In Situ Synthesis and Single Crystal Synchrotron X-ray Diffraction

Article Cite This: Acc. Chem. Res. XXXX, XXX, XXX−XXX

pubs.acs.org/accounts

In Situ Synthesis and Single Crystal Synchrotron X‑ray Diffraction Study of ht-Sn3Sb2: An Example of How Complex Modulated Structures Are Becoming Generally Accessible Published as part of the Accounts of Chemical Research special issue “Advancing Chemistry through Intermetallic Compounds”. Sven Lidin* and Laura C. Folkers Centre for Analysis and Synthesis, Department of Chemistry, Lund University, POB 124, 221 00 Lund, Sweden CONSPECTUS: Recent developments in X-ray sources and detectors and the parallel development of software for nonstandard crystallography has made analysis of very complex structural problems accessible to nonexperts. Here, we report the successful solution of the structure of ht-Sn3Sb2, an analysis that presents several challenges but that is still manageable in a relatively straightforward way. This compound exists only in a narrow temperature regime and undergoes an unquenchable phase transformation on cooling to room temperature; it contains two elements with close to identical scattering factors, and the structure is incommensurately modulated with four symmetry dependent modulation wave vectors. In this study, an attempt was first made to synthesize the title compound by in-house crystal growth in the stability region of ht-Sn3Sb2, followed by cooling to room temperature. This is known to produce mutiply twinned stistaite and elemental tin, and this sample, freshly prepared, was then reheated in situ at the single crystal materials beamline Crystal at the synchrotron Soleil. This method was unsuccessful as reheating the sample led to loss of Sn from stistaite as revealed by a change in the measured modulation wave vector. The compound was instead successfully synthesized in situ at the beamline by the topochemical reaction of single crystalline stistaite and liquid tin. A well-formed crystal of stistaite was enclosed in a quartz capillary together with a large excess of tin and heated above the melting point of tin but below the melting point of ht-Sn3Sb2. The structure was probed by sychrotron X-ray diffraction using a wavelength close to the absorption edge of Sn to maximize elemental contrast. In the diffraction patterns, first order satellites were observed, making the structure of ht-Sn3Sb2 incommensurately modulated. Further analysis exposes four q-vectors running along the body diagonals of the cubic unit cell (q1′ = α α α, q2′ = −α α -α, q3′ = -α −α α, q4′ = α -α −α). To facilitate the analysis, the q vectors were instead treated as axial (q1 = α 0 0, q2 = 0 α 0, q3 = 0 0 α) and an F-type extinction condition for satellites was introduced so that only reflections with hklmnp, mnp all odd or all even, were considered. Further, the modulation functions F(qi) were set to zero, and only modulation functions of the type F(qi′) were refined. The final model uses the four modulation functions, F(q1′), F(q2′), F(q3′), and F(q4′), to model occupancy Sn/Sb and positional modulation. The model shows a structure that comprises small NaCl type clusters, typically 7 × 7 × 7 atoms in extension, interspersed between single layers of elemental tin. The terminating layers of tin are slightly puckered, emulating an incipient deformation toward the structure of the layers perpendicular to the [001] direction in elemental tin. It is notable that this model is complementary to that of stistaite. In stistaite, two-dimensionally infinite slabs of rock salt are interspersed between layers of antimony along the trigonal [001] direction, so that the terminating Sb layers are the puckered bilayers typical for elemental Sb. Since all modulation functions are simple first-order harmonics, the structural model describes a locally disordered and most probably dynamic ordering.

■

INTRODUCTION The increased power of X-ray sources in combination with the development of faster and more sensitive 2D detectors has led to an increased awareness of the ubiquitous presence of nonBragg features in X-ray diffraction data. The power of modern synchrotron sources facilitates the quantification of diffuse scattering and weak satellite reflections and turns weak features into hard facts. A parallel development has taken place on the software side where user-friendly programs are now available to © XXXX American Chemical Society

nonexperts for treating such diffraction features in a relatively straightforward way.1 The importance of this latter development can hardly be overstated. When the concept of incommensurately modulated structures was first introduced by Dehlinger2 in 1927, it was Received: October 12, 2017

A

DOI: 10.1021/acs.accounts.7b00508 Acc. Chem. Res. XXXX, XXX, XXX−XXX

Article

Accounts of Chemical Research

a simple, one-dimensional modulation, the standard method of representation is that of a section through (3 + 1)-dimensional space where one of the coordinate axes is internal space, that is, the direction of modulation. This is an excellent way of visualizing how a particular atomic position is, or a set of atomic positions are, affected by the modulation (Figure 1).

as a curiosity. Dehlinger studied cold worked metals and noted that apart from broadening of powder diffraction peaks, the appearance of new lines indicated incipient additional ordering. He derived an essentially correct model, but his results did not receive much attention outside of the metals-working community, and the method was not easily generalized to more complex systems. Instead, the real starting-point for standardized treatment of incommensurability came with the analysis of the structure of γ-Na2CO3.3 By a lucky coincidence, P.M. de Wolff had been working on this structural enigma, while concurrently Aloysio Janner and Ted Janssen had been exploring ideas of additional dimensions in crystallography, although the original idea of the latter authors was to include time as the additional dimension, and these scientist (all Dutch) became acquainted at a crystallography conference in Japan.4 The formalism to use was thus introduced in the mid 1970s, and it was quickly adapted to a number of long-standing problems in structural science.5,6 A particular class of modulated structures that was recognized early, and long before any formal treatment was possible beyond standard translationally periodic modeling, is the composite structures. These are solid state compounds that host two or more sublattices that are mutually incommensurate and that may be describes as two crystal structures that interpenetrate and modulate each other. Examples of such compounds are canizzerite, Pb46Bi54S127, a mineral with crystal faces that do not adhere to simple rational indices,7 and the family of Nowotny chimney ladder phases.8,9 Thanks to monumental efforts, these structures were solved and to some extent understood in terms of periodic crystallography, and their special nature was appreciated, but the full elucidation of their particulars had to wait for the treatment using methods of aperiodic crystallography.10−12 Intermetallic compounds seem to be particularly prone to incommensurability, and a recent review13 presents some interesting figures. There are 21 primitive tetragonal structure types with more than 100 atoms in the unit cell reported in Pearson’s handbook,14 and out of these, 11 are composite and an additional two are modulated. Similarly, there are 12 face-centered orthorhombic structures with at least 100 atoms per unit cell, and two of them are modulated and two are composites. It is clear that modulations and in particular composite structures constitute a large portion of the complex intermetallic structures. Our interest in the field arose from studies of superstructure ordering in B8 type intermetallic compounds where incommensurability is rampant.15−18 Using the concept of incommensurability to describe compounds with large solid solubility regions or regions of phase diagrams that contain phase bundles provides an attractive alternative to standard disordered models. Incommensurate models elegantly describe ordered intergrowth of what has been considered separate phases and provide a formalism for understanding the existence of similar, closely grouped compounds.19,20 This understanding provides a basis for further synthetic work, theoretical modeling, and understanding of properties. The properties of incommensurately modulated structures are similar to but distinct from those of normal complex superstructures,21 and this is particularly relevant for understanding thermal transport in thermoelectric compounds.22−26 A particular challenge in presenting the results from studies of modulated compounds is representing them in pictorial form. We are used to seeing the structure of solid state compounds represented as atomic positions in the unit cell. For

Figure 1. Pictorial presentation of the electron density of single atomic surface in a modulated structure (Zn3−xSb227). The atomic surface for a one-dimensional modulation is a curve, the red curve showing the best fit of the maximal electron density. The vertical axis is internal space while the slightly skewed horizontal axis is the a-axis of the periodic unit cell. By following the red line of the atomic surface, we can envisage how the atomic position varies along internal space. The heavy red, straight horizontal line shows the relationship between the atomic positions in successive unit cells. In this way the plot visualizes how interatomic distances vary in the structure. In this particular case, the variation in distance is a bimodal distribution of long and short distances, and the sequence of long and short forms an aperiodic distribution.

For structures with more than one modulation direction, pictorial presentations are much more challenging. An atomic surface for a two-dimensional modulation may be represented as a surface where the basis plane is spanned by the two modulation vectors and the height above that plane represents the deviation away from the position in the average unit cell. For cases with more modulation directions representation becomes even more unwieldy and often recourse is either taken to using periodic approximations or one-dimensional representations where the other modulation vectors are kept constant. The present study of a cubic structure with an incommensurate modulation in several dimensions is an example of how a confluence of methods and equipment developments may be used to elucidate features that have previously eluded analysis. It should be stressed that the ordering patterns thus revealed are no minor changes to an otherwise unperturbed basic structure, but that these are fundamental features without which the nature of the compound becomes incomprehensible. The binary phase diagram Sn−Sb28 (Figure 2) is deceptively simple, dominated by two essential features; the highly asymmetric nature of mutual solubility and the wide composition range of the roughly equimolar stistaite phase SnSb. Stistaite was the subject of a previous paper where it was shown that the stistaite composition range may be understood in terms of a fully ordered model where Sb double layers are inserted between Sn4Sb3 slabs.29 The limiting composition at B


Article


Figure 2. Phase diagram of the system Sn−Sb. Note the large solubility of Sn in Sb, the wide composition range of stistaite, and the location of the title phase ht-Sn3Sb2.

forms domain twins, and here the unique (tetragonal) axes are aligned with the face normals of the external shape of cubic crystals. This indicates two things: first, that the structure of htSn3Sb2 cannot be retained at room temperature and, indeed, attempts to quench the sample producing the same result as slow cooling; second that the structure of ht-Sn3Sb2 is expected to have cubic symmetry, both from considering the tell-tale signs from balanced twinning and from the habit of the crystals. Clearly, crystals with a cubic external shape containing equal amounts of domain twins of two different phases to mimic cubic symmetry indicate that the high temperature phase from which they have formed is likely to have cubic symmetry.

the Sn-rich end is thus Sn4Sb3, while the limiting composition at the Sb-rich end is given by the maximum width of the Sbslabs that still support ordering between adjacent Sn4Sb3 slabs. From the phase diagram, this limit is close to SnSb2, that is, (Sn4Sb3)(Sb2)2.5, indicating that the structural information propagates through a maximum of about two double layers of Sb and that for higher Sb content, while the Sn4Sb3 slabs may still be present, their positions in the Sb matrix are uncorrelated. The substantial solubility of Sn in Sb, around 13%, is an indication that the mode of solubility may be cooperative, that is, the insertion of a small number of Sn atoms between Sb-layers leads to increased probability of further Sn insertion into this interlayer and thus leads to intercalation of Sn-layers into Sb. This is in sharp contrast to the negligible solubility of Sb in Sn where the solubility mechanism is more likely to be simple elemental exchange. There is however one more feature in the Sn−Sb phase diagram, the phase tentatively labeled ht-Sn3Sb2, with a narrow existence range in temperature from 242 to 324 °C. This phase may be grown as large single crystals from a Sn-rich melt. The crystals have a cubic habit, but single crystal X-ray diffraction at room temperature reveals that they are composed of two phases, Sn-rich stistaite and Sn. Ordered intergrowth of two or multiple phases is a phenomenon often observed in electron diffraction studies, but it is also not uncommon in single crystal X-ray diffraction. The term “single crystal” is certainly a misnomer in this context, but in want of a better term it will be used here. Identifying such behavior is not straightforward but requires careful examination of the reciprocal space patterns. In this particular case, the diffraction pattern of Sn is quite weak and did not in any way interfere with the identification of the features of the main phase, but once the reflections belonging to the stistaite phase were indexed, identification of the diffraction pattern of Sn was straightforward. The rhombohedral stistaite is present as balanced domain twins, that is, the amount present of each twin domain is the same and this retains the cubic symmetry of the collective. The rhombohedral domains have their unique (trigonal) axes aligned to the body diagonals of the external shape of the cubic crystal. The Sn also

■

EXPERIMENTAL SECTION Single crystals of ht-Sn3Sb2 were grown from tin rich melts and isolated from the flux by centrifugation.30 The crystals were sealed in capillaries and measured at an in-house single crystal diffractometer to assess crystal quality. Most crystals showed some degree of twinning, but several were single domain samples and no traces of residual Sn could be detected. This discrepancy between the previous experiments where balanced twinning was observed is probably due to a different setup of the centrifuge that makes the procedure of removing surplus flux considerably slower and allows for a considerable degree of recrystallization of the sample. It should be pointed out that the multiply twinned samples will revert to single domain crystals if stored for a month at room temperature.29 This process also removes the diffraction signal from elemental Sn. Diffraction data from the crystals were then collected at the Crystal beamline at the synchrotron Soleil. The wavelength was set to 0.42484 Å, close to the absorption edge of Sn, in order to optimize elemental contrast, and the sample to detector distance was set to 100 mm. All full measurements were made as single run phi scans 0−360° with one-degree steps with the kappa angle set to −60° and the theta angle to 26°. This setup certainly does not generate a full sphere of data, but it is a good compromise that allows the generation of data sets with a high degree of completeness in a short time and allows C



■

for the presence of a hot-air blower close to the sample but avoids heating of the detector. In a first, temperature dependent, sequence of measurements, single crystals of Sn-rich stistaite, as produced by the centrifugation method, were measured from room temperature to 370 °C, which is above the transformation temperature from stistaite to ht-Sn3Sb2. Heating was accomplished by a traditional hot-air blower. The results clearly show that the residual Sn, if indeed present, was insufficient to recreate ht-Sn3Sb2, and instead the sample shows loss of Sn as indicated by the change of the length of the q-vector. This tin concentration dependent q-vector change is evident in Figure 3. The figure contains data

Article

ANALYSIS OF DATA

The data from the beamline Crystal was acquired on a Titan detector from Oxford diffraction, and the CrysAlisPro31 suite of software was used for data reduction. This software allows for the evaluation of diffraction patterns with 3 q-vectors only, but due to the special symmetry of the diffraction pattern of htSn3Sb2 where the q-vectors adhere to the form (±α ±α ±α), it is possible to formally treat the data as (3 + 3)D. To achieve this, a coordinate system change is introduced by transforming the previous q-vectors to axial q-vectors of the form q1 = (α 0 0), q2 = (0 α 0), and q3 = (0 0 α) and adding an extinction condition hklmnp, with only m,n,p equal parity allowed. Only first order satellites were observed. The fundamental structure is primitive, although the basis structure is closely related to rock salt, making the superspace group symmetry Xm3m ̅ . The centering vectors in (3 + 3)D space are given by 0001/21/20, 0001/201/2, 00001/21/2. An alternative interpretation of the diffraction pattern is a twinned rhombohedral structure with a (3 + 1)D modulated structure. Refinement in this model is possible, but the agreement between model and data is considerably worse (satellite agreement worse than 20%). Additionally there are other compelling arguments against a twinned model. There is no indication of any further phase transitions on heating, but ht-Sn3Sb2 melts congruently at 324 °C. Further, the phase is clearly distinct from stistaite, and it appears very strange that the two should have the same structure but form two distinct, adjacent phases in the phase diagram. Stistaite formed from htSn3Sb2 is multiply twinned as expected from a transition from cubic to rhombohedral symmetry, and the elemental tin formed in this transition is also mimetically twinned according to the cubic symmetry of the ht-Sn3Sb2 phase. A cubic model was constructed in the program JANA20061 using a single atom in the basic unit cell, a = 3.0692(9) Å, and occupational modulation allowing the interchange of Sn for Sb as well as positional modulation allowing for relaxation. The modulation was described by simple harmonic functions, and the contributions from the fundamental modes (q1, q2, q3) were suppressed, allowing only contributions from the combinations q1 + q2 + q3 = q1′, -q1 + q2 − q3 = q2′, q1 − q2- q3 = q3′, and −q1- q2 + q3 = q4′. This makes the model equivalent to that of a data set where the q1−4′ vectors are defined as (±α ±α ±α). The single atomic position is refined as one fully occupied atom, with the partial occupancies of Sn and Sb being constrained so that their sum is unity and the overall composition is Sn3Sb2. The four occupational modulation waves and the four positional modulations generate a total of two refinable parameters for the modulation. The position of the single atom is fixed and the single refinable parameter for the basic structure is the free parameter for the anisotropic displacement parameter. Together with the overall scale factor, this yields a grand total of 4 refinable parameters for this very complex structure. The number of independent observations is 301, and out of these, 205 have intensities above the detection level of 3σ. The relationship among the four vectors q1′, q2′, q3′, and q4′ and the axial vectors q1, q2, and q3 is shown in Figure 6. Having a structure with four q-vectors is peculiar as the number of modulated structures with two modulation vectors is limited and the number with three or more is very small. Esmaeilzadeh mentions seven in a paper from 200132 including

Figure 3. Relationship between q-vector and composition established in ref 29 allowed one to match the development of increasingly Sb-rich stistaite with temperature. Thus, it became clear that heating a sample with only minor amounts of surface Sn produces a loss of Sn in the bulk of the sample and does not yield ht-Sn3Sb2.

from an earlier paper (ref 2, squares), together with new data from this publication (dots). Further, reciprocal space reconstructions (Figure 4) show clearly that the modulation remains (3 + 1)D since only one of the two cube body diagonals present in the figure hosts satellites, while in htSn3Sb2 all body diagonals must contain satellite reflections due to its cubic symmetry. In a second experiment, a single crystal of Sn-rich stistaite was enclosed in a capillary together with a similar amount of Sn. The sample was heated in the beam, and the diffraction pattern was monitored. There was no visible reaction at the reported lower existence range for Sn3Sb2 (242 °C), but at the melting point of Sn (250 °C) the disappearance of the powder rings for Sn was accompanied by the emergence of additional satellite reflections for Sn3Sb2. This temperature was maintained for a full diffraction experiment. Reconstruction of the reciprocal lattice reveals that satellites associated with an incommensurate modulation are present in all the [111] directions of the underlying cubic reciprocal lattice. This is compatible with a twinned rhombohedral (3 + 1)D modulated structure or with a cubic (3 + 4)D modulated structure. The q-vectors (in cubic setting) are given by q1 = (0.473 0.473 0.473), q2 = (0.473 −0.473 −0.473), q3 = (−0.473 −0.473 0.473), and q4 = (−0.473 0.473 −0.473). A section of reciprocal space is shown in Figure 5. D


Article


Figure 4. Reciprocal lattice reconstructions of the single crystal X-ray diffraction pattern of single crystalline stistaite with a nominal composition close to Sn4Sb3. Note how heating from room temperature through 300 to 370 °C produces a diffraction pattern with a decreasing distance between the first order satellites (inclined scarlet arrows). The decrease in distance may not be evident from the images, but full statistical analysis of the data shows the relationship between the q-vectors and the measurement temperature that is illustrated in Figure 3. This is directly translatable to an elongation of the q-vector, which in turn is indicative of an increasing Sb content. Note that satellites are present only along the 001 direction (trigonal setting), while the other body diagonal of the corresponding cubic unit cell (dashed line, only indicated for room temperature measurement) is devoid of satellite reflections.

Figure 6. Axial vectors q1, q2, and q3 (purple) combine to yield the modeling vectors q1′, q2′, q3′, and q4′ (green) by simple linear combinations. This is formally modeled by an extinction condition requiring all indices m,n,p to have equal parity, congruent to a face centering in normal 3D crystallography.

Figure 5. Reconstructed reciprocal lattice section of ht-Sn3Sb2. Note the appearance of new satellite reflections corresponding to a second q-vector (red arrows indicate q1 and q2). The section is generated by a cut through the face diagonal of the reciprocal cell, the 111* and the 111̅* direction supporting the q1 and q2 directions being obvious.

is not possible to constrain the local individual occupancies to stay between 0 and 1 for a (3 + 3)D modulated structure. The occupancy refinement is naturally closely correlated to the model for the anomalous scattering, and this is well-known to be difficult to establish with high accuracy33 for values close to the absorption edge. An attempt to establish values for f ′ and f ″ was made by refining the structure of stistaite using one of the (3 + 1)D data sets acquired. This leads to values for f ′ and f ″ that significantly differ from those tabulated. Using these values to refine the structure of ht-Sn3Sb2 however leads to a model where the values of the sinusoidal occupation functions for Sn and Sb overshoot unity and go below 0. In the final model, the amplitude of the harmonic modulation was adjusted to a value

various forms of modified Bi2O3, wüstite, V6Ni16Si7, and Cu9BiS6.

■

RESULTS AND DISCUSSION The final agreement between model and data was Rw(obs) = 5.2% for main reflections (42 independent with I > 3σ) and Rw(obs) = 16.6% for first order satellites (259 independent, 163 observed). There is an issue with the occupancy of Sn/Sb. The sum of the occupancies is fixed to unity for all positions, and the ratio of Sn to Sb is fixed to the nominal composition, but it E


Article


Figure 7. Structure of ht-Sn3Sb2 shown in a (100) projection. Red atoms are Sn; blue atoms are Sb. Bonds are generated for heteroatomic contacts Sn−Sb so that the NaCl-type clusters become obvious. The interleaving layers of pure Sn are also clearly seen as gaps between the roughly cubic clusters. It should be noted that the assignment of elemental species is only statistical; the occupancy function for Sn/Sb is modeled highly successfully by simple harmonic functions, and the occupancy pattern is quite disordered.

that is physically meaningful, and the values of f ′ and f ″ for Sn were refined. The sinusoidal nature of the occupancy for Sn and Sb does not furbish a clear-cut picture of the structure, but if the results are interpreted to make the dominant atomic species representative for each individual position, the picture is quite clear. The structure of ht-Sn3Sb2 consists of cubic rock salt type clusters, SnSb, roughly 7 × 7 × 7 atoms large (Figure 7). The clusters terminate with Sn layers, and the surplus Sn, making the composition Sn3Sb2 rather than SnSb, is found in layers that interleave these clusters. The arrangement is most probably quite dynamic, with a rapid exchange of Sn and Sb positions, given both the unquenchable phase transformation that occurs on cooling and the low reported decomposition temperature of the phase at 324 °C. The behavior of the phase ht-Sn3Sb2 is remarkable in many aspects. It is stable in a narrow temperature interval, and below that interval, at 242 °C, it decomposes in a topochemical reaction to form Sn-rich stistaite and solid elemental Sn. At the high end of the stability interval, the phase decomposes to stistaite and a Sn-rich melt. In this Account, it was shown that this topochemical reaction is reversible not only with the freshly prepared sample containing domains of stistaite and Sn in an intimate

intergrowth but also from a single domain stistaite crystal and a droplet of Sn. This reaction was observed to be instantaneous with molten Sn at 250 °C, but it was not observed with solid Sn at 242 °C, which may be a question of slow kinetics.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Sven Lidin: 0000-0001-9057-8233 Laura C. Folkers: 0000-0002-3424-1932 Notes

The authors declare no competing financial interest. Biographies Sven Lidin received his Ph.D. from Lund University in 1990. He was a postdoctoral fellow at the Department of Applied Mathematics, Research School of Physics, in Canberra and at the Max-Planck Institute für Festkörperforschung in Stuttgart. He was professor of Inorganic Chemistry at Stockholm University from 1996 to 2010. He is currently professor of Inorganic Materials Chemistry at Lund University. F


Article


(23) Litvinchuk, A. P.; Lorenz, B.; Chen, F.; Nylen, J.; Häussermann, U.; Lidin, S.; Wang, L. M.; Guloy, A. Optical and electronic properties of thermoelectric Zn4Sb3 across the low-temperature phase transitions. Appl. Phys. Lett. 2007, 90, 181920. (24) Nylén, J.; Lidin, S.; Andersson, M.; Iversen, B. B.; Newman, N.; Häussermann, U. Low-temperature structural transitions in the phonon-glass thermoelectric material beta-Zn4Sb3: Ordering of Zn interstitials and defects. Chem. Mater. 2007, 4, 834−838. (25) Nylén, J.; Andersson, M.; Lidin, S.; Häussermann, U. The structure of alpha-Zn4Sb3: Ordering of the phonon-glass thermoelectric material beta-Zn4Sb3. J. Am. Chem. Soc. 2004, 126, 16306− 16307. (26) Tengå, A.; Lidin, S.; Belieres, J. P.; Newman, N.; Wu, Y.; Häussermann, U. Metastable Cd4Sb3: A Complex Structured Intermetallic Compound with Semiconductor Properties. J. Am. Chem. Soc. 2008, 130, 15564−15572. (27) Boström, M.; Lidin, S. The incommensurably modulated structure of ζ-Zr3‑xSb2. J. Alloys Compd. 2004, 376, 49−57. (28) Massalski, T. B. Binary Alloy Phase Diagrams, 2nd ed.; ASM International: Materials Park, OH, 1990; Vol. 3, pp 3304−3306. (29) Lidin, S.; Christensen, J.; Jansson, K.; Fredrickson, D.; Withers, R.; Norén, L.; Schmid, S. Incommensurate Stistaite-Order Made to Order. Inorg. Chem. 2009, 48, 5497−5503. (30) Boström, M.; Hovmoller, S. Preparation and crystal structure of the novel decagonal approximant Mn123Ga137. J. Alloys Compd. 2001, 314, 154−159. (31) Agilent CrysAlis PRO; Agilent technologies Ltd, Yarnton, Oxfordshire, England, 2014. (32) Esmaeilzadeh, S.; Lundgren, S.; Hålenius, U.; Grins, J. Bi1‑xCrxO1.5+1.5x, 0.05 < x < 0.15: A new high-temperature solid solution with a three dimensional incommensurate modulation. J. Solid State Chem. 2001, 156, 168−180. (33) Chantler, C. T. ″X-ray Form Factor, Attenuation, and Scattering Tables. J. Phys. Chem. Ref. Data 1995, 24, 71−643.

Laura C. Folkers received her M.Sc. from the ETH Zürich in 2015. She is currently a Ph.D. student in Inorganic Chemistry at Lund University.

■

REFERENCES

(1) Petricek, V.; Dusek, M.; Palatinus, L. JANA2006, Institute of Physics, Academy of Sciences of the Czech Republic, 2017. (2) Dehlinger, U. On the Broadening of Debye lines in cold treated metals. Z. Kristallogr. - Cryst. Mater. 1927, 65, 615−631. (3) De Wolff, P. M. The Pseudo-Symmetry of Modulated Crystal Structures,. Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr. 1974, A30, 777−785. (4) Jansen, T. Private Communication. (5) Hastings, J. M.; Pouget, J. P.; Shirane, G.; Heeger, A. J.; Miro, N. D.; Macdiarmid, A. G. One-dimensional Phonons and phase-ordering phase-transion in Hg3‑δAsF6. Phys. Rev. Lett. 1977, 39, 1484−1487. (6) Bak, P.; Janssen, T. Symmetry of modulated phases in tetrathiafulvalene-tetracyanoquinodimethane (TTF-TCNQ) − 4dimensional and 5-dimensional superspace groups. Phys. Rev. B: Condens. Matter Mater. Phys. 1978, B17, 436−439. (7) Graham, A. R.; Thompson, R. M.; Berry, L. G. Studies of Mineral Sulfosalts: XVII − Canizzerite. Am. Mineral. 1953, 38, 536−544. (8) Flieher, G.; Vollenkle, H.; Nowotny, H. Die Kristallstruktur von Mn15Si26. Monatsh. Chem. 1967, 98, 2173−2179. (9) Völlenkle, H.; Wittman, A.; Nowotny, H. Abkömmlinge der TiSi2-structur - ein neues Bauprizip. Monatsh. Chem. 1966, 97, 506− 516. (10) Matzat, E. Modulated structure of Canizzerite. Acta Crystallogr. 1978, A34, s13−s13. (11) Lidin, S.; Lind, H.; Eriksson, L.; Larsson, A.-K.; Rohrer, F. E. On the question of incommensurability − The Nowotny chimney ladder structures revisited. Z. Kristallogr. - Cryst. Mater. 2000, 215, 650−660. (12) Rohrer, F. E.; Lind, H.; Eriksson, L.; Larsson, A.-K.; Lidin, S. Incommensurately modulated The Nowotny chimney ladder phases Cr1‑xMoxGe∼1. 75 with x = 0.65 and 0.84. Z. Kristallogr. - Cryst. Mater. 2001, 216, 190−198. (13) Sun, J.; Lee, S.; Lin, J. Four dimensional spece groups for Pedestrians: Composite structures. Chem. - Asian J. 2007, 2, 1204− 1229. (14) Villars, P., Calvert, L. D. Pearson’s Hanbook of Crystallographic Data for Intermetallic Phases, 2nd ed., ASM International: Materials Park, 1991. (15) Larsson, A.k.; Stenberg, L.; Lidin, S. Crystal-structure modulations in ε-Cu5Sn4. Z. Kristallogr. - Cryst. Mater. 1995, 210, 832−837. (16) Lidin, S.; Elding, M.; Stenberg, L. Modulations of B8-type intermetallics. Acta Crystallogr., Sect. A: Found. Crystallogr. 1996, A52, C367−C367. (17) Elding-Pontén, M.; Stenberg, L.; Lidin, S.; Madariaga, G.; PeresMato, J.-M. Structure of Mn8Sn5. Acta Crystallogr., Sect. B: Struct. Sci. 1997, B53, 364−372. (18) Elding-Pontén, M.; Stenberg, L.; Lidin, S. The ε-phase field of the Cu-In system.,. J. Alloys Compd. 1997, 261, 162−171. (19) Roof, R. B.; Elliot, G. R. B. X-Ray Diffraction Evidence for Microphases in CeCd∼4.5 Solid Solutions. Inorg. Chem. 1965, 4, 691− 697. (20) Piao, S. Y.; Palatinus, L.; Lidin, S. All the Disorder Mechanisms in the 13:58 Phases Come Together. Out of the Modulated Confusion Rises the Remarkable Phase Ce12.60Cd58.68(2). Inorg. Chem. 2008, 47, 1079−1086. (21) Wang, F.; Veremchuk, I.; Lidin, S. Tuning Crystal Structures and Thermoelectric Properties through Al Doping in ReSi1.75. Eur. J. Inorg. Chem. 2017, 2017, 47−55. (22) Nylén, J.; Lidin, S.; Andersson, M.; Liu, H.; Newman, N.; Häussermann, U. Effect of metal doping on the low-temperature structural behavior of thermo electric b-Zn4Sb3. J. Solid State Chem. 2007, 180, 2603−2615. G


In Situ Synthesis and Single Crystal Synchrotron X-ray Diffraction

Recommend Documents