Bronze Telluride (BT) in Thermoelectrics

• Bronze telluride (BT) is an Sn-modified Cu2Te derived from industrial bronze via CVD-assisted tellurization, featuring trace Sn and engineered lattice distortions.
• Controlled synthesis and microstructural analysis confirm phase-pure hexagonal Cu2Te with interlayer Sn incorporation that enhances phonon scattering and maintains p-type conductivity.
• Thermoelectric evaluation reveals a rising ZT near 1 at 500 K, underscoring BT's potential for scalable waste-heat recovery and energy harvesting applications.

Searching arXiv for the specified paper to ground the article in the cited source. Bronze telluride (BT) denotes Sn-doped copper telluride obtained by tellurizing pre-alloyed Cu–Sn bronze powders; in the reported realization, the product is chemically best described as predominantly hexagonal Cu$_2$Te with trace Sn incorporation rather than as a distinct bulk Sn-substituted ternary phase. TEM-EDS mapping quantified Sn at 0.09 at% in the tellurized product, whereas XPS did not detect Sn signals, indicating that Sn is present at very low concentration and does not substitute Cu or Te lattice sites in the structural picture advanced for this material. Instead, Sn is associated with local lattice distortions and interlayer incorporation, while the macroscopic thermoelectric response remains that of a p-type, degenerate Cu$_2$Te-based system. The term “bronze telluride” emphasizes both the industrial bronze precursor and the telluride product generated by a CVD-assisted tellurization route designed for scalable thermoelectric energy harvesting (R et al., 12 Aug 2025).

1. Definition, nomenclature, and materials identity

In this context, “bronze telluride” refers specifically to Sn-modified Cu$_2$Te synthesized from pre-alloyed Cu–Sn bronze powder. The precursor identity is central to the nomenclature: the process begins from industrial bronze, and tellurization converts that Cu-rich alloy into a telluride product while retaining only trace Sn in the final material. Structural and spectroscopic evidence supports the conclusion that BT is not a stoichiometric bronze–tellurium compound in the conventional sense, nor a Cu$_2$Te/SnTe composite, but rather Cu$_2$Te:Sn with approximately 0.09 at% Sn and no detected secondary Sn telluride phase (R et al., 12 Aug 2025).

This distinction addresses a likely misconception. The initial motivation treats Sn as a dopant that could substitute Cu and tune carrier concentration, but the reported analyses instead support non-substitutional incorporation. A plausible implication is that the functional role of Sn in BT is primarily defect- and strain-mediated rather than conventional aliovalent substitutional doping. The DFT models are consistent with this interpretation, placing Sn in interlayer positions rather than on Cu or Te sites.

The material is framed as a sustainable and scalable p-type thermoelectric candidate within the broader class of copper chalcogenides. Among these, Cu$_2$Te is noted for its degenerate semiconducting character and low thermal conductivity, and BT inherits those characteristics while introducing a controlled defect landscape through the bronze-derived synthesis route.

2. Crystal chemistry, phase constitution, and microstructure

Before tellurization, XRD of the bronze powder matched cubic Cu$_{3.64}$Sn$_{0.28}$ (ICSD 629271) and hexagonal Cu$_{10}$Sn$_3$ (ICSD 103105), establishing the precursor as Cu-rich bronze. After tellurization, XRD indexed the product to hexagonal Cu$_2$0Te with space group $_2$1 (ICSD 77055), and no secondary phases, notably SnTe, were detected. Relative to the standard lattice constant for pure Cu$_2$2Te of approximately 8.12 Å, the tellurized sample exhibited a decrease of 0.943 Å, attributed to strain-induced lattice contraction and/or Sn incorporation without direct substitution (R et al., 12 Aug 2025).

Raman spectroscopy using 532 nm excitation and a 2400 g/mm grating showed modes at 100.2 cm$_2$3, 120 cm$_2$4, 142.2 cm$_2$5, and 275.8 cm$_2$6. These were assigned respectively to Cu–Te $_2$7, Te–Te stretching associated with Te-rich local environments, Cu$_2$8Te lattice vibrations, and Te $_2$9, together supporting Cu$_2$0Te formation and local strain. XPS identified Cu 2p peaks at 932.4 eV and 952.4 eV, consistent with Cu$_2$1, and Te 3d peaks at 572.7 eV and 583.1 eV, consistent with Te$_2$2. A broad feature near 570 eV was interpreted as defect- or strain-related Te states. The absence of Sn in the XPS survey is consistent with the very low Sn concentration quantified by TEM-EDS.

Electron microscopy further refines the structural picture. SEM showed spherical bronze precursor particles, whereas the tellurized product consisted of layered Cu$_2$3Te crystals with step-like morphology. HRTEM resolved the layered structure along the (002) direction, and FFT analysis gave a lattice spacing of 0.36 nm corresponding to the (002) planes of Cu$_2$4Te. Inverse FFT analysis revealed edge dislocations and discontinuities linked to strain. These observations support the interpretation that trace Sn induces local distortions and dislocation structures rather than forming a substitutional solid solution.

The significance of this microstructure is thermoelectric. The layered morphology and dislocation-rich lattice provide a mechanism for enhanced phonon scattering, which lowers $_2$5, while the p-type transport character of Cu$_2$6Te is retained.

3. Controlled growth by CVD-assisted tellurization

BT is synthesized by a CVD-assisted tellurization of pre-alloyed Cu–Sn bronze powder. The reported reactor is a ThermoFisher Lindberg Blue single-zone furnace with a 1-inch quartz tube. Te powder of 99.9% purity (Sigma-Aldrich) is placed in an alumina boat upstream in the hot zone, while the bronze precursor is placed in separate alumina boats at the center of the furnace. Ar is maintained as carrier gas at 50 sccm throughout the process. The temperature is ramped to 750 °C at 25 °C/min and held for 10 min. H$_2$7 is introduced starting at 400 °C and continued to the end of growth to assist tellurization. Cooling is by natural cool, with the furnace opened at 400 °C to accelerate cooling. No post-processing is reported; the product is analyzed as-grown (R et al., 12 Aug 2025).

The process is paired with a thermodynamic description of phase selection. Calphad modeling was performed in FactSage using the modified quasichemical model for the liquid, compound energy formalism for solids, and Kohler–Toop interpolation for the ternary Cu–Sn–Te system. At 750 °C, Cu$_2$8Te is predicted to crystallize at lower tellurium partial pressures and Cu mole fractions above 0.8. Further lowering the Te vapor pressure can revert the system to metallic Cu–Sn phases, including BCC, $_2$9, and FCC, depending on alloy composition. The experimental combination of Cu-rich bronze, controlled Te vapor, and H$_2$0-assisted tellurization lies within the predicted Cu$_2$1Te formation window, consistent with the observed absence of SnTe.

The nucleation and growth mechanism inferred from experiment and computation is that Te vapor reacts with Cu-rich bronze particles to form hexagonal Cu$_2$2Te, while a fraction of Sn is incorporated into interlayer regions. This interlayer-like incorporation produces local strain and edge dislocations rather than substitutional doping. A plausible implication is that the synthesis route is simultaneously a phase-formation method and a defect-engineering strategy.

The approach is presented as scalable and sustainable because it uses industrially relevant bronze recoverable from recycled wastes, relies on a single-step tellurization, and employs a short dwell time at 750 °C in an Ar/H$_2$3 atmosphere. The resulting material is described as phase-pure Cu$_2$4Te with uniform Sn incorporation at trace levels and engineered defects.

4. Thermoelectric transport and performance metrics

The study uses the standard thermoelectric relations

$_2$5

The Wiedemann–Franz relation is used qualitatively in the computational analysis to interpret $_2$6 trends, but an explicit Lorenz number $_2$7 is not reported experimentally. For degenerate semiconductors and metals, $_2$8 is noted as a common value (R et al., 12 Aug 2025).

Hall measurements show p-type behavior. The hole mobility $_2$9 increases from approximately 49 cm$_2$0 V$_2$1 s$_2$2 at 302 K to approximately 55 cm$_2$3 V$_2$4 s$_2$5 at 363 K, suggesting reduced grain-boundary or impurity scattering with increasing temperature and indicating that phonon scattering is not dominant in that interval. The carrier concentration $_2$6 also increases with temperature between 302 and 363 K, consistent with thermally activated behavior in a degenerate Cu$_2$7Te system with shallow acceptors or narrow-gap excitation.

The electrical conductivity $_2$8 increases with temperature, displaying a semiconducting trend. The Seebeck coefficient $_2$9 also increases with temperature; the reported interpretation is that strain-enhanced effective mass and localized states near $_2$0 increase the energy dependence of the density of states in a way favorable for $_2$1. Although standalone $_2$2 values for BT are not listed in the main text, the device discussion cites a typical range for Sn-doped Cu$_2$3Te of approximately 70–140 $_2$4V/K. Total thermal conductivity $_2$5 decreases with temperature, which is consistent with stronger phonon–phonon and defect scattering at elevated temperature, while the degenerate or metallic character keeps $_2$6 moderate.

These trends combine to raise the power factor with temperature and produce a $_2$7 curve that rises steadily toward approximately 1 at 500 K. The improvement is attributed to the joint action of increasing $_2$8 and decreasing $_2$9. The comparative $_{3.64}$0 plot indicates that BT outperforms various Cu$_{3.64}$1Te-based materials prepared by other synthesis routes in the low-to-mid temperature range. The stated origins of this improvement are suppressed $_{3.64}$2 through Sn-induced local distortions, edge dislocations, and layered microstructure; maintained high $_{3.64}$3 due to the degenerate nature of Cu$_{3.64}$4Te and reduced grain-boundary scattering with temperature; and enhanced $_{3.64}$5 through DOS asymmetry near $_{3.64}$6 and a possible strain-related increase in effective mass.

The Pisarenko relation is not explicitly analyzed. The discussion notes only that, in degenerate semiconductors, $_{3.64}$7 typically follows $_{3.64}$8, and that the observed increase in $_{3.64}$9 with temperature and controlled $_{0.28}$0 is qualitatively consistent with DOS and $_{0.28}$1 shifts inferred from DFT.

5. Electronic-structure and thermodynamic interpretation

The computational framework combines DFT and semiclassical transport. Electronic-structure calculations were performed in SIESTA with GGA-PBE, Troullier–Martins norm-conserving pseudopotentials in Kleinman–Bylander form, and a DZP basis. Geometry optimization proceeded to forces below 0.01 eV/Å, using a $_{0.28}$2 $_{0.28}$3-mesh for structural relaxation and a $_{0.28}$4 mesh for bands and PDOS. Transport calculations used BoltZTraP2 within semiclassical BTE under the constant relaxation time approximation and rigid band approximation, with a dense $_{0.28}$5 $_{0.28}$6-mesh (R et al., 12 Aug 2025).

For pristine hexagonal Cu$_{0.28}$7Te, the optimized lattice parameters are $_{0.28}$8 Å, $_{0.28}$9 Å, and $_{10}$0, with Cu–Te bond lengths of approximately 3.82 Å and 2.74 Å, reflecting the layered structure. An orthorhombic phase with $_{10}$1 Å, $_{10}$2 Å, and $_{10}$3 Å also retains metallic character, and states near $_{10}$4 are dominated by Cu 3d and Te 5p orbitals.

Sn incorporation is modeled at interlayer positions denoted m2, m3, and m4. The m4 configuration is reported as thermodynamically more stable than pristine Cu$_{10}$5Te. In m2 and m3, metallic character is preserved and Sn 5p states contribute near the conduction band, acting as donor-like states that slightly raise $_{10}$6 and modify DOS asymmetry. In this picture, $_{10}$7 is slightly reduced because of local scattering, and $_{10}$8 is reduced accordingly. In m4, $_{10}$9 shifts downward toward the valence band, and a higher concentration of Sn 5p states appears in the valence band. This substantially enhances $_3$0 in the model, in some cases by orders of magnitude, but at the expense of reduced $_3$1, leaving $_3$2 moderate.

Two coupled mechanisms are used to explain BT’s thermoelectric behavior. The electronic mechanism is DOS redistribution around $_3$3 by Sn 5p states, which tunes the Fermi-level position and enhances the energy dependence of carrier transport. The phononic mechanism is stronger phonon scattering from local lattice distortions and altered Cu–Te bonding, which lowers $_3$4. The experimentally observed edge dislocations and layered morphology are presented as supporting evidence for the latter mechanism.

Thermodynamically, the calculations predict Cu$_3$5Te formation at 750 °C under lower tellurium partial pressure and Cu fraction greater than 0.8, while competing SnTe and metallic Cu–Sn phases are avoided within the selected growth window. This theoretical phase map is consistent with the XRD observation of phase-pure Cu$_3$6Te.

6. Device demonstration, limitations, and outlook

A device-level demonstration uses a single p–n thermocouple consisting of p-type BT and n-type galena (PbS). The schematic and photograph show a simple module with electrical and thermal contacts, but detailed leg dimensions, metallization strategy, and packaging are not provided. The open-circuit voltage increases with temperature. Two device values are reported: $_3$7 mV at $_3$8 K, corresponding to an effective combined Seebeck coefficient of approximately 90.9 $_3$9V/K, and $_2$00 mV at $_2$01 K, corresponding to approximately 80 $_2$02V/K (R et al., 12 Aug 2025).

The measured effective Seebeck is lower than a simple estimate based on

$_2$03

using the cited typical ranges $_2$04–140 $_2$05V/K for Sn-doped Cu$_2$06Te and $_2$07 to $_2$08 $_2$09V/K for n-type PbS, which would imply a combined Seebeck of approximately 290–410 $_2$10V/K. The reported explanation is interfacial contact resistance, inhomogeneity, and possible partial shorting. For this reason, output power and internal resistance were not reported, in order to avoid uncertainty.

The device result nevertheless establishes feasibility for BT as a p-type leg in a cascaded thermoelectric architecture for medium-temperature waste-heat recovery. BT reaches $_2$11 at 500 K, and the BT–PbS pair is described as thermally compatible and composed of earth-abundant elements. Stability under cycling or oxidation was not quantitatively assessed, although XPS and phase purity are taken as suggestive of good chemical stability of the Cu$_2$12Te phase after growth.

The work also delineates its own limitations. The Sn content is very low, around 0.09 at%, and incorporation is non-substitutional, so further tuning of dopant level and site occupancy may be required to maximize $_2$13. Device performance is presently constrained by contact resistance and inhomogeneity. In addition, intrinsic structural transitions of Cu$_2$14Te at high temperature and Cu$_2$15 mobility in Cu$_2$16Te with $_2$17 may affect long-term stability, motivating mitigation through dopant strategies and microstructure control.

The stated optimization pathways are correspondingly specific: dopant level and site engineering, particularly toward interlayer positions with m4-like DOS features; microstructure engineering through grain-size and texture control, porosity reduction, and hierarchical defect introduction; interface and device engineering through low-resistance metallization, diffusion barriers, precise leg geometry, and robust packaging; and process control through fine adjustment of Te vapor activity, dwell time, and H$_2$18 flow, guided by Calphad predictions. A plausible implication is that the principal contribution of BT lies not only in the reported $_2$19 value, but also in the integration of precursor selection, controlled tellurization, defect engineering, and modeling into a route for scalable copper-based thermoelectrics.