
,
J. Phys. Colloq.
44
(
C3
),
C3
(
1983
).
32.
A.
Meerschaut
,
J. Phys. Colloq.
44
(
C3
),
C3
(
1983
).
33.
S.
Kikkawa
,
N.
Ogawa
,
M.
Koizumi
, and
Y.
Onuki
,
J. Solid State Chem.
41
(
3
),
315
(
1982
).
34.
V. Y.
Pokrovskii
,
S. G.
Zybtsev
,
M. V.
Nikitin
,
I. G.
Gorlova
,
V. F.
Nasretdinova
, and
S. V.
Zaitsev-Zotov
,
Phys.-Usp.
56
(
1
),
29
(
2013
).
35.
E.
Guilmeau
,
D.
Berthebaud
,
P. R. N.
Misse
,
S.
Hebert
,
O. I.
Lebedev
,
D.
Chateigner
,
C.
Martin
, and
A.
Maignan
,
Chem. Mater.
26
,
5585
(
2014
).
36.
S.
van Smaalen
,
Phys. Rev. B
38
(
14
),
9594
(
1988
).
37.
N. B.
Bolotina
,
I. G.
Gorlova
,
I. A.
Verin
,
A. N.
Titov
, and
A. V.
Arakcheeva
,
Crystallogr. Rep.
61
(
6
),
923
(
2016
).
38.
K.
Endo
,
H.
Ihara
,
K.
Watanabe
, and
S.-I.
Gonda
,
J. Solid State Chem.
39
(
2
),
215
(
1981
).
39.
C.
Sourisseau
,
R.
Cavagnat
,
M.
Fouassier
, and
P.
Maraval
,
J. Raman Spectrosc.
21
(
6
),
337
(
1990
).
40.
C.
Sourisseau
,
R.
Cavagnat
,
M.
Fouassier
, and
P.
Maraval
,
J. Mol. Struct.
219
,
183
(
1990
).
41.
A.
Zwick
,
M. A.
Renucci
,
R.
Carles
,
N.
Saint-Cricq
, and
J. B.
Renucci
,
Physica B+C
105
(
1
),
361
(
1981
).
42.
S.
Tanda
,
H.
Kawamoto
,
M.
Shiobara
,
Y.
Sakai
,
S.
Yasuzuka
,
Y.
Okajima
, and
K.
Yamaya
,
Phys. B
284-288
(
Part 2
),
1657
(
2000
).
43.
T.
Matsuura
,
S.
Tanda
,
K.
Asada
,
Y.
Sakai
,
T.
Tsuneta
,
K.
Inagaki
, and
K.
Yamaya
,
Phys. B
329-333
(
Part 2
),
1550
(
2003
).
44.
T.
Tsuneta
and
S.
Tanda
,
J. Cryst. Growth
264
(
1-3
),
223
(
2004
).
45.
G.
Kumagai
,
T.
Matsuura
,
K.
Ichimura
, and
S.
Tanda
,
Phys. B
405
(
Suppl. 11
),
S284
(
2010
).
46.
M.
Tsubota
,
K.
Inagaki
,
T.
Matsuura
, and
S.
Tanda
,
Cryst. Growth Des.
11
(
11
),
4789
(
2011
).
47.
A. N.
Enyashin
and
A. L.
Ivanovskiĭ
,
Phys. Solid State
48
(
4
),
780
(
2006
).

Supplementary Material