# Fileset

[Chemistry of Materials---Understanding the Prospects of the Thermoelectric Performance of the YbMg2(Bi,Sb)2 Zintl Phase.pdf](https://mdr.nims.go.jp/filesets/e536c18c-b55d-4c99-94c0-1ead2b027b59/download)

## Creator

[Kushal Mehrotra](https://orcid.org/0000-0002-4940-0967), [Andrei Novitskii](https://orcid.org/0000-0002-7304-806X), [Takao Mori](https://orcid.org/0000-0003-2682-1846)

## Rights

[Creative Commons BY Attribution 4.0 International](https://creativecommons.org/licenses/by/4.0/)

## Other metadata

[Understanding the Prospects of the Thermoelectric Performance of the YbMg<sub>2</sub>(Bi,Sb)<sub>2</sub> Zintl Phase](https://mdr.nims.go.jp/datasets/e10f7282-8f34-4704-9f9b-c79aff25aef4)

## Fulltext

Understanding the Prospects of the Thermoelectric Performance of the YbMg2(Bi,Sb)2 Zintl PhaseUnderstanding the Prospects of the Thermoelectric Performance ofthe YbMg2(Bi,Sb)2 Zintl PhasePublished as part of Chemistry of Materials special issue “Honoring the Outstanding Contributions of MercouriKanatzidis to Chemistry of Materials.”Kushal Mehrotra, Andrei Novitskii,* and Takao Mori*Cite This: Chem. Mater. 2025, 37, 6782−6790 Read OnlineACCESS Metrics & More Article Recommendations *sı Supporting InformationABSTRACT: AM2X2 Zintl compounds, crystallizing in layered structures, haverecently garnered attention due to their promising thermoelectric properties. Inthis study, we explore the chemical bonding and elastic and thermoelectricproperties evolution across the full YbMg2Bi2−xSbx solid solution. The transitionfrom YbMg2Bi2 to YbMg2Sb2 leads to a continuous linear chemical-bondshortening and thus a significant enhancement in elastic moduli and soundvelocity, resulting in overall significant lattice stiffening. Simultaneously, the shifttoward more ionic chemical bonding leads to significant changes in the bandstructure, particularly an increase in effective mass and a decrease in both carrierconcentration and mobility, which in turn reduces the power factor for Sb-richsamples. However, a rapid increase in point-defect scattering causes the latticethermal conductivity to drop from ≈3 to almost 1 Wm−1−K−1 at 300 K for the intermediate compositions, thus opening new roomfor further optimization of the Sb-rich representatives of the YbMg2Bi2−xSbx. Therefore, in this work, we have demonstrated thatdespite the seemingly intrinsically higher thermoelectric performance in the Bi-rich region of the YbMg2Bi2−xSbx solid solution, theSb-rich representatives may in fact be even more promising due to better mechanical and thermal stability and greater room forfurther charge carrier concentration optimization.■ INTRODUCTIONThermoelectric (TE) devices offer a solid-state means toconvert heat into electricity, holding promise for waste heatrecovery and refrigeration applications.1 The conversionefficiency of a TE device is governed by the dimensionlessfigure of merit (zT) of a material used as a working body.2 Inturn, zT = α2σT/κtot, where α is the Seebeck coefficient, σ isthe electrical conductivity, T is the absolute temperature, andκtot = κlat + κel is the total thermal conductivity, representingthe sum of the lattice and electronic thermal conductivities,respectively.3 Achieving a high zT requires a large power factor(α2σ) and low κtot, which is not a trivial task since α, σ, and κtotare inherently interdependent.Zintl phases have emerged as promising TE materials in thiscontext, as their intrinsically low κlat and complex crystal andelectronic structures make it possible to decouple electronicand thermal transport.4,5 In particular, layered AM2X2compounds (where A is a divalent cation, M a metallicelement, and X a pnictogen) crystallizing in the CaAl2Si2-typestructure have drawn extensive interest.6 These compoundsconsist of covalent two-dimensional [M2X2]2− (M = Mg, Zn,Cd; X = Sb, Bi) anionic slabs sandwiched between A2+ (A =Ca, Mg, Eu, Yb) cation monolayers, and many exhibit excellentTE performance (for example, AZn2Sb2, Mg3Sb2, and AMg2Bi2families).7−9Many of the highest zT results among AM2X2 type Zintlcompounds were obtained via isoelectronic substitution oralloying on the cation sublattices, increasing carrier mobilityand/or inducing band convergence.4,7 AMg2Bi2-based com-pounds, in particular, have been recently reported to achievezT ≈ 1.3 at 873 K without the use of toxic elements (e.g., inEu0.2Yb0.2Ca0.8Mg2Bi2).10 YbMg2Bi2 in particular has showncompetitive thermoelectric performance in recent years (e.g.,YbMg2Bi1.58Sb0.4 with zT ≈ 1 at 873 K).11,12 Yet, Bi-rich Zintlphases generally exhibit weaker chemical bonding compared totheir Sb-rich counterparts,6,13 owing to the larger atomic radiusand lower electronegativity of Bi.14,15 Such a softly bondedlattice typically results in a low Young’s modulus, thus lowerκlat, but also poor mechanical and thermal stability.12,16,17 Inthe case of YbMg2Bi2, their weakly bonded interlayers can leadReceived: June 11, 2025Revised: July 28, 2025Accepted: August 13, 2025Published: August 26, 2025Articlepubs.acs.org/cm© 2025 The Authors. Published byAmerican Chemical Society6782https://doi.org/10.1021/acs.chemmater.5c01433Chem. Mater. 2025, 37, 6782−6790This article is licensed under CC-BY 4.0Downloaded via NATL INST FOR MATLS SCIENCE (NIMS) on September 10, 2025 at 11:52:18 (UTC).See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.https://pubs.acs.org/curated-content?journal=cmatex&ref=featurehttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Kushal+Mehrotra"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Andrei+Novitskii"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Takao+Mori"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/action/showCitFormats?doi=10.1021/acs.chemmater.5c01433&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?goto=articleMetrics&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?goto=recommendations&?ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?goto=supporting-info&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=tgr1&ref=pdfhttps://pubs.acs.org/toc/cmatex/37/17?ref=pdfhttps://pubs.acs.org/toc/cmatex/37/17?ref=pdfhttps://pubs.acs.org/toc/cmatex/37/17?ref=pdfhttps://pubs.acs.org/toc/cmatex/37/17?ref=pdfpubs.acs.org/cm?ref=pdfhttps://pubs.acs.org?ref=pdfhttps://pubs.acs.org?ref=pdfhttps://doi.org/10.1021/acs.chemmater.5c01433?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://pubs.acs.org/cm?ref=pdfhttps://pubs.acs.org/cm?ref=pdfhttps://acsopenscience.org/researchers/open-access/https://creativecommons.org/licenses/by/4.0/https://creativecommons.org/licenses/by/4.0/https://creativecommons.org/licenses/by/4.0/to microstructural instability at high temperatures (e.g., theemergence of voids or phase segregation upon long-termannealing), surface oxidation, and deterioration when exposedto air or water, ultimately degrading its transport proper-ties.11,12 These reliability challenges underscore a need tobalance intrinsic TE performance with mechanical robustnessand chemical stability in Bi-rich Zintl phases.YbMg2Sb2, in turn, is a poor conductor exhibiting very highresistivity and a large Seebeck coefficient when undoped due toits semiconducting nature.18 At the same time, YbMg2Sb2 hassignificantly shorter Yb−Sb and Mg−Sb bonds (higher bondenergy) than the corresponding bonds in YbMg2Bi2.Strengthening the bonding is counter to the conventionalstrategy of lattice softening,16 yet it directly addresses thestability and durability concerns. Consequently, strengtheningthe lattice by partial Sb substitution on the Bi sublattice is oneof the promising approaches to overcoming the processing/handling and stability issues of YbMg2Bi2. Recent studies byWang et al. and Liang et al. have provided new insights into theeffects of Sb alloying in YbMg2Bi2, demonstrating that amoderate increase in bonding strength can considerablyimprove the material’s resistance to thermal degradation andmechanical failure.11,12 Ab initio calculations and soundvelocity measurements confirm that replacing Bi with Sbstiffens the lattice in the YbMg2Bi2−xSbx solid solutions. Inaddition, Sb-alloyed samples were found to better preservetheir TE properties after long-term heating, whereas Bi-richsamples suffered noticeable performance degradation. Thisevidence illustrates that chemical-bond strengthening via Sbsubstitution can yield a more robust thermoelectric materialwithout sacrificing efficiency, providing a valuable designstrategy for practical applications. To date, studies on theYbMg2(Bi,Sb)2 system have largely focused on keeping theanion entity fixed to be either Bi or Sb, while substituting Yband/or Mg-site with isovalent elements. To the best of ourknowledge, only two studies, by Wang et al.11 and Liang etal.,12 have investigated Bi to Sb substitution. However, due tothe increased porosity in Sb-rich samples as mentioned byWang et al.,11 their investigation was limited to x = 0.4 in Sbcontent in YbMg2Bi1.98−xSbx. However, a comprehensiveinvestigation of the solid solution series YbMg2Bi2−xSbx isimportant for elucidating the continuous evolution of thematerial’s transport, structural, and mechanical properties.Such an investigation can reveal whether any intermediatecompositions offer superior trade-offs, e.g., an optimal alloythat maximizes phonon scattering while maintaining sufficientcarrier mobility, or any nonlinear effects (e.g., bandconvergence or even structural transitions) that occur atcertain Bi/Sb ratios.In this work, we present a systematic study of theYbMg2(Bi,Sb)2 solid solution covering the entire range fromYbMg2Bi2 to YbMg2Sb2. A series of polycrystallineYbMg2Bi2−xSbx samples was synthesized and characterized toevaluate how gradually substituting Bi for Sb affects thematerial’s thermoelectric and elastic properties as well aschemical bond strength. We reveal the relationship betweenanion composition and key transport properties, as well asmaterial hardness and elastic properties. The insights gainedhere not only demonstrate the tunability of YbMg2(Bi,Sb)2materials for improved zT and durability but also may guidefurther improvement in the Sb-rich end (e.g., through doping,band structure modifications, or microstructural engineering)to advance these Zintl phases toward practical thermoelectricapplications.■ MATERIALS AND METHODSSynthesis. Bulk polycrystalline samples with nominal composi-tions of YbMg2Bi2−xSbx (x = 0, 0.2, 0.4, 0.6, 0.8, 1, 1.2, 1.4, 1.6, 1.8,and 2) were synthesized using mechanochemical synthesis, followedby densification via spark plasma sintering (SPS). Ytterbium ingots(Yb, Alfa Aesar, 99.9%), bismuth chunks (Bi, 5N Plus, ≥99.999%),antimony shots (Sb, Sigma-Aldrich, 99.999%), and magnesiumturnings (Mg, Sigma-Aldrich, 99.95%) were weighed according tothe stoichiometric ratios and loaded into a stainless steel milling jaralong with two 1/2-in. diameter stainless steel balls inside an Ar-filledglovebox with oxygen and water vapor levels below 5 ppm.Subsequently, the mixtures were subjected to high-energy ball millingin an 8000D Mixer/Mill (SPEX SamplePrep). Based on previousreports11,12,19 and preliminary ball milling process optimization(Figures S1−S3), 12 h of milling with a ball-to-powder ratio of16:7 was used to ensure phase formation and reproducibility. All ofthe as-milled powders were consolidated via SPS (322 LX, Fuji-SPS,Japan) at 873 K for 5 min in a Ø10 mm graphite die under 50 MPauniaxial pressure in Ar atmosphere with heating and cooling rates of100 and 25 K min−1, respectively. The densified pellets were thenannealed at 773 K for 10 h in an evacuated quartz tube.Structural and Chemical Characterization. PXRD patternswere collected at room temperature using a MiniFlex diffractometer(Rigaku, Japan) with Cu−Kα radiation (λ = 1.5406 Å). Latticeparameters were calculated through Rietveld refinement of PXRDpatterns by using the FullProf software package.20 The microstructureand elemental composition of the sintered samples were examined byfield emission scanning electron microscopy (FESEM; HitachiSU8230, Japan) in conjunction with an energy-dispersive X-rayspectroscopy (EDS) detector (X-MaxN, Horiba Scientific, Japan).Elastic Moduli and Sound Velocity. To investigate elasticmoduli, the longitudinal (νl) and transverse (νt) sound velocities ofYbMg2Bi2−xSbx were measured by a sing-around ultrasonic velocitymeasuring instrument (UVM-2, Ultrasonic Engineering Co., Japan) atroom temperature. The Young’s (E), bulk (K), and shear (G) moduliwere calculated asikjjj y{zzzEd v vv vK d G d(3 4 )43, andt2l2t2l2t2l2t2t2== =(1)where d is the sample density measured through the Archimedesmethod with ethanol as the working fluid. In the Debye model, thephonon density of states (pDOS) scales as g(ω) ∝ ω2/v3, so that eachacoustic branch enters thermodynamic and transport integrals with aweight v−3.21,22 Accordingly, the average sound velocity va is taken asthe cubic-harmonic meanÄÇÅÅÅÅÅÅÅÅÅÅÅikjjjjjy{zzzzzÉÖÑÑÑÑÑÑÑÑÑÑÑvv v131 2al3t31/3= +(2)which properly accounts for one longitudinal and two transversemodes in proportion to their contributions to the total pDOS.23Vickers Hardness Measurement. Vickers hardness of eachsample was measured via the indentation method using a microVickers hardness tester (HMV-G31, Shimadzu Co., Japan) on thesurface of polished samples under a load of F = 1.961 N and a holdtime of 10 s per measurement. Vickers hardness values werecalculated as HV = 1.891F/(2l)2 (2l is the diagonal length of theindent). To get a constant average value and the respective error bar,at least 5 indents were collected.Thermal Property Measurement. The total thermal con-ductivity (κtot) was estimated through the well-known formula κtot =χCpd, where χ is the thermal diffusivity and Cp is the specific heatChemistry of Materials pubs.acs.org/cm Articlehttps://doi.org/10.1021/acs.chemmater.5c01433Chem. Mater. 2025, 37, 6782−67906783https://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfpubs.acs.org/cm?ref=pdfhttps://doi.org/10.1021/acs.chemmater.5c01433?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ascapacity. The thermal diffusivity χ was measured on graphite-coateddisk-shaped samples using the laser flash technique (LFA 467Hyperflash, Netzsch, Germany) and analyzed using a modified Cape−Lehman model with pulse correction.27 The Cp was approximatedusing the model proposed by Agne et al. and calculated for eachcomposition.28 The lattice thermal conductivity (κlat) was calculatedfrom κtot by subtracting the electronic contribution (κel) estimatedaccording to the Wiedemann−Franz law, where κel = σLT. L is theLorenz number derived from the Seebeck coefficient (α) as L = 1.5 +exp{−|α|/116} in the framework of the effective mass model withtransport in a single parabolic band limited by acoustic phononscattering (SPB-APS).29Electrical Transport and Hall Measurements. The Seebeckcoefficient (α) and electrical conductivity (σ) were simultaneouslymeasured on rectangular bars using a commercial apparatus (ZEM-3,ULVAC Inc., Japan) under partial He pressure. Electrical resistivityand Hall effect measurements were performed at room temperatureusing a standard five-probe configuration implemented in a physicalproperties measurement system (PPMS9T, Quantum Design Inc.).Electrical contacts were made with a 0.05 mm platinum wire andsilver paste (Ted Pella, Inc.). A thin layer of gold was predepositedonto the contact area to improve electrical contact. The Hallcoefficient (RH) was obtained from the linear fit of the magnetic fielddependence of the Hall resistivity in a range from ±5 T. The chargecarrier concentration (n) was also calculated assuming SPB-APS as n= rH/eRH, with e representing the elemental charge, and rHrepresenting the Hall factor, which depends on the chemical potentialand carrier scattering mechanism. Charge carrier mobility (μ) wascalculated as μ = σ = RH/rH. Samples were slightly polished prior toeach measurement. Between measurements, the samples were kept ina desiccator under a static vacuum of ≈2.5 × 10−3 mbar.The uncertainty of the Hall measurements was estimated to be in arange from 5 to 10%, 6% for the Seebeck coefficient, 8% for theelectrical conductivity, 11% for the thermal conductivity, and 16% forthe figure of merit zT.30 Error bars are not shown in some figures toimprove the readability.■ RESULTS AND DISCUSSIONPhase Identification and Microstructural Analysis.The room-temperature powder X-ray diffraction (PXRD)patterns of polycrystalline YbMg2Bi2−xSbx samples are shownin Figure 1a. The main diffraction peaks for all compositionscorrespond to the YbMg2(Bi,Sb)2 phase with a trigonalCaAl2Si2-type crystal structure (space group P3̅m1).24 Aweak reflection near 2θ ≈ 27° is observed in all samplesexcept for the pure YbMg2Sb2 end-member (x = 2), ashighlighted in Figure 1b. Within the detection limits of PXRD,this peak is attributed to residual Bi, typically observed as asecondary phase in this family of compounds.11,12,25,26,31,32With increasing Sb content, the main diffraction peaks shiftsystematically toward higher 2θ values (Figure 1b), indicatinga linear decrease in lattice parameters in agreement withVegard’s law (Figure 1c,d). This behavior is consistent with thesmaller atomic radius of Sb (1.45 Å) compared to Bi (1.60Å),33 and confirms the successful substitution of Sb at the Bisite, forming a complete solid solution across theYbMg2Bi2−xSbx series. The qualitative analysis was furthervalidated by Rietveld refinement (Table S1 and Figure S4),which quantitatively confirms the monotonic decrease in latticeconstants and the associated contraction of the unit cellvolume (Figure 1e). The refined lattice parameters for pristineYbMg2Bi2 and YbMg2Sb2 agree well with previously reportedvalues.12,24,25All YbMg2(Bi,Sb)2 samples exhibit a fine and densemicrostructure, typical for Zintl compounds, obtained byhigh-energy ball milling, followed by spark plasma sintering.The relative density was measured to be above 95% of thetheoretical density (Table S1) for all of the samples. EDSanalysis, in turn, confirms the presence of relatively small insize (much less than ≈1 μm) Bi inclusions, while also revealinga small amount of rather large (tens of microns) MgOFigure 1. (a) Room-temperature powder X-ray diffraction patterns for YbMg2Bi2−xSbx (x = 0, 0.2, 0.4, 0.6, 0.8, 1, 1.2, 1.4, 1.6, 1.8, 2) samples afterspark plasma sintering and annealing. (b) Enlarged section of part (a) in a 2θ range from 22 to 29°, where the Bi secondary phase has the mostintensive reflection indicated by a black solid triangle (▼). Bragg’s reflections for the YbMg2Bi2 phase, taken from ref 24 are indicated by gray tickson the top part of the figure. The lattice parameters (c) a, (d) c, and (e) unit cell volume V of YbMg2Bi2−xSbx are shown as a function of increasingnominal Sb content. The a and c parameters experience a slight linear contraction of ∼2% in good agreement with Vegard’s law, resulting in a ∼5%shrinkage of the unit cell volume V. Reference data shown in parts (c−e) taken from refs 11,12,18,24−26 (open gray symbols for polycrystals andfilled gray symbols for single crystals, respectively).Chemistry of Materials pubs.acs.org/cm Articlehttps://doi.org/10.1021/acs.chemmater.5c01433Chem. Mater. 2025, 37, 6782−67906784https://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig1&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig1&ref=pdfpubs.acs.org/cm?ref=pdfhttps://doi.org/10.1021/acs.chemmater.5c01433?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asinclusions that were not detected by XRD. As mentionedabove, such secondary phases are typically observed in thesecompounds, yet they are difficult to detect due to the highreactivity and susceptibility of the material to oxidation andreaction with water, which complicates measurements andaccurate SEM surface analysis.25 At the same time, the actualcomposition of main phase YbMg2(Bi,Sb)2 is confirmed to beclose to the nominal one, with a uniform distribution of theconstituent elements (Figures S5−S7).Composition-Dependent Elastic Properties. As men-tioned above, the unit cell volume (Figure 1e) and thusaverage bond length decrease with Sb content. As a result,Young’s modulus, as well as bulk and shear moduli, increaselinearly with increasing x due to Mg−(Bi,Sb) bonds becomingshorter13,34 and show an ≈45% increase with transition fromYbMg2Bi2 to YbMg2Sb2 (Figure 2a), which is quite noticeableand, perhaps, greater than one might expect from the 5%increase in unit cell volume. This trend also holds for the bulkand shear moduli, which increase by 13 and 57%, respectively.Although the shear-to-bulk modulus ratio increases toward Sb-rich compositions, which is typically associated with greaterbrittleness,35 our experimental observations during samplehandling suggest the opposite: Sb-rich samples were noticeablyeasier to work with, exhibiting greater compliance andresistance to fracture during polishing and cutting. This isalso corroborated by a 68% increase in Vickers hardness (HV)from 214 for YbMg2Bi2 to 360 for YbMg2Sb2 (Figure 2b). Aswell as elastic moduli and HV, average sound velocity alsoincreases from 1995 ms−1 for YbMg2Bi2 to 2786 ms−1 forYbMg2Sb2, demonstrating an ≈40% enhancement (Figure2c).36 Together, this indicates a clear tendency toward strongerbonds in Sb-rich YbMg2Bi2−xSbx, which may be attributed tothe greater electronegativity difference (Δχ) and thus higherPauling bond dissociation energy of Mg−Sb bonds (Δχ = 0.66,3.0 eV) compared with Mg−Bi bonds (Δχ = 0.3, 1.9 eV).15The increase in the Pauling bond dissociation energy andionicity of chemical bonding undoubtedly affect both theelectronic and thermal transport properties, which will bediscussed in detail in the following sections.Temperature-Dependent Thermoelectric Properties.For YbMg2Bi2−xSbx samples with x ≤ 1, σ exhibits a metal-liketemperature dependence (Figure 3a), as expected fordegenerate semiconductors.37 This also correlates with theSeebeck coefficient α, which increases almost linearly withtemperature and remains below 200 μVK−1 across the entiretemperature range studied (Figure 3b). In contrast, sampleswith x > 1 show thermally activated transport behavior, asevidenced by the decrease in σ with temperature and thecorresponding increase in the absolute values of α, indicatingnondegenerate transport behavior (Figure 3a,b). In turn, theincrease in the maximum Seebeck coefficient (Smax) and itsshift toward lower temperatures with increasing Sb contentsignifies an increase of the band gap Eg, according to theGoldsmid−Sharp relation, Eg = 2eSmaxTmax.38 The correspond-ing values of the thermal Eg, obtained from the temperaturedependence of the Seebeck coefficient (Figure 3b), 0.21 eV forYbMg2Bi2, 0.31 eV for YbMg2BiSb, and 0.38 eV for YbMg2Sb2,are in good agreement with the previous studies18,39 andconfirm the gradual increase in band gap with Sb substitution.The power factor for all of the samples (Figure 3c) iscalculated using the experimental α and σ values across theentire temperature range. The power factor shows anincreasing trend with temperature and then saturates at highertemperatures for compositions up to x = 0.8, while forcompositions beyond this, α2σ increases monotonically overthe entire temperature range. The temperature dependence ofα2σ follows that of α, while the change in magnitude withtemperature closely resembles the changes occurring in σ withT.The total thermal conductivity κtot gradually decreases withtemperature (Figure 3d), as expected.36 The maximum valuesof κtot are observed for the two end compositions among theseries of synthesized YbMg2Bi2−xSbx samples, while asignificant drop in thermal conductivity is obtained for theremaining compositions. The lattice thermal conductivity κlatof YbMg2Bi2 and YbMg2Sb2, in turn, exhibits a ∼T−1temperature dependence, while the remaining intermediatecompositions follow a ∼T−0.5 temperature dependence (Figure3e), signifying the changing scattering mechanism in thesolid−solution compositions over the pristine samples, asdiscussed in detail below. Although there is a decline in boththe power factor and thermal conductivity in Sb-substitutedsamples, the degree of reduction for κlat outperforms thereduction in α2σ, resulting in the highest zT for the x = 0.2composition. Overall, zT for all of the samples shows a steadyincrease with temperature, with the highest zT ≈ 0.9 at 723 Kachieved for the YbMg2Bi1.8Sb0.2 sample. Across the composi-tional range examined, the zT values reported in this work aregenerally consistent with those from previous studies by Wanget al.11 and Liang et al.,12 especially at room temperature,where all three studies show similar performance with onlyFigure 2. Evolution of (a) elastic moduli, (b) HV, and (c) soundvelocities with x in YbMg2Bi2−xSbx at room temperature. Note thatthe light gray lines are guides for the eye. Literature data forpolycrystalline YbMg2Bi2−x−δSbx (x ≤ 0.6) reported by Liang et al.(ref 12) is also shown for comparison (open gray symbols).Chemistry of Materials pubs.acs.org/cm Articlehttps://doi.org/10.1021/acs.chemmater.5c01433Chem. Mater. 2025, 37, 6782−67906785https://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig2&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig2&ref=pdfpubs.acs.org/cm?ref=pdfhttps://doi.org/10.1021/acs.chemmater.5c01433?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asminor variation (Figure S7f). At an elevated temperature (723K), the reported data in this work not only maintains overallagreement with literature but also demonstrates a slightlyhigher zT for the composition with x = 0.2, reaching ≈0.9(Figure S8). These observations underscore the reproducibilityof thermoelectric trends across various studies, considering theuncertainty in zT estimation.30Composition-Dependent Electrical Properties. Weanalyzed and compared the electronic properties of our databy using an effective mass model assuming a single parabolicband transport limited by acoustic phonon scattering at roomtemperature. The Pisarenko plot (Figure 4a) shows a clear shifttoward higher effective masses as Bi is progressively replacedwith Sb. The increasing density of states effective mass (md*)reveals the possible evolution of the band structure upon BiFigure 3. Temperature dependence of the (a) electrical conductivity σ, (b) Seebeck coefficient α, (c) power factor α2σ, (d) total κtot and (e) latticeκlat thermal conductivities, and (f) the figure of merit zT for YbMg2Bi2−xSbx (x = 0, 0.2, 0.4, 0.6, 0.8, 1, 1.2, 1.4, 1.6, 1.8, 2) samples.Figure 4. Seebeck coefficient α of YbMg2Bi2−xSbx samples (a) versus charge carrier concentration n (Pisarenko plot) and (b) logarithm of electricalconductivity σ (Jonker plot) at room temperature. The solid lines in part (a) represent the Seebeck coefficient calculated in the framework of theeffective mass model,37 assuming acoustic phonon scattering (r = −1/2), md* = 0.4, and md* = 0.7, respectively. The dashed lines in part (b)represent the Pisarenko formula for different weighted mobility values (μw in cm2 V−1 s−1),40 each labeled next to its corresponding line at thebottom part of the figure. Evolution of charge carrier (c) concentration n and (d) mobility μ with x in YbMg2Bi2−xSbx at room temperature. Thelight gray lines in parts (c) and (d) are guides to the eye. Literature data for other YbMg2Bi2−xSbx (0 ≤ x ≤ 0.4) reported previously (refs10,11,18,19,24−26) is also shown for comparison (open gray symbols for polycrystals and filled gray symbols for single crystals, respectively).Chemistry of Materials pubs.acs.org/cm Articlehttps://doi.org/10.1021/acs.chemmater.5c01433Chem. Mater. 2025, 37, 6782−67906786https://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig3&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig4&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig4&ref=pdfpubs.acs.org/cm?ref=pdfhttps://doi.org/10.1021/acs.chemmater.5c01433?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-assubstitution with Sb. This is further supported by the Jonkerplot (Figure 4b), which deviates from the expected slope of−86.291 μVK−1 (shift toward lower weighted mobility; μw) foracoustic phonon-limited scattering, indicating changes in theband structure and/or scattering mechanisms.40 The fact thatcarrier mobility (μ) is inversely proportional to band effectivemass (mb*) through μ ∝ (mb*)−5/2 suggests that the decrease inmobility is due to an increase in mb* with increasing x.Therefore, according to the relation md* = Nv2/3mb*, the DOSeffective mass (md*) should also increase with increasing mb*,i.e., with increasing Sb content, as is observed in Figure 4a. Theincreasing md* values with increasing Sb content is alsosupported by the insights drawn from previous DFTcalculations on YbMg2Bi2 and YbMg2Sb2 systems, whereYbMg2Bi2 is identified as a direct band gap semiconductor witha smaller effective mass (m* = 0.6) compared to the Sb-basedsystem (m* = 1).18,39Interestingly, a subtle change in the Seebeck coefficient forcompositions 0 < x ≤ 1 (110 μVK−1 for x = 0 to 130 μVK−1for x = 1) followed by significant changes in α forcompositions beyond x = 1 (158 μVK−1 at x = 1.2 to 366μVK−1 at x = 2) (Figure 4a,b) can be attributed tocorresponding changes in the charge carrier concentration(n), as shown in Figure 4c, where no change in the order of n isobserved for Bi-rich compositions while a significant reductionfollows for the Sb-rich compositions, thus influencing the trendin α with Sb content at room temperature. The decrease incharge carrier concentration, though not monotonically, uponSb content can be mainly attributed to the increase of the bandgap, accompanied by the increasing electronegativity differencebetween the cation (Yb2+) and the anionic framework[Mg2(Bi,Sb)2]2−. In addition to this, the influence of secondaryphases, in particular elemental Bi, cannot be excluded. Thereduction of elemental Bi fraction (Figure 1b), along with theincrease in the band gap, may affect the defect energetics,thereby contributing to the decrease in charge carrierconcentration.32 Thus, the observed trend in charge carrierconcentration likely reflects the combined influence ofevolution in the electronic structure and native defectenergetics. To provide a more reliable confirmation of theincrease in band gap, Eg, we used the Goldsmid−Sharprelation,38 Eg = 2eSmaxTmax, to estimate thermal Eg from theSeebeck coefficient data (Figure 3b). While this method mayunderestimate the magnitude of Eg,41 it can capture the trendwithin a series of samples. Correspondingly, the estimated Egfor YbMg2Bi2, YbMg2BiSb, and YbMg2Sb2 are 0.21, 0.31, and0.38 eV, respectively, confirming the expected trend. More-over, the increase in electronegativity difference of Mg−Sbbond (Δχ = 0.66) compared to that of Mg−Bi (Δχ = 0.3)contributes to a decrease in carrier mobility6,15,19 (μ), asshown in Figure 4d.Figure S6 shows an increasing trend in the effective mass(md*) with x, where a small but prominent peak in md* wasobserved for the composition YbMg2Bi0.6Sb1.4. At this point, itis important to note that most studies on AM2X2 compoundsassume acoustic phonon scattering (APS) as the dominantcarrier scattering mechanism when analyzing electronictransport properties.10,18 However, some works have doubtedthe dominance of APS, suggesting that it may not fully capturethe charge carrier transport behavior in these systems.25Moreover, there is a growing body of work suggesting polaroptical phonon (POP) scattering as the dominant mechanismin many materials.42,43 Correspondingly, we calculated theeffective mass by assuming both APS and POP mechanisms forcomparison. As shown in Figure S6, although the absolutevalues of the extracted effective mass differ, the overall trend,i.e., the increase of md* with increasing Sb content, remains thesame. This increase in md* with rising x can be interpreted interms of the crystal field-induced orbital splitting.44 TheAM2X2 Zintl compounds have the p-orbitals of the anionsdominating their valence band edge with an offset in theenergy (Δ) of pz and (px, py) orbitals owing to the effect ofcrystal field splitting, meaning that there is a separationbetween these two sets of orbitals on the energy scale. As thelattice parameters a and c decrease with x (Figure 1c,d), thecrystal field splitting energy (Δ) should also vary, indicatingthat the magnitude of separation (Δ) between the two sets oforbitals will vary. Considering Δ dependence on a for severalAM2X2 compounds, one may expect Δ to be more negative onthe energy scale for YbMg2Sb2 than YbMg2Bi2.44 A morenegative Δ value signifies a higher energy offset between the pzand (px, py) orbitals at the Γ-symmetry point in the bandstructure. This implies that the difference in the energy of pzand (px, py) orbitals of YbMg2Bi2 is less compared to thedifference in the energy for the same set of orbitals ofYbMg2Sb2. Therefore, we can presume the (px, py) orbitals ofYbMg2Bi2 to be closer to the valence band maxima (VBM)compared to the (px, py) orbitals of YbMg2Sb2. Hence, with theincrease in Sb content in the pristine composition YbMg2Bi2, itcan be expected that the energy difference between the pz and(px, py) orbitals of YbMg2Bi2 will slowly increase, shifting the(px, py) orbitals toward a more negative energy scale. Due to analloy composition, it is very much feasible that the shifts in (px,py) orbitals of pure Bi-based composition will eventually lead itto an overlap with the pz and/or (px, py) orbitals of Sb-richcompositions at some intermediate value of x inYbMg2Bi2−xSbx. This speculated gradual shifting of one set oforbitals toward the other set of pz/(px, py) orbitals, pushing ittoward an orbital overlap state, will gradually lead to anincrease in the overall band effective mass (mb*) and thus anincrease in the md*, followed by a simultaneous increase in theSeebeck coefficient, supporting the experimentally observedincreasing trend in α while a decreasing trend in μ with x(Figure 4).Composition-Dependent Thermal Transport. Thetotal thermal conductivity (κtot) of all samples was measured,revealing a significant reduction of approximately 61% at roomtemperature, e.g., from 3.36 Wm-1K−1 for YbMg2Bi2 to 1.20Wm-1K−1 for YbMg2Bi0.8Sb1.2 composition. Beyond x = 1.2, thethermal conductivity gradually increases (Figure 3d). A similarcompositional trend was observed in the room-temperaturelattice thermal conductivity (κlat), as shown in Figure 5. Latticethermal conductivity values of 2.8 Wm-1K−1 and 2.7 Wm-1K−1for the end-members YbMg2Bi2 and YbMg2Sb2, respectively,are in agreement with the reported values.26,45 Compositionaldependence of the κlat was further analyzed by employing theKlemens model.46 The model accounts for the disorder scalingparameter (u) obtained from the following equation.uutanlatlat01=(3)which includes the disorder scattering parameter (Γ) related tou asChemistry of Materials pubs.acs.org/cm Articlehttps://doi.org/10.1021/acs.chemmater.5c01433Chem. Mater. 2025, 37, 6782−67906787https://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfpubs.acs.org/cm?ref=pdfhttps://doi.org/10.1021/acs.chemmater.5c01433?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asuVhv22D ata2 lat0=(4)where κlat0 is the lattice thermal conductivity of the pristinesample (i.e., YbMg2Bi2), κlat is the lattice thermal conductivityof the Sb-substituted samples, h is Planck’s constant, θD is theDebye temperature, Vat is the volume per atom, and va is theaverage velocity of sound for Sb-substituted compositions,respectively. The obtained values of u and Γ are listed in TableS3. A reasonable fit for κlat is achieved, as shown in Figure 5,suggesting that point-defect scattering plays a central role inthe thermal transport of the solid solution system. The trend inthe obtained fitting parameters u and Γ suggests that the massand strain fluctuation arising due to the Sb substitution at theBi site plays a vital role in the thermal conductivity of alloyedcompositions.■ CONCLUSIONSWe report the first successful synthesis and systematicinvestigation of the full solid solution series YbMg2Bi2−xSbx(0 ≤ x ≤ 2), a case study of anion-site substitution in theAM2X2 Zintl phases. The primary aim of achieving amechanically tough/stable system with substituting moreelectronegative element at the anion site is successfully realizedover the full range of composition, as we shift from a weaklybonded structure of YbMg2Bi2 to a strong bonding network ofYbMg2Sb2, as supported by the increasing elastic moduli valueswith Sb content. A comprehensive analysis of the transportproperties supported by the room temperature Hall measure-ments has been discussed in order to understand themechanism behind the evolution of thermoelectric propertiesupon Sb content. A gradual increase observed in the effectivemass (md*) leads to the monotonic decrease of carrier mobility(μ) with Sb content, which is also supported by the observedincrease of electronegativity. Changes occurring in the bondingorbitals at the VBM due to the difference in crystal fieldsplitting energy in the two end compositions, YbMg2Bi2 andYbMg2Sb2, lead to orbital overlap at an intermediatecomposition of YbMg2Bi0.4Sb1.6. The enhancement in md* isconsistent with the increase in the orbital overlap arising fromcompositional tuning. Though the power factor decreasesmonotonically with x, the sudden drop (almost 3 times) in κlatfor YbMg2Bi1.8Sb0.2 from the pristine composition improves itsefficiency, resulting in a zT ≈ 0.9 at 723 K. A significantlyreduced lattice thermal conductivity upon alloying is analyzedusing the Klemens model, and as a result, point-defectscattering due to the increase in mass difference at the anionsite is attributed to the trend of κlat with x. Overall, the currentwork provides insight into the complex interplay amongchemical bonding, band structure, and electron transport inanion-substituted Zintl compounds. Our findings suggest thatthe midrange to Sb-rich compositions (x = 1.6 − 1.8) exhibit afavorable balance of low lattice thermal conductivity, enhancedmechanical robustness (reflected in higher elastic moduli), andhigh Seebeck coefficients (≈400 μVK−1), making themparticularly promising for targeted doping strategies aimed atoptimizing carrier concentration and further improvement ofzT.■ ASSOCIATED CONTENTData Availability StatementData will be made available upon request.*sı Supporting InformationThe Supporting Information is available free of charge athttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433.XRD data of samples prepared with different ball millingtime and corresponding temperature-dependent elec-trical and thermal transport properties of those samples,relative density and lattice parameters for allYbMg2Bi2−xSbx samples, Rietveld refinement fits forYbMg2Bi2, YbMg2BiSb, and YbMg2Sb2, SEM images forthe same compositions, evolution of the effective masswith Sb content, fit parameters of Klemens model androom temperature thermoelectric properties of solidsolution YbMg2Bi2−xSbx upon Sb content, and composi-tional dependence of zT at 723 K (PDF)■ AUTHOR INFORMATIONCorresponding AuthorsAndrei Novitskii − Research Center for MaterialsNanoarchitectonics (MANA), National Institute forMaterials Science (NIMS), Tsukuba, Ibaraki 305-0044,Japan; Email: NOVITSKII.Andrei@nims.go.jpTakao Mori − Research Center for MaterialsNanoarchitectonics (MANA), National Institute forMaterials Science (NIMS), Tsukuba, Ibaraki 305-0044,Japan; Graduate School of Pure and Applied Sciences,University of Tsukuba, Tsukuba, Ibaraki 305-8573, Japan;orcid.org/0000-0003-2682-1846;Email: MORI.Takao@nims.go.jpAuthorKushal Mehrotra − Research Center for MaterialsNanoarchitectonics (MANA), National Institute forMaterials Science (NIMS), Tsukuba, Ibaraki 305-0044,Japan; Graduate School of Pure and Applied Sciences,University of Tsukuba, Tsukuba, Ibaraki 305-8573, JapanComplete contact information is available at:https://pubs.acs.org/10.1021/acs.chemmater.5c01433Author ContributionsK.M.: Investigation, formal analysis, visualization, writing�original draft, and writing�review and editing. A.N.:Conceptualization, investigation, formal analysis, methodology,Figure 5. Lattice thermal conductivity of the YbMg2Bi2−xSbx solidsolution. The dotted line shows the predicted reduction in κlat fromthe Klemens model (eq 3) at 300 K. Literature data forYbMg2Bi2−xSbx reported previously (refs11,12,18,19,25,26.) is alsoshown for comparison.Chemistry of Materials pubs.acs.org/cm Articlehttps://doi.org/10.1021/acs.chemmater.5c01433Chem. Mater. 2025, 37, 6782−67906788https://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfhttps://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?goto=supporting-infohttps://pubs.acs.org/doi/suppl/10.1021/acs.chemmater.5c01433/suppl_file/cm5c01433_si_001.pdfhttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Andrei+Novitskii"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfmailto:NOVITSKII.Andrei@nims.go.jphttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Takao+Mori"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://orcid.org/0000-0003-2682-1846https://orcid.org/0000-0003-2682-1846mailto:MORI.Takao@nims.go.jphttps://pubs.acs.org/action/doSearch?field1=Contrib&text1="Kushal+Mehrotra"&field2=AllField&text2=&publication=&accessType=allContent&Earliest=&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig5&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig5&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig5&ref=pdfhttps://pubs.acs.org/doi/10.1021/acs.chemmater.5c01433?fig=fig5&ref=pdfpubs.acs.org/cm?ref=pdfhttps://doi.org/10.1021/acs.chemmater.5c01433?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-asvisualization, writing�original draft, and writing�review andediting. T.M.: Resources, funding acquisition, supervision, andwriting�review and editing.NotesThe authors declare no competing financial interest.■ ACKNOWLEDGMENTSThis work was supported by JST Mirai JPMJMI19A1. K.M.acknowledges the financial support provided by the JapaneseGovernment (Monbukagakusho) Scholarship (MEXT scholar-ship). A part of this work was supported by “AdvancedResearch Infrastructure for Materials and Nanotechnology inJapan (ARIM)” of the Ministry of Education, Culture, Sports,Science and Technology (MEXT) Proposal NumberJPMXP1224NM5121.■ REFERENCES(1) Snyder, G. J.; Toberer, E. S. Complex Thermoelectric Materials.Nat. Mater. 2008, 7, 105−114.(2) Snyder, G. J.; Snyder, A. H. Figure of Merit ZT of aThermoelectric Device Defined from Materials Properties. EnergyEnviron. Sci. 2017, 10, 2280−2283.(3) Ioffe, A. F. Semiconductor Thermoelements, and ThermoelectricCooling; Infosearch: London, 1957.(4) Kazem, N.; Kauzlarich, S. M. Handbook on the Physics andChemistry of Rare Earths; Elsevier, 2016; Vol. 50, pp 177−208.(5) Kauzlarich, S. M. Zintl Phases: From Curiosities to ImpactfulMaterials. Chem. Mater. 2023, 35, 7355−7362.(6) Peng, W.; Chanakian, S.; Zevalkink, A. Crystal Chemistry andThermoelectric Transport of Layered AM2X2 Compounds. Inorg.Chem. Front. 2018, 5, 1744−1759.(7) Kauzlarich, S. M.; Zevalkink, A.; Toberer, E.; Snyder, G. J. ZintlPhases: Recent Developments in Thermoelectrics and Future Outlook,Energy and Environment Series; RSC, 2016; Chapter 1, pp 1−26.(8) Liu, Z.; Sato, N.; Gao, W.; Yubuta, K.; Kawamoto, N.; Mitome,M.; Kurashima, K.; Owada, Y.; Nagase, K.; Lee, C.-H.; Yi, J.;Tsuchiya, K.; Mori, T. Demonstration of Ultrahigh ThermoelectricEfficiency of ∼ 7.3% in Mg3Sb2/MgAgSb Module for Low-Temperature Energy Harvesting. Joule 2021, 5, 1196−1208.(9) Imasato, K.; Wood, M.; Anand, S.; Kuo, J. J.; Snyder, G. J.Understanding the High Thermoelectric Performance of Mg3Sb2−Mg3Bi2 Alloys, Adv. Energy Sustainability Res. 2022, 3, No. 2100208.(10) Shuai, J.; Geng, H.; Lan, Y.; Zhu, Z.; Wang, C.; Liu, Z.; Bao, J.;Chu, C.-W.; Sui, J.; Ren, Z. Higher Thermoelectric Performance ofZintl Phases (Eu0.5Yb0.5)1−xCaxMg2Bi2 by Band Engineering andStrain Fluctuation. Proc. Natl. Acad. Sci. U.S.A. 2016, 113, E4125−E4132.(11) Wang, J.; Guo, M.; Zhu, J.; Qin, D.; Guo, F.; Zhang, Q.; Cai,W.; Sui, J. Enhanced Thermoelectric Properties of Zintl PhaseYbMg2Bi1.98 through Bi Site Substitution with Sb. J. Mater. Sci.Technol. 2020, 59, 189−194.(12) Liang, Z.; Shang, H.; Xu, C.; Shi, X.; Zhang, F.; Ren, W.; Song,S.; Ding, F.; Ren, Z. New Insights into the Effect of ChemicalBonding Strength on Thermoelectric Performance and Stability inYbMg2Bi2 toward Practical Thermoelectric Applications. Mater. TodayPhys. 2022, 28, No. 100858.(13) Isotta, E.; Peng, W.; Balodhi, A.; Zevalkink, A. Elastic Moduli:A Tool for Understanding Chemical Bonding and Thermal Transportin Thermoelectric Materials. Angew. Chem., Int. Ed. 2023, 62 (12),No. e202213649.(14) Shannon, R. D. Revised Effective Ionic Radii and SystematicStudies of Interatomic Distances in Halides and Chalcogenides. ActaCrystallogr., Sect. A 1976, 32, 751−767.(15) Tantardini, C.; Oganov, A. R. Thermochemical Electro-negativities of the Elements. Nat. Commun. 2021, 12, No. 2087.(16) Hanus, R.; Agne, M. T.; Rettie, A. J.; Chen, Z.; Tan, G.; Chung,D. Y.; Kanatzidis, M. G.; Pei, Y.; Voorhees, P. W.; Snyder, G. J. LatticeSoftening Significantly Reduces Thermal Conductivity and Leads toHigh Thermoelectric Efficiency. Adv. Mater. 2019, 31, No. 1900108.(17) Slade, T. J.; Anand, S.; Wood, M.; Male, J. P.; Imasato, K.;Cheikh, D.; Al Malki, M. M.; Agne, M. T.; Griffith, K. J.; Bux, S. K.;Wolverton, C.; Kanatzidis, M. G.; Snyder, G. J. Charge-Carrier-Mediated Lattice Softening Contributes to High zT in ThermoelectricSemiconductors. Joule 2021, 5, 1168−1182.(18) Zhou, T.; Feng, Z.; Mao, J.; Jiang, J.; Zhu, H.; Singh, D. J.;Wang, C.; Ren, Z. Thermoelectric Properties of Zintl PhaseYbMg2Sb2. Chem. Mater. 2020, 32, 776−784.(19) Zhang, Z.; Wang, X.; Liu, Y.; Chen, C.; Yao, H.; Yin, L.; Li, X.;Li, S.; Zhang, F.; Bai, F.; Sui, J.; Yu, B.; Cao, F.; Liu, X.; Mao, J.; Xie,G.; Zhang, Q. Balancing the Anionic Framework Polarity forEnhanced Thermoelectric Performance in YbMg2Sb2 Zintl Com-pounds. J. Materiomics 2019, 5, 583−589.(20) Rodríguez-Carvajal, J. Recent Advances in Magnetic StructureDetermination by Neutron Powder Diffraction. Phys. B 1993, 192,55−69.(21) Ashcroft, N. W.; Mermin, N. D.Solid State PhysicsHarcourtCollege Publishers, 1976.(22) Kittel, C. Introduction to Solid State Physics, 8th ed.; Wiley,2004.(23) Gurunathan, R.; Hanus, R.; Snyder, G. J. Alloy scattering ofphonons. Mater. Horiz. 2020, 7, 1452−1456.(24) May, A. F.; McGuire, M. A.; Singh, D. J.; Custelcean, R.;Jellison, G. E., Jr Structure and Properties of Single CrystallineCaMg2Bi2, EuMg2Bi2, and YbMg2Bi2. Inorg. Chem. 2011, 50, 11127−11133.(25) May, A. F.; McGuire, M. A.; Singh, D. J.; Ma, J.; Delaire, O.;Huq, A.; Cai, W.; Wang, H. Thermoelectric Transport Properties ofCaMg2Bi2, EuMg2Bi2, and YbMg2Bi2. Phys. Rev. B 2012, 85,No. 035202.(26) Shuai, J.; Liu, Z.; Kim, H. S.; Wang, Y.; Mao, J.; He, R.; Sui, J.;Ren, Z. Thermoelectric Properties of Bi-based Zintl CompoundsCa1−xYbxMg2Bi2. J. Mater. Chem. A 2016, 4, 4312−4320.(27) Cape, J. A.; Lehman, G. Temperature and Finite Pulse-TimeEffects in the Flash Method for Measuring Thermal Diffusivity. J.Appl. Phys. 1963, 34, 1909−1913.(28) Agne, M. T.; Imasato, K.; Anand, S.; Lee, K.; Bux, S. K.;Zevalkink, A.; Rettie, A. J.; Chung, D. Y.; Kanatzidis, M. G.; Snyder,G. J. Heat Capacity of Mg3Sb2, Mg3Bi2, and Their Alloys at HighTemperature. Mater. Today Phys. 2018, 6, 83−88.(29) Kim, H.-S.; Gibbs, Z. M.; Tang, Y.; Wang, H.; Snyder, G. J.Characterization of Lorenz Number with Seebeck CoefficientMeasurement. APL Mater. 2015, 3, No. 041506.(30) Alleno, E.; Bérardan, D.; Byl, C.; Candolfi, C.; Daou, R.;Decourt, R.; Guilmeau, E.; Hébert, S.; Hejtmánek, J.; Lenoir, B.;Masschelein, P.; Ohorodnichuk, V.; Pollet, M.; Populoh, S.; Ravot, D.;Rouleau, O.; Soulier, M. A Round Robin Test of the Uncertainty onthe Measurement of the Thermoelectric Dimensionless Figure ofMerit of Co0.97Ni0.03Sb3,. Rev. Sci. Instrum. 2015, 86, No. 011301.(31) Wang, J.; Guo, F.; Guo, M.; Zhu, J.; Zhang, Q.; Cai, W.; Sui, J.Zintl Phase Yb1−xBaxMg2Bi1.98 Compounds with Enhanced Thermo-electric Performance Caused by Cation Substitution. ACS Appl.Energy Mater. 2020, 3, 11036−11041.(32) Guo, M.; Zhai, W.; Li, J.; Zhu, J.; Guo, F.; Liu, Z.; Liu, M.; Zhu,Y.; Dong, X.; Zhang, Y.; Zhang, Q.; Cai, W.; Sui, J. HighThermoelectric Performance of CaMg2Bi2 Enabled by DynamicDoping and Orbital Alignment. Adv. Funct. Mater. 2022, 32,No. 2200407.(33) WebElements, “Periodic Table of the Elements” 2024. https://www.webelements.com/, (accessed February 2024).(34) Zeier, W. G.; Heinrich, C. P.; Day, T.; Panithipongwut, C.;Kieslich, G.; Brunklaus, G.; Snyder, G. J.; Tremel, W. Bond StrengthDependent Superionic Phase Transformation in the Solid SolutionSeries Cu2ZnGeSe4−xSx. J. Mater. Chem. A 2014, 2, 1790−1794.(35) Thompson, R.; Clegg, W. Predicting Whether a Material isDuctile or Brittle. Curr. Opin. Solid State Mater. Sci. 2018, 22, 100−108.Chemistry of Materials pubs.acs.org/cm Articlehttps://doi.org/10.1021/acs.chemmater.5c01433Chem. Mater. 2025, 37, 6782−67906789https://doi.org/10.1038/nmat2090https://doi.org/10.1039/C7EE02007Dhttps://doi.org/10.1039/C7EE02007Dhttps://doi.org/10.1021/acs.chemmater.3c01874?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.chemmater.3c01874?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1039/C7QI00813Ahttps://doi.org/10.1039/C7QI00813Ahttps://doi.org/10.1016/j.joule.2021.03.017https://doi.org/10.1016/j.joule.2021.03.017https://doi.org/10.1016/j.joule.2021.03.017https://doi.org/10.1002/aesr.202100208https://doi.org/10.1002/aesr.202100208https://doi.org/10.1073/pnas.1608794113https://doi.org/10.1073/pnas.1608794113https://doi.org/10.1073/pnas.1608794113https://doi.org/10.1016/j.jmst.2020.04.052https://doi.org/10.1016/j.jmst.2020.04.052https://doi.org/10.1016/j.mtphys.2022.100858https://doi.org/10.1016/j.mtphys.2022.100858https://doi.org/10.1016/j.mtphys.2022.100858https://doi.org/10.1002/anie.202213649https://doi.org/10.1002/anie.202213649https://doi.org/10.1002/anie.202213649https://doi.org/10.1107/S0567739476001551https://doi.org/10.1107/S0567739476001551https://doi.org/10.1038/s41467-021-22429-0https://doi.org/10.1038/s41467-021-22429-0https://doi.org/10.1002/adma.201900108https://doi.org/10.1002/adma.201900108https://doi.org/10.1002/adma.201900108https://doi.org/10.1016/j.joule.2021.03.009https://doi.org/10.1016/j.joule.2021.03.009https://doi.org/10.1016/j.joule.2021.03.009https://doi.org/10.1021/acs.chemmater.9b04131?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acs.chemmater.9b04131?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1016/j.jmat.2019.08.002https://doi.org/10.1016/j.jmat.2019.08.002https://doi.org/10.1016/j.jmat.2019.08.002https://doi.org/10.1016/0921-4526(93)90108-Ihttps://doi.org/10.1016/0921-4526(93)90108-Ihttps://doi.org/10.1039/C9MH01990Ahttps://doi.org/10.1039/C9MH01990Ahttps://doi.org/10.1021/ic2016808?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/ic2016808?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1103/PhysRevB.85.035202https://doi.org/10.1103/PhysRevB.85.035202https://doi.org/10.1039/C6TA00507Ahttps://doi.org/10.1039/C6TA00507Ahttps://doi.org/10.1063/1.1729711https://doi.org/10.1063/1.1729711https://doi.org/10.1016/j.mtphys.2018.10.001https://doi.org/10.1016/j.mtphys.2018.10.001https://doi.org/10.1063/1.4908244https://doi.org/10.1063/1.4908244https://doi.org/10.1063/1.4905250https://doi.org/10.1063/1.4905250https://doi.org/10.1063/1.4905250https://doi.org/10.1021/acsaem.0c02019?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1021/acsaem.0c02019?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1002/adfm.202200407https://doi.org/10.1002/adfm.202200407https://doi.org/10.1002/adfm.202200407https://www.webelements.com/https://www.webelements.com/https://doi.org/10.1039/C3TA13007Jhttps://doi.org/10.1039/C3TA13007Jhttps://doi.org/10.1039/C3TA13007Jhttps://doi.org/10.1016/j.cossms.2018.04.001https://doi.org/10.1016/j.cossms.2018.04.001pubs.acs.org/cm?ref=pdfhttps://doi.org/10.1021/acs.chemmater.5c01433?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-as(36) Tritt, T. M. Thermal Conductivity; Springer: New York, 2010.(37) Fistul, V. I. Heavily Doped Semiconductors; Plenum Press, 1969.(38) Goldsmid, H. J.; Sharp, J. Estimation of the Thermal Band Gapof a Semiconductor from Seebeck Measurements. J. Electron. Mater.1999, 28, 869−872.(39) Zhang, J.; Song, L.; Madsen, G. K.; Fischer, K. F.; Zhang, W.;Shi, X.; Iversen, B. B. Designing High-Performance LayeredThermoelectric Materials Through Orbital Engineering. Nat.Commun. 2016, 7, No. 10892.(40) Novitskii, A.; Mori, T. Pisarenko’s Formula for the Thermo-power. 2025 arXiv:2502.03837. arXiv.org e-Print archive. https://arxiv.org/abs/2502.03837.(41) Gibbs, Z. M.; Kim, H.-S.; Wang, H.; Snyder, G. J. Band gapestimation from temperature dependent Seebeck measurement−Deviations from the 2e|S|maxTmax relation. Appl. Phys. Lett. 2015,106, No. 022112.(42) Ganose, A. M.; Park, J.; Faghaninia, A.; Robinson, R. W.;Persson, K. A.; Jain, A. Efficient Calculation of Carrier ScatteringRates from First Principles. Nat. Commun. 2021, 12, No. 2222,DOI: 10.1038/s41467-021-22440-5.(43) Guo, M.; Liu, M.; Yuan, D.; Chen, H.; Sun, C.; Zhang, Q.; Zhu,Y.; Guo, F.; Yu, Y.; Sui, J. Strong Polar Optical Phonon Screening andSoftening Enhance the Thermoelectric Performance of ZintlCompounds. Adv. Energy Mater. 2025, 15, No. 2405024.(44) Zhang, J.; Song, L.; Madsen, G. K. H.; Fischer, K. F. F.; Zhang,W.; Shi, X.; Iversen, B. B. Designing High-Performance LayeredThermoelectric Materials Through Orbital Engineering. Nat.Commun. 2016, 7, No. 10892.(45) Zhou, T.; Feng, Z.; Mao, J.; Jiang, J.; Zhu, H.; Singh, D. J.;Wang, C.; Ren, Z. Thermoelectric properties of zintl phase YbMg2Sb2.Chem. Mater. 2020, 32, 776−784.(46) Klemens, P. G. Thermal Resistance due to Point Defects atHigh Temperatures. Phys. Rev. 1960, 119, No. 507.Chemistry of Materials pubs.acs.org/cm Articlehttps://doi.org/10.1021/acs.chemmater.5c01433Chem. Mater. 2025, 37, 6782−67906790https://doi.org/10.1007/s11664-999-0211-yhttps://doi.org/10.1007/s11664-999-0211-yhttps://doi.org/10.1038/ncomms10892https://doi.org/10.1038/ncomms10892https://arxiv.org/abs/2502.03837https://arxiv.org/abs/2502.03837https://doi.org/10.1063/1.4905922https://doi.org/10.1063/1.4905922https://doi.org/10.1063/1.4905922https://doi.org/10.1038/s41467-021-22440-5https://doi.org/10.1038/s41467-021-22440-5https://doi.org/10.1038/s41467-021-22440-5?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1002/aenm.202405024https://doi.org/10.1002/aenm.202405024https://doi.org/10.1002/aenm.202405024https://doi.org/10.1038/ncomms10892https://doi.org/10.1038/ncomms10892https://doi.org/10.1021/acs.chemmater.9b04131?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://doi.org/10.1103/PhysRev.119.507https://doi.org/10.1103/PhysRev.119.507pubs.acs.org/cm?ref=pdfhttps://doi.org/10.1021/acs.chemmater.5c01433?urlappend=%3Fref%3DPDF&jav=VoR&rel=cite-ashttps://solutions.cas.org/CASInsights_Subscribe?utm_campaign=GLO_GEN_ANY_CIS_AWS&utm_medium=DSP_CAS_ORG&utm_source=Publication_CEN&utm_content=pdf_footer